,6*^ \,'*^^**A* *-t,°'-fS'*\ 3 2 and 55. 2 Crane and Joyce: J. Soc. Chem. Ind., 1910, 29, 540. I 4 CELLULOSE containing 57 to 67 per cent of sulphuric acid, 16 to 6 per cent of nitric acid and 25 to 27 per cent of water, and nitrating for short times have prepared products with as little as 3.5 to 4.5 per cent of nitrogen. These are pasty, gelatinous masses, in- soluble in all nitrocellulose solvents, but readily soluble in solutions of caustic alkalis. When stronger acids are employed the nitrates formed con- tain more nitrogen up to a limit of about 14 per cent which corresponds to the trinitrate, GeE^C^NC^. This nitrate, which is the most explosive gun-cotton, may be prepared by treating cotton with a mixture of 3 parts of nitric acid (sp. gr. 1.5) and 1 part of sulphuric acid for about 24 hours at io° C. The unstable mixed esters, containing both nitric and sulphuric groups, are formed as an intermediate stage in the reaction and the NO3 groups finally replace the HS0 4 groups. It is con- sidered that traces of the mixed ester remaining in the finished product are often responsible for its instability. In this reaction 100 parts of cellulose yield about 170 parts of the nitrate. Nitra- tion under these conditions does not visibly alter the physical structure of the cellulose. The trinitrate is insoluble in alcohol, ether, mixtures of the two, glacial acetic acid or methyl alcohol; it is very slowly soluble in acetone. The next lower members, corresponding approximately to the dinitrate, C 6 H 8 05(N03)2, are soluble in ether-alcohol, acetic ether, acetic acid and methyl alcohol, while the mononitrate, C6H9O5NO3 is very soluble in ether-alcohol, acetic ether and absolute alcohol. A consider- able number of nitrates have been formed but it has proved very difficult to isolate any one of them in a pure condition and from a careful study of the work of G. Lunge, 1 Cross and Bevan have reached the conclusion that " the stages of nitration of cellulose are not molecular stages, but represent progressive increments of the esterifying groups in a mass-aggregate, which is the reacting unit." 2 The general properties of the cellulose nitrates are: (1) nitric 1 G. Lunge: J. Am. Chem. Soc, 1901, 23, 527. 2 Cross and Bevan: Researches on Cellulose, II, 1900-1905, 44. GUN-COTTON 15 acid may be removed by warming with alkaline solutions, the amount removed depending on the concentration of the alkali; (2) nearly all of the nitric acid is expelled by treatment with cold concentrated sulphuric acid; (3) boiling with ferrous sul- phate and hydrochloric acid drives off the nitrogen as nitric oxide; (4) alkaline sulphydrates, ferrous acetate and numerous other substances convert the nitrate into ordinary cellulose. The various cellulose nitrates find many very important com- mercial uses. Mixed with castor oil they are extensively used in the manufacture of artificial leather. In solution in ether- alcohol cellulose nitrate is employed in the manufacture of artificial silk by the Chardonnet process, the fibre being finally denitrated by treatment with ammonium sulphide to render it less inflammable. The fibrous nitrates may be reduced to plastic masses by kneading with solvents and in this condition may be formed into articles of any desired shape. As films they find use in photography, as the carrier for the emulsion; in the solid form, after the incorporation of camphor, they are spoken of as celluloid or xylonite and find innumerable uses. The nitrogen content of the nitrates for various purposes is given by Mork 1 as follows : Celluloid and films for moving pic- tures about io| to n| per cent; varnishes and lacquers n| to 12 per cent; for powder purposes 12 to 12! per cent; and for gun-cotton 13 to 14 per cent. In the higher nitrates the pro- portion of oxygen is such that upon decomposition the products are entirely gaseous and it is upon this property that their use as explosives depends. Gun-cotton. Because of the similarity between many of the methods used in paper making and those in the manufacture of gun-cotton a very brief description of the methods used in the latter industry may be of interest. Cotton is still practically the only cellulose used although it has apparently recently been demonstrated that certain grades of wood pulp will make acceptable substitutes. The cotton is obtained in the form of spinning wastes or of the short fibres 1 Mork: J. Frank. Inst., Sept., 1917. I 6 CELLULOSE from the seeds known as linters. Where spinning wastes are used the fibres are first degreased by treatment with some solvent, then boiled with caustic soda, bleached with bleaching powder solution or with calcium sulphide, washed, neutralized with sulphuric or hydrochloric acid, again washed, and finally dried. It should contain no chlorides, sulphates, oxycellulose or hydrocellulose, but it often contains mechanical impurities such as wood, string, colored threads, metal, etc., and to re- move these it is hand picked as it passes along a conveyor to a " willow" which opens out the lumps and knots. After leaving the "willow" it is again hand picked and then is dried by hot air, weighed in charges of the desired size and cooled in closed containers. Various methods of nitration are in use but all depend on the immersion of small quantities of cotton in comparatively large volumes of mixed nitric and sulphuric acid. The time of nitration varies from 30 minutes to 24 hours according to the method employed. The speed of nitration increases rapidly with rise of temperature, but the yield decreases although the nitrogen content of the product remains practically constant. After nitration the excess acid is removed by centrifugal action and the fibrous nitrate is washed by rinsing and is then boiled either with water alone or with the addition of a very little alkali in order to remove traces of free acid and to decompose and dissolve unstable impurities. This boiling operation, with intermediate washings with cold water, sometimes lasts 4 to 5 days. The next operation is that of pulping the washed nitrate; this is done in beaters very similar to those used in paper manu- facture but slightly modified in order that thorough agitation and no settling may take place. During the pulping the fibre is reduced in length and is at the same time washed continu- ously with hot water in order to remove the last traces of acid. Further washing is given by agitation, settling and removing suspended impurities after which enough alkali is added to leave in the finished gun-cotton 1 to 2 per cent of alkaline matter ACETATES I 7 calculated as CaC0 3 . The pulp is then run into moulds with bottoms of fine wire gauze and the water removed by suction and by hydraulic pressure. The slabs thus formed are used wet or after drying according to the purpose for which they are desired. For the manufacture of smokeless powder the washed pulp is screened and then dried by centrifugal action; it still con- tains much water and this is removed by treatment with ethyl alcohol. Finally a little ether, or other volatile solvent, is kneaded in which produces a paste ready for the blocking operations. The solvents are removed during drying so that very little remains in the finished product. Nitrites. When viscose silk is treated with nitrous gases in the presence of nitric acid nitrites of cellulose are formed. They are not soluble in water, alcohol, acetone, ether, chloroform, or ethyl acetate. Nitrogen is given off slowly at ordinary tem- peratures and rapidly on heating. Nitrites are liable to occur in nitrocellulose and cause rapid deterioration. 1 Acetates. Cellulose acetates are formed when cellulose is treated with acetic anhydride under quite widely varying con- ditions. The monoacetate, which is formed at no degs., is insoluble in all neutral solvents and in the solvents of cellulose. At 140 to 160 degs. higher acetates are formed, accompanied by solution in the reaction mixture, while in the presence of catalytic agents — zinc chloride, sulphuric acid, or phosphoric acid, for instance — the reactions take place at much lower temperatures, due doubtless to the formation of hydrocellulose which acetylates more rapidly than the normal cellulose. Ost, 2 who has studied the formation of acetates by three different processes, finds that all yield the triacetate, CeH^Os^HsC^, but at the same time he doubts the existence of a triacetate of normal cellulose and considers these compounds as derivatives of a series of hydrocelluloses. This feature of preliminary hydrolysis seems to be one of the essentials of acetate forma- 1 Nicolardot and Chertier: Compt. rend., 1910, 151, 719-722. 2 Ost: Z. angew. Chem., 1906, 19, 993. l8 CELLULOSE tion. Acetylation of the fibrous celluloses without appreciable structural change may be accomplished by diluting the reagents with hydrocarbons. The higher acetates are soluble in acetone, phenol and chloro- form and the solutions are of high viscosity. Boiling the ace- tates with alkaline solutions splits off the acetyl groups and the cellulose is regenerated. In contrast to the nitrates the acetates are non-explosive, and since they can be dissolved by appropriate volatile solvents to homogeneous solutions they are admirably adapted for use in the preparation of films, threads, etc. Commercial acetates are on the market in the fibrous, granular or powdered forms or in solutions of various viscosities. They are used for nearly all purposes for which the nitrate is used except for explosives and their use would be still more extended but for the fact that their cost of manufacture is appreciably greater than that of the nitrate. It is interesting to note that acetate silk is the only one in which the final product retains the ester composi- tion, since in other cases reactions take place which leave the thread simply as hydrated cellulose. This is probably the reason why wetting reduces the strength of acetate silk so much less than it does the other kinds. Cellulose-Sulphuric Esters. The action of concentrated sul- phuric acid on cellulose causes the formation of a series of esters which have been described as cellulose sulphuric acids but which are more probably derivatives of resolution products. The first stage of the reaction is, according to Stern, 1 the forma- tion of a disulphuric ester, C 6 H 8 03(S04H) 2 , which is soluble in water while its calcium barium and lead salts are insoluble in alcohol. This reaction of cellulose and sulphuric acid is of great importance in processes of esterification, where the acid acts as a catalyst, first combining with the cellulose and then being replaced by the ester forming groups. In the case of nitrates and acetates this substitution is never quite complete, and traces of SO4H remain fixed in both compounds. The 1 Stern: J. Chem. Soc, 1895, 1, 74-90. FORMATES 1 9 presence of such residues 'in the nitrate renders it unstable and its removal is one of the chief reasons for the extended boiling and washing treatments which gun-cotton undergoes. Mixed Esters. As may be concluded from the above the joint action of sulphuric acid with other esterifying agents can result in the formation of mixed esters containing SO4H groups as well as other negative groups. In this way aceto-sulphates are formed by the action of acetic anhydride, glacial acetic acid and sulphuric acid. 1 These contain from 5 to 25 per cent of combined S0 4 and may be grouped in three classes ac- cording to their physical properties. Those with most S0 4 are soluble in water, the others in acetone or dilute alcohol. In a similar way aceto-nitro-sulphates and nitro-benzoates have been formed. Benzoates. The action of benzoyl chloride in the presence of alkali hydroxides results in the formation of cellulose benzoates. The monobenzoate is formed with only slight structural changes when cellulose is treated with a 10 per cent solution of caustic soda and shaken with benzoyl chloride. The dibenzoate is formed in the presence of 15 per cent caustic soda solution, the fibrous celluloses being disintegrated, as the dibenzoate is an amorphous substance. This compound is soluble in acetic acid and chloroform. Formates. Formylated cellulose may be made by treating hydrocellulose with formic acid in the presence of zinc chloride. 2 Its slight solubility in organic solvents and its lack of thread, and film-forming ability make it rather unpromising. Accord- ing to German patents, 3 solutions of the following compounds may be used as solvents of cellulose formate: iodides and bro- mides of the alkali metals, metallic nitrates as well as those of ammonia and the alkaline earths, cupric chloride, soluble bichro- mates, alkali xanthates, aniline salts and alkali salts of aromatic mono- and polysulphonic acids. 1 Cross, Bevan and Briggs: Berl. Ber., 1905, 38 and 1859. 2 Worden: J. Soc. Chem. Ind., 1912, 31, 1064. 8 Ger. Pats. 266,600 and 267,577, July 5, 1912, and Feb. 26, 1913. 20 CELLULOSE Alkali-Cellulose. The action of solutions of sodium hydroxide of 12 to 15 per cent strength causes considerable change in the structure of fibrous celluloses, particularly cotton which is changed from a flattened twisted ribbon with a large central canal to a thickened cylinder in which the canal shows very little. When this action takes effect on cloth there is a con- siderable shrinkage both in length and width but at the same time the cloth gains in strength. The amount of shrinkage varies with the strength of solution employed; it is practically uniform for solutions of sp. gr. 1.00 to 1.10, while there is a sudden increase at sp. gr. 1.10 to 1.12. Between this point and sp. gr. 1.225 there is a relatively rapid increase in shrinkage but beyond the latter point the shrinkage again diminishes. If the goods are kept under tension during the reaction the physi- cal changes give to the material a peculiar, silky lustre. This reaction is spoken of as mercerization from the name of Mercer by whom it was first discovered. The effects thus obtained are due to a definite chemical com- bination of cellulose and caustic soda in the proportions C 6 Hi O5 to 2 NaOH, accompanied by combination with water. This compound is entirely dissociated by water, the alkali being recovered unchanged while the cellulose remains in the hydrated condition. When it is treated with alcohol equilibrium is reached when part of the alkali is removed, the residue being C 12 H 2 oOio • NaOH. 1 Cellulose-Sulpho-Carbonates. The above mentioned alkali- cellulose-hydrate, containing 30 per cent cellulose, 15 per cent caustic soda and 55 per cent water, is the first step in the forma- tion of cellulose-sulpho-carbonate which is also known as sodium- cellulose-xanthate and viscose. This compound is prepared by acting on the alkali cellulose with carbon disulphide at ordinary temperatures. In practice bleached cotton or wood pulp is treated with an excess of 15 to 18 per cent caustic soda solution and then pressed until it retains 2\ to 3 times its weight of the solution. This is then treated in a closed vessel with carbon 1 Cross and Bevan: Researches on Cellulose, II, p. 13. CELLULOSE-SULPHO-CARBONATES 21 disulphide amounting to about half the weight of the cellulose. At the end of about three hours at ordinary temperatures water is added and the mass allowed to stand for some hours to com- plete its hydration; on stirring a homogeneous solution results, which may be diluted to any desired degree. The impure compound is yellow in color due to by-products of the reaction, but the pure material, which may be prepared by treating the crude solution with alcohol or saturated brine, is obtained in the form of greenish white flocculent masses. These redissolve in water to a faintly yellow solution and from such solutions the xanthates of the heavy metals may be pre- cipitated by adding solutions of the corresponding heavy metal salts. Viscose solutions may be evaporated at low temperatures to solids which are completely resoluble in water, but if the solu- tions are heated to 70 to 80 degs. they thicken and at 80 to 90 degs. coagulation takes place very rapidly. Mineral acids neu- tralize the total alkali in the viscose and cause precipitation of hydrated cellulose while organic acids are not sufficiently strong to decompose the sulpho-carbonate. The most characteristic property of viscose is its spontaneous decomposition with formation of hydrated cellulose, caustic soda and carbon disulphide or its reaction products. With aqueous solutions of greater strength than 1 per cent cellulose this de- composition causes the formation of a jelly of the shape and volume of the containing vessel. This jelly gradually contracts with the expulsion of water. Observations on 100 c.c. of a 5 per cent solution kept in a stoppered vessel at ordinary temper- atures showed the following rates of coagulation and shrinkage. 1 (See table on p. 22.) The cellulose regenerated from viscose differs from the original in being more hygroscopic, as well as being hydrated, and in being more reactive toward bases but less so toward acid groups. Viscose finds extensive use in the manufacture of artificial silk and in the preparation of films for transparent wrappings. 1 Cross and Bevan: Text-Book of Paper Making, 3rd ed., p. 25. 22 CELLULOSE These latter are very thin, o.ooi inch, and as compared with nitrate or acetate films are much more water-absorbent. Coagulation First appearance of liquid Time in days 8th day nth 1 6th 20th 25th 30th 40th 47th Vol. of cellu- lose hydrate 83 5 72.0 58.0 42.8 38.5 Diff. from 100 c.c. = vol. expressed 16.5 28.0 42.0 57-2 61. s The Groups of Celluloses. Up to this point the remarks have applied chiefly to cotton, which may be considered as the typical cellulose, but there are also numerous other celluloses which differ more or less widely from this standard. The fibrous celluloses, for instance, are grouped, according to C. F. Cross, into three classes, depending upon their degree of resistance to hydrolytic and oxidizing actions, the amount of furfural which they yield when boiled with dilute hydrochloric acid, and their elementary composition as regards the ratio of carbon to oxygen. The characteristics of the three groups may be tabulated as follows : A Cotton group B Wood cellulose group C Cereal cellulose group Hygroscopic moisture Elementary composition C:0 Furfural 6-8% | 44.0-44.4 i 5 ° 0.1-0.4% J No active I CO groups 9-"% 43.0-43.5 5i 3-6% Some free CO groups 10-12% 41.5-42.5 53 12-15% Considerable re- activity of CO groups Other characteristics In group A are included cotton, flax, hemp, rhea (ramie), sunn hemp, etc. They are usually associated in the plant world with substances easily removed by digestion with alkalis. COMPOUND CELLULOSES 23 The purified celluloses of this group are considered chemically identical with cotton. Group B comprises celluloses obtained by the decomposition of compound celluloses, i.e., those from woods and lignified tissues in general. They may be considered as oxidized and partially hydrolyzed products and are more readily attacked by hydrolyzing agents than are the celluloses of group A . The fibres of group C are in most cases complex, both struc- turally and chemically. They are oxycelluloses and are still less resistant than the group B celluloses. They undergo gradual oxidation in dry air at a temperature of 100 degs. and become discolored. In still a fourth group may be classed those cellular, rather than fibrous, celluloses which offer low resistance to hydrolysis. These are easily resolved by boiling with dilute acids and are also soluble to some extent in dilute alkaline solutions. As the celluloses of this group are not employed in paper making no further discussion of their properties is essential in this work. Compound Celluloses. Passing from the consideration of the purified celluloses to what may be called their raw materials, it is found that plant physiologists recognize three modified or "compound" celluloses: cutocellulose, pectocellulose and ligno-. cellulose. Cutocelluloses contain, mixed with the tissues, various oily and waxy substances which render them quite water-resistant. The two principal types of these compound celluloses are cork and the cuticular tissues of leaves, stems, etc. Cork contains, in addition to oils and waxes, tannins, lignocelluloses and nitroge- nous materials. No celluloses of this type are employed in paper manufacture, hence a knowledge of their properties is only of incidental interest. The pectocelluloses may be considered either as compounds or intimate mixtures of cellulose and colloidal carbohydrates which are easily hydrolyzed by either acid or alkaline treat- ments to simpler, soluble materials. They are "saturated compounds" and therefore do not react with the halogens. Cel- 24 CELLULOSE luloses of this kind are widely distributed in the plant world and are extremely varied in composition and structural char- acter. Among the more important pectocelluloses are flax and such other non-lignified fibres as ramie, hemp, nettle fibres, sisal, esparto, bamboo, etc. Many of these are more or less associated with lignocelluloses. While many of the pecto- celluloses enter into paper making operations, it is not as pecto- celluloses, but only after the separation of the pectic constituents by means of alkaline hydrolysis. Lignocelluloses form by far the most important group of compound celluloses, so far as paper making operations are con- cerned, both because they are employed directly, as in the case of ground wood and jute, and also because they are the basic raw materials from which the greater part of the paper making celluloses are prepared. They have been studied in consider- able detail by Cross and Bevan, who worked on jute fibre, which they consider to be the typical lignocellulose. One essential feature of lignification is the formation of meth- oxyl groups, O • CH 3 , while a second ke tonic grouping, CO • CH 2 , is also a characteristic constitutional feature. This latter group doubtless plays an important part in the formation of acetic acid during hydrolysis. Woods are more pronounced ligno- celluloses than jute, having more non-cellulose constituents and yielding more methoxyl and furfural. They are still further differentiated from jute by their behavior toward cellulose solv- ents, since they yield, as a whole, to no process of solution and are, moreover, almost totally resistant to the thiocarbonate (viscose) reaction. In all but these two characteristics the lignocelluloses of jute and wood are very much alike. Jute lignocellulose is dissolved by cuprammonium and by zinc chloride either in aqueous or acid solution. Hydrolysis accompanies the solvent action, so that on precipitation the recovery of the lignocellulose is incomplete. Alkalis at high temperature attack and dissolve the lignin and the less resistant cellulose but at the same time some of the more resistant cellulose is also dissolved. The residual cellulose, COMPOUND CELLULOSES 25 in the case of jute, is very similar in composition to the normal cellulose. Acids, even in dilute solutions, rapidly disintegrate lignocelluloses at temperatures above 60 degs. Caustic soda and carbon disulphide dissolve only part of the lignocellulose, 50 to 75 per cent remaining undissolved but in such a hydrated and swollen condition that it has been known to occupy one hundred times the volume of the original fibre. This insoluble residue reacts with chlorine as does the original fibre. Oxidizing agents profoundly attack lignocelluloses with the formation at first of acid products and finally, on further oxidation, of carbon dioxide and water. Because of the difficulty of separating lignin from cellulose without causing decomposition or structural changes, its com- position is still regarded as uncertain. Klason * assigns to lignin the empirical formula C4oH 42 On and believes the evidence goes to show that lignin is not in chemical union with the cellulose. Cross and Bevan, 2 on the other hand, have concluded " that the fibre substance is not merely a mixture of cellulose with non- cellulose constituents, but that these are compacted together into a homogeneous though complex molecule by bonds of union of a strictly atomic character." They submit the follow- ing constitutional formula 3 for the lignin (lignone) of the typical lignocellulose, as representing its quantitative reactions of chlo- rination, resolution by bisulphites, production of acetic acid and estimation of methoxyl. / C °\ /°\ /°\ /OH _ HC CH-[CH 2 -CO]2-HC CH • CH • CH • CH/ )3 a TLC !cO CH3O.HC JcH.OCHa OH CeUuloses ^CH£ ^CO' Lignin has been investigated by Heuser and Skioldebrand 4 1 Klason: Beitrage zur Kenntniss der chemischen Zusammensetzung des Fichtenholzes. 2 Cross and Bevan: Cellulose, p. 134, London, 1895. 3 Cross and Bevan : Researches on Cellulose, III, 104. 4 Heuser and Skioldebrand: Z. angew. Chem., 1919, 32, 41-45. 26 CELLULOSE who prepared it from spruce wood sawdust by hydrolyzing with 42 per cent hydrochloric acid. The yield obtained was 33.12 per cent and the air dry Hgnin contained 9.25 per cent moisture and 0.485 per cent ash. It yielded no furfural but showed a copper number of 12.90 and a methyl value of 6.77 per cent. On destructive distillation it gave more charcoal and tar and less acetic acid than cotton or wood cellulose. Lignin shows a number of color reactions which are of much value in detecting the presence of incrusting matters in paper or its raw materials. Nitric acid gives a yellowish brown color; a solution of aniline sulphate in water stains lignified tissues yellow; paranitroaniline in hydrochloric or sulphuric acid solu- tion produces a bright yellowish orange color; while an alcoholic solution of phloroglucin acidified with hydrochloric acid devel- ops an intense rose red or magenta color. This latter reagent has been used as a basis for a quantitative method for estimating ground wood in papers. 1 Schwalbe, 2 however, states that the color reaction with phloroglucin fails in many cases where lignified fibres are present and considers that the results should be confirmed in every case by Cross and Bevan's chlorination reaction. One of the most characteristic reactions of lignin is that of direct combination with the halogens, and more especially with chlorine. In the presence of water, chlorine attacks lignin with the formation of a chloride, C 19 Hi 8 Cl 4 09. This chloride is bright yellow in color and is soluble in sodium sulphite solution with the production of an intense magenta red color. With jute the chlorine combining with the lignin amounts to 8 per cent of the lignocellulose and an equal amount is combined as hydro- chloric acid; with wood, on the other hand, the proportions are not equal, considerably more chlorine going to form hydro- chloric acid than corresponds with that which is used in chlo- rinating the lignin. This reaction with chlorine is simple and there is no indirect oxidation as a result of the reaction CI2 + H 2 = 2 HC1 + O. The chlorinated material is homogeneous and as 1 Cross, Bevan and Briggs: Chem. Ztg., 1907, 31, 725. 2 Schwalbe: Z. angew. Chem., 1918, 31, pp. 50 and 57. DETERMINATION OF CELLULOSE 27 the cellulose is unattacked, maximum yields of the latter are obtained. This reaction forms the basis of Cross and Bevan's method for the quantitative determination of cellulose. Determination of Cellulose. The amount of cellulose in any raw material determines the maximum yield of pure fibre which can be prepared by any of the chemical processes used in pulp making. Practically, however, the yield never reaches this the- oretical maximum because of the hydrolytic action of the cook- ing liquors and the oxidizing and solvent action of the alkaline bleaching agents used. A method for accurately determining the cellulose in wood, or any fibrous raw material, is therefore of great value to the paper maker, in that it enables him, on the one hand, to detect variations in his supplies and, on the other, to see how nearly his processes are approaching the optimum. A satisfactory method for the determination of cellulose must separate it in a form which is free from lignin and colored im- purities and which is as- pure as possible. Freedom from min- eral matter is of secondary importance since it can be allowed for or removed by treatment with dilute acids. The formation of oxycellulose and hydrocellulose must especially be guarded against since in their formation part of the cellulose is converted to soluble products and moreover they are easily attacked by many of the reagents- employed. Another consideration is that the process should involve no very complicated operations and that it should be capable of completion within a reasonably short time. Many methods for determining cellulose have been proposed, of which some do not remove all non-cellulose matter, some destroy part of the cellulose itself, and some, which give pure products, are too long and complicated for practical work. It is unnecessary to give the full details of all the proposed methods, but a very brief description may prove of interest to those who wish to study into the subject. The following outlines, to- gether with notes on the purity of the products, are taken from Renker's x work on methods of determining cellulose. 1 Renker: Uber Bestimmungsmethoden der Cellulose, Berlin, 1910. 28 CELLULOSE Konig * heats 3 grams of the sample with 200 c.c. of glycerine (sp. gr. 1.230) and 4 grams of concentrated sulphuric acid. The flask is fitted with a reflux condenser and the heating is con- tinued for just one hour at the boil (131 to 133 degs.); the residue is then washed with hot water, alcohol and ether, dried and weighed. This process strongly attacks the cellulose and the product is not pure cellulose though entirely free from pentosans. Cross and Bevan's 2 method depends upon the chlorination of the lignin by treating the moist material with chlorine gas and the subsequent solution of the chlorinated products by boiling with dilute sodium sulphite solution. The cellulose ob- tained is pure white and free from lignin, but consists, according to the terminology of Cross and Bevan, of a mixture of a and /3 celluloses, the latter being of low resistance toward hydrolytic agents and yielding furfural on boiling with hydrochloric acid. Modifications of Cross and Bevan's method in which chlorine water is used instead of the gas give lower yields and the product contains some oxycellulose. In H. Miiller's 3 process the sample, after extraction of the resins and boiling in water, is treated with 5 to 10 c.c. of dilute bromine water diluted with 100 c.c. of water. When the color indicates that the bromine is all used up, more bromine water is added until an excess of the reagent remains; this takes from 12 to 24 hours. The residue is filtered off and washed. This process is repeated until the fibrous residue is pure white. This process yields highly pure cellulose which is free from oxycel- lulose but too long a time is required, especially with woods which may require twenty repetitions of the treatment. Kla- son's 4 modification, involving preliminary treatment of the sample with calcium or magnesium bisulphite solution, gives lower yields than Miiller's original method and, although the 1 Konig: Z. Nahr. Genussm., i, 8 (1898). 2 Cross and Bevan: Cellulose, 95 (1895). 3 Hofmanns Ber. iiber die Entwickl. d. chem. Industrie, III, 27 (1877). 4 5 Internationaler Kongress fiir angewandte Chemie, 1903, 1, 309. DETERMINATION OF CELLULOSE 29 time required is less, the method is still unsatisfactory tech- nically. The Schulze-Henneberg 1 method consists in digesting the sample for 12 to 14 days at 15 C. with 0.8 part KC10 3 and 12 parts HNO3 (1.10 sp. gr.). This is followed by dilution, filter- ing and washing and then by digestion with dilute ammonia for 45 minutes at 6o° C. The residue is washed with ammonia and finally hot water. This process requires an excessive time and very careful control of the temperature, and the results obtained are low and irregular, especially with woods. Hoffmeister 2 treats the sample with hydrochloric acid (1.05 sp. gr.) and as much potassium chlorate as will dissolve during the reaction. The digestion is conducted at ordinary tempera- ture, — not over 17.5 C, — with frequent shaking for 24 to 36 hours or until all parts of the fibre have become a clear yellow. The residue is then washed, digested on the water bath for 1 to 2 hours with dilute ammonia, washed, dried and weighed. This process yields a product which is free from lignin but is some- what yellow; the cellulose is somewhat attacked and oxy cellu- lose is present. Digestion at 60 degs. with a large excess of nitric acid (10 per cent) has been proposed by Cross and Bevan. 3 When the fibre has changed to a yellow color it is washed and treated with a solution of sodium sulphite. This treatment hydrolyzes and dissolves the /3 cellulose and leaves only the more resistant a cellulose; this results in low yields as compared with the chlorination process. The products are free from lignin but contain oxycellulose and apparently also hydrocellulose. Lifschutz 4 digests the material for 14 to 16 hours at a tem- perature of 45 to 60 degs. with 10 parts of a mixture of 32 per cent sulphuric acid, 18 to 20 per cent nitric acid and 48 to 50 per cent water. With wood, very low yields are obtained and the prod- 1 Ann., 146, 130 (1868). 2 Landw. Jahrb., 17, 240 (1888). 3 Cellulose, 97 (1895). 4 Ber., 24, 1188 (1891). 30 CELLULOSE uct forms a brownish horny mass which is free from lignin but contains considerable oxycellulose. Schwalbe * treats the moist material for several hours with gases containing oxides of nitrogen, washes and heats on the water bath with a 2 per cent Na 2 S0 3 solution. The cellulose thus obtained is free from lignin but contains oxidation products and is brownish yellow. The results are very variable when woods are analyzed. Zeisel and Stritar 2 suspend the sample in dilute nitric acid and add a 3 per cent solution of potassium permanganate a cubic centimeter at a time until the color persists for half an hour; during this treatment cooling and stirring are necessary. The manganese salts are removed by SCVwater or NaHS0 3 and after washing the residue is digested with dilute ammonia. It is claimed that hemicelluloses and about 4 per cent of the cellulose are dissolved. The method gives low and variable results, forms much oxycellulose and is unsatisfactory for woods. Neutral permanganate does not remove all the lignin, while if it is used with acetic acid the product is free from lignin but contains oxycellulose. Permanganate with hydrochloric acid works fairly well with cotton, jute and sulphite, but with woods the treatment must be many times repeated and the results are low. Sodium hypochlorite was found to give low results and though the product was free from lignin it was very high in oxycellulose. 3 After a large number of experiments with the different meth- ods, Renker concludes that there is no absolutely correct method for determining cellulose but that the method of Cross and Bevan using chlorine gas is the most satisfactory. Konig and Huhn, 4 on the contrary, contend that Cross and Bevan's method gives too high results because it fails to remove pentosans and hemicelluloses which remain as impurities in the product. They 1 D.R.P., 204,460. 2 Ber., 35, 1252 (1902). 3 Renker: Bestimmungsmethoden der Cellulose, 79. 4 Z. Farben-Ind., 1911, II, 297 et seq. DETERMINATION OF CELLULOSE 31 claim that only hydrolysis followed by oxidation can free the true cellulose from all impurities. This brings up the question of what constitutes pure cellulose, which has never been settled beyond dispute and about which there are probably as many opinions as there are methods for its determination. The results obtained by Renker, in his investigation of the different methods for the determination of cellulose, are given in the accompanying table. The percentages are based on mate- rial free from ash, fat or rosin, and water-soluble substances. Method of analysis Glycerine-sulphuric acid. — Konig Chlorine gas. — Cross and Bevan Concentrated chlorine water Dilute chlorine water Bromine water. — H. Muller Bromine water. — Muller- Klason Nitric acid and potassium chlorate. — Schulze Muriatic acid and potassium chlorate. — Hoffmeister Nitric acid. — Cross and Bevan Nitrous acid. — German pat. 204,460 Nitric-sulphuric acids. — Lifschiitz Potassium permanganate and nitric acid. — Zeisel and Stritar Potassium permanganate, neutral Potassium permanganate and acetic acid . . . Potassium permanganate and hydrochloric acid Hydrogen, peroxide Sodium hypochlorite Phenol. — German pat. 94,467 Material Sulphite cellulose Per cent 74-15 97-9 97-65 98.0 98-i 96.6 98.05 98.25 97-65 90.6 98.5* 98.25 97-9 96.05 97-4 9C75 Jute Per cent Wood Per cent 55 43 35 Cotton Per cent * Inapplicable because lignin is not all removed. All things considered the Cross and Bevan method, employing chlorine gas, is doubtless the best means of determining cellu- lose, especially where woods are under consideration, since it has been proved by Heuser and Sieber 1 that lignin can be com- pletely removed by chlorination without any oxidation of the 1 Z. angew. Chem., 26 (1913), 801. 32 CELLULOSE cellulose, and because the color produced by sodium sulphite on chlorinated lignin is characteristic and forms a good indicator to show the degree of purification. If, however, the chlorine is allowed to act after the lignin is all removed the cellulose is gradually converted to oxycellulose, which is soluble in sodium sulphite, and lower yields result. According to one's under- standing of pure cellulose it may or may not be advisable to accompany this treatment with one of acid hydrolysis. Even where Cross and Be van's method has been employed the different methods of operation render the figures of different observers hardly comparable. The wood must be very finely divided and of uniform particles. If coarse pieces are present the lignin chloride formed on the surface prevents further pene- tration of the gas and long exposure is of little use unless the lignin chloride is removed. Variation in the size of the wood particles also means that a portion will be over^treated in order to insure the complete removal of the lignin from the rest, and this results in lower yields. To avoid many of the manipula- tive variations Johnsen and Hovey x worked on wood reduced to a fine " sawdust" by rasping with a suitable wood rasp, and screened to pass an 8o-mesh but not a ioo-mesh sieve. This they chlorinated in a Gooch crucible, the small plate of which was carefully covered with fine bleached cloth sewn on with cotton thread. Their method, which includes a preliminary hydrolysis, is in detail as follows: Two samples of about one gram each of air dry sawdust are weighed exactly, transferred to small flasks and heated for one- half hour with alcohol, on a water bath, filtered through the crucibles and washed with hot alcohol. The material is then transferred from the crucibles to 150 c.c. flasks and covered with about 75 c.c. of glycerine-acetic acid mixture (60 grams glacial acetic acid and 92 grams glycerine, 1.26 sp. gr.). The flasks are then heated in an oil bath at 135 C. for four hours, using glass tubes as reflux air condensers. The material is then collected in the crucibles, washed well with hot water and the 1 Pulp and Paper Mag. Can., XVI (1918), 85. DETERMINATION OF CELLULOSE 33 crucibles, after cooling, are placed in an apparatus permitting washed chlorine gas to be passed directly through the crucible and its contents. After passing chlorine through continuously for 20 minutes the free gas is removed by washing once with a cold, weak solution of sulphurous acid in water, and the cru- cibles are placed in small beakers which are filled not quite to the top of the crucibles with a 3 per cent solution of sodium sulphite. After the beakers have been heated in a water bath for three-quarters of an hour the material in the crucibles is washed with hot water, using a filtering flask, and allowed to cool. The chlorination process is then repeated three times exactly as before except that the duration of gassing is 15, 15 and 10 minutes. After the last treatment with sodium sulphite the fibres are thoroughly washed, dried at 105 C. for four hours, or to constant weight, and weighed in closed weighing bottles. The moisture in a separate sample of the original sawdust having been determined, the per cent of cellulose can be calculated over to the basis of bone dry wood. Working by this method the lignin may be completely re- moved from finely divided wood in four chlorinations totalling one hour's exposure to the gas. If the treatment with glycer- ine and acetic acid is omitted the results are 2 to 4 per cent higher, indicating the extent to which hydrolysis removes sub- stances which are not dissolved by the straight chlorination process. CHAPTER II FIBROUS RAW MATERIALS The Vegetable Cell. The structural unit of which the plant fabric is composed is the cell, and even those elements which are least cellular in appearance, as fibres, ducts, etc., are only transformed cells, or simple combinations of them. In all of the higher forms of plant life the living cell consists essentially of protoplasm surrounded by a wall of cellulose. The proto- plasm is practically transparent and colorless but it is seldom found without an admixture of other matter which gives it a granular appearance. Chemically considered it is a very com- plex substance belonging to the group of albuminoid bodies and is quite similar in composition and appearance to the albu- min which forms white of egg. The cell wall is produced from the substances contained in the protoplasm and is formed in close contact with the limiting film of the latter. At first it is an even, homogeneous film showing no obvious structure but it soon becomes modified in appearance and composition according to its location in the plant; in some cases it becomes capable of absorbing a large amount of water with a consequent increase in volume and the formation, upon warming, of a thick mucilage. In other cases it becomes mineralized by the deposition of silica or calcium salts upon or within the cell wall, the effect being a hardening and stiffening of the structure as illustrated by many of the grasses and straws. A third modification of the cell wall is known as cutinization or suberification. Ordinary cell walls absorb water freely but in certain parts of the plant they are water-repellent due to the presence of cutin or suberin. This water-proofing of the cell wall may be superficial only or it may permeate the entire structure of the wall as in the case of cork. 34 THE VEGETABLE CELL 35 A fourth change, which, from the point of view of the paper maker, is far more important than the others, is known as ligni- fication and is caused by the deposition in and upon the cell- wall of a substance somewhat similar to cellulose and which is termed lignin. This material, although generally spoken of as a single substance, is in all probability made up of several closely related substances. It forms by far the greater part of the incrusting matters which are removed by the treatment with chemicals in preparing celluloses for paper making. The chemical properties of the lignocelluloses have already been described in the preceding chapter. The thickening of the cell wall, which up to a certain age accompanies the deposition of lignin, seldom takes place uni- formly over its entire surface. As a result the fibres may show ridges, depressions, pits, rings, spirals and other markings, which are often so characteristic as to be of great assistance in the identification of the fibres in which they occur. In some cases it is easy to see that the markings are pores or fissures running from one cell to the next. Such pores and the thinner portions of the cell walls aid in the transmission of sap. In many plants gums and resins are formed by processes which are not well understood but which are probably in large part degradations of the cell walls. The products of such changes are often found as small, irregular drops either within the cells or in the spaces, known as resin ducts, arising from the confluence of a number of cells. The resins are soluble in alco- hol or in solutions of the alkalis; they are colored reddish orange by a solution of Sudan III in diluted alcohol and when stained in this way may be readily recognized under the microscope. It is only the living cells which contain protoplasm and these are found at the points where growth is taking place. The older cells, where growth has ceased, occasionally contain a little water but generally only air, more or less highly rarefied. They are still of value in the plant structure for it is from such lifeless cells that stiffness and strength are derived. Cells are present in different plants in an almost infinite diversity of form, but, $6 FIBROUS RAW MATERIALS except for those cases which assist in the identification of the fibres which they accompany, the following discussions will be confined to the long, slender pointed elements of the bast or woody tissues, which, with the single exception of cotton, are the only ones of practical interest to the paper maker. Such fibres are derived from the ordinary cells simply by growth and change of form. The paper making fibres may be divided, according to their source, into four classes. i. Seed hairs, of which cotton is the only representative. 2. Bast fibres, as hemp, flax, manila, etc. 3. Those derived from whole stems, such as straw, esparto, or bamboo, and which are associated with various cells and vessels which are not properly fibres. 4. Fibres derived from wood. Seed Hairs Cotton (Gossypium). The cotton fibre consists of a single hair-like cell, which when fully ripe is flattened and twisted. This appearance is a characteristic of fully matured cotton and is not shown by unripe fibre or that which has been injured during growth. The fibres form the covering of the cotton seed and are removed from the seed by ginning. The length of the cotton fibre varies from 2 to 5.6 cm. and the diameter from 0.0163 to 0.0215 mm.; the longest fibres are found in the Sea Island cottons, followed by Egyptian, Brazilian, American and East Indian. The cell walls of the mature cotton are thin and often present a granulated appearance, or highly characteristic cross-markings. Raw cotton has been examined by Schunck who found in it two coloring matters, both containing nitrogen; a wax similar to carnauba wax; albuminous matter and a solid crystalline fatty acid. Miiller has analyzed raw cotton with the following results : LINEN 37 Per cent Water 7. 00 Cellulose 91 .35 Fat o. 40 Aqueous extract (containing nitrogeneous substances) o. 50 Ash o. 12 Cuticular substance (by difference) o. 63 Cotton reaches the paper mill in the form of rags and spinning waste. The short fibres removed from the seed hulls by me- chanical processes are also occasionally available and can be used in high-class papers. Bast Fibres Under this heading are included the fibres found in the inner bark of various plants. These are often present in the plant in the form of bundles of considerable length and it is these fibre bundles or filaments which are of particular interest in textile work, while the paper maker is concerned only with the ultimate fibres composing these filaments. The fibres are usu- ally firmly attached to those above and below them, either by incrusting matters or by partial identity of their cell walls, so that a chemical treatment is generally necessary in separating them. The filaments, on the other hand, are isolated from the surrounding tissues by various mechanical or chemical processes, one of the commonest being retting as applied to flax and hemp. The walls of bast fibres are usually of considerable thickness and the central canal varies greatly in the different species and even in different individuals of the same species. The irregularities of the cell walls are such as to cause thickenings or knots in some cases, projections into the cell cavity or out- ward from the walls, and various other characteristic appear- ances. Variations in the amount of lignification and in the nature of the materials deposited in and on the cell walls are frequently sufficiently characteristic to assist materially in their identification. Linen. Linen is composed of the bast fibres of the flax plant Linum usitatissimum. The plant yields about 8 per cent of 38 FIBROUS RAW MATERIALS fibre which is separated by retting and is then known as flax. The ultimate fibres are 6 to 60 mm. long and 0.012 to 0.026 mm. wide with an average ratio of length to width of about 1200. They are thick walled tubes with thickened places or knots at intervals. The ends are tapered, the walls rather transparent, and the canal is small. Miiller gives the analyses of two samples of Belgian flax as follows: Per cent Per cent Water 8.60 81.99 2.37 3.62 O.70 2 .72 10.56 Cellulose 7° -55 Fat and wax 2-34 5-94 Ash 1 .32 Pectous substances 9.29 Linen or flax reaches the paper mill in the form of scutching refuse, spinning waste, threads, rags, etc. It requires much the same treatment as given to cotton except that it is necessary to employ a rather more severe bleaching process. Hemp (Cannabis sativa). The fibre is prepared by retting, from filaments which run the entire length of the stem. The ultimate fibres composing these filaments vary from 5 to 55 mm. and average about 22 mm. long by 0.022 mm. in diam- eter, the ratio of length to diameter being therefore about 1000. The fibres have very thick walls which are not very highly lignified. The ends are large and sometimes flattened and the central canal is almost obliterated. In microscopic appearance the fibres are very similar to those of flax; they show the same knots or thickenings and the same striae, but they differ from linen in having more ability to break down into fibrillae during the mechanical processes of paper making. Miiller gives the following as the analysis of a sample of raw Italian hemp. Per cent Water 8.80 Cellulose 77 • 13 Fat and Wax 0.55 JUTE 39 Per cent Aqueous extract 3 . 45 Ash o. 82 Pectous substances 9. 25 Many other plants yield fibres to which the name hemp is given but they are generally distinguished as manila hemp, sisal hemp, sunn hemp, etc. Hemp comes to the paper mill in the shape of rags, rope and cordage, etc. It is used in high-grade papers, particularly very thin sorts, where its ability to split into fibrillae makes it espe- cially valuable. Manila Hemp (Musa textilis). Manila is prepared from the outer sheath of the stems of the Musa, which is a species of banana, by stripping, scraping and drying. It is sometimes further purified by washing and beating. The ultimate fibres are from 3 to 12 mm. long and average about 6 mm.; the width varies from 0.016 to 0.032 mm. with an average of 0.024 mm. The fibres taper very gradually toward the ends; the central canal is large and very prominent, while fine cross-markings are numerous. The composition of raw manila is given by Muller as follows: Per cent Water n. 73 Cellulose 64. 07 Fat and wax o. 62 Aqueous extract o. 96 Ash 1.02 Lignin and pectous substances 21. 60 The paper maker obtains manila fibre almost entirely in the form of old ropes and cordage. It is generally given a compara- tively light alkaline treatment with lime and used in the un- bleached condition for papers where strength is of far more importance than cleanliness or color. It is occasionally bleached to a light yellowish shade for some grades of paper. Jute (Corchorus capsularis and C. olitorius). The filaments are obtained by a retting process, and are used in the manu- facture of twine, cordage, and woven goods such as burlap. 40 FIBROUS RAW MATERIALS It is in these forms and also as "butts" and "rejections" that it reaches the paper maker. The fibres of jute are about 2 mm. long and 0.022 mm. in diameter. They are thick- walled and the central canal is very variable, at times being of considerable width and then narrow- ing to hardly more than a line. The surface is quite smooth and at intervals may be noticed radial canals and joints similar to those in linen though not so pronounced. As used in paper making the fibres are not completely separated, so that bundles of fibres are of frequent occurrence. Jute has been carefully studied by Cross and Bevan who regard it as typical of lignified fibres. They consider it as a chemical unity, which they term lignocellulose, and which splits up into cellulose and other products on treatment with suitable reagents. Muller gives the composition of the raw fibre as follows: First quality Butts Water Per cent 9-86 63.76 0.38 1 .00 0.68 24.32 Per cent 12 .40 Cellulose 60.89 O.44 3-89 1 .40 20.98 Fat and wax Aqueous extract Ash Lignin or non-cellulose Adansonia (Adansonia digitata). This fibre, which is derived from the inner bark of the baobab, or monkey bread tree, of Africa, has been used to some extent in preparing papers where the very highest strength is a necessity. The composition of the bast as exported varies, as shown below: 1 Per cent Per cent Water 10.90 49-35 0.94 13-57 6.19 I9-05 13.18 Cellulose 58.82 Fat and wax 0.41 Aqueous extract 7.08 Ash 4.72 Pectous substances 15 .19 Griffin and Little: Chemistry of Paper Making, p. 128. FIBRES DERIVED FROM WHOLE STEMS 4 1 Paper Mulberry Tree {Broussonetia papyrifera). The fibres of the paper mulberry are used by the Japanese in the prepara- tion of some of their peculiar papers. They are separated from the inner bark of the tree by scraping, soaking and mace- ration in water, and are sometimes purified still further by boiling in weak alkaline solutions. As used in Japanese papers the fibres are generally unbroken. The fibres are long and slender, varying in length from 6 to 20 mm., with an average width of 0.030 mm. They are nearly transparent when viewed under the microscope, and show trans- verse jointings as well as longitudinal striae. The central canal generally shows as a well-defined line and the ends are some- times blunt and rounded, sometimes fringed. Agave. Fibres are prepared from the leaves of various species of agave by maceration or scraping. The filaments thus obtained are light colored, lustrous and comparatively stiff. Among the most common of the fibres of this class is sisal hemp or heniquen, which is largely employed for cordage, bags, etc., in which forms it reaches the paper mill. The ultimate fibres are long, of rather small diameter, taper- ing and pointed at the ends. The central canal is not promi- nent but can be seen as a narrow line in some of the fibres. The walls are thick and are characterized by many fine cross lines, close together, which are found on nearly every specimen. The fibres are harsher than paper mulberry and longer than manila. Fibres Derived from Whole Stems The fibres in this class, being produced by treatment of the entire stems, include the ultimate fibres from all structural ele- ments of the stem proper, together with cells from the epidermis and other parts of the plant. Straw. In straw pulp the bast cells or fibres form the greater part of the pulp. These are comparatively short and slender with sharp pointed ends; at quite regular intervals the walls appear to be thickened and drawn together to resemble joints. 4 2 FIBROUS RAW MATERIALS The dimensions of straw fibres vary with the kind of straw and with the conditions of growth, nature of soil, etc. They are longer than those from esparto but not so long as the fibres from spruce wood and would compare more nearly with poplar fibre in paper making value. Accompanying the bast fibres in straw pulp are numerous epidermal cells with toothed or serrated edges, and also smooth, thin-walled cells from the pithy portion of the stem. The latter vary in shape from nearly round to long, oval cells whose length is several times their width. Both types of cell aid materially in the identification of straw pulp. Esparto {Stipa tenacissima and Lygeum Spartum). The bast fibres are grouped in bundles or filaments which are resolved into ultimate fibres by the chemical processes employed. The fibres are shorter and more even than those from straw, aver- aging about 1.5 mm. in length, and the central canal is nearly closed. Serrated cells are numerous but considerably smaller than those from straw, while the smooth, thin-walled cells are absent. The chief characteristic which distinguishes esparto from straw and other fibres is the presence of small tear-shaped cells derived from the hairs on the surface of the leaves. The composition of esparto is given by Miiller as follows : Spanish African Water Per cent 9-38 48.25 2 .07 IO.19 3-72 26.39 Per cent 8.80 Cellulose Fat and wax Aqueous extract Ash Pectous substances 34.80 2.62 9.81 3-67 29.30 Cross and Be van 1 give the percentage of cellulose in air dry esparto as follows: Percent Spanish 58. o Tripoli 46. 3 52.0 45-6 1 Cross and Bevan: Text Book of Paper Making. Arzew. Oran. . BAGASSE AND CORN STALKS 43 Bamboo. Bamboo fibres closely resemble those from the straws in many of their characteristics. According to Raitt 1 the average length of the ultimate fibres is from 2.20 mm. to 2.60 mm. according to the variety, and the diameters are from 0.018 to 0.027 mm. While not so long as spruce fibres they are much longer than those from any of the deciduous trees. Accompanying the true fibres are numerous serrated cells and ovoid pith cells of various sizes and shapes. Cells similar to the tear-shaped cells of esparto have also been noticed. Many of these different cells are very small and a good proportion of them would be lost during the preparation of the pulp. The number of species of bamboo runs into the hundreds and not all of them are suited for paper making work because of the difficulties in reduction to pulp and bleaching to good color. Rushes. In many districts there are large areas of swamp- land densely covered with rushes of various kinds. Investi- gation has shown that the fibres in these are very similar to esparto and that pith cells are also present though they are far smaller than those from straw and many would be lost during the washing and bleaching of the fibre. Examination of rushes from South Carolina showed that they contained 35 per cent of cellulose and that of this about 30 per cent consisted of fine pith cells which would wash easily through a 100-mesh sieve. Rushes appear to require a more drastic cooking treatment than straw or esparto and the fibres bleach a little harder. They are, however, potential sources of fibre which can be availed of when necessity arises. Bagasse and Corn Stalks. These two materials are so similar in their fibrous characteristics and in the treatment necessary in cooking them that they may be considered as practically identical. Bagasse, the crushed stalks of the sugar cane, is produced in large quantities in the sugar industry and is gen- erally burned as fuel, though its value for this purpose is com- 1 Raitt: Indian Forest Records, Vol. Ill, Part III. 44 FIBROUS RAW MATERIALS paratively low because of its wet condition. It is estimated that seven States in the "corn belt" produce annually about eighty million tons of corn stalks of which the greater part is practically a waste. Both of these materials can be reduced to a pulp quite easily by the soda process and the pulp will bleach to a good white color with less than 10 per cent of bleach. The ash in the stalks is considerably lower than that in straw and contains much less silica, so that its effect on the recovery process is very slight. The pulp from both raw materials consists of long thick- walled fibres mixed with shorter fibres of similar structure and cells of various shapes and sizes. These include serrated cells of comparatively large size, long thick cells with rounded ends and many pith cells which are so thin-walled that nearly all become flattened during the reduction process. These pith cells are much larger than those from straw and are therefore much more difficult to remove by washing. They impart to the paper made from this fibre a hardness and rattle which are undesirable in many products, and as their separation from the rest of the fibre has proved very difficult the presence of pith has proved one of the chief stumbling blocks in the way of using either bagasse or corn stalks. Miscellaneous Materials. Among the many substances pro- posed as sources of fibre the following may be considered as falling in this class: papyrus x which grows in immense quanti- ties along the Nile and which yields about 33 per cent of easy- bleaching pulp when cooked by the soda process; asparagus 2 waste from canning factories and dry stalks at the end of the season; pea and bean vines 3 which according to Reinke yield better fibre than asparagus; cotton stalks, of which immense quantities are burned every year, but which so far have never been utilized successfully; zacaton, a Mexican grass, which has been carefully investigated by the Bureau of Plant Indus- 1 Beam: Chem. Section Bull. No. 2, Welcome Tropical Research Lab. Sudan. 2 Reinke: Chem. Ztg., 19 13, 37, 81. 3 Reinke: J. Soc. Chem. Ind., 1913, 594. woods 45 try; x tropical grasses, numerous species of which have been tried on an experimental scale by Raitt, 2 Richmond 3 and others. Many of these have been found to produce excellent fibre. Peat also might possibly be considered in this class. Numer- ous attempts have been made to produce useful fibre from peat, the treatments given ranging from an entirely mechanical pro- cess, through treatment with alkalis and acids, to fermentation processes. All methods have thus far failed to produce a fibre which can be used in anything but the coarsest products. Woods The woody tissues of plants are made up of cells which exhibit great diversity of form, size, and markings as the accompanying drawings Figs, i and 2 show. Those in which the paper maker is particularly interested are the true wood fibres, or libriform cells, and the tracheids, but many of the other cells are of inci- dental interest as helping to identify the wood from which they were derived. The wood fibres are always spindle or fibre-form and the walls are relatively strongly thickened. They never have true spiral striations; in only a few species do they show pits, which are generally elongated and oblique. Wood fibres are variable in length in different woods, ranging from 2.0 mm. to 0.14 mm., but in all cases they are the longest elements present. As an example of wood fibres may be cited the chemical pulp made from poplar wood; this contains in addition only the ducts and rarely a few of the small cells from the medullary rays. Tracheids are elongated and tapering cells, more or less ligni- fied and having peculiar markings known as bordered pits or discoid markings. Fig. 3 shows these as they appear on the surface of the fibre and illustrates how they are formed by the thin partition wall between two tracheids. These pits are so constant in number and mode of distribution that they may be 1 Brand and Merrill: Bull. No. 309, U. S. Dept. of Agriculture. 2 Raitt: Indian Forest Records, Vol. V, Part III. 3 Richmond: Philippine J. Sci., 1906, 1, 433-462. 4 6 FIBROUS RAW MATERIALS Fig. i Drawings of Wood-elements 1-7. Avicennia sp. 1. Wood-parenchyma cells; tangential section. 2. Sep- tum of a duct. 3, 4, 5. Conjugate wood-parenchyma cells. 6, 7. Portions of spirally striated libriform fibres. 8-1 1. Bast cells of Cytisus laburnum. 8. Cross-section through young bast bundle acted on by chloroiodide of zinc. 9, 10, n. Cross-sections through young bast cells similarly treated. 12. Porlieria hygrometrica; radial section of conjugate substitute fibres. 13. Jatropha mani- hot, radial section through wood. 14-18. Tectona grandis; elements separated by maceration. 14. Conjugate wood-parenchyma cells. 15. Ordinary wood- parenchyma fibre. 16. Substitute fibre. 17. Libriform fibre. 18. Septate libri- form fibre. WOODS 47 Fig. 2. Drawings of Wood-elements 19. TracheidfromTectonagrandis. 20-23. Porlieria hygrometrica. 20. Con- jugate substitute fibres in cross section. 21. Ordinary substitute fibre after maceration. 22, 23. Conjugate substitute fibres after maceration. 24-27. Cy- tisus laburnum; elements after maceration. 24. Wood-parenchyma fibre. 25. Tracheid. 26. Substitute fibre. 27. Simple libriform fibre. 28. Cross-section through cambium and youngest wood of Cytisus laburnum. 29, 30. Mahonia aquifolium; ducts. 29. After maceration. 30. Longitudinal section. 31-36. Extremeties of ducts separated by maceration from Hieracium. 37-39. Ducts from Onopordon acanthium. 40. Spirally marked duct from Vitis vinif era. 41. Libriform fibre from Jatropha manihot. 48 FIBROUS RAW MATERIALS Fig. 3. Discoid Markings on Wood Cells Pinus laricio; a. Radial walls"; b. A transverse section. Pinus sylvestris; c. Development of markings; d, e. Transverse sections of nearly perfect and perfect discoid markings. used as a distinguishing characteristic for some woods. In cone-bearing or coniferous trees, such as spruce, fir, hemlock, etc., the wood consists almost entirely of tracheids, and when sulphite fibre or ground wood from such trees is examined under the microscope the discoid markings may be very readily seen. These tracheids are generally much longer than the libriform FIBRE LENGTH 49 fibres from other woods and hence possess greater paper making value. Sap and Heartwood. The sapwood, or that of compara- tively recent growth, is usually lighter in color and contains more fermentable material than the older and denser heart- wood. Each year a layer of sapwood goes over into heartwood which becomes darker and harder from infiltration of coloring matters, resins, etc. The sapwood is generally preferred for pulp because it is more easily reduced by either the mechanical or chemical processes. In some woods, as fir and buckeye, the difference in color and hardness between heartwood and sap- wood is not in evidence. Fibre Length. The length of the fibres or tracheids in a given tree is known to vary with the position from which the sample was taken. This has been very carefully investigated for conif- erous woods by Mell, 1 who finds that the length of tracheids varies considerably not only in different parts of the same tree, but also within the same annual ring at the same distance above the ground. In both trunk and branches the average length increases from the center outward until the tree reaches its maximum height growth, after which it remains quite constant. The highest average length of tracheids in the branches is usu- ally less than in the trunk. Tracheids also vary in length according to the character of the soil and the amount of mois- ture, those from trees growing in rich, moist soil being longer than those from trees grown in dry soil. Lee and Smith 2 have also made very careful studies along similar lines of one or two trees of Douglas spruce. Their observations are in general confirmatory of those of Mell, though in some respects their conclusions are different. Because of these variable factors it is very difficult to determine the average length of tracheids or of the libriform cells of broad- leaved woods. The following table by Mell gives the length of tracheids of many of the coniferous woods of the United States; 1 Mell, Paper Trade J., June 15, 191 1. 2 Forestry Quarterly, Dec, 1916. 5° FIBROUS RAW MATERIAL the figures are the averages of many measurements of samples taken from different parts of the trunks and branches. Names of trees Amabilis fir (Abies amabilis) Balsam fir (Abies balsamea) White fir (Abies concolor) Lowland fir (Abies grandis) Alpine fir (Abies lasiocarpa) Noble fir (Abies nobilis) Port Orford cedar (Chamaecyparis lawsoniana) Yellow cedar (Chamaecyparis nootkatensis) . . . . White cedar (Chamaecyparis thyoides) Western larch (Larix occidentalis) Incense cedar (Libocedrus decurrens) White spruce (Picea canadensis) Engelmann spruce (Picea engelmanni) Black spruce (Picea mariana) Red spruce (Picea rubens) Sitka spruce (Picea sitchensis) Knobcone pine (Pinus attenuata) Sand pine (Pinus clausa) Jack pine (Pinus divaricata) Shortleaf pine (Pinus echinata) Pifion (Pinus edulis) Limber pine (Pinus flexilis) Spruce pine (Pinus glabra) • Sugar pine (Pinus lambertiana) Silver pine (Pinus monticola) Lodgepole pine (Pinus murrayana) Longleaf pine (Pinus palustris) Western yellow pine (Pinus ponder osa) f Western yellow pine (Pinus ponder osa) Parry pifion (Pinus quadrifolia) Red pine (Pinus resinosa) Pitch pine (Pinus rigida) Pond pine (Pinus serotina) White pine (Pinus strobus) Loblolly pine (Pinus taeda) Scrub pine (Pinus virginiana) Red fir (Pseudotsuga taxifolia) Redwood (Sequoia sempervirens) Bigtree (Sequoia washingtoniana) Bald cypress (Taxodium distichum) Arborvitae (Thuja occidentalis) Western red cedar (Thuja plicata) Hemlock (Tsuga canadensis) Western hemlock (Tsuga heterophylla) Length of fibres Mm.* Mm. Mm. aver. Max. Min. 3-ii 5.62 1,49 3.10 4 21 1.98 4-63 6 03 2-73 4.14 5 70 2.89 3.16 4 13 2 .07 4.02 4 96 2.b 4 3- 6 3 4 38 2.56 2.47 3 02 i-93 2 .10 2 81 i-43 2-59 3 80 i-73 4.01 4 7i 2.97 3-53 4 21 2.31 5-7i 6 94 3.06 3-48 4 13 2.89 2-97 3 63 2.48 2.87 3 72 2.31 2 .06 2 40 1 .24 2.85 4 38 1.32 5.06 6 53 3-72 5-«9 7 19 4.38 1 .96 2 56 1.49 2.94 3 9b 1 .90 4i3 5 45 2-73 4-47 S 8b 2-73 4-39 5 45 2-73 2.63 3 72 1.82 5-53 6 69 2.97 3-3 2 3 9 b 2.48 2-35 3 3° 1 .40 2 .10 2 89 1.49 4 03 4 79 3.22 3-86 4 7i 2.48 4.20 6 36 2.64 3-53 4 54 3.22 3.10 3 88 2.56 2-73 3 9 b i-73 2.68 3 3° 1.82 6.99 9 25 4-05 4.82 5 95 3-47 4.69 S 78 3-£>3 2 .02 2 39 i-39 3-87 4 54 3-i4 4.01 5 04 2.81 3 -04 3 <>3 i-73 * A millimetre is equal to about one-twenty-fifth of an inch, t From a tree growing in dry soil. MOISTURE IN WOOD 51 Among broad-leaved woods the following have been exam- ined by the author. The measurements are for the true wood fibres and the samples were prepared from wood of commercial size. All measurements are in milhmetres. Wood Length Width Max. Min. Avg. Max. Min. Avg. Beech (Fagus atropunicea) Poplar (Populus grandidentata) .... Aspen {Populus tremuloides) Cotton gum (Nyssa aquatica) Red alder {Alnus oregona) Sycamore {Platanus occidentalis) . . Red maple {Acer rubrum) Buckeye (Aesculus flava) Cucumber tree {Magnolia acuminata) Umbrella {Magnolia fraseri) Tulip tree {Liriodendron tulipifera) Sweet gum {Liquidamber styraciflua) Black gum {Nyssa sylvatica) Elm ( Ulmus americana) • Birch {Betula papyrifera) 1 .72 1 .62 1.68 2.67 1.77 2 .21 1. 19 0.92 1.30 i-47 1 59 2 .02 2.32 r.98 1.63 O.70 O.71 O.78 I .24 O.84 0.94 O.67 O.46 0.55 O.62 O.64 O.96 1. 18 I.03 O.78 1. 13 1.08 i-i5 1.85 1.23 i-57 o.93 0.62 0.86 1.08 1. 14 i-55 1.68 i-35 1. 17 0.029 0.044 0.046 O.IOO 0.038 0.033 0.028 0.026 0.036 0.041 0.035 0.036 0-035 0.021 0.042 0.015 0.020 0.020 0.028 0.014 0.016 0.014 0.014 O.020 0.013 0.021 0.022 0.015 0.014 0.014 0.022 0.028 0.032 0.066 0.027 0.024 O.020 0.020 0.029 0.027 0.029 0.031 0.026 0.019 0.025 Moisture in Wood. The cell cavities of all wood contain large amounts of air and moisture. According to Sachs the volume percentage of freshly cut fir wood is Per cent Cell walls 24. 81 Water 58. 63 Air cavities 16. 56 The moisture in wood varies with the amount of seasoning it has had and also with the kind of wood, the position in the tree, and the time of cutting. Certain kinds of wood in the living tree, as for example, white ash, black locust and the white and red cedars are comparatively dry; black ash and the oaks have about twice as much, and chestnut and buckeye about three times as much moisture as white ash; cypress and white fir also contain large amounts of water. In the hard woods the variation in moisture with the different positions in 52 FIBROUS RAW MATERIAL the tree is comparatively slight while the conifers show wide variations, the heartwood generally being very dry and the sapwood very' wet. Seasoning greatly reduces the amount of moisture present in wood but the rate of drying is not the same for all varieties, some losing moisture in one-tenth the time required by others. The term "air dry," therefore, is one which may denote almost any condition of moisture from 40 per cent down to 4 per cent of the total weight, according to the length of seasoning and the conditions to which exposed. Weight per Cubic Foot. The figure for weight per cubic foot is one which is closely connected with the moisture content, since the shrinkage in volume due to loss of moisture is not at all proportional to the amount of such loss. For this reason it is best to base the weight per cubic foot on the absolutely dry material. Determinations made by the Forest Service 1 for a number of American woods gave the following results : Kind of wood Balsam fir Red fir White fir Alpine fir Lowland fir Eastern hemlock Western hemlock Tamarack Lodgepole pine, Montana. . Lodgepole pine, California Pounds per cubic foot (bone dry) 21.50 22.25 21 .40 22.00 21-53 24.60 24.80 32 .00 25-15 23 . 20 Kind of wood Jack pine Loblolly pine, fall cut White pine Engelmann spruce, Montana Engelmann spruce, Colorado Sitka spruce White spruce White birch Poplar Black gum Pounds per cubic foot (bone dry) 24.00 28.86 20.25 24.40 21 .28 23 .60 26.4O 34.20 24.16 30-36 The following table of weights per cubic foot is also from data supplied by the Forest Service. In this case the weights are for kiln dried material and the bone dry weights would probably be about 4 per cent less. This table shows the varia- tions which may be expected in wood of the same species when grown in different localities and under different conditions. It 1 The Log of the Lab., Dec, 1916. RESINS 53 illustrates the impossibility of establishing a figure for any wood which will apply in all cases. Kind of wood Locality Pounds per cubic foot (kiln dry) Cypress, bald Louisiana 33 30-35 38-42 32 28 Douglas fir California, Oregon, Washington. Florida, Mississippi Longleaf pine Wisconsin Spruce, red Spruce, white Red alder New Hampshire New Hampshire, Wisconsin Washington 25-29 27 26 Aspen Basswood Wisconsin Wisconsin, Pennsylvania Pennsylvania, Indiana 24-26 41-43 37 24 29 35 34-37 32 41-42 34-35 27 35 Beech Wisconsin Buckeye Chestnut Tennessee Maryland Black gum Tennessee Pennsylvania, Wisconsin Wisconsin Red maple Silver maple Sugar maple Indiana, Pennsylvania Sycamore Indiana, Tennessee Tulip tree Tennessee Tupelo Louisiana Resins. Many woods contain small amounts of volatile oils generally approaching terpene (CioHi 6 ) in chemical composition. Practically nothing is known regarding their formation. From these by oxidation are formed balsams and resins, the former being regarded as mixtures of resins with volatile oils. There are also in some cases resins which contain gum or mucilage and are hence termed gum resins. The amount of resin in wood varies greatly with the different kinds and with the solvent used in its determination. With Canadian woods Richter x obtained the following results : Fresh balsam. Fresh spruce . Ether extract Per cent 0.45-0.85 0.70-1 .80 Alcohol extract Per cent I-I5-3-65 o . 70-1 . 94 1 Richter: Wochbl. Papierfabr., 44 (19 13), 4507. 54 FIBROUS RAW MATERIAL Examination of the resins obtained gave the following con- stants : Ether resin Alcohol resin Acid number Saponification number Acid number Saponification number White spruce 61 66 80 no 54 35 83 168 Black spruce Richter claims that storage of the wood decreases the ether extract and increases the alcohol extract, while Schwalbe and Grimm 1 state that seasoning or passing air over the chips reduces both ether and alcohol soluble material. The ether extract is usually lighter in color and more liquid and sticky than the alcohol extract. Part of each is soluble in petroleum ether and according to Johnsen 2 it is this portion which is responsible in large part for the trouble with pitch in sulphite pulp. The portion soluble in petroleum ether is a thick yellow liquid and appears to be of a fatty rather than a resinous nature; the insoluble part is brown and brittle. The resin in Bohemian pine and its distribution in the sulphite cellulose made from it have been studied by Sieber, 3 who found that cooking removed 4.2 per cent; the knotter, screens, sand traps, etc., took out 51.8 per cent, while bleaching removed 15 per cent. Proximate Analysis of Wood. The following table gives the analysis of a number of European woods. 4 1 Schwalbe and Grimm: Wochbl. Papierfabr., 44.(1913), 3247. 2 Johnsen: Pulp and Paper Mag. Can., 1917, 577. 3 Papierfabr., 19 15, June 18. 4 Miiller: Die Pflanzenfaser. BARK AND KNOTS 55 Wood Black poplar Silver fir. . . Birch Willow Scotch pine. Chestnut . . . Linden Beech Water Soluble in water 12 .IO 2.88 13-87 1 .26 12.48 2.65 11.66 2.65 12.87 4-05 12.03 S-4i 10.10 3-56 12.57 2 .41 Soluble in alcohol and benzine i-37 o.97 1. 14 1.23 1.63 1 .10 3-93 0.41 Cellulose 62 .77 56 99 55 52 55 72 53 27 52 64 53 09 45 47 Incrusting matter 26.91 28.21 28.74 28.18 28.82 29.32 39-14 Still more complete analyses of seven American woods are given by Schorger, 1 all percentages being based on oven dry samples. Some of his results follow: Wood Longleaf pine (Pinus palustris). Douglas fir (Pseudotsnga taxi- folia) Western larch (Larix occiden- talis) White spruce {Picea canadensis) Basswood (Tilia americana) Yellow birch (Betula hitea) . . . . Sugar maple (Acer saccharum). . °-37 0.38 0.23 0.31 0.86 0.52 0.44 Solubility of wood M S 7-15 6.50 12-59 2 .14 4.07 3-97 •4-36 a 6.32 1 .02 °o &3 22 .36 16. 11 22 .14 "•57 23.76 19-85 17.64 < 0.76 1 .04 0.71 i-59 5-79 4-3o 4.46 5-o5 4-95 7.46 6.02 10.80 10.39 19-93 24.63 21 .71 58.48 61.47 57-8o 61.85 61 .24 61.31 60.78 The percentages of cellulose given in this table by Schorger are considerably higher than the figures of most other analysts for the same, or similar, woods. They are doubtless more accu- rate because of the greater pains taken with the work and because of the more complete knowledge of the precautions necessary to prevent hydrolysis of the cellulose and its conse- quent loss during the analytical procedure. Bark and Knots. The bark serves as a protective envelope for the stem and gradually increases in thickness, as a layer is added each year. It contains long bast fibres which give strength 1 Schorger: J. Ind. and Eng. Chem., 1917, 9, 556. 56 FIBROUS RAW MATERIAL to the bark, and cork cells which, because of their impermea- bility, are admirably suited to form a protective covering for the tissues beneath. Bark often contains coloring matters and tannins, sometimes in sufficient amounts to make extraction profitable. It is only slightly acted upon by the chemical pro- cesses of pulp manufacture and for this reason is of interest to the paper maker chiefly because of the necessity for its removal. Recent experiments have led to its use on a small scale as a substitute for better grades of material in making roofing felts and similar products and it will probably eventually be quite fully utilized. The loss in barking varies greatly with the size and shape of the logs, with the care used by the men in charge of the bark- ers, and with the type of barker used. With disc barkers it may amount to 10 to 25 per cent of the rough wood. Knots are formed at the points where the branches make out from the stem or trunk. They are usually very hard and dense and are frequently highly charged with resins and coloring matters. They are partially reduced by the soda process but are almost unaffected by the sulphite process which usually fails even to soften them. Decay. The importance of decay is becoming greater every year because of the increasing cost of wood. Not only does the wood stored in the yard decay but ground wood stored in laps also suffers damage which is estimated by the Forest Service to amount to between five and fifteen million dollars a year. The decay of wood stored in piles depends on the size and form of the pile, upon temperature and humidity and upon the foun- dation upon which the pile is built. Small piles of evenly stacked wood will nOt decay because they are well ventilated and dry out readily. Neither will the wood in the interior of large piles because it is too wet; but between these two ex- tremes is a condition where the moisture is just right and the fungi flourish wonderfully. The summer, with its high tem- peratures, is the time when most of the decay takes place and practically no loss is suffered in winter. The following photomicrographs show the characteristic forms and markings of a number of the typical paper-making fibres. These photomicrographs, as well as those in Chapter X, were prepared by the Paper Section of the Bureau of Standards. :■ I Plate i Cotton (Gossypium) Magnification ioo diameters. Photographed by Bureau of Standards. Plate 2 Linen (Linum usitatissimum) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 3 Hemp (Cannabis saliva) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 4 Manila (Musa textilis) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 5 Jute (Corchorus capsularis) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 6 Paper Mulberry (Broussonetia papyrifera) Magnification ioo diameters. Photographed by Bureau of Standards, Plate 7 Sisal (Agave rigida) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 8 Rice Straw (Oryza sativa) Magnification ioo diameters. Photographed by Bureau of Standards. Plate 9 Esparto {Stipe tenacissima) Magnification xoo diameters. Photographed Bureau of Standards. by Plate io Bamboo {Bambusa arnndinacea) Magnification ioo diameters. Photographed by Bureau of Standards. Plate ii Corn (Zea mays) Magnification ioo diameters. Photographed by Bureau of Standards. Plate 12 Red Spruce (Picea rubens) Magnification 100 diameters. Photographed by- Bureau of Standards. Plate 13 Spruce Ground Wood (Picea canadensis) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 14 Balsam Fir (Abies balsamea) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 15 Jack Pine (Pinus divaricata) Magnification 100 diameters. Photographed by- Bureau of Standards. Plate 16 Hemlock (Tsuga canadensis) Magnification ioo diameters. Photographed by- Bureau of Standards. Plate 17 Douglas Spruce (Pseudotsnga taxifolia) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 18 Aspen (Populus tremnloides) Magnification ioo diameters. Photographed by- Bureau of Standards. Plate 19 Yellow Birch {Betula luted) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 20 Beech (Fagus atropunicea) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 21 Chestnut (Castanea dentata) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 22 Tulip-tree (Liriodendron tiflipifera) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 23 Sweet Gum (Liquidambar styraciflna) Magnification 100 diameters. Photographed by Bureau of Standards. V \ Plate 24 Hard or Sugar Maple (Acer saccharum) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 25 Soft or Silver Maple (Acer saccharinum) Magnification 100 diameters. Photographed by Bureau of Standards. Plate 2b Black Gum (Nyssa sylvatica) Magnification ioo diameters. Photographed by Bureau of Standards. KINDS OF WOOD 57 The actual loss of the wood itself is not the only serious feature, for the handling of such materials is expensive because of the labor required to cut out the decayed portions. As it is never possible to make a complete separation some poor wood gets into the digester, cutting down its capacity, injuring the quality of the pulp, and using up the cooking liquors uselessly. Another serious feature is the fire risk involved by the tinder- like decayed portions. It takes but a spark to ignite such material and it is said to be responsible for most of the serious pulp wood fires. The decay in wood yards can be greatly lessened by making small piles which will dry rapidly and by piling the wood on a foundation of crushed stone or gravel rather than on sod-land. Spraying the foundation with a disinfectant is also recommended by Haas. 1 Another point is the careful removal of old bark, decayed wood and the like; such materials should never be left to be covered by new wood. Kinds of Wood. The woods most generally employed in the United States for the manufacture of both ground wood and chemical fibre are spruce and poplar, but the growing scarcity of these two species has led to the use of many other kinds. The quantities of such other woods depend on the factors of price, length of fibre, ease of reduction and the relative loca- tions of the mill and the source of the wood supply. The questions of length of fibre and weight per cubic foot have already been treated and we shall now consider briefly some of the woods which are in actual use or which have been tried out sufficiently on an experimental scale to demonstrate their value. The names and ranges of the various species have been taken from Sudworth's " Check List of the Forest Trees of the United States," while the other data have been collected from widely scattered sources. The species mentioned comprise only a por- tion of those which will probably be found suitable for the manufacture of paper or pulp in some of its forms. 