’ ; SS ce ‘ ete: é i Lun © Raymond Pettibon RESEARCH LIBRARY THE GETTY RESEARCH INSTITUTE JOHN MOORE ANDREAS COLOR CHEMISTRY LIBRARY FOUNDATION va 4 BAW AOMA WM PRACTICAL COLLOID CHEMISTRY PRACTICAL COLLOID CHEMISTRY WOLFGANG OSTWALD PROFESSOR OF THE UNIVERSITY OF LEIPZIG WITH THE COLLABORATION OF DR. P. WOLSKI and DR. A. KUHN TRANSLATED BY I. NEWTON KUGELMASS, M.D., Ph.D., Sc.D. YALE UNIVERSITY SCHOOL OF MEDICINE AND THEODORE K. CLEVELAND, Ph.D. WITH 22 ILLUSTRATIONS NEW YORK E. P. DUTTON AND COMPANY INC. PUBLISHERS THE GETTY RESEARCH INSTITUTE LIBRARY PREFACE TO FOURTH EDITION HIS book has been received so favourably that three editions were exhausted within two years. The author attributes this to the dire need for experimental knowledge of colloid chemistry after a period of theoretical interest in this science. If this be the explanation, it is all the more gratifying that this manual of experimental colloid chemistry has been welcomed to such an extent. The rapid sequence of the previous editions made it impossible to keep abreast of advances in colloid chemistry. Therefore, the present edition has been revised and re- edited in many respects. The experiments in this manual were performed and the procedures tested many times during the last four or five semesters by about two hundred students under the supervision of the author and Drs. P. Wolski and A. Kuhn. As a result, previous errors have been corrected and fifteen new experiments have been added, which include Chapter X on elementary dispersoid analysis. The author has also received suggestions from colleagues and invites such in the future. Some reviewers of the book suggested an index. The author purposely omitted it in previous editions because the manual contains a systematically arranged list of the experiments in the table of contents. If the manual is to give a survey of experimental colloid chemistry, it is preferable that the student learn to recognize systematic Vv vi PRACTIGAL COLLOID CHEMIST: colloid phenomena in conjunction with the experiments. Inspection of the manual readily reveals the chapter in which certain types of experiments are to be found. This manual is, of course, no reference work, but it does give a systematic presentation of the phenomena of colloid chemistry. WO. OSTWALD Leripzic May, 1924 CONTENTS Peek eeaRoA LION OF COLLOIDAL SOLUTIONS A. CONDENSATION METHODS: OW, N Ho MAsTIc, PARAFFIN, SELENIUM SOLS . ; : RED GOLD SOL PREPARED WITH TANNIN RED GOLD SOL PREPARED WITH ALCOHOL . BLUE GOLD SOL PREPARED WITH HYDRAZINE HYDRATE : GOLD SOLS PREPARED WITH ILLUMINATING GAS GREEN GOLD SOLS PREPARED WITH ALCOHOL GOLD SOLS PREPARED WITH COMBUSTION GASES GOLD SOL PREPARED WITH A SOLID DISPERSION MEDIUM : , SILVER SOL PREPARED WITH TANNIN SILVER SOL PREPARED WITH HYDRAZINE HyDRATE . COLLOIDAL SULPHUR . ARSENIC TRISULPHIDE HyDROSOL . ANTIMONY TRISULPHIDE HyDROSOL . MERCURIC SULPHIDE HyYyDROSOL . MERCURIC SULPHIDE ALCOSOL . ; . SILVER [IODIDE HyDROSOL . SILVER CHLORIDE AND BROMIDE HYDROSOLS Z . PRUSSIAN BLUE HyYDROSOL . COPPER FERROCYANIDE HYDROSOL . FERRIC HYDROXIDE SOL . FERRIC HYDROXIDE SOL BY Hort DIALysis . . FERRIC HyDROXIDE SOL BY THE GRAHAM METHOD vill AP RW > —— na KJ Substance Added 100 200 200 400 500 600 = 00 Fic. 3. VisCOsliy Viscosity denotes the resistance which a fluid exerts against displacement of its own molecules. Glycerin has a high viscosity and ether a very low one. An approxi- mate measure of this value is the time a given volume of liquid requires to flow through a certain capillary. Relative viscosity suffices in colloid chemistry. It is proportional to the product of the time of flow, the specific gravity of the substance and the so-called ap- paratus constant, K. This is determined by standard- izing the viscosimeter in terms of distilled water. Gener- ally, the specific gravity may be neglected. If the viscosity of water, observed in the apparatus, be as- sumed unity, the relative viscosity of the colloid studied is simply the ratio of the time of flow of the colloid to that of water through the viscosimeter. The viscosities of many homogeneous liquids are independent of the 38 PRACTICAL COLLOID CHEMISE: pressures which produce the capillary flow. However, the viscosity is affected appreciably at higher pressures, of course in the direction of a faster rate of flow. The work of E. Hatschek, W. Hess, and E. Rothlin shows that the viscosities of hydrophile colloids deviate from Poiseuille’s law with slight variations. Hence, even relative viscosity measurements must be made at known or constant pressures. The Ostwald viscosimeter is a very convenient form of apparatus. It consists of a U-tube, one arm of which has a capillary and bulb. The two ends of the bulb are marked so that it may J containa definite volume (Fig. 4). In using the viscosimeter, always place the same volume of liquid in the tube, draw it into the capillary side arm above the upper mark, and measure the time of flow of the liquid between the two marks. Use a viscosimeter with a capillary of greater bore for viscous liquids, such as hydro- phile emulsoids. Viscosimeters whose capil- laries permit the volume of water between the two marks to flow out within 20 seconds are most suitable. The viscosity of a liquid varies considerably with the temperature, hence the viscosimeter should always be kept in a water-bath with two glass sides so that the flow of the lquid may be observed. The viscosimeter should be cleaned by drawing cleaning mixture and water through the capillary. The presence of gas bubbles in the capillary prevents adequate cleansing of the viscosimeter wall, and therefore inaccurate meas- urements result. This viscosimeter may be more con- veniently used by providing the side arm with a rubber stopper containing a bent glass tube. Attaching a piece of rubber tubing to the glass tube, the liquid within BiG. A; SUR AGH. CE NSION AND VISCOSITY 39 the viscosimeter may be blown into the capillary and back with each measurement. Wee oti yY sex PERIMENTS. WITH GELATIN SOLUTIONS Gelatin is a hydrophile emulsoid which has been studied more extensively than other albuminous substances. Prepare a I per cent. solution of gelatin in the following manner: Weigh 2 g. of either sheet gelatin, gelatin cuttings used for photographic purposes or gelatin powder and place it in cold distilled water. Change the water often in order to obtain a “ pure”’ solution. Weigh a beaker, add 150 c.c. of water and heat to boiling. Remove the flame and add the swollen gelatin free from wash water. Stir continuously with a glass rod until the gelatin dissolves. The decomposing effect of heat as well as the stickiness of the gelatin are thus avoided. After the gelatin dissolves, cool and weigh the beaker and contents. Add enough water to obtain 200 g. of a I per cent. solution. Pass water at room temperature through the viscosi- meter at least 30 to 40 seconds before measuring the viscosity of the gelatin. Expt. 66. Influence of the age of gelatin solutions upon viscosity—Measure the viscosity of the cooled gelatin solution directly after preparation. Measure it again in half an hour. The solution may be left in the viscosimeter, but must be drawn two or three times through the capillary before making a measurement. The viscosity of the solution increases considerably after standing an hour. A I per cent. solution is usually too viscous to flow through the viscosimeter. Expt. 67. Influence of preliminary mechanical treatment on the viscosity of gelatin solutions— 40 PRACTICAL COLLOID CHEMTS irae Allow a o-5 per cent. gelatin solution to stand in the viscosimeter for 24 hours. Measure the viscosity after drawing the solution very slowly into the capillary. Then agitate the solution by rapidly drawing it through the capillary several times or by bubbling air through it. Measure the viscosity once more and observe a considerable shorter time of flow. The structure of a very dilute gelatin solution is the cause of these phenomena. Therefore, it is necessary to run the solution through the viscosimeter two or three times before determining the viscosity of the colloid. Expt. 68. Influence of preliminary thermal treat- ment on the viscosity of gelatin solutions—Measure the viscosity of a 0-5-1 per cent. gelatin solution. Place about 200 c.c. of this solution in an Erlenmeyer flask and heat on a steam-bath. Provide the flask with a reflux condenser to prevent loss of water by evaporation, or mark the original water level, and after heating add the required amount of water. Remove every half-hour about 20 c.c. of solution and determine after cooling the viscosity of this sample.! Do not forget to replace the water lost by evaporation. To obtain more accurate results, heat a larger volume of gelatin solution and take larger test portions. Use these portions to determine the increase of viscosity with time. Plotting the viscosities obtained as ordinates against the age of the solutions as abscisse, a series of curves are obtained in which the gelatin solutions which have been heated the longest, show the lowest slope. For longer periods of heating the gelatin solutions, the slopes of the age curves approach zero. 1 The concentration of the solution can be maintained fairly constant by weighing the filled flask before and after heating as well as before and after removing samples and subsequent refilling. Seer TENSION AND VISCOSIFY AI Expt. 69. Influence of concentration upon the viscosity of gelatin solutions—Prepare the following solutions by mixing a warm I per cent. gelatin solution with warm water, or place the cold mixture on a steam- bath for 5 minutes and then cool : iinet ey. , : : Te ees eA BS So wotereiatin —. : fee a Oe cer water |. : eee Sel Ose 5 ris Allow the mixtures to stand a few hours before determin- ing the viscosity and then determine them all in sequence. Plot the curves to show the increase of viscosity with concentration and determine the gradual increase in slope. Expt. 70. Influence of temperature on the viscos - ity of gelatin solutions—Determine the rate of flow of a 5 per cent. gelatin solution at the temperature of ice water. The filled viscosimeter should be allowed to come to this temperature by letting it stand at least half an hour in the ice water. Then heat the bath to a tem- perature of 20°C. and allow the filled viscosimeter to remain one half-hour before repeating the determination. Make a third measurement at 4o°C. Plot the corre- sponding curves and observe the rapid fall of viscosity with rising temperature. Expt. 71. Influence of additions of electrolytes on the viscosity of gelatin solutions—Prepare the following mixtures : Peon per cent, gelatin -+- 20°c.c. H,O oa * * Pees 20rc,. ce IN Na oO) or KaO) for ; MgsO,. 2, ay = pe et20 GC. OwINeK IT or K Br, 4a. a * Peale C:C.t1 Oi 216. C.Or1N. HCl | = about 0-005 N HCl. 4b. oo wo se 10 C.Ce1.0 = 47¢.c7 N’ HCP about o-IN HCl. 42 PRACTICAL COLLOID CHEMIST (5%. 20 C.c. I per cent. gelatin4- 16 c.c) B30 =a cen ar 4 NaOH = about o-o1N NaOH. (ee oe Fr » +16c.c. H,0+4c.c. N NaOH =about o-IN NaOH. Mix the solutions thoroughly and allow them to stand for 24 hours. Determine rate of flow, or better still, determine the age curves in the manner described above (EXDLEOS8): The addition of sulphates to a gelatin solution increases its viscosity considerably compared with that of pure gelatin. The iodides and bromides greatly decrease the viscosity. Carbonates, phosphates, oxalates, acetates and citrates raise the viscosity. The cyanides and thio- cyanates lower it. Chlorides, nitrates, and chlorates form complex changes, in so far as they can raise or lower the viscosity according to their concentration and the age of the gelatin. Additions of acids and bases, in the small concentra- tions mentioned, increase the viscosity. Greater con- centrations lower it again. A viscosimeter having a rate of flow for water equal to 150 fifth-seconds, gives the following values : 700 for a 0:5 per cent. gelatin solution ; 3-4,000 for 0-005N HCl ; 300 for o-rN HCl; 2-3,000 for 0‘o1 N NaOH; 500 for o-1 NNaQOH, etc. Prepare the complete concentration curves for HCl, NaOH and NaCl, and determine the viscosity 24 hours after pre- paration of the mixtures. VISCOSIMETRY OF CHANGESSOR 33 AGGREGATION Expt. 72. Viscosity measurements on the coagu- lation of aluminium hydroxide hydrosoi—Measure the viscosity of a highly concentrated Al(OH), sol pre- pared according to Expts. 30 and 34. The viscosity of Sere TENSION AND. VISCOSITY 43 this sol is higher than that of water, hence use a vis- cosimeter of wider capillary bore. Mix 8 c.c. of the sol with 2 c.c. of 2N KCl! and measure the changes in vis- cosity with time. Such an experiment gave the following results : Original Value 393 fifth-seconds After 15 minutes 406 ay, ”» 55 ” 417 5) reo = 3 427 3 i Ba ae 445 :, Coagulation depends upon the concentration of the sol and that of the added solutions. After coagulation, the viscosity of the system again decreases and after 16 hours the vigorously stirred solution gives a rate of flow equal to 415 fifth- seconds. tere ee: Expt. 73. Viscos- . fs ity measurements of the setting of plaster of Paris (Wo. Ostwald and P. ‘ ik Wolski, Kolloid. Z., hes 27, 78 (1921)—Pre- te Patewae 5 per cent, suspension of finely powdered gypsum and transfer at once to a_ viscosimeter. Sedimentation may be prevented by drawing a continuous Time of Flore ———» Fa : Se, B 2555 30) SSG OS current of airthrough STMT eles the suspension. Ob- Fi: 5. 1 Use an empirically determined KCl concentration which wil produce no flocculation within a half-hour, 44 PRACTICAL COLLOID CHEMISia serve that the viscosity increases with the time as shown by Fig. 5. The process of setting may thus be followed by viscosity measurements. A viscosimeter having a diameter of 0-7-I-0 mm. and a water value of I00—150 fifth-seconds is suitable for a charge of 20 c.c. Expt. 74. Viscosimetry of the formation of potato starch paste—Suspensoid systems, such as_ coarse suspensions of starch in cold water, show relatively small increases in viscosity, approximately proportional to their concentrations. On the other hand, hydrophile emulsoids show very great increases in viscosity both absolutely and relatively, with increases in concentration. In the forma- tion of potato starch paste, which, as is well known, generally takes place between 55° and 65° C.,a suspensoid system is converted into an emulsoid one. On warming the starch suspension, the viscosity decreases in accord- ance with the known decrease in the viscosity of water, the dispersion medium. When the temperature is raised between 55°—65° C. this decrease is, however, replaced by a great increase in viscosity, the most striking criterion of the formation of paste and of a radical change in the starch-water system. In order to measure the viscosity of a dispersoid which coalesces spontaneously after the fashion of unheated starch suspensions, a viscosimeter must be used whose rate of flow is relatively large compared to its rate of sedimentation.! Suitable viscosimeters have small volumes and short narrow capillaries or larger volumes and longer capillaries of wider bore. A 0-5—I-0 per cent. starch suspension may be used in the first type of viscosi- meter, whereas a 5 to 10 per cent. starch suspension is suitable for the second type. Warm the starch suspension in the viscosimeter in a water-bath with glass walls (a large beaker). Place the 1 See Wo. Ostwald and H. Luers, Kolloid. Z., 25, 82, 116 (1919). pUnPaACee TENSION AND VISCOSITY 45 thermometer in the liquid contained in the wide side arm of the U-tube, for the temperature of the suspension lags behind that of the water-bath. Warm the starch suspen- sion rapidly to 50° C., measuring the rate of flow every 10°. Above 55° C. warm the suspension slowly at a rate of about 1° C. per 5 minutes. Lower the flame, measure the rate of flow continually and record the temperature of the water-bath after stirring. Stir the starch suspension thoroughly, by bubbling air through it before each measurement. If the concentration and the capillary bore of the viscosimeter have been suitably chosen, we find within a narrow temperature interval a change from decrease of viscosity to an increase as given in the follow- ing example : Five fer cent. suspension of commercial potato meal ; small viscosimeter, water value about 365 fifth-seconds at ay sie Cnarge, 10: C.c. Temperature. Time of Flow (T). Log. 1. See 239 2°30173 ah: 228 rahoy pcs! 56-2 226 *35411 56°60 225 *35218 DiS e274 ie Pe hers) “veg ae "34439 58-1 220 °34242 58:4 220 °34635 58:8 228 "35793 59°4 235 ops 59°5 240 -38021 59°7 244 "38739 Between 58-1° and 58-4°C. there is a reversal of the viscosity change, i.e. the formation of starch paste. A graphical representation of the formation of starch paste 46 PRACTICAL COLEOID CHEMISiE is given below in another experiment. If the viscosi- meter is not sensitive or the concentration of the starch suspension too small, the time of flow neither decreases nor increases, but remains practically constant within a range of 10°C. as in the following experiment : A 5 fer cent. suspension of potato meal; viscosimeter too large, water value about 150 fifth-seconds at 25° C.; charge, 20 €.C. Temperate soe 4I. 49 53 55 57° 59 Oba Ojo eos Time of Flow. . 112 106 103 101. 