1 Haas: J. Soc. Chem. Ind., 1910, 29, 415. 58 FIBROUS RAW MATERIAL Spruces. Red spruce, Picea rubens, ranging from Nova Scotia to North Carolina and Tennessee, and white spruce, Picea canadensis, are the common pulp woods of the East. The range of the latter is from Newfoundland to Hudson Bay, and north- westward to Alaska; southward to Northern New York, Michi- gan, Wisconsin, Minnesota, South Dakota, Montana and British Columbia. Both woods are light, soft, straight grained and fairly free from resin. Very similar in quality are Engelmann spruce, Picea engelmanni, ranging from northern Arizona through the Rocky Mountain region to British Columbia, and Sitka spruce, Picea sitchensis, extending on the coast region from Alaska to northern California. This latter species is said to be the best pulp wood on the Pacific coast. All the spruces are reduced easily by the sulphite or sulphate process but with considerably more difficulty by the soda pro- cess. The soda fibre is very hard to bleach. Because of the uniformly light color of the wood all are admirably adapted for the preparation of ground wood, Sitka spruce, however, being somewhat inferior to the others in color. Firs. Balsam fir, Abies balsamea, occurs from Newfoundland and Labrador westward to the region of Great Bear Lake and southward to Pennsylvania. Its wood is very light, soft, not strong, rather coarse grained, and not durable. It is frequently cut and used with spruce, sometimes to the extent of 20 per cent or more, in the manu- facture of sulphite. The fibre is considered somewhat inferior to that from spruce and is likely to contain more pitch. Bal- sam fir can be successfully used in the preparation of ground wood of good color but its use is not favored because of the low yield per cord and its tendency to decay. Other firs which are found largely on the Pacific coast and in the Rocky Mountain region are lowland fir, Abies grandis; white fir, Abies concolor; amabilis fir, Abies amabilis; noble fir, Abies nobilis, and red fir, Abies magnifica. The wood of most of these is light, soft and straight grained; it varies in the different species from fine to moderately coarse PINES 59 grained. All of these trees grow to comparatively large size and yield excellent pulp wood. With the proper treatment all can be satisfactorily reduced by the sulphite process with the production of long fibre similar to spruce in quality. White fir is already being so treated commercially and yields a long silky fibre of the highest quality. All of these firs can be used for preparing ground wood for news print work though the color of the product is in most cases rather inferior to that of spruce ground wood. Pines. Longleaf pine, Pinus palustris, ranges from southern Virginia to Florida and eastern Texas and northward to north- eastern Alabama and northwestern Georgia. Covering a con- siderable portion of the same territory is shortleaf pine, Pinus echinata. The wood of both is hard, strong, dense, durable and resinous and is much used for many grades of lumber. It cannot be cooked successfully by the sulphite process because of its resinous nature, but if treated by the soda or sulphate process it yields a very strong, long-fibred stock suitable for wrapping or kraft papers but not for bleaching. Jack pine, Pinus divaricata, is found from New Brunswick to New Hampshire, westward through the Great Lake region to the Rocky Mountains and south into northern Maine, New York, Indiana, Illinois and central Michigan. It is not suitable for use in the sulphite process but is largely used for the pro- duction of kraft pulp by the sulphate process. This wood also yields a fair grade of ground wood. Other pines which are either used commercially or which have been proved by semi-commercial tests to be suitable for use in the sulphate or ground wood processes are: white pine, Pinus strobus; western yellow pine, Pinus ponder osa; lodgepole pine, Pinus murrayana; loblolly pine, Pinus taeda; scrub pine, Pinus virginiana and red or Norway pine, Pinus resinosa. These pines may be reduced by the soda process but the treat- ment required is considerably more severe than that given poplar and the fibre is not easily bleached. 60 FIBROUS RAW MATERIAL Hemlock {Tsuga canadenis). Range. — Nova Scotia to Min- nesota, Wisconsin, Michigan and southward in the mountains of the Atlantic region to northern Alabama and Georgia. The wood is light, soft, brittle, not strong, crooked grained, liable to wind-shake and splinter and is not durable; its color is light brown tinged with red, the sapwood usually somewhat darker. This wood is used to a considerable extent in the sulphite process and yields a fibre very similar to spruce though the treatment has to be rather more severe than for spruce. It also yields a good grade of kraft fibre. In the ground wood process a fair grade of fibre can be produced, though it is considerably more difficult to handle than spruce. A related species, western hemlock, Tsuga heterophylla, has fine-grained, pale yellowish brown, light, soft wood which works like white pine. It yields a good grade of sulphite fibre and for ground wood is far superior to eastern hemlock. Larch or Tamarack, (Larix laricina). Range. — From New- foundland and Labrador to northern Pennsylvania, Indiana, Illinois, central Minnesota and northwestward to Hudson Bay. Wood very heavy, hard, strong, rather coarse grained, dur- able in contact with the soil; color light brown, the sapwood nearly white. Larch is reduced by the sulphite process with some difficulty and if mixed with spruce or hemlock is likely to cause chips and shives. By the sulphate process it can be made into a good grade of kraft fibre; it also yields a good grade of ground wood except for color which is a decided grayish green. Douglas Spruce (Pseudotsuga taxijolia) . Range. — From the Rocky Mountain region and northward to central British Columbia. The wood varies widely in character and grain which may be very coarse, medium or fine. The coarse-grained wood is usu- ally reddish brown, while the fine-grained is clear yellowish brown. The wood is slightly resinous and resembles pine in many of its characteristics. WHITE BIRCH 6l Douglas spruce has been found well suited for making kraft pulp and is already used commercially in the sulphite process, though it is not so easy to cook and bleach as spruce. Poplar {Populus grandidentata) . Ranges from Nova Scotia through New Brunswick, southern Quebec and Ontario to north- ern Minnesota; southward to Delaware, southern Indiana and Illinois. The wood is light, soft, not strong, close grained, compact, decays rapidly; color light brown, the sapwood nearly white. This is the wood most commonly used in the soda process; it is almost never used in sulphite mills though it is readily reduced by that process. It yields a ground wood of fair color but of rather short fibre. The three following species are very similar to poplar in character of wood and are of practically identical paper making value. Aspen {Populus tremuloides) . Range. — Southern Labrador to Hudson Bay and northwestward to the Mackenzie River and Alaska; southward to Pennsylvania, northeastern Mis- souri, southern Nebraska, and throughout the western moun- tains to northern New Mexico and Arizona and central California. Balm of Gilead {Populus balsamifera) . Range. — Coast of Alaska and valley of Mackenzie River to Hudson Bay and Newfoundland; southward to northern New England and New York, central Michigan and Minnesota, northwestern Nebraska, northern Montana, Idaho, Oregon and Nevada. Cottonwood {Populus deltoides). Range. — From Quebec and Vermont through western New England and New York, Penn- sylvania, Maryland, and Atlantic States to western Florida and west to the Rocky Mountains. White Birch — Gray Birch {Betula populifolia) . Range. — From Nova Scotia, New Brunswick, and Lower St. Lawrence River southward to Delaware and westward through northern New England and New York to Lake Ontario. The wood is light, soft, not strong, close grained, not durable; color light brown with thick, nearly white sapwood. 62 FIBROUS RAW MATERIAL This wood is easily reduced to pulp by the soda process and the fibre bleaches readily. The chief difficulty in its use is in the economical removal of the bark. Paper Birch (Betula papyrifera) . Range. — From Labrador to Hudson Bay, Great Bear Lake, Yukon River and coast of Alaska; southward to New York, northern Pennsylvania, cen- tral Michigan and Minnesota, northern Nebraska, Dakota, northern Montana and northwestern Washington. The wood is light, strong, hard, tough and close grained; its color is light brown tinged with red and the sapwood is nearly white. The bark is removed from this wood with some diffi- culty and as even the inner bark causes dirt in the pulp it must be very completely removed. Paper birch cooks by the soda process with a little more difficulty than poplar and the fibre requires slightly more bleach. It yields pulp similar to poplar and fully equal to it in quality. Red Alder {Alnus oregona). Range. — From Sitka through the coast ranges of British Columbia, Washington and Oregon to California. The wood is light, brittle, fine grained; color pale reddish- brown. By the soda process this wood cooks readily, yielding a fibre very similar to poplar. Beech (Fagus atropunicea). Range. — Nova Scotia to Lake Huron and Northern Wisconsin; south to western Florida and west to southeastern Missouri and Texas. The wood is very hard, strong, tough, very close grained, not durable in contact with soil, inclined to check on drying; color dark or light red with nearly white sapwood. Beech cooked by the soda process requires about the same, or possibly a little more severe treatment than poplar. The pulp is soft and easily bleached, though not quite so easily as poplar. Chestnut (Castanea dentata) . Range. — From southern Maine to nortwestern Vermont, southern Ontario and the southern shores of Lake Ontario to southeastern Michigan; southward SWEET GUM 63 to Delaware and southeastern Indiana and on the Allegheny Mountains to central Kentucky and Tennessee, central Ala- bama and Mississippi. The wood is light, soft, not strong, coarse grained, liable to check and warp in drying, easily split, very durable in contact with soil. It is reddish brown in color. Chestnut wood contains tannin which can be profitably ex- tracted for use in tanning or other industries. The extracted chips can be reduced quite readily by the soda process and the fibre bleaches without much difficulty. If the tannin is not removed the fibre is hard to bleach. Chestnut fibre is short and is used as a substitute for poplar. Cucumber Tree (Magnolia acuminata). Range. — From west- ern New York through southern Ontario to Illinois and south in the Appalachian Mountains to southern Alabama and north- eastern Mississippi; central Kentucky and Tennessee. The wood is soft, light, not strong, close grained and easily worked; color light yellow with nearly white sapwood. This wood reduces easily by the soda process giving a fibre closely resembling poplar in its paper making value. Tulip Tree (Liriodendron tulipifera) . Range. — From Rhode Island to southwestern Vermont and west to Lake Michigan; south to Florida, southern Alabama and Mississippi. The wood is light, soft, brittle, not strong, easily worked; its color is light yellow or brown with creamy white sapwood. Tuliptree is readily reduced by the soda process, yielding a fibre which is similar in character to poplar though generally a trifle longer. Sweet Gum (Liquidamber styraciflua) . Range. — From Con- necticut to southeastern Missouri and Arkansas; south to Florida and Texas. The wood is heavy, hard, not strong, straight and close grained, inclined to shrink and warp badly in seasoning; color bright brown tinged with red, sapwood nearly white. Sweet gum can be treated by the chemical processes about as easily as poplar: the fibre is considerably longer than poplar 64 FIBROUS RAW MATERIAL but not long enough to bring it into the class with spruce and other coniferous woods. Sycamore {Platanus occidentalis) . Range. — From south- eastern New Hampshire and southern Maine to northern Ver- mont and Lake Ontario; west to eastern Nebraska and Kansas, and south to northern Florida, central Alabama, Mississippi, and Texas. The wood is rather light, hard, coarse grained, not very strong, very hard to split. Sycamore cooks easily by the soda process and the fibre is longer and more slender than that from poplar; it is said, how- ever, to give a rather "punky" paper. Sugar Maple, Hard Maple {Acer saccharum). Range. — From southern Newfoundland to Lake of the Woods and Min- nesota; south to northern Georgia and western Florida; west to eastern Nebraska, Kansas and Texas. The wood is heavy, hard, strong, tough and close grained; in color it is light brown tinged with red. By the soda process it is reduced by about the same treat- ment given poplar; the fibre is shorter than poplar but bleaches readily. Silver Maple, White Maple (Acer saccharinum) . Range. — From New Brunswick to western Florida; west to southern Ontario, through Michigan to eastern Dakota, Nebraska and Kansas. The wood is moderately light, hard, strong, close grained, easily worked but rather brittle. It is reduced by the soda process as readily as poplar and makes a paper of practically the same quality. The fibre is a little shorter than poplar and bleaches readily. Red Maple (Acer rubrum). Range. — From New Brunswick, Quebec and Ontario to Florida; west to Lake of the Woods, eastern Dakota, Nebraska and Texas. The wood is very heavy, close grained, not strong; its color is light brown slightly tinged with red, the thick sapwood is lighter colored. BULK OF RAW MATERIALS 65 Red maple is slightly more difficult to treat than poplar, the fibre is rather shorter than poplar and bleaches a little harder. Basswood (Tilia americana). Range. — From New Bruns- wick to Virginia, Georgia and Alabama; west to Lake Superior, Lake Winnipeg, eastern Dakota, Nebraska, Kansas and Texas. The wood is light, soft, not strong, very close grained, com- pact and easily worked; color, light brown tinged with red. It is very easily reduced by the soda process and yields an easy bleaching pulp very similar to poplar. Black Gum ( Nyssa sylvatica) . Range. — From Maine to Florida and west to southern Ontario, southern Michigan, south- eastern Missouri and Texas. The wood is heavy, soft, strong, fine grained, very difficult to split; in color it is light yellow or nearly white. By the soda process it cooks nearly as easily as poplar and yields a fibre free from shives. It bleaches a little harder than poplar. Its fibre is longer than that from poplar and makes an excellent paper. A very white ground wood can be made from it but it has not sufficient strength for newspaper work. Bulk of Raw Materials. This is an important factor to be considered in the trans- portation, storage and cooking of the different materials yet very few figures have apparently been published. The follow- ing notes therefore make no claim to completeness but are merely an attempt to collect in one place what little informa- tion is available. Rags. Bales of rags as received at the mill have been found to have the following weights : Lbs. per cu. ft. Egyptian rags 37. 5 Blue cottons 26. 5 White cottons 19. 5-22. 1 When dusted and dumped into bins, but not tamped, they weigh 15.6 lbs. per cu. ft. before cutting. 66 FIBROUS RAW MATERIAL Straw. Weighings of rice straw gave the following results: Lbs. per cu. ft. As baled for shipment 1 1 Chopped and tamped 4.7 Chopped and not tamped 3.3 Esparto. Beadle and Stevens 1 give the figures for esparto as 120 cu. ft. per ton when pressed as usual or 90 cu. ft. from hydraulic presses. A boiler of 540 cu. ft. capacity (vomiting type) will hold 50 cwt. of esparto, which after cooking will occupy a volume of 300 cu. ft. Wood. The number of cubic feet of solid wood per cord and the consequent weight per cord vary greatly with the different kinds of wood and the size of the logs. Graves 2 gives the following table for the number of solid cubic feet for sticks of various diameters. Diameter of No. of sticks per cord Solid cubic feet per cord of sticks Hardwoods Softwoods Mixed Ins. 6.8 6.0 4-75 3-5 94 126 205 378 102 .40 94.72 88.32 79-36 102 .40 98.56 97-28 90.88 IO2.40 96.OO 92.16 84.48 Measurements by the author on carefully stacked poplar wood in four-foot lengths also illustrate the same point: Average diam- eter of sticks No. of sticks per cord Weight per cord, air dry Per cent moisture Weight per cord, bone dry Ins. 11 .72 7-9° 3-i8 35-5 70.5 Lbs. 4295 3610 2625 34-2 29-3 19.7 Lbs. 2828 2553 2108 Sound poplar wood when chipped, blown into a bin and leveled off but not tamped occupied a space of 259 cu. ft. per 1 Beadle and Stevens: Chem. News, 1914, 109, 302-304. 2 Graves: Forest Mensuration, p. 105. BULK OF RAW MATERIALS 67 cord of wood for an average of three tests. The chips, when placed loosely into a measuring box, weighed 10.8 lbs. per cu. ft. and if thoroughly shaken down but not tamped 13.8 lbs. per cu. ft. Spruce wood gave, in the case of one test, the following figures: One cord air dry spruce with bark weighs 4500 lbs. Same after disc barking weighs 3600 lbs. Volume of chips from above is 260 cu. ft. One cubic foot of chips weighs 13 lbs. CHAPTER III RAGS, ESPARTO, STRAW, BAMBOO Treatment of Rags. It is generally conceded that about 70 per cent of the rags used in this country are imported, while the remaining 30 per cent are collected chiefly in the larger American cities where the conditions of sorting are not con- ducive to cheapness or good results because of the high rentals for buildings and the unskilled labor which must be employed. In Europe each family keeps a separate bag for linen, woolen and cotton rags. These are collected at frequent intervals and are sorted and packed in large sheds which are rented cheaply, thus securing less costly and more satisfactory grading. The grades of rags differ in different countries and vary from time to time. Among those listed in current trade jour- nals are the following, which may be taken as generally char- acteristic. Domestic Foreign No. 1 New white shirt cuttings New white cuttings No. 2 New white shirt cuttings Unbleached cottons Fancy new shirt cuttings Light flannelettes New blue cottons New mixed cuttings New mixed cottons New dark cuttings New black cottons White linens, Nos. 1, 2, 3, 4 New light second cottons Old extra light prints No. 1 Whites Ordinary light prints No. 2 Whites • Medium light prints House standard whites Dutch blue cottons Soiled standard whites German blue cottons Thirds and blues German blue linens Black stockings Checks and blues Dark cottons The moisture in baled rags as received at the mill has been found from a long series of tests to vary from 7 to 16 per cent 68 TREATMENT OF RAGS 69 70 RAGS, ESPARTO, STRAW, BAMBOO with an average of about 10 per cent. While rags are never purchased on the basis of a definite percentage of moisture, as is wood pulp, they should be tested at intervals to see that they are not intentionally wetted with the object of defrauding the buyer. The first process in the mechanical treatment of rags is gener- ally a dusting or thrashing. The bales are opened and the loosened-up rags thrown into the thrasher, which is usually a rapidly revolving cylinder covered with teeth or spikes enclosed in an outer cylindrical casing also fitted with teeth. There are various types of dusters but the object of all is to free the rags from loose dirt and deliver them sufficiently clean for the next operation, which is that of sorting into the numerous arbitrary grades maintained at the mill in question. This second grad- ing is desirable, because they are frequently very imperfectly sorted before baling as is shown by tests on seven different lots of linens, which, when examined at an American mill, were found to contain from 10.5 to 45 per cent, by weight, of cottons. During this sorting the larger rags are usually cut into two or three pieces, the seams are opened and buckles, buttons, hooks, iron, rubber, etc., are removed. Considerable skill and judg- ment are required to sort and grade rags correctly and much of the success of subsequent operations depends on the care with which it is done. The next step in the process is that of cutting the sorted rags into pieces 2 to 4 inches square. This is generally done by machinery though for the very highest grades hand cutting is preferred because it gives cleaner stock. Machine cutting causes greater waste in the form of dust and many bits of rag are also lost; the rags are also more stringy and do not empty from the rotary as quickly as hand cut rags. After cutting, the rags are given a final dusting in some form of willow which permits the dust to pass through the wire covering while the cleaned rags are delivered into cars and are ready to be conveyed to the boilers. The object of the boiling process is to dissolve or saponify the grease, loosen up the dirt and other impurities so that they may be easily washed out, and destroy or so modify the TREATMENT OF RAGS 7 1 coloring matters that they may be easily bleached. Another im- portant duty is that of destroying any wool which was not removed during sorting. The agents used to effect these changes are caustic lime, caustic soda or a mixture of lime and soda ash. The question of which to use is largely one of personal pref- erence and of the type of boiler used. Lime is only slightly soluble in water, one part dissolving in 1500 parts of water at 212 F. or in 728 parts at 68° F. It forms insoluble compounds with the grease and other impurities in the rags and is thus removed from solution, but as soon as it is precipitated a fresh portion dissolves to take its place so that the strength of the solution is practically constant throughout the boiling. This slight solubility of the lime limits the speed of reaction and the only way to make up for it is to increase the time of treatment. On the other hand, it prevents injury to the fibres from too great concentration of alkali and is therefore more likely to give good results when the usual, unscientific, hit-or-miss methods of control are employed. Caustic soda acts in the same way as lime except that it is readily soluble in water and the compounds formed also remain in solution and are therefore more easily washed out. Because of the complete solubility of caustic soda its use subjects the rags at first to strong solutions which continually diminish in strength as the treatment continues. If the caustic soda used is the chemical equivalent of the lime generally employed the rags will probably be overcooked and tender and the yield low, but if the caustic soda is used in smaller amounts and the time of boiling properly regulated there is no reason to think that it will give inferior fibre or a lower yield than a lime cook on the same stock. Lime-boiled rags are brighter in color than those boiled with soda but this difference frequently disappears after bleaching. The rapidity with which the alkali is "used up in a soda boil is shown by the following condensed data from large scale experiments by Beadle. 1 1 Chem. News, 1901, 84, 257. 72 RAGS, ESPARTO, STRAW, BAMBOO Rags Gals. per cwt. Liquor, per cent Na 2 Per cent Na 2 on rags Free Na 2 at hours below, start being 100 per cent 1 2 3 4 S i and 2 cottons . . . 24 25-7 22—22.5 23-25 °-33 O.654 O . 55-O . 90 O.91-I .36 0.697 1.68 1 .22-1 .78 I-9I-2-45 38 48 31-70 25-80 15 24 21-43 16-24 8 19 13-26 9-19 2nd cottons 3rd cottons Linens IO-23 7-13 •-I3 The severity of the treatment given rags varies greatly in different mills and with the grade of rags and the kind of paper to be produced. The caustic soda necessary is given by dif- ferent authorities as from i to 10 per cent, while the use of lime varies from 5 to 20 per cent on the weight of the rags. The steam pressure runs from 15 to 50 lbs., with 30 lbs. as probably a fair average, while the time of boiling ranges from 2 to 14 hours, and in some cases has been as much as 30 hours. Lime necessitates longer cooks than soda but the tendency in all cases is toward the shorter cooks since the steam consumption is less and the yields and the capacity of the plant are greater. Watt 1 gives the European practice as 12 hours at 30 lbs. steam pressure using 216 to 378 lbs. of lime and 114 to 190 lbs. of 48 per cent soda ash for 4000 lbs. of rags. In general it may be said that dark-colored or very dirty rags require much more severe treatment than new cuttings or clean, light-colored stock. The presence of starch in new cuttings is said to interfere with the boiling by gelatinizing and preventing the penetration of the alkali. It has been proposed to get around this diffi- culty by the use of malt to hydrolyze and remove the starch. The rags are boiled with water to swell the starch, cooled to 6o° C. by adding cold water, and an infusion of malt is then added. In one to two hours hydrolysis is sufficiently complete so that alkali may be added and the boiling completed in the usual way. The .lime used should slake rapidly and completely and should be as free as possible from iron. It is generally conceded that 1 Watt: Art of Paper Making. TREATMENT OF RAGS 73 its value for this work is in proportion to its content of avail- able calcium oxide. In preparing it for use it should be slaked and run into the boiler through wire sieves to take out any lumps. Griffin and Little 1 give the following analysis of lime as representing an excellent grade for rag boiling. Per cent Silica, etc., insoluble in acid o. 01 Iron and alumina (Fe 2 03 and AI2O3) o. 28 Lime (CaO) 92-81 Magnesia (MgO) 2. 28 Moisture, Carbonic Acid, etc. (by difference) 4.62 100. 00 The volume of milk of lime added to a rotary boiler is gener- ally enough to fill it from one-half to two-thirds full after the rags are in. If the rags are not covered by the milk of lime, but are exposed to the steam, they may become tender and brittle. Several types of boilers are used for cooking rags; spherical or cylindrical rotary boilers and stationary boilers in which the circulation of the liquor is maintained either by a pump as in the Mather kier or by a vomiting pipe which is connected with a false bottom. In this case the steam enters under the false bottom and in passing up the pipe carries along the cooking liquor which is then distributed over the charge by baffle plates. The stationary type of boiler is fitted with a safety valve and there is therefore less danger of explosion than with the rotaries. It is claimed that there is less loss of fibre because the rags are not in motion and hence there is no rubbing off of the weaker fibres. On the other hand, they require more steam, are not suitable for use with lime cooks and take longer to discharge, since the cooked rags must all be removed by hand through manholes. Stationary boilers are largely used in Great Britain but seldom on the Continent or in this country. Of the rotary boilers the cylindrical is more generally em- ployed than the spherical, though the shape of the latter greatly assists in discharging its contents. The boilers are charged 1 Griffin and Little: Chemistry of Paper Making, p. 152. 74 RAGS, ESPARTO, STRAW, BAMBOO through manholes and with large boilers the men frequently enter them in order to pack the rags properly. The steam enters the boiler through the trunions and passes through grat- ings on the inside into the stock. The rotaries are also fitted with strainers through which much dirt passes in blowing off pressure. The rotaries usually turn at a speed of one revolu- Fig. 5. Spherical Rag Boiler tion in two to five minutes. The consequent agitation and friction of the rags cause the detachment of the lime compounds formed and allow the action to continue on the remaining fatty materials. When the cook is completed the liquor and steam are blown off through the strainers and blow-off cocks, the manhole covers are removed and the contents are discharged by allowing the boiler to revolve. Even a large cylindrical boiler will empty itself clean in this way. With competent labor the operations on a cylindrical boiler holding a charge of 5000 to 5200 lbs. of rags occupy the following times: Hours Charging rags if Blowing off liquor 2 Discharging rags 3 During the cooking of rags a small amount of ammonia is given off. Tests on a large scale showed that the following amounts were discharged. TREATMENT OF RAGS 75 o. 22 per cent ammonia (NH 3 ) from Japanese blues, o. 05 per cent ammonia (NH3) from second white stock o. 14 per cent ammonia (NH 3 ) from clean white linen o. S3 P er cent ammonia (NH 3 ) from German blues The washing of the cooked rags is usually done in large en- gines similar to beating engines. One or more cylinder washers covered with perforated metal or wire cloth, frequently old Fourdrinier wire, serve to remove the dirty water which is constantly replaced by fresh water. If the rags are allowed to stand some time with the lime on them before washing they become softer and more absorbent and are better for blotting papers. On the other hand, if they are washed at once they are harder and the color is rather better. The lime compounds formed during cooking are friable and easily washed out in most cases but occasionally enough mineral oil is present so that they collect in sticky masses on the sides of the washer, fill up the meshes of the wire on the washing drum, and appear as small globules in the rag stock itself. When lime is used the washing should always be done with cold water because of the greater solubility of lime at low temperatures. When the water from the cylinder washer comes away fairly clear the beater roll is lowered to loosen up the remaining dirt and gradually reduce the rags to half stuff. If this is done too soon the fibre loss is unduly great and the frayed out fibres catch and hold the dirt and the resulting stock is of poor color. After the rags are reduced to half stuff the bleach is added to the engine and the whole put into drainers to complete the bleaching. The amount of water used in preparing rag stock is very con- siderable. Sindall 1 states that the water required to wash the rag stock for a ton of paper is 25,000 to 35,000 gals. Beadle 2 estimates the quantity used in preparing the rags for one ton of two grades of paper to be as follows, measurements being in Imperial gallons. 1 Sindall: Paper Technology. 2 Beadle : Chapters on Paper Making, Vol. IV. 7 6 RAGS, ESPARTO, STRAW, BAMBOO Bank paper Rag paper Boiling rags 1,500 1,500 40,000 5.000 48,000 Rinsing rags Washing in engines 37.830 5,000 43.930 Washing out bleach The losses which rags undergo in making into paper vary enormously with the quality of the rags and the treatment they are given. The following table, condensed from data in Hof- mann's "Papier Fabrikation" shows the losses in preparing half stuff from different materials. Kind White linens Bagging White cotton Red and blue calico. . . Half wool and old rags Blue linen Mois- ture Per cent 5-7 5-i 1 4-5 6 6 6-7 Cutting and dusting Per cent 4-5 5-7 5 6 6 6 Cook- ing and wash- ing Per cent 7-12 II-18 n-13 12-13 38 IO-13 Beat- ing to half stuff Per cent 4-9 12-14 8-9 12-13 10 9-14 Bleach- ing Per cent 3-6 5-9 4 4-5 4 4-5 Total loss Per cent 26-33 42-55 34 40-42 64 35-45 Kilos bone dry half stuff per 100 kg. rags 67-74 58-45 66 58-60 36 55-65 Similar data from the records of an American book paper mill using rotary boilers are: Kind Sorting and tare Thrashing Cutting Total loss to finished paper Japanese blues New calico Belgian blues Whites Muslins Russian blue linens Per cent 3. Per cent 2.4 4.0 4-9 0.2 1.8 2-5 Per cent 3-6 0.4 2.8 5-o 3-i 6-7 Per cent 34-6 26.2 42 .2 25-5 4i-7 45-8 Beadle x in recording English practice gives the following figures for the losses on boiling and bleaching. 1 Beveridge: Paper Makers' Pocket Book. ESPARTO 77 Best new cottons Low grade cottons No. i cotton rags No. 3 cotton rags New unbleached cottons Percentage lost on Boiling Per cent 8.71 12.20 5-8o 12.50 2 3S° Bleaching Per cent .29 .70 .20 Esparto. Esparto, because of its high cost, has never come into competition in this country with soda fibre prepared from woods, but in Europe, and more particularly in Great Britain, it is one of the most important fibrous raw materials. Intro- duced into England in i860 by Routledge, its use increased from 16 tons in 1861 to 184,000 tons in 1884, since which time about 200,000 tons have been used every year, not including what is used on the Continent. The grass is imported in bales bound with iron or twisted bands of esparto. These bales are opened up and spread out on tables covered with iron netting to allow sand and dirt to pass through, and the grass is sorted by hand to remove roots, weeds, etc., which are more resistant to alkali than esparto and hence cause dirt and shives in the pulp. The loss in dusting and hand picking is 1 to 6 per cent. The dust contains sand and other mineral matter as well as fat or wax from the esparto itself. The analysis of fine dust is given by Beveridge * as follows : Per cent Organic matter (by ignition) 64. 6 Water (loss at 212 F.) 6_ 2 Mineral matter 2 n 2 Of the organic matter, about 90 per cent is of a waxy nature, while the mineral matter or ash gave the following analysis: 1 Beveridge: Paper Makers' Pocket Book. 78 RAGS, ESPARTO, STRAW, BAMBOO Per cent Silica (Si02) 56. 43 Calcium carbonate (CaCOs) 19. 17 Magnesium carbonate (MgCOs) 3-76 Alumina (AI2O3) 20. 57 . 99-93 Machine cleaning by willows is also resorted to and by some is considered superior to hand picking. The grass from the willows goes directly to the cookers without any further sorting, and sand and uncooked weeds, etc., are removed by sand traps and screens after cooking. The boilers used for esparto may be merely open tubs since little or no pressure is necessary but the more general practice is to employ closed, stationary boilers with some device for cir- culating the liquor. Rotary boilers are almost never used as they cause the fibres to roll up into small balls which beat out with great difficulty and are apt to cause lumps in the paper. The capacity of esparto boilers is generally 2 to 3 tons of grass. The boiler is partly charged with lye and the esparto is fed in, steam being admitted at the same time to soften it and enable a greater quantity to be packed in. The grass, which is not cut as in the case of straw, is charged through the top and after cooking is removed through a side door near the bottom. The steam pressure carried varies in different mills from 5 to 50 lbs., and the time of cooking from 2 to 6 hours; the present tendency seems to be toward the higher steam pressures. As compared with rags esparto requires a much greater amount of alkali, which is invariably soda, since lime is never used because of the formation of insoluble compounds. The amount of alkali recommended by Dunbar for different grades of grass is as follows: Fine Spanish 16. 1-17. 9 lbs. 70 per cent caustic per 100 lbs. grass Medium Spanish.. 14. 3-16. 1 lbs. 70 per cent caustic per 100 lbs. grass Fine Oran 16. 1 lbs. 70 per cent caustic per 100 lbs. grass Medium Oran. ... 14. 3-15. 2 lbs. 70 per cent caustic per 100 lbs. grass Tripoli 17. 0-17. 9 lbs. 70 per cent caustic per 100 lbs. grass ESPARTO 79 The liquor for boiling varies from 7 to 15 degs. Tw. As typical examples of actual cooks made under fairly good working conditions Beveridge 1 gives the following: Weight of charge Caustic liquor per charge, gals. . . Pounds of 60% caustic per charge Steam pressure, maximum Time under pressure Yield, unbleached air dry fibre . . Spanish 5600 lbs. I570 900 20 lbs. 2§ hrs. 44-45% Tripoli 5600 lbs. I570 1020 20 lbs. 3 hrs. 41-42% When the cook is completed the pressure is blown down and the liquor run off to the recovery system; hot water is then run into the boiler and the pressure brought up to 20 to 30 lbs. This is again blown off, the stock drained as dry as possible, re- moved, and conveyed to the washing system which usually consists of engines similar to those used in treating rags. Dur- ing washing much of the cellular matter (leaf hairs) is removed and some short fibres are lost. The bleaching of esparto is generally carried out in engines. The bleach required varies from 7 to 12 per cent of the un- bleached fibre and often a little sulphuric acid is added to the engine about half an hour after the bleach. When bleaching is completed the stock is run off on a press-pate which serves the double purpose of removing the bleach residues and de- livering the fibre in a convenient condition for further operations. The recovery of alkali from the waste liquors is conducted along the same lines as with the liquors from wood. Accord- ing to Griffin and Little 2 the silica in the ash from esparto forms silicate of soda during furnacing and thus reduces the per cent of ash recovered. A recovery of 80 per cent is con- sidered good while 85 per cent is the most that can be expected under the best conditions. 1 Beveridge: Paper Makers' Pocket Book, p. 76. 2 Griffin and Little: Chemistry of Paper Making, p. 158. 8o RAGS, ESPARTO, STRAW, BAMBOO Straw. Straw as used in paper making includes the stems and leaves of the various cereals. The composition of straws, particularly with regard to the amount of ash and its constitu- ents, varies greatly with the soil upon which they were grown. Wolff 1 gives the following analyses for different straws. Water Ash Fat and wax Nitrogenous matter Starch, sugar, gums, etc Cellulose Winter rye Per cent 14-3 3 3 5 25-7 54 Winter wheat Per cent 14.2 5-5 i-5 2 .0 28.7 48.0 Sum- mer barley Per cent 14-3 Winter barley Per cent 14-3 5 4 28 Oats Per cent 14-3 5-0 2.0 2-5 36.2 40.0 Corn Per cent 14.0 4.0 1 .1 3° 37-9 40.0 Beveridge 2 considers the cellulose as determined by Muller's method too high and gives the following percentages obtained by digesting the straw with bisulphite of soda. Per cent French wheat 41.5 Zealand wheat 40. 9 Dutch wheat 41. 6 Dutch oats . . 42. o Dutch rye 44. 7 Dutch barley 38. 3 Cross and Bevan 3 give the percentage of bone dry cellulose on bone dry straw as follows: Per cent Oats 52.0 and 53. 5 Oats, foreign 46. 5 Wheat 49. 6 Wheat, foreign So. 2 Rye, foreign 53. o From analyses of rice straw by Takeuchi 4 the following figures are taken: 1 Wolff: Landwirtschaftl. Kalender, 1869, zitiert bei Hugo Miiller, Die Pflan- zenfaser, p. 97. 2 Beveridge: J. Soc. Chem. Ind., 1894, 101. 3 Cross and Bevan: Text Book of Paper Making, 1907, p. 136. 4 Takeuchi: Bull. Coll. Agric, Tokio, 1908, 7, 619-621. STRAW Hygroscopic moisture .... Dry matter Total nitrogen ■ Crude protein Crude fat Crude fibre Crude ash Silica Dextrose Sucrose Starch and hemicelluloses Pentosans Good harvest Poor harvest Per cent Per cent 12.31 9-85 87.69 90 15 0.97 1 48 6.05 8 82 I.36 1 65 31.16 28 72 11 .42 12 35 5-39 6 13 2.25 3 28 0.79 96 14.86 18 75 14.28 16 55 Analyses by the author of two samples of American grown rice straw of the "Carolina Gold" variety showed the presence of 14.5 and 12.4 per cent of ash of which silica constituted 81.6 and 79.7 per cent respectively. The ash from various straws shows the following compo- sition: 1 Total mineral K2O Na 2 CaO MgO Fe 2 3 P2O5 so 3 Si0 2 CI Barley, aver, of 4 Oat, aver, of 8. . . . Rye, aver, of 3. . . . Wheat, one test. . . % 8.IO 7-77 4-32 3-25 % 23-75 38.37 26.28 12.16 % 1 .92 3-99 0.74 1 .00 % 7-53 4-23 11 .10 6.82 % 2.62 2-53 4-45 4.00 % 2.19 1.79 3-i9 1 .02 % 3-94 2.66 8-97 3.20 % 3-91 3.06 5-57 5-78 % 51-43 35-68 36.86 65-34 % 3-75 7-99 3-68 0.60 Straw is used for the production of two quite dissimilar fibrous materials; a coarse, yellowish, half stuff which is used for straw boards and cheap wrapping papers, and a bleached cellulose, similar to esparto in many of its properties, which enters into the composition of numerous high grade products. For the manufacture of straw boards, the straw, after dusting and cutting into short lengths, is treated with milk of lime in rotary boilers. The action of the lime is not vigorous, incrust- ing matters are not separated from the cellulose to any great extent and mineral matters originally present in the straw re- main practically untouched in the finished product. According 1 Wolff: Ashen analysen. 82 RAGS, ESPARTO, STRAW, BAMBOO to Kirchner x it is the general practice in Germany to cook the chopped straw in rotary spherical boilers of 14 cubic metres (494 cu. ft.) capacity which hold 1100 kilos (2420 lbs.) or 1500 kilos (3300 lbs.) after tamping. The steam pressure varies from 45 to 75 lbs. and the time of cooking from 4 to 5 hours. A yield of 70 to 80 per cent is obtained. In some plants the lime is slaked and strained into the boiler while in others the quicklime is added without slaking. If the lime is poor a little soda ash is sometimes added. Rye and wheat straw should not be boiled together since wheat needs more treatment than rye. For rye straw 8.5 per cent of quicklime is sufficient while for wheat 10 per cent is required. It is gener- ally reckoned that there is a loss of 30 to 35 per cent of fibre substance from the dry straw to the finished paper. According to B. Haas 2 the chopped straw and liquor are fed into the digester together and gently steamed at the same time, thus increasing the capacity of the digester fully 10 per cent. The digester should be rotated without steam to distribute the liquor, then steamed with the blow-off cock open to allow air to escape and finally steamed at 60 lbs. pressure for 2\ to 3 hours. If handled in this way it is claimed to be possible to cook wheat straw with 6 per cent of lime while under ordinary conditions it would require 13 to 15 per cent. If too much lime is used the resulting stock is " greasy," drains slowly and clogs wires and felts. In France considerable straw is said to be treated by a cold process as follows: Rye straw is cut short and put into large rectangular br^ck wells where it is just covered with dilute milk of lime. Boards are put on, weighted down with stones and the whole left for two to four weeks. It is then taken out and worked in edge runners for at least an hour. This product is harder than straw pulp cooked by steaming at high tempera- tures and as the knots are not softened the grinding must be done with especial care. Straw pulp prepared in this way 1 Kirchner: Das Papier, III, B and C. 2 Der Papier-Fabr., 12 (1914), 305-310. STRAW 8$ possesses a natural self-sizing property which is lost if the milk of lime is heated and is absent in pulp cooked by the soda pro- cess. Hot cooked stuff requires much rosin size because of the presence of lime residues which it is impracticable to com- pletely wash out. The general American practice differs somewhat from the German in that the straw is not cut and sorted before cooking. The kind most generally used is winter wheat though some use rye and oat straw. It is always received at the mill in bales which are stored in large stacks near by; these stacks are fre- quently provided with roofs and sometimes with shelter on the sides. There appears to be little deterioration of the straw, except on the very outside of the bales or in those at the bottom of the pile, and even such straw, which has become very dark in color, is used by adding a little more lime to bring it to the right color. In filling the rotaries the bales are broken up and immediately fed in without sieving, cutting or any other method of preparation; in fact as much as a quarter of a bale is some- times added without loosening up in any manner. This is done because it is believed a larger charge can be secured. As the rotary is filled about a gallon of water is added for every 2\ lbs. of straw. The rotary is then revolved and steam ad- mitted to raise the charge to '25 lbs. pressure which is main- tained for half an hour or so. The head is removed, more straw added and the charge again steamed or " wilted" as before. This is repeated three or four times so that the final charge is about double that which was first added. At this point milk of lime is added, equivalent to 10 per cent calcium oxide (CaO) on the weight of the straw, after which it is rotated and brought up to 35 to 45 lbs. steam pressure where it is held for 8 to 10 hours. The liquor is drained off under pressure, the charge cooled till practically at atmospheric pressure, and then dumped onto the floor where it is allowed to season three to five days before being used. The straw is washed about four hours in washing engines, the washer taken up and the beating proceeded with. 84 RAGS, ESPARTO, STRAW, BAMBOO The time required for one cook, including charging, cooking and cooling, is about 24 hours. The yield per ton of straw is generally figured as 1600 lbs. of board for rye, 1400 lbs. for wheat and 1200 lbs. for oat or rice straw. The treatment of straw by the soda process for the production of a high grade bleached cellulose occupies an important place in the paper industry of Europe though it is practically never undertaken in this country. The kinds generally used are wheat, oat, rye and barley straws, and of these the first two are used most extensively in England. Since straw is more highly lignified than esparto it requires a more drastic treat- ment; even the knots must be so reduced that they will bleach readily. Because of this more severe treatment and because of the presence of cellular tissues which are lost during the washing process the yield from straw is less than that from esparto. Cross and Bevan 1 give the practical yield as about 35 per cent, while Beveridge 2 says that a mixture of equal quantities of barley, oat, wheat and rye straws will yield 40 to 41 per cent of air dry bleached cellulose. The straw cellulose makes a weaker paper than esparto but it is suitable for mixing with rags or wood pulp for thin, hard, rattly papers. It tends to make the stock a wet" on the wire and imparts translucency to the papers in which it is used. In modern plants the straw is cut by rotary cutters into pieces one to two inches long and freed from dust, grain, etc., by an air blast. It is then fed into the boilers, steam and alkali added at the same time assisting in packing the charge. Contrary to the practice with esparto rotary boilers are preferred for straw because the agitation of the charge permits a penetration of the liquor which cannot be obtained in a stationary cooker because of the close packing of the wet straw. The amount of alkali used varies greatly at the different mills and with the kind of straw; it may run as low as 10 per cent or as high as 20 per cent of the weight of the straw. Experiments conducted 1 Cross and Bevan: Textbook of Paper Making. 2 Beveridge: Paper Makers' Pocket Book. STRAW 85 by the author on rice straw proved that a simple extraction with water at 20 to 25 lbs. steam pressure removed so much material from the straw that well reduced and easily bleached fibre could be produced with 60 to 75 per cent of the caustic soda necessary for unextracted straw. Barley straw is said to require 20 per cent less soda than oat, wheat or rye straw. The time of cooking is variously given as 3 1 to 8 hours and the steam pressure as 10 to 90 lbs.; there seems to be a decided tendency toward the use of the higher pressures of from 75 to 90 lbs. The cooked straw is run or blown from the boiler into wash tanks with false bottoms; these are preferred to drum washers because of the large loss of fine fibres and cellular matter which the latter cause. After washing, the straw is treated in edge- runners to crush the knots and it is then bleached in much the same way as esparto; the bleach required is from 7 to 10 per cent. The recovery of alkali is carried out in the same way as with esparto or wood but the working of the process is rendered difficult in many instances by the silica in the straw. This combines with the alkali forming sodium silicate and when the recovered ash is causticized a bulky, gelatinous precipitate of calcium silicate results. This prevents settling in the causti- cizing tanks and very greatly reduces the amount of soda which can be recovered. A patented process by Sutherland and Kynaston proposes to precipitate the silica by adding bicarbo- nate to the solution of the recovered ash; carbon dioxide can also be used. They claim to obtain in this way a granular precipitate which can be easily handled, but the process has not been an entire success. Beveridge r states that the soda lost varies with the amount of silica in the straw, with the composition of the silicate formed and with the amount of potash rendered soluble. He estimates that as much as 42 per cent of the total soda in the cooking liquor may be neutralized by the silica and claims that when the amount of silica in the straw approaches 3 to 4 per cent of its weight the recovered ash is of little value for further 1 Beveridge: Paper Makers' Pocket Book. 86 RAGS, ESPARTO, STRAW, BAMBOO digestions. The recovery in a Russian mill operating by the sulphate process is claimed to be 80 per cent. 1 Numerous other processes have been suggested for treating straw. Chlorination in stone chambers, of straw which has been partially cooked with caustic soda, followed by a treat- ment with bleach, gives a high yield of well-bleached, uniform stock, but the process is too costly and difficult to manage. Diess and Fournier 2 propose steeping the straw in acidulated water for three hours, retting by organisms cultivated from African esparto, and finally cooking with strong caustic soda solutions — 20 to 30 Be. — at 45 to 75 lbs. steam pressure for three to five hours. Reichman 3 treats the straw with caus- tic soda, washes, treats with hydrofluoric acid of i° to 2 Be. for about five hours and finally washes with dilute ammonia. Probably the only modification of the original soda process which has found any extensive use is the sulphate process and this has been applied to straw with notable success. The sulphite process is not generally applied to straw, though in isolated cases good fibre has been prepared from it in this way and practical experience has. shown that it can be employed with excellent results. The general assumption that the large proportion of silica in the straw would interfere with its treat- ment by the sulphite process is apparently not founded on fact. Bamboo. Bamboo, while not of immediate interest in this country, seems destined to hold an increasingly important place as a source of fibre because of the rapidity of its growth and the high quality of the paper which can be made from it. Raitt 4 esti- mates that with poor growth the annual yield of stems would be n tons, air dry, per acre while with luxuriant growth it may amount to as much as 44 tons. He states that in Lower Burma alone there is an area of about 20,000 square miles easily available. 1 Altaian: Chem. Ztg., 191 1, 35, 979. 2 French Pat. 403, 518. 3 English Pat. 12,059, May 21, 1909. 4 J. Soc. Chem. Ind., 1908, 27, p. 35. BAMBOO 87 Characteristic analyses of Philippine bamboos are given by Richmond as follows: : Structural bamboo Dwarf bamboo Cellulose Per cent 53-94 0.96 4.98 24-25 12 .40 3-47 Per cent 55-75 1.03 4.69 21 .27 11 .20 6.03 Fat and wax Water extract Non-cellulose or lignin Water Ash Analyses of typical absolutely dry Indian bamboos are: Cellulose Fat and wax Water extract Pectose Lignin Ash. B. polymorpha Per cent 54.71 1 -05 8-95 19-55 15-74 100 . 00 3-97 B.arundinacea Per cent 50.32 1. 17 8.48 24-39 15-64 IOO.OO 1 .60 C. pergracile Per cent 52.73 0.92 7.96 By an extended research on the five most likely Indian bam- boos, Raitt 3 has shown that by the soda process yields of 41.0 to 43.0 per cent of bleached fibre may be obtained, but the bleach consumption is high. When the sulphate process is used the yields are 42 to 44 per cent and the bleach required is much lower— 15.5 to 18.0 per cent. He finds the sulphite process is unsuited to bamboo because of the- difficulty of bleach- ing the fibre and of working with sulphite liquor in a tropical climate. Raitt overcame the difficulties previously encountered with bamboo by adopting the following treatment: (1) Culms not to be cut till the shoots of the year are full grown. 1 Richmond: Philippine J. Sci., I, 1906, 1075-1084. 2 Raitt: Indian Forest Records, Vol. Ill, Part III, p. 15. 3 Raitt: Indian Forest Records, Vol. Ill, Part III. 88 RAGS, ESPARTO, STRAW, BAMBOO (2) Seasoning for at least three months before use. (3) Crushing. (4) Extraction of starchy matter. (5) Digestion with sulphate liquor. The limiting conditions of satisfactory digestion for the five species investigated were found to be 20 to 22 per cent caustic soda (including the sodium sulphide), temperatures of 162 to 177 degs., pressures of 80 to 120 lbs., and durations of 5 to 6 hours. Old Papers. In the case of old printed papers of the higher grades, such as book and magazine papers, in which there is no ground wood, little difficulty is experienced in preparing them for use a second time. The removal of the printer's ink can be effected by ■ cooking in digesters with a caustic soda solution followed by disintegration and washing of the pulp. The alkali removes the rosin sizing and saponifies the oily constituents of the ink thus rendering them soluble and loosening the pigments so that they may be detached from the surface of the fibres and washed out. In this process, as applied to old magazines for instance, the staples are removed by mechanical means and the magazines fed into rotary digesters. The removal of the staples allows them to come to pieces sufficiently so that the alkali can pene- trate enough to reach all parts of the paper and opening up by thrashers is therefore not necessary. About 3 to 4 per cent of caustic soda, on the weight of the papers, is then added together with enough water to insure thorough saturation of the charge and it is cooked at 40 to 50 lbs. steam pressure for a number of hours, sometimes as long as 13 hours. After blow- ing down pressure the rotary is dumped and the stock allowed to drain, after which it is transferred to washers and washed and bleached in practically the same manner as rag stock. The time required for washing varies greatly with the size and condition of the washing engine, it may even go as high as 15 hours although this is not usual if the stock has been well drained. The bleach required amounts to 3 to 4 per cent of the OLD PAPERS 89 papers used and the color obtained is usually a grayish white because of the impossibility of removing all traces of the carbon from the printer's ink. This process involves a considerable loss due first to the action of the soda and steam in the digester and second to the mechanical loss of fine fibres and the mineral fillers during the washing. The combined loss from these causes will frequently amount to 35 to 40 per cent of the absolutely dry papers used. Similar results are obtained by tearing the papers up some- what and heating in large boilers with an 8 to 10 per cent solu- tion of soda ash. The temperature is maintained just below the boiling point as the object is merely a surface loosening of the ink and not complete disintegration of the paper to a fibrous stock. It has also proved possible to eliminate the rotary digester and carry out the entire process in the beating engine or washer. The printed papers, either plain or coated, are fed into the engine with water and about 2 per cent of caustic soda based on the papers. When warmed to 120 F. they disintegrate readily and in a short time the washing can be started and carried out as usual. The color of the stock obtained is better than that of the material cooked in rotaries and the loss is probably slightly less. The chief objection to the process is that the old magazines which are largely used have to be thoroughly broken up before adding to the beater. In either of these processes ground wood has to be carefully avoided because it is turned brown by the alkali, yet is not cooked enough so that it will bleach to a good color. Serious trouble has been caused at times by getting ground wood papers into the rotary and it is well to supply the sorters with phloro- glucin or paranitro-aniline so that they may test suspected papers and throw aside all those found to contain ground wood. Stock which contains ground wood, after passing through the rotaries and bleachers, has the appearance of being contami- nated with fine, brown hairs or fibres. These are very con- spicuous when such print papers are used in a white sheet. It 90 RAGS, ESPARTO, STRAW, BAMBOO is interesting to note that after cooking the ground, wood is just enough changed so that it gives no test, or at most a very- faint pink, with phloroglucin. This explains why it has proved so difficult to recover the stock from old newspapers, which consist of approximately three-quarters ground wood. The problem has, however, proved very attractive to a large number of investigators and the patents taken out are very numerous. Among the reagents which it has been proposed to use are various alkalis and alka- line salts, silicates, borates, phosphates, etc., soaps, peroxides, hypochlorites, aluminum chloride, enzymes and inert materials, such as clay, talc, fuller's earth, etc. These latter seem to be added to serve as points about which the pigments from the. ink may gather and thus facilitate their removal by the washers. Most inventors are not content with adding single reagents or even simple combinations of two or three, but in cases use as many as eight different substances at the same time. Any such combinations to be of value in treating ground wood papers must be only weakly alkaline and must be used at com- paratively low temperatures. Among the materials which have been found best for this purpose are fatty soaps and similar materials used in conjunction with soda ash and sodium silicate. Experiments have proved that the disintegration of the printed papers with the detergent in kneaders where comparatively dry conditions are maintained produces poor results. The pig- ment is set free under such conditions that it is ground into the pores of the fibres and it is then almost impossible to wash out enough of it so that a good color can be obtained. If the papers are disintegrated and then diluted in the washer before mixing with the detergent much better stock will result because the separated ink is then in such a state that it tends to rise to the surface and can be more readily removed by the washers. It is evident, then, that a method of disintegration which tends to pull the ink from the surface of the paper is superior to one which tends to grind it into the fibres. Such a procedure has been embodied in the Winestock process OLD PAPERS 91 for the recovery of old papers. The apparatus used is shown in sectional view in Fig. 6 of a machine driven by a direct connected steam turbine, M. The essential features are a propeller tube B at the bottom of a cylindrical tank A which is mounted within a chamber H; through the horizontal part of the pro- Fig. 6. Westestock Defibering Machine Courtesy of Castle, Gottheil & Overton peller tube extends a shaft which turns at 2000 revolutions per minute and upon which are mounted two propellers of different pitch. Between the propellers is a baffle plate K to stop any inclination toward a rotary motion of the stock in the tube. The papers to be treated are opened up and dusted in a rail- road duster or its equivalent and then soaked in a tank of water at about 160 F. either with or without the addition of soda ash or other detergent. They are then charged into the cham- 92 RAGS, ESPARTO, STRAW, BAMBOO ber H, together with any other additional detergents desired, such as a mixture of soda ash, caustic soda or a soap composed of tallow, soda ash, caustic potash and silicate of soda. The chamber H being full the stock overflows into the cylinder A and is circulated by the propellers back to the chamber H which it enters tangentially and through which it circulates to again enter A. The basic idea is that the paper suspended in water or a weak alkaline solution is struck by the blades at such a speed that it is unable to take up the rapid motion and is therefore pulled out or defibred while at the same time the ink is loosened from the fibres by the violent agitation. The second propeller has a greater pitch than the first so that there is cavitation between the two or a pressure on one side of each propeller and a suction on the other. The speed at which the stock circulates is estimated at 1200 ft. per minute. Since there is no cutting or grinding action in this machine there is no shortening of the fibres. The time of treatment necessary in this machine varies with the papers used and the products desired. Newspapers which are to be used for boards and from which the ink is not removed require only about fifteen minutes; if the product is to be used for white paper, and the ink must be washed out, about thirty minutes are required. Book and magazine papers need thirty- five to forty-five minutes' treatment and hard sized writings a somewhat longer time. The machine holds 700 to 900 lbs. of dry papers per charge and requires from 40 to 60 horse power. Assuming that the papers have been sorted with reasonable care and strings, bags and foreign materials eliminated, the Winestock process is suitable for all classes of papers, since the low temperatures and mild chemicals do not cause any serious discoloration of ground wood. As the apparatus liberates the ink and color from the fibres but does not remove the loosened pigment a subsequent washing is necessary. CHAPTER IV THE SODA PROCESS The principles upon which this process depends are the sol- vent power of the caustic soda for certain constituents of the wood and the hydrolysis of other constituents resulting, to a considerable extent, in the formation of products of an acid nature which are then brought into solution as salts of soda. Both of these processes neutralize the alkali and by diminishing its concentration and hydrolyzing power render it useless for further work until it is regenerated. The reactions and de- compositions involved are of a very complicated nature and the products are numerous and for the most part ill-defined and little understood. The degradation of the woody constit- uents is in general far greater than for the same constituents when dissolved by the sulphite process. Even at low temperatures alkali dissolves a very appreciable proportion of the non-cellulose constituents while if the tem- perature is raised the action is greatly intensified. Experiments on small poplar chips showed that a 3.3 per cent solution of caustic soda would dissolve 20.3 per cent of their weight by twenty-four hours' treatment at 25 C. while if the temperature were raised to 8o° C. the wood lost 31.6 per cent of its weight in the same time. Higher temperatures, such as are obtained by steaming under pressure, still further enhance the solvent power of the alkali and the speed with which it acts. In working with materials other than wood due consideration must be given to pectous substances. Working with bamboo Raitt * finds that pectose (matter soluble in 1 per cent NaOH solution at 100 degs.) is easily soluble in boiling NaOH solutions 1 Indian Forest Records, Vol. Ill, part 3, "Bamboo as Material for Paper-pulp." 93 94 THE SODA PROCESS but that it gelatinizes at the temperatures employed in digestion and is therefore likely to become mechanically attached to the cellulose and is then very difficult to wash out. Pectose, fat and wax grouped together neutralize 0.32 per cent of NaOH on the raw material for each 1 per cent found on analysis. Lignin, unlike pectose, is not soluble in weak solutions nor at tempera- tures below 130 degs. The wood most used in the soda process is poplar, at least in the northern part of the United States, but because of its in- creasing cost and scarcity other woods are frequently substi- tuted. Among these may be mentioned basswood, maple, birch, cottonwood, tulip tree, sycamore, several kinds of gum, chest- nut, beech, etc. The best results are obtained if the different kinds are treated separately but this is frequently impossible without an excessive amount of labor. If mixed woods are used it is desirable to employ those which require about the same degree of treatment and to keep the mixture as nearly constant as possible. The mixing of woods which require widely different cooking conditions invariably means diminished yield because of the overtreatment of at least one kind in order that the most difficult to reduce may be sufficiently cooked. For long fibred stock, spruce, hemlock, pine and white fir are sometimes treated by the soda process. They require more alkali and longer cooking and yield less pulp than the broad- leaved woods. The soda pulp industry is, however, using pine in rapidly increasing quantities and in this way it is possible to use a number of woods which are too resinous to be treated by the sulphite process. According to the United States De- partment of Agriculture out of a total of 843,048 cords of wood used in the soda process during 1917, 379,466 cords were poplar, 11,069 cords were hemlock, and 116,267 cords were pines of various kinds. On account of the vigorous action of the alkaline solutions less care is necessary in preparing the wood than for the sulphite process. Knots are either dissolved by the treatment or left in such condition that they are easily separated by the screens. THE SODA PROCESS 95 Contrary to the usual belief the inner bark, however, is a source of trouble and should be removed as completely as possible. It cooks with difficulty, uses up fully as much caustic soda as sound wood, and bleaches with far more difficulty than the fibre from the wood. If the two are cooked and bleached together the resulting product is liable to be contaminated with brown, stringy shives. The outer bark also uses up a large amount of caustic and in cooking breaks down into masses of cells which will not bleach to better than a yellowish brown color and hence cause dirt specks in the bleached pulp. These are not the only faults of bark, for its rough surface tends to catch dirt from outside sources, such as cinders, sand, etc., and transfer it to the pulp. It is not necessary to remove decayed portions of the wood since they are completely resolved during the cooking and do no harm unless they are of a very black nature. Such decayed wood, however, gives a very low yield of fibre and if present in large amount greatly reduces the output of the digesters, so that for this reason, at least, its use is to be avoided. The cellu- lose in partly decayed poplar wood was found by us to be 24.9 and 27.0 per cent in two samples as compared with 63 per cent for sound wood. The wood is chipped by running the logs diagonally against the face of a rapidly revolving disc from which project from two to four knives. The distance to which these extend de- termines the length of the chip, which for poplar is from three- quarters of an inch to an inch and a quarter. The chips go next to some form, of screen which separates them into three grades, dirt and very fine material, good chips, and slivers and coarse pieces. The dirt and fine stuff is waste, so far as pulp making is concerned, though it is often used as fuel in the boiler house, while the slivers and coarse pieces are either crushed or rechipped and returned to the screens. Uniformity of cooking is greatly aided by a uniform size of chips but it is not so essen- tial as in the sulphite process because of the greater penetrating power of the alkaline liquor. Fig. 7. Vertical Digester, Section Showing Inlets, Pump, and Piping (96) ^ DIGESTERS 97 The digesters used in this process are of the usual rotary or stationary types but vary greatly in size and capacity. The tendency is toward the vertical stationary type Fig. 7, and away from the rotaries since the former effect savings in floor space, power required and time of filling and emptying. The latter is a very appreciable item, as a rotary holding 3 to 3! cords requires three hours for blowing down pressure, discharging the contents and refilling with chips and liquor, while a vertical digester A. Digester B. Heater for cook- ing liquor C Strainer D. Boiler to supply steam to B E. Pump to return condensed water to boiler W/^//»: ,.',.,::? Fig. 8. Morterud Digester holding 14 to 15 cords of wood requires only one hour for the equivalent operations. The size of rotary digesters in American practice is generally about 20 X 7 ft., while the vertical digesters range from 27 to 49 ft. tall by 7 to 10 ft. in diameter; one 30 X 8 ft. will hold about 5.5 cords while one 49 X 10 ft. will hold nearly 15 cords of wood. A modified type of digester is that used in the Morterud system of cooking with forced circulation and indirect heating. Such an outfit is shown in Fig. 8. The steam in this system is not blown directly into the charge but the liquor is circulated through an outside exchange heater in which steam is the source 98 THE SODA PROCESS of heat. The condensed steam is returned to the boiler as feed water under high pressure and at a high temperature. The claims for this process are savings in coal and alkali and an increased yield because of the more even temperature of the cooking liquor. Another modification is that of the jacketed digester in which the steam for heating is between the two walls. Experience has proved that these are very hard to keep tight as the stays be- tween the walls cause very frequent leaks. No insulating cov- ering can therefore be used and the steam consumption is very high because of the great loss by radiation. The material of the digesters is either iron or steel and they are made either riveted or welded. The latter is much to be preferred, since" the alkaline solutions work their way through crevices which would be impervious to water and for this reason a riveted seam is very difficult to make perfectly tight. As the alkaline solutions are practically without action on the material of the digester no lining is necessary. In filling the digester the chips and liquor are run in at the same time and in order to get in as large a charge as possible a tamping device is sometimes employed. This is especially necessary for horizontal rotaries. The time required for charg- ing a rotary holding 3 to 3I cords is about one hour, while chips from 14 to 15 cords of wood can be run into a vertical digester in 1 5 to 20 minutes. The cooking liquor for the soda process is merely a solution of caustic soda containing a small amount of sodium carbonate. It is generally made at the mill by causticizing soda ash with quick lime. The reaction on which the process depends is Na 2 C0 3 + CaO + H 2 = 2 NaOH + CaC0 3 . This is a reversible reaction and the extent to which the soda ash is causticized depends on the dilution of the liquor. Ex- periments by Lunge 1 illustrate the effect of concentration as follows: 1 Lunge: Sulphuric Acid and Alkali, 2nd ed., Vol. II, p. 750. CAUSTICIZING 99 Liquor before causticizing After causticizing Per cent Na2C0 3 Specific gravity Per cent causticity 2 5 IO 14 20 i .022 at 15 C. i .052 at 15 C. 1. 107 at 15 ° C. 1. 150 at 15 C. 1. 215 at 30 ° C. 99-4 99.0 97.2 94-5 90.7 It is obvious that the lower the concentration the higher the causticity and that therefore a demand for a definite concen- tration sets a limit to the causticity attainable. Attempts to increase the causticity by boiling either under diminished or increased pressure have met with no success; practically all causticizing is therefore done in tanks under atmospheric pressure. The type of equipment used for causticizing varies greatly in the different mills; it may be stated to consist generally of a wrought iron tank fitted with an agitator and frequently a perforated basket near the top for holding the lime. The tank is filled with soda ash solution until the bottom of the basket is just covered; it is then brought to a boil and the lime added to .the basket very gradually. As the lime slakes it passes through the perforations while stones or unburned cores are retained and may be easily removed. If the tank is not fitted with a basket the lime is thrown directly into the hot soda ash solution and slaked in the bottom of the tank. It has been found best to boil with a steam coil rather than by blowing steam directly into the liquor. Agitation with compressed air is also not to be recommended as it reduces the causticity slightly. The time of boiling has a considerable influence on the causticity obtained as the latter increases with the duration of the boil. The size of the plant in relation to the necessary output frequently limits the time of boiling but if possible it should be continued for at least an hour; beyond this point it is doubtful if the gain in causticity will pay for the extra expense. IOO THE SODA PROCESS Greater causticity can also be obtained in many cases by in- creasing the agitation. This has been done in certain instances, where adding more wings to the agitator shaft, or increasing its speed, has enabled less lime to be used in obtaining the same amount of caustic soda. Tests on a small scale have proved that thorough agitation is to a large extent equivalent to boil- ing, and that if the agitation is complete enough the charge need not be boiled at all as a temperature of 85 degs. is ample. The causticizing operation is generally carried out in about the following manner. The soda ash and lime are boiled and then allowed to settle in the same tank. The clear liquor is drawn off by an adjustable syphon, more soda ash and water are added and the sludge again boiled up and then pumped over into a second tank. The sludge from this is flooded with weak liquor and again boiled up and settled. Fourth and fifth boilings are made with clear water. The clear liquors from the first four boilings combined make the cooking liquor, while the fifth boiling produces the weak liquor for a subsequent third boil. The Dorr Company have recently proposed to causticize and wash the lime mud in a continuous operation by machinery similar to that used in metallurgical work. The lime is crushed and continuously mixed in definite proportions with a soda ash solution; the mixture then passes through three reaction agitators which are furnished with steam coils and then goes to the first thickener. The clear liquor overflowing from this goes to the storage tank for cooking liquor, while the sludge is pumped to the second thickener where it is mixed with the overflow from a third thickener. The clear liquor from the second thickener flows to the reaction agitators, while the sludge goes to the third thickener and thence to waste or to a recovery plant for lime. A causticizing plant of this type is shown in plan in Fig. 9. A plant operated in this manner has been in operation for some time and is said to be giving good satisfac- tion. It seems doubtful if it is very much superior to a carefully supervised plant of the ordinary type. < Ph O g u u O => o PM (101) 102 THE SODA PROCESS The lime mud produced in causticizing is generally a waste product, though there is sometimes a small local demand for agricultural purposes. Attempts have been made to reburn the mud and use the lime over again and a number of plants are now operating with rotary kilns similar to those used in the cement industry. The lime mud is freed from water as much as possible by mechanical means and then enters the kiln, through which it passes in a direction opposite to the combus- tion gases. It is first dried and then heated to such a tempera- ture that the carbon dioxide is driven off and the material delivered as burned lime. A kiln 7 ft. in diameter and 120 ft. long will burn 35 to 40 tons of lime per day, while for capacities between 20 and 30 tons a kiln 6 ft. by 100 ft. is sufficient. If the mud enters at 55 per cent dry, the fuel requirements will be about 9500 cu. ft. of natural gas or 675 lbs. of coal per ton of lime burned. In order to keep the impurities down to a reasonable figure it has been found necessary to remove about 10 per cent from the circuit regularly. Where producer gas is used one passage through the kiln adds three pounds of impurities for every 100 lbs. of quick lime, but this only reduces the causticizing power of the lime 2 per cent because the impurities are held mechanically rather than chemically. When powdered coal is used instead of gas 6 per cent of impurity is added and this reduces the causticizing power of the reclaimed lime by fully 18 per cent. The lime mud and the reclaimed lime from a gas- fired kiln contain the following impurities for every 100 lbs. of available calcium oxide; as may be seen nearly all of the im- purities added in reclaiming come from the fire brick lining. Magnesia, MgO Oxides of iron and alumina, Fe203 and AI2O3. Sodium oxide, Na 2 Sulphur trioxide, SO3 Silica, Si02 Lime mud Reclaimed lime Per cent Per cent 1-3 0.9 1.6 1 .1 2.6 1 .1 0.2 0.7 Q-7 4-7 2.2 7^7 COOKING 103 The following analyses show the composition of the new lime and the recovered lime from a plant using natural gas as a fuel. Calcium carbonate, CaCOs Iron and alumina, Fe 2 Os and AI2O3 Silica, Si0 2 Calcium oxide, CaO. . . Magnesia, MgO. . . ; Undetermined New lime Recovered lime Per cent Per cent 4.48 0.15 3-91 1 .62 0.16 92.00 2.62 0.59 0.70 89.58 1.96 2.23 The recovered lime is in the form of rounded nodules ranging up to the size of a hen's egg. It is often slightly greenish or yellowish in color and slakes rather more slowly than good lime. The strength of the caustic liquor used in cooking varies from 8° to 15 Be. at 6o° F. according to operating conditions. Stationary digesters require more dilute solutions than rotaries, while wet wood necessitates increasing the strength of solution to counterbalance the water contained in the chips. If digesters are heated by jackets or closed coils, weaker liquors may be used because they are not diluted by any condensed steam. With direct heat in rotaries about 700 to 900 gals, of liquor are used per cord of wood while in digesters the liquor amounts to about 800 to 1 100 gals, per cord. The boiling operation is a very simple one, the object being to reach full pressure as soon as possible and maintain it to the end of the cook. During this period the air which collects in the top of the digester is blown off several times through the " relief pipe" so that no false pressure may be recorded on the gauges. This " relief " is usually not necessary when cooking in rotary digesters. The uniformity of the cook depends on good circulation of the liquor and in practice this is obtained in several ways. Some digesters are fitted with internal circu- lating pipes on the same principle as the vomiting pipes in rag boilers, while another very successful and positive method of io4 THE SODA PROCESS circulating is to take the liquor from below the false bottom and pump it up into the top of the digester; as both sides of the pump are under the same pressure very little power is required. The steam consumed in cooking depends on the form of the digester, on whether it is covered with an insulating covering or not, and on the size of the charge. Steaming may be con- sidered as taking place in two stages, the period in which the charge is being brought up to pressure, during which the de- mand for steam is very great, and the period at full pressure when only enough steam is required to make up for the heat lost by radiation. Records obtained with steam flow meters on three sizes of digester gave the following results in pounds of steam required. While coming up to pressure (lbs. per hr.) During period at pressure (lbs. per hr.) . Total required during cook (lbs.) 3 -cord rotary 7,500-7,800 500 12,125 6J-cord rotary 10,800 1,050 19,360 15-cord vertical digester 20,000-22,100 1,250* 37.750-44-250 * Calculated by radiation formula for steam pipes. In spite of the apparent simplicity of the process there are a number of factors which greatly influence the results, and uni- form and satisfactory cooking depends on the proper adjust- ment of these variables. Much study has been given to these factors by the author and at about the same time by the Forest Products Laboratory who published a bulletin 1 on their re- sults. The following discussion of the individual variables is based on the author's experiments which were very carefully made in special apparatus which enabled very close control of all conditions to be maintained. The experiments were all made with poplar chips and all yields are figured as bone dry fibre on bone dry wood. Fig. 10 illustrates the effect of changes in steam pressure upon the yield and the bleaching properties of the fibre when all other cooking conditions are kept constant. 1 U. S. Dept. of Agriculture, Bulletin No. 80. *T STEAM PRESSURE i°5 In this and the succeeding charts the fibres were all bleached to a standard color so that the figures are directly comparable. It is at once evident that the yield is greatly influenced by the steam pressure employed and that the decrease in yield is at a nearly constant rate for pressures between 70 and 130 lbs. Throughout this range increasing the steam pressure 10 lbs. Per cent Bleach 6 7 8 9 130 CO O On g CO CD 3 1. 17 Sodium sulphide, Na 2 S o. 37 Magnesia, MgO o. 34 Iron and alumina, Fe 2 C>3 and Al 2 03 o. 26 Silica, Si0 2 o. 17 Calcium sulphate, CaSCu o. 07 Carbon by difference, C 89. 