100,300) 100siQ0 =a (se For a more accurate determination of the temperature of starch paste formation, it 1s necessary to select a suitable viscosimeter so that the point of inflexion is observed within a very narrow temperature range. Furthermore, it is observed that in going above the temperature of starch paste formation, the times of flow determined in rapid succession are no longer constant but increase spontaneously because starch paste formation requires a certain amount of time. Such an observation can of course be utilized as an approximate indicator of the temperature of starch paste formation by deter- mining at which temperatures there are definite increases in viscosity according to three consecutive measurements made within 10 minutes. The use of a sensitive viscosi- meter is of course more accurate and more rapid.! To determine the exact temperature of starch paste formation, graphic representation of the data may be made in the following way: Plot the temperatures as abscissee and the logarithms of the times of flow as ordinates. From the data in the above table are obtained 1 Such experiments demonstrate that a definite temperature of starch paste formation in a physico-chemical sense is a practical entity. Strictly speaking, there is probably a temperature range in which the rate of starch paste formation is abnormally rapid. — Peewee LE NSION AND VISCOSITY 47 two practically straight lines (Fig. 6), which produced, intersect at a point. This experiment gives graphically a temperature of starch paste formation between 58:-2°— 58:3° C. Ifthe graph of data obtained shows a horizontal line connecting the two oblique branches, it is due to the use of an unsuitable viscosimeter or to too small a con- centration of starch suspension. In such a case the two lines may be produced until they intersect at a point which would give an approximate value of the temperature of starch paste formation. SB Expt. 75. Viscosity mea- surements of the ageing of starch paste (M. Samec). ~* —Prepare al percent. starch solution by moistening 4 g. of potato starch with a small “sss amount of water and add gradually with constant stir- ring, 200 c.c. of warm but not hot water. Dilute to twice the volume, boil on a sand-bath for 30 minutes and after cooling make the volume up to 400 c.c._ Filter and cover with toluene for protection against bacterial action. Determine the time of flow for this suspension immediately after preparation. Measure the rate of flow of this suspension every day and plot the rates of flow against age. On plotting the viscosities obtained, the curve shows a large decrease in viscosity at first and gradually becomes asymptotic after an ageing of one to two weeks. Expt. 76. Viscosity measurements on the coagula- tion temperature of an albumin solution—Separate the yolk of a fresh egg from the white. Beat the latter into a foam and allow to stand overnight. The greater 039) Zoe * 59 60° 5? 58 Temperotur —> HiG,.6; 1 Another example is given by Wo. Ostwald, Koll. Z., 12, 215 (1913). 48 PRACTICAL COLLOID CHEMEStiaa= part of the foam clears while the egg membrane remains suspended in the remainder of the foam. One egg gives about 20 c.c. of liquid, which is a mixture of albumin, and globulin. Dilute to twice the volume with a weak solution of 0-7 per cent. NaCl. The solution becomes turbid upon dilution with distilled water due to the relative insolu- bility of the globulin. Measure the change in viscosity with rise in temperature under the same conditions as those of the starch paste in Expt. 74, using the same rates of temperature increase. The time of flow decreases at first with rise in temperature until about 60° C. is reached, then either an increase or no change in time of flow occurs, depending upon the concentration of the solution and the sensitivity of the viscosimeter. The time of flow decreases again when the temperature exceeds 70° C. Observe that Temp. T. for egg white. T. for water. A 51-0 yas 233 78 55°0 292 220 W2 56:3 286 217 69 57:0 283 215 68 57°60 281 212 68 58°3 255) 212 65 59:2 275 210 65 60°3 | 270 206 64 61-0 267 205 62 61°6 268 204 64 62:6 267 202 65 62°8 207; 200 67 63°5 266 199 67 0357, 267 198 69 64°5 266 196 70 66-0 252 192 60 66:9 247 190 57 69°5 238 185 53 a et. Senet TENSION “AND VISCOSITY 4Q the solution becomes turbid during the increase in the times of flow. The viscosity change during coagulation may be made more obvious by plotting the viscosity increase (i.e. time of flow of solution minus time of flow of H.O at the corre- sponding temperature) instead of the rate of flow as observed. This may be done by plotting a temperature— viscosity curve for water in the same viscosimeter. Com- pare the rates of flow of water with those of the egg white solution at the corresponding temperatures and plot in the same manner. A simple illustration of such an experi- ment is given graphically in Fig. 7,1 wherein the differ- 60 Temperatur —> BiG.7- ences in viscosities between albumin and water are plotted for increasing temperatures. The experiment was performed with equal volumes of egg albumin and 0-7 per cent. NaCl solution. The viscosimeter was of the small type with a water-value of 365 fifth-seconds at 250. The curve and data show that the coagulation of the albumin solution takes place at about 61°C. By plotting 1 Another example is given by Wo. Ostwald, Koll. Z., 12, 214 (1913). 4 50 PRACTICAL COLLOID CHEMISTie the legarithm of the rate of flow against the temperature, two approximately straight lines are obtained which intersect to give an acute angle. The determination of coagulation temperatures is practically conclusive only when all experiments are conducted under similar con- ditions. The viscosimetric method may be used to study the kinetics of effects of added salts, acids and bases on the course and mechanism of coagulation at the critical temperature range. 1 More specific data is given in the monograph by H. Chick and C. T. Martin, Kolloidchem. Bethefte, 5, 49 (1913). IV OPTICAL PROPERTIES Ele AaAl, HETEROGENEITY r NHE optical heterogeneity of solutions is shown macroscopically by turbidity, microscopically by the so-called ultramicroscopic phenomena. Turbidity is best observed by contrast against a dark background. Better results may be obtained by placing a source of intense illumination to one side of the dark background or by holding the test-tube in a narrow beam of sunlight or a beam from a projection lamp. For ordinary purposes wrap a piece of black paper with a small hole in it around an incandescent lamp. The light from the source of illumination should not fall directly on the eye of the observer. For semi-quantitative purposes, prepare a comparison scale with milk or mastic hydrosol.!. Start with a con- centrated milk-white sol and dilute to give the required turbidity. A mastic sol is remarkedly stable for deter- minations of the turbidity of colourless sols provided an aged sol is used. Nephelometers, Tyndallometers ? and 1 Diluted milk is especially suitable as a standard since the measurements of N. Manz (Dissertation, Marburg, 1885) showed that it absorbs all wave-lengths equally. F. B. Young [PAzl. Mag. [6] 20] 793 (1910) | used diluted milk as a standard for degrees of turbidity of ether at the critical temperature. *See B. H. von Oettingen, Z. f. physik. chem., 33, I (1900) ; J. Friedlander, ibid., 38, 385, 413 (1900) ; C. Benedicks, Koll. Z., 51 52 PRACTICAL COLLOID CHEMISTRY also colorimeters may be used for more accurate deter- minations of turbidity,! which is the ratio of the light diffracted by the colloid particles to that transmitted. Expt. 77. Detection of faint turbidity by means of the Faraday-Tyndall light cone—A large number of disperse systems appear completely transparent upon superficial examination, especially in transmitted light. However, they produce a decided Tyndall effect. Examine the following sols by transmitted light and then by a narrow beam of light from a projection lantern : a red gold sol, freshly dissolved collargol, freshly prepared arsenic trisulphide or Prussian blue, as well as a I per cent. solution of potato starch paste which has been heated for thirty minutes. A cold saturated solution of cane sugar gives a bluish- white light cone. Expt. 78. Polarization of the Tyndall light cone— Place a turbid colloidal solution as mastic hydrosol in the beam of a projection lantern. Use a sol of such dilution that it will give a well-defined Tyndall cone unaffected by too an intense source of light. Examine this cone with a Nicol prism at right angles to the beam of light. The cone of light disappears or becomes dim twice 7, 204 (1910) ; Th. W. Richards, Proc. Amer. Ac., 30, 385 (1904) ; Am. Chem. J., 31, 235 (1914); 35, 510 (00@G) ee ee ibid., 35, 100 (1906); E. Schlesinger, Berl. Klin. Wochenschr., 48, 42 (1911), etc.; accurate measurements are given by W. Steubing, ‘‘ On the optical properties of colloidal gold solutions,” Dissert., Greifswald (1908); Ann. d. Physik, 26, 329 (1909) ; W. H. Keésom, Ann. d. Physik, 35, 501 (19011) ;) Comma evs. Tab., Leyden, No. 104 (1910) ; W. Mecklenburg and S. Valentiner, Z. f. Insirumenthunde, 34, 209 (1914) ; Kolloid. Z., 15, 99 (1914) ; 16, 97 (1915); F. Sekera, Koll. Z., 27,728 (1021) . hee ibid., 27, 236 (1921). 1 An apparatus like the Wilh. Ostwald-Donnan’s Colorimeter is suitable (Physico-chemical Manual, 3rd Ed., 358 (1910). OPTICAL PROPERTIES 53 in one complete revolution of the prism. The light is thus partially polarized. Perform similar experiments with a very dilute solution of a fluorescent substance, such as quinine sulphate, alkaline fluorescin or eosin and observe that no dimming of the light cone occurs by revolving the prisms. Fluorescent light in distinction from a light ray of a turbid solution, is not polarized. Expt. 79. Turbidity and degree of dispersion— Experiment and theory show that turbidity is greatest in a solution of moderate concentration of the disperse phase. Therefore, the maximum degree of turbidity does not occur in colloids, but in coarsely disperse systems. Place freshly prepared mastic, arsenic trisulphide or red gold sol in two beakers and to one add a few drops of HCl or BaCl, solution. Compare the turbidities by means of a Tyndall cone and observe that the coalescing sol shows a considerably greater turbidity. Perform the same experiments with dilute ‘‘ Congo rubin.”’ The “ pure’’ solution seldom shows a light cone in a nephelometer. Add a few drops of an electrolyte as HCl, Ba(OH),, until the solution gradually changes to a blue-violet colour as the turbidity increases. The coagulated sols show individual coarsely disperse particles when shaken. The diffracted light which these coagu- lated particles radiate is less intense than during the initial stages of flocculation. TURBIDITY PHENOMENA IN HYDROPHILE COELOIDS Changes in turbidity are not only dependent on the variation in the degree of dispersion but also on the degree of hydration. Every emergent ray from a colloid solution whether produced by refraction or reflection, is due to a distinct variation in the optical relations between. 54 PRACTICAL COLLOID CHEM the disperse phase and the dispersion medium. This difference is less the more hydrated is the disperse phase. The difference increases if dehydration of the disperse phase takes place. Variations in the degree of dispersion and hydration frequently occur at the same time. Marked variation in the turbidity of a colloid solution may occur with but slight changes in external conditions. Expt. 80. Changes in turbidity of aqueous gelatin solutions with concentration—Prepare a_ series of gelatin solutions! of the following concentrations by diluting with warm water : 65 4 -30 27°35) 50. 30 Sper After the solutions have cooled allow to stand over- night in an ice-box. The maximum turbidity occurs not in the most concentrated solution but in the one of medium concentration such as 2-3 per cent. Gelatin solutions prepared at a lower temperature show the turbidity maximum more distinctly. Soak a thick sheet of gelatin or a transparent sheet of glue in a beaker filled with water. After a few hours, compare the soaked swollen portions with those still unaffected. Observe that a considerable increase in turbidity has occurred in the portion swollen by the water. Dry a piece of 30 per cent. gelatin jelly in an oven or ina desiccator over H,SO,. Do not use too high a temperature when drying with heat on account of the tendency of the gel to liquefy. The 30 per cent. gelatin jelly is very turbid, but it becomes less so with gradual loss of 1 The gelatin is purified by washing for 2-3 days with continu- ously flowing water or by frequently replacing with distilled water, and the weight of the gelatin determined before and after swelling with the added precaution that none of the gelatin is lost during the washing... The latter may be realized, in the author’s experi- ence, by using boiled porous sacks. The experiment may at times be carried out with unwashed gelatin. OPTICAL PROPERTIES 55 water and finally almost transparent when the original thickness of the gelatin is attained. Therefore a gelatin- water mixture of various concentrations may have two degrees of maximum turbidity. Expt. 81. Effect of dehydration on turbidity of silicic acid gels—Prepare according to Expt. 25 a clear aqueous solution of silicic acid by mixing two parts of 2N acetic acid with one part of water, and after cooling the mixture, carefully add one part of Io per cent. water- - glass.1_ Place the greater portion of this gel in a desiccator over concentrated H,SO,. This dries in the course of I to 2 weeks at a rate which may be determined by periodic weighings. Observe that with a water content of 35-55 per cent., the apparently clear jelly containing some gas bubbles gradually becomes turbid. Generally, the centre of the jelly mass first shows an opalescence which gradu- ally extends in all directions. This illustrates the sudden “transitions ’”’ of the silicic acid gel. The turbidity dis- appears after longer desiccation and the jelly becomes as transparent and as firm as glass.2_ If a dried piece of jelly is placed in a flask with moistened filter paper and the flask sealed with a stopper, the hardened gel frequently disintegrates with a noise as a result of internal stress. Such gels usually show a new “ transition ”’ after rehydra- tion but not to as marked a degree. : Expt. 82. Gelation and _ turbidity—Allow one portion of a 2-3 per cent. gelatin solution to solidify at room temperature, another portion in an ice-chest and 1 Acetic acid yields as a rule clearer gels than HCl. 2 Good results have not always been obtained by the author, as J. M. van Bemmelen has already reported. Occasionally the author has been able to observe the phenomenon on drying the silica gel, previously washed with HCl, in the air and also warming it gently. Sometimes, in spite of apparently similar conditions, the phenomenon has not been observed ; however, it may have been missed during the night. 56 PRACTICAL. GOELOID, CHEMisi a keep the remainder liquid at about 30°-40° C. The solution which solidifies most rapidly at a low temperature is the most turbid and that solidifying at room temperature is more turbid than the fluid solution at an elevated tempera- ture. Fill two beakers with the same 2-3 per cent. gelatin solution, liquefy both portions by placing them in hot water for about 30 minutes. This liquid jelly is much more turbid than the solid portion. Expt. 83. Ageing phenomena and _ turbidity— Let a I per cent. solution of starch paste age in an ice- chest (Expt. 75). The freshly prepared solution shows a bluish-white opalescence. Its turbidity increases con- siderably upon ageing and in the course of 2—3 weeks it becomes white and opaque. A considerable increase in turbidity of the jelly is observed after 24 hours. Expt. 84. Influence of electrolytes upon the turbidity of gelatin jellies—Prepare about 150 c.c. of 2-3 per cent. gelatin solution. This concentration of jelly shows the maximum turbidity (Expt. 80). Pour ro c.c. of this solution into a number of test-tubes and add to successive portions a drop of 2N solutions of NaOH, KI, KNCS, KCl and Na,SO, so that a o-oIN solution results. Place these tubes in the ice-box, as well as two control tubes which contain no additions. The following series of decreasing turbidities are observed after 1 to 2 days: Control, KI, KNCS, KCl, Na,SO], Gis siege All electrolytes in the concentrations given above pro- duce a decrease in the turbidity of the jelly. Acids and bases exert the strongest effect. Expt. 85. Critical turbidity—A distinct turbidity maximum is observed at an intermediate stage of mixing two liquids soluble in one another to a limited extent (J. Friedlander, V. Rothmund). Prepare a mixture of OPTICATS “PROPERTIES 57 about 36 parts of colourless solid phenol and 64 parts of water. At room temperature, there are two layers, which form molecular disperse immiscible systems. Heat the mixture to 70° C. and shake continuously until the turbid emulsion becomes clear. Continue to shake the solution and allow it to cool slowly. At first a very slight tur- bidity occurs which appears as a colour phenomenon or as an opalescence. This is indicative of the emulsoid state.! On further cooling the turbidity increases con- siderably and a coarsely disperse emulsion appears which gradually separates into two distinct layers. ULTRAMICROSCOPY According to the theory of microscopy, particles appear- ing geometrically similar are greater than the wave-length of hight to which they are exposed. Such particles may be differentiated from one another provided they are spaced at intervals greater than one-half the wave-length of the hght by which they are illuminated. The smaller colloidal particles have the dimensions of about o-Im, and hence cannot be optically distinguished with an ordinary microscope. However, it is possible to distinguish single colloidal particles without their individual geometric forms by means of an intense lateral illumination and not by transmitted light. By this method, particles which are considerably smaller than the wave-length of light may be recognized individually, for they reflect the light in all directions and are consequently self-illuminative. An ultramicroscope consists of an intense Faraday- Tyndall cone which strikes a microscope with a special . attachment to make colloid particles visible. Water for ultramicroscopy—Water with the least 1See Wo. Ostwald, The World of Neglected Dimensions, 8th edition (Dresden, 1922), p. 72. t 58 PRACTICAL’ COLTOID: CHEMiSai amount of optical impurities is obtained by storing a large volume of distilled water at a uniform temperature for a long period of time and then siphoning off the upper portion of the water. Glass and hard rubber stoppers give ‘‘ optical dust ’’ which may be prevented by covering the stoppers with paraffin or tinfoil. Ultrafilters im- prove the water considerably, especially if the water is carefully excluded from the air after filtration. The number of dust particles present is usually small and the experimenter soon learns to recognize their presence. Expt. 86. Suspensoids—Ultramicroscopic experi- ments may be easily performed on the separate particles of suspensoid colloids, if the distance between each particle is relatively great. The colloidal solution must be very dilute so that the particles are at sufficient distances from each other, to obviate mutual reflection phenomena which would blur the particles when viewed individually against a dark background. Mastic hydrosol—A very dilute sol prepared accord- ing to Expt. 1 shows a large number of white, intensely illuminated particles in rapid Brownian movement. The interference of aggregated particles is certainly appreciated if observations are made after alternate additions of the concentrated solution and distilled water. A dark back- ground is necessary for the easy detection of the individual particles. Black India ink gives an image similar to mastic sol when using a much greater dilution. The background cannot be made so dark because of the presence of hydrated or protective colloids. Both the black and colourless particles reflect white light. Gold hydrosol—Observe and compare ultramicro- scopically the red and blue gold sols prepared according to Expts. 