05 No use for this material, which will even begin to take care of the amount made, has ever been developed. The most promising field seems to be as a fuel, since the heating value of the dried waste is 14,000 to 14,500 B.T.U. Losses. The chief loss to which attention should be paid is that of soda. This occurs at numerous points, all of which should be subjected to careful scrutiny. The loss in the lime mud should be checked by routine analy- ses. These will indicate whether the process is being run care- 136 THE SODA PROCESS fully and may prove that a different lime can be used which will permit better settling and cleaner washing. The loss in the waste lime sludge will be about 2.0 per cent of its dry weight under average conditions. The loss in washing out the black liquor from the fibre may be studied by analyses and volume measurements of the final washings. Such studies indicate that under normal conditions this loss will amount to 1 to 1.5 per cent of the total soda used in the digesters. There is also a loss due to the presence of a small amount of soda in the washed fibre; this has been found to be about 1.5 per cent of the soda used. Under abnormal conditions, as when the recovery plant cannot keep up with the wash pits, there is likely to be a much greater loss in the final wash water. The losses in burning the black ash are hard to determine because the process is a continuous one and reliable measure- ments of the materials going to and coming from the furnaces are seldom made. Moreover the loss is chiefly in material carried up the stack mechanically or because of a slight vola- tilization of soda. Spence * has studied the loss up the flues and gives the soda lost per twenty-four hours as follows for different sizes of incinerators. Lbs. 14-foot rotaries 450-1000 16-foot rotaries 650-1300 30-foot rotaries 2500-4000 In the leaching of the black ash there is a loss of soda which may run from a few tenths of a per cent up to four or five per cent for unsatisfactory conditions. While it would seem an easy matter to determine this loss it has proved to be a very difficult proposition because the nature of the material and the way it is handled make it almost impossible to obtain a fair sample. Tests and Analyses for the Soda Process. The analytical work necessary for the control of the soda process is of a com- paratively simple nature. The materials which should be tested 1 Spence: Paper, 1919, 619. TESTS AND ANALYSES 137 either regularly or at intervals are the cooking liquor and black liquor, soda ash, lime, lime mud, black ash and black ash waste. The cooking liquor contains caustic soda as the essential in- gredient, but it also contains soda ash and in some cases salt, which renders its valuation by a simple hydrometer test some- what misleading. The actual grams per liter of caustic soda should be determined by titrating 10 c.c. with normal acid using first phenolphthalein and finally methyl orange as indi- cator. The methyl orange shows the total alkali while the phenolphthalein shows all the caustic soda and half the car- bonate. Twice the difference between the two, subtracted from the methyl orange reading, gives the acid to neutralize the caustic soda and this multiplied by 4 gives the grams per liter of NaOH. The causticity of the cooking liquor should also be recorded; this is obtained by dividing the cubic centi- meters of acid required for the caustic soda by the volume necessary for the total alkali. Probably most practical mill men would say that the Baume test is the only one necessary but for the reasons given above this test alone may lead to entirely incorrect conclusions. Soda ash is ordinarily one of the purest of commercial chemi- cals yet it is desirable to test it occasionally when received and if it is stored in bulk it is sometimes necessary to determine the amount of moisture which it has taken up. For moisture a representative sample of 1 to 2 grams is accurately weighed into a platinum crucible and dried for three-quarters of an hour over a gas flame which is so adjusted that the bottom of the crucible just shows a faint redness when shielded from strong- light. After cooling in a desiccator it is quickly reweighed, the loss being calculated as percentage of moisture. For the de- termination of alkali present the dried sample is dissolved in water and titrated with normal acid using methyl orange as indicator. The percentage of sodium carbonate in the original sample is calculated by the following formula C. C. acid X 0-053 Weight of sample before drying 138 THE SODA PROCESS The lime used in causticizing should be regularly tested in order to see that its quality is kept up to a reasonable standard. A chemical analysis will not necessarily show what results it will give in practice but a simple causticizing test made under conditions similar to those of actual work will give very valu- able information. From each car of lime received as fair a sample as possible should be taken by going all over the car and taking portions from the top, middle and bottom of the load. This should be selected to represent both the fine and the coarse material and as soon as the entire sample is taken it should be crushed and quartered down as rapidly as possible to avoid the absorption of moisture. From this final sample a weighed amount is taken and boiled for exactly one hour with water and a weighed amount of dry soda ash which is in excess of the amount the lime can causticize. The amount of water used is so taken that the final solution at the end of the test is about the strength of that used in practice. After the boiling is completed the sludge is allowed to settle and the clear liquor is titrated with both phenolphthalein and methyl orange as in the case of cooking liquor. Knowing the causticity from this titration and the weight of soda ash taken the amount of lime required to causticize 100 lbs. of dry soda ash may be calculated by this formula: Weight of lime used X 100 Per cent causticity X weight of soda ash used The settling quality of the lime may be ascertained in this same test by taking a sample of the rapidly boiling mixture just at the end of the test and, without giving it time to settle, filling a 100 c.c. graduate just to the upper mark. By noting the cubic centimeters of clear liquor at fixed time intervals the relative settling qualities of the various limes can be compared. The lime mud which settles in the causticizing and washing tanks is generally a waste product and in order to see that too much caustic socla is not thrown away it should be tested at intervals for the amount of alkali present. A representative sample of the dried mud is weighed out, placed in a small por- TESTS AND ANALYSES 1 39 celain dish and moistened with strong ammonium carbonate solution. It is next evaporated to dryness, heated over a low flame until no odor of ammonia can be noticed, and finally leached out repeatedly with boiling distilled water until all the soluble alkali is removed. The combined leachings are then titrated with standard acid, using methyl orange as indicator, and the results calculated to precentage of alkali based on the dry lime mud. In the case of black liquor it is at times desirable to know the strength both in total alkali and in free caustic soda. For total alkali evaporate a measured volume to dryness in a plati- num dish and ignite over a flame till the organic matter is com- pletely carbonized. Cool, extract several times with hot water and pour the extracts through a small platinum cone. Put the cone in the dish together with the residual wet carbon, cover with a filter paper which just fits inside the dish and quickly ignite over a gas flame. The filter paper prevents loss by spattering and enables the carbon to be burned off without waiting for it to be dried first. Again cool and extract with hot water, add the extract to the first and titrate with acid in the presence of methyl orange. Calculate the results to grams of sodium carbonate per liter or pounds per gallon. For the determination of free caustic soda in black liquor add 25 c.c. of the latter to 400 c.c. of water and 15 c.c. of barium chloride solution (400 grams per liter) in a beaker. Titrate directly with standard acid using a dilute solution of phenol- phthalein on a spot plate as an indicator. The acid should be added quite slowly and the end point may be considered as that point at which no pink color develops within two minutes after mixing one or two drops of the liquid with the indicator. Owing to the presence of soluble coloring matter and to the precipitate thrown down by the barium chloride the end point is not very sharp but with a little practice it is fairly easy to get concordant results. A sharper end point is obtained by filtering off or settling out the barium precipitate and using only the clear liquor, but this introduces a distinct error due 140 THE SODA PROCESS to the loss of the caustic soda occluded by the precipitate. In the method outlined above this is kept within the sphere of action of the acid and more accurate results are obtained. Analyses of black ash and black ash waste are usually con- fined to simple determinations of the amount of soluble alkali present. In the case of black ash this may be leached out with hot water and titrated as usual in the presence of methyl orange. With black ash waste the volume of carbon is relatively so large that it is well to burn off most of it in a platinum dish before attempting to leach out the soda. CHAPTER V THE SULPHATE PROCESS The sulphate process is similar to- the soda process in that the cooking liquor is alkaline, but it differs from it by replacing the alkali lost with sodium sulphate instead of soda ash. In, rare cases both are used but the sulphate is always in the greater amount and it is from the use of this material that the process gets its name. Actually it is a misnomer, and it would be better to call it the sulphide process because of the important part played by sodium sulphide in the cooking liquor. This sulphide is derived from the sulphate by reduction in the recovery of the alkali and it is this feature which introduces the greatest deviation from the soda process. There is more or less confusion in the use of the terms " sul- phate" and "kraft" as applied to the process and products. Sulphate may be considered as a general term applying to any cooking process in which the loss of alkali is made up by adding sodium sulphate, while kraft is that subdivision of the sulphate process in which the pulp is intentionally undercooked in order to produce very strong stock. The products of the sulphate process vary according to the cooking conditions from the dark brown, unbleachable kraft fibre to a soft, easy bleaching stock. The latter may be used in making white papers where it gives a soft, pliable sheet in comparison with the harder and more rattly paper from sulphite. The principal use of the sulphate process however is in the preparation of kraft stock. The sulphate process has two chief advantages over the sul- phite process: the chemicals used can be recovered and the wood used may be of a highly resinous nature. It is this second fact which gives the process its widest application. As compared 141 142 THE SULPHATE PROCESS with the soda process it gives somewhat higher yields and employs a cheaper source of alkali. Its one disadvantage lies in the extremely disagreeable odors due to the organic sulphur com- pounds formed during cooking and recovery. This has been found serious enough to limit its use to sparsely inhabited local- ities. As the process is an alkaline one there is no appreciable action upon the digester plates and it is not necessary to use any lining. The equipment used is very similar to that for the soda process though the digesters are, as a rule, smaller, yielding from two to three tons of pulp per charge. There is a tendency in Europe towards the use of digesters which rotate on their short axes. The sizes of these are variously reported as 1 8 to 45 cubic meters capacity (635 to 1590 cu. ft.). The digesters should always be welded and not riveted, as the latter are bound to develop leaks from the continual expansion and contraction to which they are subjected. The cooking process employed depends on whether kraft fibre or easy bleaching stock is to be produced. For kraft fibre in which a dark color is desired a portion of the cooking liquor is made up of black liquor containing practically no caustic soda. The proportion of this black liquor, as might be expected, varies considerably, some mills using as little as 27 per cent while in others it amounts to nearly 60 per cent of the total liquor. The volume of liquor used is about 45 to 50 per cent of the capacity of the digester and it tests 8° to 12 Be. at 6o° C. if indirect heat- ing is employed or 18 to 23 Be. when steam is blown into the charge. Beveridge 1 states that the total volume of liquor should not be less than 150 cu. ft. per 2000 lbs. of air-dried pulp produced, or per 1.60 cords of wood. The relative volumes of white and black liquor would then depend on the strength of the former as delivered by the causticizing system, any excess over the volume of white liquor necessary to give the required amount of alkali being made up by black liquor. The alkali required in making kraft fibre is said to be 640 lbs. per ton of pulp. 1 Beveridge: Paper, 1918, 22, 21. RELATIVE VALUE OF ALKALIS J 43 The cooking time is from i| to 6 hours and the steam pressure employed varies from no lbs. to 135 lbs., the highest pressure being held for only i| to 2 hours. In practice it is seldom neces- sary to cook at more than no lbs. pressure. For easy bleaching fibre no black liquor is used and if the cooking liquor has to be diluted water only is employed. The other cooking conditions are about the same as they are for kraft fibre. A modification of the usual cooking process is that proposed by Ungerer 1 in which the cooking liquor is passed through a series of digesters until exhausted. The original installations were of small digesters and the process has never been very extensively used. The chief trouble seems to have been in keep- ing the liquor heaters in repair, while the advantages claimed were more uniform, stronger and easier bleaching fibre. The fol- lowing figures 2 show the strength of the various constituents in a definite quantity of the liquor as it passes the digesters in series. Digester Total soda as Na 2 Soda combined with organic matter Soda as NaOH Soda as Na2C0 3 2 3 4 5 6 7 874 729 657 611 605 598 251 543 579 552 558 582 298 31 22 9 3 325 155 56 5° 44 16 From experiments in the Forest Products Laboratory 3 it has been proved that increasing the amount of either the caustic soda or the sodium sulphide decreases the yield and that the former has about twice as much influence as the latter. The carbonate and sulphate of sodium are apparently without effect on the wood. This has been confirmed by observations under •actual working conditions and Beveridge 4 expresses the opinion 1 Ungerer: Papier Ztg., 22 (94), 3360. 2 Knosel: Papier Ztg., 22 (97), 3470. 3 Wells: Paper, Sept. 24, 1913, p. 15. 4 Beveridge: Paper, 19 18, 22, 21. 144 THE SULPHATE PROCESS that the caustic soda unites with the organic matter first and when exhausted, or nearly so, the sulphide comes into play. The composition of sulphate cooking liquor, according to several different authorities, is given below in tabular form, the quantities of the various substances being expressed as grams per liter. Authority Na 2 C0 3 NaOH Na 2 S Na 2 S0 3 Na 2 S0 4 Deg. Be M. Muller * Schacht * Schacht * Heuser f Klein J 24.00 36.00 45-05 7.48 15.00 10.00 45.00 80.60 77.80 61.80 62 .00 55- 00 13.00 I3-SO 11.25 25.12 22 .00 30.00 2.00 7-25 8.19 3-78 3.00 3-00 14.00 15.10 12.18 4-52 5.00 i 7 '.8 18.5 * Kirchner: Das Papier, p. 109. t Heuser: Papier Ztg., 1910, p. 1511. t Klein: Papier-Fabr., 1914, p. 628. From the investigations of E. Heuser 1 on beechwood it appears that comparatively little sodium sulphide is used up during the process. The following table shows the strength of the cooking liquor used in three of his cooks and the percentages of the indi- vidual constituents consumed. Na2S added in grams per liter NaOH added in grams per liter. . . . Na2C03 added in grams per liter. . . Percentage consumption of Na 2 S. . . Percentage consumption of NaOH . Percentage consumption of Na2COs I 2 19-50 24.70 27-75 48.55 18.00 22 .00 7-95 15.60 69.80 51.20 39.00 6.40 21 .40 28.20 9-9° 7 .00 63-80 2.80* * Percentage Increase In all three cooks the Na 2 S0 4 present increased very slightly during the cook. The materials necessary to produce one ton (2000 lbs.) of sulphate pulp vary quite widely in different mills; the following seem to be the limiting values: 1 Heuser: Wochbl. Papier-Fabr., 1913, p. 2209. YIELD OF FIBRE 145 Wood 177-247 cu. ft. Coal S50-770 lbs. Sodium sulphate 320-395 lbs. Lime 530-660 lbs. The washing, screening and bleaching of sulphate pulp differ in no essential detail from the treatment accorded soda pulp. It is most general to use diffusers to wash the black liquor from sul- phate pulp but there is no reason why open pans could not be used and there is considerable difference of opinion as to the rela- tive advantages of the two methods. It is claimed that the size of the plant influences the choice, diffusers being satisfactory for small installations and open pans for large ones. As it is desirable to have five or six diffusers to each digester the cost of installation is high in plants of large capacity. Moreover their life is com- paratively short, — about ten years, — due to the pitting and eat- ing away of the steel. The yield of fibre by the sulphate process, as already stated, is greater than by the soda process. According to Kirchner 1 the yields for spruce (Fichte) and fir (Kiefer) are: Soda Sulphate Sulphite Spruce Per cent 29.7-32.8 28.0-29.3 Per cent 32.8-35.9 29.3-32.0 Per cent 37-50 Fir Experiments by the author gave the following yields per cord, assuming that there are 100 cu. ft. of solid wood per cord: Soda Sulphate Poplar (Populus sp ) Pounds 1007 699 790 786 Pounds 1 130 773 862 White pine (Pinus strobus) Pitch pine (Pinus rieida) Spruce (Picea sp ) 956 1 Kirchner: Das Papier, IV, 358. 146 THE SULPHATE PROCESS In addition to the greater yield the sulphate fibre bleaches considerably easier than soda fibre from the same wood. The blow-off gases from sulphate cooks have been the subject of much investigation in Sweden and Germany. Bergstrom and Fagerlind 1 have found in them methyl mercaptan, dimethyl sulphide, dimethyl disulphide, methyl alcohol, ammonia, turpen- tine, rosin oil, hydrogen sulphide, ammonium sulphide, and acetone. The evil odors of the process are due in large part to the first two compounds of which the mercaptan is far the worse. According to Klason and Segerfeld 2 about 100 grams of mer- captan are produced per ton of wood treated in making easy bleaching pulp while ten times as much may be obtained in kraft cooks. Pine yields about twice as much as spruce. According to Falk 3 the condensed materials obtained per ton of cellulose from pine wood are as follows: In oily portion In aqueous portion Mercaptan Dimethyl sulphide . . Dimethyl disulphide. Turpentine Distillation residues . Methyl alcohol Ammonia Kgs. 0.062 0.927 0.103 8.487 0.721 Kgs. 0.06 0.17 0.05 0.92 5 -oo 0.18 From work done at the Billingsfors mill the condensed steam from the digesters was found to yield the following quantities for every ton of finished pulp made. Lbs. Turpentine 17.6 (from fir) Turpentine 2.2 (from pine) Methyl alcohol 1 1 . o Methyl mercaptan 2. 2 Methyl sulphide 6. 6 Methyl bisulphide 0.2 Ammonia 0.4 1 Papier-Fabr., 1909, 7, 7, 27, 78, 104, 129. 2 Papier-Fabr., 1911, 9, 1093-1099. 3 Papier-Fabr., 1909, 7, 469-472. BLACK LIQUOR 147 Bergstrom 1 states that in 191 2 five sulphate mills were recover- ing methyl alcohol. Pine and spruce yield about the same amount, 15 kgs. per 1,000 kgs. of cellulose, and of this 5 kgs. are collected in the condensed vapors while further amounts may be obtained from the vapors formed during evaporation of the black liquors. The cost of such recovery is slight and the process in no way interferes with regular operations of the mill. Klason 2 has carried out extended investigations of the sulphate process in the attempt to eliminate the odors. He found that the gases could be almost entirely freed from mercaptan by passing through solutions of various metallic salts but that the only metallic mercaptides which were completely odorless were those of the noble metals. Caustic soda will also absorb mercaptan but not methyl sulphide. The gases, separated from entrained liquid, can be rendered harmless by passing them under a furnace grate but explosions must be guarded against. Oxidizing agents, as bleach or permanganate, will destroy the odors but they also oxidize the alcohol and the quantity necessary is therefore exces- sive. The black liquor from the sulphate process has been inves- tigated repeatedly. According to Klason and Segerfeld 3 of the organic matter present 54.3 per cent is lignin; 2.5 per cent fatty and resin acids; 3.7 per cent formic acid; 5.2 per cent acetic acid and 30.3 per cent lactonic acids. Of the sulphur originally present as alkali sulphide 51.8 per cent was combined with lignin, 15 per cent expelled as volatile compounds, 16.8 per cent remained as alkali sulphide and 16.4 per cent was unac- counted for. The committee appointed by the Finnish Government 4 in 1908 gives the following composition for a black liquor with a specific gravity of 1.140 from a Swedish mill. 1 Papier-Fabr., 1912, 10, 677. 2 Papier Ztg., 1908, 33, 3577. 3 Papier-Fabr., 1911, 9, 1093-1099. 4 Papier Ztg., 1910, 35, 2744. 148 THE SULPHATE PROCESS Per cent Sodium carbonate (Na2COa) 2. 75 Sodium hydroxide (NaOH) o. 45 Sodium sulphide (Na 2 S) 1.76 Sodium sulphate (Na 2 S04) 1.21 Sodium sulphite (Na 2 S03) 0.16 Sodium chloride (NaCl) o. 17 Sodium thiosulphate (Na2S 2 03) o. 14 Soda combined with organic acids 2. 25 Organic acids n. 71 Water, etc 79- 40 100. 00 An interesting proposal for handling the black liquor is that of Rinman. 1 He concentrates the liquor to about 90 grams per liter of Na 2 0, adds salt (NaCl) to a strength of about 40 grams per liter and precipitates the humus matters with carbon dioxide at a temperature of 75 C. The humus matters may be washed with water, provided no sodium sulphide is present, and may then be subjected to destructive distillation. The solution, freed from humus, may be treated by the ammonia-soda process for the recovery of the alkali as bicarbonate. The original plan, to recausticize and use it over again several times before sending it to the reclaiming system, has not proved practical because of the accumulation of substances not precipitated by carbon dioxide. As an alternative to this process Rinman suggested mixing the black liquor, after evaporation, with lime or caustic soda and distilling destructively in presence of steam at a temperature of 400 C. This would result in the formation of fuel gases, acetone, alcohol, ketones, hydrocarbons, cresols, etc., and the residual carbon would contain the alkali which could be leached out and recovered. An experimental plant at the Stora Kopparsberg mill gave the following yields per ton of pulp made from pine or fir: Lbs. Pure acetone 39. 6 Motor spirit 59-4 Motor oil 121. o 1 Rinman: Papier-Fabr., 191 2, 10, 39 and 101. SODA RECOVERY 149 These proposed methods are rather complicated and have not been generally adopted, much the greater proportion of the alkali being recovered by the usual methods. These consist of evaporation by any of the well known multiple effect evaporators, by disc evaporators or combinations of the two. When both are used the multiple effect takes the weak liquor and brings it up to about 2o°'Be., then it passes to the disc evaporator which concen- trates it to about 35 Be. According to Beveridge x in a well equipped mill the water which the recovery process must handle per 2000 lbs. of pulp is derived from the following sources: Cubic feet Per cent Water from chips 47.6 42.48 67-33 3I-50 25 .1 Water from steam 22 .5 Water from white liquor 35 .6 Water used for washing 16.6 188.91 After concentration to about 35 Be. the liquor, which contains approximately equal weights of water and total solids, goes to the rotary furnaces. These are similar to the furnaces used in soda mills but are sometimes as long as 30 to 35 ft. As the liquor works forward in the furnace it becomes more concentrated and finally burns to a moist, black mass still containing a considerable amount of organic matter and water. In this form it is dis- charged in close proximity to the smelting furnace into which it is to be fed. During the passage of the liquor through the furnace rings are sometimes formed which hold it back while allowing the soda near the discharge end to become melted by the heat from the smelting furnace. When the ring finally breaks and the liquor runs into the melted mass explosions are very apt to occur. After being discharged from the rotary furnace the mass is mixed with enough sodium sulphate to replace the alkali lost and is then shoveled by hand into the smelting furnace. This is a chamber about four feet square by seven feet high, lined with soapstone to which the molten alkali does not adhere, and so 1 Beveridge: Paper, 1918, 22, 349. IS© THE SULPHATE PROCESS constructed that oxidation of the sodium sulphide is reduced to a minimum. Combustion in the smelting chamber is provided for by an air blast entering through tubes which are water-cooled or sometimes made of platinum. The condition of the entering ash and the blast must be most carefully regulated to insure proper results. Moreover the furnace must be watched closely to see that the blast nozzles do not burn off, due to a temporary stoppage of the water, or that the furnace does not become clogged and allow a considerable quantity of the melt to collect. Either of these conditions is likely to result in severe explosions. The reaction which takes place in the smelter is Na 2 S0 4 + 4C = Na 2 S + 4CO. This is to a certain extent reversible as the oxygen supplied in the air blast oxidizes a small amount of the sulphide back to sulphate. For this reason it is not possible to produce a liquor entirely free from sulphate, the proportion depending on con- ditions in the smelter. The melt, if rich in sulphide, is of a red- dish color. The composition of various samples is as follows: 1 Authority M. Muller W. Schacht. ........ Klason and Segerfelt Kgs. sul- phate per 100 kgs. melt 23-7 11 .0 8-10 8-10 20-22 20-22 20-22 Composition of melt in per cents Na 2 CO ; 56.60 71.40 80.26 74.20 59-42 62 .07 68.37 61.73 NaOH 0.40 O.50 I .04 1 .60 0.20 2 .20 3-5° Na 2 S 22 .60 II .60 7-i5 9-5° 14.00 17-75 13-75 21.50 Na 2 S0 3 7-33 NazSO, Insol- uble * Papier-Fabr., ign, 9, 1093-1099. Beveridge 2 gives the relationship between the melt, its solu- tion, and the causticized liquor as follows: 1 Kirchner: Das Papier, 105, 279. 2 Beveridge: Paper, 1918, 22, 21. *► REDUCING ODORS 151 Melt, per cent Solution of melt, grams per liter Causticized liquor, grams per liter Na 2 C0 3 Na 2 S Na 2 S0 4 NaOH Si0 2 Insol. in water. 62 .44 26.96 4.26 2-3S 6.18 212.85 104.17 iS-84 13-8 28.7 7.92 From the smelting furnace the melt flows in a red hot con- dition to dissolving tanks containing water or the dilute washings from the lime sludge in the causticizing room. When the solu- tion has reached the desired density it is discharged into the causticizing system. This solution generally has a greenish tint from which it is termed " green liquor " to distinguish it from the " white liquor " obtained after causticizing. The dissolving tanks must be cleaned out occasionally because a certain amount of mud collects in them; this mud should be used in the caus- ticizing system in order to utilize the soda which it contains. The gases from the recovery furnaces have been analysed by Klason x and found to contain 13 mgs. of methyl mercaptan and 4 mgs. of hydrogen sulphide per cubic meter. Careful blowing off from the digester reduces this by about one-third by removing more from the black liquor. The gases from the calciners can- not be washed commercially on account of the great size of the scrubbers required and the slight solubility of the impurities in water. On account of their great dilution only a small part could be burned by passing under the grates of the boilers. In order to reduce the formation of bad odors to a minimum the digesters should be blown off until the condensed distillate has no odor, the black liquor should be evaporated to such a point that it will not cool the fires enough to cause distillation or incomplete combustion, and the calciner should be given plenty of air. Lead acetate paper may be used as a test in controlling the running of the calciner. 1 Klason: Papier Ztg., 1908, 33, 3619. 152 THE SULPHATE PROCESS The methods for the analysis of the liquors and the recovered ash in the sulphate process are more complicated than for similar products in the soda process. The following procedure for white liquor has been worked out in the Forest Products Laboratory l and found to be accurate except in the presence of polysulphide. (a) Total Alkali. Two c.c. of the liquor are titrated with half normal acid using methyl orange as. an indicator. This gives the acid equivalent to the Na 2 C0 3 , NaOH, Na 2 S and \ Na 2 S0 3 . (b) NaOH and Na 2 S. To 2 c.c. of the solution contained in a 100 c.c. flask add 20 c.c. of a 10 per cent solution of BaCl 2 and make up to the mark with boiling distilled water; shake for a few minutes and allow to settle; cool and draw off 50 c.c. of the clear liquid and titrate with half normal acid, using methyl orange as indicator. (c) Sodium sulphide + Na 2 S 2 3 + Na 2 S0 3 . Find by trial the approximate amount of standard iodine necessary to react with 2 c.c. of the liquor. Using about half a c.c. less than this amount of iodine in 200 c.c. of water add 2 c.c. of the liquor, acidify with acetic acid and complete the titration with iodine using starch as an indicator. This shows the iodine equivalent to the Na 2 S, Na 2 S 2 3 and Na 2 S0 3 . id) Na 2 S 2 3 + Na 2 S0 3 . To 5 c.c. of the solution in a 250 c.c. graduated flask add an excess of an alkaline solution of zinc chloride; make up to the mark, shake for a few minutes, and allow to settle; draw off 50 c.c. of the clear solution, and neutralize with sulphuric acid using methyl orange as an indicator. Titrate this solution with tenth normal iodine using starch as indicator; decolorize by add- ing one drop of sodium thiosulphate solution and titrate to neu- trality with tenth normal sodium hydroxide. The number of c.c. 1 Paper, Feb. 23, 19 16, p. 30. METHODS OF ANALYSIS I53 multiplied by 0.0042 gives the amount of Na 2 S0 3 in the sample and this figure divided by 0.0063 gives the iodine value of the sodium sulphite. Subtract this from the iodine titration previ- ously obtained, which will give the iodine equivalent to the so- dium thiosulphate present. Calculations. c -d gives the c.c« of iodine for sodium sulphide. a - b gives the c.c. for Na 2 C0 3 and § Na 2 S0 3 . The titration in (b) expressed as Na 2 0, minus the sodium sulphide as Na 2 0, gives the Na 2 equivalent to the NaOH present. This same procedure may be employed for the examination of the recovered ash in which case 50 grams of the sample are dis- solved in about 400 c.c. of freshly distilled water and after shaking repeatedly for two hours made up to 500 c.c. This does not include the determination of Na 2 S0 4 which may be accomplished by acidulating a measured volume of the solu- tion with HC1, boiling till all H 2 S is driven off and then pre- cipitating the sulphate with BaCl 2 . The BaS0 4 is filtered off, washed, ignited and weighed and from this weight that of the Na 2 S0 4 may be calculated. For a quick method of analysis for the control of the cooking liquor Heuser 1 recommends the following: Dilute 10 c.c. of the liquor to 100 c.c. (1) Treat 10 c.c. of the dilute liquor with an excess of — iodine 10 solution and determine the excess with — thiosulphate using starch as an indicator. Calculate to Na 2 S. (2) Heat 10 c.c. of the dilute liquor with an excess of — Ha 2 S0 4 10 till the odor of H 2 S is gone and then titrate back with — Ba(OH) 2 10 using phenolphthalein as indicator. This gives Na 2 S, NaOH and Na 2 C0 3 . 1 Private communication. 154 THE SULPHATE PROCESS (3) Treat 10 c.c. of the dilute liquor with 10 per cent BaCl 2 N solution and titrate with — H 2 S0 4 . This gives NaOH and Na 2 S. 10 Calculations. (3) - (1) = NaOH. (2) - (3) = Na 2 C0 3 . (1) = Na 2 S. Moe l uses a silver nitrate solution containing 87.89 gms. AgN0 3 per liter for determining sulphides. When 1 c.c. is titrated the c.c. of AgN0 3 used represent pounds of Na 2 S per cubic foot of liquor. He gets the end point directly in the solu- tion by shaking to coagulate the silver sulphide, and noting the point at which no further precipitate is formed by a drop of the silver nitrate. Oliver 2 works in a somewhat similar way but uses an ammoni- acal silver nitrate solution. The liquor to be tested is filtered, made ammoniacal and boiled. The end point is determined by filtering, adding more silver nitrate and repeating till a drop causes only a slight opacity. The examination of black liquor for total alkali may be per- formed exactly as in the case of liquor from the soda process. The method for caustic soda, given under the soda process, if applied to sulphate black liquor will show both the caustic soda and the sodium sulphide present so that a correction for the latter is necessary. For the determination of sodium sulphide in black liquor the Forest Products Laboratory recommends the following method: Prepare a standard zinc solution by dissolving 16.746 grams of pure zinc in a small excess of nitric acid and adding ammonia till the precipitate formed is completely redissolved. The solu- tion is diluted to 2000 c.c. and enough ammonia must be present to keep the zinc from precipitating at this dilution. As indicator use a solution of nickel ammonium sulphate made alkaline with ammonia. 1 Paper, Aug. 12, 1914, p. 19. 2 Paper, July 22, 1914, p. 20. METHODS OF ANALYSIS 155 For the determination 50 c.c. of black liquor are diluted to 1000 c.c. and 20 c.c. taken for the test. This is diluted to about 100 c.c. and the zinc solution run in from, a burette. The end point is determined by noting whether a precipitate of black NiS is formed when a little of the solution being titrated is added to three drops of the ammoniacal NiS0 4 solution on a spot plate. Each c.c. of the zinc solution equals 0.010 grams Na 2 S. CHAPTER VI THE SULPHITE PROCESS The first patent relating to the use of sulphurous acid in pre- paring pulp from wood was that granted to B. C. Tilghman in 1867. 1 This was followed in 1869 by a supplementary patent covering the treatment of fibrous materials at ordinary pressures, and it is upon these two patents that subsequent modifications of the process are based. The specifications of the original patent show that Tilghman had worked out his process in an experimental way with great care and thoroughness and that he fully understood all its possibilities. In its practical application, however, difficulties of an engineering nature arose which proved too serious for him to overcome. After his failure the process was taken up by Ekman and Fry in Sweden, by Mitscherlich in Germany, by Francke, Graham, Ritter, Kellner and numerous other investigators and manu- facturers and the engineering difficulties were gradually over- come so that finally the process was established on a firm footing. Theory of the Sulphite Process. Probably the best explana- tion of the sulphite cooking process is that it is primarily a hydro- lytic splitting of the cellulose-lignin esters followed by secondary decompositions of the lignin. Water alone at high temperatures will hydrolyze lignocellulose rendering part soluble, but in the presence of an acid, such as sulphurous acid, the reaction pro- ceeds much more rapidly and at a lower temperature. For- tunately cellulose itself is comparatively stable under such condi- tions but that it is by no means unattacked has been shown by Tauss and others. When a substance is present which will com- bine with the products of hydrolysis and remove them from the 1 U. S. Pat. 70,483, Nov. 5, 1867. 1S6 WOOD AND ITS PREPARATION 1 57 sphere of action the reaction will proceed to the limit fixed by the constitution of the lignocellulose. Such substances are the sulphites which during the reaction form true chemical com- pounds with the aldehydic products of decomposition. The organic acids formed during the decomposition of the lignin also combine with the sulphites and thereby set free an equivalent amount of sulphurous acid. During the latter part of a cook this causes a steady increase in the amount of gas which necessi- tates " blowing off " to prevent too great pressure. According to Klason * lignin is a glucoside-like body of which a part is sugar-like or cellulose while the other part is of an aromatic nature. This contains an oxypropylene group, -CH= CH - CH 2 OH, as well as methoxyl, hydroxyl and aldehyde groups. It is closely related to coniferyl alcohol. In the sulphite process sulphurous acid is taken into the double bond = C = C II + H 2 S0 3 = I = C = CS0 2 OH and forms the calcium salt of lignin sulphonic acid, CisHjgOsSCa \. Secondary reactions may cause the formation of calcium sul- phate, either from the decomposition of the calcium bisulphite according to the reaction : 3 Ca (HS0 3 ) 2 = 3 CaS0 4 + H 2 S0 4 + S 2 + 2 H 2 or from reduction of tannin with consequent oxidation of sul- phurous acid. This latter reaction explains the difficulty in using wood rich in tannin. Wood and Its Preparation. One of the chief requisites for wood to be used in the sulphite process is freedom from any excessive amount of rosin, and of nearly equal importance is the even distribution of that which is present. If the rosin is local- ized at certain points the wood at those places will remain hard 1 A. Klein: Papier Ztg., 1906, p. 474. 158 THE SULPHITE PROCESS and will cause shives in the pulp from the portions which are easily reduced. In the sulphite process the liquor has not nearly the solvent power of the alkali used in the soda process and any bark, decayed portions or knots which go into the digester are likely to appear as dirt in the finished pulp. This applies to the light-colored inner bark if the product is to be used unbleached, since it takes on a dark color and shows in the pulp as dark fibres. For bleached pulp, however, this inner bark is harmless, since it bleaches quite as readily as the rest. The removal of the bark is accomplished practically in a num- ber of ways. Hand peeling in the woods is not so usual for spruce though it is quite general for the poplar used in the soda process. Disc barkers of various types are most commonly used; in these the log is held by hand or machinery against knives fastened to a rapidly revolving disc. These knives remove much wood as well as bark, particularly where the sticks are crooked. The amount of such loss depends on the size of the logs as well as the care with which the work is done; the sound wood lost may amount to 8 to 20 per cent of the original weight of the logs, in addition to the 8 to 10 per cent of true bark; the total loss, therefore, amounts to 16 to 30 per cent of the logs as received. The drum barker, which is a device to remove the bark by rubbing the logs against one another and the sides of the drum, saves all this loss of wood. It is, however, expensive to install, requires more power than other barkers to do the same amount of work, and in winter the water used in it has to be heated, which adds to the expense. This type of barker is suitable only where very clean pulp is not essential or where bleached pulp is to be made as it does not remove all the thin inner bark. It is best applied to wood which has been in the water two or three months as the bark is then more readily removed. Of the drum barkers the intermittent give better results than the continuous. With any system of barking the logs should be inspected on the conveyor as they go to the chipper, and any with bark remain- ing in cracks or around knots put at one side to be cleaned by WOOD AND ITS PREPARATION 159 hand. Some operators clean very little by hand but put the defective wood at one side to be cooked separately into second grade stock. This necessitates a very thorough cleaning up of all apparatus afterwards. The barked wood next goes to the chippers which should be run slowly enough to produce even chips. The length of chip Fig. 21. Pulp Wood Chipper depends on the method of cooking as well as the kind of wood. For hemlock they should be § to f inch long while spruce is chipped about as follows: For Mitscherlich unreachable i\ to if inches long. For quick cook, newsprint stock f to i inch long. For easy bleaching f to f inch long. The coarse chips are generally further reduced in size by some sort of crusher and are then screened to remove coarse pieces, which are rechipped, and sawdust, which goes to waste. The saw- dust from chipping should not amount to more than 3 per cent. The knots may be removed from the chips in several ways; one very successful process is to blow them against an inclined wire 160 THE SULPHITE PROCESS screen through which the dust and fine material passes while the good chips slide down the screen and are collected on a conveyor. By proper adjustment of the air blast the knots, being heavier than the good chips, fall short of the screen and are collected by a separate conveyor. Another method, which also makes use of the difference in weight of the good chips and knots, is that of passing the dry chips through a trough of water. The good chips float and are removed by a conveyor which passes across the surface of the tank, while the knots, being heavier than water, sink to the bottom and can be removed by scrapers. Uniformity and cleanliness of chips are essential to clean pulp and good yield and all chips cooked in the same charge should be of one kind of wood and as nearly as possible of the same age and moisture content. Much difference of opinion exists as to the relative desirability of wet or seasoned wood. Many prefer dry, well seasoned chips, claiming more rapid penetration of the liquor and better yield; others x state that green wood or that which has lain in the water for some time is reduced more easily than seasoned wood. Experiments by the author tend to confirm this latter opinion, as in certain cases woods which could not be successfully treated dry readily yielded to the acid if they were saturated with water before cooking. The capacities of wood handling machinery vary enormously under the conditions of operation but they may be said to be approximately as follows. Circular and band saws up to 50 cords per hour. Drum barkers , 1-4 cords per hour. Disc barkers 1 . 2-1 . 4 cords per hour. Chippers 10-15 cords per hour. Flat screens yV cord per sq. ft. per hour. Liquor Making. The liquor or acid used in the sulphite process is an aqueous solution of sulphurous acid in which lime, or some other base, has been dissolved; the final result is there- fore a solution of the bisulphite of the base containing an excess of sulphurous acid. In practice this is prepared in two different 1 Griffin and Little: Chemistry of Paper Making, p. 189. PREPARATION OF SULPHUR DIOXIDE l6l ways: by passing the gas through water in which the base is dissolved or suspended, or by bringing the gas into contact with comparatively large lumps of the carbonate of the base which are moistened by a continuous flow of water. The first effect of this treatment is the solution of the gas in water forming the true sulphurous acid: H 2 + S0 2 = H2SO3. This then attacks the base according to one of the following Ca (OH) a + H 2 S0 3 = CaS0 3 + 2 H 2 0. or • CaC0 3 + H 2 S0 3 = CaS0 3 + H 2 + C0 2 . If the base is soda or magnesia the sulphites stay in solution but if lime is used calcium sulphite is precipitated as fast as it is formed since it is very insoluble, one part requiring about 800 parts of cold water for its solution. In the first method, working with milk of lime, the formation of sulphite continues until all the lime is precipitated; further passage of the gas then causes this to redissolve with formation of the bisulphite. When limestone or dolomite is used, as in the second method, its surface grad- ually crusts over with the sulphite which is again brought into solution as more gas is dissolved. Both of these reactions doubt- less take place simultaneously when there is a liberal supply of gas. Of the available bases for making sulphite liquor soda has the advantage of forming a stable bisulphite and the cellulose pre- pared with such liquor is very pure and easy to bleach. Magnesia possesses these qualities, but in a lesser degree, while lime is likely to make the fibre hard and harsh from formation of calcium monosulphite. Preparation of Sulphur Dioxide. Sulphur dioxide for use in the absorption system is prepared by burning either sulphur or iron pyrites and with proper care good results can be obtained by either method. The choice of the two processes depends to a great extent on local conditions, such as t^ie relative cost of the two materials, the possibility of disposing of the spent oxide from the pyrites, the floor space available, etc. ; it is much more of a financial problem than a technical one. 162 THE SULPHITE PROCESS Sulphur burning was formerly carried on very largely in the retort type of furnaces into which the sulphur was fed through a door at one end. This caused gas of irregular composition due to the sudden rush of cold air into the furnace and through the apparatus. This type of furnace was suitable for very pure sul- phur, particularly Sicilian, but when there was a tendency for the sulphur to form an oily surface scum during burning good results were not obtained. This trouble can be overcome by Fig. 22. Rotary Sulphur Burner installing rakes which travel slowly back and forth, breaking up the scum and presenting fresh surfaces for combustion. With this device the flat burners give excellent results. Modern sulphur burners are of two quite dissimilar forms — one a horizontal rotary furnace and the other vertical and sta- tionary. In the rotary burner, illustrated in Fig. 22, the sulphur enters from the hopper at the front end and gradually works toward the back. As the burner revolves the sulphur is carried up the sides and thus presents a large and continually renewed surface for combustion. A highly successful modification of the sulphur feed is to introduce it in the molten condition through a pipe entering at the axis of the burner. A continuous supply of * PREPARATION OF SULPHUR DIOXIDE 163 molten sulphur may be obtained by placing the crude sulphur in a steam jacketed tank and also surrounding the delivery pipe with a steam jacket so that the sulphur will not solidify before it Fig. 23. Vesuvius Sulphur Burner Courtesy of Valley Iron Works Company reaches the burner. Users of this type of burner claim that it is the best if it is fitted with self-feeding device and a large com- bustion chamber because it makes stronger and more uniform gas and starts and stops very quickly in case of shutdowns. 164 THE SULPHITE PROCESS "¥" Fig. 24. Vesuvius Sulphur Burner, Sectional View Courtesy of Valley Iron Works Company The stationary type of burner is illustrated in Figs. 23 and 24. It consists of a vertical cylinder lined with fire brick and fitted with four combustion trays or shelves. The sulphur is introduced through a melting chamber at the top and flows downward from PREPARATION OF SULPHUR DIOXIDE 165 one tray to the next. The ash and residue from the sulphur col- lect at the bottom and if at any time they accumulate on the shelves they can be flushed downward by a sudden increase in the sulphur feed. This burner has no moving parts and therefore requires no power. The makers claim that a burner a little less than 7 ft. high over all and occupying a floor space 6 ft. square has a capacity of 9 tons per twenty-four hours. It is the exper- ience of numerous operators that burners of this type are difficult to run satisfactorily. The most important factor in sulphur burning is the regulation of the air supply. One pound of sulphur requires for its complete combustion just one pound of oxygen, which is the amount con- tained in 53.8 cu. ft. of air. If much more air is admitted, and particularly if it contains much moisture, there is a formation of S0 3 and sulphuric acid which causes loss of both lime and sul- phur. According to Frohberg x the maximum formation of S0 3 takes place at 400 to 500 C. while at 900 to 1000 C. it is again broken up into sulphur dioxide and oxygen. He recommends running the furnaces as hot as possible to produce rich gas which should then be at once sprayed with cold water to reduce its temperature to 90 to ioo° C. In one modern American mill the best temperature, measured at the outlet of the combustion chamber, is considered to be 620 C, while variations from 480 to 68o° C. sometimes occur. According to Cosier 2 poor cooling may be responsible for excessive formation of SO3, weak cooking acid and low free SO2. Acid saturated with CaS0 4 may carry into the digesters as much as 55 lbs. per ton of chips. This is largely precipitated at the temperature of cooking and may cause poor penetration and necessitate a high temperature to finish the cook. The ehmina- tion of S0 3 can be assured by employing the Cottrell electrical precipitation process or filtering the gas through sawdust as recommended by Paulson. 3 1 Chem. Ztg., 1914, 38, 126. 2 Cosier: Paper, 1918, Feb. 13, p. 19. 3 Paper, 1917, 21, Oct. 3, p. 32. 1 66 THE SULPHITE PROCESS Any overheating of the burner is likely to cause vaporization of unconsumed sulphur which passes along with the gases until it reaches the colder portions of the system where it condenses. Sublimation of sulphur also occurs when too little air is admitted after the burner has become thoroughly heated up since there is then too little oxygen present to combine with all the sulphur vapor. The sublimation of sulphur is likely to lead to the formation of thiosulphuric and polythionic acids which at the temperature of cooking again break down with the liberation of sulphur. Free sulphur may also appear in the acid from direct contamination with sublimed sulphur. Klason claims that a liquor may contain as much as 250 mgs. per liter of sulphur as thiosul- phuric acid before sulphur is formed in the digester and as under normal conditions the fresh liquor contains only about 3 mgs. per liter of sulphur the danger from this source is not very great. The conditions under which a burner is working may be judged from the appearance of the flame. When operating satisfac- torily it is blue, sometimes tipped with white; if it shows brown fumes of unconsumed sulphur vapor it indicates that the furnace is too hot, probably from the use of too much air, and that there is danger of sublimation. Attached to, or immediately adjoining, nearly every type of sulphur burner is a combustion chamber which the gas enters as soon as it leaves the burner proper. This is so arranged that more or less air can be admitted at will through appropriate dampers and in this way sublimed sulphur carried along from the burner can be completely burned to sulphur dioxide. The burning of pyrites in the old type of burners was con- siderably more difficult to control than the burning of sulphur and it could be worked advantageously only where the burners could be grouped together in sufficient numbers to insure gas of even composition. Many of these difficulties have been over- come by modern mechanical furnaces of which the Herreshoff fur- nace is a type. These burners (see Fig. 25) usually have five shelves over which the pyrites is raked in succession by mechan- ically operated rakes. The shafts and arms are hollow and are PREPARATION OF SULPHUR DIOXIDE 167 cooled by a current of air supplied by a fan; part of the hot air thus produced is used in the lower parts of the chamber and materially assists combustion. The heat produced by the oxida- tion of the sulphur and iron is sufficient for carrying on the opera- tion and once the furnace is in good working condition no fuel is required. The spent pyrites or cinders leaves the furnace with I per cent to 4 per cent of sulphur. The separation of dust is Fig. 25. Herreshoff Pyrites Burner particularly important where pyrites is used. This was formerly done by passing the gas through long chambers of large area so that the velocity of the gas should be slight. In modern practice it is more successfully accomplished by passing the gas through towers into which water is sprayed; this not only removes dust i68 THE SULPHITE PROCESS but also takes out SO3 and cools the gas. Provided the wash water is discharged at 175 F. (8o° C.) the loss from dissolved SO2 is very slight. The gas from sulphur burners operating under satisfactory conditions generally contains 14 to 18 per cent of S0 2 by volume; the maximum which it can possibly contain is 21 per cent. In the case of pyrites the theoretical maximum is 16.2 per cent of SO2 and it generally runs about 10 to 14 per cent. As a rule when burning sulphur about 2 to 3 per cent is converted into S0 3 while with pyrites as much as 13 per cent may be lost in that way. After leaving the combustion chambers the gas is conveyed through iron pipes to the coolers. Up to this point the gas is hot and dry and has little action on iron but for the cooler and all pipes beyond lead should be used. The cooler generally con- sists of lead pipes through which the gas passes back and forth. These pipes are placed in a trough through which water flows, or are so arranged that a thin film of water trickles over them. The cooling surface should be about 15 sq. ft. per ton of daily produc- tion which is sufficient to bring the gas nearly to the temperature of the water in summer. In winter the gas should be cooled to about 55 F. (12.8 C). Regular and uniform cooling of the gas is very important as the rate of absorption and the quantity of gas dissolved depend very largely upon the temperature. The following table 1 shows how rapidly the quantity of gas absorbed decreases with rise in temperature. 1 vol. of water dis- 1 vol. of solution Temperature solves SO2 contains SO2 Degs. C Vols. Vols. 79-79 68.86 10 56.65 SI-38 20 39-37 36.21 30 27.16 25.82 40 18.77 17 .OI Absorption Apparatus. The apparatus in which the bisul- phite solution is prepared depends on whether the base is used 1 Schonfeld: Ann., 95, 5. ABSORPTION APPARATUS in suspension or in lumps of con- siderable size. In the first class come the apparatus of Partington, McDougall, Frank, Burgess, Steb- bins, Barker, etc., while in the second class are the systems of Flodquist, Mitscherlich, Kellner, Ekman, Jenssen, etc. As the tower is the oldest and in some ways the simplest form of absorp- tion apparatus it will be con- sidered first. The essential feature of the Mitscherlich system is a high tower, usually circular in section and built of wood or of cement lined with acid resisting tile. They vary from 6 to 10 feet in diameter and from ioo to 150 feet in height and generally taper slightly toward the top. Fre- quently four or more towers are built together and the whole sur- rounded with a wooden structure with stairs, platforms and stone hoist. A water tank, supplied with cold water, surmounts each tower. The stone is supported on strong oak beams placed about 6 to 10 ft. from the bottom and below these, and about a foot above the acid outlet, are other beams set close together to catch pieces of stone which pass the upper timbers. Frequently the tower is divided into sections 170 THE SULPHITE PROCESS by timber gratings to assist in the filling and the regulation of the absorption. Some towers act as chimneys and no artificial draft is necessary while with others it is desirable to place a fan between the burners and the towers and a steam exhaust at the top of the towers. Ritter-Kellner towers are constructed in pairs; the acid from the bottom of one is pumped to the top of the second while the gas from the top of the second is led into the base of the first. These towers are smaller than the Mitscherlich towers and have the advantage of avoiding undue loss of gas. Further development along this same line is in the direction of the multiple tower system where towers about 20 ft. high are worked in groups of six to eight. The acid passes through these in succession in one direction and the gas in the other. This system is easy to charge with stone, and permits regulation of the gas temperature between towers, which is important in maintaining a constant ratio of base to acid. Its disadvantage lies in the necessity for so many small acid pumps. Such a system is cheaper to install but more costly in repairs than the high tower system, and the latter is replacing all others in European plants. In any of these systems the towers are filled with lumps of limestone or dolomite. In European works a special soft lime- stone is preferred but sufficiently pure material of this nature is not available in this country and ordinary dense stone is used. A stone low in magnesia and as free as possible from dirt, iron and silica is preferred. Since marble is practically all calcium car- bonate and is of uniform structure it is highly satisfactory for use in towers. A typical analysis of a suitable stone is as follows: 1 Per cent Loss on ignition 43 ■ 6 3 Iron and alumina o. 74 Calcium oxide 54- 10 Magnesium oxide 0.82 Silica Q-59 99- 88 1 Cooper: Paper, 1918, 22, 721. ABSORPTION APPARATUS 171 The water is discharged over the stone at the top by spray pipes or some similar device and in passing downward forms a thin film on the surface of the lumps. In order that it may be properly distributed the inside of the tower is fitted with wooden rings at intervals which prevent the water from running down the walls without moistening the stones. The gas from the burners enters the base of the tower under the grating and pass- ing upward over the moist limestone is very rapidly absorbed by the downward flowing film of water. In working with tower systems several difficulties are likely to be encountered. It is hard to secure a uniform distribution of the water as it descends the tower or a proper spread of the ascend- ing gas. This tends to form channels which increase rapidly after their first appearance. The lower lumps of stone are in contact with the strongest gas and so dissolve more rapidly than those in the upper part of the tower; this tends to form arches which finally break, letting the stone above settle so compactly that it may impede the passage of the gas. This trouble is over- come in modern installations by tapering the tower toward the top, by dividing it into sections which are packed with stone separately and by proper periodical inspection. In the lower part of the tower crusts of sulphate or of monosulphite of lime sometimes almost stop the flow of gas; the latter is particularly apt to form if the gas is weak or insufficient water is used. Irreg- ular acid is also likely to result if the temperature varies since, as already shown, the solubility of the gas decreases rapidly with rise of temperature while at the same time the base is much more quickly dissolved thus changing the proportion between free and combined acid. Apparatus for use in acid making by the other system, where the base is in solution or suspension, generally consists of a series of tanks with accompanying piping or in other installations of towers divided into sections by partitions. A frequent arrangement consists of three tight tanks fitted with agitators and with pipes so arranged that the gas enters the bottom of the first tank and passes upward through the solution; from the 172 THE SULPHITE PROCESS ■V +■> a> SB 5 1 —~ ABSORPTION upper part of this tank the un- absorbed gas passes to the bottom of the second and so on through the system. The gas may be either forced in under pressure or caused to pass through by an exhaust fan attached to the last tank. It is general to place the tanks at different levels so that after drawing off the finished acid from the first tank the contents of the second and third may each be run down one stage by gravity. Fresh milk of lime is charged into the third, or upper, tank, and the finished liquor leaving the lower tank should be quite clear. The Burgess and the Barker systems are examples of the towerlike form of tank appa- ratus. The Burgess apparatus is generally one high tank di- vided into three parts by hori- zontal partitions. It is fitted with a hollow shaft and arms through which the gas passes and is mixed with the milk of lime. The Barker apparatus illustrated in its relation to the other equipment in Fig. 27, is a high tank divided into three or more compartments by hori- zontal perforated partitions. The construction of the tower is shown in Fig. 28. The milk APPARATUS 5ecr/oN A- A 173 Qcmcnt Qacxinq Sreci. tank Aoo 'psoof Brick Fig. 28. Barker Tower, Sectional View 174 THE SULPHITE PROCESS of lime enters the upper compartment in a continuous flow and meets the gas bubbling up through the perforated false bottom. The weak liquor passes through an overflow pipe to the next compartment where it absorbs more gas, and so on through the four sections. From the lowest one it flows onto a distributing plate which allows it to trickle down into the absorbing tower where it meets strong gas and is strengthened to the desired composition. The gas enters near the bottom under a perforated distributing plate above which is the absorption tower filled with stoneware filling to increase the surface exposed. This appa- ratus, as well as any good tower system, working in conjunction with a proper reclaiming system, will produce an acid with i per cent combined and up to 6 per cent total S0 2 . The quality of the lime or dolomite used for liquor making is of the greatest importance and its value increases with the amount of magnesia which it contains. For use in milk of lime systems it should be well burned, should slake easily and should be as free as possible from silica and iron. Air slaked or poorly burned limes are not so readily acted on as those of good grade and they are likely to vary so in composition that it is almost impossible to keep the correct proportion of base to acid. The following analyses show the composition of several extensively used limes from different parts of the country. Calcium oxide, CaO Magnesium oxide, MgO Alumina and ferric oxide, A1 2 3 and Fe20 3 Sulphur trioxide, SO3 Insoluble in HC1, sand, etc Silica soluble in acid, Si02 Loss on ignition, H 2 0, CO2, etc Massachu- setts Per cent 56.02 40.10 0.57 O.II 0.94 0.47 i-43 99-64 Ohio Per cent 5 8.6l 40.25 O.I2 0.1S 0.07 0.15 99.86 New- Brunswick Per cent 55-96 37-98 I.23 0,16 I-5I 99-65 It frequently happens that lime of the above quality cannot be obtained on time so that it is necessary to use high calcium ABSORPTION APPARATUS 1 75 lime for a certain period. It has been found by experience with the Barker system that this causes no serious inconvenience nor does it require any radical change in the cooking system. The opinion seems to be gaining ground that a high calcium lime can be used in the milk of lime system with practically the same results as with dolomitic lime. Lime for use in the absorption system is first slaked either in tanks provided with agitators or in troughs. One authority recommends a tank with a perforated plate on which the lime is dumped after filling the tank with water to one inch above the surface of the plate. The lime is sprinkled on top with water and when it steams freely is quickly covered with water. The milk of lime produced is diluted sufficiently so that it can be strained through brass sieves of 60 meshes to the inch and the strained material run to storage tanks fitted with agitators where it is diluted to the proper consistency. The milk of lime must be cold when it goes to the absorption system. According to Beveridge x the finished milk of lime has a specific gravity of 1.0075 and contains 6.31 gms. per liter of CaO and 4.19 gms. per liter of MgO. This corresponds to about 91 lbs. of lime of the composition given to a thousand gallons of acid. As the lime is never entirely dissolved and some is lost as sulphate and mono- sulphite in cleaning the apparatus more than the above quantity has to be used, the amount depending on the quality of the lime, the type of apparatus, etc. There is still much dispute as to the relative advantages of the tower and milk of lime systems. The latter are no longer com- mon in Europe where towers are generally employed, but they are much used in America. The advocates of the tower systems claim that they are simpler to operate, require less power and cost less for upkeep. The cost of lime is also less as the unburned stone is cheaper than burned lime. Obermanns 2 states that towers require 50 to 60 horse power as compared with 225 horse power for a three tank system, and that the sulphur per ton of 1 Beveridge: Paper Makers' Pocket Book, p. 99. 2 Obermanns: Paper, 1918, Feb. 13, p! 100. 176 THE SULPHITE PROCESS pulp is 240 lbs. as against 300 lbs. for the tank system. For a 100-ton mill these savings, together with that for the cheaper lime, would amount to $42,000 to $52,000 per year. Textor 1 compared the systems from a thermochemical standpoint and con- cludes that an acid of 2.00 per cent free and 1.60 per cent com- bined S0 2 will entail a rise of 6.4 C. if made in towers from cal- cite, and 16.1 C. if made in a milk of lime system from a high magnesia lime without the use of steam in slaking or of cooling water in making up . the tanks. Considering the claim that towers produce stronger acid it is to be noted that they operate at nearly atmospheric pressure whether the gas is passed through by pressure or vacuum. Tank systems, such as the Barker, gen- erally operate by means of a vacuum pump and if this were changed to force the gas through under pressure the strength of the acid would be increased by 1 5 to 20 per cent. The losses which occur in making sulphite liquor are those due to dirt, ash and moisture in the sulphur, to sublimation, and to the formation of sulphuric acid. The loss from the first four causes should never exceed 5 per cent, but that from formation of sulphuric acid is likely to amount to very much more unless careful control is maintained. The formation of monosulphite, which is removed with the sediment in the storage tanks or in cleaning the absorption system, may also cause considerable loss. Such wastes should be examined before being thrown away as it often pays to work them over. The acid made in either the tower or the tank system varies very widely in different mills according to the kind of fibre being made and the method of enriching the liquor with the relieved or recovered gas. There is no essential difference, however, between the two systems with regard to the liquor produced. Harpf 2 gives the composition of liquor of 4.5 Be. from Mitscherlich towers as follows: Per cent Total S0 2 3-397 Free S0 2 2. 098 Combined SO2 1 • 299 1 Textor: Paper, 19 18, Feb. 13, p. 60. 2 Harpf: Dissertation, 1892. PUMPING AND STORAGE 1 77 The acid from either tower or tank system is stated by Thorne 1 to have the following composition before being enriched by the recovered gas. Per cent Total S0 2 2. 60 FreeS0 2 ... 1.60 Combined SO2 1 • 00 The finished acid when ready for the digesters is also very variable in strength. From reports on a number of American mills the following figures have been selected as representative: Total S0 2 Free S0 2 Combined SO2 I 2 3 4 Per cent Per cent Per cent Per cent 4-3° 4.09 5-50 3-8o 3.10 2.46 4-5° 2.40 I .20 1.63 1 .00 1 .40 Per cent 6.24 5.22 1 .02 Acid of the strength of No. 5 can only be obtained during warm weather bjf employing artificial refrigeration; in this particular case the temperature of the acid ready for the digesters was 72 F. (22. 2° C). Pumping and Storage. Where the liquor is to be discharged directly into the digester, a steam injector may be used for transferring it, but if nothing is to be gained by heating the liquor an injector is too expensive and it also causes considerable loss of sulphurous acid. The best method of handling is with rotary pumps of acid-resisting bronze, which should be so placed that the acid flows to them under a slight head. If a pump is so placed that a foot-valve is necessary on the suction pipe continual trouble will be caused by the crystallization of monosulphite in the working parts. Storage tanks for liquor are generally of wood, either Southern pine or Douglas fir, without lining. They should be made tight with water or steam before any liquor is admitted so that no monosulphite may crystallize between the staves. The tanks 1 Thorne: Pulp Paper Mag. Can., March 15, 1915, p. 173. 178 THE SULPHITE PROCESS should be covered to prevent escape of gas but the covers need not be perfectly air-tight since there is little loss in strength, either through escape of gas or Oxidation to sulphate, when the liquor is stored in quantity. The tanks should be fitted with gauge glasses to show the depth of liquid and the delivery pipes should draw from near the bottom but high enough to avoid sediment. The tanks should be so located that the sediment may be easily washed out when necessary. Digesters and Digester Linings. The acid liquor used in the sulphite process acts so destructively on iron that some form of lining is necessary to protect the digester shell. According to Griffin and Little x wrought iron suffers most severely, steel resists somewhat better and cast iron suffers least of all. Even the modern acid-resistant cast irons, which are extensively used in concentrating acids, have proved to be too easily attacked to be safe for use in the sulphite industry. Tests by the author on a number of such cast irons gave the following results: Loss in grams per square inch 18 hours in ' cold 4 hours more at 90 C 18 hours more in cold Ordinary cast iron O.6410 0.2615 O.2362 0.327S O.5410 Acid resistant No. 1 0.1096 0.0990 O.2138 O . 2030 O.I 184 O.1797 Acid resistant No. 2 Acid resistant No. 3 Acid resistant No. 4 Lead is the only common metal which satisfactorily resists the action of the acid liquor and this is due very largely to the formation of a surface film of insoluble lead sulphate which acts as a protective coating for the metal beneath. Lead, however, has certain properties which have prevented its successful use in digester lining in spite of the immense amount of time and money which have been spent in the attempt. Its coefficient of expan- sion is 0.0000297 while that of iron is only 0.0000123, so that on 1 Griffin and Little: Chemistry of Paper Making, p. 232. DIGESTERS AND DIGESTER LININGS 1 79 heating the digester the lead lining tends to become too large for the shell. This trouble is still further increased by the fact that lead which has been expanded by heat does not quite return to its original size on cooling but remains permanently larger. This causes " crawling" and " buckling" and cracks are apt to appear wherever short turns are made. There is also in vertical digesters a gradual creeping downward of the lead due to its own weight; this causes the upper part to become thinner and finally give way. Even uniting the lead to the iron by melting it on over a flux of zinc chloride, a true soldering process, did not prove entirely satisfactory. Griffin * states that such linings remained clean until about two hundred cooks had been made, then star- shaped defects showing cracks and hard crystals appeared and multiplied so fast that they could not be cut out and repaired. Finally black scabs of lead sulphide formed in large masses and the whole lining became worthless. Many other very ingenious methods for controlling the lead were tried but the task was finally given up as hopeless. Bronze digesters, built of cast sections, were tried at one time but were found to be more or less acted on by the liquor with the formation of black scales of oxide and sulphide of copper. Heating to the temperatures used in cooking considerably re- duces the strength of bronze and after several disastrous explo- sions the use of such digesters was abandoned. The Salomon-Briingger digester consisted of an inner shell of welded steel and an outer shell also of steel but riveted. The protective coating is obtained by admitting sulphite liquor into the digester which has been previously heated by steam in the jacket at about 40 lbs. pressure. This treatment causes the deposit of a hard, impervious crust of sulphite of lime which gradually increases in thickness with each succeeding cook. This coating did not prove to give adequate protection and the method was never in extensive use. The Mitscherlich lining is interesting as being the first in 1 M. L. Griffin: J. Soc. Chem. Ind., 1898, 216-220. 180 THE SULPHITE PROCESS which bricks were used. It consists first of a coating of tar and pitch applied directly to the shell, then a thin lining of sheet lead with the edges burned together and finally two courses of dense vitrified bricks with tongues and grooves. These were sometimes laid in Portland cement. The Preston lining consists of bricks of Scottish clay backed with a mixture containing clay and lead mixed to the consistency of bread dough with silicate of soda. This must be applied to a perfectly clean shell. Modern digester linings are generally of acid-proof bricks backed with cement next to the shell. The bricks are 2 to 3 ins. thick while the cement backing is about an inch thick. In some cases a lead lining is applied next the shell and upon this the ce- ment backing is laid. This is seldom done, however, as it is cus- tomary to pierce the digester shell with numerous small tell-tale holes so that the location of cracks in the lining may be approxi- mately known. The cement for the backing and pointing the first layer of bricks varies more or less in different localities. Steffanson 1 gives its composition as one part of cement, and two parts of crushed and sifted acid-proof brick with enough asbes- tos added to render it non-brittle; this is mixed to the desired consistency with 4 Be. silicate of soda. The last layer of bricks is pointed with litharge and glycerine and the bricks should be set half an inch apart to make repairs easy. Another formula 2 for pointing the inner layer of bricks is 5 parts litharge, 2 parts cement and 3 parts quartz sand, all measured by volume. After mixing these materials dry, they are moistened with glycerine to the right consistency for use. This should be mixed in small quantities only and used quickly as the mixture retains the proper consistency for only a short time. Digesters pointed with this mixture have been operated up to eighteen months without repairs and in some cases the pointing has proved more durable than the bricks so that the latter have worn down, leaving a raised network of cement exposed. 1 Steffanson: Pulp Paper Mag. Can., May 20, 1914, et seq. 2 Private communication. DIGESTERS AND DIGESTER LININGS l8l The bricks used for digester lining should be very hard, dense and well annealed. If soft or under-burned they are apt to crack from changes in temperature and pieces then come away in the pulp. When tested by immersing in water for twenty-four hours they should not absorb more than 2 per cent of their weight. In general the form of digester used is that of a vertical cyl- inder with conical top and bottom. The total length is about three times the diameter, and the lower cone is about 60 degs. while the upper is about no degs. Digesters of other forms are of course used in some of the older mills and horizontal digesters are sometimes used for the Mitscherlich process. The size of digesters has gradually increased; formerly a capacity of 4 tons of fibre per charge was considered large while now 18 tons or more is not uncommon. The following table by Cor- coran x gives the approximate capacities of sulphite digesters of standard construction and lined with the usual brick and cement lining: 1 Corcoran: Paper, 22, 1918, 406. 182 THE SULPHITE PROCESS Capacity of Standard Sulphite Digesters with Standard Lining9 Size of digester Thickness of lining, Contents, Capacity- contents, Gallons of Cords of Diameter, ft. Height, ft. ins. cu. ft. tons fibre acid wood 8 24 8 610 i-33 3,000 2.48 8 30 8 840 i-75 3,787 3-24 IO 28 8 I-3I9 2.66 6,000 4.96 IO 30 8 i>397 2.90 6,525 5-34 IO 37 8 1, 850 3-85 8,663 7.16 IO 40 8 2,024 4-5o 9,450 7.81 II 30 8 1,672 3-48 7,330 6.47 II 37 8 2,896 4.60 10,125 8-37- II 40 8 2,416 5.00 11,250 9 30 II 42 8 2,563 5-33 12,000 9.92 II 45 8 2,784 5-75 12,937 10.69 12 3° 9 2,015 4-i3 9,282 7.67 12 35 9 2,457 5-io 11,470 9.48 12 40 9 2,879 6 .00 13,500 11 .16 12 45 9 3,272 6.80 15,300 11 .64 12 48 9 3,572 7.40 16,650 13.76 14 38 9 3,8i9 7-9o i7,775 1464 14 42 9 4,320 9.00 20,250 16.74 14 45 9 4,678 9-75 2i,934 18.13 14 47 9 4,924 10.50 22,950 18.97 14 48 9 5,046 10.64 23,625 19-53 14 5o 9 5,392 11 .20 25,000 20.83 IS 40 10 4,682 9-75 2i,934 18.13 IS 42 10 4,964 10.33 23,250 19.22 IS 45 IO 5,388 11 .20 25,200 20.83 IS . 47 IO 5,671 11.80 26,550 21.31 IS 5° IO 6,096 12 .40 27,900 22.06 IS 54 IO 6,652 13-75 3o,937 25-57 16 45 IO 6,146 12 .80 28,800 24.80 16 48 IO 6,680 13-75 30,937 25-57 16 So IO 6,952 14.40 32,400 26.78 16 54 IO 7,598 15.80 35,55o 29.38 16 60 IO 8,565 17.80 40,050 33 -io 16 64 IO 9,210 19.00 42,750 35-34 17 56 IO 9,o74 18.80 42,300 34-96 17 60 IO 9,8i4 20.25 45,900 37-94 17 64 IO io,5S2 21 .80 49,050 40.44 17 70 IO 11,660 23.00 5i,75o 42.78 The digester space required to produce 2000 lbs. of dry pulp is given by Steffanson as follows: Cu. ft. Mitscherlich process 400-425 Quick cook process 450 For easy bleaching pulp 475-500 BOILING 183 Boiling. The method of making a boil depends on whether the Mitscherlich or the quick cook process is being employed. The former is very generally used in Europe but in America is not nearly so common as the quick cook or Ritter-Kellner process. In the Mitscherlich process the digesters are either horizontal or vertical stationary boilers. The cooking is all done by steam admitted to coils of hard lead or copper pipe placed in the bot- tom of the digester. The standard procedure is to fill the digester with chips and then steam gently for several hours with direct steam, the water condensing being allowed to run to waste as a brownish liquid. Care must be used during this period to avoid steam pressure in the digester as temperatures much in excess of ioo° C. are likely to burn the wood. After steaming all valves except that leading to the liquor tanks are closed and the partial vacuum formed by the cooling of the digester and the condensation of the steam draws the cold liquor in rapidly. The object of this steaming and subsequent admission of cold liquor is to obtain thorough penetration of the chips by the liquor and so prevent floating and burning. Steam is now admitted to the coils and the temperature raised to no C. as rapidly as possible, although this may require as much as twelve hours because of the large size of the digester. When pressure is reached it is relieved by opening a valve for^ a few minutes to get rid of air; this is repeated two or three times in the next hour. Since the relief from Mitscherlich cooks contains no liquor no separator is necessary and the relieved gas can go at once to the reclaiming system. The temperature in the digester is gradually raised to about 120 C. which is maintained throughout the cooking period; during this time the pressure should not exceed 80 lbs. About an hour before the end of the cook the steam is shut off and the pressure gradually reduced to 50 lbs. by relieving gas; this must not be done too rapidly or the pulp may not be thoroughly reduced by the time the sulphur dioxide is gone. The contents of the digester are then discharged as usual. The old method 1 84 THE SULPHITE PROCESS of emptying horizontal digesters was to admit cold water as soon as the liquor had been discharged, the object being to cool and wash the pulp which was finally removed by shovelling. The liquor used in the Mitscherlich process is about 3.5 to 4.5 per cent total S0 2 with 0.9 to 1.24 per cent combined. The steam used in the coils is at 60 to 100 lbs. pressure: 75 lbs. in the coils gives about 90 lbs. in the digester. The actual time of cooking varies enormously in different mills. It was origi- nally about eighty hours, but this has been greatly reduced by raising the temperature of cooking and by using some direct steam to bring the charge up to pressure quickly. In this latter case space must be left in the digester to allow for con- densation. In modern practice the total time is about twenty- five to forty-five hours. Steffanson * states that for bleached pulp the cook is usually not more than twenty-four hours, while Beveridge 2 subdivides the time as follows for a digester which has to be emptied by hand. Hours Filling 2 Steaming 4 Filling with liquor 2 Boiling 35 Blowing off pressure 3 Washing twice 6 Emptying, etc : 5 Total Jj The particular advantages of the Mitscherlich process are strong fibre and high yield because of the comparatively weak acid and the low temperature of cooking. It is stated by Bache-Wiig 3 that temperatures over 13 5 C. cause the forma- tion of hydrocellulose with consequent lower yield and loss of strength. The Ritter-Kellner, or quick cook process, in which the steam is blown directly into the digester, is the one most generally 1 Steffanson: Pulp Paper Mag. Can., May 20, 19 14, et seq. 2 Beveridge: Paper Makers' Pocket Book. 3 Private communication. BOILING 185 used in this country. The digester is usually charged with chips as fully as possible since the settling during the first part of the cook suffices to cover them completely with liquor. Steam- ing of the chips is seldom resorted to, though it is an advantage, as it drives out air, moistens the chips uniformly and makes better pulp. Moreover the partial vacuum formed in the digester when steam is shut off hastens the running in of the acid and makes it possible to perform this operation without removing the digester head. The most satisfactory point to admit the acid is at the bottom of the digester; if pumped in from below instead of on top of the chips it tends to loosen them up and helps the circulation. The steam inlet for cook- ing generally ends in a coil around the sides near the bottom and is perforated in such a way as to direct the steam up the sides, thus giving a downward current in the center and good circulation. There is also provided a small jet just at the bottom of the digester to cook the chips in the lower part of the bottom cone. The steam used in cooking may be superheated or ordinary saturated steam, the latter being used in far the greater num- ber of cases. Tests by Andrews 1 using superheated steam with a temperature of 500 F. at the digesters indicated that it gave a little more uniform product and that a somewhat stronger acid could be prepared. No difference in yield could be de- tected and the volume of liquor and the fuel required were the same as with saturated steam. The steam required for cooking is much greater than for a soda cook of a corresponding number of cords. This is due to the continual relief of gas and steam through the coolers into the recovery system. Andrews calculates the saturated steam for a 14 X 47 ft. digester holding 16 cords of rossed wood as about 60,000 lbs. per cook. In another plant with digesters of about the same capacity steam flow meter records during a period of four months showed that the average steam consump- tion per cook was 75,940 lbs. 1 Andrews: Paper, Feb. 20, 1918. 1 86 THE SULPHITE PROCESS In steaming a cook it is very important that the pressure be brought up slowly as otherwise a high temperature may be reached before the liquor has had time to penetrate the chips and their centers will be found hard and of a red or brown color. The time from the start until 75 lbs. pressure is reached varies in different mills from two to four hours. The pressure, however, does not afford a reliable indication of conditions within the digester since the actual steam pressure is augmented by that of the gas set free during boiling and in some cases the indicated pressure may be almost wholly hydrostatic due to the filling of the digester by condensation. The temperature is therefore the real factor to be watched and this should be taken at a point about one-third of the way down the digester. The best method of keeping track of the temperature is by means of some form of recording thermometer as this gives a permanent record of each cook from start to finish. No hard and fast rule for cooking can be given and each mill has its own particular method which is generally the result of gradual evolution rather than the application of scientific knowl- edge. As an example of the procedure in cooking easy bleach- ing pulp of high quality Steffanson * gives the following schedule: Steam in such a way as to reach 75 lbs. pressure in two to three hours; open relief and bring temperature to 240 F. (115.5 C.) in about an hour. Close both steam and relief valves for an hour and a half, then turn on steam and open relief very slightly. The maximum temperature of 300 F. (149 C.) should be reached in ten hours with the maximum pressure still 75 lbs. Now shut off steam but not relief, allow the pressure to drop to 50 lbs. in one to two hours and discharge into the blowpits. At the end of the cook the liquor should test 0.05 per cent total S0 2 . If the cook is blown at a pressure much in excess of 50 lbs. some partially cooked chips will be blown to pieces and cause shives. 