2-7. Smaller particles are usually found in the red sol rather than in the blue sol unless the red sol is Piety PEUIGAL PROPERTIES 59 prepared in the presence of protective colloids. Red sols occasionally show larger particles, which probably are soluble aggregates of smaller particles. A highly disperse dilute gold sol cannot be further resolved ultramicro- scopically, but gives only a diffused light cone. Other experiments on suspensoids—Colloidal silver and other metal sols! ; Prussian blue, metallic sulphides, organic dyes such as indigo, alkali blue, alizarin in paste form are suspensoids which are suitable for ultramicro- scopic experiments. Quantitative studies of the dimensions of particles cannot be based upon the size and intensity of the illuminated spots observed. For experiments’ on approximate determinations of particle size, see H. Siedenkopf and R. Zsigmondy, Ann. d. Phys., 10, 16 (1903) ; G. Wiegner, Koll. Beth., 2, 213 (1911). Particles of miscroscopic dimensions, those of 0-2” and more in diameter, are recognizable by the formation of refraction figures, such as concentric circles, V- or Y-shaped light haloes and other complicated light figures. Typical colloidal suspensoid particles produce comparatively bright and approximately circular light areas. Emulsoids—Non-hydrated emulsoids such as_ oil- water sols (Expt. 1) give the same ultra-images as suspen- soids. The difference in optical constants between the disperse phase and dispersion medium, which are neces- sary for ultramicroscopic recognition of the particles, disappears -with increasing hydration. The undiffer- entiated light cones always become dimmer with increas- ing hydration of the suspensoid and may practically dis- appear. Such negative results are due to smaller particles in the emulsoid previous to hydration. One must dis- tinguish between an optical amicroscopy and a dimen- sional amicrescopy. A negative ultramicroscopic obser- 1 Compare the preparations given in Chapter IX. 69 PRACTICAL COLEOID CHER ai vation does not prove the presence of a highly disperse emulsoid. Many solutions may be recognized as colloid systems in consequence of diffusion, dialysis, ultra- filtration, etc., yet they give only a negative or diffuse ultramicroscopic image. Starch paste is a good example of such an emulsoid. Ferric hydroxide sols—These sols are transitions between suspensoids and emulsoids, and illustrate emul- soid properties even better than solutions of egg white, gelatin, etc. Ferric hydroxide does not contain so many coarse impurities as the viscous albumin and gelatin sols. Sols which prove to be typical colloids by dialysis or ultrafiltration show at first an increase in light intensity, but on further dilution they show a light cone not resolv- able into single particles. The commercial Fe(OH), or preferably that prepared in Expts. 21 and 22, is suitable for these experiments. Solutions of egg white, gelatin, silicic acid, stannic acid, etc., usually show a diffuse cone containing much “optical dust.’”’1 A 0-5-I-0 per cent. solution of potato starch which has been heated at 100° C. for 30 minutes appears relatively clear. Starch paste shows a consider- ably diffuse light cone, in which only few “ dust ’’ particles are imbedded. If the starch sols are heated at 100° C. for about Io minutes, they coagulate and show a greyish- white irresolvable light cone. Nevertheless, the starch particles are not so coarse to settle on standing. This is an example of the above-mentioned fallacy of assuming dimensional amicroscopy from optically clear images given by sols whose particles are irresolvable. In previous experiments the starch particles were apparently so 1 Optical and chemical purity are not necessarily equivalent. Sodium hydroxide may be prepared with the greatest precaution from fresh metallic sodium and yet may not be optically as clear as an old caustic solution whose impurities have settled. OPTICAL PROPERTIES 61 strongly hydrated, especially in their external layers, and were so thickly aggregated that the optical transition between disperse phase and dispersion medium was practically constant. Dyes such as safronine, night blue, etc., usually contain so many impurities that they tend to destroy the typical image of an irresolvable light cone. Freshly prepared, very dilute silicic acid and solutions which contain great excess of acid appear optically clear. This also applies to serum and egg albumin solutions to which a few drops of HCl or NaOH have been added. ViiaoonOsCOPIC CHANGES OF STATE The large number of experiments on_ turbidity phenomena described in the previous paragraphs (77-85) may also be performed ultramicroscopically by observing changes of state in colloidal solutions. Expt. 87. Ultramicroscopy of gelation—According to the experiences of the author a solution of “ pure”’ gelatin is essential for an accurate study of the changes of state. This solution is best prepared by washing a 2-3 per cent. gelatin for several days. This lukewarm ! solution shows a grey-white Tyndall cone containing many impurities and showing Brownian movement. Choose an ultramicroscope bulb with a clesed stopcock, wash with alcohol to prevent the cloud- ing of the window, fill it with gelatin solution and place in an ice-chest for 24 hours. The ultramicroscopic imege first shows a considerably increased illumination of the whole light cone; later, a great number of light rays appear, which orient themselves to form a coarse structure in an apparently regular manner. 1 At higher temperatures the sealing wax which binds the cover- glasses to the cuvette melts. 62 PRACTICAL, COELOID CHEMTSii For a control experiment,! liquefy the gel by moistening the bulb with water at 40°-50° C. for a few minutes. Observe the considerable decrease in the intensity of the light cone, the disappearance of the Brownian movement. These effects may also be observed when the gelatin solution is heated to a higher temperature in order to completely disintegrate the coarse aggregates formed by gelation. By studying the process of gelation with a strong lighting apparatus, one may observe the gradual transitions from Brownian movement to oscillatory motion to complete immobility of the particles, and finally, to the formation of larger aggregates (W. Menz, W. Bachmann). Expt. 88. Ultramicroscopy of the ageing of starch pastes—Ultramicrcscopic observations may be made together with viscosimetric experiments on the ageing of a I per cent. starch paste (Expt. 75) so as to co-ordinate changes in turbidity with the variations in its viscosity. The presence of particles possessing Brownian movement is especially evident with the cccurrence of complexes in aged starch sols. Brownian movement is absent or only very slight in cold freshly prepared starch pastes. Expt. 89. Ultramicroscopy during coalescence— Fill the ultramicroscope with a very dilute mastic hydrosol or black India ink. Copy the image and count the number of particles in a portion of the field bounded by an ocular grating. Use the centre of the optical field for the observations. Add to the bulb about two drops of HCl or BaCl, and thoroughly mix the contents by pouring the whole solution into a small beaker and back again into the bulb. After standing a few minutes there is a decrease in the Brownian movement with the formation of larger 1 The cuvette should be carefully washed after these experi- ments by means of tepid warm solutions of KI, KCN or KCNS, which are good solvents of gelatin. OPTICAL PROPERTIES 63 irregular aggregates. Another count shows a decrease in the total number of particles as a result of aggregation. Similar experiments with red gold sols show a change of colour when the particles aggregate and hence decrease in number (Expt. 60). The coagulation of dilute ferric hydroxide sols with a drop of NaOH is striking. Slightly disperse granular particles are seen to coalesce in groups which become quite distinct from one another and finally unite to form very large flakes. A similar coagulation process may be observed by flocculating 0-or per cent. sols of Congo rubin and antimony sulphide (Expt. 13). Toe ONS Or THE PLANE OF POLARIZED PiGat BY COLLOIDS Hydrated colloids such as egg white, gelatin, tannin, starch paste, etc., strongly rotate the plane of polarized light. This phenomenon is very interesting and yet has been little investigated. An ordinary saccharimeter with a sodium flame, intensely illuminated to overcome the turbidity of the solution, is suitable for observing this phenomenon. Expt 90. Optical rotation by gelatin solutions (H. Trunkel)—The degree of optical rotation of a gelatin solution is as variable a property as its viscosity. The degree of rotation increases with increasing concentra- tion, decreasing temperature and is also dependent upon the age of the gelatin solution. The effect of age upon optical rotation may be determined as follows: Fill a 200 mm. polarimeter tube with a freshly prepared gelatin solution of a concentration such that the intensity of the light source may suffice. The experiment should be per- formed at constant temperature especially if the polari- meter 1s provided with a water-jacket. Record the degrees of optical rotation every hour and observe the constant 64 PRACTICAL ‘COLLOID CHEMIST increase at high concentrations and low temperatures. The degree of rotation generally reaches a maximum after 2-4 days. The following results were obtained at room tempera- ture by using a clear 5 per cent. solution of gelatin, a 200 mm. polarimeter tube and sodium flame :— t a Rotation. A Time in Hours. | Observed Degree. | Calculated Degree.| 12 Per cent. 0°47 7:97 7°97 0-0 A hy (eer 7°94 oa 2°48 9°72 7°49 + 0°23 Bez IO'l5 10°09 -+ 0:06 4°47 10°27 10°42 — o13 5°55 10°63 10°79 — O16 The calculations conform to the equation a= K¢?”, in which ¢ is the time and K and m are constants. Expt. 91. Optical properties of vanadium pent- oxide sols (H. Diesselhorst and H. Freundlich)—The freshly prepared sol, Expt. 33, does not appear to have optical properties. A vanadium pentoxide sol, after standing a few weeks, when stirred with a glass rod, shows silky streaks. These streaks are yellow in reflected light and dark in transmitted light. The sol, which can be diluted to give a bright brown colour in transmitted light, is placed in a cuvette between two crossed Nicol prisms. On stirring, a brightening of the field of vision is observed, after which dark clouds reappear. A sol kept for six months is so sensitive that it glows on shght stirring. This is the characteristic behaviour of a fluid crystal. SEICAL PROPERTIES 65 COLOUR OF COLLOID SOLUTIONS The colour phenomena in colloid solutions are brought about in two ways. Colloid particles show a selective absorption of light rays possessing certain wave-lengths. Some of the colour phenomena are due to the small size of the particle, which radiates laterally a considerable amount of light. This type of radiation is selective and hence various colloids reflect different coloured light rays. These conditions of colloidal state account for the double colour phenomena often occurring in colloids, wherein the difference in colour depends whether the light is transmitted or reflected. Selective adsorption no doubt gives the colour to transmitted light, while selective radiation is responsible for the colour of reflected light. Often, the colour of the reflected light is complementary to the colour of the transmitted light. Selective adsorption and selective radiation, together with the degree of dispersion, the orientation as well as the shape and mass of the particles play a considerable rdle in the colour change. CeEOUnReOte COLOURLESS COLLOIDS ” Colourless substances are those which absorb ultra-red or ultra-violet light rays. If these substances are dis- persed in colourless dispersion media, they show the usual colour phenomenon of opalescence irrespective of their individual properties. These sols impart a yellow or red colour to reflected light. The best example of such an opalescence is the cloudy sky at sunrise and sun- set due to transmitted light and in the daylight due to light reflected from a dark background. This opalescence results from the retardation of the light of shorter wave- lengths by the colloid particles, that is, they retard the 5 66 PRACTICAL COLLOID, CHINES ian blue and violet rays more than the longer yellow and red rays. The yellow and red light rays pass through the colloid with the least amount of retardation by the particles, while the blue and violet rays are strongly refracted or radiated. Thus, hght is dispersed into its different wave-lengths so that the longer waves passing through produce absorption colours, while the shorter waves are deflected laterally, producing refraction colours. Opalescence may be distinguished from fluor- escence by the fact that opalescent light is, while fluor- escent light is not, polarized (Expt. 78). Expt. 92. Opalescent solutions—A beautiful opal- escent solution may be prepared in the following way: Pour about 50 c.c. of a o-1 per cent. alcoholic solution of mastic or colophonium into 200-300 c.c. of distilled water or prepare a sulphur sol according to Expt. II. Add in small portions 100-200 c.c. of boiling distilled water to about 50 c.c. of a 0-5-I-0 per cent. filtered solu- tion of dried egg albumin preserved in a 0-9 per cent. NaCl solution. To prepare a sol of fresh egg white, as in Expt. 103, beat the white to a foam and allow to stand. Separate the clear fluid from the membranous foam. Dilute this clear liquid with four times its volume of o-g per cent. NaCl solution and slowly add 300 c.c. of boiling water. Other beautiful colour phenomena obtained from colourless dispersoids are the so-called Christiansen diffraction colours observed in the NaCl gel prepared in Expt. 27. Furthermore, polymerized cinnamic acid ethyl esters ; many liquid crystals ; fine suspensions of glass, quartz, NaCl etc., in mixtures of organic solvents, which have almost the same coefficient of refraction, also show these colour phenomena. See B. C. Christiansen, Ann. d. Physik. 23,298 (1884) ; 24, 439 (1885). Asimple example of this phenomena is an aqueous saturated OPTICAL PROPERTIES 67 solution of H.S, in which decomposition has begun. By holding the coalescing solution against the light, a distinct violet absorption colour may be observed. NaCl gel produces the two colours, yellow and bluish green ; and cinnamic ester the colours green and red at room temperature, and yellow and blue at higher temperatures. COLOUKS OF COLLOIDAL METALS Colloid metals show a great variation in colour phenomena. M. Faraday pointed out that the degree of dispersion is largely responsible for colour formation. Expt. 93. Polychromism of gold sols—Expts. 