1 Steffanson: Pulp Paper Mag. Can., May 20, 1914, et seq. ■ BOILING 187 In another mill making an easy bleaching sulphite for use in writing papers the following schedule is in effect: Bring to 78 lbs. pressure in three hours and then start relief. Bring temperature to 228 F. at 4th hour. Bring temperature to 246 F. at 5th hour. Bring temperature to 262 F. at 6th hour. Bring temperature to 278 F. at 7th hour. Bring temperature to 290 F. at 8th hour. Bring temperature to 298 F. at 9th hour. Start to reduce pressure when 5 c.c. of liquor require 2.7 c.c. of iodine and lower to 50 lbs. at blow, which is when 5 c.c. of liquor require 0.7 c.c. of iodine. Fig. 29 shows the general method of recording the conditions Time Fig. 29. Sulphite Cooking Chart within the digester during the cook. Observations are taken at frequent intervals and plotted on the chart in such a way as 1 88 THE SULPHITE PROCESS to show the gauge, gas and steam pressures. This gives a graphic representation which can be followed easily by the workmen. The chart given is for a quick cook easy-bleaching sulphite. Fig. 30 is the temperature record taken during such a cook. During the progress of a cook the heat and the chemical re- actions taking place within the digester cause more or less gas ^.., *$B _- — 1 — -—JJfy*' . 4^ « .-■- \ IM^ ^— $&&£ Fig. 30. Temperature Record of a Sulphite Cook to be evolved which results in a gradual building up of the pres- sure. This gas pressure is not injurious to the fibre provided the temperature is carried at the right point, but it is the general custom to reduce it by "relieving" or blowing off some of the gas either at intervals or continuously. Study by Schwalbe 1 of 1 Schwalbe: Wochbl. Papier-Fabr., 1913, 44, 2786. RELIEVING GAS 189 the relieved gases shows that when blown off at no° C. they contain no oxygen, indicating that all originally present in the digester has already been used up. He recommends relieving air when the temperature reaches 75 C. as the bisulphite is relatively stable at that temperature and the loss of gas will consequently be less. The relief at times also contains much liquor and for this reason it is sometimes passed through a separator of some kind so that the gaseous portion may be con- veyed to the liquor tanks to bring up the strength of the raw acid to the proper cooking strength. This recovery process is of the first importance in reducing the consumption of sulphur per ton of pulp. The amount of relief depends partly on the strength of the acid used since the stronger the acid the greater will be the amount of gas evolved on heating. The kind of cook, direct or indirect steam, and the temperature of cooking also have much influence on the amount of relief. If too much gas is blown off the pulp will be burned and whenever burned pulp is obtained and the temperature has not risen above 320 F. (160 C.) it is an indication that the acid was originally too weak or that too much gas was blown off. 1 If it is discovered during a cook that the usual method of relief will allow too much S0 2 to escape before the cook is done it may be remedied by closing the relief, drawing off some of the liquor from the bottom of the digester, and then again steaming. Some mills use this method regularly, relieving only to get rid of air and to reduce pressure before blowing. 2 In practice the relieving of a digester is done from one or both of two points, either through the top or through the side. It is claimed by Wimmer 3 that side relief is a help in systematic cooking and also aids in the recovery of gas and the reduction of sulphur per ton of pulp. To use the side relief to the best advantage he recommends relieving from the top for about one and one-half to two hours, or until the temperature is 120 to 1 Griffin and Little: Chemistry of Paper Making, p. 252. 2 Steffanson: Paper, June 3, 19 14, p. 19. 3 Wimmer: Paper, Jan. 19, 19 16, 15. 190 THE SULPHITE PROCESS 130 C. (248 to 266 F.), then, closing the top entirely, relieve from the side for one and one-half to two hours. After this top relief can be started again and only dry gas will ordinarily be obtained to the end of the cook. On large digesters there should be from three to five hours of dry gas. . As already stated the composition of the acid liquor varies greatly in different mills and with the nature of the fibre being made. Schlick x states that the strength of the pulp produced is in indirect proportion to the strength of the acid used. Ex- periments by the Forest Products Laboratory 2 appear to indi- cate that with a total SO2 content of 5 per cent, increasing the amount of combined SO2 above 1.0 per cent, has very little influence on the time of cooking while decreasing the combined acid below 1.0 per cent increases the speed of cooking. As the combined SO2 is decreased below 1 per cent there is an increase in the amount of screenings and in the bleach required. Other authorities place the lower limit for combined S0 2 at 0.75 per cent and agree that going below this point increases the bleach required. Decreasing the temperature of cooking between the limits of 146 C. and no° C. tended to increase the yield and decrease the screenings and the bleach required due to the more even cooking. Frohberg 3 claims that the rapidity of the diges- tion, other things being equal, is dependent solely on the con- centration of the free S0 2 and that in order to hasten the process dry wood and liquor rich in free S0 2 should be used. The present tendency is toward the use of acid of high strength, par- ticularly in free S0 2 . An increase in this factor enables the time of cooking and the final temperature to be reduced. The higher test for the digester acid means a higher blowing test at the end of the cook in order to obtain equal yields. It used to be stated that the best acid contained 75 per cent of its total S0 2 in the free state while now it is considered that 80 to 85 per cent is a better figure. Much difference of opinion exists 1 Schlick: Paper Pulp Mag. Can., 1915, 13, 227. 2 Private communication. 3 Wochbl. Papier-Fabr., 1910,41, 1179-1182. PROGRESS OF THE COOK 191 regarding the relative value of acids of different compositions and much more work will have to be done before the influence of all factors can be definitely established. In the cooking process irregularities are likely to occur from very obscure causes. Overcooking may be caused according to Klason : if too little lime is used. The minimum amount nec- essary to saturate the lignosulphonic acids is 22.5 grams per kilo of wood and if less than this amount is used the pulp is overcooked and charred. Overcooking may also result from decomposition of the liquor with formation of free sulphur according to the reaction 3 S0 2 = 2 S0 3 + S. The sulphur acts catalytically, producing further decomposition of the calcium bisulphite, and the sulphuric acid combines with the lime as CaS0 4 so that not enough base is left to combine with the sulphonic acids which polymerize and eventually the fibre is attacked and darkened. Selenium in the cooking liquor will also act catalytically and with far greater power than free sulphur. Torgerson and Bay 2 have proved that it is not the selenium directly but the simultaneous presence of dust that causes the trouble. This dust acts as an energetic contact sub- stance promoting catalytic action of the selenium. This action is recognized by a sudden fall in the S0 2 and lime contents of the liquor after the digestion has proceeded for some time. The best way to determine the progress of a cook is to titrate portions of the liquor with — iodine solution which will show the 10 amount of total S0 2 . In taking samples from the digester for this test a cooler should be used, otherwise much of the S0 2 will be lost. Another test which is frequently employed is to remove samples of the liquor and treat them with a mixture of strong ammonia and water in equal parts in test tubes. This 1 Klason: Wochbl. Papier-Fabr., 19 10, 41, 464 et seq. 2 Torgerson and Bay: Papier-Fabr., 1914, 12, 483. 192 THE SULPHITE PROCESS causes the precipitation of calcium monosulphite which is at first light and voluminous but which decreases in quantity as the cooking proceeds. When a certain volume of precipitate is reached the cook is considered finished. This test is usually stated to show the amount of lime present as sulphite but according to Oman x it is virtually a test for S0 2 since if enough of the latter is present all of the lime will be precipitated. By taking the sample from the digester through a cooler and adding some calcium chloride to the ammonia used the test may be made to indicate the amount of SO2 present. The cooks in charge of the digesters also judge of the condition of the cook, to a certain extent, by the color and odor of samples of the liquor removed from the digester at frequent intervals toward the end of the cook. Recovery of Gas. The relief of gas and liquor from the digester during cooking causes a very large loss of sulphur dioxide and the recovery of this is an important item in keeping the cost for sulphur at a low point. This is accomplished by passing the relief through separators and coolers from which the gas is taken to the acid storage tanks where it is absorbed and brings the acid to the desired strength for cooking. The liquid portion from the separator is either allowed to go to waste or is mixed with the acid from the acid system previous to strengthening the latter with relieved gas. In Thome's 2 recovery system the separated and cooled gas from the digester is passed into the bottom of a tower filled with wood blocks over which the acid from the acid system is passed, the strength- ened acid thus formed goes to the storage tanks. The liquor from the separator contains about 1 per cent of SO2 and after cooling it is sent to the acid system with the water or milk of lime as the case may be. This system is equally applicable to the limestone or milk of lime systems and by its use it is said to be possible to strengthen an acid of 2.60 per cent total and 1 per cent combined S0 2 up to 5.50 per cent total with no in- 1 Oman: Teknisk Tidskrift, 1916, 46, 4. 2 Thome: Pulp Paper Mag. Can.., 1915, p. 173 (March 15). BLOWING AND WASHING 193 crease in the amount of combined; moreover this is accomplished at a temperature of 35 C. (95 F.) with no loss of gas. In blowing down pressure before discharging the digester, the gas can be recovered by passing the gas and steam under the false bottom of a tower in which there is a continuous shower of water. This absorbs the gas and soon becomes heated by the steam to such a temperature that it can no longer hold the gas in solution. This liberated gas, together with that coming from the digester, soon increases to such an extent that the water supplied can no longer absorb it and there is delivered from the top of the tower a constant stream of pure gas which can be used in the acid system. In a few instances attempts have been made to recover the gas liberated when the digester is blown into the blow pits, but the volume of gas and steam which must be handled in a very short time is enormous and it is very doubtful if the results obtained pay for the expensive equipment and the cost of oper- ating. In one installation where this is being tried there has been no reduction in the amount of sulphur used per ton, indi- cating that the recovery is not a paying proposition. The sulphur consumption in practical work is generally fig- ured from the weight of sulphur fed to the burners and the tons of pulp produced. Bryant x states that in commercial work the sulphur consumption varies from 235 to 400 lbs. per ton and that the amount theoretically necessary, not taking the formation of sulphuric acid into account, would be 184 lbs. Schwalbe 2 gives the sulphur consumption as 9 to 10 kgs. per 100 kgs. fibre (180 to 200 lbs. per 2000 lbs. fibre) and thinks that it should be possible to work with as little as 8 kgs. (160 lbs. per ton). Less sulphur per ton is used in the processes where in- direct steam is used because less loss is incurred by dilution and relieving liquor. Blowing and Washing. As already stated the contents of the digester are usually discharged into the blow pit under a pres- 1 Bryant: Paper, Jan. 28, 19 14. 2 Schwalbe: Chemie der Cellulose, p. 530. 194 THE SULPHITE PROCESS sure of about 50 lbs. or less. The pipe from the digester to the blow pit may be of copper or even cast iron since the pitch in the pulp covers the inside of the pipe and prevents corrosion. This pipe is so arranged that the stock is discharged against a target placed in one end of the blow pit, to prevent wear on the pit walls. This target may be of bronze or of hard cast iron; the latter is generally used as it is cheaper, and offers more resistance to mechanical wear than bronze. As the acid at this stage is comparatively weak and the blowing is immedi- ately followed by washing the chemical resistance of the cast iron is sufficient for this purpose. Blow pits are of various shapes but consist essentially of tanks with false bottoms through which the waste liquor may drain. Modern American mills often use perforated tiles in the pit bottoms. They are frequently of reenforced concrete lined with wood of a resinous nature such as Southern pine or Doug- las fir. The washing is done by a stream of water from a hose, or by means of sprinkler pipes, and requires several hours. A novel wash pit arrangement is that proposed by Kuhn. 1 The entire floor is covered with drainer tiles and the wash water enters from below these tiles and is thoroughly mixed with the stock by air or steam forced through a series of perforated pipes laid on the surface of the tiles. After thorough mixing in this way the wash water is allowed to drain off by opening the appropriate valves. If desired this process can be repeated. The claim is made that the washing of 15 tons of dry fibre can be accomplished in 37 minutes as follows: 15 minutes for filling with hot water, 15 minutes for washing with air or steam and 7 minutes for draining. The treatment of the pulp after washing is largely a mechani- cal one, to remove dirt, knots, slivers, and uncooked or partly cooked chips, by means of rifners, screens, etc. The screenings thus removed amount to 3 to 8 per cent of the pulp produced. During these mechanical purification processes the stock is very 1 Kuhn: Papier-Fabr., 1915, 13, 725 and 744. ROSIN IN PULP 195 largely diluted; one authority gives the concentration of the stock as follows: In rifflers 250 parts water to 1 part pulp In coarse screens 125 parts water to 1 part pulp In fine screens 150 parts water to 1 part pulp The unbleached sulphite fibre found on the market shows wide variations in color, strength and physical properties. Its chemical composition also varies more or less as is proved by the following table. Analysis of Unbleached Spruce Sulphite Fibre * Per cent 6-45 0.65 1-52 81.51 9-87 Moisture, loss at 100 ° C Mineral matter (ash) Hydrocellulose, etc., soluble in alkali Cellulose Non-cellulose (lignin) by difference.. Per cent Per cent Per cent 6. i S 6.70 6-57 1 .00 0.45 o-33 2-53 2 .26 4-25 85-32 89-74 88 . 12 5.00 O.85 o-73 * Griffin and Little: Chemistry of Paper Making, p. 268. The ash in sulphite spruce fibre has been given considerable study by Richter. 1 He finds that it rarely exceeds 1 per cent and is usually about 0.5 per cent. Silica, generally amounting to about one-third of the total ash, is probably fixed as calcium or magnesium silicate during boiling as the silica present in the wood is too little to account for so much in the fibre ash. The percentage of iron in the ash showed no constant relationship to the total ash nor to any characteristics of the pulp. The rosin in sulphite has been given much attention because of its possible relation to rosin spots in the paper made from it. Herzberg gives the following percentages of rosin extracted by ether : Bleached Unbleached Mitscherlich pulp Per cent 0.44 o.43 Per cent 0.58 °-59 Ritter-Kellner pulp Other investigators find considerably higher amounts as follows : 1 Richter: Wochbl. Papier- Fabr., 1913, 44, 1776. 196 THE SULPHITE PROCESS Ether soluble Alcohol soluble Total Foreign sulphites * . . American sulphites * Unbleached fibref. . Bleached fibre | Per cent 0.70-1.33 0.65-1 .21 0.82-0.98 o . 66-0 . q8 Per cent 0.10-0.22 0.15-0.62 0.18-0.82 0.23-0.62 Per cent 0.90-1.43 0.86-I .52 1. 05-1. 67 1 . 00-1 . 60 * Richter: Wochbl. Papier-Fabr., 1913, 44, 4507. t Schwalbe: Wochbl. Papier-Fabr., 1913, 44, 3247. Both Richter 1 and Schwalbe 2 claim that the ether soluble portion is responsible for trouble with rosin spots. The total material contains both a fatty and a resinous constituent and Schwalbe has shown experimentally that neither constituent alone will produce rosin spots but that both together will give distinct and characteristic spots. The best means of avoiding rosin troubles seems to be seasoning the wood, which reduces the rosin soluble in ether and alcohol. Changes in the char- acter of the rosin extracted from the wood indicate that storage of chips for two to three weeks in the open air is as effective as storing the logs for two years. Richter 3 gives the following figures for rosin in cellulose from the same wood wet and dry. Rosin in cellulose Moisture in wood Ether soluble Alcohol soluble Total Per cent 49 ° 0.5 35-o 5° Per cent 1 .00 0.38 I-3I 0.88 Per cent O.II 0.26 0.31 0.36 Per cent 1 .11 0.64 1 .62 1 .24 Bleaching largely eliminates rosin troubles and with well- bleached pulp trouble from this source is seldom encountered. For the demonstration of rosin in sulphite fibre, Klemm 4 uses a strong solution of Sudan III in a mixture of three parts alcohol 1 Richter: Wochbl. Papier-Fabr., 1913, 44, 2486. 2 Schwalbe: Wochbl. Papier-Fabr., 1914, 45, 2926. 3 Richter: Wochbl. Papier-Fabr., 1913, 44, 4621. 4 Klemm: Wochbl. Papier-Fabr., 1911, 42, 967. MODIFIED SULPHITE PROCESSES * 197 and one part water. The sample is reduced to a pulp with water, drained and the moist fibres treated with the dye. The excess is removed by blotting paper and the fibres mounted for observation in water. The rosin will be found stained orange red while the fibre is uncolored. Modified Sulphite Processes. Numerous modifications of the sulphite process have been tried out and patented from time to time. It would be impossible to enumerate all of these but a few may be mentioned as showing the trend of modern investi- gations. Eichmann 1 first subjects the wood to the action of gaseous SO2 and then boils with sulphite liquor as usual. Moore and Wolf 2 charge the digester with chips and liquor as usual, close the head and then inject gaseous S0 2 , air being allowed to escape. The charge is allowed to stand without heat- ing to allow the SO2 to penetrate and steam is then passed in until the cook is complete. The injection of S0 2 gas is repeated during the steaming to make up for the dilution by condensation. Morterud 3 takes the liquor from under a false bottom in the digester, passes it through a heater and back into the top of the digester, thus maintaining good circulation and cooking with weaker liquors because there is no subsequent dilution by con- densation. This process is giving excellent satisfaction in the sulphate pulp industry and is claimed to be equally applicable to the sulphite process. Sammet and Merrill 4 have obtained a United States patent for cooking with gaseous ammonia, sulphur dioxide and steam instead of the usual liquor. Somewhat similar is the process of Tyborow- ski 5 who causes ammonia to react with sulphite liquor, thus precipitating calcium monosulphite, which is removed, and 1 German Patent 184,991, May 31, 1906. 2 U. S. Patent 1,119,977, Dec. 8, 1914. 3 German Patent: 286,074, Class 55b, Dec. 28, 1913. 4 Paper Trade J., March 14, 191 2, p. 46. 5 U. S. Patent 621,692, June 16, 1914. 198 THE SULPHITE PROCESS obtaining a cooking liquor containing ammonium sulphite and free ammonia. The claims for nearly all of these modified processes are much alike, viz.: shorter time, lower temperature, lighter colored, stronger and easier bleaching fibre, greater yield, etc. By-products and Waste Liquor. The sulphite process offers opportunities for the recovery of by-products which are now lost during the period of relieving or in the waste liquor. Bergstrom * states that the vapors condensed from the digesters yield an aqueous distillate containing methyl alcohol, acetone, aldehyde and traces of acetic and formic acids together with a brown oil floating on the surface. This oil contains 7 per cent boiling between 150 and 160 C, 55 per cent boiling from 160 to 190 C. and 17 per cent boiling between 190 and 210 C. The portion boiling between 160 and 190 C. consists largely of cymene. The specific gravity of the various fractions varies from 0.845 to 0.951. The same author 2 states that in the Ritter-Kellner process 8 to 10 kgs. of methyl alcohol are formed per ton of easy bleaching cellulose produced; of this about 3 kgs. may be obtained from the relieved gases. If the waste lyes are distilled in a continuous column apparatus the distillate contains methyl alcohol, acetaldehyde, acetone, oils and S0 2 as well as small quantities of formic and acetic acids. The problem presented by the waste liquors of the sulphite process is one which is not only interesting from a chemical standpoint but has also attracted much attention because of its bearing on stream pollution. In some cases, particularly in Europe, mills have even been obliged to close because it has been found impossible to purify the waste liquors sufficiently to com- ply with the legal requirements. While the industry in the United States is not confronted with quite such serious con- ditions it is only a question of time before much more complete purification will be demanded, so that the desirability of an early 1 Bergstrom: Papier-Fabr., 1912, 10, 359. 2 Bergstrom: Papier-Fabr., 1912, 10, 677. WASTE LIQUOR 199 solution is very evident. The present method of purification, if it may be called such, is merely to separate the fibres and neutral- ize the free acid with lime. The material thus removed is com- paratively small in amount considering that for every ton of fibre produced there is a ton of organic matter dissolved in the waste liquor. The total amount of this waste in the United States was estimated by the Geological Survey to be a billion pounds annually as long ago as 19 13. The discharge of such vast quantities of waste into streams renders the water injurious to health and makes it unfit for boiler use. It aids in the develop- ment of algae which may even grow in such quantities as to choke the streams. Under certain conditions it may cause the develop- ment of hydrogen sulphide with accompanying loss of oxygen in the water and consequent death of animal and vegetable life. Considerations of the nature outlined above as well as the desire to obtain useful and valuable products from a waste of such enormous magnitude have led many investigators to take hold of the problem. Walker x describes the waste liquor as a dark, reddish brown fluid of a specific gravity of about 1.05, and having a peculiar, not unpleasant odor. Among the constitu- ents present he mentions sulphur dioxide, sulphur trioxide, free sulphur, calcium and magnesium lignin sulphonates, pentoses and pentosans, mannose, dextrose, galactose, free furfural, traces of vanillin or vanillin-like body and small quantities of terpene- like substances. Waste liquors obtained in cooking hemlock wood have the following composition, according to Bryant : 2 Grams per liter Pounds per ton of pulp Total solids 115.00 105.36 9.64 7-83 0.76 2999 2748 251 204 20 Loss on ignition Ash Total sulphur Sulphur as SO3 The specific gravity of this liquor was 1.0425. 1 Walker: J. Soc. Chem. Ind., 32 (1913), 389. 2 Bryant: Paper, 1914, Jan. 28. 200 THE SULPHITE PROCESSS Klason 1 calculates that for every ton (2202 lbs.) of dry fibre produced the waste liquor contains the following: 600 kgs. (1320 lbs.) lignin. 200 kgs. (441 lbs.) sulphur dioxide combined with lignin. 90 kgs. (198 lbs.) CaO combined with lignin sulphonic acid. 325 kgs. (717 lbs.) carbohydrates. 15 kgs. (33 lbs.) proteins. 30 kgs. (66 lbs.) rosin and fat. According to Krause 2 the principal constituent in the waste liquor is the calcium salt of lignin-sulphonic acid. Ritter-Kellner liquor is darker and contains more furfural and generally more sugars than Mitscherlich liquor. Wood boiled in the autumn contained about twice as much sugar as wood obtained in the spring. Very careful analyses of liquor from autumn cut wood gave the following figures: Mitscherlich process Ritter-Kellner process Furfural Per cent O.OI 0.40 0.21 1.48 0.47 0.48 0.28 O.OI Per cent 0.02 Pentosans 0.29 0.49 1-47 0.41 0.48 Hexosans Total sugars Pentoses Mannose Levulose 0.25 O.OI Galactose Dextrose Trace According to Johnsen 3 the volume of liquor which can be obtained without special apparatus is 740 to 800 gallons per ton of pulp, while Haegglund 4 claims to obtain 960 gallons. Great differences of opinion also exist as to the rate of formation of fermentable sugar during the cook. Krieble 5 states that most of the sugar is formed before the end of the seventh hour and that 1 Klason: Papier-Fabr., 1909, 26, 627, 671, 703. 2 Krause: Chem. Ind., 1906, 29, 217. 3 Johnsen: Pulp Paper Mag. Can., 16, 1918, 314. 4 Haegglund: Pulp Paper Mag. Can., 15, 1917, 1185. 5 Krieble: Paper, 23, 1919, 753. WASTE LIQUOR 2 OI part of the fermentable material is destroyed if the temperature rises above 145 C. after that time. Haegglund, on the other hand, claims that only a little sugar is formed during the first six or eight hours but that it increases rapidly on longer cooking, the rate depending on temperature and composition of the cooking acid. Hoenig * claims that no organic acids except formic and acetic are present and that the ratio of these is 1 : 1.56. He finds 2.15 to 9.08 grams of volatile acid per liter. The waste liquor, according to Walker, 2 yields brominated and chlorinated products; it contains active carbonyl and methyl groups and is a strong reducing agent. On addition of alcohol the chief constituents are precipitated as a dark, gummy mass which becomes brittle on drying. This may also be obtained by salting out with sodium chloride or by treating with concentrated mineral acids or lead acetate. It is almost impossible to purify this substance because of its colloidal nature and its limited solubility in the usual organic solvents. The attempts to utilize the materials in this waste liquor have been very numerous and many patents have been issued covering all kinds of industries. A complete enumeration of all the patents in detail would occupy too much space and moreover it has been well covered by Miiller 3 up to the year 191 1 and also by Johnsen and Hovey. 4 A brief outline of the more important uses, or proposed uses, is therefore all that will be attempted here. As a binder the waste liquor, either in its original strength or concentrated by evaporation, has been tried for various purposes. As a road binder liquor at 1.13 sp. gr. has given very fair service when sprinkled upon the streets. While it is not water-resistant, the roads to which it has been applied appear to resist the action of rains fully as well as those to which crude oil has been applied. This use has only local interest because of the cost of transporting 1 Hoenig: Chem. Ztg., 19 12, 36, 889. 2 Walker: J. Soc. Chem. Ind., 32 (1913), 389. 3 Miiller: Literatur der Sulfit-Ablauge. 4 Bulletin 66, Dept. of Interior, Canada, 1919. 202 THE SULPHITE PROCESS the relatively dilute liquors. In preparing briquettes from waste coal it has met with some success, the briquettes being hard and making excellent fuel in ordinary grates or in smelting furnaces. An advantage which it possesses for this work is that the briquettes do not soften on heating and hence hold their shape well in use. On the other hand, the high percentages of ash and sulphur are detrimental in some cases. The briquetting of pyrites, wood waste, iron ores and other materials has also been successfully carried out by means of the concentrated waste liquor. Another use as a binder is in the iron foundry where it is mixed with the sand in preparing the moulds. Stutzer 1 gives the following as the composition of a sulphite waste liquor and two concentrated products made therefrom: Original waste liquor Wood extract Cell pitch Dry matter Ash constituents I2.l8 I.44 O.87 O.85 O.24 . 63.88 2 64 0.50 4.80 o.iS 82.79 14.90 Lime Total sulphur S0 2 8.50 5 87 0.85 In the preparation of concentrated products either iron or copper apparatus may be used, but if the former is employed the acid in the liquor must be neutralized by lime. Direct evapora- tion in copper is preferable if the product is to be used in tanning. A multiple effect evaporator is generally employed to bring the liquor to 35 Be. and the final concentration is performed on drums one of which will bring it up to 6o° Be. while a second will convert it to a solid. The pitch appears as a black opaque resin but is soluble in water. About 1 kg. of dry pitch is obtained from 10 kgs. of waste liquor and for each ton produced about a ton of coal is required. It has been proposed by Knosel 2 to prepare a fertilizer from the waste liquor by evaporating to about 25 Be. and mixing with 1 Stutzer: Papier Ztg., 1911, 36, 5. 2 Knosel: German Pat. 128,213. USES FOR WASTE LIQUOR 203 about an equal weight of ground Thomas slag. Analyses of this product show that practically all the phosphoric acid is in the citrate' soluble form. The use of sulphite waste liquor in the sizing of paper has been proposed by Mitscherlich 1 who mixed the liquor with gelatin solution and separated the precipitate formed. This was then dissolved in weak alkali and added to the paper stock in which it was precipitated by alum. A sizing process of another kind is that of Klason 2 in which the waste liquor is used instead of alum as a precipitant for silicate of soda. Neither of these processes has ever come into extensive use. Stutzer 3 has investigated the possibilities of waste liquor in the preparation of cattle feed and asserts that in each kilo, con- taining 120 grams per liter of solids, there are 550 calories which can be made available by feeding. His proposed treatment is to evaporate 100 liters to 50 liters in a vacuum, mix with 0.5 kg. of formaldehyde and ground limestone and then filter. The filtrate, after further evaporation, is mixed with molasses and 6.25 kgs. of peat to give 45 kgs. of cattle food. In the dyestuff industry it has been used as the basis for the manufacture of sulphur dyes, in the reduction of indigo and in the preparation of indanthrene and similar dyes. The sodium Hgnin sulphonate prepared from it has been employed to replace tartaric acid in mordanting wool. Strehlenert 4 has proposed the following method for the prepa- ration of lignin and the recovery of sulphur dioxide. A little acid sodium sulphate is added to the fresh hot liquor and the calcium sulphate which precipitates is separated. The hot liquor is then run into digesters and heated to ioo° C. ; air is pumped in until the pressure reaches 18 atmospheres, when the temper- ature rises 20 because of the chemical reactions taking place. Heating is continued to 160 C. Between 160 and 170 C. is 1 Mitscherlich: German Pat. 54,206, 1890. . 2 Klason: Z. angew. Chem., 22 (1909), 1423. 3 Stutzer: Z. angew. Chem., 22 (1909), 1999. 4 Strehlenert: Papier-Fabr., 1913, II, 645, 666. 204 THE SULPHITE PROCESS the critical point at which decomposition of the lignin sulphonic acid begins. This reaction causes the temperature to rise about 2o° more and it is finally forced up to 200 C. The time after reaching ioo° C. is about 40 to 60 minutes. If properly con- ducted this procedure causes the evolution of sulphur dioxide, which can be recovered, while the lignin is precipitated in granular form and can be used as fuel after partial drying. The sulphur recovery is claimed to be 25 to 30 kgs. (55 to 66 lbs.) per ton of pulp and the lignin enough to supply the entire fuel requirements of the pulp mill. The use of the dried waste liquor as a fuel is suggested be- cause of the large amount of combustible matter which it con- tains. The dry material is light, powdery and has a heating value of about 6000 B.t.u. per pound. ■ Experiments in burn- ing this in the same manner as powdered coal 1 gave excellent results and the ash formed in easily accessible places and showed no tendency to fuse. Rinman 2 mixes the waste liquor with lime to make 22 to 25 grams per liter of calcium oxide and boils first for five hours at a low temperature and finally at 180 C. The precipitate of cal- cium sulphite and humus is filtered off and treated with sulphur dioxide to recover the calcium bisulphite. The alkaline filtrate is evaporated to 40 Be., more lime is added, the mass evapo- rated to dryness and finally destructively distilled in presence of steam. The products of this latter process include acetone and low and high boiling oils. The preparation of a material for use in making insulating substances or artificial leather is patented by Trainer. 3 The waste liquor is evaporated to 30 Be. and then heated with an acid, preferably after adding an aldehyde such as formaldehyde. Extracts for use in tanning leather are prepared in consider- able quantities by processes involving neutralization with lime, concentration, and subsequent separation of the organic and 1 Paper, Dec. 19, 1917. 2 Papier Ztg., Apr. 4, 1915. 3 Trainer: German Pat. 197,195, Feb. 20, 1906. . USES FOR WASTE LIQUOR 205 inorganic materials. The extracts obtained contain practically no true tannins but do contain materials which are taken up by the skins and act as fillers. While these extracts are not suit- able for tanning alone they find a legitimate use as additions to other true tanning substances. The preparation of alcohol l from waste lyes has probably attracted more attention than any other method of utilization and a number of commercial plants are already in successful operation. The two principal processes now in use are the Swedish, which is a combination of the Wallin and Ekstrom proc- esses, and the Norwegian or Landmark method. The principle of all processes is the fermentation of the sugars present followed by the distillation of the alcohol formed, and one of the prin- cipal difficulties encountered has been due to the poisoning of the yeast by traces of sulphur dioxide. The Swedish process uses a tempered yeast which is capable of resisting this action. The liquor is first neutralized by calcium carbonate arid the last traces of acidity by calcium hydroxide, it is then cooled, settled and run to the fermentation vats where the yeast is added. It is fermented at 27 C. for four or five days and then distilled. The raw alcohol contains 92 to 93 per cent of ethyl alcohol, 3 to 4 per cent methyl alcohol and small amounts of cymol, acetone and aldehyde. The yield of 100 per cent alcohol by this process is said to be 74 liters per ton of dry sulphite, and the cost about 12 cents per U. S. gallon 180 proof. In the Norwegian process the fermentation is aided by a nutrient and easily fermentable medium prepared from milk or whey. To the milk an equal volume of sulphite liquor is added and a small amount of muriatic acid and the precipitated ligno-casein filtered off. The filtrate is then added to the waste liquor and the mixture evaporated to a concentration of about 15 per cent. It is next neutralized by powdered limestone, cooled to 2 7 C. and the yeast added; in this case ordinary 1 Tartar: J. Ind. and Eng. Chem., 1916, 226; Hedalen: Pulp Paper Mag. Can., 1916, 176; Kiby: Chem. Ztg., 1915, 39, 212 et seq.; Segerfelt: Papier. Ztg., 53, 2518, 2558. 206 THE SULPHITE PROCESS brewers' yeast has proved entirely satisfactory. After four or five days' fermentation it is ready to be distilled. The yield by this process is claimed to be 91.2 liters of 100 per cent alcohol per ton of dry sulphite, and the cost about 9 cents per U. S. gallon 180 proof. According to McKee l the so-called poisoning of the yeast is not due to sulphur dioxide but to the lack of oxygen. He has patented a process in which the hot waste liquor is cooled by blowing air through it and then placed in closed fermenta- tion tanks where it is kept Agitated during the fermentation period by a slow current of air. Under these conditions fer- mentation proceeds satisfactorily with ordinary yeast even in the presence of very considerable amounts of sulphur dioxide. Loss of alcohol is prevented by scrubbing the exit gases from the fermentation tanks with unfermented liquor. According to Stalnacke 2 three mills in Sweden were produc- ing in 1916 about 660,000 gals, of 100 per cent alcohol annually, while the total possible production was about 6,868,000 gals. Considering the equipment necessary he states that for a mill making 55 tons of cellulose the fermentation vats should hold 264,000 gals, and the distilling apparatus should be capable of handling about 2400 gals, per hour. While the preparation of alcohol from the waste liquor is a profitable undertaking it does not solve the problem of the dis- posal of the liquor since the spent fermentation residues are nearly as objectionable as the original waste. Other substances which it has been proposed to recover from the waste liquor are antiseptic materials, calcium sulphite, cal- cium sulphate, coniferin, cymol, acetic acid, furfural, levulinic acid, oxalic acid, sulphur, turpentine, lignorosin, vanillin, etc. Many of these can be obtained only in small amounts and the demand and the prices obtainable are not such as to make the undertaking attractive. 1 McKee: Paper Trade J., 1918, Aug. 22, p. 42. 2 Stalnacke: Paper, Apr. 5, 1916. TESTS AND ANALYSES 207 While much progress has been made there is still a tremen- dous amount of work to be done before the utilization of this enormous amount of waste can be considered satisfactory. The quantity of material to be handled would seem to indicate that such investigations should be directed towards the production of large amounts of products at small profits rather than the preparation of small quantities of substances of relatively high value per pound. Tests and Analyses for the Sulphite Process — Sulphur. Sulphur as obtained at present on the American market is of a very high degree of purity but it is nevertheless well to examine it occasionally. Moisture may be determined by weighing a 3 to 5 gram por- tion into a weighing bottle, drying to constant weight at 70 to 8o° C, cooling and again weighing. The heating should not be prolonged beyond the time necessary to reach constant weight and the temperature must not be allowed to rise too high. To determine non- volatile substances a 10 gram sample is cautiously heated in a porcelain crucible on a sand bath until nearly all of the sulphur has volatilized (ignition of the sulphur must not take place). Cover the crucible with a perforated lid and pass into it pure, dry hydrogen gas until all sulphur has escaped; cool and weigh as non- volatile matter. The inorganic ash may be determined by igniting the non- volatile matter with access of air and reweighing after cooling. Selenium may be determined according to the method of W. Smith 1 as follows: Weigh 30 to 50 grams of the finely ground sulphur into an Erlenmeyer flask, add a few cubic centi- meters more bromine than the grams of sulphur used and allow it to stand 15 minutes. Transfer to a 100 c.c. separatory funnel and shake vigorously with 40 c.c. of bromine water for one minute. Separate the sulphur bromide from the aqueous solu- tion and pour the latter through a wetted filter paper. Add 1 W. Smith: J. Ind. Eng. Chem., 1915, 7, 849. 208 THE SULPHITE PROCESS about 2 c.c. of bromine and 40 c.c. of bromine water to the sulphur bromide, and repeat the extraction four times, keeping the last extract separate. Treat the last extract in the same way as the combined extracts, using proportionate parts of potassium iodide and hydrochloric acid and if the presence of selenium is proved repeat the extraction as often as necessary. The combined extracts are boiled till clear and any remaining free bromine is removed by careful additions of powdered potassium metabisulphite or sulphite until the solution just becomes colorless. Dilute to about 250 c.c, add 15 c.c. of hydrochloric acid and about 5 grams of potassium iodide and boil; this completes the precipitation of the selenium, and gradually converts the red to the black form. The free iodine is removed by a few cubic centimeters of potassium sulphite solution; the solution is boiled for 20 minutes, filtered through a tared Gooch crucible, the selenium washed with hot water, and dried at ioo° C. until constant in weight. Burner Gases. In order to control the burners properly the gases should be tested once an hour, or oftener if the burners are not working satisfactorily. The sulphur dioxide may be determined by means of an Orsat apparatus using caustic soda in the absorption pipette. The gas is measured in the burette over water which soon becomes saturated with S0 2 and then introduces only a very slight error. Excess air may be estimated by determining the oxygen by means of alkaline pyrogallate after first absorbing the SO2 in caustic soda. This test gives valuable information regarding leaks in the apparatus between the combustion chamber and the absorption system. Sulphur trioxide in the gases is very difficult to determine with accuracy. Richter 1 passes the gas, at a rate of 1000 c.c. in 20 to 25 minutes, first through a hard glass sampling tube surrounded with an iron jacket and then through a tube 30 cms. long which is filled with garnets and bits of porcelain and cooled 1 Papier-Fabr., 11, 1913, p. 610. TESTS AND ANALYSIS 209 with ice. The gas is measured by the amount of water de- livered by the syphon bottle which is used to induce its flow. After passing 2 to 5 liters of gas the tube is washed out by drawing pure air through it and finally is washed into a beaker with water to remove the S0 3 . This is then determined gravi- metrically by precipitation as BaSCV The generally accepted place to sample the gases is in the main pipe between the cooler and the absorption apparatus. Besides these chemical tests it is customary to record the temperature of the gases as they leave the combustion chamber and again after passing the cooler. Some form of recording pyrometer is desirable for the former work while an ordinary chemical thermometer is satisfactory for the latter. "Acid" or Bisulphite Liquor. Samples of the liquor delivered by the absorption apparatus are generally tested once an hour and the liquor in the storage tanks whenever a digester is filled. For total S0 2 a 1 c.c. sample is taken by means of a pipette and diluted with 200 to 300 c.c. of water in a white porcelain N bowl. A few drops of starch solution are added and then — 10 iodine solution is run in until a blue color appears. N For free, or available, S0 2 a 1 c.c. sample is titrated with — 10 caustic soda, using phenolphthalein as an indicator. The combined S0 2 is the difference between the total and free S0 2 . The tests given are those generally made by the mill foremen and are intended to give comparative rather than absolute results. They are subject to errors because of the small size of the sample, and because the pipette is not usually washed out and hence a little of the liquor always remains adhering to the glass. It is customary to read from a chart the amount of either free or total S0 2 corresponding to the volume of caustic soda or iodine used. This is generally spoken of as per cent, but it is not exactly that as the test made in this way takes no account of the specific gravity of the liquor. 2IO THE SULPHITE PROCESS For accurate analyses of the liquor a sample of 10 c.c. should be diluted with recently boiled and cooled water to ioo c.c. and 10 c.c. of this used in the tests. Sulphates in the liquor may be determined by placing 10 c.c. in a covered beaker, adding an excess of strong hydrochloric acid and boiling for some time until the odor of S0 2 is no longer noticeable. The sample is then diluted and the S0 3 determined gravimetrically as BaS0 4 . Bases may be determined by the following method given by Griffin and Little. 1 Ten to twenty cubic centimeters of the acid are treated with H 2 S0 4 in slight excess in a platinum dish, and evaporated to dryness. The dry mass is cautiously ignited until no more fumes come off, and the residue cooled and weighed as sulphates of calcium and magnesium. Treat the sulphates with 5 c.c. of water and one or two drops of hydrochloric acid and break up all lumps with a stirring rod. Rinse into a beaker with as little water as possible, add one or two drops of strong sulphuric acid and alcohol equivalent to twice the volume in the beaker. Allow it to stand with occa- sional stirring for an hour or more, then filter off and wash the precipitate first two or three times with 60 per cent alcohol and finally with 40 per cent alcohol as long as anything is removed by the treatment. The residue on the filter is dried, ignited, and weighed as pure calcium sulphate. Selenium. As a practical test for the presence of selenium in injurious quantities, Klason and Mellquist 2 seal up some of the liquor in glass tubes, from which air has been expelled by carbon dioxide, and heat for 15 hours at 13 7 C. in a bath of boiling xylene. If appreciable amounts of selenium are pres- ent, the composition of the liquor will be changed by this procedure. Waste Liquor. At the end of the cook the liquor in the N digester is tested for total S0 2 by means of — iodine. The 10 1 Chemistry of Paper Making, p. 414. 2 Klason and Mellquist: Papier-Fabr., 19 13, 145. TESTS AND ANALYSIS 211 sample is generally drawn from about one-third the way down the digester and no attempt is made to retain the S0 2 by cooling the liquor as drawn. Of this hot liquor a 5 c.c. sample is gen- erally tested. For the determination of free S0 2 in waste liquor, Stutzer 1 recommends the following: Place 25 c.c. of the liquor in an Erlenmeyer flask, add 25 c.c. of standard alkali, then 1 gram of ammonium chloride or nitrate and immediately connect with a condenser. Boil for just 20 minutes and catch condensate in 25 c.c. of standard acid. The acid not neutralized by this treat- ment is equivalent to the free acid in the waste liquor. 1 Stutzer: Papier Ztg., 1911, 36, 5. CHAPTER VII GROUND WOOD OR MECHANICAL PULP The preparation of ground wood pulp brings in comparatively little of a chemical nature yet its close relationship to the rest of the industry makes it desirable to include a discussion of the methods and general principles involved. The commercial man- ufacture of ground wood is generally not conducted according to any fixed standards of practice, each superintendent or man- ager having his own theories about the best methods of operating. For this reason little reliable information was available until the Forest Service undertook the collection of data and it is upon their results x that much of the present chapter is based. The present method of manufacture has been in use for a long time and except for increased size and capacity of grinders it has changed but little since its introduction in 1867. The process consists briefly in pressing blocks of wood against the surface of a grindstone which is also supplied with water to remove the pulp as fast as it is made. The stones are usually about 54 ins. in diameter by 27 ins. face, but in some recent installations have been as large as 60 ins. in diameter by 48 ins. face. Up to a few years ago natural quarried stones only were used, but many mills are now experimenting with artificial stones. The wood to be ground is placed in pockets in the housing of the stone in such a way that the logs are parallel with the shaft on which the stone is mounted. Pistons operated by hydraulic pressure then force the wood against the stone until it is reduced to pulp. The pockets are tjien recharged and the process re- 1 Forest Service Bulletin, Experiments with Jack Pine and Hemlock. Forest Service Bulletin 127, Grinding Spruce for Mechanical Pulp. Forest Service Bulle- tin 343, Ground-Wood Pulp. 212 GROUND WOOD OR MECHANICAL PULP 213 peated. There are generally three or four pockets for each grinder and as they seldom become empty all at once they are filled as necessary and the process is thus continuous. There is now on the market a magazine grinder equipped with two pockets into which the wood is fed automatically. The Voith Magazine Grinder, shown in part sectional elevation Fig. 31. Three-pocket Grinder (1) Grindstone (2) Shaft (3) Steel flanges (4) Casings or side frames (5) Bridge-trees (6) Studs to support (7) Pockets (8) Hydraulic cylin- ders (9) Piston rods (10) Pressure foot in Fig- 33 is of this type. The advantages of this grinder are said to be (1) increased capacity, (2) charging is automatic and not subject to the irregularities of manual feeding, (3) constant load on motor or turbine and (4) decreased cost of attendance. The pulp coming away from the stones collects in pits under the grinders and from these pits it flows to screens with one- fourth to three-fourths inch perforations which remove slabs, knots, large splinters, etc. For satisfactory operation of these 214 GROUND WOOD OR MECHANICAL PULP screens the stock should be diluted to at least i per cent dry matter. The stock passing these coarse screens goes next to the regular screens of centrifugal or diaphragm type, and for this operation it should be still further diluted to about 0.25 Fig. 32. Three-pocket Grinder Sectional Elevation per cent dry. It is desirable that the diluted stock be passed through a riffler or over a sand settler before going to the screens as much wear on the latter is thus avoided. The screened stock is then thickened to the proper consistency for use in the beaters by means of filters, wet machines, or some other equiva- GROUND WOOD OR MECHANICAL PULP 215 lent device, or if it is to be sold in the form of "laps" these are made on a wet press. The chief factors which enter into the production of mechani- cal pulp from any species of wood are: (1) Surface of stone; whether rough or smooth, sharp or dull, or of coarse or fine grit. Suspension IJin for swinging magazine out to narg stone Charging Floor -" L C ' ' "c , \-a~- J Fig. 33. Voith Magazine Grinder (2) Pressure employed in forcing the wood against the re- volving stone. (3) Peripheral speed of the stone. (4) Temperature of grinding and thickness of stock in the grinder pit. (5) Physical condition of the wood. 216 GROUND WOOD OR MECHANICAL PULP The condition of the surface of the stone depends on several factors among them being the size and sharpness of the indi- vidual particles of grit, the ease with which the binding mate- rial wears away and the manner of dressing the stone. This latter operation is performed by working across the face of the stone small steel rolls or burrs of various designs which roughen the surface and form depressions through which the ground wood can escape. A great deal of attention has been given to the designing of these burrs but it appears that practically the same quality of pulp can be obtained under like conditions of pressure, speed and temperature if the surface is brought to the same condition of sharpness of grit, regardless of whether the design of the markings is straight cut, spiral or diamond point. The important thing, so far as quality is concerned, is to give the particles of grit the correct treatment, rather than to form a deeply grooved surface. The horsepower per ton of pulp varies inversely with the sharpness of the stone, while the production varies directly with the sharpness. Immediately after sharpening a stone, therefore, the rate of production is high and the power con- sumed low, while as the stone becomes dull the former de- creases and the latter increases. The condition of the surface of the stone appears to have very little influence upon the yield per cord unless it has been made so extremely sharp that more screenings are formed and possibly more fine fibre lost in the white water. With such a surface the fibres are actually ground to pieces and in some instances they are so short and fine that it is almost impossible to remove the lap from the wet machine press rolL Deep grooving of the surface of the stone causes more rapid production of pulp but at the expense of quality, while better pulp is produced by a less sharp stone and a greater application of power. Paper prepared from this latter pulp has greater strength than that from pulp ground on very sharp stones. Next to the surface condition of the stone the factor most influencing quality is the pressure at which the wood is forced GROUND WOOD OR MECHANICAL PULP 217 against the stone. For any given cylinder pressure this varies greatly with the length and diameter of the logs, and further variations are caused by the binding of the wood in the pockets and by fluctuations in the water pressure when the pistons are raised or lowered. The result of increasing the pressure is to increase the power required by the grinder and decrease the power consumption per ton of pulp made. This latter effect is less noticeable on sharp than on dull stones. This result is interesting because it suggests that by carrying a high pressure and using only part of the pockets the power con- sumption per ton can be reduced, or in other words these con- ditions permit the production of a larger quantity of pulp during times of low water, without sharpening the stone to an unusual degree. It has been found that the yield of pulp per 100 cu. ft. of solid wood increases with increase of pressure. The screenings also increase, but not so fast as the total yield so that there is a net gain of good fibre. The strength factor, or the bursting strength per square inch divided by the weight per ream, which indicates the quality of the pulp, decreases quite rapidly with increasing pressure. The peripheral speed of the stone is given little attention in most commercial plants. When the pressure on a pocket of the grinder is removed the speed increases greatly, which counteracts to a certain extent the decreased production due to the smaller number of pockets in use. While this is rather beneficial than otherwise there are conditions of operation which require a fairly constant speed, and the use of a governor is therefore desirable. As would be expected the power to the grinder varies directly with the speed; this is also true, and to an even greater extent, of the production in twenty-four hours. With constant power to the grinder the production in twenty-four hours is practically constant, regardless of whether the pulp is produced at low pressure and high speed or at high pressure and low speed. The strength of the paper is greater with pulp produced at high pres- sure and low speed than with that made at low pressure and high 218 GROUND WOOD OR MECHANICAL PULP speed. The yield per cord and the quality of the pulp are only slightly influenced by the speed. The effect of the temperature at which mechanical pulp is produced has long been a controversial point between European and American manufacturers. The general American practice is to operate at high temperature and it is claimed that pulp so produced has longer and stronger fibres, is considerably tougher than cold-ground pulp and works "freer" on the paper machine. Cold-ground pulp, on the other hand, is said to be finer, more free from shives and to give a better closed sheet of greater opacity than hot ground pulp. Another factor which is claimed to have an important influence on the paper produced is the thickness of the pulp in the grinder pit, and in actual operations this varies from extremely thick to comparatively thin. Investigation of these two points shows that varying the tem- perature from hot to cold has little effect upon the power con- sumption or power to grinder, but the production in twenty-four hours is somewhat higher when grinding hot. With thick stock in the grinder pit the power required to rotate the grinder with- out load is greater than with thin stock, but the difference, when calculated to the basis of power consumption per ton of pulp, becomes negligible. Neither of these two factors of temperature and thickness of stock influences the yield per cord of wood. The quality of pulp, however, is affected, that produced at high tem- perature being long-fibred, while a fine-fibred stock is more easily secured by the cold grinding process. The physical condition of the wood, apart from any changes induced by boiling or steaming, has a very appreciable influence on the results obtained. For green wood the average power con- sumption is lower than for seasoned wood while the rate of pro- duction is higher. The diameter of the bolts used and the rate of growth of the wood have very little effect upon either the power consumption or the rate of production. Rapidly grown wood, as compared with that which has grown slowly, yields considerably less pulp which is softer though of about the same GROUND WOOD OR MECHANICAL PULP 219 strength as that from slowly grown trees. It is generally recog- nized that green or freshly cut wood gives a better product than seasoned wood, and in the case of white fir McNaughton 1 has shown that pulp from young trees 18 ins. or less in diameter is whiter and stronger than that from old trees of about 40 ins. diameter. The commercial efficiency in converting rossed wood to pulp under ordinary conditions averages about 88 per cent. Of the remaining 12 per cent about 2 to 7 per cent is lost as screenings and in the white water as wood fibre, while the remaining 5 to 10 per cent must be in the white water as water soluble organic or inorganic materials. The changes induced by boiling or steaming the wood before grinding very profoundly influence the product obtained. The color of the pulp is darker than that from unsteamed wood and the fibres are much longer and better separated, resulting in a stronger product. The changes in color and physical character of the pulp are practically identical, provided the temperature and duration of the cooking are the same, whether the logs are steamed or boiled while immersed in water. Steaming has the advantage over boiling in that less heat is necessary and the con- densed liquors are drawn off in concentrated form which is a ben- efit where recovery of by-products is attempted. As soon. as steaming starts the formation of acid commences and increases up to the end of eight hours treatment. Both acetic and formic acids are produced, in the ratio of six acetic to one formic. Spruce gives 0.213 per cent of acetic acid and pine yields similar amounts. After two hours steaming reducing sugars appear and eventually amount to about 0.25 per cent of the dry wood. The pressure and duration of steaming are important factors to control since they have a great influence on the color, strength and yield of pulp. Increasing both time and steam pressure increases the strength of the pulp but makes it much darker in color while at the same time the yield is much decreased because of the greater solvent power of the water. 1 Paper, Nov. 1, 1916, p. 13. 220 GROUND WOOD OR MECHANICAL PULP A study of the various factors in the grinding of steamed spruce wood has brought out a number of very interesting facts. The power required per ton of pulp is at least 25 per cent greater than that used in grinding untreated wood, and the maximum power per ton is reached when the wood is cooked for six hours. This holds true for cooking pressures between o and 75 lbs. gauge pres- sure. With the same length of cooking, wood which is treated at high pressure requires more power per ton of pulp than that which has been cooked at lower pressures, while with a fixed amount of power to the grinder the amount of pulp produced is less at high pressure and low speed than it is at low pressure and high speed. The pulp made by grinding steamed wood can be used for different purposes, depending on the nature of the grinding process. If a sharp and coarse stone is used a large number of shives will be present and the pulp will serve for the manufacture of box boards or similar products. When ground to a finer state, and mixed with a small amount of chemical fibre a bogus kraft paper can be produced which will serve for a cheap wrapping paper. Tests on papers made from steamed and unsteamed woods show that the steamed pulps give a higher percentage of stretch than the unsteamed even though the latter are mixed with 20 per cent of sulphite spruce fibre. Like chemical pulps steamed ground wood is considerably influenced by beating treatments and variations in the latter cause marked variations in the strength of the paper. With prolonged beating the paper becomes more brittle but gives higher strength tests. The boiling or steaming of wood results in the formation of a natural size from some of its constitutents and this sizing action is particularly noticeable in the production of pulps from the hardwoods — birch and aspen — which are not naturally pitchy. All paper produced from cooked woods, pulped by the mechanical process, shows the characteristic water-resistant qualities and hardness of hard sized papers. While spruce is the standard wood for the manufacture of mechanical pulp the supply of this wood is decreasing so rapidly GROUND WOOD OR MECHANICAL PULP 221 that some substitute must be found. With this end in view a long series of tests has been made on a practical scale by the Forest Service * and the pulps made run into paper and tested for printing qualities on newsprint presses. The following table gives the common and scientific names of the woods used, the yield of bone dry fibre from ioo cu. ft. of solid rossed wood and the color rating, No. i being the best color and No. 23 the poorest. Common name Scientific name Bone dry fibre per 100 cu. ft. Color rating Balsam fir Red fir White fir Alpine fir Amabilis fir Grand fir Noble fir Eastern hemlock. . . . Western hemlock. . . Tamarack Western larch Lodgepole pine Lodgepole pine Western yellow pine. Jack pine Loblolly pine White pine Engelmann spruce . . Sitka spruce White spruce Aspen White birch Black gum Abies balsamea. . . . ' ' magnified .... " concolor " lasiocarpa. . . ' ' amabilis ' ' grandis ' ' nobilis Tsuga canadensis . . . " heterophylla. . Larix laricina " occidentalis . . . Pinus murrayana l . . ' ' murrayana 2 . . " ponder osa. . . . " divaricata. . . . ' ' taeda " strobus Picea engelmanni . . . ' ' sitchensis ' ' canadensis .... Populus tremuloides . Betula papyrifera. . . Nyssa sylvatica Lbs. 1910 1915 2000 2060 1870 i95o 1920 2030 2160 2620 2100 1920 2140 2060 2200 2500 1885 2000-2250 2100 2400 2200 2950 2600 9 20 10 1 13 5 15 19 22 21 23 14 7 16 17 4 1 Wood from California. 2 Wood from Montana. In these trials very little difficulty was experienced in pro- ducing pulp from the woods tested. With the conifers grinding could be done under practically the same conditions employed for spruce. All the substitutes, with the possible exception of noble fir and amabilis fir, require the use of more power per ton of pulp than does spruce. The best results were generally obtained by grinding on a somewhat dull stone with high pressure and rather slow speed. 1 Bull. No. 343, U. S. Dept. of Agriculture. 222 GROUND WOOD OR MECHANICAL PULP Of the woods tried all of the firs yield pulp suitable for news- print purposes; hemlock gives a short fibre and much care is necessary in grinding; western hemlock is much superior to eastern. Tamarack gives a good quality of pulp except for its color which is grayish green, while western larch yields a very inferior pulp, shivey and of poor color. The pines yield pulp which could be used for newsprint work though there is a tend- ency toward softness. The one exception is loblolly pine which gives an inferior pulp which would find use only as a filler. Of the hardwoods aspen gives a satisfactory pulp if a large amount of power is employed in grinding. When mixed with spruce it operates very well. White birch yields a short, but very fine fibre, which has a pinkish tinge; it could be used as a filler in certain grades of paper. Black gum gives a fibre resembling in many ways that from white birch. It is very short but forms a tougher sheet than coniferous fibres of the same length. This pulp is not promising for newsprint paper but could be used as a filler or mixed with pulp of a better grade. The ground wood process has received much attention from investigators with the object of producing fibre which could be used in making news paper without the addition of sulphite, Bache-Wiig * has patented a process whereby the blocks of wood are heated in a solution of salt and then ground as usual. As, in many cases, the salt does not penetrate the entire block the resulting pulp is a mixture of untreated and treated fibres. The claims for this process are that less power is required and the fibre is longer and stronger and of better color. According to another process 2 the wood blocks are placed in a digester which is partially evacuated, then treated under pressure with sulphur dioxide and finally cooked with water, salt solution or bisulphite liquor. Wood thus treated gives, on grinding, a pulp suitable for making news paper without the addition of sulphite. Henckel 3 proposes cooking the logs with a caustic soda solu- 1 U. S. patent 913,679. 2 Bache-Wiig: Paper, 1916, No. 21, p. 18. 3 Henckel: Austrian patent 34,816. GROUND WOOD OR MECHANICAL PULP 223 tion of 3 to 5 Be. for three or four hours under pressure and then grinding as usual. The claims for this process are better fibre, less power and greater yield. It is difficult to see how the yield can be increased by treating with a solution having such strong solvent powers as a caustic soda solution. A modification which has recently attracted much attention is that for producing white ground wood by the Enge process. 1 According to this process the logs are placed in a boiler which is then completely filled with water. The temperature is raised by direct or indirect steam to 176 to 257 F., steam is shut off and hot or cold water pumped in to raise the pressure to 147 lbs. or more, which is maintained for five to eight hours. The higher the pressure carried the higher can the temperature go; good conditions are 230 F., and six hours at 147 to 176 lbs. per square inch. Following this treatment the grinding is completed on ordinary stones as usual. The advantage of this process is somewhat doubtful, for although the pulp can be made into news paper alone, yet the extra cost for steaming, labor, etc., just about counterbalances the saving in cost of sulphite fibre. With any method of grinding close control of the process is essential if a uniform product is to be made. A very simple method for obtaining such control is to examine the image of the fibres when thrown on a screen by means of a lantern. The pulp may be mixed with a little aniline dye so that a sharper image may be obtained. A little of the pulp is then placed between two glass plates in such a density that the individual fibres may be observed, and its image thrown upon a white screen. The best magnification is about 40 diameters and when enlarged to this extent it is very easy to tell the relative proportions of "fluff," long fibre, shives, slivers, etc. This method is in operation in one of the most progressive mills in Canada and is very highly recommended. Samples from each stone are examined every two hours and the condition of the stone and the work it is doing are thus accurately known, making it possible to sharpen the 1 German patents 20,860 and 20,932, E VII, 55a. 224 GROUND WOOD OR MECHANICAL PULP stone when it is necessary and keep the pulp much more nearly uniform. Another control test which is in actual use and is giving very good results is that of the sediment tester described by Fishburn and Weber. 1 In this test a mixture of 5 grams bone dry pulp and 500 c.c. of water is placed in a graduated tube with a woven wire bottom and the time required to drain down to a definite mark is noted. The standard time for news grade wood-pulp is 80 seconds and if it is ground in such a way as to give a test of 70 seconds trouble is experienced on the paper machine. The bleaching of wood pulp cannot be performed by hypo- chlorites or other oxidizing agents since all the lignin and other incrusting matters are still present. Its color can be consider- ably improved by treating with sulphurous acid or a bisulphite. This is regularly done in European practice by moistening the pulp with a solution of sodium bisulphite as it is wound up on the press roll of the wet machine. According to Schutz 2 about 2 to 3 per cent is used. The action requires time and the color is not so bright as that of bleached chemical pulp. All woods are not equally susceptible to improvement in color by this process; hemlock and tamarack, for instance, are not so good as poplar and spruce. 1 Paper, Oct. 11, 1916, p. 13. 2 Schutz: Paper, Feb. 4, 1917, p. 64. CHAPTER VIII BLEACHING All of the commercial processes for isolating the fibrous cellu- loses for paper making fail to produce a perfectly pure material since it is always associated with a small amount of the in- crusting matter originally present in the raw material and gen- erally also with coloring matters, which either escaped destruc- tion during the process or were formed by it. The object of bleaching is the destruction or removal of such undesirable impurities so that the natural white color of the pure cellulose may become evident. The process of bleaching is essentially one of oxidation and the success attained depends on the fact that the accompanying impurities are attacked and resolved into soluble products much more easily than the comparatively inert cellulose of which the impure fibre is largely composed. Many different oxidizing agents can be utilized in the bleach- ing process and a number of them have been applied with more or less success, but practically all commercial work is performed by chlorine, or some of its compounds, which in the presence of moisture tend to liberate oxygen. Chlorine. Chlorine is a greenish yellow gas, darkening in color as the temperature rises. It has a pungent and irritating odor and cannot be inhaled as it attacks the membranes of the throat and nose. Its atomic weight is 35.457 and a liter of it at o° C. and 760 mm. pressure weighs 3.1 691 grams. One vol- ume of water at 15 C. and 760 mm. pressure absorbs 2.37 volumes of chlorine. At 15 C. and 6 atmospheric pressure it is converted into a clear yellow liquid of specific gravity 1.33, which is not miscible with water. When perfectly dry it does not attack iron, which enables it to be stored and shipped in 225 226 BLEACHING wrought-iron cyclinders. Dry chlorine is also devoid of bleach- ing properties as may be shown by passing the dry gas over a piece of litmus paper, or a cloth dyed a delicate shade; so long as moisture is absent no bleaching action takes place, but on the addition of water the color is at once destroyed. Gas Bleaching. In bleaching with chlorine gas the material to be treated must be placed, while in a moist condition, in a receptacle which is capable of being tightly closed. It is then subjected to the action of the gas, either generated by means of manganese peroxide or obtained from some other source. If the material to be bleached is very shivey, gas bleaching mate- rially assists in the production of clean stock as it tends to resolve the woody matter more completely than does bleaching powder solution. It is also very efficient in removing metallic particles. Gas bleaching has in the past been applied to rag and rope stock, wood pulp, straw, etc., but little information is available to show how much of the gas is actually used up in the bleaching. Beadle and Stevens 1 cite a case of a cotton rag half-stuff which normally required 12 per cent of bleaching powder but in which a better color was produced by treatment with 2 per cent of chlorine followed by 2 per cent of bleach. In this case 2 per cent of chlorine replaces 10 per cent of bleach- ing powder, or is equivalent to 3.5 per cent of chlorine in the form of bleaching powder. Experiments by the author on a sample of soda hemlock which could not be satisfactorily bleached with hypochlorite solution showed that 4 per cent of chlorine followed by 5 per cent of bleach gave a good color and that in this case 2 per cent of chlorine was equivalent to 3 per cent in the form of hypochlorite. In spite of its good bleaching efficiency the process is seldom used because of the difficulty of maintaining tight apparatus and the general in- convenience involved; it is never employed in this country. Hypochlorites. When chlorine is passed into an aqueous solution of an alkali, or alkaline earth, a hypochlorite or hypo- 1 J. Soc. Chem. Ind., 1914, p. 727. HYPOCHLORITES 227 chlorous acid is formed according to the equations: 2 KOH + 2 CI = KOC1 + KC1 + H 2 KOH + 2 CI = KC1 + HOC1. If it is passed into a suspension of calcium carbonate, hypo- chlorous acid only is formed: CaC0 3 + H 2 + 4 CI = CaCl 2 + C0 2 + 2 HOC1. The first bleaching compound known, eau de Javel, was made in 1789 at the Javel works near Paris by passing chlorine gas into a solution of crude potassium carbonate. In 1798 Tennant patented a bleach liquor prepared by passing the gas into caustic soda or milk of lime and this method is still very largely employed where chlorine can be produced cheaply and the bleach solution used on the spot. According to Higgins 1 hypochlorites and hypochlorous acid bleach because of their readiness directly to produce oxygen and to a lesser extent because of the generation of nascent chlorine. As the bleaching with hypochlorous acid proceeds, hydrochloric acid accumulates and reacts with the hypochlo- rous acid according to the equation, HOC1 + HC1 = H 2 + Cl 2 . When a hypochlorite is used the acid formed is used in liberat- ing more hypochlorous acid. It has been shown experimentally that the removal of free hydrochloric acid from either hypo- chlorous acid or chlorine water by means of calcium carbonate results in an acceleration of the bleaching action. Higgins 2 has also shown that the addition of calcium, barium or sodium chloride, or of sodium or potassium fluoride to a hypochlorite solution causes an initial increase in rate of bleaching, but that this acceleration soon ceases and the normal rate of bleaching ensues. This action is due to the formation of nascent chlorine: HOC1 + NaCl ±± NaOH + Cl 2 . The calcium chloride formed during the action of the bleaching powder is negligible but when calcium chloride is added during the process it always has 1 J. Soc. Dyers and Colorists, 19 14, 30, 326. 2 J. Soc. Chem. Ind., 1913, 32, 350. 228 BLEACHING a stimulating effect. This effect of adding salt probably ex- plains in part the claims for greater efficiency which many observers have made for electrolyzed salt solutions since the latter always contain a large proportion of undecomposed chloride. Bleaching Powder. The bleaching solutions first prepared by Tennant proved difficult to keep and transport and in 1799 he introduced a great improvement by absorbing chlorine gas in slaked lime, thus forming bleaching powder, which is still the most important commercial bleaching agent. The quality of the lime used in making bleaching powder is of importance, a fat lime which slakes quickly and gives a fine, light powder being most suitable since it absorbs the gas more quickly and gives a better keeping powder than a poorer lime. Careful slaking is essential since the total moisture in the chlo- rine and the lime should be about 28 per cent or about 4 per cent over that necessary to give calcium hydrate, Ca(OH) 2 . Well- made bleaching powder should be a pure white powder which, if of high strength, may contain some lumps. These, however, should be of the same quality as the powder and should not contain hard cores of calcium hydrate. In the air it absorbs moisture and carbon dioxide and is finally converted into a sticky, grayish mass. According to Lunge the composition of bleaching powder is best expressed by the formula, Ca Sulphite 1 n-5 40.0 n-5 12 .0 11. 7 18.0 14 18 8 Temperature C 18 Time in hours Per cent of bleach consumed 0-5 1 .0 i-5 2 .0 3° 4.0 5-o 6.0 7.0 8.0 63.8 70.O 72.5 82.5 5i-3. 55-3 71.7 73-i 77-5 96.7 33 -o 44 66.0 70.0 80.0 3° 55 78 90 20 33 43 49 56 63 70 11 .0 71 .0 IOO. O 1 The Testing of Wood Pulp. Weight Lost on Bleaching. The loss in weight which all fibre undergoes on bleaching, because of the oxidizing and dissolving power of the bleach solution, is in accord with the observed facts regarding the influence of temperature. Since the higher tem- perature causes quicker action and gives poorer color due prob- ably to chlorate formation it would be expected that the attack on the fibre would be less with a consequent smaller loss in 1 The Testing of Wood Pulp. WEIGHT LOST ON BLEACHING 2 37 weight. That such is the case is proved by the following results of experiments in which the fibre was bleached at different tem- peratures but with the same amount of bleach until the latter was completely used up. Fibre Percentage bleach used Loss in weight due to bleaching at 20° C. at 40° C. at 70° C. Sulphite 13.O "5 Per cent 2.31 I.89 I.89 Per cent 2 .27 1.66 Per cent E. B. sulphite 1 -97 Soda poplar 1.64 If the fibre is treated with an excess of bleach, which is not all used up when the desired color is reached, then an increase in temperature would be expected to increase the loss in weight and Simonsen 1 has found this to hold true in the case of a sulphate pulp which lost 4.5 per cent of its weight when bleached at 20 C. and 6.1 per cent at 30 C. The amount of bleach used exerts a great influence on the loss in weight during the process and the chemical composition of the bleach solution probably also has a considerable effect. It is hardly to be expected that sodium, magnesium and calcium hypochlorites would all cause the same loss in weight and the presence of alkali or acid during the process would certainly have an influence. With calcium hypochlorite the effect of increasing the amount of bleach used is illustrated by the following data: Fibre Sulphite E. B. sulphite Soda poplar. . . Per cent 35% bleaching Loss in weight due powder used to bleaching Per cent 13 2 .27 22 1 5i 32 5 06 5 1 66 15 3 59 25 5 72 "•5 1 94 20.5 3 69 3°-5 8 °3 1 Paper Trade J., Feb. 12, 19 14. 2 3 8 BLEACHING This has an important bearing on the cost per ton of bleached fibre, since in the production of very white stock not only is the amount of bleach increased, but the loss during the process also rises so that more fibre has to be used to produce a ton of finished product. In commercial fibres the loss due to the chemical .95 .90 .85 .80 .75 .70 ,€5 .60 .55 .50 .45 .40 .35 .30 .25 .20 .15 .10 FROM SINDAI 1 & RACDN // // \\ "TESTING OF WOOD PULP""' P 120 LOV1BOND COLOR READINGS Bleached Hot Cold r \\ \ \ \\ \v* -Yel low \ \ \ \ \ \\ V \ \ ' — \ V ed V l \ \ \ \ \ XX k Blu el \ \ \ \ \ \ \ > \\ > N ' " s N \ 1 \ \ ^« \ \. \ \ X - >^\ % _..-- .05 • -^ ^~ 2 4 6 8 10 12 14 16 18 20 22 24 % Bleach used Fig. 34 action of the bleach may drop as low as i per cent for very easy bleaching stock or rise as high as 5 per cent for fibre taking 20 to 22 per cent of bleach. In one case where a fibre lost 12 per cent of its weight when bleached and washed it was discovered that it had been washed in the blow pits with sea water and that about one half of this loss was due to the soluble constitutents left by the salt water. USE OF BACKWATER 239 The rate at which bleaching eliminates the color of sulphite fibre is shown graphically in Fig. 34, which is plotted from Sindall & Bacon's book on the Testing of Wood Pulp. The measurements were made by means of Lovibond's tintometer on samples of pulp which had been treated with progressively larger amounts of bleach. This shows very plainly the comparatively rapid disappearance of the red and blue and explains why bleached pulp generally has a slightly yellowish tone. No very positive statement can be made in regard to the bleach required by various classes of paper making fibres since it depends so largely on the way the unbleached fibre was pre- pared as well as on the bleaching method employed and on the color obtained. At the present time the following figures are approximately correct: Rags 2-5 Sulphite spruce 5-20 Soda fibres 8-15 Use of Backwater. The character of the water used in break- ing up the pulp preparatory to bleaching has an appreciable influence on the amount of bleach required. In many cases it has been recommended, and has been the practice, to use the water drained from the bleached pulp to make up another charge for bleaching. This is not to be recommended since it is attended with an increase in bleach consumption and a lowering of the color of the pulp. The greater bleach consumption is caused by the organic matter in solution or suspension, which continues to use up bleach when further portions are added. In a number of tests by the author this "yellow water" removed from engines of bleached pulp which were originally furnished entirely with fresh water was found to use up for each liter of water from 0.94 gram of 35 per cent bleach in six hours to 1.53 grams in eighteen hours at 35 C. Since the ratio of water to stock was 30 : 1 in these tests it is evident that this bleach consumption would be equivalent to 2.8 per cent to 4.6 per cent of bleach on the weight of the fibre. Sindall and Bacon l found that at 3 2 C. the bleach 1 Testing of Wood Pulp. 240 BLEACHING used up by this residual liquor was 35.5 lbs. for 10,000 lbs. of water while at 49 C. it amounted to 112 lbs. These quantities would be considerably increased if the water were used over and over, thus permitting an increase in the concentration of the organic matter. The use of "back" or "yellow" water for furnishing the unbleached stock is therefore to be recommended only when it is comparatively rich in unused bleach and when the products of its action can be washed out before the final bleaching is commenced, as under any other circumstances its use is likely to cost more than it saves. An intermediate washing to remove the soluble products formed by bleaching is beneficial as it enables a better color to be produced than would otherwise be possible, especially with hard-bleaching pulps. When one-third of the bleach is used up before such washing and the rest after- ward it has been found in actual practice that about 90 lbs. of bleach gave the same results as 100 lbs. when all is added at once. The bleaching of paper stock is performed in engines, chests or drainers according to the equipment available or the personal preference of the operator. Rag stock is most often treated with the bleach solution in the engine and then emptied into large tanks or chests provided with false bottoms. In these the bleaching is completed and the exhausted bleach solution then allowed to run away; the stock may then be treated with dilute sulphuric acid to still further improve the color and finally washed with water. The process may be hastened by heating the contents of the beater, best before the bleach is added, since otherwise it is likely to cause local overheating with loss of bleach and formation of oxycellulose. Addition of acid also hastens the bleaching by liberating hypochlorous acid which is more vigorous in its action than hypochlorite. A number of acids have been proposed for this work, from sulphuric to acetic or even carbonic which it has been proposed to lead into the beating engine just under the roll. The known accelerating influence of agitation or beating is due doubtless in part to the presence of carbonic acid in the air. Only a small amount of acid should be used, as it is constantly regenerated, and it should be added in a SYSTEMS OF BLEACHING 241 highly dilute condition as strong acid tends to liberate chlorine and cause the production of yellow chlorinated products. Treat- ment with acid is especially useful in bleaching shivey flax or linen and in treating rags which have been boiled in alkali and hence have accumulated basic matters which might be injurious at some future stage. Rags may also be bleached before reducing to half stock by washing the boiled rags and treating them in a tumbler with bleach solution; or they may be piled up in chambers and the warm bleach circulated through them. In the latter case they must be quickly drenched with cold water and transferred to the beaters as the fibres become tender if allowed to stand in contact with the bleach. Neither of these methods is so satisfactory as that of bleaching the half stock. Systems of Bleaching. Bleaching systems in general may be divided into two classes, one a rapid bleach in which an excess of fairly strong liquor is added and the excess removed as soon as the stock has reached the desired color; the other a slower process, using only a very slight excess and allowing it to prac- tically exhaust. The rapid bleach necessitates the use of the excess bleach liquor removed from the bleached stock, which has been shown to be uneconomical; it also requires more careful control since the higher temperature usually maintained and the stronger bleach solutions are more apt to cause oxidation of the cellulose and a consequent weakening of the fibre. The slower process is more economical of bleach but requires much more space for a given output and in the choice of a method the space available for the plant is frequently the deciding factor. In either method the amount of water used should be kept as low as pos- sible since this saves both bleach and time. The proportion of water to fibre will necessarily vary with the type of apparatus employed but with large chests furnished with good agitators it should be not much greater than 30 to 1. If the bleaching is done entirely in engines somewhat less water can be used, while in the Belmer bleaching apparatus the stock may be run at 5I to 6| per cent density. 