2-7 give the methods of preparation of red, violet, blue and green gold sols. A simple experiment which successively produces all the gold sol colours mentioned is similar to the method described in Expt. 3, using alcohol and a reducing agent. Use a large volume, 100-150 c.c. of boiling water, and add in the manner described above, 5-I0 c.c. of a o-or per cent. gold salt solution, and the same amount of alcohol. Warm the mixture until the red colour is developed. Pour a test portion of the hot sol into a small Erlenmeyer. Keep the main portion boiling continually and add, drop by drop, more gold salt solution, without any addition of alcohol. Continue to heat until a violet to blue sol appears and remove another test portion. To prepare a green sol, add 10-20 c.c. more gold salt solution to the remaining 50—I00 c.c. of hot violet sol. Observe that in such a series of variously coloured gold sols, prepared from the same original solution, the increase in turbidity follows the order of colour change from red to green. Expt. 94. Polychromism of silver sols—A series of coloured silver sols ranging from bright yellow, through the various shadings of red to blue and bluish black, 68 PRACTICAL COLLOID: CHEMISi is prepared in the following manner: First prepare three solutions: 0o-oIN AgNO;, 0-o001M hydroquinone [C,H,(OH).] and 0-o1N sodium citrate. Make the latter solution by titrating o-5N citric acid with an equal volume of 0:-5N NaOH until added phenolphthalein just assumes a pink tinge. All solutions, especially the AgNOs, should be neutral. Use freshly prepared hydroquinone solution. A very slight excess of alkali is necessary for reduction. Make the following preliminary experiments in order to standardize the hydroquinone solution. Add 2 c.c. of hydroquinone and 4 c.c. of sodium citrate to 2 c.c. of silver nitrate solution. If the mixture does not develop a faint yellow colour after Io seconds, add a drop of dilute NH,OH to 100 c.c. of the sodium citrate solution and repeat the experiment. Should the yellow colour not appear in 10 to 15 seconds, add two drops of NH,OH to the citrate solution and continue to do so until the desired reaction takes place. The presence of much NH,OH soon destroys the coloration by floccu- lating the silver sol. After standardization of the citrate solution, place 2 c.c. of the AgNO, solution into ten well-cleaned test-tubes, and add the following mixtures of hydroquinone and sodium citrate. Exp. No. I 2 iS 4 5 6 z 8 9 10 Hydro- quinone |5 drops.|7 drops,|10 drops.|14 drops.| I C.c.|1°4 ¢.c.| 2 ¢.c. |2°8c.c.| 4 C.c. | 5°6 C.c. Citrate .| 16. ¢:c. | r1¢.¢. | 8:c:c) | 5°6.¢.c. |4 C.C)2-8ie Cl aese em reac cee x4 drops If no reaction 1s evident in the first two or three tubes, a few more drops of NH,OH may be added to them with- out danger of flocculation. All the mixtures first assume a yellow or red colour, but they gradually develop a graded series of colours toward the blue of the last tube. After BEEiCAL PROPERTIES 69 3 hours, the tubes have the following colours: 1, bright yellow; 2, yellow; 3, orange-yellow; 4, orange; 5, red-orange; 6, red; 7, red-violet; 8, violet; 9, blue- violet; 10, blue. The last sol must be diluted with water in order to make the colour more distinct. The sols remain stable a few days, when all the colours gradu- ally change to blue. Green sols cannot be prepared in the cold by this method. Pour the above mixtures all together, stir and heat to boiling. A yellow sol appears before the colours fade. If solution 10, which finally becomes blue, is heated at the point where it is faintly red, a greenish coloured sol forms. Continued heating produces a yellow sol and finally changes it toa blue one. By this method, approximately any shade of green between blue and yellow may be prepared. Observe the increase in turbidity in the series ranging from yellow to blue. The continuous change in colour from yellow to blue corresponds to a change in the ab- sorption maximum of the shorter to longer wave-lengths with a decreasing degree of dispersion. This is a general phenomenon in colloid chemistry illustrating the relation between colour and degree of dispersion. Expt. 95. Polychromism of sulphur sols (R. Auerbach)—Mix Io c.c. of 1-33 per cent. phosphoric acid (Io c.c. of commercial H;PO, and dilute to 150 c.c.) with Pomc oieO05 1 Naj>.0,. After a few minutes, the absorption colour becomes yellow; reflected colour, blue; then the absorption colour becomes red; the reflected colour very turbid and greyish white. Later a dark blue or occasionally green shade appears, and finally flocculates to give a white, coarsely disperse sulphur sol. Expt. 96. Colour changes in gold sols during flocculation—tThe relation between colours of sols and 70 PRACTICAL “COLLOID CHEMIST the size of their particles is evident by the sudden change of red gold sol into a violet or blue upon addition of an electrolyte which produces flocculation. Place in large test-tubes or Erlenmeyer flasks equal volumes of a red gold sol and add to respective portions a drop of dilute HCl, NaCl, BaCl,, etc. After a few seconds, the red sol suddenly changes into a violet or blue sol. The occurrence of turbidity in the Tyndall cone, the ultramicroscopic image and the ultimate appearance of flocculation show that the gold sol forms greater complexes during the sudden colour transition. Reversible colour changes of colloidal gold in the presence of casein have been shown by R. Zsigmondy, Nachr. d. Gottinger Ges. d. Wiss., January, I9g16. The silver sol prepared in Expt. 94 shows the sudden change of colour upon flocculation with electrolytes. The colour of the silver sol after complete flocculation is usually black. Expt. 97. Colour changes in Congo-rubin sols— The particles of this dye sol have diameters between those of colloids and molecular dispersoids. It may be sud- denly transformed to a blue-violet or blue solution not only upon addition of acid but also by the addition of any neutral salt or even alkaline substances. The dye behaves like a red gold sol in many respects and it may be used as a gold sol substitute. The colour transition of Congo rubin is reversible by dilution, by raising the temperature, by addition of alcohol, etc. The colloidal changes in this dye are observed in the following ex- periment : Use a o-o1r per cent. solution of Congo rubin. Place 10-20 c.c. of Congo-rubin solution in a large number of test-tubes and add to respective tubes a few drops of some common chemical reagents. All electrolytes cause colour transition except strong alkalies and NH,OH. OPTICAL PROPERTIES 71 Warm the Congo-rubin solution, coloured blue-violet by a small amount of electrolyte, until the solution turns red. Pour half of this hot solution into a cold test-tube and cool further in a stream of water. The red solution turns blue once more, To a large beaker of distilled water add, drop by drop with constant stirring, a solution of Congo-rubin blue dye. After a few minutes the violet tinged solution changes to bright red. Add 2-3 c.c. of a blue Congo-rubin solution to an equal volume of methyl or ethyl alcohol. The red colour again appears with a simultaneous disappearance of turbidity. For further experiments with Congo rubin, see numbers 159, 177 and 179. Expt. 98. Colour and degree of dispersion— From the preceding experiments, the relations between colour and degree of dispersion of variously coloured gold, silver and Congo-rubin sols have been illustrated. This relation is important especially in the theory of colours of substances in the colloid state. The following experi- ments demonstrate this relation : Ultrafilter according to Expt. 57, red, blue, green gold sols, yellow, red and blue silver sols, red and blue rubin solutions. The first members of these series pass un- changed through the 2-4 per cent. collodion filter. How- ever, the ultrafilter retains the blue and green sols, while the behaviour of the intermediate series varies. A mixed colour sol is changed by ultrafiltration, so that a sol of another and purer colour constitutes the ultrafiltrate. Adda few drops of dilute acid or neutral salt, baryta, etc., to a Congo-rubin solution, so that a violet shade just appears, and ultrafilter this solution. The filtrate consists of the red coloured sol. Pour together the different coloured silver sols obtained in Expt. 94 and ultrafilter the mixture. A highly disperse 72 PRACTICAL COLLOID CHEMIST yellow silver sol is obtained as an ultrafiltrate from the dark grey mixture. The relation between colour tone, turbidity, and ultramicroscopic images of colloids was brought out in the above experiments. Compare the ultra-image of the red Congo-rubin solution, and the image of the blue solution, upon addition of electrolytes. Expt. 99. Ultramicroscopic colours—The colour of the particles visible in an ultramicroscope is due to the selective absorption, selective refraction and radiation. Lateral radiation colours of colloid particles may be observed microscopically in very dilute solutions. When using large quantities of concentrated solution, the colloid layer lying above the light cone acts as a light filter. The radiated colours of the single particles are altered by the absorption colours of the entire colloid. Concentrated ferric hydroxide sol studied under the ultramicroscope gives an intense brownish-yellow cone. If the same solution is diluted, the colour cone becomes greenish white. Likewise, a very concentrated red gold sol often gives a brown Tyndall cone and upon dilution a pure green. Concentrated Prussian blue sols show a violet cone and upon dilution give a_ brownish-yellow cone. The ultra-colours of colloid particles are often com- plimentary to their absorption colours provided very dilute solutions are compared with one another. Red gold sols usually give a green, blue or brown-yellow Tyndall cone. Polychrome ultra-images are often obtained when sols are examined under an ultramicroscope. A green-grey silver sol prepared by mixing coloured sols, has a wide range of particle size which gives a number of radiation colours. Complex reflection colours due to coarser micrcscopic particles may also be observed. Another meiealt) PROPERTIES 73 example is given by brownish-red, commercial colloidal selenium. A polydisperse system contains particles of various sizes which give a grey colour to the solution due to their individual absorption colours, when viewed by the naked eye. Ultramicroscopic inspection of this sol shows the individual colours. These phenomena indicate a close relationship between the colour of a sol and its degree of dispersion. V ELECTRICAL PROPERTIES OST colloid particles migrate when an electric M current is passed through their sols (Electro- phoresis). This indicates that the particles possess electric charges. From the capacity of trans- porting electricity it follows at the same time that colloid solutions must have a conductivity of their own, apart from the conductivity of the dispersion medium and the ordinary electrolytes contained in it. Typical colloid particles generally possess a relatively large number of unit charges, 30 to 40 in contrast to the ions of ordinary electrolytes (acids, bases and salts). Moreover, the charge on colloid particles can be both positive or negative. Therefore, colloidal gold and ferric hydroxide may occur as either anions or cations. Such is not the rule in mole- cular disperse solutions. The changes in conductivity with concentration, temperature, etc., partially follow laws different from those applicable to molecular dis- perse electrolytes. Transitions of “‘ colloid electrolytes ”’ (McBain) to “‘ molecular electrolytes ’’ may occur as in the case of ferric hydroxide sols, according to W. Pauli and J. Matula. The term “ion’’ which has been applied to electrically charged particles in gases must be used, 1 The exceptions are so-called ‘‘ amphoteric electrolytes,” as hydrolytic products of proteins (leucine, alanine, etc.), alkaloids, caffeine, theobromine, etc. 74 ERECTRICAL PROPERTIES 75 therefore, in a broader sense to include electrical charged colloid particles. The methods of studying the electrical charges of colloid particles may be arranged in the three following groups : U-tube method—The simplest apparatus for the detection of the sign of the charge on colloid particles consists of an U-shaped glass tube. The tube is 15 to 20 cm. high and has an inside diameter of 2 to 3 cm. Two pieces of platinum or silver wire, which are twisted into an horizontal spiral, serve as electrodes. These are stuck through two cork stoppers. Bore a hole in each stopper or cut’a groove along the side so that any gases which are formed may escape. Fill the tube with the colloid to be studied and pass a r10-volt direct current from the main through the solu- tion. If the colloid is coloured, electro- phoresis may be detected by a gradual disappearance of colour from one side of the tube after 15 minutes. The sign of the charge on the particle may also be deduced from this phenomenon. If the colloid is colourless, pass the current through it for 30 minutes. Shut off the current, remove the stoppers without disturbing the liquid and pipette off the upper- most 10-20 c.c. of liquid in each side of the tube. De- termine the concentration of the colloid in each portion of the liquid by the methods previously given. Accurate studies on the rates of electrophoresis may be made if polarization and. electrolysis are eliminated at the electrodes. Polarization may be overcome if an U-tube is used which has a constriction in the middle of both arms (Fig. 8).1_ Fill the lower bulb with colloidal solution to the upper ends of both capillaries. Prepare 1 U-tubes of this form have first been devised by J. Billitzer. Bice 3. 76 PRACTICAL COLLOID (CHEMIST icy, two small plugs of filter paper and place one in each capillary so that the tubes are stoppered. Invert the U-tube and wash out the upper parts of the arms with distilled water. Fill the rest of the tube with distilled water to serve as an electrode fluid. Since distilled water has a low conductivity, quite a length of time may elapse before considerable electrophoresis has taken place. A very dilute electrolyte makes a suitable electrode solution. If possible, such an electrode should be chosen which exists in small amounts in the colloid to be studied. Use a dilute solution of KI or AgNO, in the experiments on silver iodide sols, and in the study of ferric hydroxide sols, use a dilute solution of FeCls. The best method is to use the ultrafiltrate of the colloid studied, for the electrode fluid. The ultrafiltrate may be ob- tained in sufficient amounts by simple ultrafiltration. Itis not necessary that the electrode fluid touch the stoppers when they are in place. Maintain the fluid in both arms at the same level. Carefully re- move the paper plugs with tweezers. If this is done correctly, a sharp surface boundary remains in the con- striction of the arm. A tube of the dimensions given requires a current of 80 to 110 volts. Use the conversion apparatus given in Fig. 9 for more accurate studies. The middle part, 3, is filled with the colloid to be studied, until the liquid rises slightly above the stopcocks A and B. Close A and B and wash the cathode and anode arms 2 and 4. Fill the arm tubes with a conducting liquid ELECTRICAL PROPERTIES o whose concentration is the same as that of the colloid studied, in order to obviate diffusion potentials. Add the attachments 1 and 5, leaving the tap C open, and fill them with the same conducting liquid by means of a Io c.c. pipette. After removing all air bubbles, place the apparatus in a support and close stop- cock C. Add about 0-2 g. of CuCl, to the cathode solution and about 1 g. of NaCl to the anode solution so as to prevent polarization. Thoroughly mix the NaCl and the anode solution by moving the silver electrode to and fro. Use a plated copper foil or wire for the cathode and coat all but the lower end with paraffin. This precaution allows the current to enter the solution at points where the cathode is in contact with the dissolved copper salt. Before passing the current, open tap C, until the level of the fluids in both tubes is the same, then close again. Open A’and B and observe, every few minutes, whether a displacement of the middle layer occurs without the flow of current.. If such does not occur, the current may be started. ; Ultramicroscopic methods for investigation of electrophoresis—Measurements of the electrophoretic rate of displacement of individual particles are made by means of the ultramicroscope. Construct a small glass chamber on the object support of the microscope according to the process of The. Svedberg. Place two small pieces of glass on the right and left sides of the object support and two rectangular cover-glasses at the front and back. Seal all joints. A current of 4 to 6 volts is led into the chamber by two thin platinum or silver foils fastened on the right and left sides and connected to the conducting wires by pinch contacts. The chamber is finally closed with a cover-glass. Illumination from the front is fur- nished by an ultramicroscope. These experiments are 78 PRACTICAL COLLOUD CHESS. described by Svedberg and Anderson in Koll. Zeitschr., 24, 156 (1919). This method eliminates disturbing influences of the walls of the vessel. The electrophoretic rate of movement of the single particles may be observed and measured by means of a specially prepared microscopic bulb. Draw . out a piece of glass tube about I-5 cm. in diameter, so that the diameter of the constricted part is about equal to that of the bulb outlet. Bend the narrow part of the tube at right angles and cut off the larger end so that its upper end is level with the top of funnel attached to the bulb. Tightly fasten this tube to the bulb outlet with a short piece of rubber tubing and bind with wire to hold it in a perpendicular position. The bulb now has a funnel at both ends. Carefully introduce flexible o-5 mm. platinum or silver wires into the bulb through each of the two openings. The ends should be in the same plane and o-5-1-:0 cm. from each other. Use a current no greater than Io volts.1 Obviate overheating by passing the current through the sol by means of a pinch contact and only for short periods of time. Electro- phoresis is manifest by the particles, capable of Brownian movement, continually moving in one direction through the influence of the electric current. Try the experi- ment with a mastic sol prepared in Expt.1. Its particles are always charged negatively. Observe the direction of migration of the particles in the ultramicroscope. Measure the relative rates of certain particles passing over a given area, taking the time with a stop-watch. Other portions of the colloid solution may be studied by tipping the cuvette back and forth. This tipping will also mix the 1 The nearer the electrodes are to each other, the smaller the voltages necessary. Conversely, the effect may be increased with a given small voltage by carefully approximating the elec- trodes. PEeciRICAL PROPERTIES 79 products of electrolytic dissociation and render them harmless. The velocities of electrophoretic moving particles are calculated from the following formula : P 2a eS —, t a where s is the distance in centimetres covered in ¢ seconds, P the potential difference in volts, d the distance in cm. between the electrodes, 6 a constant characteristic for each sol, which may be defined as the electrophoretic mobility of the colloid particles studied. Experimentally pee Se a Be Its value usually hes between I and Io x 1073 ick cee sec. volt and is approximately of the same magnitude as the mobility of ordinary ions. Capillary method for the study of electrophoresis —This method involves the characteristic behaviour of certain colloidal solutions to rise by capillarity in filter paper and will be described in Expt. 105. Galvanic couples (W. Biltz)—Make a galvanic couple by soldering the ends of a strip of zinc 7 cm. long and r cm. wide to the end of a copper strip of the same size and bend both strips at right angles to the soldered joint. Dip the ends of the galvanic element into the sol to be studied. The suspensoid sol flocculates in a short time, the positively charged particles adhere to the copper strip or remain in its vicinity, while the negative particles adhere to the zinc strip. In strongly hydrated systems, such as egg-white solutions, the electrophoresis requires a longer time, but ultimately gives the same results. 