242 BLEACHING Recognition of the advantage of using as little water as pos- sible is shown in Dobson's bleaching process in which the dry sheets of fibre are added to the bleach solution in a drum-like vessel which is then closed and rotated at two to four revolutions per minute for about three hours, when the bleaching is completed. The claims for this process are that it saves time, power and floor space and that the stock needs no washing or draining and is ready to use as soon as the bleaching is completed. While this process might be applied to wood pulp it is obvious that rags which have been cooked with lime could not be satisfactorily treated by it. The bleaching of fibres prepared by the soda process, whether from esparto, straw or any of the numerous woods now used, follows in general the same course as rag stock. Since the alka- line cooking treatment gives a well reduced fibre with a tendency toward an alkaline or basic condition it is safe to use a little acid in the process. This is best added in a well diluted condition after a considerable part of the bleaching has taken place and even then it must be used with care as these celluloses are more easily attacked than cotton or linen. When the color of these fibres has reached the desired point the bleach residues should be quickly removed lest the stock "go back" in color; this is especially likely to take place if the temperature is much over 30 C. The removal of the exhausted bleach may be readily accomplished by running off the stock on some form of wet machine or on a press-pate. The treatment of sulphite fibre differs slightly from that accorded soda fibre since it has already a tendency toward acidity because of its method of preparation. Also it is usually not so thoroughly freed from lignin and incrusting matters as is soda fibre and hence is more liable to take up chlorine with the forma- tion of yellow chlorinated compounds. As acid increases this tendency it is not generally used with sulphite fibre, though it may be successfully employed by adding it about an hour after the bleach, washing the bleached fibre, which is frequently orange colored, and rebleaching with if to 2 per cent of bleach followed JUTE AND MANILA 243 by a little acid. This treatment is unnecessarily complicated as most sulphite which is cooked with the idea of making into bleached fibre can be satisfactorily treated in a single process. As with wood pulp prepared by the soda process bleached sulphite is very apt to "go back" in color, or turn yellowish, if it stands in contact with exhausted bleach liquors, especially at elevated temperatures. For this reason it is quite important that the temperature be kept as low as is consistent with the desired rapidity of bleaching and that the fibre be freed from bleach residues as soon as the process is completed. In handling sulphite fibre it is frequently observed that a rose color develops on adding bleach. This is also caused by ferric chloride, potassium ferricyanide, mercuric chloride, potassium permanganate and potassium bichromate. 1 The exact reason for this phenomenon is not known but as the color produced is roughly proportional to the amount of bleach which the fibre requires it may be taken as giving an indication of the bleaching properties of the sample in question. This rose color is only transitory, as it disappears very quickly as the bleaching pro- ceeds, and has no permanent effect on the color of the fibre. Jute and manila fibres are particularly difficult to bleach since, in order to preserve their strength, they are generally only lightly treated with milk of lime and hence arrive at the bleaching process in a highly lignified condition. For this reason, and because the production of a high white seriously injures the strength of the stock, they are seldom bleached beyond a good cream shade. To hasten the action 4 to 5 per cent of alum is sometimes added but if the stock is heated this is particularly likely to cause the formation of yellow chlorinated compounds which defeat the object of the process. The bleaching of these fibres is usually conducted in the beater but in some cases they are dumped into drainers in which the last part of the bleaching proceeds slowly for several days. The chlorinated compounds mentioned, when treated with a solution of sodium sulphite, develop a strong magenta color. They are easily soluble in alkalis and may be 1 J. Soc. Chem. Ind., 1896, 467. 244 BLEACHING removed from the fibre by treating with a dilute soda-ash solution. Ground Wood. The treatment of ground wood, or mechani- cal pulp, so as to improve its color, is a problem which has occu- pied the attention of many investigators. Since the ground wood contains all the constituents of the wood itself, except a very slight amount of water-soluble matter, it is obvious that any treatment with hypochlorite would be out of the question, since 40 to 50 per cent of the weight of the stock would have to be oxidized and dissolved before any good white color could develop. The first effect of adding hypochlorite to ground wood is the production of a red or brown shade which persists until nearly all of the incrusting matter has been destroyed. The color of ground wood may be somewhat improved by treating it with sulphurous acid, or bisulphite solution. This can- not be considered a true bleaching process as the coloring matter is not destroyed but merely masked temporarily and the color returns on exposure to the air for some time. Considering the results obtained sulphurous acid is too costly and the reagent usually employed is calcium or sodium bisulphite. One method of treatment is to employ about 2.5 per cent on the weight of dry fibre, diluted with 20 to 30 times its weight of water, with which solution the fibre is flooded from below so as to drive out the air. After allowing to stand for some time the solution is washed out. Another procedure is to spray the fibre with bisul- phite solution as it is taken from the wet machine and allow the whitening to take place in the moist laps. In this case no final washing is given. Antichlors. It is very, essential that the bleached stock going to the beaters should contain no active chlorine, which may readily be determined by means of iodide of starch solution. This may be prepared by boiling up a little starch with water and adding a few crystals of potassium iodide. A little of this test solution sprinkled on the pulp will, if the latter contains any bleach, develop a blue color varying from a faint color almost to a black according to the amount of bleach present. It is always ANTICHLORS 245 best to test the stock being bleached rather than the water squeezed from it since if much bleach is present in the latter it may destroy the blue, while if the bleach is nearly exhausted the liquor sometimes fails to give a test though the stock still shows its presence. The difficulties encountered and the time required in washing out the last traces of bleach from the pulp have led to the use of various chemicals to reduce and render harmless any which may remain at the completion of bleaching. Such chemi- cals, from the nature of the work which they perform, are called antichlors. The one most commonly employed is sodium thiosulphate, Na 2 S 2 3 5 H 2 0, or "hyposulphite of soda" as it is generally called. This is added to the engine as a dilute solution after the stock has come to color. It acts on the bleach according to the following reaction: 2 Ca0 2 Cl 2 + Na 2 S 2 3 + H 2 = 2 CaS0 4 + 2 NaCl + 2 HC1, in which the products formed are calcium sulphate, or "Pearl Filler," common salt and hydrochloric acid. According to Griffin and Little 1 if the solutions employed are very dilute the decomposition may take place in another direction as : Ca0 2 Cl 2 + 4 Na 2 S 2 3 + H 2 = 2 Na 2 S 4 6 + 2 NaCl + 2 NaOH + CaO The products of either of these reactions are apt to exert a de- structive influence on the machine wires and their presence in the paper is fully as serious as that of the bleach which they are intended to eliminate. The use of thiosulphate is therefore not to be recommended. Safer antichlors to use are sodium sulphite, Na 2 S0 3 , and calcium sulphite, CaS0 3 , which according to the reaction, Ca0 2 Cl 2 + 2 Na 2 S0 3 = CaS0 4 + Na 2 S0 4 + 2 NaCl or Ca0 2 Cl 2 + 2 CaS0 3 = 2 CaS0 4 + CaCl 2 , give products which are much less harmful than those from 1 Chemistry of Paper Making. 246 BLEACHING thiosulphate. Because of the slight solubility of calcium sulphite the reaction in this case takes place rather slowly but it possesses the advantage that any excess acts as a filler and is in most cases quite harmless. Other antichlors which have been proposed are the ordinary sulphite liquor used in the manufacture of wood pulp; the mix- ture of calcium thiosulphate and polysulphide prepared by boil- ing sulphur with milk of lime; and hydrogen peroxide. The first of these is a rapid and efficient antichlor and in many cases it tends to brighten, temporarily, the color of the fibre. It must be used with considerable caution as any excess tends to set up an acid condition in the pulp with consequent injury to wires, driers and even the paper itself. The lime-sulphur mixture is probably even more injurious, since a considerable amount of free sulphur is precipitated on the fibres during the reaction and this, because of its finely divided condition, is gradually oxidized to free sulphuric acid, which renders the paper brittle by reason of the formation of hydrocellulose. The free sulphur also causes tendering of the machine wires through formation of metallic sulphides. Hydrogen peroxide is the safest of all the antichlors since it forms only water and free oxygen. It is, however, too expensive for commercial use. There are times when the use of an antichlor is of assistance but the regular employment of such an agent indicates inefficient bleaching methods since expense is being incurred in destroying a portion of the bleach which should be employed in doing useful work. In any well regulated mill it should be possible to elimi- nate the use of antichlors. Washing Bleached Pulp. The washing of bleached pulp to remove chlorides and the products of antichlor action is generally considered very necessary if durable paper is to be made from the stock. While this is probably true with regard to excess of antichlor and may have some influence in the case of the chlorides formed by the reduction of the hypochlorite yet it is felt that for the general run of book and magazine papers too much stress is laid on this point. This opinion is based on experiments carried PERMANGANATE BLEACHING 247 out with very thoroughly washed soda and sulphite pulps, and with part of the same lots which had been only very slightly washed after bleaching. These conditions were chosen as repre- senting the best and worst which were likely to occur in actual manufacturing operations. The two lots were beaten, sized and made into paper in the same way and pieces of the paper were then exposed to sunlight for three months and to temperatures of 8o° to oo° C. for eight days. Neither of these tests showed greater discoloration in one case than in the other and both became brittle and unsized to about the same extent. The conclusion that slight washing is not likely to prove injurious to the dur- ability of the paper is put forward with some hesitation as it runs counter to generally accepted theories but from the above experiments no other conclusion can be reached. Washing will however improve the color of the bleached fibre by removing the yellow, soluble products of the bleaching action, and will, in large measure, prevent the brownish discoloration which is sometimes noticed on the edges of wet bleached pulp after storing for some time. According to Griffin and Little this is caused by the concentration of calcium chloride, due to more rapid evaporation on the edges, till it becomes strong enough to act on the fibre and form colored decomposition products. Our experiments with thoroughly washed, pure filter paper and a solution of chemically pure calcium chloride have shown that even relatively strong solutions fail to cause any discoloration even under very severe conditions of storage. The reason for these brown edges in commercial pulps is the incomplete removal of the soluble organic matter which is brought to the surface and edges of the pulp by capillary action and there concentrated by evaporation till its color becomes noticeable. Permanganate Bleaching. Permanganate of potash has often been proposed as a bleaching agent in place of the hypochlorites, and all sorts of claims have been made regarding its alleged superior efficiency. In presence of acid permanganate reacts according to the equation 2 KMn0 4 + 3 H 2 S0 4 = K 2 S0 4 + 2 MnS0 4 + 3 H 2 +5 O, 248 BLEACHING while in neutral solution the reaction taking place is 2 KMn0 4 + H 2 = 2 KOH + 2 Mn0 2 + 3 O. Since the presence of acid causes too powerful an attack on the fibre it is necessary to bleach in neutral or slightly alkaline solution. Permanganate bleaching is extremely rapid and there is obviously no chance for the formation of chlorinated compounds. The permanganate should be added to the stock in dilute solution in order to avoid local formation of oxycellulose. When the bleaching has been completed it is necessary to remove the brown manganese peroxide by means of some reducing agent, most conveniently sulphur dioxide, or an acid sulphite. Beadle con- siders the sulphur dioxide to act as follows: Mn0 2 + S0 2 + 2 H 2 = Mn(OH) 2 + H 2 S0 4 , Mn(OH) 2 + H 2 S0 4 = MnS0 4 + 2 H 2 0. The Mn(OH) 2 being white is practically invisible and if the re- action goes no further may remain in the pulp and cause it to go back in color by absorbing oxygen and turning brown. All manganese peroxide should therefore be converted to the sul- phate. As all the alkali formed during the reduction of the permanganate must be neutralized before the manganese per- oxide can be dissolved it is evident that a considerable saving of sulphur dioxide can be effected by using some cheaper acid for this purpose. Beadle found by experiments on rags x that one pound of per- manganate did the same work as 10 lbs. of bleaching powder and concluded that the oxygen of the two substances acted quite differently. Our own experiments with sulphite and soda fibres show that in bleaching effect one pound of potassium per- manganate is equivalent to 1.854 lbs. of 35 per cent bleaching powder, while according to the oxygen evolved one pound should equal 1.93 lbs of bleach. It is evident, in the case of chemical 1 Chapters on Paper Making, Vol. II, p. 117. EFFECT OF BLEACHING ON STRENGTH OF STOCK 249 wood fibres, that the oxygen evolved by the two substances has the same bleaching power. The use of permanganate for conducting the entire bleaching process is more expensive than bleaching by hypochlorite and the necessity of acid for removing the manganese peroxide still fur- ther increases the cost. Even when the greater part of the bleaching is done by hypochlorite and the final treatment only is with permanganate the increase in cost is out of all proportion to the gain in whiteness. For these reasons permanganate is seldom employed in commercial work. Sodium peroxide and perborates, persulphates, etc., have also been proposed as bleaching agents but they are never used in this country. In fact von Possanner x has shown that sodium peroxide is of little value for bleaching rag stock, since even when excessive amounts are used only a partial bleaching takes place. Moreover, if employed in too strong solutions, it attacks the fibre and forms oxycellulose. Effect of Bleaching on Strength of Stock. The bleaching of paper stock induces in it a two-fold change since it affects both its physical and chemical properties. Experiments made by Frohberg 2 led him to conclude that bleaching very greatly reduced the folding strength in comparison with that of paper made from the unbleached fibre. In nine samples of sulphite which he tested this reduction in strength varied from 29.8 to 62.7 per cent of the strength shown by the unbleached fibre. The breaking length was affected much less than the folding strength, being reduced by only about 7 to 12 per cent. A slight overbleaching was found to reduce the strength and durability of the paper to a still greater extent. Our own very carefully conducted experiments entirely con- tradict those of Frohberg. Starting with the unbleached fibre, samples were beaten and made into hand mould sheets under standard conditions. Portions of the same fibres were then bleached with two different amounts of bleach, the lowest being 1 Wochbl. Papierfabr., 44, 3161. 2 Ibid, 44, 3599. 250 BLEACHING that which would give a good standard white shade, and the bleached fibre beaten and made into sheets as before. The air dry sheets were then subjected to folding tests on the Schopper folder, and to bursting tests on the Ashcroft tester. The results on three different fibres are as follows: Sulphite No. I Folding test . Ashcroft test. Unbleached i-9 18.3 13 per cent bleach 15-8 35-2 Folding test. Ashcroft test. Unbleached i-5 21.3 1 1. 5 per cent bleach O.O 20.2 22 per cent bleach II. 9 30.9 Sulphite No. 2 Unbleached S per cent bleach IS per cent bleach Folding test 2.9 21 .1 15.O 34-o 4-4 32.4 Soda, poplar 20.5 per cent bleach 0.0 9-2 With the sulphites the bleached fibres possess greater strength than the unbleached even when bleached to an extremely white color, while the strength of the soda fibre is reduced by bleach- ing. A partial bleaching with hypochlorite followed by per- manganate was found in every case to give a stronger fibre than if the bleaching were carried to the same point by hypo- chlorite alone. In this connection it is interesting to note the results obtained by O'Neill x on cotton cloth before and after bleaching. He 1 Griffin and Little: Chemistry of Paper Making, p. 288. EFFECT ON CHEMICAL PROPERTIES 251 measured the weight required to break a single thread with the following results: Before bleaching After bleaching No. 1 cloth, weft threads Grains 1 7 14 3140 3407 3512 Grains 2785 1 " warp " " 2 " " " 3708 '3 " " " 4025 In three out of four cases there is a distinct increase in strength due to bleaching. In our opinion it is safe to conclude that the folding and bursting strength of sulphite is not injured by bleaching but that with soda poplar the bleaching process does occasion a loss in strength. Effect on Chemical Properties. The two changes in the chemical properties of the fibre which are most likely to be caused by bleaching are the formation of oxycellulose and chlori- nated cellulose. If acid is used hydrocellulose is also likely to be formed locally. Griffin and Little 1 claim to have found between 5 and 6 per cent of chlorinated cellulose in a bleached sulphite pulp of good color, while Cross and Bevan 2 showed that during the bleaching of wood pulp and esparto chlorine combined with the residual non-cellulose constituents, forming chlorinated com- pounds some of which remain fixed on the fibres after washing. According to these statements the presence of chlorine in a bleached pulp would not necessarily indicate insufficient wash- ing. Contrary to these opinions Schwalbe 3 finds that when cellulose is bleached by hypochlorites, whether in acid or alka- line solution, the amount of chlorine which combines with the substance of the fibre is quite negligible. In the case of sul- phite cellulose there is, however, a small quantity of chlorine 1 Chemistry of Paper Making, p. 286. 2 J. Soc. Chem. Ind., 1890, 450. 3 Chem. Ztg., 1907, 31, 940-941. 252 BLEACHING absorbed. Further work by the same author x was carried out with both calcium and sodium hypochlorite, in neutral and acid solutions and in industrial proportions and in excess. After bleaching the sulphite cellulose was washed, dried and carefully analyzed. The total chlorine in the bleached pulps from these tests varied from 0.051 to 0.114 per cent and of this 0.047 to 0.071 per cent was present in combination with organic matter, the rest being combined with the constituents of the ash. In spite of these figures Schwalbe concludes that no appreciable chlorination of the pulp takes place and that the products of bleaching are removed during the process of washing. Besides the slight chlorination which may take place during bleaching there is a quite appreciable formation of oxycellulose. An approximately quantitative estimation of the amount of hydro- and oxy-cellulose formed may be made by determining the loss in weight of the thoroughly washed, bleached fibre when boiled in 0.25 per cent solution of caustic soda, or by means of Schwalbe's "copper number" method. 2 This latter method is a determination of the amount of copper reduced by the fibre on boiling with Fehling solution under carefully regu- lated conditions and the "copper number" is the number of grams of copper which would be so reduced by 100 grams of absolutely dry fibre. Using the two methods to study the effect of various bleach- ing conditions the following data were collected from experi- ments on two samples of sulphite and one of soda. All figures are based on the materials as weighed out for analysis and are not reduced to a common basis. These figures demonstrate that the more bleach used, and hence the whiter the product, the greater will be the attack on the cellulose itself. They also show clearly that permanganate causes much less oxycellulose formation than an equivalent quantity of hypochlorite. 1 Z. angew. Chem., 1908, 21, 302-303. 2 Z. angew. Chem., 23, 924 (1910), or Chemie. der Cellulose, p. 625. EFFECT ON CHEMICAL PROPERTIES 253 Copper number Loss in weight on boiling with 0.25 per cent NaOH Sulphite No << << << << «« << << 11 Sulphite No z 1 1 1 1 2 2 2 2 un { unbleached } } 2.56 2.8l 3-5i 7-13 4-13 3-i8 2.36 4-99 7-99 0.72 3 -°4 6.70 10.30 4.20 Per cent 13 per cent bleach 22 ' " " 32 " " /13 " t 5.4 " KMn0 4 ... . unbleached 10.9 14.2 21 .7 15-3 13.0 5 per cent bleach 10.2 11 11 tit " " 17.4 11 11 2< " " 25.2 Soda poplar 11 11 \ / \ S 1 ; §60 |l6 -+- \ \ / \ \ 1 M y * 1 40 15 1 /\ 1 1 / \ \ / / \ \ 30 \ S- \ \ / \ / 14 20 r -A sh \ / \ / I \ / / s « ' 10 13 ' '" 10 15 Roll Nos. 20 25 30 Fig. 36. Effect of Filler on Sizing blanc fixe and. precipitated chalk. Many of these are sold under trade or fancy names and frequently at prices which do not correspond to their actual value. In addition to these, certain colors, as ochre, lead chromate, ultramarine, etc., are sometimes used in sufficient quantity to affect the bulk and ash of the paper, but as these are used primarily for coloring, and their loading effect is merely incidental, they will not be considered here. 292 LOADING AND FILLING MATERIALS The qualities to be considered in judging a filler are color, fineness, absence of grit, mica, etc., solubility, specific gravity and chemical composition. The chemical analysis of a filler, as a means of checking individual shipments of the same material, is of little value, except in special cases where for some reason a certain constituent must be proved absent, or else present in definite amount. It is essential, however, to know the chemical composition of the various fillers, as this makes it possible to tell from the analysis of a paper ash what kind of filler was used, and it is also necessary if it is desired to know the amount of filler in the paper, as many fillers, on igniting the paper to an ash, lose water which would normally be retained by them in the finished paper. The specific gravity is of importance, as it shows how the filler will affect the bulk of the paper, while the presence of grit in any considerable amount indicates that the wires, felts and jackets will be subjected to unnecessary wear and that the paper may be defective to such an extent as to cause serious trouble in printing. When making heavily colored papers in which soluble dyes are used, the type of filler should be selected only after consideration of the dye to be used, for different fillers have different absorptive capacities for the various dyes. Proper selection of the filler will therefore be of considerable assistance in obtaining satisfactory absorption of the dye and in reducing its loss. The relation of fillers to dyes will be discussed in Chapter XL The proportion of the filler added to the engine which appears in the finished paper is spoken of as the retention. This may vary from 30 to 90 per cent, though the latter figure is only reached under very exceptional circumstances and a retention of 50 per cent is generally considered satisfactory. Many factors, other than the filler, influence its retention and it is impossible to estimate their effects, except in a general way. The kind of stock and the extent of its beating, the speed of the paper machine, the pull on the suction boxes, the amount of filler added and the thickness of the sheet must all be taken into consideration. Slow or " greasy" stock, a light suction and a thick sheet all RETENTION OF FILLERS 293 tend to give high retention, while in the case of sulphate of cal- cium the retention increases with the amount added. A study of the loss of filler taking place at different parts of the paper machine was made in a German mill x during the running of a rotary press print paper of 50 grams per square meter, made from stock containing 25 per cent sulphite and 75 per cent ground wood. The ash in the air dry stock at various points was as follows: Per cent Stock from the chest 20. o Stock just before suction boxes 19.0 Stock after the couch 15.0 Stock after first press 14. o Stock after second press 13.5 Finished paper 13-0 The apparently slight loss in the drainage through the wire is due to the fact that the white water was used over again and the ash in the stock actually flowing onto the wire would therefore be greater than that in the chest. Working with sulphite, and with sulphite and soda furnishes, Kress and McNaughton 2 found the retention to decrease slightly as the amount of clay added was increased. This is contra- dictory to the results given in the subjoined table from observa- tions by the author. They also found, as would be expected, that increasing the thickness of the sheet — or the ream weight — increased the retention. It was also increased as the amounts of size and alum were increased, and by greater hydration of the stock in the beater. The following figures from tests made under the observation of the author show what may be expected from various fillers when used in manufacturing high grade book papers. 1 Paper: 1916, Sept. 20, p. 13, from Wochbl. Papierfabr. 2 Kress and McNaughton: Paper, Oct. 3, 1917. 294 LOADING AND FILLING MATERIALS Filler Percentage in finished paper Percentage retention China clay 39-1 32.2 26.7 25.6 i5-i 24.8 19. 1 12 .0 9-7 22.3 10.8 10.2 24.9 24.2 27.9 76.2 74.6 68 6 n it n n 66.5 57-6 38.7 40.9 36.6 54 -o 72.4 46.1 S2.4 64.4 51-2 39 -o 11 ii Precipitated chalk ii ii ii ii Asbestine " Pearl finish Crown filler Blanc fixe Clay. Clay is a soft, friable, sectile, white substance showing irregularly shaped particles under the microscope and possessing only a moderate plasticity when mixed with water. It is a mixture of hydrated silicates of alumina containing particles of quartz, mica and felspar and its chemical composition is approx- imately Al203-2Si02'2H 2 0. It is formed by the weathering of felspathic minerals and the presence of mica indicates its granitic origin. Those clays which remain overlying the rock from which they were formed are known as residual, while those which have been conveyed to a distance by eroding influences are called sedimentary. The former are usually more free from iron and of better color than the latter. Good grades of clay are found in England, France, Bohemia and the United States, but in spite of our own deposits the greater part of that used in this country is still imported. This is probably due in large part to insufficient care or improper methods of preparation, and it is thought that these will shortly be overcome so that domestic clays will play a much more important part in American manu- facture. Clay is prepared for the market by washing it with a stream of water through long sluices of riffles, in which the sand and mica are caught, into tanks or basins in which the fine clay is CLAY 295 allowed to settle. The water is then drawn off and the clay dug out and dried when it is ready for market. Instead of these crude methods the fine clay and water is sometimes passed through filter presses, and there has recently been proposed an electrolytic method of purification and separation which is claimed to give a product perfectly free from grit and much dryer than that from any filter press. Even the very finest washed clays contain a small proportion of sand and mica and there is always some moisture present. For English clays up to 12 per cent is permissible though it more frequently runs under than over that figure. It is not well to have a clay too dry as it causes loss in handling and the dust is bad from a hygienic standpoint. Freezing moist clay causes no permanent change in composition or physical properties so that clay which has been frozen can be used with perfect safety after it has been thawed out. The following analyses give a good idea of the composition of different grades of clays; numbers one to four being English clays tested by Remington x while five and six are American clays. No. of clay 1 2 3 4 5 6 Coatings High grade papers News White printings Silica, Si02 Alumina, AI2O3 Ferric oxide, Fe 2 03. . . Lime, CaO Magnesia, MgO Alkalis, K 2 46.21 39.82 O.38 0-45 O.IO 0.23 12.81 47.60 38.26 0.55 O.42 0.20 0.57 12 .40 46.46 37-40 2.00 0.86 0.21 1 .26 II. 81 45-92 38.43 0.71 1. 18 0.21 0.78 12.77 45-67 37.86 I.48 O.05 O.OI 0.80 13.22 43-36 40.54 O.90 0.08 0.38 0.88 Total water 13-86 Grit per cent 100.00 0.09 IOO.OO O.32 100.00 2.22 100.00 2.28 99-09 100.00 The similarity of these analyses is quite noteworthy and it is evident that, apart from the ferric oxide which seriously influences J. Ind. Eng. Chem., 3, 555. 296 LOADING AND FILLING MATERIALS the color, the qualities by which a clay may be judged are largely physical. It is stated by one authority that if the grit rises over 2 per cent the clay would be objectionable for high grade news- paper. The total water in these clays is very largely that chemically combined. This is not driven off on drying at ioo° C. but is expelled at a red heat or at the temperature of the usual ash determination. For this reason the percentage of ash as deter- mined in a paper loaded with clay should be divided by 0.88 in order to convert the ash into the approximate amount of actual filler; this is neglecting the ash from the fibrous materials and from the size and alum which is usually insignificant in compar- ison with the filler. Clay loses all hygroscopic moisture at ioo° C. and all combined water at about 500 C. Clay which has been dried at ioo° regains its plasticity on soaking in water. In this connection it should be noted that some air dry clays contain a small amount of moisture which is driven off on drying at ioo° C. and again absorbed on exposure to the atmosphere. In a number of high grade clays this moisture was found to amount to 0.43 to 0.77 per cent of the weight of the air dry clay. The size of particles varies greatly in different clays and the proportion of coarse and fine particles may be quite different in clays which are very similar in appearance. Tests of a number of clays by separation of the particles which settle if ins. in different lengths of time gave the following results: Settling in 1 minute " " 1-10 minutes " " 10-120 " " " over 2 hours. . Domestic UXL GAC GH Per cent Per cent Percent Percent 2 4 26 13 20 24 IO 35 3° 37 28 25 48 35 36 27 AR Percent 44 25 12 19 In the case of UXL clay the sizes of the particles separated in this way were found to be m •v.A'' =' *-.,, 1 '. i {> «t ."''■** •',;•%«."'' .•^.'•4 '>.v- f.'i *§^%' v #^ Plate 27 Domestic Filler Clay. Magnification 100 diameters. Photographed by Bureau of Standards. ?'%£? . "T" i^fe "i * ' X sr . ;; " ft* w > Plate 28 English Coating Clay. Magnification 100 diameters. Photographed by Bureau of Standards. CLAY 297 Settling in 1 minute o. 028 m.m. 1-10 minutes o. 007-0. 018 m.m. 10-120 minutes o. 0035-0. 007 m.m. over 2 hours o. 0015 m.m. The specific gravity of clay is variously given as 2 to 2.86. Observations by the author on six different samples of thoroughly dried English clays showed them to range from 2.568 to 2.634. The weight per cubic foot of clay as ordinarily received varies according to the method of handling from 53 lbs. when the fine material is run in loosely to 84 lbs. when the clay is tamped in hard. This is for clay with about 7.5 per cent of moisture. Clay is usually added to the stock in the beater soon after it is furnished so that it may be well distributed before the alum is added. The old way was to add the dry clay direct to the beater but the more progressive mills are now mixing the clay with water and straining before adding to the stock, thus avoid- ing much dirt. In preparing domestic clays for use in this way a little ammonia or sodium phosphate is sometimes found to be of great assistance in helping the clay to work up and in keeping it in suspension. In testing this clay mixture for uniformity it should be noted that the consistency of a clay-water mixture is dependent on the physical qualities of the clay as well as upon the amount used. Of six clays which were mixed in such a way that 100 c.c. of the mixture contained from 19.6 to 19.9 grams of dry clay the Baume readings varied from 16. 5 to 23 showing that this method cannot be used for daily control if the clay supply varies. A satisfactory control test for the clay mixture supplied to the beaters may be made by determining the weight of a definite volume of the mixture in comparison with an equal volume of water. A 250 c.c, glass stoppered, ungraduated flask is a con- venient volume to use. First clean, dry and weigh the empty flask including the stopper; then fill it completely with distilled water, insert the stopper carefully so that no air is trapped, dry the outside of the flask and weigh again. Once these two weights are established for a given flask they may be consid- 298 LOADING AND FILLING MATERIAL ered as constants. The test therefore consists simply in filling the flask with the clay mixture, cleaning the outside, especially around the stopper, and weighing. Assuming the specific gravity of clay to be 2.6 the calculation of the weight of dry clay in the flask would be as follows: Let x = weight of dry clay in flask y = weight of water in flask w = weight in grams of flask full of clay mixture c = capacity of flask in cubic centimeters = weight of water it will hold / = weight of empty flask in grams. Then x = w —f — y x y =c ~77 2.0 X = w — / — c -\ 2.0 _ 2.6 (w —f) — 2.6 c . 1:6 X = I.625 w ~ I -^> 2 5 (/ + c). Since / and c are constants it is only necessary to make a single weighing of the flask full of clay mixture in order to be able to calculate the amount of dry clay it contains. This can be converted readily into pounds per inch in any measuring tank which may be used. This method of testing has been employed in mill control work for a number of years and if the flask is weighed to a tenth of a gram the results are found to check closely with those ob- tained by drying and weighing a definite volume of the clay mixture. Gypsum. This is a natural calcium sulphate, CaS0 4 • 2 H 2 0, and is prepared for use as a filler by grinding, or it may be cal- cined, finely ground, washed and dried quickly to prevent its reabsorbing water and becoming hard and compact. The ground, PEARL HARDENING 299 uncalcined mineral has the form of plates. Three-quarters of the water of crystallization of gypsum is driven off at a tem- perature of 120 C. and the calcined gypsum which is formed takes up water readily and sets to a firm mass. If, however, the gypsum is heated to over 160 C. there is formed an an- hydride which absorbs water very slowly and hence may be rapidly washed. This anhydride on long soaking in water changes in form from the irregular particles of the ground mineral to small needle-shaped crystals. This change takes place so slowly that it is doubtful if it is completed during the time elapsing between the furnishing of the engine and the running of the stock into paper. Gypsum is generally added directly to the beating engine and results in the production of rather soft paper. Its use also tends to fill up the felts due to the crystallization of CaS0 4 • 2H2O. All gypsum is soluble in hydrochloric acid, or in 400 to 500 parts of water, which causes serious loss and poor retention unless the back-water is used over again. This solubility has not been found to interfere in any way with the rosin sizing in the engine in which it is used and in certain cases it is even claimed to be beneficial. The loss on ignition of calcined gyp- sum is only 0.5 to 2 per cent, while that of the uncalcined mate- rial is nearly 21 per cent. Pearl Hardening is an artificial, hydrated calcium sulphate made by precipitating a solution of calcium chloride with sodium sulphate. It may occur in two forms, flat, tabular crystals or minute, needle-shaped crystals. Its specific gravity is 2.39 and it loses 21 per cent of its weight on ignition. Other names for calcium sulphate preparations, either hydrated or anhy- drous, are Pearl White, Crown Filler, Pearl Finish, Annaline, Alabastine, etc. Some of these contain considerable water be- sides that chemically combined, and extravagant claims are fre- quently made as to their advantages both as to the finish imparted to the paper and the amount retained. These claims are to be taken with a grain of salt as all possess essentially the properties of gypsum or pearl hardening. 300 LOADING AND FILLING MATERIALS Precipitated Chalk or calcium carbonate is occasionally used as a filler particularly in very thin papers of the Bible class such as are now used so largely for the printing of dictionaries and encyclopaedias. It may be added directly to the engine but much better results are obtained by mixing a solution of cal- cium chloride with the stock and, when it has become thoroughly incorporated, precipitating it with a solution of sodium car- bonate. Care should be taken that the quantities used are approximately equivalent as otherwise serious losses may occur; a slight excess of either one has not, however, been found to be injurious to the paper though a large excess of calcium chloride would cause it to absorb moisture and become limp and lifeless. Calcium carbonate seriously interferes with the rosin sizing and even when the size is precipitated by alum before the filler is added it has been found practically impossible to produce a well-sized sheet. The advantage of precipitated chalk lies largely in its color since it imparts to the paper a much whiter color than can be obtained by the use of clay. It also gives the paper a characteristic velvety feel though it does not take such a high polish on supercalendering as does a clay-filled paper. Talc. Talcum or Spanish chalk is a hydrated silicate of magnesium. A part of the magnesia is nearly always replaced by alumina so that it may be regarded as a double silicate of magnesium and aluminum with the magnesium largely in ex- cess. It is very soft, has a characteristic soapy or greasy feel and is usually of a creamy or greenish white shade. Its specific gravity is 2.6 to 2.9. It is very resistant to acids and solutions of alkalis and also to heat, losing no water below a red heat. In preparing it for the market the stone is sorted according to color and then ground and graded either by an air blast or by bolting. The product is not so fine as clay and the grit is usu- ally greater. It is often adulterated with heavy spar or more often with ground limestone. Talc improves the printing qualities and the feel of the paper and gives it a rag-like appearance. It is said that 20 per cent Plate 29 Crown Filler. Magnification 100 diameters. Photographed by Bureau of Standards. Plate 30 Talc. Magnification 100 diameters. Photographed by Bureau of Standards. Plate 31 Asbestine. Magnification joo diameters. Photographed by Bureau of Standards. HEAVY SPAR AND WITHERITE 301 of talc will give the same results as 30 to 40 per cent of lower grade china clay. Asbestine, or agalite, is a fibrous talc which occurs as an alter- ation product of tremolite. It is of American origin and is ex- tensively prepared in St. Lawrence County, N. Y. It is nearly pure (95 to 97 per cent) magnesium silicate and being in the form of rodlets attaches itself well to the fibres and gives good retention. It does not impart quite so high a finish to the paper as does talc. The best grades are free from sand, nearly free from iron and almost snow white. This material loses up to 1.3 per cent on drying at ioo° C. and from 0.5 to 7 per cent on ignition. The following analyses are fairly representative of the composition of asbestine or agalite. 1 2 3 Silica, SiC>2 60.59 0.13 0.23 1. 16 61.89 1.36 0.44 Alumina, AI2O3 Ferric oxide, Fe203 O.31 O. IO Manganese oxide, MnO Calcium oxide, CaO 4.21 30.70 1 .40 100.00 Magnesia, MgO Water 34-72 3-77 100.60 32.12 4-3° 98.84 Heavy Spar and Witherite are sometimes used as fillers. The former is a naturally occurring barium sulphate, BaS0 4 , while the latter is carbonate of barium, BaC0 3 . In preparing the heavy spar the foreign stone is broken off and the spar is first ground dry, then wet and finally washed. Neither of these materials is very satisfactory as their high specific gravity, 4.2 to 4.5, causes much loss of filler and also settling on the wire so that the two sides of the sheet are not alike and the wire side is particularly destructive to pens and type. If barium sulphate is to be used as a filler it should be precipitated in the engine by adding first barium chloride and then a solution of sodium sulphate and great care should be exercised that loss does not occur through an insufficiency of the latter. 3° 2 LOADING AND FILLING MATERIAL Testing Fillers. The color of fillers is best determined by- comparison with standard samples of similar materials. As the moisture present has a very great influence on the color, espe- cially with clays, the materials compared should be equally dry. This is readily assured by drying at ioo° C. before comparing. One method, which is extensively used, is to mix the material to a paste with water on a glass plate and after drying to compare with the standard, similarly treated. A bet- ter method and one which can be rapidly carried out is to use a block of wood in the top of which there are shallow compart- ments separated by knife edges coming flush with the upper sur- face of the block. Such a test- ing block is shown in Fig. 37. Into one of these compartments the standard material is pressed by means of a polished steel spatula and in the next the filler to be compared is placed. The knife edges permit of very close contact of the two samples, the polished spatula gives a good surface and the small size of the block allows it to be easily handled so that it may be held in any position with regard to the light and the samples examined from all sides. In many cases it is well to note how much a filler can be im- proved by the addition of blue. This may be readily done by grinding the dry filler in a mortar with a very little ultramarine and then comparing the color with the same material unblued or with the standard. To determine whether a clay has been blued, it may be moist- ened with turpentine to form a rather thin paste on a porcelain plate, and then compared with samples which are known to be blued and unblued and which have been treated in the same manner. The artificially colored clays are said to give a bluish green color but as it is claimed there are a number of natural Fig. 37. Block for Testing Color of Clays TESTING FILLERS 303 English clays which give this same color with turpentine the test cannot be considered as conclusive. A better method is carried out as follows: In one of two similar white porcelain dishes place a measured amount of freshly prepared saturated lime water and in the other dish an equal amount of distilled water. Then into each of these liquids dust, from the end of a knife or spatula, a little at a time, equal amounts of the clay. After allowing to stand for a few minutes the excess liquid should be siphoned off and the moist clay examined. If the clay has been artificially blued the lime water will remove the bluing so that the two samples will appear quite different after this treatment. Although neither of these tests is entirely satisfactory, yet if both are applied it is possible to get a fairly accurate idea whether the sample has been arti- ficially blued or not. The grit in clay or other loading material may be determined in a number of ways. One roughly quantitative test is to place a little of the clay in the mouth when the grit may be readily detected between the teeth. A more accurate procedure and one which may be easily duplicated by independent operators is to weigh out a sample of the clay, place it on a standard mesh screen and wash it with a spray of water until the water running away is perfectly clear. The residue on the screen is then dried, weighed, and reported as grit. The size of the sample may be varied according to the preference of the operator but 20 grams has been found to be a convenient amount. Some form of spray head which delivers a fine spray under a consid- erable pressure will be found very satisfactory for washing the clay on the screen. It is also desirable to make the test on both 200-mesh and 300-mesh screens in order to get a more comprehensive idea of the character of the grit. A flotation process which gives excellent comparative results may be carried out by means of a large, wide mouthed, bottle fitted with an inlet tube for water and a siphon whose inlet is always kept near the surface of the liquid by means of a float. This siphon is fitted with an automatic stop to prevent its en- 3°4 LOADING AND FILLING MATERIALS tirely emptying the bottle. The clay, or other filler to be tested, is mixed with water and placed in the bottle and water is forced in rapidly through the inlet tube until it rises to a mark near the neck. The mixture is allowed to stand a definite time and the siphon then started; this process is repeated until at the end of the settling period the water is perfectly clear for a definite distance down from the filling mark. Any material settling more than this distance in the standard time is con- sidered as grit and its quantity is determined by drying and weighing. It is evident that the dimensions of this apparatus Tube Support \ Siphon Tube-H Hose Connection — »■ 1 -tWater Inlet \ Li£ Rubber Hose \~ Glass Receptacle Graduations 2" apart Fig. 38. Apparatus for Determining Grit in Fillers may be arranged to suit the convenience of the operator but those of an outfit which gives very satisfactory results with clay, blanc fixe, asbestine, etc., are shown on the accompanying sketch, Fig. 38. This method cannot be employed with calcium sulphate because of its slight solubility. The grit, when separated by any method, should be examined by means of the microscope as its appearance reveals much as to the quality of the filler and gives some idea of the trouble it is apt to cause. Thus the grit from clay may be either sand or mica, and, while the former causes wear on the wire and some- times pinholes in the paper, the latter may make shiny spots TESTING FILLERS 305 which are apparent on looking across the sheet and also give trouble in printing. Talc, as stated above, is often adulterated and hence its color is not a sure criterion of value. It should be tested for gypsum, calcium carbonate, mica, iron and sand. The usual test for calcium carbonate is by boiling with acid and noting the loss in weight, but if the acid used is too strong the test may be erroneous because the solvent action includes other portions of the filler. Good grades of talc should not contain over 1 to 2 per cent of oxides of iron or over 3 to 4 per cent of CaC0 3 . Talc may be readily distinguished from clay by moistening with a little cobalt nitrate solution and warming over a flame; the talc gives a pinkish color and the clay a strong blue. CHAPTER XI COLORING The importance of coloring matter to the paper industry is not always evident to the casual observer but it is at once realized when it is understood that very few papers are made without coloring matter of some kind. This is of course self-evident in the case of heavily colored papers, or even pronounced shades, but it is equally true of white papers, for very few of these are made of the natural color of the fibre. It may even be said that color- ing, as applied to the production of shades of white, is of much more importance than in the making of colored specialities, for the production of the former runs into far greater tonnage. It is also true that the maintenance of uniformity in a tinted white requires more attention and care than in the case of deeper colored papers, for a slight error in the amount of color used will make an appreciable difference in the resulting shade, while with deeper colors an error of the same magnitude will have hardly noticeable results. Variations in the color of the fibres used are also much more serious in the case of tints than they are with deeper colors. In coloring paper it is usually required that the shade of a sample be matched. When such a sample is submitted previous runs of paper should be looked over to see if anything of the same or a similar color has been made before. It is often the case that the records of former orders will show just about what coloring matters should be used and an engine can then be colored up and the paper compared with the sample as soon as it gets over the machine. When the shade desired is an entirely new one the use of a small beater, of a pound or two capacity, will be found very convenient. By using a weighed amount of 306 COLORING 307 stock in such a beater, sizing and loading as usual, and noting carefully the quantities of colors used, the amounts necessary for the larger engines may be easily calculated. The stock pre- pared in the small beater should be made into sheets on a hand mould and dried on a steam heated drying cylinder. The sheets may then be compared, either before or after calendering, with the sample submitted. When such records or equipment are not available a small part of the sample should be reduced to a pulp with a little water, being careful not to use so much that coloring matter is lost by being washed out. The first engine is then colored up to match this wet sample. Whichever of these methods is used a sample of the paper should be taken from the winders, as soon as the color gets thoroughly over the machine, and compared with the sample submitted. If the shade is not right then the proper changes should be made at once, both in the stock in the chest and in the other engines which are being prepared. In estimating the amount of dye to use due attention must be paid to the beating of the stock and the calendering of the finished paper. The more hydrated the stock has become from prolonged beating the less dye will be required for a given shade. Calen- dering and supercalendering also darken the shade. When a colored order is to be made it is desirable to start in the morning so that plenty of daylight may be available for match- ing the shade. The nature of the light used for color compar- isons is important and a subdued north light is generally con- sidered best. This should be from a window not affected by reflections from neighboring buildings. Whatever light is selected as a standard, it should be used for all color work as the same results cannot be obtained if south light is used one day and north light the next. Difficulties from variations of light are entirely avoided if one of the color matching outfits, or so-called "daylight lamps," is used. These vary more or less in quality but several excellent ones are on the market. To give the best results a lamp of this type should be used in a dark room. In comparing colors the first impression should be decisive as 308 COLORING the eye becomes less sensitive by prolonged staring. If doubt exists after the first glance rest the eyes by closing them, or by looking at some distant object, and then make a second com- parison. When papers are being examined they should be folded to such an extent that their thickness prevents any light from being transmitted through them since it is the light reflected from the surface which it is desired to compare. It is also well to change the samples from side to side — that in the right hand being transferred to the left, and vice versa — since the relative positions of the sheets has an influence on their apparent colors. In the coloring of paper the materials used may be divided into two general groups, the pigments, which are for the most part insoluble materials, and the dyes, which are generally employed in solution. Each of these classes has certain advantages and disadvantages which must be taken into consideration in select- ing the coloring matter to be used in any particular lot of paper. Pigments. Pigments are as a rule very fast to light and have the added advantage that they increase the weight of the paper by acting as fillers. They are not generally so brilliant as the dyes and have been in most cases replaced by the latter. They have properties, however, which make them valuable for certain papers and they should not be overlooked because they are more or less old fashioned. Both pigments and paste colors may with safety be added directly to the beater though heavy colors such as the canary and orange pastes may advantageously be thinned with a little water to prevent the settling of lumps and to insure their thorough mixing with the stock. While coloring with pigments is a purely mechanical operation it is at the same time necessary to pay due attention to the nature of the materials used and to those of the substances with which they come in contact; otherwise trouble will be caused by using simultaneously substances which are injurious to each other. Natural mineral colors are obtained from numerous natural deposits but before being of value they must be ground and separated in some way from any coarse or gritty particles. The PIGMENTS 3°9 fineness of their particles never equals that of the pigments pro- duced by chemical means but the finer they are ground the better results they will give. In addition to this mechanical treatment some of the earth colors are also treated chemically and in some cases various shades are obtained by heating the colors to certain temperatures. The shades of the natural mineral colors are usually of a subdued rather than a brilliant nature but so far as permanence is concerned they are not equaled by any other class of colors. Among colors of this class which are of interest to the paper maker are ochres, and red and brown earth colors. Whites, which would also come in this class, such as clay, gypsum, blanc fixe, etc., are discussed in the chapter on fillers. Ochres depend for their coloring power upon ferric oxide or hydrated ferric oxide, and various shades from yellow to brown are found upon the market. The best are finely divided powders, soft to the touch, and possessing plastic properties; dark-colored brands of this nature are generally richest in coloring matter. Ochres are sometimes mixed, or "topped," with chrome yellow to produce more brilliant shades; such products possess the defects of chrome yellow and if used without proper precautions are likely to cause trouble. The red earths owe their coloring power to the presence of amorphous ferric oxide. This is the chief ingredient in red hem- atite, which is the basis for numerous colors. Other reds are obtained by heating to redness, clays which contain hydrated ferric oxide. This class of colors includes a number of "red oxides," varying from yellowish to bluish red, and also Pompeian and Venetian reds which are usually of less strength than the "oxides." Among the brown earth colors are "velvet," "umber" and "chestnut" brown which depend upon burnt ferric hydrate for their coloring power. True umber consists mostly of manganese silicate which is greenish brown in its natural state but becomes a rich deep brown on burning. These natural mineral colors are largely used in the production of wall paper, for which purpose their subdued shades, their fast- 310 COLORING ness to light and their resistance to atmospheric influences render them especially suitable. Artificial mineral colors are still used quite extensively in the coloring of paper, though, as already stated, many of them have been replaced by aniline dyes. Among those which are still relatively important are chrome yellow, Prussian blue and ultramarine. Chrome yellow may be obtained in the paste form from the manufacturers of pigments or it may be prepared directly in the engine by adding first nitrate or acetate of lead and when this is thoroughly mixed following with a solution of potassium or sodium bichromate. This determines the precipitation of lead chromate upon the fibre, in a very finely divided state. The color produced is influenced by alkalis, very small amounts of which are sufficient to darken the shade. Heat also influences the shade to a marked extent which necessitates very careful handling of the paper on the driers if irregular results are to be avoided. As the lead salts used are readily soluble in cold water no heating is necessary at this point and for the purest yellows the size and alum solutions should also be cold when used. The choice between the ready made paste color and that prepared in the beaters is largely a question of personal preference. The paste colors are comparatively simple to use and the matching of shades is much facilitated when they are employed; on the other hand the production of the color in the beater aids in the fixing of the color and in the obtaining of even shades. Chrome yellow is very fast to light but is destroyed by hydro- chloric acid. As the lead salts are dangerous poisons, their use is not to be recommended, and whenever possible the substitution of yellow dyes would appear to be good policy. Chrome yellow may be converted into chrome orange, or basic lead chromate, by treatment with caustic soda or hot milk of lime. This color can be used for unsized papers only as it reverts to chrome yellow in the presence of aluminum sulphate. Prussian blue is classed with the mineral colors because of its iron content. It may be produced directly in the beater by add- PIGMENTS 311 ing ferrous sulphate followed by potassium ferrocyanide; the white precipitate which is first formed is rapidly oxidized by exposure to air with the formation of the blue color. This action may also be hastened by the addition of bleaching powder or acid to the beater. As the ferrocyanide is the more expensive of the two ingredients any excess is carefully avoided and it is customary to use three parts of ferrous sulphate to two of the ferrocyanide although two parts of the sulphate are usually sufficient for the full development of the color. Prussian blue which has been oxidized in the beater sometimes causes the paper to turn red after some time. This can be avoided by washing the stock nearly free from acid or better by employing a Prussian blue prepared outside the beater and washed before use. Paper colored with Prussian blue has the peculiarity that exposure to sunlight partially decolorizes it; the full blue shade is, however, again developed when the paper is kept in the dark in contact with the oxygen of the air. Prussian blue is affected by alkalis, particularly caustic soda, which destroys the color with the formation of ferric hydrate; treatment with acid restores the color. When using it in the beater it is well to see that it is added while the reaction is slightly acid due to the presence of alum. Soluble Prussian blue is produced when a ferric salt is added to an excess of a solution of ferrocyanide. The soluble product also results if Prussian blue is boiled in a ferrocyanide solution. It is soluble in water but is precipitated by salts. Ultramarines are formed when aluminum silicate is calcined with sodium sulphide. In actual practice the sulphide is formed from the action of sulphur and carbon upon sodium sulphate or carbonate. Ultramarines are manufactured in various shades of blue from a greenish to a reddish tone and there are even pure greens which, however, find little use in coloring paper. They are made by heating mixtures of pure clay, sodium sulphate, sodium car- bonate, sulphur, silica and charcoal; the mixture after heating is finely ground and washed. The proportions of the ingredients 312 COLORING vary with the different manufacturers but in general three grades are made as follows : i . Sulphate ultramarines are those made with sodium sulphate. They are the palest, are greenish in tint and are most easily attacked by alum. 2. Soda ultramarines low in sulphur are pure blue and darker than the sulphate ultramarines. 3. Soda ultramarines high in sulphur and silica are the darkest and have a reddish tinge. They are the most resistant to alum. The finished ultramarine contains sodium, aluminum, silicon, sulphur and oxygen; its actual constitution is not known and no theory so far proposed accounts for all its properties. Ultra- marines are absolutely fast to light and are not changed by expo- sure to the atmosphere or by weak alkalis. They are decom- posed by mineral acids with evolution of hydrogen sulphide and destruction of the color. They are darkened by moisture which is taken advantage of by unscrupulous dealers who add water, glycerine or molasses to make them appear of greater strength. For this reason ultramarine should not be purchased on the basis of its appearance but actual coloring tests should be made. The different products vary in the fineness of their particles and the ease with which they mix with water. Both of these points should be considered since if they are not satisfactory spots or color streaks are likely to appear in the paper. When used for tinting in the production of white papers ultra- marines give bright effects which are hard to equal with other coloring materials. If they are used for deeper colors the two sides of the sheet are apt to vary in shade because of the loss of pigment in passing over the suction boxes. Another pigment which is still used occasionally for gray or black papers is lamp black. If used in large proportions it tends to cause streaks and specks in the paper and to make it smut badly. Its low specific gravity makes it difficult to handle with- out getting it all over the beater room and uniform results are hard to obtain because the shade depends so much on the length ARTIFICIAL ORGANIC COLORS 313 and nature of the beating which the stock has had. Lamp black is being replaced by various mixtures of soluble dyes. Natural Organic Colors. Colors of this class, of a vegetable or animal origin, were formerly much used in coloring paper but they are no longer employed to any extent because of the better and cheaper results obtained with the coal tar colors. For this reason brief mention of their names and properties is all that seems desirable. Annatto is derived from the fruit of the annatto tree and gives shades of orange. It is costly and fugitive. Turmeric is prepared from the roots of Curcuma tinctoria. It dyes paper pulp a direct yellow which is fast to acids but is sensitive to light and alkalis. Weld is obtained from the blossoms of Reseda luteola. It gives various shades of yellow according to the mordant used. Quercitron is the powdered bark of Quercus tinctoria. It gives yellow shades when used in the same manner as weld. Safflower is obtained from the petals of Carthamus tinctoria. It gives pinks of great beauty but they are fugitive to light and air. Redwoods. Various species of Caesalpinia have been used for the dyeing of pink and reddish shades which are not very fast to light. Cochineal is obtained from the cochineal insect and was formerly used as a pink for toning white papers. Cutch or Catechu is the dyestuff obtained from Mimosa catechu. It gives shades of brown which are fast to light, acids and alkalis. Logwood is obtained by extracting Haematoxylon campe- chianum. It is used for blacks in conjunction with iron salts and tannic mordants. Artificial Organic Colors. This class includes all the so-called coal tar colors or aniline dyes. They are superior to other color- ing matters in brilliancy and purity of shade, coloring power, 314 COLORING solubility and ease of application and their chief drawback is lack of fastness to light. In using these dyes the method of working depends on the fibre to be dyed and the nature of the coloring matter used. A knowledge of the methods employed in textile dyeing is of con- siderable assistance though the impossibility of thorough washing in the beater renders impractical the use of many of these methods. Moreover since vegetable fibres only are used in paper making any methods which are applied to wool or silk are only of abstract interest to the dyer of paper. It must be borne in mind that the different fibres in mixed stock may have different affinities for the coloring matter and by taking it up in different degree cause an uneven or variegated appearance. If clear light shades are desired only bleached pulp may be used while the heavy,. deep colors can with advantage be obtained on unbleached stock. Fillers which take up the dyes assist in obtaining even shades and whenever possible the filler should be selected with this object in view. To allow the filler to absorb the greatest amount of color it should be added after the dye but before the size. The combination of the filler and the dyestuff is regarded by some as a chemical phenomenon while others consider it purely a physical action depending on the ability of the filler to form colloidal solutions. The amount of dye taken up by different fillers has been determined by Heuser 1 who gives in the following table the percentage of the added dyestuff taken up by the filler when 10 grams of the latter are treated with 0.4 gram of color. 1 E. Heuser: Wochbl. Papierfabr., 1914, 2288 and 2470. ARTIFICIAL ORGANIC COLORS 315 Color Malachite green. . . Crystal violet Manchester brown. Safranine Chrysoidine 96 99 97 84 96 66 60 58 68 25 Diamine green 58 Dianil blue 96 Diamine violet Diamine heliotrope 60 Diamine purpurine 59 Alkali blue .... Acid magenta. . Ponceau Cotton scarlet. Napthol yellow. Asbes- tine Blanc fixe Bohemian earth China clay Kaolin 72.75 64. 72 41 55 Talc 49-95 60.56 40.62 30.06 35-52 35-26 49-89 54-27 60.12 25-45 50.00 60.76 69.11 65-85 60.48 The following figures by H. Strom x show the color absorbed in grams by 1 gram of filler from 100 c.c. of 0.1 per cent solution of the dyestuff . Malachite green Safranine, cone Paper Scarlet ex Dianil yellow R Dianil red R Eosine ex. 5 B Paper deep black, cone. Asbestine 0.005786 O.007752 0.003715 0.005786 0.008039 o . 000900 0.005970 Blanc fixe .003715 .004877 .002251 .006712 .OIOOOO .002764 .007277 Talcum O.005058 0.007752 0.003237 0.004518 O .006526 O.002764 0.004877 Kaoli O.008425 O.009941 0.001914 o .005603 o .006526 0.002936 O.C03633 China clay 0.008425 0.009941 0.003237 O.004518 o . 009404 0.002593 0.006712 Not all dyes of the same class are taken up to the same extent by the same filler. Acid dyes can be removed almost com- pletely from fillers by washing with hot or cold water; basic colors fix themselves on silicates but even basic colors can be washed out of blanc fixe. Apart from all questions of the theory of dyeing the coloring Strom: Wochbl. Papierfabr., 44, 4516. 316 COLORING of paper pulp is not only a question of forming and fixing colored precipitates where pigments are concerned but also of fixing the soluble colors firmly on the fibres by means of mordants. Coal tar colors which form no precipitates with metallic salts and which are not fixed on the fibres when the pulp is acidified are of no use in coloring paper. The mordants in some cases serve to fix the color upon the fibre and make it more fast to washing, light, etc., while in other cases they combine with the dye as an essential constituent without which it would be uncolored or a worthless shade. Mordants are of two general classes, acid mordants, such as tannic acid and the fatty acid compounds used for fixing basic dyes; and basic mordants, consisting of the hydrated oxides of the heavy metals as tin, copper, chromium, iron, aluminum, etc., which serve for fixing the acid dyes. Basic mordants are employed in the form of soluble salts, such as the sulphate or acetate of aluminum, which react with the fibre with the deposition of the base which then attracts and fixes the color. Time is required for this reaction and different mordants give different colors with the same dye. Cotton has little affinity for ordinary metallic salts but if they are present in very basic con- dition it may decompose them with the loose fixation of metallic hydroxides. Tannin on the other hand has a direct affinity for cotton and may be still more firmly fixed by the use of tartar emetic or glue. Linen is similar to cotton but is even more difficult to dye. While not strictly a mordant rosin size gives to fibres some of the properties of animal fibres and enables them to take up many colors without the use of any other mordant. The use of mordants, other than rosin sizing, is not nearly so general in the paper industry as in textile work and in many mills they are never employed. The water used in dyeing operations may have a considerable influence on the results. Finely divided vegetable impurities have little effect on either colors or mordants but inorganic im- purities are much more serious. Hard water due to carbonates or bicarbonates of calcium or magnesium may cause partial pre- cipitation of basic colors; if it is necessary to use such water for DIRECT COTTON COLORS 317 dissolving basic colors it should be corrected by adding a very slight excess of acid, preferably acetic acid. Salts of iron in the water are particularly bad as they discolor the fibres and act as mordants with the production of bad shades. Dyes should not be added to the beater in the dry state except in very special cases, as sooner or later trouble with color specks will be encountered. They should be dissolved in soft, or con- densed, water and strained through a hair sieve or through wet flannel before being used. The amount of water necessary varies greatly but in general is more for basic than for acid dyes; with some of the former it may be necessary to use as much as 200 lbs. of water for 1 lb. of color. Most colors may be heated nearly to boiling without danger, but a few, as auramine, methyl green, etc., are injured by boiling and should not be heated above 160 to 170 F. If the color separates from solutions which have been made some time it may be redissolved by heating and stirring. Some colors are insoluble or slightly soluble in water and in this case equal parts of methyl alcohol and water may be used. The dyestuffs are variously classified by different writers. Direct or substantive colors are those which color the fibres directly without the use of a mordant; they are fully developed colors and always give the same shade, either weaker or stronger according to the amount used. Mordant or adjective colors are those which must be treated by chemical means in order to develop the true colors. This group forms with metallic oxides insoluble precipitates or lakes on the fibre. Colors which are of interest to the paper maker may be divided into four principal groups as follows: (1) direct colors, (2) basic colors, (3) eosines and rhodamines and (4) acid colors. The grouping of the acid colors separately is for practical rather than scientific reasons since the dividing line between the acid and the direct cotton colors is not at all sharp. Direct Cotton Colors. These may be used on unmordanted fibres in a neutral or alkaline condition and they can be used mixed with each other in the same bath. Acid colors may usually 318 COLORING be mixed with the direct colors and used together though basic colors should never be mixed with direct colors either dry or dis- solved. Dyeing with the direct colors is best done at the boiling temperature though it may also be done warm or even cold. If the stock is not heated the backwater is apt to be colored. Salt to the extent of 75 lbs. to a 1,000-lb. beater is also desirable to assist the fibre in taking up the dye. Under most conditions the colors tend to bleed from the fibres when they are mixed with white fibres so that they are not satisfactory for granite papers. This can be prevented to a certain extent by topping with basic colors and it is also claimed that by adding 10 lbs. of Glauber salt per 100 lbs. of fibre and boiling for three quarters of an hour fastness to water can be insured. The direct cotton colors are usually precipitated by lime and magnesia and water containing these substances should be corrected by boiling with soda ash. These colors are particularly desirable for blotting papers and tissues where sizing cannot be used; they are also equally ser- viceable for sized papers. They, exhaust well and a colorless backwater is usually obtained. They vary greatly in fastness to light, some being fully as fugitive as the basic colors while others are among the fastest colors known. Certain of the direct blues are increased in light resistance by adding a little copper sulphate to the beater after the dye has been taken up by the fibre. When 1 per cent or less of the dyestuff has been used 2 per cent of cop- per sulphate should be added but if over 1 per cent of dye has been employed an equal weight of the sulphate will be sufficient. Basic Colors are salts of organic bases of artificial origin, the base containing the color bearing group. Most commercial basic colors are hydrochlorides, though sulphates, acetates, oxa- lates, nitrates or even double salts of hydrochloric acid and zinc chloride are also met with. In rare cases the color base is used. All basic colors are decolorized by reducing agents as zinc and hydrochloric acid. Some are decomposed into other substances so that the color cannot be regenerated but with most a colorless or "leuco compound" is formed which is easily oxidized to the EOSINES AND RHOD AMINES 319 original color. In dyeing with basic colors the salts decompose, the basic part combining with the acid present in the fibre or fixed thereon by mordanting with tannic acid. The nature of the mordant or of the fixing metal does not greatly affect the shade of any given dye. Basic colors have very great tinctorial power and are generally of pure and brilliant shades. They are fugitive to light but because of their great coloring power they are extensively used where permanence is not of the utmost importance. The affinity of all fibres for basic dyes is not the same, so when mixtures are being treated uneven dyeing may result; this can be avoided to a great extent by adding a little alum before the color or by adding the dye in a very dilute condition. When the fibres have a considerable affinity for the color, as with sulphite and jute, the dyeing may be done with the aid of rosin and alum or even alum only but for absolute fastness, as for use in making granite papers, mordanting with tannin is necessary. With unbleached sulphite there is a very strong tendency to absorb the color and to avoid uneven dyeing it is well to use the color in very dilute solution. For ground wood pulp basic colors are especially suitable and the dyeing should be done hot. Basic colors may be used mixed with each other but never mixed with either acid or direct colors since precipitation re- sults. If it is necessary to use both classes of color they should be dissolved separately and added separately to the beater. Many of the lakes formed by basic and direct colors are decom- posed at 70 C, or even below, so where both are used it is well to avoid high temperatures. The use of acid and basic colors together gives colorless backwater; to obtain the best results it is well to use the basic dye first and top with the acid color. The exhaustion of the color is also aided by the filler which readily takes it up at comparatively moderate tempera- tures. Eosines and Rhodamines. Commercial eosines are alkali salts of chlorine, bromine, iodine or nitro-substitution products of fluorescein or some of its derivatives, while rhodamines are 320 COLORING basic hydrochlorides of organic bases. Rhodamine B, for in- stance, possesses both acid and basic properties while the eosines have the properties of feeble acids. All of this group form leuco compounds on reduction and regenerate more or less rap- idly on oxidation but not always to the same compound as the halogen is frequently removed by the reducing agent. The eosines are mostly soluble in soft water or dilute alcohol, but with hard water insoluble lakes are formed so that such water must be corrected with soda ash. Rhodamines may be dissolved as with basic colors. Solutions of the eosines show more or less fluorescence particularly in alcoholic solution or in the presence of ammonia. Eosines are fast to alkalis but not to mineral acids; they are also affected by alum to a harm- ful extent so that an excess of the latter should be avoided. They have very slight affinity for vegetable fibres and their fixation is due largely to the sizing. They form lakes with metallic salts, those with lead being particularly brilliant; to obtain the best shades, therefore, sugar of lead may be used in the beater. Eosines dye shades from yellowish to bluish red; they are remarkably brilliant but are very fugitive. Rhodamines are also very brilliant and are much faster. They are not much used on cotton because they cannot be permanently fixed on the fibre without impairing their brilliancy. Acid Colors. Acid colors are of three groups (i) nitro com- pounds, (2) azo compounds and (3) sulphonated basic colors. They are decolorized by reducing agents but are differently affected; nitro compounds are converted into amino compounds from which the coloring matter cannot be regenerated; azo compounds decompose with breaking up of the azo group and cannot readily be linked up again; while the sulphonated basic colors form leuco compounds from which the color can be regenerated. Acid colors can De mixed with each other to produce com- pound shades. They are not suitable for unsized papers as they have comparatively little affinity for fibres and cannot be fixed VARIOUS APPLICATIONS 321 on cotton or linen to resist washing. Rosin sizing is essential for good results especially for deep shades and better fixation is obtained if the color is added long enough before the size to insure thorough mixing before the latter is added. As a rule acid colors have comparatively little tinctorial power. They are quite variable in fastness to light but are generally more fast than basic dyes. Miscellaneous. Not mentioned in any of the previous groups are certain organic pigments especially the indanthrenes. These belong to a class of vat dyes which in textile work yield shades which are remarkably fast to light. While a number of vat colors are used in textile work the only ones which have been tried for coloring paper are the blue indanthrenes. These are sold in the form of paste colors, being insoluble in water and dilute acids or alkalis. They are used in the same way the pigments are and have the common fault of all paste colors that the strength of the color is continually changing because of the loss of water through evaporation. The colors obtained with indanthrene, at least where tints are concerned, are not particularly bright and as a class they have little to recom- mend them except their great fastness to light. Various Applications. The foregoing remarks apply chiefly to the coloring of paper stock in the beater. The coloring of coating to be applied to the surface of the paper follows similar lines, except that the colors used must in general be fast to alkalis since satin white contains considerable amounts of free lime and the casein is usually treated with an excess of alkali in dissolving. When pigments are used in coating they must be very finely divided and easily broken down into their ultimate particles since lumps of color will crush when calendering and form spots or streaks. A third method of coloring is that known as stuffing, padding, or calender staining. Here the color is applied to the surface of a web of paper as it passes through the calenders. Soluble colors are necessary for this work and for absorbent, or very slightly sized, paper the solution may be made with water. If 322 COLORING the paper is hard sized, the color may be dissolved in denatured alcohol to make it penetrate the paper slightly. Acid dyes are much used for this type of work. For the production of heavy colors a saving of dye is some- times made by coloring the stock partly in the beater and when the paper is partly dried on the first section of driers running it through a vat of the color solution, then through squeeze rolls and finally over the rest of the driers. Testing Colors. The tests which it is desirable to apply to colors are those which will show whether the material is of the same shade and strength as the standard. In the case of new samples it is necessary to compare their color and tinctorial power with dyes already in use, and also to apply tests to show whether they are sufficiently fast to the chemicals with which they will come in contact. The most satisfactory method of testing colors, though by no means the quickest, is to color up a weighed amount of stock which has been beaten, sized and loaded in a small beater. Sheets made from this stock and dried will then give a perma- nent record of the shade and strength of the color in question. This method can be applied equally well to dyes and pigments but it is not very satisfactory for tints, such as the blue- or pink-whites, of the book or writing paper class. With soluble colors the following method has proved satis- factory in most cases. Prepare a solution of one part of the dye in iooo parts of water, noting carefully any insoluble resi- due. From this solution remove two samples of 50 c.c. and 25 c.c. respectively, place them in separate beakers of about 300 c.c. capacity, and dilute each to exactly 250 c.c. From a stock supply of dry, bleached, sulphite fibre prepare strips of the same size which will fit easily into the beakers of the diluted dyes. Plunge these strips into the solutions for exactly one minute, remove, drain and air dry. They can then be com- pared with the standard strips for shade and strength. For some of the very strong basic colors it is well to use 25 and 10 c.c. instead of 50 and 25. It is obvious that this test is TESTING COLORS 323 comparative only and that the quantities and volumes may be varied to suit the wishes of the observer. With pigments the following test has proved very convenient. Weigh into a porcelain cup 100 grams of clay and 3 grams of the pigment. Mix this with 70 to 75 c.c. of water and then add 60 grams of a casein solution containing exactly 12 grams of casein. Mix the contents of the cup very thoroughly and by means of a brush or some sort of a scraping device apply a very thick coat of the mixture to small sheets of white paper. When these sheets are dry, preferably air dried, they are com- pared with the standard sheets for color. This test also is comparative only and to make it of value the same clay, casein and paper must be used in every case. Fastness to chemicals may be determined by treating a dilute solution of the dye with small quantities of the chemicals with which it will come in contact. By making the tests with solu- tions of definite strength and using the same volumes each time the effects of the chemicals on the different dyes can be more readily compared. It is well to make the test first by noting the effect of standing several hours in the cold and finally by bringing the solutions just to a boil. For use in vulcanized fibre the colors should be fast to zinc chloride solution. This is best determined by coloring some of the stock with the dye to be tested, passing the sheet through a bath of zinc chloride of the correct strength, and finally washing. This will show whether the color is affected and also whether it will bleed during the washing of the finished product. Fastness to light may be determined by exposing strips of paper, colored with the dye to be tested, in a printing frame in such a way that part of the strip is protected from the light. If daylight is used it is difficult to duplicate the test because of the variation in the quality of the light from day to day and at different times of the year. This can be avoided by using artificial light rich in the ultra violet rays which are the ones most active in the fading of colors. Such apparatus intensifies 324 COLORING the fading effect and enables results to be obtained in a com- paratively short time. It is very seldom that it is necessary to determine what dye was used in coloring a given sample of paper, usually it is suffi- cient to be able to match the shade. The determination of the dyestuff used is a difficult matter at best and is greatly com- plicated when more than one dye was used, which is generally the case with colored papers. If it is necessary, the scheme proposed by A. G. Green 1 for the determination of dyes on vegetable fibres is probably the best one to follow. The iden- tification of the dyes themselves may be carried out according to the schemes of A. G. Green 2 and S. P. Mulliken. 3 The general properties of the different classes of dyes have already been described and it is usually sufficient to determine to what class or group the sample belongs, since this information fixes the dye method to be used. It is often desirable to know whether a commercial product is a simple dye or a mixture. If the dyes have been mixed in the powdered state, the fact that it is a mixture may be ascer- tained by taking a small sample on the end of a spatula and blowing it onto a piece of wet, white blotting paper. Each speck of color gradually dissolves and the various colors of a mixture show very plainly and can be tested by chemical means. When the coloring matters' have been mixed in solution and evaporated together this test fails. For such cases the dye can be tested by making a succession of dyeings of wool or cotton skeins in the same dye bath. If the dye is a mixture the first and last skeins will differ in shade. 1 Green: J. Soc. Dyers and Colorists, 1907, 252. 2 Green: J. Soc. Chem. Ind., 1893, p. 3. 3 Mulliken: Identification of Pure Organic Compounds, Vol. Ill, Commercial Dyestuffs. CHAPTER XII COATED PAPERS The class of papers variously known as coated or glazed, and in England as art papers, has been developed within compara- tively recent years in response to the demands of the printers for a paper on which half-tones could be reproduced to good advantage. The essential feature of such papers is a thin layer of mineral matter and adhesive applied to the surface of an ordinary sheet of paper, the function of the mineral matter being to form the surface for printing, while the adhesive is merely added to hold this mineral matter on the paper and prevent its being removed by the ink. The coating covers the individual fibres on the surface of the paper and in addition rills in any hollows or irregularities between them so that when the paper is calendered there results a fine, smooth, even, and con- tinuous surface which permits the finest dots of the half-tone screens to take perfectly. Such papers are used for lithographic work, for magazine and other printing and especially for the high class half-tones in catalogues and advertising matter. While coated papers possess advantages from the standpoint of the printer, they also have certain defects which are of a quite serious nature. The body stock is frequently made of inferior materials while the coating, due to its nitrogenous nature, is peculiarly subject to decay and to the attacks of insects, especially when stored in warm, damp places. The paper is also very heavy and bulky so that books made from it are difficult to handle and, moreover, it has very poor folding or bending qualities and is therefore much more liable to injury than uncoated papers. It is by no means certain that coated papers will not deteriorate within a comparatively short time 325 326 COATED PAPERS to such an extent that matter printed on them will be valueless and for this reason they should never be used for documents of permanent or historic value. The paper, or body stock, to which the coating is applied, need not be of the highest class since it is entirely covered by the coating, and hence inferior materials are frequently used in its manufacture. In fact in Europe ground wood is often one of the principal ingredients though it is a finer quality of ground wood than is generally made in this country. A practical limit is set to the use of low-grade stock by the fact that the coating is more or less translucent and does not prevent dirty or shivey stock from showing in the finished paper. Some of the qualities demanded in paper for coating are regular formation, softness and pliability, freedom from fuzz and cockles, uniformity of finish, and medium sizing. The regular formation is essential if uniform finish is desired in the coated paper, and softness and pliability are necessary if the paper is to run well on the c oaters since a hard, rattly sheet will not lie flat, tends to curl on the edges and does not take the coating well. Each little fibre which stands up from the sur- face in a fuzzy sheet seems to attract the coating so that it causes a mottled effect; this is particularly noticeable when the coating is tinted since the color seems to concentrate itself round the base of each fibre. One of the most important qual- ities to be considered is the sizing of the sheet since if it is too hard sized the coating tends to lie on the surface and be streaky, while if it is too slack sized the adhesive will be absorbed, per- mitting the coating to be weak, or necessitating the use of an abnormally large amount of adhesive. Defective sizing is much more serious when glue is used than when casein is employed since the temperature of the drying lines is generally high enough to keep the former liquefied and thus permit its absorption. The coating mixture is nearly always applied to the paper by machinery, several different types of apparatus being used. In one the mixture is transferred from a roller revolving in a trough to a felt or brush, and from this to the surface of the paper; APPLYING THE COATING 327 in another the paper passes under a roller which is immersed in a tank of the coating mixture and the excess is removed by- squeeze rolls. In either case the coating is immediately smoothed out and brought into good contact with the paper by means of brushes working across its surface with a reciprocating motion. The first of these brushes is comparatively coarse, but they gradually become finer till the last, which is very fine, to eliminate the marks made by the first. The arrangement of this coating machine is in- dicated in Fig. 39. After leaving the brushes the paper passes through a drying gal- lery heated by steam coils or a current of hot air and when thoroughly dry is reeled up ready for calendering. The amount of coating applied by this process varies with the purpose for which the paper is made and may be anything from a wash coat up to 35 per cent of the weight of the fin- ished paper. According to the type of coating machine used, both sides may be coated at one operation with the same color or each side may be coated separately with the same or different colors. 328 COATED PAPERS The adhesives used in coated paper are chiefly glue and casein, though starch and albumen have been used to some extent and many others have been proposed from time to time. The amount used varies with the different mineral matters, satin white requiring much more than blanc fixe, while clay occupies an intermediate position. If too much adhesive is used the paper tends to curl, the color is not so good and the coating is less porous so that when printed the ink is apt to offset and to crawl or mottle instead of lying flat. Such paper also takes a lower finish on calendering than when less adhesive is employed. From the point of view of the printer, as well as for reasons of economy, it is therefore desirable to use as little adhesive as possible. On the other hand if too little is used, the coating will "lift" or "pick" when printed, especially if heavy or tacky ink is used. This imposes a minimum beyond which the adhesive cannot be reduced, while as a matter of fact this minimum is seldom even approached because the many variable conditions render it necessary to employ a large factor of safety. In good work it is seldom possible to use less than 15 lbs. of dry casein to 100 lbs. of dry clay and much more frequently 18 to 20 lbs. are used, while with glue the quantity employed is about 20 to 25 lbs. The influence of the amount of casein in the coating on the time required for linseed oil to saturate the paper, which may be taken as a measure of the rapidity with which ink will pene- trate, is well shown by the following experimental data. Grams of casein per 100 grams clay Time in seconds for oil to saturate Grams of casein per 100 grams clay Time in seconds for oil to saturate IO IS 20 20 37 125 25 3° 780-1020 Does not penetrate 1 The mineral matter employed in coating paper may be clay, blanc fixe, satin white, talc, or a number of other substances and they may be used singly or in mixtures of two or more. FINISHING COATED PAPERS 329 The character of the finished paper depends very largely upon the mineral matter used and many different effects can be produced by a careful blending of the various substances. Satin white gives the smoothest coating and the highest finish of any of these materials, clay gives a lower finish, while blanc fixe takes less polish than either, thus by a proper selection nearly any desired finish may be obtained. The qualities demanded in a coating substance are good color, freedom from grit and the property of working up to a good, fluid mixture when the adhesive is added. For tinted or colored papers it should also be free from all traces of acid since this might injure the shade. The color demanded of a coating material is in general much brighter than that which a filler is expected to possess and the price paid is correspondingly higher. In addition to the two main ingredients in the coating mix- ture, other substances are frequently used in smaller amounts for special purposes. Soaps or waxes, either in solution or in the form of emulsions, are added with the object of improving the finish; antifroth oils are used to prevent excessive foaming of the coating mixture; glycerine, or some of its substitutes, is added to give the coating increased pliability or folding power and salt is sometimes employed to reduce the curling tendency of the paper. Nearly every manufacturer has in use some such modification of the regular process which he generally considers a secret, though it often proves to be quite widely known. There is thus opened a very wide field for the employment of chem- istry in the coating industry and it is probably safe to say that it offers more and harder problems than any other department of paper making. The finish imparted to coated paper depends on the way in which it is calendered as well as upon the materials of the coat- ing; the damper the paper the higher the finish. There is a limit, however, to the amount of moisture which can be advan- tageously left in coated paper which is to be calendered, since if it is too high the paper tends to crush or blacken and the color is seriously injured. The amount of moisture which will cause 330 COATED PAPERS this varies with different papers and with the pressure put on the calenders so that no invariable limit can be set. It is, however, safe to say that if the paper contains much over 5 to 6 per cent of moisture, exceeding care will have to be used in calendering it. The finish of paper may be considered as com- posed of two factors, smoothness and shine. The former is essential to good printing while the latter is not, and since it is highly inartistic as well as seriously injurious to eyesight, it would seem well to avoid it as far as possible. This is being done by producing dull finish or mat papers in which the smooth- ness is imparted by a light calendering, which is done in such a way that little friction is employed, so that the polishing effect is slight. Another method is to give a very thin wash coat, on the coating machine, over a paper which has been thoroughly smoothed. This produces a peculiarly velvety surface which takes half-tone effects with very beautiful results but possesses the slight defect that it is easily scratched, a slight stroke with the finger nail sufficing to cause a distinct mark. Careful selec- tion of materials which do not readily take a high finish, such as blanc fixe, barytes, precipitated chalk, etc., materially assists in the minimizing of this defect and also aids in the production of dull finish papers of the first class. Although papers with a very high glossy finish are still largely demanded, yet these dull finish papers are rapidly growing in favor and it is anticipated that they will become more popular as they become more widely known. The printing qualities of coated papers are largely influenced by the kind of adhesive used as well as by the amount. Glue coating possesses the property of taking ink especially well while casein is slightly more difficult to handle. This difference caused trouble when casein was first introduced and its general use was delayed because of the prejudice of the printers. It is often claimed that the reason for this difference lay in the acid reaction of glue coated papers and the alkalinity of those coated with casein. This explanation is probably erroneous since it has been found that casein coated papers, except those contain- GLUE 33 1 ing satin white, are normally acid to litmus in spite of the fact that an excess of alkali is almost always used in preparing the casein solution. Except with inks which are exceedingly sensi- tive to acids or alkalis the reaction of the papers is probably of very little importance since printers practically never make any difference, in regard to the ink used, between ordinary coated papers and those containing considerable satin white which are strongly alkaline. Glue. Of the adhesives used for coating paper glue was the first, and for a long time practically the only one, and although it has been largely superseded by casein it is still used to some extent. It was formerly the custom, in many cases, for the consumer to manufacture his own glue but it is better practice to purchase it of some reliable dealer as the supply obtained in this way is likely to be more uniform. For convenience in handling, the ground glue is to be preferred to the sheet glue, since it requires very little time for soaking and the solution can be quickly prepared. The quality best suited for coated paper work is a good grade of hide glue but it is undoubtedly true that much inferior material is used, either intentionally because of the apparent saving in cost, or unintentionally be- cause of its substitution by unscrupulous dealers. Poorly made glue is apt to give trouble by frothing and it probably lowers the quality of the paper in which it is used though from the numerous variables entering into the manufacture it is frequently impossible to locate the cause of inferiority with certainty. Glue for coating paper should be of good color, free from objectionable odor, of good strength, nearly neutral in reaction and for some purposes free from grease. The grease if present in appreciable amount tends to make " birds'-eyes " in the coat- ing and thus cause defective printing. Too much grease also lowers the clay carrying power or strength of the glue. If on the other hand, the glue is entirely free from grease, the coating dusts on the calenders and the paper will not take a good finish. This reason for dusting was suspected by one manufacturer who made his own grease-free glue and when a small amount 332 COATED PAPERS of fat was added to the glue solution the trouble entirely disappeared. A rapid semi-quantitative test for grease may be made by coloring the glue solution intensely with some aniline dye and then brushing it lightly onto white paper. If grease is present, spots or " birds'-eyes " will form and the number of these is roughly proportional to the amount of grease. This test also gives an indication of the way the coating mixture will spread under the action of the brushes. The best quantitative method for the determination of fat is that of Kissling which is carried out by dissolving 20 grams of the sample in 150 c.c. of water containing 10 c.c. of hydrochloric acid (sp. gr. 1.20). This is heated three to four hours on the steam bath, using a reflux condenser, and after cooling the fat is extracted by means of petroleum ether, which is then evaporated off and the residue dried and weighed. Good grades of glue may contain 0.1 to 0.6 per cent of fat, but it is safer to use those which run nearer the lower figure. The presence or absence of acidity in glue is of even more importance than the question of fat since in many instances the colors used are affected by the acid to such an extent as to cause serious variations in shade. The acidity of glue may be determined with fair accuracy by dissolving 1 gram of the sample in 500 c.c. of water, adding a few drops of phenolphthalein solu- N tion and titrating with — alkali. This gives all free acid, 10 organic as well as mineral, and since a number of acids may be present it is well to express the total quantity as the equivalent percentage of sulphuric acid. The percentage of acid in glues is quite variable, often running as high as 1.2 per cent, but for use with delicate colors it should not be over 0.2 to 0.3 per cent. The strength of glue is often considered to be proportional to the firmness of the jelly which it forms, but experience has shown that the amount of clay which a sample will hold is not always in accord with its jellying powers. A much more reliable test for strength is described under casein. CASEIN 333 Coated papers in which glue is used are generally slightly acid in reaction and they possess certain qualities of surface and porosity not present in casein-coated papers. The change from glue to casein was delayed for this reason since it involved also a change in inks and the technic of printing, but the lower price of casein enabled it to force its way gradually in till now it has almost wholly replaced glue in all ordinary grades of coated papers. Casein. Casein is a nitrogenous body which is present in milk to the extent of about 3 per cent by weight. It may be separated from milk either by the action of acids or rennet, but the rennet casein is relatively insoluble in alkalis so that it is out of the question for coated paper work. Acid caseins may be prepared by the action of any of the mineral, or the stronger organic acids; or the milk may be allowed to curdle spontane- ously from the formation of lactic acid. This latter procedure gives the so-called "self-soured" casein, while if acid has been used its name is usually attached to denote the method of preparation, as muriatic casein, sulphuric casein, etc. In this country acid is generally added to the skim milk, in South America much is made by self-souring, while in Europe both processes are used. Commercial casein is prepared from milk which has been freed as much as possible from fat by means of cream separa- tors. This skimmed milk is warmed to 49 to 50 C, the acid is added and when the curd has settled the whey is drawn off. The curd is then washed by hot water, drained on racks or boards, shredded, spread on wire bottom trays and dried in a current of warm air. The dried curd is then ground to any desired degree of fineness. If a high grade product is to be made great care must be taken that the wet curd is not kept so long that it has a chance to decompose or mould and the drying temperature must be closely regulated, since if it goes too high the casein becomes orange-brown and difficultly soluble. The kind of acid used also exerts a considerable influence on the character of the product; that made with muriatic acid gives 334 COATED PAPERS thick solutions which tend to foam and to become jelly-like on cooling, the self-soured produces thinner solutions which, on cooling, retain their fluidity to a notable extent, while that made by means of sulphuric acid occupies an intermediate position. The green curd from any of these methods can be used to good advantage in preparing the solutions for coating, but owing to its poor keeping qualities it is only obtainable locally. The nitrogen in pure casein has been found by various in- vestigators to range from 15.12 to 15.74 per cent, hence the per- centage of nitrogen found in any sample multiplied by the aver- age factor of 6.48 will give the percentage of ash-, fat-, and moisture-free casein. Commercial caseins have been found to contain about 13 per cent of nitrogen on an average so that the factor to convert percentage of nitrogen into commercial casein would be X 7-69. The results of examinations of samples of casein from both German and American markets are given below and show the variations to be expected in good commercial material. The German samples were tested by Hopfner and Burmeister, 1 while the American samples were analyzed in the author's laboratory. German American Max. Min. Avg. Max. Min. Avg. Per cent Moisture Fat Ash Nitrogen on total sample NaOH to neutralize to litmus .... 10.50 2 .06 4-95 13-55 7.27 O.23 3-53 12.52 9-23 0.85 4.07 12.99 12.3 9-9 4-5' '3-83 5-4 trace 1 .0 1.30 9-65 3-68 3-29 2.51 The specific gravity of commercial casein as received ranges from 1. 3 1 to 1.34 and the weight per cubic foot is 38.6 lbs. if run loosely into the container or 47.8 if tamped. The moisture in casein has been found to vary with the humidity of the surrounding atmosphere as is shown by the results of tests on two standard commercial samples. 1 Chem. Ztg., 36, 19 12, 1053. CASEIN 335 Per cent humidity Moisture in Sample 1 Sample 2 40.6 47-5 52-7 62.2 7.27 8.27 9 13 10.53 7 .20 8.06 8.99 10 -S3 This explains why a sample of casein containing 12 to 14 per cent of moisture when received will, if spread out in a thin layer, lose weight very rapidly at first until the moisture is reduced to 7 to 8 per cent and then will gain or lose in weight according to atmospheric conditions. Soluble caseins which were formerly on the market in large quantities, and are still met with occasionally, are merely mix- tures of ground casein and some alkali as. borax, soda ash, etc. An analysis of a representative sample gave the following results: Per cent Moisture 12. 2 Casein (free from fat, ash, and moisture) 67. 1 Soda ash, Na2C03 8.6 Borax, Na 2 B 4 07, 10 H 2 3.1 Fat 1.3 Ash insoluble in water 9-3 101. 6 Such caseins are easy to handle but they are likely to keep poorly as the alkali may absorb water and act destructively on the casein, thus impairing its solubility and decreasing its strength. They also limit the user to the kind of alkali already mixed with the casein and do not permit its proper adjustment to the work in hand. A third point against them is that a cheap alkali is paid for at the price of casein so that the user can make quite a saving by buying both the alkali and the casein separately and mixing them himself in the proportions he desires. Casein is essentially of an acid nature and hence requires treatment with an alkali to bring it into solution. It forms two 336 COATED PAPERS series of salts, those neutral to phenolphthalein being called basic casemates, while those neutral to litmus are known as neutral casemates. The casemates of sodium, potassium, lith- ium and ammonia are readily soluble, those of calcium, barium and strontium are much less so while those of the heavy metals are insoluble. The alkalis which have been proposed as solvents are am- monia and ammonium carbonate, and hydroxide, carbonate, silicate, borate, sulphite, and phosphates of sodium. Of these ammonia and the carbonate, borate and phosphates of sodium are the ones most frequently used while caustic soda is generally avoided because of its drastic action if used in excess. If care is used caustic soda is one of the best solvents, being cheap, quick in its action, and giving as good solutions as the more expensive solvents. The relative amounts of these alkalis re- quired to give a neutral. solution vary with the different caseins as the following figures from a large number of tests made with high-grade commercial solvents will show 1 lb. borax, Na 2 B 4 07 • 10 H 2 = 0.1736 — 0.2330 lb. NaOH average 0.2020 = 1.062 — 1. 1 59 lb. Na 3 P0 4 - 12 H 2 average 1.130 = 1. 1 58 — 1.872 lb. Na 2 HP0 4 - 12 H 2 average 1.419 In the case of the phosphates the reaction evidently ceases when dihydrogen sodium phosphate is formed and this supposi- tion is confirmed by the fact that the latter salt has no appre- ciable solvent action even when used to the extent of 50 per cent of the weight of the casein. Practical application of these solvents brings out certain marked differences; ammonia is one of the quickest solvents and is especially useful in dissolving the last traces of the casein, the phosphates give thinner solutions especially with muriatic caseins, while borax is a reliable general solvent for most work but cannot be used with satin white. In any case a consider- able excess is to be avoided as it is wasteful, darkens the color and with most caseins causes a very marked thickening of the CASEIN 337 solution. The full strength of a casein is developed if the alkali used is sufficient to give a solution neutral to litmus and while a moderate excess does not cause loss of strength, this does take place if a large excess is used, particularly if the heat- ing is prolonged. The amount of alkali required to give a neutral solution varies greatly with different caseins ; with borax the range may be as great as from 7 to 18 per cent. A diffi- cultly soluble casein may be improved by soaking it in 10 parts of water acidulated with acetic acid and then washing. This improves its solubility and reduces the amount of alkali re- quired. In very many caseins there is a small amount of insoluble matter which generally takes the form of white flakes resembling skins or envelopes from which the casein has been dissolved. In some cases these will dissolve on the addition of more alkali, but usually it is impossible to ehminate them in this way. They are light but bulky so that they appear to be present in large amount while in reality the percentage is very small, the worst sample which has been noticed containing only 1.64 per cent by weight. Under the microscope the great majority of these white flakes are found to consist of the hyphae of moulds so the old theory that they were particles of albumen precipitated by overheating of the milk is untenable. Additional proof of their nature has been obtained by moistening casein and allowing it to mould, then drying and grinding it to its original fineness. Four samples thus treated showed white flakes amounting to 15 to 80 times those originally present and all had the same microscopic appearance as those present before moulding. These flakes cause trouble by working up into the brushes of the coat- ing machines and after they have collected in quantity, dropping off onto the paper thus making spots which sometimes stick to the calender rolls and break the paper. They also cause the brushes to work "dead" and not spread the coating well. These flakes, as well as all other undissolved foreign matter or casein, can be easily and completely removed by centrifugal clarifica- tion of the casein solution and the good portion of a casein 338 COATED PAPERS which is dirty or partially insoluble can thus be obtained in fit condition to use. Solutions of casein are very liable to become spoiled, especially in the summer months, and such solutions cannot be used both because of their extremely bad odor and because the strength of the casein is very largely destroyed. Many substances have been proposed as preservatives for casein solutions, among them camphor, essential oils, salicylates, benzoates, hexamethylene tetramine, etc., but few of them are worth the added expense. The proper selection of alkalis has a large influence on the rate of spoiling, borax acting as a good preservative, while ammonia, phosphates or caustic soda permit rapid spoiling. Tests on the same casein with different solvents gave spoiled solutions in forty-eight hours when ammonia, caustic soda, or trisodium phosphates were used while if borax was employed the solution was good at the end of 16 days. Other experiments proved that 3 to 4 per cent of borax on the weight of the casein was sufficient, when mixed with other alkalis, to preserve the solu- tion as long as it would ever be necessary under normal working conditions, provided ordinary care was employed in keeping the tanks and mixers clean. For this reason it is a wise pre- caution to use a small amount of borax in preparing casein solutions. Another substance which acts as a preservative but which is used primarily as a waterproofing agent is formaldehyde. This is obtained commercially as a strong aqueous solution of the gas and should always be diluted with about ten times its volume of water before adding to the coating mixture. Over 2 per cent of the strong solution, based on the weight of the casein, will cause thickening while i| to 2 per cent can be added safely if it is well diluted and stirred in. This amount is sufficient to make the coating waterproof or washable, while even smaller quanti- ties produce very noticeable waterproofing effects. When small amounts of formaldehyde are used the waterproofing is slight at first but gradually develops so that paper which has been stored several weeks will show better results than when first made. TESTING CASEIN 339 Many other substances are capable of rendering casein insol- uble but as most of them cause curdling or thickening of the casein solution they cannot be applied when the coating is spread. If casein coated paper is moistened with solutions of salts of iron, lead, copper, aluminum, magnesium, zinc, etc., the casein is rendered insoluble to such an extent that the paper may be safely washed. The use of any of these materials means putting the paper through one more process and as the desired properties can be more cheaply attained by formaldehyde none of them are now employed; they do, however, possess possibilities worthy of future investigation. Under normal conditions of storage caseins are apt to become infested with worms, the larvae of either moths or beetles, and the deterioration thus caused is very marked. Caseins affected in this way require less alkali to give neutral solutions and in some cases may lose as much as 50 per cent of their original strength in one year. The period of greatest deterioration is found to be in the summer, when the worms are most active, while in winter very little change takes place. Obviously the best conditions for lengthy storage would nearly approach those of cold storage plants. The tests which may be applied to casein are of two general classes, chemical analyses for certain constituents as moisture, ash, fat, or nitrogen and empirical tests to show certain proper- ties as solubility, strength, and alkali required. Moisture may be determined by drying in a thin layer at ioo° to 105 C. for about two and a half hours, cooling, and weighing in a closed vessel to prevent reabsorption of moisture. All caseins are not equally sensitive to heat but none have been found which, under these conditions, suffer decomposition which is serious enough to cause any appreciable error. In fact most caseins may safely be heated much longer at this temperature, or even as high as 115 C, without vitiating the results. The ash in casein may be determined by burning 2 to 3 grams in a silica dish; platinum should not be used because of the presence of phosphates which might be reduced and injure the 340 COATED PAPERS platinum. In some cases the ash is gray and infusible while in others it melts to a clear, glassy mass. Rennet casein burns out readily and gives a high ash, while acid caseins are much harder to ignite free from carbon. Fat may be estimated by extracting the finely ground sample in a Soxhlet apparatus with ether and petroleum ether, evapora- ting the solvent, and drying and weighing the fat. A more accurate and rapid method is as follows: Soak 2 grams casein in 6 c.c. water in a small beaker and after about half an hour add with constant stirring 9 c.c. of concentrated sulphuric acid (sp. gr. 1.84). Pour the solution into a Babcock skim milk bottle and wash out the beaker with 5 c.c. water and 5 c.c. H 2 S0 4 . Fill to the base of the neck with dilute sulphuric acid (4 c.c. water and 5 c.c cone. H2SO4) and whirl in the Babcock centrifugal five minutes. Fill with dilute acid, whirl two minutes and while still hot read the fat on the graduated neck. The reading mul- tiplied by nine gives the percentage of fat in the casein. The secret of this method is in getting the concentration of the acid just right since if it is too strong the fat will char while if it is too weak the casein will not all be in solution and a reading will be impossible. Nitrogen is best determined by the usual Kjeldahl method. The factors for converting this to casein have been given above. These chemical analyses, while giving information as to the purity of the sample in question, do not show how it will work in practice, whether it is completely soluble, how much alkali it will require, whether the solution will be thin or thick, or whether the casein has good adhesive strength. For the manufacturer of coated paper these points are far more important than the chemical data and while such tests are for the most part com- parative they are quite simple to carry out and are sufficiently accurate for practical purposes. The alkali required, the consistency of the solution, and its completeness may be determined at one operation as follows: Soak 50 grams of the finely ground casein in 200 c.c. of water in a weighed No. 3 breaker for about half an hour then add a weighed TESTING CASEIN 341 amount of the solvent and heat on the steam bath with constant stirring. The amount of solvent should be less than will be required to give a neutral solution and the kind is immaterial so long as it is capable of being accurately measured; borax and triso- dium phosphate are very satisfactory or a standard caustic soda solution may be added from a burette in place of weighing the dry solvent. When the alkali added has all been used up test the solution by dipping in it a moistened piece of blue litmus paper. If the reaction is acid add more alkali and repeat the heating on the steam bath. This operation should be continued until the solution reacts neutral to litmus, which will be when it turns red litmus slightly blue and blue litmus slightly red. If at this stage the solution is not complete it is well to add more alkali and see if this excess will bring into solution all of that which is insoluble at the neutral point. This test indicates the minimum amount of alkali which can be used with the casein in question, shows whether it is good enough to use without clarifying the solution and by the consistency of the solution tells the expe- rienced man much as to the kind of alkali to use and the manner of running it on his coaters. The test is of such a nature that, so far as the consistency of the solution and its completeness go, it does not lend itself readily to numerical expression. For this reason the interpretation of the test requires experience and this cannot be imparted by words. The strength of a casein or the amount of clay which a given weight of casein will hold on the surface of the paper is best deter- mined by a method approximating actual coating operations. The apparatus is simple; a porcelain cup (without a handle), a brass plate, a steel scraper so shaped that a thin coating of the clay and casein may be applied to a sheet of paper, a copper stirrer flattened and bent at one. end so that the contents of the cup may be thoroughly worked over, and a set of scales capable of weighing down to 0.1 gram. One hundred grams of clay, pre- viously dried at ioo° C, are weighed out into the cup and soaked up with 70 c.c. water. The casein solution which was prepared in the solubility test is so adjusted, by evaporating or adding 342 COATED PAPERS water, that each gram of dry casein is equivalent to 5 grams of solution. The clay and water are now worked over with the copper stirrer, the whole balanced up on the scales, 30 grams of casein solution added and the mass stirred until homogeneous. A sheet of paper is then laid on the brass plate, a little of this coating mixture placed on one end and a thin but even coating applied by means of the scraper. What coating mixture remains on the scraper is returned to the cup which is again balanced up, 5 grams more casein solution added and another sheet spread. This is repeated until the dry casein used amounts to 12 grams per 100 grams of clay. The coated sheets, which are marked from 6 to 12, according to the amount of casein, are allowed to dry and then by looking through the sheets places of uniform thick- ness are selected and marked on each. Short pieces of high grade sealing wax are then melted on one end, either in a gas name or on the surface of a steam heated box, and applied to the marked places with a firm pressure. They are allowed to become thor- oughly cold and are then removed from the paper by a steady vertical pull. If enough casein is present the surface of the wax will be covered with fibres to the very edge, if too little has been used the coating comes away from the paper without pulling off any fibre, while there is usually an intermediate case in which the center of the wax shows fibre and the edges clay only; this would be considered just on the line between weak and strong. This is an excellent comparative test but it does not show the actual amount which can be used practically, but for several reasons indicates a higher strength than the same casein shows on a large scale. It will be found, however, that if this test shows one lot to be weaker than another it will be necessary to use more of the former than the latter when it is put on the coaters. The test is also influenced by the kind of clay, thickness of coating, nature of paper, etc., so that it is necessary to use standard materials in carrying it out. This means that, while it gives very valuable comparative results in the hands of experienced persons and under standard conditions, it is not safe to compare results obtained in different laboratories. STARCH 343 Albumen. Both blood- and egg-albumen are similar in some respects to casein yet differ from it by losing their solubility if heated to about 75 C, while casein may be treated with boiling water and still be soluble when the correct proportions of alkali are used. Solutions of either of these may be prepared by stirring the albumen into warm water to which a little spirits of ammonia has been added. The temperature should not be over 20 C. and the stirring must be frequent enough to prevent the albumen from collecting and sticking on the bottom of the container. Other substances are also used as assistants in dissolving albumen, as borax, magnesium sulphate, etc. The strength of albumen is very nearly equal to that of casein, 24 to 28 parts of albumen doing the work of 22 to 24 parts of casein. Coatings prepared with albumen are not rendered so water- proof by formaldehyde as are casein coatings; they may, however, be made washable by heating, preferably in the presence of steam. Mixtures of casein and albumen when treated in this way give washable coatings which are good for chromo and leather papers because of their capacity to absorb printing ink. While under certain conditions albumen gives a higher finish to paper than does casein yet it is seldom used because of its high price and in many cases its disagreeable odor. Starch. As an adhesive for coated paper work starch has many good points; it is clean, of good color, without odor, non- nitrogenous and hence not liable to putrefactive decomposition, has good strength and is cheap. The different starches, such as corn, wheat, cassava, potato, etc., have quite distinct character- istics yet all are sufficiently alike so that they may be discussed as a class. The simplest way to prepare an adhesive from starch is to stir the dry starch into 8 to 15 times its weight of water and heat to the boiling point. Different starches vary as to the temper- ature at which they gelatinize and the thickness of the paste they produce but with reasonable amounts of water all are too thick to use on the ordinary coating machine. This difficulty may be 344 COATED PAPERS overcome by modifying the starch by chemical treatment so that it cooks thinner while at the same time losing nothing in strength but in many cases actually gaining in adhesive properties. Such modified starches may be produced by treatment with acids, acid salts, oxidizing agents, etc., under very varied conditions and the patents taken out along such lines are almost innumer- able. The products of these treatments resemble the original starches in appearance but as stated give much thinner solutions and also have, according to their method of preparation, minor differences which sometimes cause trouble on the coaters. A well made modified starch for coated paper work should give a light colored, thin solution when one part of starch is boiled with four parts of water. This solution should not thicken too much on cooling or at least should thin down to its original consistency on reheating. While there are many such products only a few give satisfactory mixtures with clay as most are lack- ing in the property which glue and casein possess of making the clay into a very fluid suspension. The lack of this quality causes the coating mixture to work "dead" or draggy and the paper is apt to show brush marks, or if these results are to be avoided so much water has to be added that the drying lines are overtaxed and the capacity of the machine is reduced. The strength of starch has been found to be slightly less than that of casein so that about 20 to 25 lbs. are required to do the work of 18 to 20 lbs. of casein. Coated papers prepared with starch do not take such a high, glossy finish as casein coated papers, partly because of the larger amount of adhesive used and partly because of the nature of the starch itself. Even the addition of considerable amounts of wax does not enable it to take so good a finish as casein coated paper. This does not prevent half tones taking well on it for its surface is sufficiently smooth and even to print well. A characteristic feature of starch coated paper is the porosity or absorbent power of the surface. This seems to be greater than with glue or casein coatings so that the printing ink tends to sink in rather more and the resulting cut, especially with color work, is slightly dull. CLAY 345 This trouble can be overcome by a proper adjustment of the ink but at present it is delaying the general introduction of starch coating, though otherwise the paper works well, running well on the presses, permitting rapid work and requiring no slip-sheeting. Miscellaneous Adhesives. Other materials have been pro- posed from time to time as assistants to, or substitutes for, casein and among these may be mentioned glutin, viscose, shellac, algse, vegetable gums, mucilages, etc. While these may be, and probably are, used in small quantities or for special purposes, as for instance shellac in the manufacture of imitation leather papers, their employment is by no means general and it is not probable that they will ever seriously compete with casein. Clay The nature and properties of clay have been discussed in the chapter on fillers and the methods of testing which are given there apply equally well to coating clays. The principal differences between filler and coating clays are in color and fine- ness, the coating grades being whiter and finer and containing less grit. These differences are not by any means along hard and fast lines since the higher grade filler clays are sometimes used for coating, while for certain kinds of high class book papers, good grades of coating clay are used in the beating engines. The presence of grit in clay for coating is more serious than in a filler since it is sure to appear on the surface of the paper where it inter- feres with the finish and may even cause trouble in printing, especially in lithographing. In this process it is said to etch the stones or plates so that the portions which should remain white are not entirely ink resistant thus permitting the ink to be trans- ferred to the paper when it is not desired and giving tinted or mottled backgrounds. The fineness of the clay has a large influence on the finish which the paper will take on calendering, the finer the particles of the clay the higher the gloss which is imparted. Unfortunately the amount of casein, or other adhesive, required increases quite rapidly as the size of the clay particles decreases so that the gain in finish due to a fine clay is in part offset by the effect of more adhesive. Unexpected or unknown changes in the fineness of 346 COATED PAPERS the clays used are without doubt responsible for some of the cases of weak coating, especially when the amount of casein used is kept as near the minimum as possible. With ordinary clays from 15 to 18 parts of casein are required for 100 parts of clay and with this casein any of the ordinary solvents as borax, soda ash, ammonia or mixtures of these solvents may be used with good results. Blanc Fixe and Barytes. Both of these are chemically barium sulphate, BaS0 4 ; blanc fixe being prepared by precipitation while barytes is the natural mineral ground and bolted to any desired degree of fineness. The best grades of blanc fixe are prepared from witherite (BaC0 3 ) by dissolving in muriatic acid, filtering and precipitating with sulphuric acid. The precipitate is washed practically free from acid and put on the market either dry or as a paste containing 25 to 30 per cent of moisture. Cheaper grades of blanc fixe are produced as by-products in the manufacture of hydrogen peroxide, etc., and appear to be only slightly inferior to that from witherite in color and cleanliness. Under the microscope blanc fixe is seen to consist of extremely fine crystals, which are very uniform in size. If large irregular shaped pieces are present it may be taken as an indication of adulteration with barytes or of very careless handling of the solutions before precipitation. The grit in blanc fixe, as determined by the flotation test described in the chapter on fillers, should not exceed 0.2 to 0.3 per cent and it should con- sist almost entirely of small lamps of the blanc fixe itself which have not broken down during the test. This test also will show the presence of barytes though not its amount. The reaction of blanc fixe varies commercially from neutral to decidedly acid; both appear to give equally good results under ordinary working conditions as the acid is neutralized by the excess of alkali in the casein. Barytes, being a ground mineral, gives much larger amounts of grit than blanc fixe, the flotation test showing from 8 to 15 per cent for different commercial grades. Its particles are much coarser and more variable in size than those of blanc fixe, and SATIN WHITE 347 it is usually quite inferior to the latter in color. For these reasons it should not be used in the highest class of papers. Blanc fixe is one of the whitest of the minerals used in coating paper and can be used in the very best of products. It does not take such a high finish as clay or satin white and is especially serviceable in making dull finish coateds since it has less tendency to scratch than clay or satin white. Since barium sulphate is practically insoluble it cannot react with casein solutions so that almost any solvent can be used in preparing the latter. Both blanc fixe and barytes require much less casein than does clay and both tend to settle out of the coating mixture more rapidly than clay on account of their high specific gravity. Satin White. This material consists essentially of calcium sulphate and aluminum hydrate, formed by the interaction of slaked lime and aluminum sulphate; 3 Ca (OH) 2 + Al 2 (S0 4 ) 3 = 3 CaS0 4 + Al 2 (OH) 6 . The alum used may be potash or ammonia, or aluminum sulphate itself may be used, and the character of the resulting product depends very largely on which is employed. The slaked lime is frequently used in excess and the amount of this excess together with the quality of the lime has a very great influence on the working properties and color of the product. The following analyses of two commercial sam- ples of satin white show its approximate composition after being dried at ioo° C. Sulphur trioxide, SO Alumina, AI2O3 Total lime, CaO Loss on ignition .... No. 1 No. 2 Per cent Per cen 28.9 29 -3 13 -9 12.3 39- 1 39-7 17-5 18. 5 99-4 99.8 Both these samples contained 24 to 26 per cent of free lime while other samples tested for this substance have shown that it may run as high as 30 per cent, or the satin white may even be perfectly neutral. 348 COATED PAPERS In the commercial preparation of satin white the slaked lime is mixed with alum solution, or in some cases the undissolved alum, and after sufficient agitation to insure a uniform reaction the paste is diluted, strained through fine wire gauze, 130 to 140 mesh, and run into a filter press where it is washed with clear water. The resulting paste containing about 30 per cent of bone dry material is the satin white of commerce. When a standard method of manufacture has once been established it should be strictly adhered to, since changes in the kind or proportions of the ingredients may cause differences which may not appear in the analysis yet which will cause serious trouble in the coating room. In fact it cannot be safely predicted from the analysis how a lot will work and the only sure way is to give it a trial on a practical scale. The properties imparted to coated paper by satin white are high gloss on calendering and clear white color. Because of the presence of aluminum hydrate the coating becomes rather dense and brittle so that papers with much satin white have rather poor folding qualities; this is especially true with very heavily coated papers. The amount of casein required to hold a given amount of satin white is much greater than for the same amount of clay, prob- ably in most cases nearly one and one half times as much. Special precautions have to be used in mixing the two or the result is a thick, curdled mass which cannot be spread. In order to avoid this difficulty the casein solution may be prepared with an ex- cess of soda ash or with a mixture of sodium phosphate and ammonia and the satin white should be mixed with a little ammonia before the casein is added. When phosphates are present with satin white heating causes thickening so that hot casein solution should not be mixed with satin white, nor should the mixture be heated. For working with this material temperatures of 30 C. or below are desirable. If these precau- tions are observed very little trouble will be encountered. Accessories. Under this heading may be mentioned those materials which are used in small amounts or for special pur- ACCESSORIES 349 poses and which do not form an essential part of the coating itself. Soaps and waxes are added for the purpose of improving the finish obtained on calendering. Among those used may be men- tioned, beeswax, carnauba wax, stearic acid, paraffin, Japan wax, white soap, lard oil, etc. Many different recipes are used for mixing the ingredients but most of them are suspensions or emulsions of one or more substances in a soapy medium. Japan wax has the property of saponifying very readily and is quite generally used to hold the other materials in suspension. As an example of this type the following formula may be cited; to about 250 gals, of hot water add 50 lbs. each of Japan wax, para- ffin and stearic acid and then 22 lbs. borax and a little ammonia. When stirred until the waxes are melted this forms a creamy white permanent emulsion which is ready to use in the coating mixture. While it is undoubtedly true that such materials assist in obtaining the high finish on glace papers repeated trials have demonstrated that they are of doubtful value for the ordinary grades of coated papers and it has been proved that at times some of them may be responsible for poor results in printing. Closely connected with this class of materials are those added to prevent undue frothing of the coating mixture; in fact some of the substances sold to improve the finish are also claimed to reduce the froth. Anti-frothing substances may have widely different characteristics as for instance wood alcohol, Turkey red oil, fusel oil, skim milk, and gasoline. Probably no one of these will work in every case as froth varies in its character and cause but it is believed that a little gasoline added to the frothing mix- ture from time to time will give the best results. This sub- stance should be used with great caution on account of the danger of fire. For softening the coating and increasing its pliability glycerine is frequently recommended. Its beneficial effect is supposed to be due to its hygroscopic nature but tests have shown that com- mercial glycerines absorb very little moisture from the air so that the increase in weight of the paper would be practically only that 350 COATED PAPERS of the glycerine added. Tests by a large manufacturer indicate that amounts up to 2 per cent of the weight of dry clay in the coating have practically no effect on the folding properties of the coated paper and if 5 per cent is used the printing qualities of the paper are seriously injured. The glycerine substitutes, mostly invert sugars, are even less hygroscopic than glycerine itself and hence would have even less effect. The use of any of these substances is therefore of doubt- ful advantage and their use is not to be recommended. CHAPTER XIII WATER Water which is pure, in the sense that it contains no foreign matter of any kind, is never found in nature, so that from the manufacturing standpoint, as well as from that of sanitation, water must be considered with reference to the amount and kind of the impurities which it contains. These may include solid, liquid or gaseous substances and they may be either in suspension or solution. Both mineral, or inorganic substances, and organic materials may be present, and the latter may be derived from decaying vegetation or from minute living organ- isms. Some of these substances may have a great influence on the quality of the paper, while others, which are harmless for this purpose, are decidedly bad if used in boilers. The quality of the available water supply is therefore a vital consideration in connection with the manufacture of paper. The comparative readiness with which the quality of the water affects that of the paper is still further emphasized by a consideration of the very large volumes necessary for the vari- ous manufacturing operations. Griffin and Little 1 estimate the amount required in making a ton of paper at 50,000 to 200,000 gals., or about 200 to 800 times the weight of the paper pro- duced. Others have estimated that in American fine paper mills, making linens, bonds and ledgers from rag stock, the water used amounts to 1000 gals, per pound of dry paper made. 2 Reliable data for water consumption are rarely available and no generally applicable figures can be given because its use depends so largely on local conditions, as purity, quantity avail- 1 Chemistry of Paper Making, p. 330 (1894* 2 Paper: June 21, 1916, p. 12. 3Si 352 WATER able, cost of pumping, necessity for filtration, reuse of back- water, etc. In some cases it may be necessary to purify the waste waters in order to avoid stream pollution and this tends to reduce the amount of water used. Considering these enor- mous volumes, and the fact that the paper stock forms a very effective filter, the result of very small amounts of injurious sub- stances may be easily imagined. Compared with this problem that of the boiler-house supply for steam raising is relatively unimportant. Waters may be broadly classified as (i) rain, (2) surface and (3) ground waters. Rain water, if properly collected, is the purest form of natural water, though it always contains gases and impurities from roofs, products of combustion, etc. Be- cause of the relatively small amounts available it is of no prac- tical importance as a paper making supply. Surface waters include those of brooks, rivers, ponds and lakes. These waters pick up impurities of various kinds according to the nature of the soils over which they flow and they are also contaminated by mineral substances derived from springs which discharge into them. They generally contain less mineral matter but more organic matter than ground waters, and, particularly in the case of river waters, are likely to vary greatly in composi- tion at different periods of the year. Suspended matter is usu- ally present in greater or less amount and in the case of swamp waters there is usually a yellowish color due to the peaty soil over which the water has passed. Waters of this type are likely to contain plant and animal life which may impart con- siderable color to them. Ground waters are those which have percolated through a considerable depth of soil and the under- lying porous strata. Such waters, derived from springs and deep wells, are usually clear and colorless but they contain more dissolved mineral matter than do surface waters. Soft waters are those which contain little of those mineral substances which are capable of decomposing soap, while hard waters are those which possess this property to a marked de- gree. The most common cause of hardness is the presence of . WATER 353 lime salts, either the sulphate or carbonate, the latter being much the more general. The salts of magnesium have an even greater effect than those of calcium but they are not so often present. The sulphates of both calcium and magnesium are soluble in pure water but the carbonates require the presence of carbon dioxide to enable them to dissolve as bicarbonates. The necessary carbon dioxide is derived from the air, from the decay of vegetable matter or from subterranean sources. Bicar- bonates form what is termed "temporary hardness" since on boiling the carbon dioxide is driven off, causing the precipita- tion of calcium carbonate and the softening of the water. Sul- phates cause "permanent hardness" since they are not affected by boiling. Soft water is said to be desirable for the washing of stock because it has a greater solvent power than hard water. If the latter is employed for washing sulphite fibre there is a tendency for insoluble calcium resinates to be deposited on the fibre, thus rendering the product unnecessarily hard to bleach, while in the soda process the lime salts may be precipitated as car- bonate or sulphate and carry down coloring matters with the same result. 1 Water which is very hard because of the pres- ence of calcium bicarbonate may also be injurious to sizing, though the presence of calcium sulphate is harmless. For other purposes in paper making as in the boiling of either sulphite, soda or rag stock, or in bleaching, or furnishing an engine, the importance of soft water is greatly overestimated since the materials employed will immediately harden the softest water. In making colored papers the quality of the water may affect the results obtained. Carbonates cause precipitation of the salts of iron, tin and aluminum, which are sometimes used as mordants, and reduce their effectiveness. Sulphates have little or no action. Neither carbonate nor sulphate will cause trouble when using acid colors as the amount of acid used far outweighs any alkali present in the water. Iron in the water dulls almost all mordant colors. It is also very injurious in the manufac- 1 Beveridge: Paper, Oct. 30, 19 18. 354 WATER ture of photographic paper, particularly if it is present in such form that it may deposit in the pipes, reservoirs, etc., as such sludge may break away at times and cause endless trouble. 1 The most important quality of water, from a paper making point of view, is its color. The purest natural waters are clear and colorless when examined in comparatively thin layers, while in large masses they have a bluish tint. Surface waters on the other hand show all variations in color from colorless, through yellowish and reddish tints, to the deep brown of swamp water. This color is due largely to decaying vegetable matter and where the decay has proceeded far the color is very permanent in character. The color derived from plant growths, princi- pally algse, is usually most pronounced in the .summer months. They are green or bluish green, require light for their develop- ment and thrive best in ponds or reservoirs where there is little movement to the water. These lower forms of plant life are very sensitive to copper sulphate and a treatment with a very small amount of this material is sufficient to destroy them. This method is often used, even in the case of a water supply for drinking purposes. At times the color of a water is greatly affected by the sus- pended matter which it carries; this is a very variable factor, its greatest effect appearing when heavy rains have washed much finely divided soil into the streams. Soluble mineral matter has, as a rule, little effect on the color, even the soluble salts of iron are seldom present in sufficient concentration to cause any perceptible color. Trouble is caused, however, when for any reason these iron salts are precipitated as hydroxide. Practically all waters, especially surface waters, consume small quantities of bleaching powder, the amount depending upon the source of the water. In most cases the loss of bleach from this cause is slight but if much organic matter is present, as in swamp waters or those largely contaminated with trade wastes or sewage, the loss may be appreciable. Griffin and Little 2 1 Anon: Paper, Jan. i, 1919. 2 Chemistry of Paper Making, p. 333 (1894). BOILER SCALE 355 give the bleach consumed by three waters as 1.77, 1.16 and 3.87 grains per gallon but do not state the source of the water or the conditions of the test. Water which is to be used for boiler purposes should gener- ally be soft and free from suspended matter, for substances either in solution or suspension will accumulate and form mud or scale. Such a deposit may be hard and dense or loose and porous according to the substances present. Dense scale is the more serious and water which causes it should either be avoided or treated in some way. Such a scale causes a large loss of fuel by preventing the transfer of heat to the water and when it is of any considerable thickness there is danger of overheating the boiler locally with consequent damage to the tubes or plates. Hardness, however, is not always an indication of scale forming power for both calcium chloride and magnesium sulphate make water hard but do not form scale. One of the most frequent causes of scale is calcium carbonate. This is present as bicarbonate but at the temperature of the boiler is again broken down into calcium carbonate and carbon dioxide. This same reaction takes place with magnesium bicar- bonate and if both are present in the water they will be found together in the scale. The precipitated carbonates are at first loose and powdery but if the boiler is blown off without cooling the flues, the precipitate is likely to bake into a dense, hard scale. Under these conditions the magnesium appears in the scale as hydroxide. Calcium carbonate is not so likely to bake onto the plates as the magnesium salt. The following analysis x shows the general characteristics of a carbonate scale, though the relative proportions are likely to vary quite widely in different samples. Per cent p er cen t Carbonate of lime 75. 85 Silica 7. 66 Sulphate of lime 3 . 68 Oxides of iron and alumina 2. 96 Hydrate of magnesia 2. 56 Organic matter 3. 64 Chloride of sodium o. 45 Moisture . 20 1 Griffin and Little: Chemistry of Paper Making, p. 334 (1894). 356 WATER If the hardness of a water is due to calcium sulphate no pre- cipitation takes place until the solution becomes saturated due to concentration; then a crystalline scale deposits. This is at first CaS0 4 • 2 H 2 0, but it begins to lose its water of crystal- lization at about 260 F. and becomes more insoluble. These actions tend to produce a hard scale which may accumulate to a considerable thickness. A certain amount of scale may be formed in the boilers even when the water is soft, though the quantity is almost never enough to be serious. The following analyses show the com- position of a soft water which proved excellent for boiler pur- poses and also of the scale which formed during its use. Water * Constituents Parts per million Per cent of dry matter Silica (Si02) Iron (Fe) Calcium (Ca) Magnesium (Mg) Sodium and potassium (Na + K) . Carbonate radicle (C0 3 ) Bicarbonate radicle (HCO3) Sulphate radicle (SCU) Nitrate radicle (NO3) Chlorine (CI) Total suspended solids Total dissolved solids 3-8 0.04 3-2 0.6 4.2 0.0 14.0 3-6 °-5 1.6 7-4 25.0 i5-5 0.2 i3- 1 2.4 17. 1 28.5 14.7 2 .0 6-5 Scale Moisture and organic matter . Silica (SiO-2) Ferric oxide (Fe 2 3 ) Alumina (A1 2 3 ) Calcium oxide (CaO) Magnesium oxide (MgO) Carbon dioxide (CO2) Sulphur trioxide (SO3) Per cent 17.16 35-64 4-32 4.28 30.97 5-59 1 .40 Q-9 2 100.28 Analysis by U. S. Geological Survey. Water Softening. Many so-called "boiler compounds" have been proposed for use in the prevention of scale and while some WATER SOFTENING 357 may work well under certain conditions, none are of universal application and many are useless or even injurious. If the water is bad enough so that a "boiler compound" seems de- sirable, it is in most cases preferable to give it a preliminary softening treatment. For waters containing the bicarbonates of calcium and mag- nesium the method most generally employed is that of Clark, which consists in adding slaked lime to react according to the equation : CaH 2 (C0 3 ) 2 + Ca(OH) 2 = 2 CaC0 3 + 2 H 2 The lime is added either as lime water or milk of lime, the former being preferable because it is more easy to control. This treatment removes all the bicarbonate as well as the added lime. This method was originally applied in large tanks where the water could remain quiescent for a considerable time. The lime water was added in such an amount that, after thorough mixing, the water gave a yellow or brown color with silver nitrate, then more water was added until the silver nitrate test gave no color. This test should always be applied in order to make sure that an excess or deficiency of lime is not caused by variation in the water supply. Temporary hardness of course is all that is removed, and even that not completely, as a little calcium carbonate remains in solution and magnesium bicarbonate is not so completely removed as that of calcium. Salts of iron and some organic matter are also removed. Various modifications of this process have been tried out in the attempt to shorten, or do away with, the time required for settling. The Porter-Clark process employs a filter press for the removal of the precipitate. Gaillet and Huet cause the water to take a zig-zag course to promote sedimentation and often use both lime and caustic soda to remove both temporary and permanent hardness according to the reactions CaH 2 (C0 3 ) 2 + 2 NaOH = CaC0 3 + Na 2 C0 3 + 2 H 2 0, CaS0 4 + Na 2 C0 3 = CaC0 3 + Na 2 S0 4 . 358 WATER The Archbutt-Deeley process blows air through the treated water after mixing it with the sludge from a previous precipi- tation. This hastens the sedimentation so that the clear water may be drawn from the top by a floating arm. Other proposed modifications employ barium hydroxide, sodium oxalate, mag- nesium hydroxide, etc. The latter is not always to be recom- mended as it may result in the formation of magnesium chloride and consequent corrosion of the boiler. An entirely different process is the so-called permutite pro- cess. Permutite is an artificial zeolite formed by fusing together silica, alumina or china clay, and sodium carbonate. On ex- traction with water a crystalline body is obtained of the approxi- mate composition, Si0 2 , 46; A1 2 3 , 22; Na 2 0, 13.6; H 2 0, 18.4. Filtration of water through this material removes the calcium and magnesium thus: Na 2 Al 2 Si 2 8 + CaC0 3 = Na 2 C0 3 + CaAl 2 Si 2 8 , Na 2 Al 2 Si 2 8 + CaS0 4 = Na 2 S0 4 + CaAl 2 Si 2 8 . It is claimed that the whole of the hardness is removed and that the calcium and magnesium zeolites formed can be recon- verted into the original permutite by treating with a strong solution of salt. A manganese permutite can also be used for the removal of iron. Filtration. The principal action in filtering water is that of straining to remove suspended mineral and vegetable matter. It is, however, much more complex than a mere straining since the water undergoes chemical and biological changes, particu- larly in slow sand-filtration. Among these changes oxidation is one of the chief, due to the fact that in passing through porous earth an enormous surface is exposed. In natural filtration, during the transformation of surface into ground waters, this action may proceed so far that all the organic matter originally present is oxidized leaving the water clear and colorless. In continuous mechanical filtration no such oxidation takes place and the removal of organic matter has to be secured by other. means. Many forms of mechanical filters have been devised, some WATER ANALYSIS 359 working by gravity and some under pressure. With all of them some provision is made for cleaning the filtering medium, which is usually sand, either by reversing the flow of water, by me- chanical agitation with rakes, by compressed air or some other means. The frequency of such cleaning depends on the quality of the water, and if this varies much, due to the time of year, storms, etc., the intervals between cleaning will also vary. If the stream from which the water is drawn is contaminated with fibre, sawdust or other coarse mechanical impurities the effi- ciency of the filters may be greatly increased by first running the water supply over a revolving screen similar to the moulds of a wet machine. Water may be filtered directly, in which case only suspended matter will be removed, or it may first be treated with a little alum. The slight alkalinity of most waters causes the precipi- tation of hydrated alumina which on standing for some time collects in flocks of appreciable size, gathers together finely divided suspended matter and at the same time takes out much of the organic coloring matter. The precipitate forms a film over the filtering medium and aids in removing bacteria and other minute organisms. If much loam or silt is present a little lime is sometimes added before the alum to insure a suffi- ciently bulky precipitate. The amount of alum necessary de- pends on the hardness of the water. It is customary to add it mechanically by some sort of device controlled by the flow of the raw water so that the proportion added may be the same at all times. After the addition of the alum it is desirable that the water remain for thirty minutes to an hour in a settling basin where heavy sediment may settle and the chemical reac- tion between the alum and the bases present in the water may take place. It is then ready to pass to the filter. Water Analysis. The following analytical methods have been taken in large part from "Standard Methods for the Examina- tion of Water and Sewage," J to which reference should be made if more complete details are desired. 1 Am. Public Health Assn., Boston, 1917. 360 WATER Sampling. Care must be taken that the sample is truly rep- resentative of the liquid to be analyzed. If for any reason variations are likely to occur, as for instance because of periodic contamination by trade waste, a sample obtained by mixing together several portions taken at different times is likely to be more representative than one taken all at one time. The amount required for the ordinary chemical, physical and micro- scopical analysis is not less than two liters, and for special tests larger quantities may be required. The samples should be collected in glass-stoppered bottles which have been previously cleaned with sulphuric acid and potassium bichromate or with alkaline permanganate, followed by a mixture of oxalic acid and sulphuric acid and by thorough rinsing and draining. The stoppers and necks of the bottles should be protected from dirt by tying cloth or thick paper over them. The time which may safely elapse between the collection of the sample and its analysis depends on the character of the sample and the tests to be made. No exact limits can be fixed but it is considered that the analysis should be begun within the following times. Physical and chemical analyses Microscopic examinations Ground waters Hours 72 48 12 Hours 72 24 Fairly pure surface waters Polluted surface waters . . . In general the shorter the time elapsing between the collection of a sample and its analysis the more reliable will be the ana- lytical results. Turbidity. The turbidity of water is due to suspended mat- ter such as clay, silt, finely divided organic matter, microscopic organisms, etc. The standard for turbidity adopted by the United States Geological Survey, 1 is a water containing 100 parts per million 1 F. D. West: Proc. 111. Water Supply Assoc, Vol. VI, pp. 49-51, 1914- COLOR 361 of silica in such a state of fineness that a bright platinum wire 1 mm. in diameter can just be seen when the centre of the wire is 100 mm. below the surface of the water and the eye of the observer is 1.2 meters above the wire. This observation should be made in the middle of the day, in the open air but not in sunlight, and in a vessel so large that the sides do not shut out the light so as to influence the results. The standard may be prepared by sifting dry Pear's "precipitated fuller's earth" through a 200-mesh sieve and suspending 1 gram of the mate- rial in 1 liter of distilled water. This has a turbidity of 1000 but should be tested by diluting with nine times its volume of water and trying out with the platinum wire apparatus. If not exactly right it may be adjusted by adding more silica or more water. This method requires a rod with a platinum wire 1 mm. in diameter inserted in it 1 in. from one end and projecting from it at least 25 mm. Near the other end of the rod and 1.2 meters from the wire is fixed a small ring, directly above which the observer places his eye when making the observation. This rod is graduated so that when lowered into the water to be tested, as far as the wire can be seen, the level of the water on the graduated scale indicates the turbidity. Turbidity may also be determined by the candle turbidi- meter, 1 which consists of graduated glass tubes with flat pol- ished bottoms, enclosed in a metal case. This is supported over an English standard candle so that the distance from the bottom of the tube to the top rim of the candle shall be 3 ins. The observation is made by pouring the sample of water into the tube until the image of the candle flame just disappears. This test should be carried out in a darkened room or with a black cloth over the head. An electric light may conveniently be substituted for the candle as it avoids any deposit of soot or moisture on the tubes. The calibration of the instrument may be made by means of the silica standards already mentioned. Color. The "true color" of water is that due to substances in solution while the "apparent color" is that of the original 1 Tech. Quart., Vol. XIII, pp. 274-279, 1900. 362 WATER unfiltered sample and includes also any color caused by sus- pended matter. A convenient standard for color 1 is prepared by dissolving 1.246 grams of potassium platinic chloride (PtCl 4 • 2 KC1), con- taining 0.5 gram of platinum, and 1 gram crystallized Cobalt chloride (C0CI2 • 6 H 2 0), containing 0.25 gram of cobalt, in water with 100 c.c. concentrated hydrochloric acid and diluting to 1 liter with distilled water. This solution has a color of 500 and by diluting in Nessler tubes standards of o, 5, 10, 15, 20, 2 5> 3°> 35? 