80 PRACTICAL COLLOID CHEMISTRY The sign and magnitude of the electrical charges in colloidal solutions are extremely unstable and vary to a considerable extent. Investigations show that colloid particles bearing positive and negative charges, may occur in the same colloid solution. This will be demon- strated in the following electrophoretic experiments. Indefinite results are sometimes obtained due to the great variability of the electrical properties of the colloidal particles rather than to the method used. Expt. 100. Positive and negative colloids—Study the electrophoresis of 0:25 per cent. ferric hydroxide sol, prepared in Expts. 20-22 by the U-tube method. Dialysis of the sol may be performed in a parchment paper cup. Use parallel connections and perform the same experiment with mastic hydrosol (Expt. 1). A considerable repulsion of the ferric hydroxide colloid by the anode, occurs after Io to 15 minutes and the particles gather at the cathode. The ferric hydroxide sol is there- fore positively charged. Conversely, there is a repulsion of mastic sol by the cathode. The colloid is therefore negatively charged. Test with the galvanic couple described above, the type of electric charge on As,S,; and Fe(OH); sols pre- pared in Expts.. I2, 20-22, Examine the =precipiace removed from the inside of both arms of the couple with filter paper. The sulphide has migrated to the zinc strip and the hydroxide to the copper strip. Observe that with a short galvanic element, all posi- tively charged particles, ions as well as colloids, migrate to the nobler metal, in this case, copper. If such an element is used as a source of current for an electro- phoresis experiment in an U-tube, the motion of particles would occur as described in Fig. 10. Connect the platinum electrode with the zinc strip for a cathode and with the copper for an anode. The positive Fe(OH). Peel RICAL PROPERTIES 8I sol will migrate from the anode to the cathode. Colloidal hydroxides often possess positive charges; colloidal metals and sulphides, negative charges. Expt. lor. Changing the sign of the charge on colloid particles by varying the mode of preparation —Prepare two silver iodide or silver bromide sols in the following manner : molwA. = Dilute 3 cic. o-:IN, KI, or KBr with foc 0) In a second vessel dilute 5 feet. OIN AgNO, Mi ts: Pateeonc.c. of H.0. i With strong agitation, slowly pour the KI solu- tion into the AgNO, solution, and not in the reverse order. Sol B. Dilute 4c.c. of o-IN AgNO, with 15 c.c.of H.O. Inanother vessel, dilute 5 c.c. of o-IN KI with 40 c.c. of H.O. Pour the aque- ous AgNO, into the KI eS lunor Sol A is found to be positively, and sol B negatively charged when the electrophoresis is studied with an ultramicroscope. Theory states that the common ion, present in an excess when the colloid is formed, imparts its charge to the colloid particles. If the AgNO, is poured into the KI, the negative anion, I~, is first present in excess and 6 @ positive Colloid © negative - # FIG, ro. 82 PRACTICAL COLLOID CHEMTSa aa it imparts its negative charge to the AgI sol. If the KI is poured intothe AgNO,, the positively charged cation Ag is first present in excess and it imparts a positive charge to the Agl sol. Expt. 102. Positive and negative ferric hydroxide sols—The commercial ferric hydroxide sols and those prepared in Expt. 21 are positively charged according to electrophoresis experiments, that is, U-tube method. Negatively charged ferric hydroxide sols may be prepared in the following way: Add, drop by drop, a saturated solution of FeCl, to 100 c.c. of 2N ammonium carbonate, shake continually until the precipitate first formed has dissolved to give a dark reddish-brown, clear fluid. If the solution begins to darken, wait a few minutes, so as not to add an excess of FeCl;. This colloid eventually loses its electrical charge upon dialysis. Determine the sign of the charge on the colloid particles in an undialysed sol by capillary, U-tube method, etc. Pour 100 c.c. of o-orN FeCl, solution into 150 c.c. of 0-OIN NaOH solution. A bright yellow opalescent sol appears which seems to be less hydrolysed than the usual ferric hydroxide sols. It may be examined with an ultramicroscope. Expt. 103. Influence of (H*) and (OH) ions on the sign of the electrical charge on eg¢-white sus- pensoid particles (W. B. Hardy)—Prepare as in Expt. 92, a sol of fresh egg white coagulated by heat. Allow it to stand a few days so that it may better tolerate the addition of acid. This sol reacts neutral toward litmus paper, and when examined by the U-tube method, a weak negative charge is sometimes observed. ‘Acidify a small portion with acetic acid so that litmus paper shows a distinct acid reaction. Make another portion alkaline with NaOH. Perform a double experiment with these sols in two U-tubes or in two Michaelis BeECTRICAL PROPERTIES 83 apparatus. The acidified sol shows a decided cation migration and hence it is positively charged. Conversely, the alkaline sol shows a distinct anion migration hence, negatively charged. These sols are suitable for an ultramicroscopic study of electrophoresis. Expt. 104. Changes in the electric charges of ferric hydroxide sols by filtration (T. Malarski)— Filter a dilute, dialysed commercial ferric hydroxide sol five times. Use a fresh filter paper each time and add to it some shreds of filter paper in order to increase its effect. Study the electrophoresis of unfiltered and filter sols simultaneously, using parallel connections. The un- filtered sol shows a sharply defined clear portion in the vicinity of the anode arm ; the filtered sol shows none or only a diffuse brightening. The unfiltered solution gives a thick dark-brown precipitate; while the filtered sol gives an extremely voluminous bluish-white precipitate in the cathode arm. According to T. Malarski, Koll. Zeitschr., 23, 113 (1918), filtration of positive sols through negatively charged filter paper should first decrease the amount of positive electricity in the sol and finally charge it negatively. The effect of the filter paper on the electric charge of the sol may be accentuated by repeated filterings. Expt. 105. Detection of electrically charged colloid particles through capillarity (F. Fichtner, N. Sahlbom)—Allow some coloured hydrosols to ascend strips of filter paper. Different properties are observed depending upon the charge of the sol. Metals and sulphide sols show a considerable separation between disperse phase and dispersion medium in the rising portion of the sol. Ferric hydroxide sols prepared in Expt. 21 show a distinct separation of both phases after a short rise. The colourless dispersion medium continues to ascend while the dispersed phase remains behind, 84 PRACTICAL’ COLLOID CHEMI aii becomes concentrated and flocculates to form a sharp boundary. Fichtner and Sahlbom claim a negatively charged colloid will ascend the strip of filter paper un- separated from its dispersion medium, while the positively charged colloid is separated. The explanation of these various properties of colloids lies probably in the assump- tion that the filter paper wetted by the water carries a negative charge. While a negatively charged capillary adsorbs a similarly charged colloid particle unaffected, the oppositely charged particles are held back and collected so that they finally clog the capillary. Examine by capillarity a commercial ferric hydroxide sol and one preparedin Expt. 94. The former, positively charged, ascends only 1-2 cc. and flocculates, giving a sharp irregular-shaped border line; the negative sol ascends almost unseparated and at nearly the same rate as the dispersion medium. Examine by capillarity a 0-2 per cent. solution of “night blue’? and a o-2 per cent. solution of “alkali blue.’ Distinct differences are obtained if a concentrated “alkali blue’’ solution is compared with the dilute “night blue’’ solution. Addition of NaOH to “alkali blue’ and HCl to “night blue” emphasizes the differ- ences between the properties of the two dyes. While the “alkali blue” ascends readily and therefore appears negative, the “night blue”’ rises a short distance and precipitates. Therefore it appears to be positively charged. These differences may be detected more quickly by dropping some hydrosol on to a filter paper. Examine the two ferric hydroxide sols by this method. After the drops have spread, hold the filter paper against a light. A positive sol forms a broad colourless ring surrounding a coloured centre portion. A negative sol gives a spot which is uniformly coloured to the outer edge. : PeLeCERICAL PROPERTIES 85 Expt. 106. Capillarity with prepared filter paper —Soak some filter paper with Al(OH); sol, prepared in Expt. 32, and allow to dry in a warm place. Study the capillarity of this prepared paper. The Al(OH), sol is no longer positively charged. Therefore, the “ night blue’’ solution separates upon absorption, while the “alkali blue’”’ flocculates at the line of contact between the paper and liquid. The sign and magnitude of the electric charge on colloid particles are characteristics which are more stable than the degree of dispersion of the colloidal suspension. However, simple filtration through a filter paper (Expt. 104) may suffice to change these properties. The capillary method may be used with care for the detection of the kind of charge on colloid particles. VI EXPERIMENTS WITH GELS ELS are disperse systems which show both solid and liquid properties. As solid bodies, they possess a relative stability in shape and elasticity, especially toward rapid changes in form. However, they behave as fluids toward continued mechanical stress. They gradually assume the shape of any new container due to stress caused by their own weight. Diffusion of mole- cular disperse substances in dilute gels proceeds practic- ally with the same velocity as diffusion in a pure dispersion medium. Expt. 107. Mechanical properties of pastes—The changing liquid and solid properties of pastes may be used to conveniently demonstrate the properties of real gels. Grind 5 g. of potato starch in a mortar with 4 c.c. of water. Tip the mortar and allow the paste to flow in a continuous stream on to a glass plate. Smaller amounts give the drop form characteristic for liquids. Quickly rub the paste in a mortar with a spatula. The paste breaks into shell-shaped fragments showing sharp fracture surfaces. Therefore the paste behaves like a liquid by slowly altering its surface and behaves as a solid when a relatively large stress is applied. . Gels are formed as a result of change in the following properties of a colloid system: (1) concentration, (2) temperature, (3) degree of hydration usually attended by chemical changes, (4) formation of insoluble precipitates. 86 EXPERIMENTS WITH GELS 87 Many gels having great elasticity and temporary rigidity, such as gelatins, may be classed as liquid-liquid systems. There are probably gels having the structure liquid-solid and solid-liquid. Silicic acid gels show emulsoid_ pro- perties in the first stages of formation and on ageing they show a suspensoid structure. This structure corresponds to the changes of elasticity with the age and to the first occurrence of crystalline Lauegrams in the aged jelly. ee ero LON Gelation includes the formation of a hydrated emulsoid, reversed by heating. Externally, the process of gelation corresponds to a large increase in the viscosity of the liquid sol and associated with the properties of solid substances such as displacement, elasticity, rigidity of form, etc. The process of gelation may be expressed in the following terms: (1) Time of gelation. Concen- tration and temperature being constant, the gel formation, similar to all colloidal changes of state, requires a certain time. (2) Gelation concentration at a certain time and temperature. Jelly formation occurs only above a cer- tain concentration. (3) Solidifying temperature. Time and concentration being constant, jelly formation occurs only below a certain temperature. A fourth stipulation may include the softening temperature of gels, which is usually higher than the solidifying temperature. There- fore, these softening temperatures do not coincide with the melting and freezing temperatures of crystalline sub- stances. Viscosimetry of dilute gelated solutions is a convenient method for studying the process of gelation. The pro- perties of concentrated solutions may be extrapolated from these results with considerable accuracy. The properties of a gel may be defined in terms of the condi- 88 PRACTICAL COLLOID CHENGSit: tions under which it exists—concentration, temperature, time, etc.,—at which the colloid will no longer flow from the container when it is inverted. Expt. 108. Determination of gelation concentra- tion and time—Prepare in the manner described in the chapter preceding Expt. 66 a 12 per cent. gelatin solution and dilute it with warm water so as to give solu- tions of 12, 8, 6, 4, 3, 2, 1°5 and 1-0 penec ape to c.c. of each solution in a test-tube, and place a ther- mometer in the test-tube containing the 4 per cent. solution. Quickly cool all the tubes to room temperature or to 10° with cold water. By carefully inclining the tubes held in a test-tube rack, determine the time elapsed before the gel in each tube ceases to flow. This experi- ment furnishes a series of periods of gelations at various concentrations. Plot a time-concentration curve. The plotted curve may be used to interpolate the gelation con- centration which gives a gel at room temperature in less than anhour. Repeat the same experiments by warming the gels on a water-bath or a hot plate. Plot the corre- sponding time-concentration curve of gelation at constant temperature and by interpolation find the concentration of the solution which will give a gel in an hour. The ‘‘normal’’ gelation concentration, that is, the concentration at which a solution solidifies within half an hour, depends upon the mode of preparation. The concentration of the solution is approximately 2 per cent. at 15° C. and usually nearer I per’cent- sata ane Perform the same experiments with agar, soap solution and caproic acid. Agar is a carbohydrate and unlike gelatin contains no albumin. The normal gelatin con- centration of agar is usually less than that of gelatin. Expt.109. Determination of solidifying and soften- 1 More accurate results are of course obtained by use of a thermostat. EXPERIMENTS WITH GELS 89 ing temperatures—The temperature at which a gelling solution solidifies or melts depends essentially upon the rate of temperature change. The slower a gel is cooled, the higher the temperature at which it solidifies. The slower a gel is warmed, the lower the temperature at which it melts. A constant rate of temperature change, such as 1° C. per minute, must be used in order to obtain com- parable results. A cooling rate of 40° to 30° C. in 10 minutes might be called the normal rate, while cooling in 50 minutes would be a o-2 normal rate of temperature change. This rate of temperature change, 1° C. Liquefaction in 5 minutes, is especially desirable for the following experiments : Use thin-walled test-tubes of equal Solidification size or if possible, metal test-tubes. Provide a water-bath (glass tank of 2 to 3 litres capacity) with stirrer and fill the tank with water at about 40° C. Punch holes of such a size in a piece of stiff paper or sheet metal so that it rigidly holds the test-tubes. Place the perforated sheet above the water-bath. Liquefy the series of eight gelatin solutions used in the previous experiment by placing them all in the water-bath heated to 60° or 70° C. Leave the tubes in the bath and allow it to cool to 40°C. Place a thermometer in each test-tube and determine to within 0-1° C. when the temperatures of the solutions have reached that of the water-bath. Prepare one vessel containing ice and water and another containing hot water. After the contents of the test-tubes have reached the temperature of the water- bath, note the time and with constant stirring, cool the bath at a rate of 2° C.in 5 minutes. Regulate the cooling Temperature —> Concentration —> PIG. iT. go PRACTICAL’ COLLOID CHEMistih by means of cold and hot water. Determine at what temperature the contents of each tube solidifies by fre- quently inverting them. Plot the concentrations against the solidification temperatures. Determination of the liquefying temperature— - Allow all the tubes to stand overnight in an ice-chest, and the following day place them in a water-bath at 5°C. When the contents of the tubes attain the tempera- ture of the water-bath, heat it at a rate of 2° C. per 5 minutes. Determine the temperatures at which the gels soften in the manner described, and plot the concentra- tion-temperature curve. Comparing the solidifying and liquefying temperature- concentration curves, it appears that the liquefying tem- peratures are far above the solidifying temperatures (Fig. Ir). The following table gives an experiment personally con- ducted wee the author :— Solidification. Liquefaction. | L—Ss es | Graphic In- ‘Solid. Temp.