4°> 5°> 60 and 70 should be prepared. These tubes should be of such diameter that the graduation mark is 20 to 25 cm. above the bottom and of such uniformity that the dis- tance from the bottom to the graduation of the longest tube shall not be more than 6 mm. greater than that of the shortest tube. The tubes should be protected from light and dust when not in use. The color of the sample is measured by filling a standard Nessler tube to the height equal to that in the standard and then comparing by looking vertically downward through the tubes upon a white surface placed at such an angle that light is reflected upward through the liquid. Water with a color greater than 70 should be diluted before testing, and water con- taining suspended matter should be filtered unless the apparent color is desired, in which case unfiltered water should be used. Nitrogen. This occurs in water in various forms as ammonia, nitrites, nitrates, etc., and its determination has been standard- ized by numerous methods. It is of very great importance when considering water from a sanitary standpoint, as its pres- ence in any appreciable quantity indicates pollution, but as it is of minor importance in the manufacturing operations of paper making the methods for its determination will not be discussed. Oxygen Consumed. This is determined by heating 100 c.c. of the water with 10 c.c. of dilute H 2 S0 4 and 10 c.c. of a stand- ard solution of potassium permanganate, adding 10 c.c. of a 1 Am. Chem. J., Vol. XIV, pp. 300-310, 1892. SUSPENDED MATTER 363 standard ammonium oxalate solution and titrating the excess with the permanganate. It is considered by some as an indication of the amount of carbonaceous matter present but where pollution, or contami- nation with trade wastes, has taken place it will also include such materials as nitrite nitrogen, ferrous iron, sulphides, etc. As a substitute for this determination it is suggested that a direct determination of the bleach consumed by the water under conditions of temperature and time similar to those of actual fibre bleaching operations would be of much more value to the paper maker. This can be carried out by adding a known vol- ume of a standard hypochlorite solution and after allowing it to stand for a definite time determining the amount remaining; the difference represents the amount consumed. Residue on Evaporation. Ignite and weigh a clean platinum dish, and measure into it 100 c.c. of the thoroughly shaken sample. Evaporate to dryness on the water bath, heat in an oven at 103 C. or 180 C. for one hour, cool in a desiccator and weigh. The increase in weight gives the total solids or residue on evaporation. The results should be expressed as parts per million and the temperature of drying should be given in the report. Fixed Residue. Ignite the total solids in the platinum dish at a low red heat. Cool, moisten the residue with a few drops of distilled water, dry in the oven, cool in a desiccator and weigh. For the greatest accuracy an electrically heated muffle furnace should be used. The loss on ignition is the difference between the total residue on evaporation and the fixed residue on evaporation. Suspended Matter. This may be determined directly by means of an asbestos lined Gooch crucible or indirectly by cal- culation from the difference between the total solids in filtered and unfiltered portions of the same sample. A test which gives valuable information regarding the paper making value of water is made by treating a definite volume of the water with enough dilute aluminum sulphate solution to 364 WATER precipitate all coagulable matter. After allowing it to stand a sufficient time for coagulation to take place, it is filtered through a Gooch crucible or through a tared filter paper and the weight of the precipitate determined by difference after drying. The amount and character of the precipitate indicate whether the water will prove satisfactory without filtration or whether it will be necessary to treat it with alum and filter before using. Hardness. A water is said to be hard when it contains in solution mineral constituents, which form insoluble compounds with soap. The most accurate method of ascertaining total hardness is to calculate it from the amounts of calcium and magnesium found by analysis of the sample. If appreciable amounts of iron or other metals are present these must be included in the calculation. The total hardness expressed as CaC0 3 equals 2.5 Ca plus 4.1 Mg. Titration with standard soap solution is a method often em- ployed for determining total hardness, though in reality it merely measures the soap consuming power of the water. The soap solution is prepared by dissolving 100 grams of dry, white Castile soap in 1 liter of 80 per cent alcohol. After standing several days dilute 100 c.c. to 1 liter with 70 per cent alcohol and standardize against a solution of calcium chloride contain- ing the equivalent of 0.2 gram CaC0 3 per liter. This is done by placing 20 c.c. of calcium chloride solution in a 250-c.c. glass- stoppered bottle and diluting to 50 c.c. with distilled water which has been recently boiled and cooled. To this the soap solution is added from a burette, 0.2 to 0.3 c.c. at a time, shak- ing vigorously after each addition, until the lather remains unbroken for five minutes over the entire surface of the water while the bottle lies on its side. Repeat this test, using 15 c.c, 10 c.c. and 5 c.c. of the calcium-chloride solution and finally distilled water alone, and from the results plot a curve showing the relation of various quantities of soap solution to correspond- ing quantities of calcium carbonate and hence to parts per million of hardness. IRON 365 In testing water 50 c.c. of the sample are placed in the bottle and the test carried out exactly as in the standardization of the soap solution. If magnesium is present a false end point may- be obtained. To see if this is the case, read the burette when the end point has apparently been reached and then add 0.5 c.c. more of the soap solution. If the end point was caused by mag- nesium the lather will disappear and the titration must be continued until the true end point is reached. Temporary Hardness. Temporary hardness is due to mag- nesium or calcium carbonates which are held in solution as bicarbonates by dissolved carbon dioxide, and which are partly precipitated when the latter is driven off by boiling. It is most accurately estimated by determining the alkalinity by titration with N/50 acid, using methyl orange as indicator, both in the original water and again after boiling, cooling, making up to the original volume with boiled distilled water and filtering. The difference between the two titrations represents temporary hardness which would include iron bicarbonate. Alkalinity or Acidity. Alkalinity or acidity may be deter- mined by titration with N/50 acid or sodium carbonate re- spectively. Indicators for alkalinity include phenolphthalein, methyl orange, lacmoid and erythrosine while for total acidity phenolphthalein should be used. Chlorides. The chlorides in water are due to common salt, which may come from mineral deposits in the earth, from wind- borne ocean vapors or from sewage and trade wastes. It may be determined by titrating a measured volume of the water with standard silver nitrate, using chloride-free potassium chromate solution as indicator. This shows a faint reddish color when the end point is reached. If the water has a color greater than 30 it should be decolorized by shaking with a little washed aluminum hydroxide and allowing it to settle. The clear portion is then used for the test. The results should be expressed as parts per million of chlorine. Iron. Iron may be present in both the ferrous and ferric states. In ground waters it is usually in the ferrous condition 366 WATER and combined with carbonic or sulphuric acid or with organic matter. After exposure to air it is frequently present as a col- loidal hydroxide. Silt-bearing waters often contain much iron in suspension, while waters contaminated with sewage effluents and manufacturing wastes contain various forms of iron of different degrees of solubility. The total iron present may be determined as follows: Evapo- rate ioo c.c. of the water to dryness and ignite at a low red heat. After cooling add 5 c.c. of concentrated hydrochloric acid, moisten the surface of the dish and warm two or three minutes, again moisten the inner surface of the dish with the acid and wash the solution into a 50-c.c. Nessler tube. Dilute to 50 c.c, add 3 drops of N/5 potassium permanganate solution and then add 5 c.c. of potassium sulphocyanide solution. Com- pare the color produced with that from solutions of known iron content similarly treated with sulphocyanide. The fixed residue from a previous determination may be dis- solved in hydrochloric acid, converted to sulphates by evapora- tion with sulphuric acid, and treated as in the modified Stokes and Cain method as applied to the analysis of alum. Manganese. The determination of manganese may be car- ried out colorimetrically if the water contains less than ten parts per million but if more than this amount is present it may be preferable to use a volumetric or gravimetric method. The amount of water used should contain not more than 0.2 mg. of manganese. Add to it 2 c.c. of nitric acid (concentrated acid diluted with an equal volume of water), and boil down to 50 c.c. Precipitate the chlorides with silver nitrate, being sure that a slight excess is used, and filter. Add about 0.5 gram ammonium persulphate crystals and warm the solution until the maximum permanganate color is developed, which usually takes about ten minutes. At the same time prepare standards containing 0.2, 0.4, 0.6 c.c, etc., of the standard manganous sulphate solution to about 50 c.c. and treat them in exactly the same way as the standard. Transfer sample and standards to Nessler tubes and compare the color at once. MANGANESE 367 The standard manganous sulphate is prepared by dissolving 0.288 gram of the purest potassium permanganate in about 100 c.c. of distilled water, acidifying with sulphuric acid, boiling and just discharging the color with a dilute solution of oxalic acid. After cooling and diluting to 1 liter, 1 c.c. of the solution contains 0.1 mg. of manganese. CHAPTER XIV TESTING WOOD PULPS In the purchase of wood pulp it is very desirable to have some means of judging its quality, as otherwise the material obtained is quite likely to be entirely unsuited to the grade of paper being made. The usual crude tests, such as feeling of the sheet, tearing it and noting the length of fibre, biting the fibre to note its hard- ness, etc., at best give but a slight idea of its suitability for the work in question and if intelligent purchases are to be made more definite methods of testing must be devised. Accurate sampling and testing are also very essential in the case of moisture determinations, since pulp is purchased on a 10 per cent moisture basis and large sums of money are frequently at stake. The following methods are those which have been gradually worked out during years of experience but they are given as a working basis rather than as final and unalterable procedures, since it is certain that many of them are capable of considerable improvement. Moisture in Wood Pulp. Baled Pulp. The following is the official method of sampling and testing baled pulp as adopted by the American Paper and Pulp Association and the Association of American Wood Pulp Importers. "All tests must be made by a chemist duly authorized and approved by the joint committee representing the Association of American Wood Pulp Importers and the American Pulp and Paper Association on one side and the Scandinavian Wood Pulp Associations on the other side, and must be made strictly in accordance with the following instructions — otherwise the com- mittee reserves the right to withdraw the approval of any chemist at any time. 368 LOCATION OF BORINGS 369 "Before proceeding to the weighing and sampling the chemist must ascertain that not less than half of the parcel in question is available. " Number. Not less than 5 per cent, nor more than 10 per cent of the entire shipment, but not less than ten bales shall be sam- pled. Samples to be drawn only from sound and intact bales, from different sections of the entire shipment, and analyst shall be careful to observe that no unusual con- ditions prevail in the selection of the bales. The accurate weight of all bales sampled to be ascertained by sworn weigher before samp- ling, or, wherever sworn weigher is not avail- able, by a competent person who must make sworn affidavit that weights are correct, and no other bales than those weighed to be sampled, and whenever bales are numbered, the number is to be given in addition to the weight. - "Method of Sampling "Depth of Boring. The sample shall be taken by boring into a bale to a depth of 3 ins. (7.62 cm.) with a special auger which cuts a disk about 4 ins. (10.16 cm.) in diameter. "Selection of Disks. The disks shall be removed, and ten of them taken as a sample, FlG - 4°- Cutter for these to be selected as follows: Sampling Pulp 1 disk 2nd sheet from the wrapper. 2 disks 1 in. (2.5 cm.) deep. 3 disks 2 in. (5.05 cm.) deep. 4 disks 3 in. (7.62 cm.) deep. " Location of Borings. The holes to be bored shall be so located that in five successive bales they will represent a portion extend- ing diagonally across the bale. Each bale to be bored but once. The first hole to be bored at the corner, the edges of the cut being 370 TESTING WOOD PULPS at a distance of one inch from the edge of the bale. The second cut shall then be made half way between the location of the first cut and the center of the bale, the third bale shall be cut at the center, the fourth bale half way between the center and the corner, and the fifth bale in the opposite corner in a position correspond- ing to the first. "All samples must be either weighed immediately after being drawn from the bales by accurate scales, or, when this is imprac- ticable, must be put into airtight vessels, made of metal or glass, with ground-glass or metal stoppers, and due care must be used in the transportation of such samples until they can be properly weighed at the laboratory of the chemists. The entire bulk of samples selected from the bales must be dried out for the test. The temperature in the drying oven shall not exceed 21 2° F. "Chemists must have proper and adequate equipment for weighing and sampling the bales, and for the weighing and drying of samples. "All sampling of pulp must be done by or supervised by the approved chemist personally, or by his bona fide assistants — each chemist to file with the committee a complete list of his bona fide assistants who will do the sampling, such list to have the approval of the committee. The chemist will be held respon- sible for the correct sampling by his approved assistants. The committee shall at any time have the privilege of investigating the sampling done by chemists or their assistants. " Every test certificate shall clearly state the name of the person who did the sampling. "The test certificates hereafter shall be uniform and in accord- ance with forms to be approved by the committee, a sample draft of which will be furnished by the committee to each chemist." This procedure can also be applied to rolls, making the first boring 1 in. from the end, the second half way between the end and the middle, the third at the middle, and so on. This method of testing is the quickest of any which has been proposed and it leaves the bales or rolls in good shape to be stored or shipped but it is not necessarily the most accurate. In LAP PULP 371 the opinion of the author it favors the seller of the pulp since the percentage of moisture decreases more rapidly on the outside than in the centre and the loss in the weight of the bale does not keep pace with the decrease in the percentage of moisture as shown by the sample taken. This is particularly true of bales put aside for a retest and stored in a comparatively dry place for several weeks ; such re tests therefore nearly always show a greater net weight of air dry pulp than do the original tests. The quarter-sheet method of sampling consists essentially of opening up the bales and taking a quarter of a sheet from different parts of the bale in such a way that the combined samples would cover the entire area of the bale. It is variously applied by dif- ferent workers; some take two quarter-sheets from each bale, others more, the extreme being ten quarter-sheets starting with the second sheet from the wrapper and spacing the others equally between the second sheet and the centre of the bale. On the whole it is believed that the quarter-sheet method is more accu- rate than the disc method already described but that it is unprac- tical because of the large amount of time and labor involved and because it leaves the bales in no condition for storage or further shipment. Lap Pulp. The methods of testing pulp for moisture have been exhaustively investigated by the Technical Section of the Canadian Pulp and Paper Association. 1 For lap pulp from ordi- nary wet machines or Rogers wet machines they tentatively recommend taking a strip 3 ins. wide clear across the sheet and to the full thickness of the sheet. Such a sample should be. taken for each 2000 lbs. of wet stock in the shipment. Tests by this method in comparison with the entire sheets dried out prove that the two methods give practically identical results and that the strip method is therefore accurate. For hydraulic pressed pulp they recommend the wedge system as proposed by Woodruff. 2 They have proved that this method gives more accurate results than the strip method in which 1 Slack: Pulp Paper Mag. Can., XVII, 1919, 265. 2 Woodruff: Paper, Oct. 3, 1917, p. 86. 372 TESTING WOOD PULPS 2-in. strips are cut from the centre of the lap to the outer edge, the cut portions forming a cross for each four samples. In the wedge system for hydraulic pulp a template is employed to mark the laps. This template has an angle of 9 degs. and at its apex is attached a disc divided into forty parts. This disc is placed at the centre of the lap and the wedge marked out by a pencil along the edges of the template. The lap is put one side and the next marked in a like manner, but moving the wedge of the template to the second position on the disc. Forty laps are thus marked, giving a total sample equivalent to one entire lap. The marked laps are then taken to a circular saw and the wedges cut out. These wedges may be split in halves if it is necessary to reduce the bulk of the sample. Strength or Beating Test. This method of testing was first described by the writer in 191 5 x and was the outcome of an attempt to demonstrate experimentally the variations in the strength of sheets made from different sulphite fibres after they had been given the same beating treatment. The method first proposed is as follows : The fibre to be tested is allowed to become air dry and two lots of 50 grams each are then weighed out. One of these is soaked up in 1000 c.c. of water and reduced to pulp by rubbing with the hands; it is then rinsed into a small pebble mill with just 1000 c.c. more water. The mill is now closed and allowed to revolve at 60 revolutions per minute for exactly an hour, at the end of which time the entire contents are emptied upon a §-in. mesh sieve and the pulp washed off the stones into a pan placed beneath the sieve. The pan is then filled to a definite mark — 27 liters total contents, and four sheets are made on a hand mould, taking two dips for each sheet and reversing the hand mould between dips so that the sheets may be of uniform thick- ness on the two edges. After pressing between felts in a copying press the sheets are air dried and are then tested for strength with an Ashcroft tester. Five bursting tests are made on each sheet along one of the diagonals and the average of the twenty 1 Paper, Nov. 10, 19 15. STRENGTH OR BEATING TEST 373 tests taken as representing the beaten strength of the sample. If it is desired to compare the strength before and after beating the second sample of 50 grams is broken up into a pulp as before and made into sheets from the same volume of water as in the case of the beaten pulp. This method has been slightly modified * by the Committee on Sulphite Pulp of the Technical Association of the Pulp and Paper Industry but the changes they propose are largely to shorten the time required and do not alter any of the essential features of the test as originally proposed. Many factors influence the results of this test and in order to get concordant results careful control is necessary. This does not lessen the value of the method but it means that intelligent super- vision is essential if the best information is to be derived. Some of the vital factors will be briefly mentioned. The size of the pebble mill as well as its volume and charge of pebbles must be carefully considered. Two mills of the same make often vary enough in size and capacity so that very different beating effects are obtained. The only satisfactory way seems to be to assume one to be right and adjust the charge of pebbles, fibre and water in the other so that the same strength test will be given by the beaten fibre. If the results in two plants are to be directly comparable it will of course be necessary to stand- ardize their pebble mills on the same lots of fibre. This is not a serious inconvenience and is not found to detract much from the value of the test since in most cases comparative results only are wanted. The proportions of fibre and water must be closely adhered to, as comparatively slight variations cause appreciable differ- ences in the bursting strength. This is also true of the time of beating, which should be regulated with care and not allowed to vary more than a minute or two. Factors having minor influence on the tests are the temperature of the water in the pebble mill and the relative humidity of the 1 Paper, Nov. 8, 1916. 374 TESTING WOOD PULPS atmosphere when the bursting tests are made. For ordinary rou- tine tests consideration of these factors may be neglected. A variable having great influence in many cases is that of the moisture in the pulp when taken for the test. Wet pulp will not give the same test as it will after being air dried. Experi- ments on one lot of pulp showed that the beaten strength was highest in the wet lap and decreased as the moisture diminished; the change from 70 to 5 per cent moisture reduced the bursting strength from 64.7 lbs. to 40 lbs. This change cannot be reversed by wetting dry pulp, which indicates the reason for air drying all samples before testing so that all may be on the same basis. For the most accurate work it would be very desirable to dry all samples at a constant humidity since this would insure more concordant results. One of the changes which it was proposed to make in the original method was to press the samples of wet pulp to a constant moisture and use this pulp as a basis for the test. This shortens the time required but it does not permit com- parison of the wet pulp with one received in the air dry condition. A point of very great importance is the manner of forming the sheets to be tested. If a hand mould is used the weight of the sheets varies considerably depending on whether the fibre is "free" or "greasy" after beating. It has been proposed to get around this difficulty by calculating the breaking test per unit weight of sheet but this is not an entirely satisfactory method. All such variations can be eliminated by using some form of sheet making machine in which a constant volume of stock is used and none is lost by overflowing the edges as happens with the deckle of the hand mould. Different methods of employing this test may be used; the original idea was to beat for a fixed time and note the differences in the strength developed, but Hatch l has adopted the modifica- tion of beating a number of samples for different lengths of time in order to see what maximum strength would be developed and the time necessary for its production. For other interesting data relating to the beating test reference 1 Paper, Oct. 3, 1917. COLOR COMPARISON OF BLEACHED PULPS 375 is made to the work of Mansfield and Stephenson l and of Sutermeister. 2 This method of testing gives satisfactory results with sulphite, soda, and sulphate pulps but does not appear suitable for ground wood since the latter is not affected by the beating treatment to nearly so great an extent as the others. Color Comparison of Bleached Pulps. The method outlined here has been taken, with slight modifications, from the report of the Committee on Sulphite Pulp of the Technical Association of the Pulp and Paper Industry. 3 The grading of the shade of a piece of bleached pulp usually leads to very uncertain results. Two pieces of pulp, which under certain conditions of light appear identical in shade, under different light conditions appear to be very different in shade. Very often bleached pulp which is satisfactory in shade under artificial light, appears quite yellow in daylight. Again, different pieces of the same pulp often appear to be of different shade, when graded together under identical light conditions. This latter phenomenon can be due only to one cause: differences in the surfaces of the two pieces of the same pulp. The problem to be solved therefore appeared to be twofold. 1. To devise some means of obtaining constant illumination conditions; and this illumination to be of such a quality that slight differences in shade could be easily distinguished. 2. To devise a means whereby (a) the surface of different pulps could be made the same, or (b) the effects of different surfaces of different pulps could be minimized. Many so called daylight lamps were tried, including both the bulb and arc types. Finally the arc type of color comparator made by the General Electric Company was adopted as giving 1 Pulp Paper Mag. Can., Oct. 1, 1916. 2 Paper, Dec. n, 1918. 3 Paper, Nov. 8, 1916, 19, No. 9. 37 6 TESTING WOOD PULPS the most constant artificial light of all those tried out and a light whereby one could most easily distinguish small differences of shade. The procedure adopted consists in attaching six disks of the pulp to be tested to the front of six wheels of varying degrees of whiteness. The wheels and pulp are then revolved at high speed (2,500 revolutions per minute) and the operator judges to which of the standardized color wheels the pulp is nearest in shade. The shade of pulp is then declared to be that of the wheel to which it is nearest in color. The test being made always under the same conditions of illumination both variable factors of light and pulp surface are rendered constant and thereby the remaining variable, the actual shade of pulp, can be easily determined. Description of Apparatus 1. Cutting Press. The disks of pulp are cut with a press and dies of the type F0 2 , made by the Ferracute Machine Company, of Bridgeton, N. J. The dies are so arranged that a disk 3! ins. in diameter is cut with a f in. hole in the center, at one operation. 2. Revolving Disk Machine. Fig. 41 gives details of the con- struction of this apparatus. The method of preparing the differ- ently shaded wheels is as follows: Different combinations of plaster of Paris, magnesia and chromate of potash are made into a paste with water, this paste poured into the brass wheel, allowed to set and then planed down to a smooth surface. It is recognized, of course, that one fixed standard of color shade is not applicable as a standard to all kinds of pulp. For example: bleached pulp when run from a drying machine has a radically different shade from the same pulp run through a wet machine and hydraulically pressed. Yet both pulps on being made into paper would give identical color shades to the paper. Hence, it is necessary to employ different color mixtures on the revolving disk machine for different types of pulp. Again the standard of pulp color shade demanded by some manufacturers of paper is entirely different from that of other manufacturers. Hence it is not possible to recommend a formula applicable to (377) u (378) DAYLIGHT LAMPS 379 all cases. The formula given here is applicable to bleached drier pulp and the color 95, as named in this report, when applied to this product is satisfactory to the great majority of paper manufacturers in this country. The different wheels are numbered as follows, reading from left to right: 100, 95, 90, 85, 80, 75. The formulas are as follows : No. of wheel Water Plaster of Paris Chromate of potash Magnesia, powdered IOO 120 107 O.O 20 95 120 107 O.0245 20 9° 120 214 O.1490 O 85 120 214 O.4660 O 80 120 214 O.6120 O 75 120 214 0.7770 O The plaster of Paris and magnesia are first thoroughly mixed together, then the water added and the paste thoroughly stirred up and poured into the wheel. When pouring the wheels containing the chromate, the latter is first dissolved in the water, the solution added to the plaster of Paris, the whole thoroughly stirred up and poured into the wheel. After pouring, the wheels are allowed to set forty-eight hours and then revolved by means of the motor, and turned down smooth with a sharp-edged cold chisel and finished with fine sandpaper. A sewing machine belt was tried out to drive the machine, but a satisfactory fastener could not be found which at the speed employed would not pull out after a short time. It was found that I in. rope with a spliced joint stands up very well. 3. Daylight Lamps. Two of these, manufactured by the General Electric Company, are used. They are known as their direct current multiple color matching outfits. The whole apparatus, of course, is placed in a dark room, where no light rays can penetrate to interfere with the light of the lamps. 380 TESTING WOOD PULPS This method of testing is a noteworthy attempt at putting the designation of color on a numerical basis in a practical way and it is in successful operation in several mills. The equipment is however rather expensive, and if several grades of pulp, such as soda poplar, sulphite spruce, and sulphate pine, are to be examined a large number of standard discs must be kept on hand and the preparation of the filling for such discs is by no means easy and is often quite unsatisfactory. It is the belief of the author that fully as reliable comparisons can be made by using hand mould sheets as explained under "Bleaching Qualities." It is of course desirable that such com- parisons should be made under a "daylight lamp" in a dark room but if this is not available a north light should be used and tests made at as nearly as possible the same time each day. Bleaching Qualities. The method used in determining the bleach required by a sample of fibre depends on the point of view of the observer and should in general be made to conform as closely as possible to actual operations in the plant in question. The method preferred by the author is as follows : Two 50-gram samples of the fibre are weighed out and at the same time a small sample is taken for a moisture test. From this moisture test it is possible to calculate the air dry fibre (10 per cent moisture) corresponding to the 50-gram samples and upon this air dry weight the figures for bleach are based. One of the samples is now broken up in a little water in a 6-in. by 8-in. battery jar until a uniform pulp is obtained and this is then diluted to about two liters with water. From a solution of bleaching powder, whose strength has previously been determined, a volume is measured out which corresponds to the percentage of bleach which it is estimated the fibre will require. This is added to the jar of pulp and the contents kept agitated and at about 35 to 40 C. until the bleach is just exhausted. The fibre is then thrown upon a screen of 70-mesh wire and washed with a heavy stream of water to remove the bleach residues and break up any knots or balls of fibre which may have formed. The pulp is next made into hand mould sheets which are dried on a small steam heated BLEACHING QUALITIES 381 cylinder and compared with sheets of pulp of the standard color. If the color is found to be much different from the standard a second test is made using more or less bleach as the first test indicates. Since bleached fibre slowly changes in color it is quite essential to obtain some permanent standard but so far efforts along this line have not been entirely successful. Plates of dull porcelain have been suggested but the difference in surface texture between the porcelain and the fibre sheets would probably render the comparison of their colors quite unsatisfactory. Attempts have also been made to use slabs of plaster of Paris tinted to the right shade and given the desired surface texture by pressing between felts. Great difficulty was experienced in getting just the right color in such slabs. The most practical method seems to be to cut out a large number of samples from the same lot of unbleached pulp and -determine with great care the amount of bleach required to give the standard white color. After this is determined one sample is bleached each month, the sheets made therefrom serv- ing as the standard for the month following. In plants where practically this same method of bleaching is carried out the above scheme of testing is found to give results in very close agreement with those of actual practice. In other plants where an excess of bleach is added and the residual bleach washed out after the color is brought up to the desired shade a method of testing employing these same principles should be used. In such tests a known amount of bleach is added and that remaining at the end of the test is determined by analysis; from these data the per cent used by the fibre may be calculated. This method of testing is quicker than the one first given but it is necessary to compare the colors of the bleached fibres in the wet state before the yellow products of the bleaching action have been removed. Bichter x has suggested a method of testing sulphite fibres which depends on the comparison of the colors developed on treat- ing samples of unbleached fibre with nitric acid. Weighed sam- 1 E. Richter: Proc. Eighth Int. Cong. Appd. Chem. 382 TESTING WOOD PULPS pies of the fibre to be tested, and also of one whose bleaching properties are known, are treated with 13 per cent nitric acid solution for about an hour; portions of the acid are then drawn off and compared in a colorimeter of any standard construction. Knowing the bleach required by one fibre it is possible to esti- mate that necessary to produce the same color in the other sample. This method is quick and may possibly be reasonably accurate with fibres cooked under nearly the same conditions but with samples from a number of different mills it proved to be only approximate, some of the results being as much as 10 to 15 per cent higher or lower than the figures obtained by actual bleach- ing trials. Moreover it does not appear to be applicable to soda and sulphate fibres, at least when a sulphite fibre is used as a standard for comparison. Loss in Weight on Bleaching. This test, taken in connection with the bleaching test, throws considerable light on the value of a pulp, since it shows what loss in weight will be suffered by the fibre because of the oxidizing and solvent action of the bleach, apart from the mechanical loss in the bleaching process, which should be practically the same for all fibres. The method is as follows: Weigh out two samples of about 2 grams each and determine the moisture in one by drying to constant weight at ioo° to 105 . The second sample is broken up to a pulp in a little water by rubbing between the thumb and fingers and the pulp transferred to a small flask. This procedure requires care and skill but after a little experience it can be done with no loss of fibre. Bleach solution is now added from a burette, the amount being regulated to give the percentage which the bleaching test proved would give the standard white color in the sample being tested. The flask is now kept at about 35 to 40 until the bleach is exhausted, when the fibre is transferred to a Gooch crucible, washed very thoroughly with hot water, dried and weighed. The difference in the bone dry weight of the unbleached and bleached fibre, which shows the loss due to the chemical action of the bleach, may conveniently be expressed as percentage Of the former. NATURE AND AMOUNT OF DIRT 383 Sedimentation Test. This test is one which has not yet been standardized but which has been applied to the study of pulps by many different observers most of whom used home-made ap- paratus of some sort. The principle of the test is that the longer the fibres have been beaten, or in the case of ground wood the finer it is ground, the more slowly it will part with its water. The general method of making the test is to place a known amount of fibre, reduced to a uniform pulp in a definite volume of water, in some sort of receptacle with a perforated bottom through which the water can drain and on which the fibre settles to form a mat. A valve below the false bottom prevents the escape of the water until the desired time and makes it possible to record the time of outflow. In Fishburn and Weber's appara- tus 1 the receptacle is a graduated glass cylinder and the test consists in" noting the time required for the water level to drop 9! ins. In the Riegler-Schopper 2 tester, Fig. 42, the water drain- ing from the stock falls into a chamber having two discharge out- lets of different dimensions and at different levels. The amount of water issuing from the large orifice, which is at the higher level, gives an indication of the degree of beating or the natural slowness or quickness of the stock. The smaller opening, which discharges under a practically constant head, acts as a sort of automatic cut-off and by taking care of the last slow drainage from the test sample it makes the differences in the results more marked. A test of this nature, using a long tube as a sedimentation chamber, has been used in testing sulphite pulps and it is claimed 3 that it shows very marked differences between long and short fibred stocks. Nature and Amount of Dirt. The dirt in a sample of pulp can be successfully investigated if a sheet of it is wet and examined by transmitted light. A convenient outfit consists of a box painted white inside and fitted with two or more electric light 1 Fishburn and Weber: Paper, Oct. 11, 1916, 13. 2 U. S. Pat. 1,193,613, Aug. 8, 1916. Paper, Aug. 30, 1916. 3 "Snowshoe": Pulp Paper Mag. Can., Sept. 5, 1918, XVI, 793. 384 TESTING WOOD PULPS I <» =T (% kj) ts m { NATURE AND AMOUNT OF DIRT 385 bulbs. The top of the box consists of a glass plate upon which the pulp rests. This box may be of any convenient size but the sample of pulp should completely cover its top so that no glare from the lights enters the observer's eyes. If a box is made to take a sheet 6\ X 8 ins. and the dirt is counted on both sides a simple multiplication by 2 gives the dirt count per 100 sq. ins., which has been found a satisfactory method of expressing the results. The sheets to be examined should be of nearly uniform thickness so that the count may be on practically the same quantity of pulp in every case. If desired the dry sheets can be weighed before testing and in this way the dirt count can be placed on a weight basis. If the material in question is not received in the form of sheets it may be made into sheets on a hand mould before examination. It will be found convenient to dry these completely and again wet them before examination; this will enable the wet sheets to be handled with much less trouble than if they are taken direct from the hand mould in which case they tear so readily that they are extremely difficult to turn over. In making this test it is well to dig out with a needle the specks of dirt which are not directly on the surface, since particles which appear to be dirt, will, in some cases, prove to be small rolls of fibre. It is also desirable to record the results under at least two headings, dirt and shives, and in some special cases it is even well to classify them in such groups as cinders, bark, iron rust, etc., as this will often enable the chief cause of the trouble to be located. The results obtained in this test depend almost entirely on the opinion of the observer as to what constitutes dirt and since the opinion of different observers often varies widely it is plain that the tests should all be made by one person in order to avoid the personal factor. Records of this nature are valuable in showing whether the product of any plant is being kept up to standard and they are also helpful in selecting pulps in the market when purchases are in question. CHAPTER XV PAPER TESTING In the following chapter are given the methods in common use in the testing of paper, together with some of the more unusual tests which are occasionally called for. It has been attempted to give these methods concisely but in sufficient detail to enable them to be carried out without difficulty by any technically-trained man who has had some experience in the paper industry. It should not be forgotten, however, that the intelligent interpretation of the results depends in many cases upon the judgment of the observer and that good judg- ment can only be replaced to a slight extent by even the most detailed directions. The essential features of a number of these tests have been taken from the report of the Committee on Paper Testing of the Technical Association, 1 to which acknowledgement is hereby made, and the student is referred to this report for more com- plete details of many of the methods. There are also included a considerable number of tests which are not mentioned in this report but which practical experience has shown to be of value in the study of paper. For a more detailed study of the factors influencing many of these tests reference should also be made to Herzberg's "Papierprufung," 2 where the subject is exhaust- ively treated. The methods of testing are grouped under three heads: microscopical examinations, physical testing, and chemi- cal analysis. Microscopical Examinations The microscope offers practically the only means for deter- mining the kinds of fibres and the relative proportions of each 1 Paper XXV, 693, 739, 777, 831 and 877, Dec. 10, 17, 24, 31, 1919, Jan. 7, 1920. 2 Herzberg: Papierprufung, Berlin, 1907. 386 ESTIMATION OF FIBRE CONTENT 387 in a sample of paper and it is therefore the universal method for making a fibre analysis. The method is rapid, and in the hands of an expert is fairly accurate, but it is probably not desirable to attempt to estimate percentages much closer than 5 per cent. The accuracy attainable depends upon two factors (1) the kind of mixture under examination, i.e., whether two or more fibres, and how beaten and (2) the experience and care of the observer. As a rule, the opinion of one thoroughly trained and careful analyst is more exact than the average judgment of several inexpert men who have spent collectively much more time on the examination. The microscope used should be of some standard make so that attachments may be added from time to time as seems desir- able. The particular kind of microscope and whether monocu- lar or binocular is largely a matter of personal preference. This also holds true of the magnification employed, some advocating 45 diameters and others 120 or even as high as 160 diameters. An instrument with two-thirds and one-sixth inch objectives and 1- and 2-inch eyepieces will be found to give as great a range of magnification as is necessary for all ordinary paper- mill work. A substage condenser and iris diaphragm should by all means be included and it is very desirable to add a stage micrometer and an eyepiece micrometer. Estimation of Fibre Content. In order that the sample taken shall be representative several small pieces should be clipped from various parts of the sheet, or if several sheets are available small pieces should be taken from each. These should be placed in a dish, small flask, or any suitable container, covered with 0.5 per cent caustic soda solution, and heated to the boiling point in order to dissolve sizing or binding materials. The samples are then washed several times in water, rolled into a ball and kneaded between the thumb and finger and then re- duced to a pulp by shaking vigorously in a test tube about half full of water. A small sample is then removed from the test tube by means of a needle, placed on a microscope slide and the water removed by touching the drop with a piece of 388 PAPER TESTING folded filter paper of ordinary quality. The fibres are covered with a drop or two of Herzberg's stain, carefully separated by the aid of microscope needles, so that they will not lie too much in a bunch, and they are then covered with a cover glass. The slide is then ready for examination by means of the microscope. In preparing the sample for examination a number of pre- cautions must be observed if good results are to be obtained, and a discussion of some of these important points will be help- ful, at least to those who are just beginning the work. There are several methods of removing a representative sample from the test tube. Spence and Krauss l prepare a rather dilute pulp in a test tube about f X 8 ins. and for re- moving the sample use a glass tube 10 ins. long by -jVin. in diameter. This tube is fitted with a small rubber bulb at one end while the walls at the other are rounded to present a smooth surface. The test tube is well shaken, the dropper quickly inserted 2 ins. below the surface, two bubbles of air expelled and about half an inch of mixture drawn into the tube. This entire portion is transferred to slides, making four drops in all, and the water removed by evaporation in an air bath. The slides are then ready to stain and examine. Another and more usual method is to prepare the mixture in a smaller test tube, say f X 6 ins., and after shaking vigorously remove the sample by inserting a microscope needle and taking out a small bunch of the fibres. This method is better for long-fibred stock such as bond, ledger and writing papers, while the first method is safer for ground-wood papers and others where very short fibres are present. A modified form of the second method, which makes it appli- cable to all grades of paper, is to prepare the mixture of such density that on shaking the test tube and then placing it up- right, small clots of fibre will remain adhering to the walls of the tube above the liquid. The sample is then obtained by removing all the fibre included in one of these small clots. Removal of water from the sample before staining may be 1 Spence and Krauss: Paper, 20, 1917, p. 12, May 23. ESTIMATION OF FIBRE CONTENT 389 done with filter paper or by drying as already outlined, or a blotting paper of firm texture may be pressed down directly on the drop on the slide. This removes the water and leaves the fibre adhering to the slide ready for staining. If a firm blotter, free from lint, is used and any loose fibres are removed by blowing upon its surface just before applying it, no contami- nation of the sample need be feared. Another method is to place the sample removed by the needle from the test tube directly upon a piece of suitable filter or absorbent paper; this takes up the water and the fibre can then be transferred to the slide. The use of very thin cover glasses for covering the stained specimen is only to be recommended when very high magnifi- cation is to be used. In nearly all fibre analysis it is quite sufficient to use a second thin microscope slide as a cover. This permits several fields to be prepared on one slide and eliminates much breakage. The slide as a cover glass also possesses an- other advantage over the lighter and thinner form: if the sample is well loosened up with the needle so that no lumps remain and the slide is then dropped onto it from a distance of half an inch or so a very even distribution of the fibres can be obtained. The Herzberg stain, which is very generally used, is made up according to the following formula: Solution A. 20 grams zinc chloride dissolved in 10 c.c. of distilled water. Solution B. 2.1 grams potassium iodide and 0.1 gram iodine dissolved in 5 c.c. of distilled water. Prepare the two solutions separately, mix and allow the mix- ture to stand several hours or until all sediment has settled out. The clear liquid is then decanted and is ready for use in staining fibres; it should be kept in a brown glass bottle or else in the dark. This stain gives different colors with different kinds of fibres; ground wood, jute, flax tow, uncooked manila hemp and practically all highly lignified tissues are colored yellow or lemon yellow. Thoroughly cooked and bleached soda and sulphite pulps as well as bleached straw and esparto fibres are colored 390 PAPER TESTING blue or navy blue. Cotton and linen rags, thoroughly cooked and bleached manila hemp and some of the Japanese fibres are colored wine red. It is very essential that this stain be so made up as to give satisfactory colors on the different fibres. Its quality should be proved by staining a mixture of fibres known to contain about equal proportions of rag, bleached sulphite and bleached soda fibres. If the stain is satisfactory, the soda pulp should stain a dark blue color, while the sulphite, because of its thinner walls, will be a light blue, and the rag fibres will be red or wine- red. If the blue is not clear but tends toward the violet, too much iodine is present and more water or zinc chloride should be added. A stain which gives the best results with ground wood will not always be entirely satisfactory with mixtures of well-bleached fibres and if widely varying papers are to be examined, it is well to keep on hand a number of stains so ad- justed that one suitable for any grade of fibre is available. One should be prepared to give a bright lemon yellow on ground wood; a second should give the colors already mentioned on sulphite, soda and rag fibres, etc^ A stain which is considered by some to be better than the Herzberg stain is made up as follows: Solution A. 1.3 grams iodine and 1.8 grams potassium iodide in 100 c.c. of water. Solution B. A clear, practically saturated solution of calcium chloride. In using this stain apply a drop or two of solution A to the moist fibres on the microscope slide. After a minute or so re- move the stain by means of a blotter and immediately put on a drop or two of solution B. Pull the fibres apart and distrib- ute them by means of needles as before and drop on a cover glass or thin microscope slide. Any excess of solution B should be removed by absorbing it with moist blotting paper. This stain is also selective in its action, the colors produced being as follows : ESTIMATION OF FIBRE CONTENT 391 Red or brownish red: Cotton, linen, hemp, ramie. Dark blue: Bleached soda pulps from deciduous woods. Bluish or reddish violet: Bleached sulphite fibres and the thoroughly cooked part of the unbleached sulphite. Greenish: Jute, manila and the more lignified fibres in un- bleached sulphite. Yellow : Ground wood. As with the Herzberg stain this one should be adjusted by trial on known mixtures of fibre until it shows satisfactory differ- ences in color. The two solutions should be protected against evaporation and dust but light does not change their staining properties to any extent. In estimating the fibre content of a paper no account is taken of the size, clay, alum, etc., which may be present; the paper is therefore considered as being all fibre and the sum of the per- centages of the various kinds is made to equal 100 per cent. The estimation should be based on the examination of at least two samples of fibre removed from the test tube and in some cases where especial accuracy is desired it is well to examine four separate slides. A set of standard samples containing known percentages of the different ingredients is very useful in making comparisons with unknown samples and should be used occa- sionally to refresh the judgment of the analyst. There are two methods of determining the percentage of the various fibres, one is the "count method," while the other is known as the "estimation method." In the first method the fibres of each kind are counted and from the figures obtained the percentage of each is calculated. The estimation method depends upon the comparison of the unknown sample with standard mixtures of known composition. The accuracy of this method depends upon the experience of the analyst and upon continual reference to the known standards. It is prob- ably fully as accurate as the count method, is considerably quicker, and is much easier to teach to a beginner. For these reasons it is usually preferred to the count method. In using 392 PAPER TESTING either method it is quite obvious that the analyst must be thor- oughly familiar with the size, shape, and markings of the various fibres and with the appearance of any ducts, cells, or foreign matter which habitually accompany them. Without such knowl- edge it is useless to attempt to make a fibre analysis. A third method has recently been proposed by Spence and Krauss 1 to enable an estimation to be made of the different kinds of fibres in a mixture containing such fibres as hemlock, beech, poplar, birch, maple, etc. Four samples are made up by the method of Spence and Krauss already given. Each slide is examined under the microscope and the lengths of the various fibres are measured in terms of the diameter of the observed field. The magnification recommended is 160 diam- eters and an adjustable stage is essential in order to cover sys- tematically the entire sample on each slide. After all four samples have been examined the figures are added together to get the total length of each kind of fibre present. This total length multiplied by a weight factor, which varies with different kinds of wood, gives a set of directly comparable results which may be converted into the per cent of each kind of fibre present. The weight factors given are as follows: rag pulp i; hemlock pulp 0.870; poplar pulp 0.454; birch pulp 0.652; beech pulp 0.525; maple pulp 0.365. Where a mixture of fibres from deciduous woods is under examination the number of counts of each kind to which the weight factor is to be applied may be determined from an examination of the number of ducts present. It was found that the proportion of ducts and fibres in different woods was as follows : 1 Spence and Krauss: Paper 20, 19 17, May 23, p. n. UNBLEACHED SULPHITE DETERMINATION 393 This method is too slow to be used where many routine analyses are to be made daily but its accuracy recommends it particularly for settling cases of dispute between different authorities. Unbleached Sulphite Determination. The method given is that worked out by Bright. 1 "The principle of the method is first to stain the fibres with Cross and Bevan's ferric ferricyanide solution, which colors the unbleached sulphite green, on account of the lignin contained in it, and leaves the bleached sulphite colorless. This alone gives a good distinction, but by subsequently staining with a red substantive dyestuff , the green of the unbleached is changed to a very pure blue, the bleached being colored red, thus giving a most striking contrast. "The problem is to adjust the treatment with the two solu- tions to bring out the sharpest contrast. If the treatment with red is too severe, some of the unbleached fibres are likely to be colored purplish, or in extreme cases take on a dull, dirty red color. On the other hand, if the treatment with ferric ferricy- anide is continued for too long a time or at too high a tempera- ture, the reagent has a tendency to decompose and form a deposit on the slide as well as on the bleached sulphite, so that the latter turns a dull purplish color when subsequently stained with red. "The results depend on three factors — namely: (1) the con- centration of the solution, (2) the temperature at which each is applied, and (3) the length of time each is allowed to act. " The solutions are prepared according to the following pro- cedure: Ferric Ferricyanide Sol. A. — N/ioFeCl3 — 2.7 grams FeCl3 6H2O per 100 c.c. Sol. B. — N/10 K 3 Fe (CN)6 — 3.29 grams per 100 c.c. " After diluting to the mark with distilled water, filter through dry filters into clean, glass-stoppered bottles — protect from dust. 1 Bright: Paper 20, 1917, Aug. 29, p. 11. 394 PAPER TESTING " Equal volumes are mixed fresh whenever the reagent is used. Substantive Red Benzopurpurin 4B extra (Bayer Co.) gm. 0.4 Oxamine brilliant red BX (Badische Co.) gm. o. 1 Distilled Water c.c. 100 " Have the water hot and stir in the dyes slowly. " The stain is placed in a tall, narrow, cylindrical beaker, which is set into a water bath. The slides are suspended in the beaker by a clamp which holds them at their upper ends, the clamps resting across the top of the beaker. A thermometer is sus- pended in the beaker of stain beside the slides. The beaker should be as small as possible so as not to use up too much stain at one time. " Staining with ferric ferricyanide is done as follows: " Mix equal volumes of two fresh solutions and heat to 35° C. regulating the waterbath so that it will remain within 1 deg. of the temperature named, for 15 minutes. The dry slide is then dipped in water to moisten it uniformly, so that air bubbles will not be formed when it is immersed in the stain. If air bubbles are formed the fibres under the bubbles will not be stained. If dipping in water still leaves bubbles, they can be removed by blowing across the slide from the edge. The slide is then suspended in the stain and left there for 15 minutes at 35 C. It is then removed and washed by dipping in and out of a beaker of distilled water six times and repeating the proc- ess in a fresh beaker of water. The slide can then be placed wet into the red solution, but it is perhaps best to dry it out so that the fibres will be stuck on tightly again in case they have been loosened to any extent by the treatment. " Staining with the substantive red solution is done as fol- lows: " A fresh solution is heated to 45 C, and the slide, after mois- tening and excluding bubbles as before, is suspended in the solu- tion for five minutes at 45 C. and immediately washed in two beakers of distilled water. • MACHINE DIRECTION 395 "The slide is then dried and a cover glass placed on with a drop of balsam. " To get the clearest, brightest results, distilled water must be used throughout, and the staining solution must be fresh. The two solutions for ferric ferricyanide will keep well if placed in separate bottles. Equal volumes are mixed together immedi- ately before using. The red solution should be made freshly each time for the best results, as it gets thick and stringy on standing, especially when it is being heated up continually. " This method of staining will in general give a distinction be- tween pure cellulose fibres and those which contain lignin. Rags, bleached sulphite, soda pulp or any thoroughly bleached materials are stained red, while unbleached sulphite, ground wood, jute, or any lignified materials are stained blue." Physical Tests Machine Direction. Several methods are available for de- termining the machine direction in a sample of paper. It may sometimes be ascertained by mere inspection of the sheet, as the formation noted on looking through it is often conclusive to the trained observer. The usual machine wire imparts to the sheet of paper a "wire mark" consisting of a series of diamond-shaped marks, the long diagonal of which points in the machine direction. If the wire mark is sufficiently prominent so that its direction can be determined this will establish the machine direction. If the paper is well sized and a circular piece is cut out and moistened on one side by floating on water, it will tend to roll up into a cylinder whose axis is in the machine direction of the sheet. If the paper is unsized it will become entirely soaked through on floating on water and will not curl up. This may be avoided by sizing the paper with an alcoholic solution of rosin, or with a solution of gelatine in water, drying and then making the test. Another method of determining the machine direction is to cut two narrow strips of the paper, one from either direction, 396 PAPER TESTING place these one over the other and hold them upright in the fingers. They will droop over of their own weight and if they cling close together the under strip is in the machine direction while if the under strip falls away from the upper the latter is in the machine direction. The form of the break made by the Mullen tester shows the machine direction, as the longest, or chief, line of rupture is always across the sheet. This is shown in Fig. 43, which illus- Fig. 43. Lines of Rupture in Mullen Test trates four typical ruptures and in which the arrows indicate the machine direction. Wire or Felt Side. In many cases this may be determined very easily by a simple inspection but in some papers the wire marks do not stand out at all plainly. Sometimes they may be made more prominent by plunging the sample for a moment into water and draining or blotting off the excess. The moist- ure causes the fibres to expand, thus undoing the work of the calenders and restoring the texture of the sheet as it left the machine wire. Inspection of a sheet thus dampened will often WEIGHT OF SAMPLE 397 show that the wire marks stand out plainly, where before they were indistinguishable. This method very often proves satis- factory even for coated papers. Weight of Sample. The weight of a sample of paper is usu- ally expressed as so many pounds per ream of a given size, but the number of sheets per ream and the standard size of the sheet are more or less variable in the different branches of the in- 1 V «W> ■-■-•-'" P: Fig. 44. Beam Type of Paper Scale dustry. Where a full sheet of paper is available the weight per ream may be conveniently found on scales of the type shown in Fig. 44. The folded sheet is placed on the hooks and the weight per ream is shown by the position of the sliding poise on the beam. This instrument, which is largely used for mill work, is usually graduated for reams of both 480 and 500 sheets. Two representative scales of the quadrant type are shown in Fig. 45 and Fig. 46. With these the weight per ream in pounds ^•~^^~-~ ' : GBBfl3ffl ' ■ 1 \ V-0m ^^^WF X /-/ ■ ^^: M^fe> ^^sjH^K^^^^^SjT - C ^-< •--. *©* • ■ '• JO W^- ,;..i Fig. 45. " Quick-stop " Paper Scale Courtesy of B. F. Perkins &° Son, Inc. (398) Fig. 46. Basis-weight Scale Courtesy of Thwing Instrument Company THICKNESS 399 per 500 sheets is read off directly on the scale and no movement of weights is necessary. In the case of very small samples a piece of known area should be weighed on a chemical balance and the ream weight calculated by the following formula: (Wt. in grams) X (1-103) X (area of trade size desired) Area of sample in square inches = Weight on trade size desired. Thickness. The thickness of a paper may best be deter- mined by means of a spring micrometer having a hand that Fig. 47. Thickness Gauge Courtesy of B. F. Perkins & Son, Inc. travels around a circular dial which is graduated in thousandths of an inch. One of the many such instruments is shown in 400 PAPER TESTING Fig. 47. This type of tester should not be read closer than half a thousandth as this is about their limit of accuracy. Thickness may also be measured by means of an ordinary micrometer caliper but the amount of pressure is not easily con- trolled unless it is supplied with a ratchet device which prevents an excessive pressure being applied. This is not so accurate as the spring type but is useful for approximate work around the mill. It is advisable to have all thickness gauges tested before use. This may be done by securing a set of standard sheet-metal leaf gauges which range from 0.001 to 0.015 in. These should be used occasionally to make sure that the instrument for measuring thickness remains accurate. It has been suggested by the Committee on Paper Testing that the relative compactness of papers be calculated as follows for purposes of comparison. Thickness in thousandths of an inch X 10,000 Weight 25 X 40,500 = Relative compactness. Bulk. The " bulk " of a paper is the thickness of a certain number of pages and is a factor which must be taken into con- sideration in planning a book which must not exceed a definite thickness. It is usually measured by making up a "dummy" or cutting out short strips, piling them up to the required num- ber and measuring its thickness with the ordinary graduated sliding clamp. In using this the pressure used must be speci- fied as heavy, medium, or light and this introduces an element of uncertainty into the results. The so-called "pressure bulker," Fig. 48, made by B. F. Perkins and Son has been made to eliminate these troubles. In this instrument the "dummy" is put under a definite pres- sure which is read in pounds per square inch on the dial and the thickness in inches is read on a scale at the side. Opacity. The opacity or translucency of a paper may be measured by the "Contrast ratio" method using a Martens OPACITY 401 photometer in a specially constructed box. 1 This method in- volves a determination of the difference in photometric bright- ness, or contrast, between a black and white spot when covered with the material to be tested. Fig. 48. Pressure Bulker Provided this apparatus is not available, good comparative tests can be made by cutting a small, sharp-edged opening in a piece of cardboard and placing this over a source of intense light. Pieces of the paper are then laid over the opening one 1 Bureau of Standards Circular No. 63. 402 PAPER TESTING at a time^ and the number of sheets required to completely obliterate the light is noted. By determining the thickness per sheet the absolute thickness of the paper required to obliterate the light can be calculated. Gloss or Glaze. An instrument to measure the gloss or glaze of paper has been devised by L. R. Ingersoll. 1 It was found that light specularly reflected from paper at an angle of 57.5 degs. was almost completely plane-polarized and working on this basis the instrument was designed to measure the gloss Fig. 49. Ingersoll Glarimeter by determining the fraction of the light reflected from the paper at an angle of 57.5 degs. which is polarized. In using the in- strument, light from a tungsten lamp is reflected from the paper through a diaphragm, part of the opening of which is covered with a small Nicol prism and the rest with a piece of smoked glass which absorbs about one-half the light. The beam then passes through the eyepiece which is another Nicol prism mounted to rotate in a divided circle which reads with the aid of the vernier to five minutes of arc. On looking into the eyepiece a divided field is seen and a measurement is made 1 L. R. Ingersoll: Electrical World, 63, 1914, 645. TENSILE STRENGTH 403 by turning the divided circle until the dividing line between the two parts of the field disappears. From the average of a num- ber of such measurements the percentage gloss may be found by consulting a table which accompanies the instrument. Fig. 49 gives a perspective view of the Ingersoll glarimeter. In using this instrument it must be remembered that the results are comparable only when papers of the same color are considered, since a black paper having the same glossy surface as a white paper would reflect an entirely different per cent of light, due to the absorption of the black color. Since the great majority of papers examined are white, this limitation is not a serious objection. Tensile Strength. The tensile strength of a paper is deter- mined by the load, in pounds, required to break a strip of it. The tensile strength machine best known in the paper industry is the Schopper tensile machine, illustrated in Fig. 50. In this device a strip of paper 15 mm. (approximately H in.) wide by 180 mm. long (approximately 7 T \ ins.), is clamped at each end and the clamps are moved apart until the strip is broken. A suitable device indicates the pull in kilograms (ap- proximately 2.2 lbs.) required to break the strip. It is recom- mended that the load in kilograms per 15 mm. width strip, be converted into pounds per inch of width by the following formula : (3.73) X (Tensile strength in kg. per 15 mm. width) = Tensile strength in lbs. per 1 in. width. A tensile strength factor may be determined by the following formula : ( Tensile strength in lbs, per 1 in. width X Weight 25 X 40,500 / = Tensile strength factor. The usual factor for tensile strength is known as the breaking length. This is the length of a strip which, if suspended at one end, would break of its own weight. The following formula may be used to determine the breaking length of a sample: 4°4 PAPER TESTING (Tensile strength per i in. width) X (13,889) (Weight of a sheet 25 X 40,500) = Breaking length in yards. Fig. 50. Schoppee. Tensile Machine Stretch. The amount of elongation at the instant of rupture of a strip of paper under tension is measured on the S chopper tensile strength machine. The result is figured as a per cent of the total, original length. Bursting Strength. There are two general types of apparatus used to determine bursting strength. One is of the hydraulic TENSILE STRENGTH 40S Fig. 51. Mullen Tester Fig. 52 . District or Columbia Paper Tester 406 PAPER TESTING type, in which the paper is clamped against a rubber diaphragm, through which the pressure is applied to a circular area of the paper measuring one square inch. The Mullen tester, Fig. 51, and the District of Columbia paper tester, Fig. 52, are of the hydraulic type. The second type of bursting strength apparatus is of the spring operated metal plunger design, in which the paper is clamped between annular rings, through which a spring operated plunger is forced. The Ashcroft tester, Fig. 53, is the only one of the above type now on the market. Fig. 53. Ashcroft Tester The bursting strength to be of greatest use must be expressed in terms of the weight of the sample. This ratio of strength to weight may then be directly compared with the strength ratio of any other paper, if the same standard size sheet is used in each case. The strength ratio is expressed as a percentage. Bursting strength X 100 Strength ratio = TTT . . Weight m pounds (on a size 25 X 40,500) Folding Endurance. The folding endurance of a paper is measured by a machine in which a strip of paper is caused to FOLDING ENDURANCE 407 fold back and forth upon itself until it is worn through. The Schopper Folding Machine, which is the only device so far Fig. 54. Schopper Folding Machine made to carry out this test, is shown diagrammatically in front view, and from above in Fig. 54. In making a test the slot cut in the thin metallic plate fastened 408 PAPER TESTING to the end of the shaft (13), is brought exactly in line with the jaws (7) and a strip of paper cut accurately, 15 mm. in width and 100 mm. in length, is firmly clamped between the jaws and through the slot. The paper is then put under tension by pull- ing out the shafts (4) to which the jaws are fastened by an intermediate spring (3). The shaft (13) is given a reciprocat- ing motion by the revolution of the wheel (19), and the paper is folded and refolded over the edges of the slot in the thin metallic plate and around the rollers (12), the revolving of which eliminates friction between the paper and the rollers. Twice in each revolution of the drive wheel (19), the jaws are pulled out to what may be termed the maximum position and the paper is then under a tension of 1000 grams. When the slot in the metallic plate is in line with the jaws the test strip is under the minimum tension of approximately 730 grams. Most papers stretch slightly under these tensions, which are thereby reduced more or less, with the result that a somewhat higher folding number is obtained. The paper is weakened in the crease made by the repeated folding and finally broken by the tension. The number of double folds is automatically recorded on the dial (18). The folding strength of paper is dependent not only upon the strength and durability of the paper, but also is very largely influenced by the relative humidity. To perform this test in the most accurate manner it is therefore necessary to keep the relative humidity constant for all tests. This can only be done by the use of a room where the humidity is under control. Where such a room is not available then note must be made of the per cent relative humidity of the air at the time of the test. No tests should be attempted when the humidity is either very high or very low. A relative humidity between 65 and 70 per cent is most easily attained throughout the year and is the standard humidity recommended. The folding factor is determined by the following formula: Folding endurance t- 1 t r ^ ,„ T . , & r = Folding factor. (Weight 25 X 40,500) TEARING TEST 409 The folding factor will vary between about 0.1 and 200. The standardization of the folding tester and the accuracy of the results obtained have been very carefully investigated by Veitch, Sammet and Reed 1 who consider it valuable for indi- cating the probable durability of paper and its suitability for specified purposes. Tearing Test. This method of testing paper is not yet stand- ardized though much interest has recently been shown along this line and several methods have been proposed for determining the tearing strength of paper. Case 2 cuts a strip z\ ins. by 1 in. which is slit lengthwise for a distance of 2J ins., leaving a distance of \ in. to be torn. The strip on one side of the slit is fastened in a clamp and to the other strip is suspended a small bucket into which water is allowed to flow until the two strips are torn apart. The weight of the bucket and water in grams is taken as the tearing resist- ance number. The S chopper tensile machine was used by Wells 3 who removed the weight from the arm and used only the arm and the milled screw with the pawls held out of action. To eliminate the local variation in a single sheet, several samples from the same sheet were torn at once, half being cut across and half with the grain. A speed of tearing of i| ins. per minute was used and readings taken every five seconds while making a tear of 3 ins. This method was found to give good check results provided the speed of tearing was constant. A tearing tester recently devised by Thwing is shown in Fig. 55. A small sample of paper is cut and punched by a special die and attached to two pins one of which is attached to a movable weight on an arm carrying a recording pen while the other is attached to a motor driven, sliding record card holder. This holder is caused to move away from the pin on the weight arm, thus tearing the paper along the line of perforations. The record 1 Veitch, Sammet and Reed: Paper, 20, 1917, May 30, p. 13. 2 Case: J. Ind. Eng. Chem., 11, 1919, 49. s Wells: Paper, 23, 750, Feb. 12, 19 19. 4io PAPER TESTING shows graphically the force in grams required to tear the paper between each two perforations, thus giving five peaks, the average of which is taken as the tearing strength. Fig. 55. Thwing Tearing Tester Courtesy of Thwing Instrument Company Absorbency. The absorbent power of a paper is generally measured by suspending strips vertically with their lower ends dipping into water and noting the height to which the water rises in ten minutes. The average for several strips cut in both directions should be taken as the figure for absorption. This method is criticized by Reed l because it employs water instead of ink but still more because it is unaffected by the bulk or weight of the paper. He proposes allowing 1 c.c. of a standard ink to fall from a pipette upon the surface of a 4-in. square of the blotting paper which is placed over a tumbler or beaker of 1 Reed: J. Ind. Eng. Chem., 10, 1918, 44. PERMEABILITY TO AIR 411 such size that the edge of the spot will not touch the glass. The time for the complete absorption of the ink is recorded. In this test attention must be paid to the temperature of the ink, the delivery time of the pipette, the distance of the tip above the surface of the paper, and the amount of liquid used. All of these factors must be standardized if comparable results are to be obtained. Volumetric Composition. The determination of the volume composition of a paper is at best only an approximation but it is at times desirable to carry it out. The weight of a cubic centi- meter of the paper is first ascertained by calculation from the thickness of the sample and the weight of a measured area. The percentage by weight of the various materials present, fibres, clay, size, etc., is then determined in the usual way and from this the weight of each in a cubic centimeter of the paper is calculated. The weight of each substance in grams divided by its specific gravity gives the volume occupied by it, and the sum of all of these volumes subtracted from 1.0 gives the volume of air per cubic centimeter of paper. This method is fairly accurate when only fibres, clay and rosin are present but when other sub- stances are added, as in coated papers, the problem becomes more complex and the results less reliable. If the volume of air per cubic centimeter of paper is the only information needed it may be obtained by determining the actual specific gravity by weighing in air and then in oil of known dens- ity exactly as in making specific gravity determinations in water. It will be found necessary to expose the paper, submerged in oil, to reduced pressure for some time in order to be sure that all air is removed and replaced by oil. Permeability to Air. No simple and accurate apparatus for measuring this property of paper is available. Herzberg * has made measurements by passing air through a definite area of the paper under standardized conditions and measuring the amount passed by a gas meter provided with a special overflow to equalize the level on the two sides of the meter. This method 1 Herzberg: Mitt. k. Materialpruf, 1915, 33, 142-144. 412 PAPER TESTING doubtless gives reliable comparative tests but the apparatus is not of a kind which is generally available. In the case of waxed or waterproofed papers it has been pro- posed by Seiter x to use the paper as a diaphragm in a dialyzing apparatus, placing ferric chloride above the paper and potassium ferrocyanide below. The time required for the development of Prussian blue is a measure of the porosity of the paper. Grease-Proof Properties. The surest method of determining whether a paper is grease-proof is to place the sample on a piece of white paper, pour on it a small quantity of oil of turpentine and rub it around with a bit of absorbent cotton. If the white paper becomes stained with the oil the sample is not grease- proof. A rough test for parchment papers is to heat the sample mo- mentarily over a flame and note the formation of blisters. Pin holes will prevent the formation of blisters while on the other hand the more impervious the surface the more blisters will be formed. This test is not absolutely reliable, as some papers which blister are not grease-proof, while some which do not blister are satisfactory in this respect; it is, however, reliable in the ma- jority of cases. Degree of Sizing. For book and magazine papers which are ordinarily not sized very hard a sufiiciently accurate test may be made by floating a small piece of the paper on a bath of ink and noting the time required for the ink to penetrate to the upper surface. In this test a standard ink should be used and it should be thrown away when used once and not returned to the bottle. The temperature of the ink has a very great influence on the time of penetration and it should be maintained within half a degree throughout the tests. The personal factor of course plays a very important part in this test and to make results strictly comparable they should all be obtained by the same per- son. Attempts have been made to eliminate the personal equation by measuring the conductivity of the paper as it is gradually 1 Seiter: Chemist-Analyst No. 21, April, 19 17. DEGREE OF SIZING 413 penetrated by an electrolyte. Okell 1 applies the solution of the electrolyte to both sides of the paper in a specially con- structed cell which permits of slight but constant pressure on both sides of the sheet. A slightly modified form of this cell has also been employed by Clark and Durgin. 2 The results obtained are interesting and warrant its use for scientific investigations but the apparatus does not appear well adapted to the rapid work required where a large number of routine tests must be made in the shortest possible time. For high grade papers, such as surface sized writing papers, the notation test does not indicate sufficiently well the quality of the sizing. For such papers the method proposed by Sammet 3 should be used. This method involves the drawing of a strip of paper over the surface of an iron tannate ink and allowing it to drain and dry naturally. Upon examination of the surface with a low power microscope, a well sized paper will show no indication of the fibre having absorbed the ink. Any variation in the depth of color on the surface will indicate a lack of uniform sizing. This test may be still further developed by erasing the surface with an ink eraser (a spun glass eraser is most suitable) and again dipping the sheet as before. A paper well sized throughout the sheet will show little or no additional absorption of ink at the erased spot. This test is only comparative but may be valuable to a mill in checking the daily progress. The ink used for the above test is made as follows: Tannic acid (dry) 23.4 grams Gallic acid (crystals) 7 7 grams Ferrous sulphate 30 grams Dilute hydrochloric acid (U. S. P.) 2 <. o c.c. Pheno1 1 . o gram Blue Dye (Bavarian Blue S. & J. No. 478) 2.2 grams Water to make up to 1000 c.c; allow to settle, and decant from any sediment. Note. — Any water-soluble aniline blue, as methylene blue, may be used in place of Bavarian blue. 1 Okell: Paper, April n, 1917, 20. 2 Clark and Durgin: Paper, 22, 1918, 223. 3 Sammet: Bureau of Chem. Circular No. 107, also Paper, 10,1913, No. 9, p. 15. 4 I4 PAPER TESTING Chemical Tests Moisture in Paper. In most cases the moisture may be de- termined easily and accurately by weighing a sample into a weighing bottle with a ground-glass stopper, drying at ioo° to 105 C, closing the bottle, cooling in a desiccator and again weighing. The loss in weight represents moisture. In no case where the work demands any degree of accuracy should the weighing of the dried paper be done in the open air because the dry paper absorbs moisture very rapidly from the surrounding atmosphere. This procedure is inaccurate in some cases where substances are present which lose water of crystallization, or of constitution, at the temperature named. This is particularly true of papers where calcium sulphate is used as a filler or where satin white is used in the coating. In these cases drying at ioo° to 105 C. expels three- fourths of the water of crystallization of the calcium sulphate and this loss gives an entirely fictitious value to the hygroscopic moisture. In such cases the only expedient seems to be to dry the samples to constant weight in a desiccator over sulphuric acid. Ash Determination. A sample of one gram of the paper is ignited in a weighed dish, over a burner, or in a muffle, until all carbon is burned off. The dish is then cooled in a desiccator and weighed; the increase in weight represents ash and by mov- ing the decimal point two places to the right it may be expressed at once as percentage of the paper taken. If the ash is to be cooled and weighed in the dish the ignition may be made in porcelain, platinum or any material not changed in weight by heating to a bright red heat, but platinum will be found to be the cheapest in the long run because it cools quickly, is practically constant in weight and above all because the ignition is more rapid than in porcelain or silica. Because of the greater speed of combustion a shallow dish is much preferable to a crucible and a still further gain may be made by covering the dish with a curved piece of platinum foil which reflects the ASH DETERMINATION 415 heat downward but allows the entrance of plenty of air. With four such dishes it is easily possible to weigh out twelve samples, ignite, cool, weigh and record the results within an hour. The sample of paper need not be weighed closer than 0.005 gram except in cases where extreme accuracy is desired and then the sample should be weighed in the bone dry condition in a weighing bottle. Meker burners have been found very satisfac- tory for ash determinations, being considerably more rapid than the ordinary Bunsen. During burning care must be taken that no ash is blown out of the dish by strong air currents as this is likely to take place especially with the light, fluffy ashes from unloaded or lightly loaded papers. This same danger must be guarded against when removing the covers from the desiccators in which the dishes are cooled. The use of desiccators may be avoided and some time saved by pouring the ignited ash into a counterpoised aluminum pan as soon as the dish is cool enough to avoid danger of loss from convection currents. The ash will cool almost instantly and may be weighed at once. The ash as finally obtained includes all non- volatile and incom- bustible matter in the paper. It may come from at least five sources: 1, materials in the pulps employed; 2, the loading or filling materials used; 3, substances used in coating or surface sizing; 4, mineral coloring matters or pigments; and 5, ash due to alum and size. It is possible that a paper may have an ash content as high as 5 per cent without being loaded but if this figure is exceeded it is safe to say some filler has been employed. In this connection it is interesting to note the following percen- tages of ash in fibrous raw materials as given by Wrede. 1 Stock Percentage ash Bleached linen half stuff o. 1 2-1 .'86 Bleached cotton half stuff o. 24-0. 79 Unbleached cotton half stuff o. 24-1. 12 Sulphite, unbleached o. 48-1 . 25 Soda o. 36-1.40 Adansonia . . .- 5. 70-7. 19 Japanese fibres 2.5 1 Wrede: Paper, Jan. 31, 191 2. 41 6 PAPER TESTING In coated papers the ash, of course, includes that from the coat- ing as well as from the body stock. If it is desired to examine each separately the coating may in most cases be removed as described under "amount of coating." The difference between the total ash and that in the body stock gives that present in the coating. If it is desired to calculate the original amount of filler used, or to make any computations regarding the minerals in the coating it is necessary to determine the nature of the mineral matters and in certain cases the amount of each present. This is rendered necessary because the different minerals lose different percentages of their weights in passing from the air dry to the ignited condition; some of the ground minerals used as fillers lose only about i per cent, clay shows a loss of around 1 2 per cent while crystalline calcium sulphate loses nearly 21 per cent. The complete quantitative analysis of an ash is a rather complicated process which can be carried out successfully only by a skilled chemist. The following simple tests may be of use where a com- plete chemical analysis is not considered necessary. Boil a little of the ash with water, filter and to the filtrate, acidified with hydrochloric acid,' add a few drops of barium chlo- ride solution. A white precipitate indicates sulphates. Warm a little of the ash with dilute hydrochloric acid filter and make slightly ammoniacal. Filter off any precipitate, which may have formed and add a little ammonium oxalate solution. A white precipitate indicates calcium. In order to be conclusive both these tests must be pronounced as small amounts of sulphates may be derived from the sizing materials and a little calcium is almost always present in clay. Evaporate a small sample of the ash to dryness after having moistened it with hydrochloric acid. Treat the dry residue with a little strong hydrochloric acid, take up a little of the moist material on a loop of platinum wire and hold it in a non-luminous gas flame. Calcium will impart a red color and barium a green color to the flame. If both are present the red color will show as soon as the wire is placed in the flame while the green will show with considerable persistence after the red has disappeared. SIZING MATERIALS 417 Boil the material moistened with strong hydrochloric acid in the above test, with a little water, filter and add a slight excess of dilute ammonia. A whitish, flocculent precipitate indicates alumina from the clay used. No simple test for magnesia, indicating the presence of talc, asbestine or similar minerals, can be given, as the qualitative test for this element is one which requires more than the ordinary skill. Retention. By retention is meant that per cent of the entire loading material added to the beater which appears in the finished paper. To ascertain this determine the following facts: P = weight of pulp added in pounds C = weight of clay added in pounds A = per cent of ash in finished paper A p = per cent of ash in pulp W c = per cent of water of composition of clay M p = per cent moisture in pulp M c = per cent moisture in clay. The per cent of clay used would then be — - — > or with greater IOO C (i - Me) accuracy P(i-lf,) ' The retention may be calculated by the following formula?, the second being the more accurate. Retention = Retention = C (100 - A) iooP(A -K) C(ioo -A+K) In this last formula K is the per cent of filler not derived from the loading added; an average value, which may be applied in the above formula, is 0.50. Sizing Materials. In its broadest sense the term " sizing " is applied to a number of materials used in the beater, in the coating and in surface sizing, and satisfactory qualitative tests 4 i 8 PAPER TESTING must be able to distinguish between these various substances and also to show whether a paper is tub (surface) sized or not. Starch may be, and frequently is, used for all three purposes and is applied either raw or cooked in the beater and cooked only in the other two cases. The universal test for starch is to apply a dilute iodine solution to the paper when a blue to violet color will appear if starch is present. It is well to" confirm this test by boiling some of the paper with a little water, filter- ing and testing the filtrate, after cooling, with a few drops of iodine solution. This is necessary because hydrocelluloses, which are only slightly soluble in boiling water, also give a blue color when brought into direct contact with iodine solution. Microscopic examination will show whether the starch granules have been burst by boiling or whether the starch was used with- out cooking. If the paper to be tested is torn so that it splits on the edge before being moistened with the iodine solution it is generally possible to tell whether it is surface sized or not. If it is surface sized only, the interior of the sheet will remain white while the surface will turn blue; if, however, consider- able starch was used in the beater, this is in part cooked and drawn to the surface by the heat of the driers so that the paper has the appearance of being surface sized when in reality it was not. Microscopic examination of the papers after treating with iodine will sometimes enable an opinion to be formed though it is seldom possible to prove positively in such a case whether the paper is surface sized or not. Casein may be detected in paper by moistening the sample with Millon's reagent and warming gently either over a flame or over an open steam bath. If casein is present a brick-red color will develop. In the case of coated paper in which much satin white is used, the alkali present determines the formation of a yellow color. In this case proof may be obtained by moist- ening the paper first with dilute nitric acid, to neutralize the alkali, and then applying the Millon's reagent as before; tested in this way satin white coated papers will give the usual red color. Casein may also be detected by boiling the paper with SIZING MATERIALS 419 water and a few drops of ammonia, filtering and adding to the filtrate dilute acetic acid very gradually. Casein will precipi- tate when the solution becomes very faintly acid, but it may redissolve on adding a considerable excess. This test is also given, though usually less strongly, by rosin, so the precipitate should be tested with Millon's reagent to confirm the presence of casein. Casein is seldom used except in the coating; cases of surface sizing or of its use in the beaters are very rare. Glue is sometimes used as an adhesive in coating papers and in rare instances in the beaters; the better grades known as gelatines are used in surface sizing. If glue is present alone it may be detected by boiling the sample of paper in water, filter- ing if necessary, and adding a little dilute tannic acid solution; a grayish, flocculent precipitate indicates glue. Casein is also precipitated by tannic acid and the presence of starch prevents the precipitation of glue so that when either casein or starch is present there is apparently no means of proving the presence or absence of glue. Rosin is used almost exclusively in the beaters to impart waterproof properties to the paper. There is no single test of a simple nature which will demonstrate positively the presence or absence of rosin and any judgment regarding it must be based on the indications of a number of different tests. If a little ether is dropped onto a sheet of paper and allowed to evaporate there will be formed, in the case of rosin-sized paper, a ring of rosin at the edge of the zone where the ether evapo- rated. This will be absent in most unsized papers, and it will, of course, be formed in any paper which contains any ether soluble material besides rosin. Another test is made by boiling a little of the paper for a few minutes in glacial acetic acid and pouring the acid into a little distilled water. A pronounced turbidity indicates rosin, but a slight opalescence may be caused by other soluble substances and must be disregarded. A third test is that known as the Raspail reaction. If a drop of concentrated sulphuric acid be placed on the paper and a 420 PAPER TESTING grain or two of sugar added a pronounced raspberry red color will develop with rosin-sized papers, while with unsized papers the color formed is brownish with only a trace of pink. This red color is also formed when albuminous materials are present so they must first be proved absent before the test can be con- sidered indicative of rosin. Rosin Determination. The amount of rosin in a sample of paper may be determined most accurately by the ether-alcohol method described by Sammet. 1 "Cut five grams of paper into strips approximately one-half inch wide and fold them into numerous small crosswise folds. Place the folded strips in a Soxhlet extractor and fill with acidu- lated alcohol diluted to approximately 83 per cent, made by adding to 100 c.c. of 95 per cent alcohol 15 c.c. of acidulated water containing 5 c.c. of glacial acetic acid to 100 c.c. of dis- tilled water. Place the Soxhlet flask directly in the boiling water of a steam bath and extract by siphoning from six to twelve times, according to the nature of the paper. Wash the alcoholic extract of rosin, which may contain foreign materials, into a beaker and evaporate to a few cubic-centimeters on a steam bath. Cool, take up in about 25 c.c. of ether, transfer to a 300-c.c. separatory funnel containing about 150 c.c. of distilled water to which has been added a small quantity of sodium chlo- ride to prevent emulsion, shake thoroughly, and allow to separate. Draw off the water into a second separatory funnel, and repeat the treatment with a fresh 25-c.c. portion of ether. Combine the ether extracts, which contain the rosin and any other ether-soluble material, and wash with 100-c.c. portions of distilled water until the ether layer is perfectly clear and the line between the ether and the water is sharp and distinct. Should glue which is extracted from the paper by alcohol interfere by emulsifying with the ether, it may be readily removed by adding a strong solution of sodium chloride to the combined ether extracts, shaking thoroughly and drawing it off, repeating if necessary, before washing with distilled water. Transfer the washed ether 1 J. Ind. Eng. Chem., 5, 732, Sept., 1913. CHLORINE 421 extract to a weighed platinum dish, evaporate to dryness and dry in a water oven at from 98 to ioo° C. for exactly one hour, cool, and weigh. This length of time is sufficient to insure complete drying. Prolonged heating causes a continual loss of rosin. Paraffin Determination. Weigh out enough of the paper to give a weighable amount of paraffin, — one to two grams should be sufficient, — and place the sample in a Soxhlet extraction apparatus, or an Erlenmeyer flask, fitted with a reflux condenser. Cover with gasoline or carbon tetrachloride and extract until the paraffin is all dissolved; if the Erlenmeyer flask is used a second extraction with a fresh amount of solvent will probably be neces- sary. Evaporate the solution to dryness and weigh the residual paraffin. If there is a tendency for the paraffin to creep over the edge of the dish it may be more satisfactory to weigh the paper after extraction and consider the loss in weight as paraffin. Either gasoline or carbon tetrachloride is satisfactory from the standpoint of solvent power but the latter is to be preferred because of its non-inflammability. It is also superior to chloro- form because its fumes are not likely to produce anesthesia. Chlorine. As generally spoken of in the mill " chlorine " means free chlorine or more often hypochlorites. These are tested for in the stock in the beater by adding to a small sample of it a few drops of potassium iodide starch solution. If they are present a characteristic blue color will result, its depth being somewhat proportional to the amount of chlorine present. Where the finished paper is to be examined it is best to moisten it with distilled water on a glass plate and test it with starch iodide paper. It is sometimes desired to determine the total chlorine present, including that as hypochlorites, inorganic chlorides and organic chlorides. This may be accomplished by moistening a weighed sample of the paper with a solution of chlorine-free sodium car- bonate, drying it in a platinum crucible and then igniting cau- tiously until all the organic material is completely reduced to 422 PAPER TESTING carbon. The soluble salts are then leached out and the chlorides in the solution are determined by titration with N/io silver nitrate solution. Free Acid Determination. Weigh 10 grams of the paper to be tested, tear into small pieces, place in a porcelain casserole and cover with a small amount of distilled water. Heat gently for an hour over water bath or electric hot plate. Pour off water and wash with small quantities of distilled water, adding same to water extract. Make up to ioo c.c. according to directions given on page 103 of Cohn's Indicators and Test Papers. The solution is then poured into a 100-c.c. Nessler tube (long form). A similar tube is filled with 100 c.c. of distilled water to which has been added two drops of the litmus solution. To the former is then added tenth normal standard solution of caustic soda until the color matches the sample. The acidity is then expressed in terms of S0 3 . Sulphur Determination. Sulphur may be present in paper in a number of different forms and from several causes. White papers may be toned with ultramarine, which nearly always contains sulphur, and which evolves hydrogen sulphide in the presence of alum or acid. The use of sodium thiosulphate as an antichlor or the presence of improperly prepared sulphite fibre may also account for the presence of sulphur compounds. The users of tissue paper for wrapping silver demand that the paper shall be free from sulphur and a method for its determina- tion, based on the stain produced on lead acetate paper, has been worked out as follows: 1 The apparatus consists of a 500-c.c. round bottom flask with a neck about 2 ins. long and 1 in. in diameter. The mouth of this neck is ground to a flat surface and on this is placed a glass tube about 4 ins. long and an inch in diameter, the lower end of which is also ground flat to fit tightly upon the upper surface of the neck of the flask. The whole is so arranged that after placing a piece of filter paper between the two ground surfaces, 1 Sutermeister: Pulp Paper Mag. Can., 15, 1917, 1021. SULPHUR DETERMINATION 423 the tube and flask can be securely clamped together so that all gas generated in the flask must pass through the filter paper and then up through the superimposed glass tube. The procedure for the testing of tissue papers is as follows: A sample of 25 sq. ins. is taken and its weight determined. It is then shaken up in a wide mouth glass-stoppered bottle with 10 c.c. of distilled water; when partial disintegration has taken place, another 10 c.c. of water is added and the shaking continued until the paper has been completely reduced to pulp. The larger part of the pulped mass is now transferred to the flask described above, and the residue which is left in the bottle is rinsed into the flask with a mixture of 10 c.c. of sulphur-free phosphoric acid and 20 c.c. of water. Prepare turnings from the highest grade, pure stick zinc, which must be free from sulphur and arsenic. Treat one gram of these turnings with 10 c.c. of a dilute solution of copper sulphate containing about 0.002 gram actual copper. After a few minutes all the copper will have deposited and the turnings are then thoroughly washed to remove every trace of zinc sulphate. The turnings are added to the flask and a wad of cotton inserted in its neck. Between the two ground glass surfaces is then clamped a piece of filter paper about 2 ins. square which has been perforated with small pinholes about | of an inch apart and which just before use is moistened with several drops of lead acetate solution. Finally a loose wad of cotton is placed in the tube above the paper. The flask is placed on the steam bath and allowed to stay, with occasional shakings, for an hour. The. filter paper is then removed from the neck of the flask and air dried. It is best compared with the standard test pieces by placing them side by side on a piece of white paper and covering them with a thin piece of clear, white glass. The standard test pieces are pre- pared by using sulphur-free cotton in the flask instead of the disintegrated paper and adding to this definite volumes of a very weak solution of sodium thiosulphate whose strength is accurately known. The sulphur-free cotton is prepared by 424 PAPER TESTING boiling absorbent cotton in weak caustic soda solution and wash- ing thoroughly with distilled water. The sensitiveness of this test is such that the presence of o.oooooi gram of sulphur in the flask will give a distinct color on the lead acetate paper. From tests of a considerable number of papers which have been found satisfactory in actual practice it has been proved that tissue paper is safe for wrapping silver- ware if it does not contain more than 0.000002 gram of sulphur per 25 sq. ins. of paper (about 0.25 gram). In using this method certain precautions are necessary if reliable results are to be obtained. The sample to be examined must be kept away from all dust and laboratory fumes and in handling it in preparing the sample the fingers should be scrupu- lously clean. Any perspiration on the fingers will be sufficient to give an incorrect result and it was found by sad experience that if the fingers were run through the hair while handling the sample for analysis a strong test for sulphur was obtained. It is quite obvious that any sample which is submitted for tests should be taken from a freshly opened package of paper and that it should be handled as little as possible. To make a test of a sample which has been lying around an office for some time and which has been handled repeatedly is merely a waste of time. The purity of the zinc employed must be ascertained by very careful tests and the same is true of the phosphoric acid. Other acids should not be substituted for phosphoric acid unless they are first proved to be absolutely reliable and it has been demon- strated that sulphuric and hydrochloric acids are not entirely safe. The standard test papers, which are prepared with known amounts of sulphur, must be freshly made each time the test is carried out since they are not sufficiently permanent even when kept in tightly stoppered bottles in the dark. Amount of Coating. It is sometimes desirable to determine the percentage of coating on a sample of paper or the amount applied per ream. To do this cut a sample of some definite size and weigh in the air dry state; soak this sample for a few minutes GLUE OR CASEIN DETERMINATION 425 in a dilute solution of ammonium hydroxide, remove it to a glass plate and loosen the coating by brushing carefully with a flat camel's hair brush moistened with the ammonia solution. Fi- nally wash in running water, air dry, and weigh. The difference in weight represents the coating removed. In some cases the coating is so thoroughly waterproofed that this treatment fails to remove it; in such cases there is no known method for determining its amount. Where the coating is partly waterproofed it may be desirable to increase the strength of the ammonia or to warm it slightly. Where the coating is not easily removed the brushing must be done with great care lest a con- siderable loss in weight be caused by removal of fibrous material. It is sometimes a question of judgment as to whether a greater error is caused by failure to remove traces of coating or by rubbing off some of the fibres. Glue or Casein Determination. There is apparently no quan- titative method known for the determination of these substances when they are present together. Both materials contain nitro- gen. If only one be present and the nitrogen content of the original material as added to the paper be known then by means of the nitrogen determination the content of glue or casein may be estimated. For the determination of nitrogen weigh out from three to five grams of the paper, cut into small pieces and place in a Kjeldahl digestion flask. Add 10 grams potassium sulphate, 0.7 gram mercury and 25 c.c. of concentrated sulphuric acid; place the flask in an inclined position in a hood with a good draft and heat below the boiling point of the acid until frothing has ceased. Increase the heat until the acid boils briskly and continue the digestion until the solution is colorless or pale straw yellow. Allow the contents of the flask to cool and add 30 c.c. of a 4 per cent solution of potassium sulphide. Next add 50 c.c. of caustic soda solution (saturated solution), or enough to make the reaction strongly alkaline, pouring it carefully down the side of the flask so that it does not mix with the contents. Connect the flask at once with the condenser, shake to mix the contents of the flask, 426 PAPER TESTING and distil until all ammonia has passed over. Collect the dis- tillate in a flask containing a known volume of standard acid, and at the end of the test titrate the excess of acid with standard alkali, using sodium alizarin sulphonate or methyl red as indica- tor. The end of the condenser tube should dip below the surface of the acid, the distillation should require about 45 mins. and the distillate should amount to about 200 c.c. If a few pieces of granulated zinc or pumice stone are added to the distillation flask bumping during the distillation may be avoided. As a check on the purity of the reagents a "blank test" should be made using the same amounts of chemicals as in the regular test but adding no paper. Substract the volume of alkali required to neutralize the dis- tillate from the volume required by the blank; the difference is the number of cubic centimeters of standard alkali equivalent to- the ammonia. This number of cubic centimeters of tenth nor- mal alkali multiplied by 0.0014 gives the grams of nitrogen in the sample taken. To convert this to casein multiply by the factor 6.3, and for glue use the factor 5.6. Since the percentage of nitrogen varies in different lots of casein and glue these factors should be determined, wherever possible, on the actual materials used in the paper being tested. Unbleached Fibres. Besides the microscopical method de- scribed elsewhere it is possible to detect unbleached sulphite fibre by chemical means. If a sheet of paper containing a small amount of unbleached fibre is moistened with Millon's reagent and then warmed in the steam escaping from a steam bath it will be found that the unbleached fibres will show up very distinctly as brownish hairs. If no unbleached fiber is present the brown hairs are absent. From the proportion of fibres stained brown it is possible to make an approximate estimate of the amount of unbleached fibre present, provided its quantity is not large. Where the sheet is mostly unbleached, or where groundwood is present in considerable amount, the entire surface becomes brown and no estimate of the proportion of unbleached fibre can be made. GROUND WOOD PULP 427 Groundwood Pulp. For the qualitative determination of groundwood a number of stains are available and find quite general application. Aniline sulphate stains groundwood a yellow color, though the test is not quite so sensitive as that with phloroglucinol. The aniline sulphate solution should be prepared by dissolving 5 grams in 50 c.c. of distilled water and acidulating with one drop of concentrated sulphuric acid. A very satisfactory stain is made by dissolving 1.0 gram of paranitroaniline in 405 c.c. of distilled water and 30.5 c.c. of sulphuric acid (sp. gr. 1.84). With groundwood this gives an intense orange color which develops without drying the sample. This stain possesses an advantage over phloroglucinol in that the acid is not volatile and that it does not deteriorate so rapidly. One of the most generally used stains is prepared by dissolving 5 grams of phloroglucinol in a mixture of 125 c.c. of distilled water and 125 c.c. of concentrated hydrochloric acid. This solu- tion should be kept in the dark as light affects its staining proper- ties to some extent. With groundwood this stain causes a ma- genta color, the depth of which is approximately proportional to the amount of groundwood present. A very light shade, however, does not necessarily prove the presence of groundwood, as partly cooked jute, undercooked unbleached sulphite and some other fibres are also"slightly colored. The reaction with phloroglucinol has been employed by Cross, Bevan, and Briggs 1 as the basis of a quantitative method for estimating groundwood. The necessary solutions are: 1. 2.5 grams of pure phloroglucinol dissolved in 500 c.c. of hydrochloric acid, sp. gr. 1.06. 2. 1 c.c. of 40 per cent formaldehyde in 500 c.c. of hydrochloric acid, sp. gr. 1.06. The paper under examination should be rasped to a loose pow- der but size need not usually be extracted. Two grams of the powdered sample are dried at ioo° C, weighed, transferred to a. dry flask and covered at once with 40 c.c. of phloroglucinol solu- 1 Cross, Bevan and Briggs: Papier Ztg., 32, 1907, 4113 and 4479. 428 PAPER TESTING tion. The flask is stoppered, shaken, and allowed to stand several hours or best over night. The solution is next filtered through a very little cotton placed in a funnel and 10 c.c. are measured out for titration. This is diluted with 20 c.c. of hydro- chloric acid (sp. gr. 1.06), warmed to about 70 C. and the formaldehyde solution added from a burette in lots of 1 c.c. at a time. Allow to stand two minutes after each addition and then remove a drop without filtering and place it on a strip of partly sized news paper. After ten seconds shake the drop off and see if a red color is produced; if it is, add more formaldehyde solution and toward the end of the reaction reduce the amount added each time to 0.25 c.c. Non-development of a red color indicates the end of the reaction. The test paper is sensitive to a solution of one part of phloroglucinol in 30,000. A' control test should be made with 10 c.c. of the phloroglucinol solution in exactly the same way and the absorption of the groundwood calculated from the difference of the two titrations. It should be based on the dry material taken for analysis. In carrying out this test it is essential that pure phloroglucinol be used and it is well to obtain a considerable supply so that its uniformity may be assured. The proportion of sample to phloro- glucinol solution must be kept constant and the temperature must be maintained at 70 C. during the titration. The absorption numbers of various fibres have been found to be as follows: Percent Groundwood 7. 87-8. 15 Groundwood, brown 5. 52 Sulphite, bleached o. 90-1. 00 Sulphite, unbleached o. 90-1. 03 Esparto o. 50 Cotton o. 20 Assuming an absorption number of 8 per cent for groundwood and 1 per cent for sulphite then ry IOO (P — I.o) ti = — 1 8.0 — I.O where H = per cent of groundwood

2200° 1440° -l86. 1° 5-69 6.23 28600 .O4O559 40° < 3 6o° 554° vol. 950 1420° sublimes at sublimes at 850 268 ° 2200 in vacuo 7 8 9 6.27 3-37 1.82 4-44 8-57 6.16 6-93 5-83 2.89 2 .22 1 .76 6.28 4.40 8.02 5-42 6.08 9260 .0177 .04 1 346 40° f { 3500° -7-3° 321 28.45° 805 ° sublimes at sublimes at sublimes at 635° — 102 ° 58.7 765-9 670 13 15 146000 25400 95000 .2213 .O43069 .O339482 40° 27-100 .O4O54 40° 40 ° 40 ° 35oo ° 35oo° 35oo ° i7 t8 .O4O786 .O4OI18 19 13950 "33-6° .O21978 0-10 1520 1478° 1950 fio8 3 ° 1 1065 ° (in air) 2200 23 24 25 "6 83200 .O4I236 40 ° 5-95 640600 .7198 .O41678 40 ° 2310°, 2000 in vacuo 27 *8 -223 -187 29 3° 3 T 5-53 5-34 6.23 30.15° 916 1280 1062 ° < — 269° vol. 1350 >i9oo 2530 , 2000 in vacuo -268. 75 54100 200 468000 33 .7003 .033386 .O4I470 o-ioo * K = the number of grams of water which can be raised from o° to 1° C. by the heat which passes through a cubic centimeter of the substance in one second when the temperature of the opposite sides of the cube are maintained at a difference of l° C. t Reciprocal of the resistance in ohms of a centimeter cube of the substance. 45° APPENDIX PHYSICAL CONSTANTS Name Sym- bol Atomic weight. = i6 Molec- ular weight Specific Gravity. Water = I. Air=i (A). Hydrogen = i (D.) Atomic vol. At.wt. Sp.Gr. Specific heat at o° C. Hydrogen, gas. liquid 3 4 5 6 7 8 9 io ii 12 13 14 IS 16 17 18 19 20 21 22 23 24 25 26 27 28 29 3° 31 32 33 34 35 36 37 38 39 40 Indium Iodine, gas solid Iridium, spongy. crystalline .... Iron, pure wrought steel gray pig white pig. Krypton, gas. . . liquid Lanthanum Lead Lithium Magnesium Manganese Mercury Molybdenum. . . Neodymium. . . . Neon Nickel Nitrogen, gas. . . liquid Osmium Oxygen, gas. . . . liquid Ozone Palladium Phosphorus, yel. red liquid Platinum Potassium Praseodymium. . Radium Rhodium Rubidium Fe Fe Fe Fe Fe Kr Kr La Pb Li Mg Mn Hg Mo Nd Ne Ni N N Os O O 3 Pd P P P Pt K Pr Ra Rh Rb 1 .008 1 .008 114. 8 126.92 126.92 193- 1 193- 1 55-84 55-84 55- 55- 55-84 82.92 82 .92 139.0 207 .20 6.94 24.32 54-93 200.6 96.0 144-3 14.01 14.01 190.9 16.00 16.00 2 .016 2 .016 253-84 82 .92 200.6 28.02 32 .00 32 .00 48.00 106.7 31-04 3*-°4 3!-°4 195 39.10 140.9 226.0 102 .9 85-45 124.16 124.16 0.06949 A. —252.83° O.O7IO5 745.52 mi „ mm. 7.12¥ 8.72 A. 4.948 170 15-86 22 .42 7-85-7-88 7.86 7 . 60-7 . 80 7.03-7.13 7-58-7-73 /2.818 A.\ l 40. 78 D. J 2.155- 1520 6-1545 "■337i£H o.534 200 i-69-i-75 7.42 13 -5953 s 10 . 281^ 6.9563 r 0.695 A. 1 I9.96 D.J 8.6-8.93 0.96737 A. o.8o42- 195 - 5 ° 22 .48 1. 10535 A. i.ii8i- 182 - 5 ° 1.658 A. 1 1 . 4-1 1 . 9 1.831 180 2.296 160 i.764 44 - 3 ° 2i.i6¥° 0.8621 200 6-4754 1.4 16. 1 25-7 12 .2 8.6 7-i 7-i 7-3 7-9 7-3 3 -4io 0.05695 0.0336 2060 0.04852 0.0323 1162 1130 0.1066 0.1050 38.5 22.6 18.3 13.00 14. 1 7-4 14.8 9-3 04485 0310 8366 0.2456 0.1217 0.03346 0.0659 i-53 2 6.7 0.1084 0.2438 17-4 8-5 0.031 13 0.2175 14-3 9.2 0.0592 17 .0 0.202 13-5 0.1829 17.6 9.2 0.0323 44-7 0.1729 21.8 8.5 55-78 0.05803 0.0802 PHYSICAL CONSTANTS OF THE ELEMENTS 451 OF THE ELEMENTS u 6 Electrical conduc- tivityf at o° C. Thermal conductivity K* at o° C. Ag=i.oo Linear coefficient of expansion Melting point, °c, Boiling point, °C. T 3-44 6.05 6.56 4.27 6.16 .O33270 At°C. -259° -252. 5° ?. 3 II9SOO . O441 7 40° 155° 700 ° S 6 .O4837 — 190-17 II4. 2° 2250° I9SO° IS30 l600° 1375° 1275° 1075° -169 184.3S 7 6.23 6.50 6.32 5-96 5-87 .O4O700 40 ° o°-ioo° o°-ioo° o°-ioo° 40 ° 8 13 IOOO .1665 .2070 .1300 .O4I182 .O4II .O4II .O4I061 2450 ° 10 63000 1 I0200- \ I 1300 TT T2 .1490 13 -iSi-7° TS 6.23 352 5.86 5 -98 6.70 6.69 6-33 8io° 327°' 186 650 1260 -38. 85 . 2500° 840 ° -253° 145 2 ° -210. 5 16 17 18 19 20 2T 50400 I I 9000 230000 .0836 .O42924 40 o°-i8o° 40 ° 1525° >i400° 1120 1900 357-33° .3760 .O42694 10630 .0148 .O3182 o°-ioo° ?.?. 23 24 -243° 6.36 3-42 144200 .1420 •O4524 .O4I279 40 ?5 -195-5° afi 21 5-95 3-48 105300 .O40657 40 2700 -227 ^ •O4563 -182. 7 20 ^0 decomp. 270 i5So° 44. i° 725° -119 Si 6.32 6.26 5-67 97900 .1683 .O4II76 40° o°-44° 32 290 350° (yel.) ^!S S4 ss 6.29 6.51 91200 150500 .1664 . O40899 40° o°-50° 1753° 63-S 940 ° 700 ° 1970° 39° ^6 757-5° S7 S8 ^o 5-97 . O40850 40° 40 696 ° 45 2 APPENDIX PHYSICAL CONSTANTS Name Sym- bol Atomic weight 0=i6 Molec- ular weight Specific Gravity. Water = i. Air=i (A). Hydrogen=i (D). Atomic vol. At. wt Sp. Gr. Specific heat at o° C. Ruthenium, spon. . melted cryst Samarium Scandium Selenium, amorph. monoclinic hexagonal Silicon, amorph. . . cryst Silver Sodium Strontium Sulphur, amorphous soft, yellow rhombic monoclinic plastic Tantalum Tellurium, amorph cryst Terbium Thallium Thorium, amorph. cryst Thulium Tin, gray rhombic tetragonal Titanium Tungsten Uranium Vanadium Xenon, gas liquid Ytterbium Yttrium Zinc Zirconium , amorph cryst Ru Ru Ru Sm Sc Se Se Se Si Si Ag Na Sr S S Sa S/3 St Ta Te Te Tb Tl Th Th Tm Sn Sn Sn Ti W U V Xe Xe Yb Yt Zn Zr Zr 101 .7 101 .7 101 .7 150.4 44.1 79.2 79.2 79.2 28.3 28.3 107.8, 23 .00 87.63 32 .06 32 .06 32 .06 32 .06 32 .06 181. 5 127-5 127.5 159.2 204.0 232.40 232.40 168. s 118. 7 118. 7 118. 7 48.1 184.0 238.2 5i-o 130.2 130.2 173-5 88.7 65-37 90.6 90.6 8.6 11. 4 12.268° 7-7-7,* II. 8 8.9 8-3 19.4 O.061 I 633-6 633.6 633.6 256.48 256.48 256.48 256.48 256.48 255 -o 255-0 4.26-4.28 250 4.47 250 4.8 250 2.00 2.49!°° 10.53 0.97I2 200 2-54 i-9556°° 2 .046 2.05-2.07°° 1.958 1 .92 i 4 . 49 if° 6.oi5 2 °° 6.27 18.5 17.7 16.5 14.2 11. 4 10.2 23.6 34-5 16.4 15-6 15.6 0-09533 o . 08401 o.24 21 ° O.I697 220 0-0559 0.2813 11.85 n.ool? 11.23 5-! 6.53-6.56 7 . 2984 150 4-5o 17 - 5 ° 18.77 18.685V 6.025IS J4.22 A. 1 1 63. 5 D.J 3 . S2 -i».io 65-37 3 .8o 15 ° 7.142 160 4-15 6 . 40 180 20.4 0.1728 0.1809 0.1902 o . 0301 7 0525 0-0475 17.2 21. 1 20.7 0.0326 20.3 18.2 16.3 10.7 9.8 12.8 8-5 0545 0-0559 0559 0.1125 0.0336 0280 o . 1 240 37-o 23-4 9.2 21.8 14.2 0.09356 o . 0660 PHYSICAL CONSTANTS OF THE ELEMENTS 453 OF THE ELEMENTS u CD X> s S S & 3&x Electrical conduc- tivity^ at 0° C. Thermal conductivity K* at o° C. Ag=i.oo Linear coefficient of expansion Melting point, C. Boiling point, °C. At°C. >i 9 5o° 2000 ° 2000 ° I350 1200° 50° (softens) i7o°-i8o° 217 • ? 3 4 5 6 6.21 .O4O963 40° 7-55 6.65 690° 690° 690° 35°°° 35°°° 1955° 877-5° 7 8 O43680 40° 9 TO 6.06 4.82 6.04 6.47 200-15600 681200 21 IOOO 40300 .O4O763 .O4I921 .O472 40° 40° o°-50° 1420 961.5° 97-6° 900 ° >I20° ignition pt. 255 114. 5° 119.25° II 12 13 14 t6 I .OOO •365 444.6 444-6° 444.6° 444.6° 444.6° 5-54 5.80 6.10 5-46 6.69 6.07 •047433 • O46413 i3°-5o° 40° 17 t8 19 60600 O408 ■O41675 . O43440 40° 0°-20° 2900° 44 6° 452° 1 390° 1390° ?T 46600 32 23 24 25 06 6.65 56800 .O43021 40 ° 302° >i7oo° 1280° ?7 6.49 6.65 6.65 5-4i 6.18 6.68 5 90 stable <2o° stable >i7o° 232° 1795° 3267° 800 ° 1720° — 140° ?8 >22 7 5° 1450-1600 29 76600 .1528 ■O42234 40° ^T 32 3^ 34 — 109. 1 ° ^5 ?fi 1800° 1250° 419° 1 500° 2350° 37 38 ^0 6.12 186000 •2653 .O42918 40 ° 918° /|0 5-98 454 APPENDIX VAPOR PRESSURE OF WATER According to Regnault °c. p_ Inches of mercury Pounds per square inch °c. °F. Inches of mercury Pounds per square inch o 32.0 0.181 0.0890 42 107 .6 2 .404 1 .216 i 33-8 0.194 0.0955 43 109.4 2-533 1.244 2 356 0.209 0.1025 44 in .2 2.669 1. 312 3 37-4 0.224 O.I IOO 45 113. 2. 811 1. 381 4 39-2 0.240 0.1 180 46 114. 8 2-959 1-454 5 41 .0 0.257 0.1263 47 116. 6 3-H4 1-530 6 42.8 0.276 O.I354 48 118. 4 3.276 1 .609 7 44.6 0.295 0.1452 49 120.2 3-444 1 .692 8 46.4 0.316 O.I55I 50 122 .0 3.62 1.78 9 48.2 0-338 0.1657 5i 123.8 3-8i 1.87 IO 50.0 0.361 0.1773 52 125.6 4.00 1 .96 ii 51.8 0.386 0.1893 53 127.4 4.20 2 .06 12 53-6 0.412 0.2023 54 129.2 4.41 2.17 13 55-4 0-439 0.2158 55 131-0 463 2 .27 14 57-2 0.469 0.2303 56 132.8 4-85 2-39 15 59-0 O.500 0.2456 57 134-6 5-09 2.50 16 60.8 0-533 0.2618 58 136.4 5-33 2 .62 17 62.6 0.568 0.2789 59 138.2 5-59 2-75 18 64.4 0.605 0.2970 60 140.0 5-86 2.88 19 66.2 0.644 0.3162 61 141. 8 6.14 3.01 20 68.0 0.685 0.3363 . 62 143.6 6.42 3.16 21 69.8 0.728 0.3577 63 145-4 6.72 3 30 22 71 .6 0-774 0.3802 64 147.2 7.04 3-46 23 73-4 0.822 . 4040 65 149.0 7 36 3-62 24 75-2 O.873 0.4289 66 150.8 7.70 3-78 25 77.0 0.927 0.4554 67 152.6 8.05 3-95 26 78.8 0.984 0.4833 68 154-4 8.41 4-13 27 80.6 I.944 0.5126 69 156.2 8-79 4-32 28 82.4 1 .106 0.5434 70 158.0 9.18 4-51 29 84.2 1 .172 0.5759 7i 159.8 9-58 4-7i " 3° 86.0 1 .242 0.6101 72 161. 6 10.00 4.91 31 87.8 I.3I5 0.6461 73 163.4 10.44 512 32 8 9 .6 1.392 0.6838 74 165.2 10.89 5-35 33 91.4 1-473 0.7234 75 167.0 11.36 5-58 34 93-2 1-558 0.7655 76 168.8 11.84 5.82 35 95 1.647 0.810 77 170.6 12-35 6.06 36 96.8 1.740 0.855 78 172.4 12.87 6.32 37 98.6 1.838 0.903 79 174.2 13-40 6.58 38 100.4 1. 941 0.954 80 176.0 13.96 6.85 39 102 .2 2.049 1 .007 81 177-8 14 -54 7.14 40 104.0 2.162 1 .061 82 179.6 15-14 7-44 4i 105.8 2.280 I .121 83 181. 4 15-75 7-74 Reprinted from D. Van Nostrand's Chemical Annual, Fourth Issue, 1918, with the permis- sion of the editor and publishers. VAPOR PRESSURE OF WATER 455 VAPOR PRESSURE OF WATER {Continued) °c. °F. Inches of mercury Pounds per square inch C. °F. Atmos- pheres Pounds per square inch 84 183.2 16.39 8.05 129 264.2 2.592 38.II 8S 185.O I7-05 8-37 130 266.0 2 . 671 39.26 86 186.8 17-73 8.71 131 267.8 2 . 753 40.47 87 188.6 18.43 9 05 132 269.6 2 . 836 41.68 88 190.4 19.16 941 133 271.4 2 . 921 42.93 89 192.2 19.91 9.78 134 273.2 3 008 44.21 90 194.O 20.69 10.16 135 275.0 3 097 45-52 9 1 195.8 21.49 10.56 136 276.8 3 188 46.87 92 197.6 22.31 io.95 137 278.6 3 282 • 48.24 93 199.4 23.17 n.38 138 280.4 3 378 49 65 94 20I .2 24.04 n. 81 139 282 .2 3 476 51.06 95 203 .O 24-95 12 .26 140 284.0 3 576 52.55 96 204.8 25.89 12 .71 141 285.8 3 678 54-07 97 206.6 26.85 13-19 142 287.6 3 783 55-6o 98 208.4 27.85 13.68 143 289.4 3 890 57i6 99 2I0.2 28.87 14.18 144 291 .2 4 000 58.79 J 29-92 l 1. 000* 145 293.0 4 113 60.44 100 2I2.0 14.70 146 294.8 4 227 62.13 101 213.8 1 .036* 15-23 147' 296.6 , 4 344 63.86 102 215.6 1.074* 15-79 148 298.4 4 464 65.62 103 217.4 1 .112* i6.35 149 300.2 4 587 67.41 104 219.2 1. 152* 16.94 150 302 .0 4 712 69.26 1 OS 221 .O 1. 193* 17-53 iSi 303-8 4 840 71.14 106 222 .8 1-235* 18.15 152 305-6 4 971 73.06 107 224.6 1.278* 18.78 153 307.4 5 104 75.02 108 226.4 1 .322* 19.44 154 309.2 5 240 77-03 109 228.2 1.368* 20.11 155 311-0 5 380 79.07 no 23O.O i-4i5* 20.80 156 312.8 5 522 81 .22 III 23I.8 1.463* 21.51 157 314.6 5 667 83.29 112 233-6 i-5i3* 22 .24 158 316.4 • 5 815 85-47 n.3 235-4 1.564* 22 .99 159 318.2 5 966 87.69 114 237.2 1. 616* 23.76 160 320.0 6 .120 89.96 us 239.0 1 .670* 24-55 161 321.8 6 .278 92.27 116 240.8 1 .726* 25-73 162 323-6 6 •439 94-63 117 242 .6 1.782* 26 .20 163 325.4 6 • 603 97.04 118 244.4 1. 841* 27 .06 164 327.2 6 •770 99-50 119 246.2 1 .901* 27.94 .165 329.0 6 •940 102 .01 120 248.0 1 .962* 28.85 166 330.8 7 •114 104.56 121 249.8 2 .025* 29.78 167 332-6 7 .291 107.18 122 251.6 2 .091* 3° -73 168 334-4 7 •472 109.84 123 253-4 2.157* 31.70 169 336.2 7 .656 112-53 124 255-2 2 .225* 32.70 170 338.o 7 .844 115.29 125 257.0 2.295* 33-72 171 339-8 8 .036 118. n 126 258.8 2 .366* 34.78 172 341.6 8 • 231 120.98 127 260.6 2.430* 35-86 173 343-4 8 •430 123.90 128 262 .4 2.515* 36.97 174 345-2 8.632 126.87 ♦Atrnc spheres. 456 APPENDIX VAPOR PRESSURE OF WATER {Continued) °c. °F. Atmos- pheres Pounds per square inch °C. °F. Atmos- pheres Pounds per square inch 175 347 -o 8.839 129.91 203 397-4 16.364 240.54 176 348 8 9.049 133.OO 204 399-2 16.703 245-49 177 35° 6 9.263 136.15 205 401 .0 17.047 250.53 178 352 4 9.481 139-35 206 402 .8 17.396 255.67 179 354 2 9-703 142.62 207 404.6 17-751 260.88 180 356 9.929 145-93 208 406.4 18. Ill 266.18 181 357 8 10.150 I49-32 209 408.2 18.477 271-55 182 359 6 10.394 152.77 210 410.0 18.848 277.01 183 361 4 10.633 156.32 211 411. 8 19 .226 282.58 184 363 2 10.876 I59-84 212 4136 19.608 288.21 185 365 11. 123 163.47 213 415-4 19.997 293.92 186 366 8 H-374 167.17 214 417.2 20.391 299 . 72 187 368 6 11 .630 170.94 215 419.0 20.791 305.57 188 37° 4 11.885 174.76 216 420.8 21 .197 3«-57 189 372 2 12.155 178.65 217 422 .6 21 .69O 317.62 190 374 12.425 182.61 218 424.4 22 .027 323-78 191 375 8 12 .699 186.63 219 426.2 22.452 330.0I 192 377 6 12.977 190.72 220 428.0 22.882 336-30 193 379 4 13 .261 194.88 221 429.8 23-3I9 342.70 194 381 2 13 • 549 199-13 222 431-6 23.761 349.21 195 383 13.842 203.43 223 433-4 24.2IO 355-81 196 384 8 14-139 207.81 224 435-2 24.666 362.50 197 386 6 14.441 212 .25 225 437 -o 25 .128 369 • 29 198 388 4 14-749 216.77 226 438.8 25-596 376.17 199 39° 2 15 .062 221.37 227 440.6 26.071 383-15 200 392 15-380 226.04 228 442.4 26.552 390.22 201 393 8 15-703 230.79 229 444.2 27.O4O 397-40 202 395-6 16.031 235-6i SODIUM CHLORIDE SOLUTION AT 15 ° Gerlach Specific Gravity Per cent NaCl Specific Gravity Per cent NaCl Specific Gravity Per cent NaCl Specific Gravity Per cent NaCl 1 .00725 1. 01450 1 .02174 1 .02899 1 .03624 I .04366 I .05108 I 2 3 4 5 6 7 I. 05851 I.06593 1-07335 I .08097 I.08859 I .09622 1-10384 8 9 10 11 12 13 14 1 .11146 1 .11938 1 .12730 1 13523 I-I43I5 1-15107 I-I593I 15 16 17 18 19 20 21 1 16755 i-i758o 1 .18404 1 .19228 1 . 20098 1 . 20433 22 23 24 25 26 26.395 Reprinted from D. Van Nostrand's Chemical Annual, Fourth Issue, 1918, with the permis- son of the editor and publisher. SODIUM CARBONATE SOLUTION AT 15' 457 SODIUM CARBONATE SOLUTION AT 15° Lunge 1 liter contains grams Specific Degrees Baume Per cent Na 2 C0 3 Per cent Na 2 C0 3 .io H 2 Gravity Na 2 C0 3 . Na 2 C0 3 .io H 2 1 .007 1 .0 0.67 I.807 6.8 18.2 1 .014 2.0 i-33 3.S87 13-5 36.4 1 .022 3-1 2.09 5-637 21 .4 57-6 1 .029 4.1 2.76 7-444 28.4 76.6 1.036 5-1 3-43 9.251 35-5 95-8 I 045 6.2 4.29 11.570 44.8 120.9 1.052 7.2 4-94 I3-323 52.0 140.2 1 .060 8.2 5-7i 15 .400 60.5 163.2 1 .067 9.1 6-37 17.180 68.0 183.3 1 075 10. 1 7 .12 19.203 76.5 206.4 I.083 11 .1 7.88 21 .252 85.3 230.2 1 .091 12 .1 8.62 23.248 94 253-6 1 .100 13.2 9-43 25-432 103.7 279.8 1. 108 14. 1 10.19 27.482 112 .9 304'5 1 .116 151 10.95 29-53 2 122.2 329.6 1. 125 16. 1 11. 81 3I-85I 132.9 358.3 I 134 17. 1 12 .61 34.009 143 385.7 1 .142 18.0 13.16 35-493 150.3 405-3 1. 152 19. 1 14.24 38.405 164. 1 442.4 Reprinted from D. Van Nostrand's Chemical Annual, Fourth Issue, 1918, with the permis sion of the editor and publishers. 458 APPENDIX CONCENTRATED SODIUM CARBONATE SOLUTION AT 30° Lunge 1 liter contains grams Specific Degrees Per cent Per cent Gravity Baume Na 2 C0 3 Na 2 C0 3 .io H 2 Na 2 C0 3 Na 2 C0 3 .io H 2 1 .142 18.O 13-79 37.21 157-5 425.0 1.152 19. 1 14.64 39-51 168.7 455-2 1 .162 20.2 15-49 41-79 180.0 485.7 1 .171 21 .2 16.27 43-89 i9°-5 514.0 1. 180 22.1 17.04 45-97 201 .1 542.6 1 .190 23.I 17.90 48.31 214.0 577-5 1 .200 24.2 18.76 50.62 225.1 607.4 1 .2IO 25.2 19.61 52.91 237-3 640.3 T.220 26.1 20.47 ■ 55-29 249.7 673.8 I-23I 27.2 21 .42 57.8o 263.7 7II-5 I .241 28.2 22.29 60.15 276.6 746.3 1.252 29.2 23-25 62.73 291 .1 785.4 I.263 30.2 24.18 65.24 305-4 824.1 I.274 31.2 25.11 67.76 3I9-9 863.2 I.285 32.2 26.04 70.28 334-6 902.8 1.297 33-2 27.06 73.02 35i-o 947.1 I.308 34-i 27.97 75-48 365-9 987.4 Reprinted from D. Van Nostrand's Chemical Annual, Fourth Issue, 1918, with the permis- sion of the editor and publishers. BAUME AND SPECIFIC GRAVITY 459 EQUIVALENT OF DEGREES BAUME (AMERICAN STANDARD) AND SPECIFIC GRAVITY AT 60 ° F. Degrees Baume = 145 — ~ — p — f° r Liquids Heavier than Water Degrees Specific Deg rees Specific Degi ■ees Specific Deg ■ees Specific Baume Gravity Bau me Gravity Bau me Gravity Bau me Gravity O.O I .0000 7 I .0262 4 1-0538 I I.0829 1 I .0007 8 I .0269 5 I -054S 2 I.0837 2 1 .0014 9 I .0276 6 I-0553 3 I.0845 3 I. 0021 4 I .0284 7 1-0561 4 1-0853 4 1 .0028 1 I .0291 8 I.0569 5 I. 0861 5 1-0035 2 I .0298 9 I.0576 6 I .0870 6 I .0042 3 I . 0306 8 I.0584 7 I.0878 7 1 . 0049 4 I -0313 1 1.0592 8 I.0886 8 1 OOS5 5 I .0320 2 I 0599 9 1 . 0894 9 1 .0062 6 I .0328 3 . I.0607 I2 1 . 0902 I 1 .0069 7 1-0335 4 1. 0615 1 1 .0910 1 1 .0076 8 I.0342 5 I .0623 2 1 .0919 2 1 . 0083 9 1 0350 6 I . 0630 3 1 .0927 3 1 . 0090 5 I-0357 7 I.0638 4 I-0935 4 1 .0097 1 LO365 8 I . 0646 5 I.0943 5 1-0105 2 I.0372 9 I.0654 6 I.0952 6 I .0112 3 I.0379 9 I .0662 7 1 .0960 7 I .0119 4 I.0387 1 I.0670 8 I .0968 8 1 .0126 S I -0394 2 I .0677 9 I.0977 9 1-0133 6 I . 0402 3 I.0685 13 I.0985 2 1 .0140 7 I . 0409 4 1.0693 1 1.0993 1 I. 0147 8 I. 0417 5 I .0701 2 I .1002 2 1. 0154 9 I .0424 6 1 . 0709 3 I .IOIO 3 1.0161 6 I.0432 7 I .0717 4 1 .1018 4 1. 0168 1 I -0439 8 1.0725 5 I .1027 S 1. 0175 2 I.0447 9 I-0733 6 I • 1035 6 1-0183 3 I.0454 IO 1. 0741 7 1. 1043 7 1 .0190 4 I .0462 1 I.0749 8 1-1052 8 1. 0197 5 I .0469 2 I-0757 9 1 . 1060 9 1 .0204 6 I .0477 3 I .0765 14 1 . 1069 3 1 .0211 7 I . 0484 4 I.0773 1 1 .1077 1 1. 0218 8 I .0492 5 I .0781 2 I. 1086 2 1 .0226 9 I .0500 6 1 .0789 3 I . 1094 3 1 -°?33 7 I .0507 7 1.0797 4 1.1103 4 1 . 0240 1 1-0515 8 1 . 0805 5 I .nil 5 1 .0247 2 I .0522 9 1. 0813 6 1 .1120 .6 1-0255 3 I.0530 II 1 .0821 7 1 .1128 Reprinted from D. Van Nostrand's Chemical Annual, Fourth Issue. 1918. with the permis- sion of the editor and publishers. 460 APPENDIX EQUIVALENT OF DEGREES BAUME (AMERICAN STANDARD) AND SPECIFIC GRAVITY AT 6o° F. {Continued) Degrees Specific Degi •ees Specific Degi ■ees Specific Deg- rees Specific Baume Gravity . Bau me Gravity Bau me Gravity Bau me Gravity .8 1.1137 2 1. 1526 6 1. 1944 28 .0 1-2393 9 1. 1145 3 1 -1535 7 I-I954 .1 1 . 2404 15 1. "54 4 1 -1545 8 1 .1964 .2 1 .2414 1 1 .1162 5 1 -1554 9 I-I974 •3 1.2425 2 1 .1171 6 1-1563 24 1 . 1983 •4 1 . 2436 3 1.1180 7 I-I572 1 I 1993 •5 1 . 2446 4 1.1188 8 1.1581 2 1 .2003 .6 1-2457 5 1.1197 9 1.1591 3 1 .2013 • 7 1 . 2468 6 1. 1 206 20 1 .1600 4 I .2023 8 1.2478 7 1 .1214 1 . 1 .1609 5 1 . 2033 9 1 . 2489 8 1 .1223 2 1 .1619 .6 1 . 2043 29 1 . 2500 9 1 .1232 3 1. 1628 7 1-2053 1 1-2511 16 1 . 1 240 4 1-1637 8 I . 2063 2 1.2522 1 1 . 1 249 5 1-1647 9 1 . 2073 3 1.2532 2 1-1258 6 1-1656 25 1 . 2083 4 1-2543 3 1 .1267 7 1. 1665 1 1 . 2093 5 1-2554 4 1. 1275 8 1-1675 2 1 .2104 6 1.2565 5 1 .1284 9 1. 1684 3 I .2114 7 1.2576 6 1. 1293 21 1 .1694 4 1 .2124 8 1.2587 7 1 .1302 1 1-1703 5 1. 2134 9 1.2598 8 1.1310 2 1 .1712' 6 1. 2144 30 1 . 2609 9 1.1319 3 1 .1722 7 1. 2154 1 1 .2620 17 1. 1328 4 1.1731 8 1 .2164 2 1 .2631 1 1 -1337 5 i-i74i 9 1. 2175 3 1 . 2642 2 1-1346 6 1. 1750 26 I. 2185 4 1.2653 3 1 -1355 7 1 .1760 1 1. 2195 5 1 . 2664 4 1-1364 8 1 .1769 2 1 .2205 6 1.2675 S 1 -1373 9 I.I779 3 1 .2216 7 1 . 2686 6 1.1381 22 1. 1789 4 1 .2226 8 1.2697 7 i-i39° 1 1. 1798 5 I .2236 9 1.2708 8 1 1399 2 1. 1808 6 I.2247 31 1. 2719 9 1 .1408 3 1.1817 7 1.2257 1 1.2730 18 1.1417 4 1. 1827 8 1 .2267 2 1.2742 1 1 .1426 5 1-1837 9 1 .2278 3 1-2753 2 1 -1435 6 1. 1846 27 1.2288 4 1.2764 3 1 . 1444 7 1. 1856 1 1 .2299 5 1-2775 4 i-*453 8 1. 1866 2 1 . 2309 6 1.2787 S 1 .1462 9 1. 1876 3 1-2319 7 1 .2798 6 1-1472 23 1. 1885 4 1.2330 8 1 . 2809 7 1 .1481 1 1-1895 5 1 . 2340 9 1 .2821 8 1 . 1490 2 1 . 1905 6 1-2351 32 1.2832 9 1. 1499 3 1.1915 7 1 .2361 1 1 . 2843 19 1.1508 4 1. 1924 8 1.2372 2 1.2855 1 1-1517 5 1 1934 9 1 • 2383 3 1.2866 BAUME AND SPECIFIC GRAVITY 461 EQUIVALENT OF DEGREES BAUME (AMERICAN STANDARD) AND SPECIFIC GRAVITY AT 6o° F. (Continued) Degrees Specific Degi •ees Specific Degi ees Specific Degi •ees Specific Baume Gravity Bau me Gravity Bau me Gravity Bau me Gravity •4 I .2877 8 1. 3401 2 I.3969 6 1-4588 5 I . 2889 9 I-34I4 3 I-3983 7 1 .4602 6 I.2900 37 1.3426 4 I.3996 8 1. 4617 7 I .2912 1 I-3438 5 I .4010 9 I.4632 8 I.2923 2 1 -345i 6 I .4023 46 I .4646 9 1-2935 3 I-3463 7 I 4037 1 1 .4661 33 I . 2946 4 1-3476 8 I .4050 2 1 .4676 1 I.2958 5 1.3488 9 I .4064 3 1 .4691 2 I .2970 6 1. 3501 42 I .4078 4 1 .4706 3 I .2981 7 I-35I4 1 I .4091 5 1. 4721 4 I . 2993 8 I-3526 2 I .4105 6 I-4736 5 I.3004 9 1-3539 3 I .4119 7 1 -4751 6 I. 3016 38 1 -3551 4 I-4I33 8 1 .4766 7 I .3028 1 1-3564 5 I .4146 9 1. 4781 8 I.3040 2 1-3577 6 I .4160 47 1.4796 9 I-305r 3 1-3590 7 I-4I74 1 1 .4811 34 I.3063 4 1 .3602 8 I. 4188 2 1 .4826 1 I-3°75 5 1-3615 9 I .4202 3 1 .4841 2 I.3087 6 1.3628 43 I .4216 4 I-48S7 3 1.3098 7 1-3641 1 I.4230 5 1.4872 4 I.3110 8 I-3653 2 I.4244 6 1.4887 S I .3122 9 1.3666 3 I.4258 7 1 .4902 6 i-3i34 39 1.3679 4 I .4272 8 1 .4918 7 1-3146 1 1.3692 5 I .4286 9 1 -4933 8 1-3158 2 I-3705 6 I.4300 48 1.4948 9 1-3170 3 i-37i8 7 I-43I4 1 1.4964 35 1. 3182 4 1 -3731 8 I.4328 2 1.4979 1 I-3I94 5 1-3744 9 1-4342 3 1-4995 2 1 .3206 6 1-3757 44 I-4356 4 1. 5010 3 1-3218 7 1.3770 1 1 -4371 5 1 .5026 4 1.3230 8 I-3783 2 1-4385 6 1-5041 S 1.3242 9 1.3796 3 1-4399 7 I-5057 6 1.3254 40 1-3810 4 I-44I4 8 I-5073 7 1 .3266 1 1-3823 5 1 .4428 9 1.5088 8 1.3278 2 1.3836 6 1 .4442 49 1-5104 1 9 1. 3291 3 1.3849 7 1-4457 1 1. 5120 36 1-3303 4 1.3862 8 I-447I 2 I-5I36 1 I-33I5 5 1.3876 9 1 . 4486 3 I-5I52 2 1-3327 6 1.3889 45 1.4500 4 1-5167 3 I-3329 7 1.3902 1 I-45I5 5 i-5i83 4 1-3352 8 1. 3916 2 1.4529 6 I-5I99 5 1-3364 9 1.3929 3 1-4544 7 I-52I5 6 1-3376 41 1-3942 4 I.4558 8 1-5231 7 I-3389 1 1-3956 5 1-4573 9 1-5247 APPENDIX HYDROCHLORIC ACID Be. ° Sp. gr. Tw. ° Percent HC1 Be. ° Sp. gr. Tw. c Percent HC1 Be. ° Sp. gr. Tw. ° Percent HC1 I.OO 1.0069 1.38 1.40 16.0 1 . 1240 24.80 24-57 20.8 1-1675 3350 32.93 2.00 1. 0140 2.80 2.82 16. 1 1 . 1248 24 96 24-73 20.9 1 . 1684 33 68 33 12 3-00 1.0211 4.22 425 16.2 1 . 1256 25 12 24.90 21.0 1 . 1694 33 88 33 31 4.O0 1.0284 5-68 5-69 16.3 1 . 1265 25 3o 25.06 21. 1 1 . 1703 34 06 33 So 500 10357 7-14 7-15 16.4 1 - 1274 25 48 25-23 21.2 I-I7I3 34 26 33 69 5-25 1.0375 7-5o 7-52 16. 5 1 . 1283 25 66 25-39 21.3 1 . 1722 34 44 33 88 5.5o 10394 7.88 7.89 16.6 1 . 1292 25 84 25 .56 21.4 1 • 1732 34 64 34 07 5-75 1. 0413 8.26 8.26 16.7 1 - 1301 26 02 25.72 21.5 1 . 1741 34 82 34 26 6.00 1.0432 8.64 8.64 16.8 1.1310 26 20 25 89 21.6 1 ■ i75i 35 02 34 45 6.25 1.0450 9.00 9.02 16.9 1.1319 26 38 26.05 21.7 1 . 1760 35 20 34 64 6.50 1.0469 9-38 9.40 17.0 1 . 1328 26 56 26.22 21.8 1. 1770 35 40 34 83 6.75 1.0488 9.76 9-78 17. 1 1 • 1336 26 72 26.39 21.9 1 1779 35 58 35 02 7.00 1.0507 10.14 10.17 17.2 1 - 1345 26 90 26.56 22.0 1 ■ 1789 35 78 35 21 725 1.0526 10.52 10.55 17-3 1 ■ 1354 27 08 26.73 22.1 1. 1798 35 96 35 40 7-5o 10545 10.90 10.94 17-4 1 - 1363 27 26 26.90 22.2 1 . 1808 36 16 35 59 7-75 1.0564 11.28 11.32 17-5 1 ■ 1372 27 44 27.07 22.3 1.1817 36 34 35 78 8.00 1.0584 11.68 11. 71 17.6 1 . 1381 27 62 2724 22.4 1 . 1827 36 54 35 97 8.25 1.0603 12.06 12.09 17-7 1 1390 27 80 27.41 22.5 1. 1836 36 72 36 16 8.50 1.0623 12.46 12.48 17.8 1 ■ 1399 27 98 27-58 22.6 1 . 1846 36 92 36 35 8.75 1.0642 12.84 12.87 17-9 1 . 1408 28 16 27-75 22.7 1 . 1856 37 12 36 54 9.00 1.0662 13-24 13 26 18.0 1.1417 28 34 27.92 22.8 1 . 1866 37 32 36 73 9-25 1. 0681 13-62 13 .65 18. 1 1 . 1426 28 52 28.09 22.9 1187S 37 5o 36 93 9So 1. 0701 14.02 14.04 18.2 1 ■ 1435 28 7o 28.26 230 1. 1885 37 70 37 14 9-75 1. 0721 14.42 14-43 18.3 1 • 1444 28 88 28.44 231 1. 1895 37 90 37 36 10.00 1. 0741 14.82 14-83 18.4 1 1453 29 06 28.61 232 1 . 1904 38 oS 37 58 10.25 1. 0761 15-22 15-22 18.5 1 . 1462 29 24 28.78 233 1.1914 38 28 37 80 10.50 1. 0781 15- 62 15-62 18.6 1.1471 29 42 28.95 23-4 1 ■ 1924 38 48 38 03 10.75 1.0S01 16.02 I6.OI 18.7 1 . 1480 29 60 29 13 23-5 1 ■ 1934 38 68 38 26 11.00 1. 0821 16.42 16.41 18.8 1 . 1489 29 78 29.30 236 1 . 1944 38 88 38 49 11.2s 1. 0841 16.82 16.81 18.9 1 . 1498 29 96 29.48 23-7 1 ■ 1953 39 06 38 72 11.50 1. 0861 17.22 17.21 19.0 1 . 1508 30 16 29.65 23.8 1 ■ 1963 39 26 38 95 11-75 1. 0881 17,62 I7.6I 19. 1 1.1517 30 34 29.83 23-9 1 • 1973 39 46 39 18 12.00 1.0902 18.04 I8.OI 19.2 1 . 1526 30 52 30.00 24.0 1 • 1983 39 66 39 41 12.25 1.0922 18.44 I8.4I 19-3 1 1535 30 70 30.18 24.I 1 ■ 1993 39 86 39 64 12.50 1 0943 18.86 18.82 19.4 1 • 1544 30 88 30-3S 24.2 1.2003 40 06 39 86 12.75 1.0964 19.28 19.22 19-5 1 1554 31 08 30.53 24-3 1 ■ 2013 40 26 40 09 13 00 10985 19.70 19.63 19.6 1 - 1563 31 26 30.71 24-4 1 . 2023 40 46 40 32 13 25 1 . 1006 20.12 20.04 19-7 1 . 1572 31 44 3090 24-5 1 ■ 2033 40 66 40 55 13.50 1 . 1027 20.54 20.4S 19.8 1 . 1581 31 62 3i-o8 24.6 1 • 2043 40 86 40 78 13.75 1 . 1048 20.96 20.86 19.9 1 ■ 1590 31 80 31 .27 24-7 1 ■ 2053 4i 06 41 OI 14.00 1 . 1069 21.38 21.27 20.0 1. 1600 32 00 31-45 24.8 1 . 2063 41 26 4i 24 14- 25 1 . 1090 21.80 21.68 20.1 1 . 1609 32 18 31.64 24.9 1 . 2073 4i 46 4i 48 l4-5o 1. mi 22.22 22.09 20.2 1.1619 32 38 3182 25.0 1 . 2083 4i 66 41 72 14-75 1.1132 22.64 22.50 20.3 1 . 1628 32 56 32.01 25 1 1 . 2093 4i 86 41 99 15 00 1.1154 2308 22.92 20.4 1 . 1637 32 74 32.19 25-2 1. 2103 42 06 42 30 1525 1.1176 23-52 23 33 20 5 1. 1647 32 94 3238 25-3 1.2114 42 28 42 64 15-50 1.1197 23.94 23-75 20.6 1 . 1656 33 12 32.56 25-4 1. 2124 42 48 43 01 1575 1.1219 24.38 24.16 20.7 1. 1666 3332 32-75 255 1. 2134 42.68 43 -40 Specific Gravity determinations were made at 6o° F., compared with water at 6o° F. From the Specific Gravities, the corresponding degrees Baume were calculated by the follow- ing formula: Baume = 145 - g^ Baume Hydrometers for use with this table must be graduated by the above formula which formula should always be printed on the scale. Atomic weights from F. W. Clarke's table of 1901. O = 16. ALLOWANCE FOR TEMPERATURE 10-15° Be. 15-22° Be. 22-25° Be. 1/40° Be. or .0002 Sp. Gr. for 1° F. 1/30° Be. or .0003 " " " 1° F. 1/28 Be. or .00035 i° F. Authority — W. C. Ferguson. The above table was prepared under the supervision of the Manufacturing Chemists Associa- tion of the United States and adopted by the Association as standard for United States practice. Reprints of each table may be obtained from the Secretary of the Association, 84 State St., Boston, NITRIC ACID 463 NITRIC ACID Be. ° Sp. Gr. Tw. ° Percent HN0 3 Be. ° Sp. Gr. Tw. ° Percent HNO3 Be. ° Sp. Gr. Tw. ° Percent HNO5 10.00 1. 0741 14.82 12.86 23.00 1. 1885 3770 30.49 36.00 1-3303 66.06 52.30 IO. 25 1. 0761 15.22 13 18 23 25 1.1910 38.20 30 86 36 25 1-3334 66.68 S2.81 IO.SO 1. 0781 15.62 13 49 23 5o 1 • 1934 38.68 31 21 36 So 1.3364 67.28 53-32 IO-7S 1. 0801 16.02 13 81 23 IS 1 ■ 1959 39-18 31 58 36 75 1-3395 67.90 53.84 II.OO 1. 0821 16.42 14 13 24 00 1 . 1983 39-66 3i 94 37 00 1.3426 68.52 54-36 II. 2S 1. 0841 16.82 14 44 24 25 1 . 2008 40.16 32 3i 37 25 1. 3457 69.14 54-89 11. So 1. 0861 17.22 14 76" 24 So 1 ■ 2033 40.66 32 68 37 50 1.3488 69.76 55-43 11.75 1. 0881 17.62 IS 07 24 75 1.2058 41.16 33 05 37 75 1.3520 70.40 55-97 12.00 1.0902 18.04 15 41 25 00 1 . 2083 41.66 33 42 38 00 1 -3551 71.02 56.52 12.25 1.0922 18.44 15 72 25 25 1 . 2109 42.18 33 80 38 25 I-3S83 71.66 5708 I2.SO 1.0943 18.86 16 OS 25 50 1 ■ 2134 42.68 34 17 38 50 1.3615 7230 57-65 12.75 1.0964 19.28 16 39 25 75 1 . 2160 43.20 34 56 38 75 1-3647 72.94 58.23 13 OO 1.0985 19.70 16 72 26 00 1 . 2185 43-70 34 94 39 00 13679 73.58 58.82 13 25 1. 1006 20.12 17 OS 2fi 25 1.2211 44.22 35 33 39 25 1.3712 74.24 59-43 13 So 1 . 1027 20.54 17 38 26 50 1 . 2236 44.72 35 70 39 50 1-3744 74-88 60.06 13-75 1 . 1048 20.96 17 71 26 75 1 . 2262 45-24 36 09 39 75 1.3777 75 54 60.71 14.00 1 . 1069 21 38 18 04 27 00 1 . 2288 45-76 36 48 40 00 1. 3810 76.20 61.38 14-25 1 . 1090 21.80 18 37 27 25 1 . 2314 46.28 36 87 40 25 1.3843 76.86 62.07 14- So inn 22.22 18 70 27 50 1 . 2340 46.80 37 26 4o 50 1.3876 77.52 62.77 14-75 1.1132 22.64 19 02 27 73 1 . 2367 47-34 37 67 40 75 I-3909 78.18 63.48 iS-oo 1.1154 23 08 19 36 28 00 12393 47-86 38 06 41 00 1 ■ 3942 78.84 64.20 15.25 1.1176 23-52 19 70 28 25 1 . 2420 48.40 38 46 41 25 1-3976 79-52 6493 IS So 1.1197 23-94 20 02 28 50 1 . 2446 48.92 38 85 41 50 1 . 4010 80.20 6567 15-75 1.1219 24.38 20 36 28 75 1 ■ 2473 49 -46 39 25 41 75 1.4044 80.88 66.42 16.00 1 . 1240 24.80 20 69 29 00 1 . 2500 50.00 39 66 42 00 1 . 4078 81.56 67.18 16.25 1 . 1262 25.24 21 0.5 29 25 1 . 2527 50.54 40 06 42 25 1.4112 82.24 679S 16.50 1 . 1284 25.68 21 36 29 So 1 ■ 2554 51 08 40 47 42 So 1. 4146 82.92 68.73 16.75 1 . 1306 26.12 21 70 29 75 1 . 2582 51-64 40 89 42 75 1.4181 83.62 6952 17.00 1. 1328 26.56 22 04 30 00 1 . 2609 5218 41 30 43 00 1. 4216 84.32 7033 17 -25 1 - 1350 27.00 22 38 30 25 1 • 2637 52-74 41 72 43 25 1. 4251 85.02 71. IS 17.50 1 • 1373 27.46 22 74 30 5o 1 . 2664 53-28 42 14 43 50 1.4286 8S.72 71.98 17-75 1 • 1395 27.90 23 08 30 75 1 . 2692 53-84 42 58 43 75 1. 4321 86.42 72.82 18.00 1 . 1417 28.34 23 42 31 00 1. 2719 54.38 43 00 44 00 I.43S6 87.12 73 67 18.25 1 . 1440 28.80 23 77 31 25 1 . 2747 5494 43 44 44 25 1.4392 87.84 74 S3 18.50 1 . 1462 2924 24 11 31 50 1 ■ 2775 5550 43 89 44 50 1 . 4428 88.56 75-40 18.75 1 . 1485 29.70 24 47 31 75 1 . 2804 56.08 44 34 44 75 1.4464 89.28 76.28 19.00 1 . 1508 30.16 24 82 32 00 1 . 2832 56.64 44 78 45 00 1.4500 90.00 77.17 19-25 1 • 1531 30.62 25 18 32 25 1 . 2861 57-22 45 24 45 25 1.4536 90.72 78.07 19- 5o 1 -1554 3108 25 S3 32 50 1 . 2889 .57.78 45 68 4.5 50 1-4573 91.46 79 03 19-75 1 • 1577 31-54 25 88 32 75 1. 2918 58.36 46 14 45 75 1 . 4610 92.20 80.04 20.00 1. 1600 3200 26 24 33 00 1 . 2946 S8.92 46 58 46 00 1 . 4646 92.92 81.08 20.25 1 . 1624 32.48 26 61 33 25 1 ■ 2975 59 -So 47 04 46 25 1 . 4684 93-68 82.18 20.50 1 . 1647 32.94 26 96 33 5o 13004 6a. 08 47 49 46 50 1. 4721 94.42 8.3 33 20.75 1.1671 33-42 27 33 33 75 1.3034 60.68 47 95 46 75 1-4758 95.16 84.48 21.00 1 . 1694 33-88 27 67 34 00 13063 61.26 48 42 47 00 1.4796 95.92 85.70 21.25 1.1718 3436 28 02 34 25 13093 61.86 48 90 47 25 1.4834 96.68 86.98 21.50 1.1741 34-82 28 36 34 50 1. 3122 62.44 49 35 47 50 1 . 4872 97-44 88.32 21.7S 1.176S 35-30 28 72 34 75 1. 3152 63.04 49 83 47 75 1 . 4910 98.20 89.76 22.00 I-I789 3578 29 07 35 00 1. 3182 63.64 50 32 4« 00 1.4948 98.96 91 35 22.25 1 . 1813 3626 29 43 35 25 1. 3212 64.24 5o 81 48 25 1.4987 99-74 93 13 22.50 1 • 1837 36.74 29 78 35 5o 1 3242 6484 5i 30 48 5o 1 . 5026 100 . 52 95-11 22.75 1 . 1861 37.22 30 14 35 75 1-3273 65.46 SI 80 Specific Gravity determinations were made at 6o° F., compared with water at 6o° P. From the Specific Gravities, the corresponding degrees Baume were calculated by the follow- ing formula: ' , 145 Baume=I45 -Sp7Gr: Baume Hydrometers for use with this table must be graduated by the above formula, which formula should always be printed on the scale. Atomic weights from F. W. Clarke's table of 1901. O = 16. ALLOWANCE FOR TEMPERATURE: At io°-20° Be. — 1/30 Be. or .00029 Sp. Gr. = 1° F. 20°-30° Be. — 1/23 Be. or .00044 " " = i° F. 3o°-40° Be. — 1/20 Be. or .00060 " " = 1° F. 40°-48.5° Be. — 1/17° Be. or .00084 " " = i° F. Authority — W. C. Ferguson. The above table was prepared under the supervision of the Manufacturing Chemists Associa- tion of the United States and adopted by the Association as standard for United States practice. Reprints of each table may be obtained from the Secretary of the Association, 84 State St., Boston. 464 APPENDIX AQUA AMMONIA Be. ° Si ). Gr. Per cent NH 3 Be. ° Si . Gr. Per cent NH 3 Be. ° Si . Gr. Per cent NH 3 10.00 1 0000 .00 16.50 9556 11. 18 23.00 9150 23-52 . 10.25 9982 • 40 16.75 954o 11.64 23-25 9135 24.01 10.50 9964 .80 17.00 9524 12.10 2350 9121 24-50 10.75 9947 1. 21 17-25 95o8 12.56 23.75 9106 2499 11.00 9929 1.62 17-So 9492 13-02 24.00 9091 35-48 11.25 9912 2.04 I7-7S 9475 13-49 24.25 9076 2597 11.50 9894 2.46 18.00 9459 13-96 24.50 9061 26.46 11.75 9876 2.88 18.25 9444 14-43 24-75 9047 26.95 12.00 9859 3-30 18.50 9428 14.90 35 00 9032 27.44 12.25 9842 3-73 18.75 9412 15-37 35-25 9018 27.93 12.50 9825 4.16 19.00 9396 15.84 35.50 9003 28.42 I27S 9807 4-59 19.25 938o 16.32 35-75 8989 28.91 1300 979o 502 i9-5o 936S 16.80 36.00 8974 29.40 13-25 9773 5-45 19 -75 9349 17.28 26.25 8960 29.89 13 5o 9756 5-88 20.00 9333 17.76 26.50 8946 30.38 13- 75 9739 6.31 20.25 93i8 18.24 36.75 8931 30.87 14.00 9722 6.74 20.50 9302 18.72 37.00 8917 3136 1425 9705 7.17 20.75 9287 19.20 37.25 8903 31.85 14 So 9689 7.61 21. OO 9272 19.68 37.50 8889 32.34 14-75 9672 8.05 21.25 9256 20.16 37-75 8875 32.83 15.00 9655 8.49 2I.50 9241 20.64 38.00 8861 33-32 15 -25 9639 8.93 21.75 9226 21.12 38.25 8847 33-81 1550 9622 9.38 22.00 921 1 21.60 38.50 8833 34 -3o 15-75 9605 983 22.25 9195 22.08 28-75 8819 34-79 16.00 9589 10.28 22.50 9180 22.56 29.00 8805 35-28 16.25 9573 10.73 22.75 9165 23.04 Specific Gravity determinations were made at 6o° F., compared with water at 6o° F. From the Specific Gravities, the corresponding degrees Baume were calculated by the follow- ing formula: -o ' J 40 Baume = 5 — ■= 130. bp. Gr. Baume Hydrometers for use with this table must be graduated by the above formula, which formula should always be printed on the scale. Atomic weights from F. W. Clarke's table of 1901. O = 16. ALLOWANCE FOR TEMPERATURE The coefficient of expansion for Ammonia Solutions, varying with the temperature, correction must be applied according to the following table: Corrections to be added for each Corrections to be subtracted for each degree degree below 6o° F. above 6o° F. Degrees Baume 40° F. 50° F. 70° F. 8o°F. 90° F. ioo° F. 14 .015° Be. .017° Be. .020° Be. .022° Be. .024° Be. .026° Be. 16 .021 " 023 " .026 " .028 " .030 " .032 " 18 .027 ' .029 ' 031 ' • 033 ' ■035 ' 037 ' 20° ■033 " .036 ' ■ 037 " .038 " .040 ' .042 ' 22° •039 ' .042 ' ■043 " • 045 ' 047 ' 26° 053 " • 057 " • 057 " • 059 " Authority — W. C. Ferguson. The above table was prepared under the supervision of the Manufacturing Chemists Asso- ciation of the United States and adopted by the Association as standard for United States practice. Reprints of each table may be obtained from the Secretary of the Association, 84 State St., Boston. SULPHURIC ACID en u 03 OOT so An > . •a ° a M g-s S'-S.S • ■3 ° el " O-S p_ °F. O I 0000 0.0 O.OO 62.37 O.OO 0.00 32.0 37 I . 3426 68.5 43-99 8374 47.20 39-53 -60 I I 0069 1 4 I.02 62 80 I 09 0.68 31 2 38 1 -3551 71.0 45 35 84.52 48.66 4I-I3 -53 2 I 0140 2 8 2.o8 63 24 2 23 1. 41 30.5 39 I.3679 73-6 46 72 85.32 50.I3 42.77 -47 3 I 0211 4 2 3.13 63 69 3 36 2.14 29.8 40 I. 3810 76.2 48 10 86.13 5I.6I 44-45 -41 4 i 0284 5 7 4.21 64 14 4 52 2.90 28.9 41 1.3942 78. 8 49 47 86.96 53 08 46.16 -35 5 i 0357 7 1 5.28 64 60 5 67 3-66 28.1 42 1 . 4078 81.6 50 87 87.80 54-58 47-92 —31 6 i 0432 8 6 6.37 65 06 6 84 4-45 27.2 43 1. 4216 84.3 52 26 88.67 56.07 49.72 -27 7 i 0507 10 1 7-45 65 53 7 99 5.24 26.3 44 1. 43S6 87.1 53 66 89.54 57-58 5156 —23 8 i 0584 11 7 8.55 66 01 9 17 6.06 25.1 45 1.4500 90.0 55 07 90.44 59 09 5344 —20 9 i 0662 13 2 9.66 66 5o 10 37 6.89 24.O 46 1 . 4646 92.9 56 48 9135 60.60 55.36 -14 IO I 0741 14 8 I0.77 66 99 11 56 7-74 22.8 47 1.4795 95-9 57 90 92.28 62 . 13 57-33 — 15 II i 0821 16 4 II.89 67 49 12 76 8.61 21.5 48 1 ■ 4948 990 59 32 9323 63.65 59-34 -18 12 I 0902 18 13 OI 68 00 13 96 9-49 20.0 49 1. 5104 102. 1 60 75 94.20 65.18 61.40 —22 13 I 0985 19 7 14-13 68 5i 15 16 io.39 18.3 50 1.5263 105.3 62 18 95.20 66.72 63.52 -27 14 I 1069 21 4 15-25 69 04 16 36 11.30 16.6 51 1 ■ 5426 10S.5 63 66 96.21 68.31 65.72 —33 IS I 1 154 23 I 16.38 69 57 17 5S 12.23 14- 7 52 1. 5591 in. 8 65 13 97.24 69.89 67.96 -39 16 i 1240 24 8 17-53 70 10 18 81 13- 19 12.6 S3 l.S76i 115- 2 66 63 98.30 7i.5o 70.28 -49 17 i 1328 26 6 18.71 70 65 20 oS 14.18 10.2 54 1-5934 118. 7 68 13 99 -38 73 11 72.66 -59 18 i 1417 28 3 19.89 71 21 21 34 1520 7-7 55 1.6111 122. 2 69 65 100.48 74-74 75- 10 19 i 1508 30 2 21.07 71 78 22 61 16.23 4.8 56 1 . 6292 125.8 71 17 101.61 76.37 77.60 £0 20 I 1600 32 22.25 72 35 23 87 17.27 + 1.6 57 1 . 6477 129- 5 72 75 102 . 77 78.07 80.23 ■* 21 I 1694 33 9 2343 72 94 25 14 18.34 - 1.8 58 1 . 6667 133. 3 74 36 10395 79-79 82.95 tu 1 PP ' 22 I 1789 35 8 24.61 73 53 26 41 19.42 - 6.0 59 1.6860 137- 2 75 39 105.16 81.54 85-75 - 7 23 I 1885 37 7 25.81 74 13 27 69 20.53 — 11 60 I- 7059 141. 2 77 67 106 . 40 8335 88.68 +12. 6 24 I 1983 39 7 2703 74 74 29 OO 21.68 -16 61 1 . 7262 145- 2 79 43 107.66 8S.23 91.76 27.3 25 I 2083 41 7 28.28 75 36 30 34 22.87 -23 62 1.7470 149. 4 81 3o 108.96 87.24 95.06 39- 1 26 I 2185 43 7 29-53 76 00 3i 69 24.08 —30 63 1.7683 153-7 83 34 110.29 89-43 98.63 46. 1 27 I 2288 45 8 30.79 76 64 33 04 2532 -39 64 1 . 7901 158.0 85 66 111.6s 91.92 102 . 63 46. 4 28 I 2393 47 9 3205 77 3° 34 39 26.58 -49 64i 1-7957 159- 1 86 33 112.00 92.64 103.75 43- 6 29 I 2500 5o 33 33 77 96 35 76 27.88 -61 ! '-i' 1 . 8012 160. 2 87 04 112.34 93.4o 104.93 41- 1 30 I 2609 52 2 34.63 78 64 37 16 29.22 -74 64! 1.8068 161. 4 87 81 112.69 94.23 106 . 19 37-9 31 I 2719 54 4 35-93 79 33 38 55 30.58 -82 65 I. 8125 162.5 88 65 113-05 95- 13 107.54 33- 1 32 I 2832 5<"> 6 37.26 80 03 39 98 32.00 -96 65! 1. 8182 163.6 89 55 113.40 96.10 108.97 24. 6 33 1 2946 58 9 38.58 80 74 41 40 33 42 -97 65; 1 ■ 8239 164.8 90 60 113.76 97.22 110.60 1.3-4 34 1 3063 6l 3 39.92 81 47 42 83 3490 -91 '">5] 1 . 8297 165.9 91 80 114. 12 9851 112.42 — I 35 1 3182 63 6 4127 82 22 44 28 36.41 -81 66 1 • 8354 167. 1 93 19 114.47 100.00 114.47 -29 36 1 3303 66.1 42.63 82.97 45-74 3795 -70 * Calculated from Pickering's results, Journal of London Chemical Society, vol. 57, p. 363. F. Approximate Boiling Points 50 Be., 295 F. 6o° ' ' 386 ' 6i° ' ' 400 ' 62° ' ' 415° ' 6.3 U ' ' 432° ' 6 4 u ' ' 451° " 65° ' ' 485° " 66° ' ' 538° ' Allowance for Temperature At 10° Be., .029° Be. or .00023 Sp. Gr. = 1° 00034 00039 00041 00045 00053 00057 00054 Specific Gravity determinations were made at 6o° F., compared with water at 60° F. From the Specific Gravities, the corresponding degrees Baume were calculated by the fol- lowing formula: " 20° " • 036° " 30° ' ■035 " 4O " .031° " 50° " .028° " 6o° " .026° " 63 " .026° " 66° " ■0235 Baume = 145 — 145 Sp. Gr. Baume Hydrometers for use with this table must be graduated by the above formula, which formula should always be printed on the scale. 66° Baume = Sp. Gr. 1.8354. 1 cu. ft. water at 60° F. weighs 62.37 lbs. av. Atomic weights from F. W. Clarke's table of 1901. O = 16. H 2 S0 4 = 100 per cent. H,S0 4 O. V. 60 ° O. V. 93.19 100.00 119.98 60° 77.67 83.35 100.00 So° 62.18 66.72 80.06 Acids stronger than 66° Be. should have their percentage compositions determined by chemi- cal analysis. Authorities. — W. C. Ferguson; H. P. Talbot. _ The above table was prepared under the supervision of the Manufacturing Chemists Asso- ciation of the United States and adopted by the Association as standard for United States practice. Reprints of each table may be obtained from the Secretary of the Association, 84 State St., Boston. (46S) 466 APPENDIX COMPARISON OF METRIC AND CUSTOMARY UNITS FROM i TO 9 1. Length Inches Millimeters Feet Meters Yards Meters (in.) (mm.) (ft.) (m.) (yd.) (m.) 0.039 37=1 1=0.304 801 1=0.914 402 0.078 74=2 2=0.609 601 2=1.828804 0.118 ii=3 3=0.914 402 3=2.743205 o.lS7 48=4 4 = 1.219 202 4=3-657607 0.19685=5 5 = 1.524003 5=4.572009 0.236 22=6 6=1.828804 6=5.486411 0.275 59 =1 7=2.133604 7=6.400 813 0.31496=8 8=2.438 405 8=7.315215 o.354 33=9 9=2.743205 9=8.229616 1= 25.4001 3.28083=1 1.093 6n=l 2 = 50.8001 6.561 67=2 2.187 222=2 3= 76.2002 9.84250=3 3.280833=3 4 = 101.6002 13- 123 33=4 4.374 444=4 5 = 127.0003 16.404 17=5 5.468056=5 6 = 152.4003 19.68500=6 6.561 667=6 7 = 177.8004 22.96583=7 7.655278=7 8=203.2004 26.246 67=8 8.748889=8 9 = 228.6005 29 527 50=9 9.842 500=9 Area Square Square inches centimeters (sq. in.) (cm. 2 ) Square Square feet meters (sq. ft.) (m.2) Square Square yards meters (sq. yd.) (m.2) 0.15500=1 1=0.092 90 1=0.8361 0.31000=2 2=0.18581 2=1.6723 0.465 oo=3 3=0.278 71 3=2.5084 0.62000=4 4=0.371 61 4 = 3-3445 0.775 oo=5 5=0.46452 5=4.1807 0.93000=6 6=0.55742 6 = 5.0168 1.08500=7 7=0.65032 7=5.8529 1.24000=8 8=0.74323 8=6.6890 i.395oo=9 9=0.836 13 9 = 7.5252 1= 6.452 10.764=1 1.1960=1 2 = 12.903 21.528=2 2.3920=2 3 = 19-355 32.292=3 3.5880=3 4=25.807 43.055=4 4.7839=4 5=32.258 53 819=5 5.9799=5 6=38.710 64.583=6 7.1759=6 7=45.161 75-347=7 8.3719=7 8 = 51.613 86.111=8 9.5679=8 9=58.065 96.875=9 10.7639=9 From Circular No. 47, U. S. Bureau of Standards. 3. VOLUME Cubic Cubic Cubic Cubic Cubic Cubic Cubic feet meters yards meters inches Liters feet Liters (cm. 3 ) (cu. ft.) (m.3) (cu. yd.) (m.») (cu. in.) (1.) (cu. ft.) (1.) 0.061 02=1 1=0.028 317 1=0.7646 1=0.016 386 7 1= 28.316 0.12205=2 2=0.056 634 2=1.5291 2=0.032 773 4 2= 56.633 0.18307=3 3=0.084 95i 3=2.2937 3=0.049 160 2 3= 84.949 0.24409=4 4=0.113 268 4 =3 0582 4=0.065 546 9 4=113.265 0.305 12 = 5 5=0.141 585 5=3.8228 5=o.o8i 933 6 5 = 141.581 0.366 14=6 6=0.169 9°2 6=4.5874 6=0.098 320 3 6 = 169.898 0.427 i6=7 7=0.198 219 7=5.3519 7=0.114 707 7=198.214 0.488 19=8 8=0.226 536 8=6.1165 8=0.1310938 8 = 226.530 0.549 21=9 9=0.254853 9=6.88io 9=0.147 480 5 9=254.846 1= 16.3872 35 314=1 1.3079=1 61.025 = 1 0.035315=1 2= 32.7743 70.629=2 2.6159=2 122.050=2 0.070 631=2 3= 491615 ioS943=3 39238=3 183.075=3 0.105 946=3 4 = 65.5486 141.258=4 5-2318=4 244.100=4 0.141 262=4 5= 81.9358 176.572=5 6-5397=5 305.125=5 0.176577=6 6= 98-3230 211.887 = 6 7.8477=6 366.150=6 0.211 892=6 7=114.7101 247.201=7 9- 1556=7 427.175=7 0.247 2o8=7 8=131.0973 282.516=8 10.4635=8 488.200=8 0.282 523=8 9 = 147.4845 317.830 = 9 11.7715 = 9 549.225=9 o.3i7 839=9 4. Capacity Liquid Measure Dry Measure U. S. Milli- U. S. U. S. U. S. fluid liters liquid Liters liquid Liters U. S. Liters gallons (gal.) (1.) ounces (fl. oz.) (ml.) pints (pt.) (1.) quarts (qt.) (1.) quarts (qt.) (1.) 0.033815=1 1=0.473 17 1=0.94633 0.264 i8=l 0.9081 = 1 0.067 629=2 2=0.94633 2 = 1.89267 0.52836=2 1.8162=2 0.101 444=3 3 = 1.41950 3 = 2.83900 o.792 53=3 2.7243=3 o.i35 259=4 4 = 1.892 67 4=3-785 33 1.05671=4 3.6324=4 0.169 074=5 5=2.36583 5=4.731 67 1.32089=5 4-54oS=5 0.202888=6 6=2.83900 6 = 5.67800 1-58507=6 5.4486=6 0.236 703=7 7=3.312 17 7 = 6.62433 1.84924=7 6.3567=7 0.270 518=8 8=3-785 33 8 = 7-57o66 2. 113 42=8 7.2648=8 O-304 333=9 9 = 4-25850 9=8.51700 2.37760=9 8.1729=9 1= 29.573 2.1134 = 1 1.056 71 = 1 1= 3-785 33 1 = 1.1012 2= 59-146 4.2268=2 2. 113 42=2 2= 7.57066 2 = 2.2024 3= 88.719 6.3403 = 3 3.170 13=3 3 = 11.35600 3=3-3036 4=118.292 8-4537=4 4.226 84=4 4=15.141 33 4=4.4048 5 = 147.865 10.5671=5 5283 55=5 5 = 18.92666 6 = 5.5060 6=177.437 12.6805=6 6.340 26 = 6 6 = 22.711 99 6=6.6072 7 = 207.010 14-7939=7 7396 97=7 7=26.49733 7 = 7.7084 8=236.583 16.9074=8 8.453 68=8 8 = 30.282 66 8=8.8096 9=266.156 19.0208=9 95io39=9 9=34.06799 9=9.9108 From Circular No. 47, U. S. Bureau of Standards. (467) 468 APPENDIX 5. Mass Troy ounces Grams (oz. t.) (g.) Avoirdu- pois Grams ounces (oz. av.) (g.) Avoirdu- Kilo- pois grams pounds (lb. av.) (kg.) 0.032 151 = 1 0.035 274=1 l=o.453 59 0.064301=2 0.070548=2 2=0.907 18 0.096 452=3 0.105 822=3 3=1.36078 0.128 603=4 0.141096=4 4=1.814 37 0.160 754=5 0.176 370=5 5=2.26796 0.192 904=6 0.211 644=6 6=2.721 55 0.225055=7 0.246 918=7 7=3.175 15 0.257 206=8 0.282 192=8 8=3.628 74 0.289 357=9 0.317 466=9 , 9=4.08233 1= 31-103 1= 28.350 2.204 62 =1 2= 62.207 2= 56.699 4.409 24 =2 3= 93 -3io 3= 85.049 6.61387 =3 4=124.414 4=113.398 8.81849 =4 5 = 155-517 5=141.748 11.023 ii=5 6 = 186.621 6=170.097 13227 73=6 7 = 217.724 7=198.447 15432 36=7 8=248.828 8=226.796 17.63698=8 9=279-931 9=255.146 19.841 6o=9 From Circular No. 47, U. S. Bureau of Standards. INDEX Abietic acid, 265 Absorbency of papers, 410 Acid colors, 320 in paper, determination of, 422 sulphite cooking liquor, 160 effect on metals, 178 strength of, 176, 184, 190 systems of making, 169 testing, 209 thiosulphuric acid in, 166 Acids, action on cellulose, 8 Adansonia, 40 Adhesives for coating, 328, 331, 333, 343; 345 Albumen, 343 Alcohol from soda cooks, 117 sulphate process, 147 sulphite process, 198, 205 Alder, 62 Alkali cellulose, 20 Alum, 269, 277 amount necessary in sizing 270, 280 cake, composition, 278 composition, 279 decomposition by celluk>se, 7 effect on glue, 259 free acid in, 280 iron in, 279 manufacture, 278 method of adding to beaters, 281 substitutes, 269 testing, 282, 286 use in clarifying water, 359 Aluminum resinate, 269 sulphate (table), 444, 445 Ammonia (table), 464 Analyses, see Testing AniHne dyes, 313 absorption by fillers, 314 adding to beaters, 317 classification of, 317 dissolving, 317 mordants for, 316 Annatto, 313 Antichlors, 244 Antifroth oils, 329, 349 Appendix, 444 Asbestine, 301 Ash, determination in papers, 414 in fibrous materials, 415 qualitative analysis of, 416 Ashcroft tester, 406 Asparagus for pulp, 44 Aspen, 61 B Bache-Wiig process for ground wood, 222 Backwater, use in bleaching, 239 Bagasse, 43 Balm of Gilead, 61 Bamboo, 43, 86 compositon of, 87 pectose in, 93 treatment recommended, 87 yield of stems per acre, 86 fibre, 87 Bark, 55 in soda process, 95 sulphite process, 158 Barker acid system, 173 Barkers, 158 Barytes, 346 Basic colors, 318 Basswood, 65 469 47° INDEX Bast fibres, 37 Baume and specific gravity (table), 459- 461 Beating test for wood pulps, 372 Beech, 62 Binder from sulphite waste liquor, 201 Birches, 61 Bisulphites (see also acid), 160 losses in preparing, 176 Black ash, 133 furnaces, 132, 149 leaching tanks, 133 waste, 135 testing, 140 Black gum, 65 Black liquor, 123 composition of. 123, 127, 147 evaporation of, 127 table, 444 testing, 139 use in sulphate process, 142 utilization of, 125 Blanc fixe, 346 Bleach required by fibres, 123, 380, 381 water, 354 Bleached pulp, blueing of, 253 color change on storing, 254 oxycellulose in, 252 washing, 246 Bleaching, 225 apparatus, 240 determination of degree ot, 11 Dobson process, 242 effect of acid, 240, 242 temperature, 234 on chemical properties, 251 strength of fibres, 249, 250 ground wood, 244 Hermite process, 233 jute, 243 principles of, 234 rate of color change, 239 systems, 241 use of antichlors, 244 backwater, 239 weight lost on bleaching, 236 Bleaching with chlorine gas, 226 permanganates, 247 peroxides, perborates, etc., 249 Bleaching powder, 228 action on metals, 231 deterioration, 229, 231 dissolving, 230 sludge from, 230 testing, 254 Blow pits, 194 tanks, 118 Blowing down pressure, 193 Boiler compounds, 356 scale, 3S5 . Boiling wood for grinding, 219 Bricks for digester finings, 181 Bright's method for unbleached sul- phite, 393 Bulk of fibrous materials, 65 paper, determining, 400 Bulker, pressure, 401 Burgess acid system, 173 Bursting strength of paper, 404 Calcium hypochlorite, color of solu- tions, 232 solid, 232 Calender staining, 321 Casein, composition, 334 detection in paper, 418 deterioration of, 339 determination in paper, 425 influence on penetration of oil, 328 insoluble matter in, 337 molding of, 337 preservation of solutions, 338 properties and preparation, 333 sizing, 286 soluble, 33s solvents, 336 testing, 339 waterproofing, 338, 339 Cattle feed from waste sulphite liquor. • 203 INDEX 471 Causticity of soda cooking liquor, 109 Causticizing methods, 98 Cell, structural unit of plant, 34 Cellulose, 1 acetates, 17 action of salts on, 7 and acids, 8 alkalis, 9, 20 ferments, 12 water, 3 benzoates r 19 chlorinated, 251 commercial meaning, 2 composition and constitution, 2 compound, 23 compounds of, 13 copper number of, n, 252 cuto-, 23 decomposition of, 8, 12 destructive distillation of, 13 determination of, 27 esters, 18, 19 formates, 19 groups of, 22 hydrates, 4 ligno-, 24, 156 nitrates, 13 nitrites, 17 normal moisture, 4 oxidation of, 10 pecto-, 23 peroxide, n physical properties, 1 solvents for, 5 sulpho-carbonate, 20 xanthate, 20 Cement for digester linings, 180 Chardonnet silk, 15 Chestnut, 62 Chips, length of, 159 Chlorinated cellulose, 251 Chlorine gas, 225 in bleached fibre, 252 in paper, 421 preparation by electrolysis, 232 Chrome yellow, 310 Clark process for water softening, 357 Clay, 294 composition of, 295 fineness of particles, 296 for coating, 345 specific gravity, 297 testing, 297, 302 Coated paper, 325 body stock for, 326 brittleness of, 440 finish on, 329 picking of, 328 printing qualities, 330, 344 waterproof, 338 Coating, coloring of, 321 determining amount on paper, 424 glycerine in, 329, 349 method of applying, 326 minerals used, 328 oils, soaps and waxes, 329 Cochineal, 313 Color standards for bleached pulp, 379, 381 water, 362 Coloring, 306 comparing colors, 307 matching shades, 306 mineral colors, 308, 310 mordants, 316 organic colors, 313 pigments, 308 vulcanized fibre, 323 Colors, acid, 320 basic, 318 direct cotton, 317 eosines and rhodamines, 319 pigments, 308 testing, 322 Combustion chamber, 166 Cooking liquor, testing, 137, 152, 209 Cooking sulphite, following progress in, 191 irregularities in, 191 Mitscherlich process, 183 precautions, 186 recording conditions, 187 472 INDEX Cooking sulphite, Ritter-Kellner pro- cess, 184 schedules for, 186 superheated steam for, 185 Coolers for sulphur dioxide, 168 Copper number of cellulose, n, 252 Corn stalks, 43 Cotton, 36 Cotton stalks, 44 Cottonwood, 61 Cucumber tree, 63 Cuprammonium hydrate cellulose, 5 Cutch, 313 Cuto-cellulose, 23 Cutter for sampling wood pulps, . 369 Cutting press for color disks, 376 Cymene, 198 D Daylight lamps, 307, 375, 379 Decay of wood, 56 effect on soda process, 95 Defects in printing, 435 Densities of soda stock, 122 stock in bleaching, 234, 241 sulphite stock, 195 Digesters, soda, 97, 98 sulphate, 142 sulphite, 178 bronze, 179 Mitscherlich, 183 Salomon-Briingger, 179 sizes of, 181 Direct cotton colors, 317 Dirt in pulps, 383 Disk machine for color comparisons, 376 District of Columbia paper tester, 405 Dobson bleaching process, 242 Dolomite for acid making, 174 Dorr causticizing process, 100 Douglas spruce, 60 Drying sized papers, 273 Ducts, proportion to fibres, 392 E Eau de Javel, 227 Electrolyzing salt, 232 Electrolytic bleach, 233 Electrotypes, 431 Elements, table of physical constants, 448-453 Enderlein's evaporator, 128 Enge process for ground wood, 223 Eosines, 319 Esparto, 42, 77 alkali required, 78 bleaching fibre from, 79 boilers, 78 bulk of, 66 cooking conditions, 79 dust, 77 recovery of alkali, 79 Evaporators, Enderlein's, 128 Torion, 127 Yargan, 128 Zaremba, 131 Ferments, action on cellulose, 12 Fert ilizer from waste sulphite liquor, 202 Fibre length of woods, 49 Fibres, estimation in paper, 387 by method of Spence and Krauss, 392 Fibrous materials, bulk of, 65 Fillers, 290 absorption of dyes by, 314 effect on sizing, 290 effect on strength of paper, 290 losses in process, 293 materials used, 291 retention of, 292, 417 testing, 302 Filters, 358 Firs, 58 Folding endurance of paper, 406 Formaldehyde in coatings, 338 Freeman process for soda cooks, 115 Fuel from waste sulphite liquor, 203 INDEX 473 Furfural, n Furnaces, recovery, 132, 149 Gas bleaching, 226 pressure in sulphite cooks, 188 relieving, 189 recovery, 192 Gelatine, 258 testing, 261, 332 Glarimeter, 402 Gloss of paper, 402 Glue, 258 detection of, 419 determination of, 425 for coating, 331 sizing, 287 Glycerine in coating, 329, 349 Grease-proof qualities of paper, 412 Grinders for wood pulp, 212, 213 Groundwood pulp, 212 Bache-Wiig process, 222 bleaching, 224 condition of stone, 216 wood, 218 efficiency of grinding, 219 estimation in paper, 427 from boiled or steamed wood, 219 in print papers, 89 sand settlers, 214 screens for, 213 speed of stones, 217 stones for making, 212 temperature of grinding, 218 testing, 223 woods for, 221 Gun-cotton, 15 Gypsum, 298 H Halftone plates, 430 standard depths, 431 screens for different papers, 433 Hardness in water, cause, 353 determining, 364 effect on sizing, 353 Heartwood, 49 Heavy spar, 301 Hemlock, 60 Hemp, 38 Hermite electrolytic bleach process, 233 Herreshoff pyrites burner, 166 Herzberg stain, 389 Hydrocellulose, 4, 8 Hydrochloric acid (table), 462 Hypochlorites, 226, 233 Hypochlorous acid, 227 Incinerating furnaces, 132, 149 Indanthrenes, 321 Ingersoll glarimeter, 402 Ink, choice of, 434 double tone, 442 drying, 438 for sizing test, 413 mottling of, 441 offsetting, 437 Jute, 39 bleaching, 243 K Knots, 56, 94, 159 Kraft fibre, 141 Lamp black, 312 Larch, 60 Leaching tanks for black ash, 133 Lignin, 25, 35, 157 reactions with sulphurous acid, 157 sulphonic acid, 157 Ligno-cellulose, 24, 156 Lime, 175 for causticizing, 103 rag boiling, 72 mud testing, 138 recovery in soda process, 102 474 INDEX Lime, slaking, 175 testing, 138 waste, 102 Limestone for acid towers, 1 70 Linen, 37 Linings for sulphite digesters, 178 Lithography, 432 smutting of ink, 440 Loading and filling, 290 Loft drying, 260 Logwood, 313 M Machine direction, 395 Magazine grinder, 213 Manila hemp, 39 bleaching, 243 Maples, 64 Mechanical pulp, 212 Melt from sulphate recovery, 150 Mercaptans, 146 Metric and ordinary units (table) , 466- 468 Microscopic examination of paper, 386 Milk of lime acid systems, 173 Mineral colors, 308, 310 Mitscherlich acid tower, 169 cooking process, 183 digester lining, 179 sizing process, 288 Moisture estimation in lap pulp, 371 paper, 414 wood pulps, 368 Monosulphite of calcium, 171, 176, 177, 192 Mordants, 316 Morterud digester, 97 Mullen tester, 405 Multiple effect evaporation, 128 N Newspaper, fibre from, 90 Nitric acid (table), 463 Nitro-cellulose, 13 O Ochres, 309 Odors from sulphate cooks, 142, 146, 151 Offsetting, 437 Opacity of paper, 400 Oxide colors, 309 Oxycellulose, 10, 252 Padding, 321 Paper, recovering fiber from printed, Paper mulberry, 41 scales, 397, 398 Paper testing, 386 absorbency, 410 acidity, 422 ash, 414 bulk, 400 bursting strength, 404 chlorine, 421 coating, 424 fibre content, 387 folding endurance, 406 gloss, 402 grease-proof qualities, 412 groundwood pulp, 427 machine direction, 395 microscopic examination, 386 moisture, 414 opacity, 400 paraffin, 421 permeability to air, 411 physical tests, 395 retention of filler, 417 rosin estimation, 420 sizing, 412, 417 stretch, 404 sulphur, 422 tearing strength, 409 tensile strength, 403 thickness, 399 unbleached sulphite, 393, 426 volumetric composition, 411 weight per ream, 397 wire side, 396 Papyrus, 44 INDEX 475 Paraffin, determination in paper, 421 Parchment paper, 6 Pea vines, 44 Pearl hardening, 299 Peat, 45 Pebble mills, 373 Pecto-cellulose, 23 Pectose, 93 Perborates in bleaching, 249 Permanganate bleaching, 247, 252 Permeability of paper, 411 Permutite water softening process, 358 Peroxides in bleaching, 249 Picking of coated papers, 328, 439 Pigments, 308 Pimaric acid, 265 Pines, 59 Pitch from waste sulphite liquor, 202 Poplar, 61 Porion evaporator, 127 Precipitated chalk, 300 Preston lining, 180 Printing, 429 defective, 435 inks, 434 paper for different types of, 432 Print-papers, 88 Prussian blue, 310 Pyrites, 161, 166 burner, 166 Quercitron, 313 R Rags, alkali used in boiling, 71 bleaching, 240 boilers for, 73 boiling, 70 bulk of, 65 dusting and sorting, 70 grades of, 68 lime for boiling, 72 losses in treating, 76 starch in, 72 washing of boiled, 75 Recovery of sulphur dioxide, 192 Redwoods, 313 Register in printing, 442 Relieving digesters, 103, 117, 189 Resins in plant cells, 35 woods, 53 rubber, 288 Retention of fillers, 292 Rhodamines, 319 Rinman's black liquor process, 125, 148 Ritter-Kellner acid towers, 170 cooking method, 184 Rosin, amount used in paper, 271 detection in paper, 419 determination in paper, 420 extraction from stumps, 265 in sulphite pulps, 195 properties of, 264 recovery from soda cooks, 117 size making, 265 testing, 275 Rosin sizing, 264 additions to, 267 and hard water, 272 composition, 266 defects, 274 effect of sunlight, 275 emulsifier for, 268 precipitation of, 269 troubles, 272 Rotogravure process, 433 Rubber resins for sizing, 288 Rushes, 43 Safflower, 313 Salomon-Briingger digester, 179 Sand settlers, 122 Sapwood, 49 Satin white, composition, 347 preparation, 348 Schopper folding machine, 407 tensile machine, 404 Sedimentation test, 383 Seed hairs, 36 476 INDEX Sizing (see also Rosin), 257 alum used, 277 casein, 286 defects of, 274 determining degree of, 412 kind of, 417 effect of drying, 273 filler, 272, 291 hard water, 353 sunlight on, 275 from sulphite waste liquor, 203 glue, 287 ink for testing, 413 Mitscherlich process, 288 reactions in, 270 requirements of different papers, 257 rosin, 264 rubber resins, 288 silicate-starch, 263 starch, 262 surface or tub sizing, 258 viscose, 287 Smelting furnaces, 149 Soap in coating, 329, 349 tub sizing, 259 Soda process, 93 alkali required, 115 black liquor, no, 123 boiling operations, 103 caustic soda consumed, 112 circulation of liquor, 103 cooking liquor preparation, 98 digesters, 97, 98 discharging digesters, 118 effect of caustic added, 107 causticity of liquor, 109 concentration of liquor, 106 steam pressure, 104 time under pressure, 108 lime recovery, 102 liquor per cord, 103 loss of fibre, 122 modified forms, 115, 116 principles of, 93 rate of reaction, no relief of digesters, 103, 117 Soda process, rosin recovery, 117 soda losses, 135 recovery, 126 steam lost on blowing, 119 required, 104 sulphur in, 115 wash pits, 119 washing black stock, 120 wood preparation, 95 woods used, 94 yields from various woods, 114 Soda lost in process, 135 recovery, 126 Sodium carbonate (tables), 457, 458 chloride (table), 456 Specific gravity and Baume (table), 459-461 Spruces, 58 Starch, detection in paper, 418 for coating, 343 sizing, 262 modified, 344 retention, 263 Starch-iodide indicator, 256 Steam lost on blowing digesters, 119 required in soda cooks, 104 sulphite cooks, 185 Steaming wood, 220 Straw, 41, 80 alkali recovery, 85 boards, 81 bulk of, 66 cellulose, 84 chlorination of, 86 composition of, 80 cooking, 81, 83, 86 packing in boilers, 82 retting, 86 Stretch of papers, 404 Stuffing, 321 Sudan III, 35 Sulphate process, 141 black liquor, 147 by-products from, 148 composition of liquor, 144 cooking, 143 INDEX 477 Sulphate process, digesters, 142 odors from, 142, 146, 151 recovered ash, 150 recovery furnaces, 149 smelting furnaces, 149 soda recovery, 149 value of alkalies, 143 yields, 145 Sulphite acid (see Acid) fibre, composition of, 195 process, 156 absorption apparatus, 168 cooking, 183, 191 digesters and linings, 178 modified processes, 197 pumping acid, 177 reactions of liquor making, 160 storing acid, 177 sulphur dioxide preparation, 161 theory of, 156 woods for, 157, 160 "turpentine," 198 waste liquor, 198 alcohol from, 205 cattle feed from, 203 composition of, 199 destructive distillation, 204 fertilizer from, 202 fuel from, 204 pitch from, 202 sizing, 203 tanning materials from, 204 testing, 210 use as binder, 201 volume obtainable, 200 Sulphur, 161 determination in paper, 422 in soda cooks, 115 per ton of sulphite, 193 sublimation, 166 testing, 207 use of molten, 162 Sulphur burners, 162 air supply, 165 automatic feed, 162 color of flame, 166 Sulphur burners, flat, 162 rotary, 162 temperatures in, 165 * Vesuvius, 163 Sulphur dioxide (table), 446 absorption apparatus, 168 cooling, 165, 168 preparation, 161 recovery, 192 solubility in water, 168, 446 testing, 208 Sulphur trioxide in burner gases, 165 Sulphuric acid (table), 465 and cellulose, 6 Sunlight, effect on sizing, 275 Superheated steam for sulphite cook- ing, 185 Sweet gum, 63 Sycamore, 64 Sylvic acid, 265 Tables: Aluminum sulphate, 444, 445 Ammonia, 464 Baume and specific gravity, 459-461 Black liquor, 444 Hydrochloric acid, 462 Metric and ordinary units, 466-468 Nitric acid, 463 Physical constants of the elements, 448-453 Sodium carbonate, 457, 458 Sodium chloride, 456 Sulphur dioxide solubility, 446 Sulphuric acid, 465 Temperature conversions, 447 Vapor pressure of water, 454 Talc, 300 Tank systems for acid making, 171 Tanning material from waste sulphite liquor, 204 Tearing test for paper, 409 Temperature comparisons (table), 447 Tensile strength of paper 4.03 478 INDEX Testing acid, 209 alum, 282, 286 black ash and black ash waste, 140 black liquor, 139 bleaching powder, 254 burner gases, 208 caseines, 339 clay, 297, 302 colors, 322 cooking liquor, 137, 152, 209 fillers, 302 gelatine and glue, 261, 332 groundwood pulp, 223 lime, 138 lime mud, 138 paper, 386 rosin and rosin size, 275 soda ash, 137 sulphur, 207 waste sulphite liquor, 210 water, 359 wood pulp, 368 Thickness of paper, 399 Thiosulphuric acid in sulphite liquor, 166 Three color work, 431 non-register, 442 Thwing tearing tester, 409 Tower systems of acid making, 169 comparison with milk of lime sys- tems, 175 difficulties in operating, 171 Tracheids, 45 Tub sizing with gelatine, 258 starch, 262 Tuliptree, 63 Turmeric, 313 Turpentine, sulphite, 198 U Ultramarine, 311 Umber, 309 Unbleached sulphite estimation, 393, 426 Ungerer's cooking process, 143 Vapor pressure of water (table), 454 Venetian red, 309 Viscose, 20, 287 silk, 21 Volume composition of paper, 411 Vulcanized fibre, colors for, 323 W Wash pits for soda fibre, 119 Washing black stock, 120 bleached pulp, 246 rag stock, 75 sulphite fibre, 194 Water, 351 alum treatment, 359 analysis, 359 bleach consumed by, 354 classification of, 352 filtration, 358 for dyeing, 316 iron in, 353 requirements in paper making, 75, 3Si, 353 sampling, 360 scale from, 355 soft and hard, 352, 364 softening, 356 suspended matter in, 354, 363 Waterproof coated paper, 338, 343 Waxes in coating, 329, 349 Weight per ream, 397 Weld, 313 Willesden products, 6 Winestock process for old papers, 90 Wire side of paper, 396 Witherite, 301 Wood, 45 barking, 158 capacities of machinery for handling, 160 chipping, 159 decay of, 56 fibre length, 49 INDEX 479 fibres, 45 for grinding, 221 kinds used, 57 moisture in, 51 preparing, 94, IS7 proximate analysis, 54 requirements for sulphite process, 157 resins in, 53 weight per cord, 66, 114 cubic foot, 52 Wood pulp testing, 368 beating test, 372 bleaching qualities, 380 borer for sampling, 369 color comparisons, 375 Wood pulp testing, dirt count, 383 loss in weight on bleaching, 382 moisture determinations, 368 sedimentation test, 383 Yaryan evaporators, 128 capacities, 132 Zacaton, 44 Zaremba evaporator, 131 Zinc chloride and cellulose, 5 333 90 % •^ a* *i

.y bV u •>Va\ ♦* *♦ /j >*<$? vSS^^ \*Sy %/ # 95?\o> \* .. * e "° * u <$v \^ > k0- ^ ^ V o o " • * ■* ^ A^ ^ % .,-.• k,<" £**V IS , * v !,« ^»_-V A 6 ^V *"* A A ***?_£&-.%>. /4V&\ ^i8fr\ / '^9 ' *•&. © vJwLv^* av r^. » Sol ^ *° • » * A