|Solid. Time Lig. ~ Temp. Liq. Time | terpolation. =S, Minutes. | = L, Minutes. | 12 tee2T 5 O |: SSkaaeee 63 OI 8 | 19:9 8 | 2626 60 7:0 6 | 179 18 de Sauee 55 8-3 4 14:0(?) 37 | ae 45 9°5 3 12°5 49 | ora 30 10°8 2 6:9 73 [- boee 19 II‘9g Be aESS eee seen os Oo 13°6 I | — —— = eae ere | The difference between solidification and liquefying temperatures of agar gels is considerably greater than for gelatin gels. EXPERIMENTS WITH GELS QI Expt. 110. Influence of preliminary thermal treatment on gelation—Divide a 12 per cent. gelatin solution into two portions. Place one portion in an ice- chest and the other in an Erlenmeyer flask. Fit the flask with a cork stopper provided with a capillary tube to prevent vaporization and set in a warm place.! After 24 hours bring both solutions to a temperature of 40° C. The gel formed in the ice-chest quickly liquefies and the heated solution cools. Determine, as in Expts. 108 and 109, the time of gelation as well as the solidifying and liquefying temperatures. Considerable difference exists between the properties of these solutions. This phenomenon is readily observed if a 2-3 per cent. gelatin solution is used. These previously treated solutions cooled to room temperature show a distinct difference in the time of gelation. The cooled portion solidifies in 1-2 hours, while the heated solution of the same concentration requires many times that amount for solidification. “Expt. 111. Influence of acids and alkalies on selation—Prepare the following gelatin solutions : A. .9 c.c. of 3 per cent. gelatin solution + 1c.c. H,O BeG.C:.c- ie fs - + 1cC.c. DC! . 20 Giese. 7 Af A +1c.c. 2N HCl 0. C.C x a + 1 c.c. N Na0H Io OQ G.c: Me rs sa + 1c.c. 2N NaOH Place all the solutions in a water-bath at 40° C. for 10-20 minutes and determine the time of gelation or the gelation temperature as in previous experiments. These additions of acids and alkalies retard gelation measured in the manner described, yet the smaller con- 1 The same result may be obtained by initially heating to boiling and packing the vessel adequately in cotton, paper, sawdust, etc., in accordance with the principle of the fireless cooker. Q2 PRACTICAL COLLOID CHEMiat centrations also produce an increase in viscosity accord- ing to Expt. 71. It is yet to be determined by a more detailed study of gelation whether small concentrations of acids and alkalies exert a gelating influence, or whether an extrapolation of viscosity measurements on the process of gelation is incorrect. Expt. 112. Influence of salts upon gelation— Prepare the following series : 10 c.c. of 6 per cent. gelatin solutions and 10 c.c. of normal solutions of the salts: potassium sulphate, citrate, oxalate, chlorate,’ chloride, carbonate, nitrate, bromide, cyanide, sulphocyanide, iodide, salicylate, etc., sulphates of potassium, sodium, ammonium, magnesium, calcium,? aluminium, zinc, copper, iron, etc. Maintain all tubes at 40°C. for Io to 20 minutes and allow to cool to room temperature. Determine the time required for gelation after the solutions have reached room temperature. These salts give the following series when arranged according to increasing time of solidifica- tion : The potassium salts affect the time of gelation in the order given above ; potassium carbonate and the salts preceding it in the series increase the rate of gelation to a greater extent than those following the carbonate when compared with the gelation solution as a control. Cyanides, sulphocyanides, iodides and salicylates of the concentrations given above and at higher concentrations practically retard gelation. These salt-effect series are obtained in weak acid as well as in weak alkaline solutions of gelatin. The cations with sulphate as anion markedly decrease the rate of gelation. There appears to be no sulphate 1 Since KCl1O, is not so soluble, the 0-6 g. of gelatin should be dissolved in 20 c.c. of 0-5N KCIO, solution. 2 Saturated CaSO, = 0:03N at 20°C. EXPERIMENTS WITH GELS 93 which will retard gelation at this concentration. The cation effect in individual cases varies with the concen- tration of the salt as well as with the acid or basic reaction of the gelatin solution. The cation effect for weakly acid gelatin solutions gives the series : ode vie, Cu, K, NH,, Al, Fe; and for weakly alkaline gelatin solutions, eemeenescdsoNa, Me, NH,, Al, Fe, K. The salts used were all 0-5N concentration except for CaSO,, which was a saturated (0-2 per cent.) solution. This ionic series, representing their relation effect on gelatin, occurs in colloid chemistry as well as in general physical chemistry and is referred to as the Hofmeister series. Expt.113. Influence of non-electrolytes on gela- tion—Determine the time of gelation of the following Hyieiiitess(>..|.Levites) : + I g. urea + 1 g. thiourea + 1 g. furfural + 1 g. chloral hydrate. + 1g. methyl alcohol + 1 g. ethyl alcohol + 1 g. propyl alcohol -- 1 g. isobutyl alcohol. C. 9 c.c. of 6 per cent. gelatin solution -++ I g. cane sugar. D. 9c.c. of 6 per cent. gelatin solution without addition (control). AYO ¢.c. of 6 per cent. gelatin tn B. 9 c.c. of 6 per cent. gelatin ton The mixtures in series A lengthen the time of gelation, compared with the control. In series B the higher alcohols likewise retard gelation in proportion to increase in molecular weight. Cane sugar at the concentration given accelerates gelation. For the theory of gelation, see Expts. 85 and 87. 94 PRACTICAL COLLOID CHEMISTRY B. SWELLING Swelling involves the absorption of a liquid by a solid to forma gel. The process of swelling, like other changes in colloidal state, requires time. The rate of absorption of the liquid is greater at first but gradually decreases. A substance capable of swelling cannot absorb an unlimited amount of liquid—there is always a swelling maxi- A B mum. Swelling usually occurs only within a cer- tain temperature range, beyond which the absorb- ing substance changes into a colloidal solution. Gelatin dissolves in boil- | ing water without first forming a gel. Solution = 100 ¢.c. C= c Equili- x Original briumfcon-|) Amount concentra- centration | adsorbed tion in after 11¢.C. joe c.c.ofo-1N; adsorp- of o-1N (c—x) Kerem oh NaOH ition in c.c..NaOH per per 100 of o-1N_ |100 c.c. of Experi- Calcu- c.c. of | NaOH per} solution. mental. lated. solution. |100 c.c. of filtrate. 12:0 13 4:7 | 0863 | 0°672 4°7 4°5 26:0 18-9 yet E277.) 0-851 he 7*I 52°3 42°3 10-0 1'626 I*000 10-0 10°5 105°2 90°5 14°7 T°957 | 1167 | 14°7 15°3 220°8 198-2 22:6 Z°207 1°354 22:6 224 444°8 AII‘7 33°1 2°O15 I°520 eta 32°1 / foots, nh = 0°40. SeareOd ==. 2, ;) Solution = 100 C.c. L272 3°6 8-6 0°556 | 0°934 8:6 9:0 24°4 10°8 13°6 1°033 1°134 13°6 14°4 49°4 27°8 224 1°431 I°350 2254 21°4 100°4 66:8 33°06 1°825 1°526 33°06 BEtO 198°5 I51°5 47°0 2:180 12672 47°O 45°0 385°8 320°6 59°2 2°514 1+772 59:2 62:6 Fee STS. n=O Ag. equation represents cases which seldom occur in practice and assumes that only the disperse phase and not the dispersion medium is adsorbed. Nevertheless, such is the general case. A discussion of this equation is given by Wo. Ostwald and R. de Izaguirre in Koll. Zeitschr., 30, 279 (1922). Expt. 144. Adsorption of Crystal ponceau and methylene blue by wool; L. Pelet-Jolivet (electro- 126 PRACTICAL’ COLLOID” CHEMI iia chemical adsorption)—Use a 0-2 per cent. aqueous solu- tion of crystal ponceau and a 0-05 per cent. solution of methylene blue to prepare the following mixtures : Crystal ponceau: 1. ro c.c. of solution. DE ONGEC te », -+10drops2N HCl. Ast LO CzOver, » +10 drops 2N NaOH. Methylene blue: 4. Io c.c. of solution. 5. IOCc. ,,° 4, — igharops Netier 6. IOC.C. ,,. 4,5) == LOU eee Place a few white wool fibres in these mixtures, leave them 20 to 30 minutes at room temperature and then wash thoroughly with cold water. Observe that the wool fibres are not affected by the basic or neutral solutions but are distinctly coloured by the acidified dye, crystal ponceau. The reverse occurs in the methylene blue solution. There is an indistinct colour adsorbed from the fibre by the acid dye, a distinct colour with the neutral dye and a very intense colour with the alkaline solution. These results may be explained by electrical or electro- chemical considerations. The wool is charged negatively in the presence of an excess of (OH) ions and positively in excess of (H*) ions, This is the caseuwithma is relatively indifferent substances (J. Perrin). Crystal ponceau is an acid and methylene blue a basic dye. If the wool carries a definite positive charge in acid medium, the oppositely charged anion of the acid dye, crystal ponceau, isadsorbed. Conversely, in an alkaline medium, the wool is negatively charged and the positively charged cation of the basic dye, methylene blue, is adsorbed. Wool is a weak acid and exists in water in the form of a massive negatively charged anion in association with the hydrogen ion. In a neutral medium, the wool should show an alkaline reaction toward both dyes ; the experi- ADSORPTION 127 ment shows this to be the case.!' In contrast with the adsorption of acetic acid by charcoal, this electro-chemical adsorption is irreversible (L. Michaelis). Use filter paper strips instead of wool. Choose a shorter time of reaction or a more dilute solution, else the difference in the reaction is insignificant. Expt. 145. Specific dye adsorption by silicic acid and aluminium hydroxide gels—The pure commercial preparation “‘ asmosil’’ or Patrick’s “ silica gel’’ may be used as the SiO, gel, while the commercial native alumina is a suitable Al(OH), gel. These gels may be prepared by pouring water glass into concentrated HCl and aluminium chloride into ammonium hydroxide. Thor- oughly wash the two gels. Fill two test-tubes or flasks with a few grams of SiO, gel, two others with Al(OH); gel. Test the properties of the gels with a 0-01 per cent. solution of methylene blue and a corresponding solution of “‘ patent blue.”’ Allow the mixtures to stand a while and decant or wash into a filter. The silicic acid gel has irreversibly adsorbed the “‘ methylene blue ’’ and has not at all or only faintly adsorbed the “ patent blue.” Con- versely, the aluminium hydroxide gel has_ strongly adsorbed the “ patent blue,” but has only faintly adsorbed the methylene blue. B. ADSORPTION AT THE INTERFACE OF TWO LIQUIDS Expt.146. Adsorption of colloidal copper sulphide at the interface, water-chloroform (W. Biltz)—Pre- pare two copper sulphide hydrosols as follows : A. Add 1-2 c.c. of a dilute solution of copper sulphate 1 Further data is given by L. Pelet-Jolivet and co-workers in Koll. Z., 2, 225 (1905), which contains a complete bibliography, or in The Theory of Dyeing, by the same author, Dresden, IgTto. 128 PRACTICAL COLLOID CHEMISTRY or copper chloride to a mixture of 90 c.c. H,O + Io c.c. of freshly saturated hydrogen sulphide water, until a weakly turbid dark-brown sulphide sol is attained. B. Toasimilar mixture of water and hydrogen sulphide, add drop by drop, 1-2 c.c. of a dilute solution of copper ammonium hydroxide. Prepare the latter by mixing the CuSO, solution used in A with NH,OH until the resulting precipitate again dissolves to give a clear solution. A dark-brown sol is obtained, but it is clear in contrast to the sol prepared in A. Shake 15 c.c. of both sols with 2-3 c.c. of chloroform. Sol A is completely decolorized after shaking for a few seconds. The copper sulphide is adsorbed at the inter- face, water-chloroform, and sinks with the chloroform. Sol B is only slightly decolorized after much longer shaking. The reaction of the sols is the cause of this difference in behaviour. Sol A is acid and sol B is alkaline as a result of their modes of preparation. If sol B is weakly acidified ~so that it is not flocculated within 1 to 2 hours, it may be adsorbed like sol A. To obtain the sol in the upper layer of the mixture, the experiment should be performed with benzol. Expt. 147. Adsorption of gelatin at the interface, water -benzol—Prepare a dilute gelatin solution which gives a distinct precipitate with the tannin test—IOo c.c. of solution + 1 c.c. of 10 per cent. tannin solution + I c.c. dilute H,SO,. ) 2) —— 50 >”? 2:0 €.C. ae +8:-0 ss ye =—100 f AsO C1 es -+-6:0 * is ==200 (f) % K, citrate. 0:6c.c.2°0N K, citrate+-9°4 c.c.H,O-+10c.c.As,S, sol=62+5 milli- mols 1°25 C.c, ;, +8-75 _ * == 1255 Pao bs - +7+5 i) Fs ca Cees 50 C.C. ¥ +5:0 fi A seo SOO ere The concentration of the electrolyte which causes a distinct turbidity immediately or after one hour may be regarded as the flocculation value. Thus, in the previous experiment, the sol did not immediately become turbid with KCl at 25 millimolar concentration, but distinctly so with a concentration of 50 millimols. The flocculation value les between the two concentrations and may be accurately determined by the use of inter- mediate concentrations. The figures in heavy type show that in the above example, the flocculation values of the three chlorides are approximately as 80:0-8:0-08 or as 1000: 10: TI. The trivalent Al*** ion is by far the strongest electrolyte, since the least concentration is required to flocculate the COAGULATION AND PEPTIZATION 139 sol. K.SO, and potassium citrate behave like KCl and their flocculation values are approximately of the same magnitude. -HCl is not really a strong flocculant and remains far below the flocculation values for CaCl, and AICl;. These results indicate that the cation or positively charged ions are responsible for the flocculation values for a negative arsenic trisulphide sol and that the activity of the cations greatly increases with their valence. The following tables show the flocculation values of other electrolytes. It indicates also that the valence of the cation does not play the only role. Determine, for example, the flocculation value of the univalent morphine chloride (mol. wt. 321°5). Other sulphide sols, negatively charged metal sols and mastic sols, behave like the As.5; sols. HILOGCCULATION OF AS,5, SOLS. According to ite Le and H. Picton. According to According to H. Schulze. Electrolyte. H. Freundlich. Univalent cations, c = millimols per litre. CAcciG acidu. .. ~ | “Ca.-14900) | — —- Prete ee) ss | Ca. 1290) — — (eoraliteachl | 4.) ‘Ca, 373) — — etipeti eee es | Ca. 275) — —- eelomeltrate., .°.. = _ > 240 Meacetatee ihn i.) S| — — IIO PEDO en ss | — 124°4 — TG Ns oe a —- 109°0 — [ELON ar A i 185°4 — 58-4 4+ K,Fe(CN), eee | 181-2 — = Sodium acetate > . | 154°3 — — LNG SCs Saas 1510 123° 65°60 % pot. oxalate . . | 13172 ~ — COE = LT7°6 1047 50:0 Ss) 0 aia 1090 137°4 — LGR) Sie 107°3 102°2 — eee we a, 5 | — 1 7°O — * Kgtartrate. . . | 104°3 —- — 140 PRACTICAL COLLOID: CHEMIS ties FLOCCULATION OF AS,S3 SOLS (continued). Electrolyte. According to H. Schulze. According to S. E. Linder and H. Picton. According to H. Freundlich. 4 K,Fe(CN), NaNO, KCl KO. NH INOS. NH,I NH,Br NETGG! Pers NaBr. Na Cl ena et } (NH,),SO,. 4+ 11,50; . HNO, EtG] HI HBr : Guanidine nitrate 1135074 ; Strychnine nitrate . Aniline chloride. p.-Chloraniline chlor- 1G ae) a Morphine chloride New fuchsine Bivalent cations, c = millimols per litre. MgsoO, : [Pe(NH,4) (SO4)» MnsSO, ; Bes; Cos0, ZnO, NiSO, CaSO, Nil. CdCl, Feces Co(NO3)» 100°5 100°4 979 277 205 Bao 3°03] 2% zy ty | 1°86 1°88 2°64 T1o0°8 [97-9] 73°9 73°9 73°9 62:9 IOI‘O I09:0 103°5 95°8 92°4 ues 58°7 5755 56-0 I:60 —- 2°10 2°02 2°02 I-96 1°68 1°65 I-60 [752 1-46 1°42 1°37 30°1 30°8 16°4 8-0 2:52 T°08 0°425 O-114 o-8Io COAGULATION AND PEPTIZATION I4I FLOCCULATION OF As,S; SOLS (continued). Electrolyte. According to H. Schulze. Zils CaCl, : Ca(HCO.,) >» Gabre MgBr, CoCl, Sr(NO,), Ca(NO,), pit Gans Cu(NOs) > BaCl, MgCl, Ba(NOs)¢ CdCl, UO,(NO3). CdBr, CdSO, CuSO, Al(NO,), NH ,Fe(SO,), KCr(SO,)» KAI(SO,) » KFe(SO,)» NH,Al(SO,), According to Sp. Linder | and H. Picton. According to H. Freundlich. Trivalent ca 0-310 0-123 1°34 feo +31 “31 -20 20 29 23 23 18 ‘T4 ‘T4 “OI Le ee ce Pe ee | 0°954 0°924 O:oII 0:899 C322 0°225 tions. 0:216 O°154 0°136 0:080 0°074 0°074 0°074 0:062 O:102 0:092 0°040 0:040 0:685 0-649 142 PRACTICAL. COLLOID *CH Riis Expt. 154. Electrolytic flocculation of copper sulphide hydrosol—Prepare a copper sulphide sol (Expt. 146) by dropping dilute copper ammonium hydroxide into dilute H.S water and pour 20 c.c. of the mixture into an Erlenmeyer flask. A drop of this sol is placed upon a filter paper by means of a glass rod. If a thin uniform light brown spot forms, the sol is negative (Expt. 105). If the sol does not spread uniformly but forms a “mirror” or a small spot with sharp edges surrounded by a larger circle, it is too coarsely disperse and the experiment does not apply. Add from a burette containing normal KCl or MgCl,, successive, small definite amounts of electrolyte, shake, and after each addition, test on filter paper. A concentration is soon attained at which the sol no more spreads uniformly over the paper, but forms a “ mirror’’ with sharp edges and is surrounded by a ring of the colourless dispersion medium. This concentration of electrolyte added may be accepted as the precipitation value. If a series of mixtures are prepared simultaneously as . in Expt. 153, the precipitation value may be determined by the formation of a sharply defined zone on suspended strips of paper. Expt. 155. Electrolytic flocculation of a gold sol —Metals in the colloidal state often show a sudden colour change as a first indication of flocculation. Red colloidal gold turns blue-violet to blue ; yellow or brown colloidal silver changes to red, violet and blue respectively (Expt. 94). A coarsely disperse flocculation appears in a short time as a sequence to the colour changes. Determine as in Expt. 153 the concentration of HCl, MgCl,, and AlCl, necessary to transform a red gold sol into a blue-violet within ten minutes. Expt. 156. Electrolytic flocculation of Congo rubin—Flocculations accompanied by sudden colour COAGULATION AND PEPTIZATION 143 changes are followed more conveniently with Congo rubin than with gold sols, when studying the gold number. According to Expt. 97, Congo rubin in a o-or per cent. solution suddenly changes to blue-violet with almost all electrolytes of certain concentrations. Flocculate the Congo-rubin dye solution with baryta or saturated sodium hydroxide solution. The colour change and flocculation produced by these alkalis is not a chemical change due to the liberation of a different coloured acid dye, since acids also yield similar results. High hydroxyl ion concentrations do not produce this chemical change. Nevertheless, the colour change and precipitation of the dye take place. For quantitative determination of the flocculation values as with the As,5; sol, Expt. 153, pipette equal volumes, e.g. I c.c. of the dye solution into a series of clean test-tubes. Prepare salt mixtures similar to those in Expt. 153 and make them up to a volume of 9 c.c. Mix the dye with the salt solutions by repeatedly pouring the contents of the tubes back and forth. The salt concentration which produces flocculation shows a distinct colour change toward red-violet or violet-blue after an hour. The indication of flocculation by colour transition becomes well defined with a little practice. A solution for colour comparison may be prepared either by mixing methyl violet or azoblue with acid fuchsine or by using a Congo-rubin sol at the colour transitions of floccula- tion. These absolute precipitation values vary with the preparation of the dye. Nevertheless, they all show a relative difference in flocculating values similar to inorganic sols illustrated in the table below :1— 1¥For further data, see Kolloidchem. Beihefte 12, 94 (1920). 144 PRACTICAL COLLOID: CHES sa Fl , Molar precipitation occulation value; jitre of sol floc- Electrolyte. in millimols culated by a litre per litre. of electrolyte.! 1 OE on ee Meee SE x. pn 95°9 10°4 MgCl, ihe 3 hee Gites, ane 1:67 597°7 PIG Sol i eet spe 0°245 4082: Na SO poo ve] = os eee aes 612 16°4 MesO ole. Vt eee 0°394 2538: 2 Al,(SO,) 9) +002 7 0°03 33333) The effect of the cations on flocculation values is as evident here as with the As.S,; sols; while polyvalent cations have greater flocculating power, the anion must not be neglected. Sulphates flocculate more strongly than chlorides. Expt. 157. Flocculation of ferric hydroxide sol— Use a positively charged ferric hydroxide sol, prepared in Expt. 22, or a commercially prepared sol freed from excess of chloride ion by warm dialysis. Applying the same methods as in the previous experiments, especially Expt. 153, determine the flocculation values of NaOH, KCl, CaCl,, AICl,, K.SO,, K,-citrate. The following data are representative of an experiment performed by the author : Fe(OH), sol, dialysed. Fe,0,; content = 0-506 per cent. (a) NaOH. 1:0 c.c.0-oIN NaOH-+9:0c.c. H,O+10¢.c. Fe(OH), sol=o-5 milli- mols ALOK eR 3 a +8:0C.Cc, He a ==1-0 AOC, ie +6:0C.C¢. Re e == 2-O0ls) 8-0 C.C. ie +2:0C.C. PP * A rUTs 1 The term “ Molar precipitation in litre of sol flocculated by a litre of electrolyte ’’ refers to the number of litres of colloid which can be flocculated by a mol of the electrolyte. COAGULATION AND PEPTIZATION 145 oye Ia GIE 0-5 cc, 2N KCl+-9-5c.c. H,O--10c.c, Fe(OH), sol= 50 millimols TOC. 6 » +9:0C.C. ne ee —100 * ar OCR, » —+8:0C.c. * .; — 200 - 4:0 C.C. » +6:0C.¢. 7 KP == 400 re (c) 4 CaCl). O*5 c.c. 2N CaCl,--9-5 c.c. H,O-+10c.c. Fe(OH), sol= 50 milli- mols TO C.C, » +9:0C.C. a bs ==100 - 2°OC.C. » +8:0C.c. iy ry —200 ig 4°0 C.C, » +6:0C.c. Pe a = 400 ov (d) 4 AICI. Pee tele oo cc. 11,.0-toc.c. Fe(OH),sol= 300 milli- mols 4°0 C.C., we t-O°0 C.c. 2 #3 = 600 4, 8-0 C.Cc. ns +2:0C.c. a Fs —1200 2 (2) K,S5Q,. T:0c.c. 0-oIM K,50,-+9-0c.c. H,O+10c.c, Fe(OH), sol=o0°5 milli- mols =O C.C, uP +8:0C.C. a Zs =e 1; ) BO Cc a +6:0C.c. re Bs == 27s 80 C.C. ie +2-0C.c. ‘ * =S4°O 3, (f) Kg, citrate. T:0C.c. 0:005M K, citrate-++9-0c.c. H,O+10¢c.c, Fe(OH); sol= 0-25 millimols +8:-oC.c. es = 0-5 millimols BO GC: ts +6:0C.C, Fe - = 1-0 millimols Z-OE.C, Ly} Compare these flocculation values with those for As.S; (Expt. 153). KCl in both cases flocculated at concentrations of the same order of magnitude. On the other hand, the flocculation value of bivalent CaCl, and especially of trivalent AlCl; are considerably lower with As.S; sol; the flocculation value of CaCl, is almost the same as that of KCl for Fe(OH), sol, and that of AICI, is greater. The flocculation values of K,5O, and potassium 10 146 PRACTICAL COLLOID CHEMISTRY citrate are of the same order of magnitude as that of KCl for As,S;, but a considerably smaller concentration is necessary for the Fe(OH), sol. The HCl has a strong flocculating activity for the As.S;, while NaOH is par- ticularly active for the Fe(OH); sol. Thus, for the Fe(OH), sol, the anion determines the flocculation values, while the cation determines the flocculation values for the As,S;. The behaviour of Congo rubin resembles more that of the As.S, sol, yet the simultaneous action of both ions of the flocculation electrolyte may be dis- tinctly recognized. The aluminium hydroxide sol behaves like the ferric hydroxide sol. Expt. 158. ‘‘ Abnormal series ’’ with mastic sol —This term refers to the phenomenon whereby the same electrolyte may have a flocculating or non-flocculating effect upon a given sol, depending upon its concentration. It may be logical to suppose that a dilute flocculating electrolyte would have a still greater effect if it were added in a larger concentration. However, particularly with multivalent electrolytes, this assumption does not hold. There is a flocculation range of concentrations referred to as the first flocculation zone; then follows a range of concentrations in which no flocculation occurs referred to as the non-flocculation zone, and finally the recurrence of a second flocculation zone. Prepare a mastic sol by pouring Io c.c. of a 5 per cent. alcohol solution of mastic into go c.c. of water. Dilute this concentrated sol ten times (0:05 per cent.) and filter. Pipette 5 c.c. of the sol into each of a large number of well-cleaned test-tubes. Pour 10 c.c. of a molar solution of aluminium chloride into a ten c.c. graduated cylinder. Add half of this solution to the first test-tube containing the mastic sol and shake the mixture. Fill the graduated cylinder with water, halving the concen- COaauentiON AND PEPTIZATION 147 tration of the aluminium chloride and add 5 c.c. to the second tube of mastic sol, etc., as illustrated in the table below :— FLOCCULATION OF 0:05 PER CENT. Mastic HyYDROSOL BY AICI, OF VARIOUS CONCENTRATIONS.. | AICI, concentra- | AICl, concentra- | No.| tion in millimols | Flocculation after || tion in millimols _ Flocculation after per litre of 24 hours. ‘| per litre of 24 hours. mixture, mixture. | I 500 Completely flocculated|| 12 0°25 | Not flocculated 2 | 250 ” ” 13 OrI25 | By 9 S| 125 te 5 14 0:064 | Flocculated 4 64. ” 53 15 | 0°032 | Completely flocculated 5 2 ? ” 16 | 0-016 Bi) 9) 6 | 16 ve ¥ 17 | 0°008 | Somewhat turbid a 8 Slightly flocculated 18 | 0°004. _ Not flocculated 8 | 4 Somewhat turbid 19 | 0002 ie s com 2 Not flocculated 20 | 0-001 aes a IO | I 6 ah pets 0:0005 a, of rr | 0°5 ae x } 22 0:00025 |» a Beginning with the smallest concentration of AICI; solution, and gradually increasing it, the first flocculation appears between 0-008 and 0-064 millimolar concentration. Then a non-flocculation zone extends to a concentration of 4 millimols. Flocculation again appears with an in- crease in the concentration of the AIC]; unto the second concentration zone.? An insight into the theory of this striking phenomenon is obtained by an electrophoretic experiment, utilizing the ultramicroscopic method first with the sol alone, without addition of salt, and then with the sol + AICI, at a concentration within the non-flocculation zone. The sol possesses a negative charge, while the sol contain- ing AICl, has a positive charge in the non-flocculating zone.2 The electric charge on the mastic hydrosol may 1 Tt is understood that on repeating the experiment there will be changes in the absolute concentration values. 2 The positive sol in the non-fluctuating zone appears to give 148 PRACTICAL COLLOID CHEMIST: be neutralized by the addition of AICI, of definite inter- mediate concentrations. With smaller AlCl, concen- trations, a mastic sol behaves like a negative As,S; sol, while with higher AICl, concentrations, it behaves like a positive ferric hydroxide sol. Expt. 159. Influence of temperature on the flocculation of Congo rubin (compare with Expt. 97) —Add 5-10 c.c. of a normal KCl solution to 50 c.c. of a O-OI per cent. solution of Congo rubin and pour some of the mixture into three test-tubes. Place one tube in ice water, allow the second to remain at room temperature and place the third in a water-bath at about 50° C. Flocculation by the electrolyte occurs first at o°C., much later at room temperature, and does not take place at 50°C. REVERSIBILITY OF FLOCCULATION. OF SUSPEN- SOIDS Most electrolytes precipitate suspensoids irreversibly. It is impossible to wash out the electrolyte and again change the gel intoa sol. In some cases the irreversibility of flocculation is not so much a property of the colloids as that of the flocculating electrolyte. Thus, colloidal silver, according to S. Odén and E. Ohlon [Zeztschr. fur physik. Chem., 82, 78 (1913)], may be reversibly floccu- lated by ammonium nitrate. Colloidal sulphur, which is classed between typical suspensoids and emulsoids, is reversibly flocculated by most alkali salts. The reversi- bility of such flocculations may be demonstrated con- veniently by Congo rubin, described in Expt. 97. up its charge easily. On passage of current no cation migration is noticeable in the first 20-30 seconds, but after I-2 minutes the direction of migration becomes evident. SAE AlION AND PEPTIZATION 149 The phenomenon of peptization, i.e. washing the gel as described in Expts. 30-34, 1s an example of the reversi- bility of electrolyte-flocculated suspensoids. Expt. 160. Flocculation of suspensoids by dialysis —In the preparation of colloid solutions described in the iNapter preceding Expt. 30, the presence of a small amount of electrolyte is necessary for the stability of most suspensoids. If the maximum concentration of these “‘ sol-forming’’ ions is exceeded, the colloid floccu- lates. Hence, caution must be taken in many cases not to carry the dialysis too far. Suitable examples of flocculation by dialysis are: first, mercury sulphide hydrosols prepared from Hg(CN). (Expt. 14) ; second, copper sulphide hydrosols prepared from copper ammo- nium hydroxide (Expt. 146); third, cadmium sulphide hydrosols (Expt. 3); fourth, positive sols, particularly ferric hydroxide sols (Graham). MDialyse 40-50 c.c. of the Fe(OH); sol in an analytic dialyser (Expt. 54). Compare with the undialysed sol kept at the same tem- perature. The Fe(OH), sol flocculates in the dialyser after 24 hours. Concentrated Fe(OH); sols form jelly- like precipitates. Expt. 161. Flocculation by an electric current— An electrophoresis experiment continued for a long time causes the colloid to flocculate on the electrode to which it is attracted. This phenomenon may be clearly ob- served by ultramicroscopic electrophoresis. Study the electrophoresis of silver and mastic sols and allow the electrical contact to last only a few minutes. Coarsely disperse, strongly reflecting flocculates may be observed, accompanied by the appearance of dark patches, 1.e, colloid-free spaces in the field of vision. For flocculation by adsorption, see Expt. 146. 150 PRACTICAL’ COLLOID CHEMI B.. FLOCCULATION OF EMUESO RE. The flocculation of hydrated emulsoids is characterized by the very high concentrations of neutral salts required. The reason for this difference in salt concentration is that the flocculating electrolyte not only coalesces the particles into greater aggregates, but also causes a dehydration of the colloid particles, 1.e. a partial separation of the dis- persion medium adsorbed by these particles. Expt. 162. Qualitative demonstration of suspen- soid and emulsoid flocculation—Pour 50 c.c. of an As.S; sol into an Erlenmeyer flask. Add 50 c.c. of a clear egg-white solution to a second flask. The egg white may be prepared by diluting the fresh product 5 times with a 0-7 per cent. NaCl solution or by using a 2 per cent. solution of dried albumin in a 0-7 per cent NaCl solution. Add 5 drops of a saturated solution of ammonium sul- phate to the As,S, sol and turbidity results immediately. A similar addition to the egg white produces no turbidity. A large amount, such as 20—30 c.c. of ammonium sulphate solution, will produce turbidity and finally flocculation. Certain electrolytes separate the water of hydration from the colloid particles. This is a specific property of individual salts and for which general rules are not yet known. The aggregation and condensation of the dehydrated colloid particles in an aqueous dispersion medium involve electrical and electrochemical factors. The electrical charge plays just as important a role in emulsoids as in suspensoids. Electrically neutral albumin sols, such as serum albumins, are flocculated by neutral salts, alcohol, etc. (Wo. Pault). The electrolytic flocculation of albumin sols, previously studied, show complex relations. The following general- izations have been established. There are two large classes of albumin sols, the isostable and isolabile albumin sols. COAGULATION AND PEPTIZATION I51 Isostable albumin sols are stable at the “‘ isolectric points,”’ that is, in a state of complete electrical neutrality. Serum and egg albumin, hemoglobin, gelatin belong in this class. Isolabile albumin sols, when in a state of electrical neutrality, are no longer colloidally soluble and hence flocculate. Such sols are globulin, casein, stable in weakly acid or alkaline solutions. Albumin sols are usually amphoteric, that is, they may be either positively or negatively charged, depending upon certain conditions. They are more easily charged than suspensoids by the addition of a small amount of alkali or acid. In alkaline media the sol is negatively charged, while in acid media it is positively charged. The behaviour of these sols is in some respects similar to suspensoid sols. The differences between these sols is that the absolute precipitation _ values of these emulsoids are smaller; the influence of oppositely charged ions is more pronounced and floccula- tion is partly reversible. While the neutral emulsoids (genuine albumin) show the Hofmeister series, especially the cation series; rather indefinitely by flocculation experi- ments, the charged protein sols definitely show the series (R. Hoéber). The Hofmeister series is reversible, depend- ing upon an alkaline (negative) or acid (positive) sol medium. Expt. 163. Acid and alkaline flocculation of casein sol (an isolabile albumin sol)—Add 3-5 g. of powdered casein to 100 c.c. of a o-o1N NaOH solution, shake the mixture repeatedly, and allow to stand for 24 hours. The saturated casein sol freed from undissolved casein by filtering, shows a very weak alkaline reaction towards phenolphthalein. Pipette 2 c.c. of the casein sol into a series of test-tubes and determine in the usual 1 The former protein sol corresponds in behaviour to the sul- phide sols, while the serum albumin behaves like silicic acid sol in inorganic systems. 152 PRACTICAL COLLOlUMD CHE ane manner the concentration of HCl and NaOH at which the sol becomes turbid or flocculates. Start with o-o1N HCl and gradually decrease the concentration. ‘The first acid flocculation of the sol occurs with a mixture of 2 c.c. of casein sol and 3:0 c.c. o0-o1N HCl+ 5 c.c. HO. This acid flocculation value is equivalent to about 0-0025 mols. The isolectric point is attained at this concentration. According to L. Michaelis, the (H*)-ion concentration of the isolectric point is equal to 2-4 xX I0°. A second flocculation occurs with higher acid and alkali concentra- tions. The flocculating concentrations for HCl is about 0-25 mols, 1.e. about 100 times greater concentration of acid than that required for the first flocculation point, and for NaOH it is about 5 mols. These flocculation values may be determined more accurately by starting with N HCl in one case and with 8N NaOH in the other. Expt. 164. Neutral salt flocculation of hamo- globin (an isostable albumin sol)!—Use the powdered preparation ; if the hemoglobin is in the form of lamelle, grind it in a mortar before use. Dissolve 2 g. in 100 c.c. of water by first grinding the powder with a little water in the mortar in order to lessen lump formation. A better method is to sift the powder by brushing it through a wire screen into a beaker containing water constantly stirred. Filter the solution and proceed as in the pre- viously described flocculation experiments by pipetting 1 Jn the experience of the author, coagulation experiments on hemoglobin are particularly suited for this important chapter on the colloid chemistry of proteins. The material is easily obtain- able in uniform composition ; it gives relatively clear solutions of greater concentration than serum or egg albumin; it may be dissolved in any desired concentration and acidified or made faintly alkaline without inducing coagulation. The flocculation value is relatively low, so that working with extremely concen- trated acid solutions or with salts is obviated. The flocculation value may be determined with great accuracy. COAGULATION AND PEPTIZATION 153 2 c.c. of the hemoglobin solution into a series of test- tubes. Prepare the salt mixtures in another series of tubes and make them up to a volume of 8c.c. Thorough mixing may be obtained by pouring the added solutions back and forth several times. The point at which a distinct turbidity is observed immediately after mixing may be taken as the flocculation concentration. Tur- bidity is recognized by comparing with a control tube. Determine the flocculation values of a number of elec- trolytes on the neutral, weakly alkaline and weakly acid hemoglobin solutions. The following example gives the approximate flocculation values usually obtained :1 I. ELECTROLYTIC FLOCCULATION OF HA#MOGLOBIN K,-citrate, 2N = 0-66 mols. 2c.c. hem. + [2 c.c. K,-citrate + 6c.c. H,O] = 0-134 molar 2c.c. hem. + [4 c.c. K,-citrate + 4c.c. H,O] = 0:267 molar Ke 20; IN = 0°5 molar. 2 c.c. hem. + [4c¢.c. K,SO,+ 4 ¢c.c. H,O] = 0-2 molar 2c.c. hem. + [8 c.c. K,SO, + 0c.c. HO] = 0-4 molar Keacetate, 2N = 2 molar. Zeoicelizem, ——|2c.c. K-acetate + 6 c.c. H,O] == 0-4 molar 2c.c. hem. + [4 c.c. K-acetate + 4c.c. H,O] — 0:8 molar 1 Jn the author’s knowledge, previous investigations on the neutral salt flocculation of hemoglobin are as yet not available. In the above experiment the heavy-typed figures are only approxi- mate flocculation values, 154 PRACTICAL |\COLLOID CH ii ana KG] ANI ==aamolatie= saturaveds 2c.c. hem. + [4 c.c. KCl1+4c.c. H,O] = 1-6 molar 2-c.c. hem. + [6 c.c. KCl + 2 c.c) HO} 2-4maias 2c.c. hem. + [8 c.c. KCl-+ 0 c.c. H2O) == 3-2 molar KNO,, 4N = 4 molar = saturated. 2c.c. hem. + 8c.c. KNO,; = 3-2 molar. Flocculation value > 3-2 molar KCNS, saturated — ca. 14 molar ; no flocculation. For three additional sulphates and chloridés, the follow- ing flocculation values are shown in a similar manner : (NH,).5O, = 0-09 molar CaCl, = 0-004 molar NasoQy O40 i. MgCl, = 0-004 us TH35O" == DO aS AlCl, 7-33 2s Arrange in series the flocculation values obtained for _ the potassium salts: citrate, sulphate, acetate, chloride, nitrate, sulphocyanide. The cations with the sulphates give the series: NH,, K, Na, Li. The chlorides orithe alkaline earths show extraordinarily small flocculation values. 3 II. FLOCCULATION OF ELECTRONEGATIVE HAMOGLOBIN Final concentration = 0:03N NaOH. 2 cc. hem.+ [2 cc. saturated’ (capes (NH,).50,-++ 6 cc. H,O+6 drops 1N NaOH] = 0°8 - molar, immediate flocculation. 2 .c.c. hem,-+ [8 c.c. saturated’ Ca- escuela NH,CNS +6 drops IN NaOH]=ca. 6 molar; no flocculation. III. FLOCCULATION OF ELECTROPOSITIVE HA#MOGLOBIN 2c.c. hem. + [2 c.c. molar (NH,).50,-+ 6 c.c. H,O -+- 6 drops NH,Cl] = 0-2 molar ; immediate flocculation. COAGULATION AND PEPTIZATION 155 eee ening (2 C.c. 0-2 molar NH,CNS-+ 6 c.c. H,O + 6 drops NHC1] = 0-04 molar ; immediate floccu- lation. } The experiments with alkaline and acid hemoglobin give the following flocculation values: Negative sols— sulphate, 0:8; sulphocyanide, >6-0. Positive sols— sulphate, 0-8; sulphocyanide, 0-04. The sulphate floc- culates negative sols more readily than sulphocyanide. However, the sulphocyanide reacts more strongly than the sulphate toward positive sols. These results show a reversal of the Hofmeister series, depending on the sign of the charged sols. Compare the behaviour of As.S; and Fe(OH), [Expts. 140 and 141]. Determine the floccula- tion values of the above series, using 0:03N alkaline and acid solutions. The following table is a summary of the flocculation values found by F. Hofmeister for egg white with potassium and sodium salts 1 :— FLOCCULATION VALUES. Hemoglobin Egg white (K-salts). (Na-salts). Mols per litre. | Mols per litre. Ri ceee eres SC. 0:27 | 0°56 PeRPIT AONE R EE ee ke —— 0:78 DOIALGnee eR we 0-4 | 0:80 Nhs rr 0:8 1°69 (CUD ie TS) ee 2°4 3°62 OLGA ec ae | 5°42 Gy RRS Eee os a — | 5°52 TOUCHE Sl Ca. 5 very large Bilpnocyanidg. 6... = 1A | very large 1 Cited after R. Héber, Physical Chemistry of Cells, 4th Edition, I914, p. 308. 156 PRACTICAL COLLOID, CHEMISi i= Determine the cation series, SO,> as anion, with alkaline and acid hemoglobin in the same way. The series obtained is as follows : Alkaline Li* (1 mol} ~>NH,* (0°6)) > Kass) Acid Lit (0-13 mols) — NH,? (0-04) — K® (0-025). A greater sensitivity is shown by the acid sols, yet the cation series remains the same with the alkali and acid concentrations used. This does not apply to other con- centrations of acid or alkali. Concentrated hemoglobin solutions are also flocculated by additions of acid or alkali mixed with neutral salts. Expt. 165. ‘* Irregular series ’’ with dialysed eg¢ white—Use an albumin sol freed from globulin and salts by dialysis. Prepare by the above procedure 2 c.c. of albumin and 8 c.c. of aqueous salt solution, using molal lead nitrate as follows : 2c.c. albumin + 8c.c. M Pb(NO;). = 0-8 molar. 2c.c. albumin + [4 c.c. M Pb(NO;).+ 4 c.c. H.O] —'O-4 TiGlateetc An experiment in which an old dialysed preparation was used, gave the following results after two hours : o-8 molar * strong flocculation) 6-4 milliniols ) a oa +Clear 0-2 7 1°60 ye weakly turbid O'l - turbid 0:8 y turbid 105 ater ) 04 7 weakly turbid 0:025 -clear 0-2 ; clear 0:0125 ,, ) orl Expt. 166. Influence of temperature on the elec- trolytic flocculation of gelatin solutions—Add enough saturated ammonium sulphate solution to 50 c.c. of 0-5-I per cent. aqueous gelatin at room temperature until the first appearance of a faint turbidity. Clear the solution by adding a few drops of water. Pour the COAGULATION AND PEPTIZATION 157 mixture into three test-tubes. Place the first in an oven, the second in an ice-chest and allow the third to remain at room temperature. After 24 hours, the solution at the higher temperature has remained clear. A faint tur- bidity appears in the one at room temperature and a strong turbidity or flocculation in the tube which was placed in the ice-chest. If solutions 2 and 3 are warmed, the flaky precipitate dissolves to produce a slightly turbid liquid, which precipitates again on cooling. Compare with Expt. 159 on Congo rubin. Expt. 167. Flocculation of hydrated globulin by electrolytic extraction—Egg white, next to albumin, contains considerable globulin. Globulin, like many sus- pensoids, is colloidally soluble in the presence of certain small amounts of electrolyte. Not only (H*™) and (OH) ions, but particularly neutral salts have a dispers- ing action upon globulin. Prepare the following mixtures of natural egg white and distilled water : Seeccrces white 5 c.c. H,O em. Grins ta 7 5 COs, Eee Ce 64 phe 75 CC. yy G20 GAC? 4; SE gt ele by Raa Ok Ce Observe that increasing dilution produces a constantly increasing turbidity, and if the solution is diluted ten times, practically all the globulin precipitates. Pour a few c.c. of fresh clear egg white into an analytic dialyser enclosed in a vessel. Guard against bacterial growth by adding chloroform. Dialyse for 1-2 days and change the water often. Large globulin aggregates appear within the dialyser. Compare analogous experi- ments with Fe(OH), (Expt. 160). Expt. 168. Reversible and irreversible electrolytic flocculation of eg white—Flocculate a mixture of 2 c.c. of egg white and 8 c.c. of salt solution with 158 PRACTICAL COLLOID CHEMISTRS ammonium sulphate (about one molar) and with calcium chloride. Allow the precipitate to settle, wash by décanta- tion with distilled water or pour a few drops of the turbid mixture into a beaker containing distilled water. The ammonium sulphate precipitate redissolves, while barium, calclum and _ strontium salts, but not magnesium salts, produce irreversible flocculation (Wo. Pauli). Analogous to the suspensoid flocculation, reversibility or _ irreversibility of the process depends less upon the nature of the colloid and more upon the flocculating medium. If the albumin receives a charge by addition of acid or alkali, then the flocculation by ordinary alkali salts be- comes irreversible. Repeat the above experiment with ammonium sulphate, using a faintly acid and faintly alkaline albumin successively and observe that on longer standing, the flocculation becomes increasingly irrever- sible. Expt. 169. Alcohol flocculation of hemoglobin— Determine the flocculation value of ethyl alcohol upon neutral hemoglobin. Use 2 c.c. of 2 per cent. hemo- globin and 8 c.c. of alcohol-water mixture. The floccula- tion values are usually between 20 per cent. and 40 per cent. by volume of alcohol. Perform the same experi- ment with weakly acid (0-03N) and weakly alkaline (0-03N) hemoglobin. Observe that the electrically charged hemoglobin, which is strongly hydrated, may be flocculated by a rather high alcohol concentration. COAGULATION OF. DIALYSED EGG WHITE BY HEAT The coagulation of albuminous substances by heat involves chemical changes of denaturization which accom- pany phenomena of flocculation. The chemical and colloidal processes may be differentiated from one another in such a way that under certain conditions the albumin COAGULATION AND PEPTIZATION 159 may be denaturized by heating without any flocculation resulting. However, the colloidal process of flocculation may be produced by cooling (Wo. Pauli and H. Handow- sky). Expt. 170. Coagulation of dialysed egg white plus KCNS by heat—Dialyse an egg-white solution from globulin and salts.‘ To the dialysed egg white add sufficient potassium sulphocyanide to make the solution approximately 2N and boil a few minutes. Allow to cool, pour half the mixture into an analytic dialyser and change the wash water frequently during the first hour. Usually, a strong turbidity appears in the dialyser after a few hours, while the undialysed mixture has remained clear. The same results may be obtained by using KI instead of KCNS. Denaturization rather than flocculation results in the presence of sulphocyanide. Flocculation may be produced by the removal of this salt. This is analogous to the flocculation of globulin by dialysis of natural egg white. | Expt.171. Influence of electrolytes on the coagula- tion temperature of dialysed egg white—The simplest method for the determination of coagulation temperature is the optical method based upon the appearance of turbidity (see page 132). The coagulation temperature depends upon the rate of heating, as in the experiments upon gelation, corresponding to Expt. 108. A _ larger “normal’”’ rate of temperature change, 1° C. per minute, is advisable for this experiment. The solutions are con- veniently heated in a small test-tube placed in a beaker of water. The results may be reproduced accurately to at least half a degree, after a few trials. Determine the coagulation temperature in the presence of neutral salts, such as the potassium salt used in Expt. 1 The above experiment cannot be performed with non-dialysed globulin containing egg white. 160 PRACTICAL CGLLOID CHEMIST 153, with a final concentration of 0-5, 0:25 and o-I25N. The following example gives approximate values !: no addition 60° nitrate 65° citrate Thee bromide 62553 acetate 70 iodide 60° chloride oy sulphocyanide 60° Observe that all salts which exert any influence raise the coagulation temperature (Wo. Pauli).2 Furthermore, the Hofmeister series appears again. R. H6ber finds a reversal of the Hofmeister series on using similar concen- trations with dialysed egg white 3: acetate 62:0 bromide 66-5 _ chloride 62-0 iodide 76:5 nitrate 66-4 sulphocyanide fe Other salt concentrations produce an irregularity in the series so that with egg white and o:15N mixtures a reversal of the ionic series occurs (R. Héber). Using concentrations of iodide, cyanide, sulphocyanide, etc., and raising the temperature produces no flocculation. Expt. 172. Theory of emulsoid precipitation— Emulsoids are liquid-liquid systems. The following experiment proves the applicability of this definition -(K. Spiro, Wo. Pauli). Mix a hot 5 per cent. gelatin solution with sufficient powdered sodium sulphate, such as 2-2:5 M solution, to form a milky flocculate. Place the test-tube containing the gelatin-salt mixture upright in an incubator or water-bath. Heat at a temperature of 35°-50° C. for 24 hours. The precipitate settles, but on account of its high water content at the above tem- perature it flows into a completely coherent yellow layer, 1 An old preparation preserved in toluol was used. 2 All salts decrease the turbidity of gelatin solutions according tO REx pina: ’R. Héber, Hofmeister’s Betiydgem i ies eo COAGULATION AND PEPTIZATION _ 161 which is often clear and at least transparent at the edges. The precipitate behaves like a fluid when the vessel is tipped. The disperse phase is still a liquid under the experimental conditions described. A liquid gel may be obtained more rapidly if a 1-2 per cent. gelatin solution is flocculated with 0:5N salicylic acid and allowed to stand for 30 minutes at 35°—40° C. The precipitate, depending upon the amount flocculated, settles to the bottom of the tube in the form of drops or as a coherent phase. There is no doubt that a part of the neutral salt action consists in a dehydration of the colloid particles. There- fore the disperse phase of the colloid in the above ex- periment is richer in water when uncoagulated and must have the properties of liquid drops. C. OPPOSING INFLUENCE OF COLLOIDAL SOLUTIONS Ly -BLOCCULATION OF TWO COLLOIDS Hardy’s rule states that oppositely charged colloid particles flocculate each other. This is also true of coarsely disperse, colloidal, or molecularly disperse par- ticles of opposite charges. Flocculation of two oppositely charged colloids may be simultaneously produced by mixing them. This type of flocculation is characterized by the fact that it may occur only when the ratio of the concentrations of the two colloidal solutions lies within certain narrow limits. Often many trials must be made in performing such experiments before optimum floccu- lation concentrations are found. Such concentrations produce a completely clear supernatant liquid, due to complete flocculation. Prepare a dozen mixtures of night blue and Congo red at optional concentrations and record II 162 PRACTICAL COLLOID, CHEAIS iia the proportions. A systematic procedure must be used for the determination of flocculation optima. Reciprocal flocculation of colloids may often be regarded as reciprocal adsorptions. The precipitates formed are a particularly important class of adsorption compounds. These precipitates differ from ordinary chemical precipi- tates in that their components are not necessarily com- bined in stochiometrical proportions (compare Expts. 178 and 179). Expt. 173. Reciprocal flocculation of arsenic trisulphide and ferric hydroxide sols—Prepare an As. Sol from 0-5 per cent. As.O,; according to Expt. 153 and dialysed ferric hydroxide by Graham’s method. The flocculation optimum may easily be found when using such sol mixtures. The following table gives a personally conducted experiment. The results were noted after 24 hours. As.S, sol; content -0-5-per cent, referredsipe se. Fe(OH), sol; content 0-5 per cent. referred to Fe,O3;. i cc. As.$5;-+ 9 c.c. Fe(OH \se icleameiram Tidy Pr Sa. ep eas - faintly turbid 550 ” ” + 5°0 ” o turbid Obs ay 19 0 poe eae a precipitate, turbid, brown super- natant liquid A) ee So tO ee, . completely floccu- lated, clear col- ourless super- natant liquid Seay ton » + Io drops a precipitate, turbid, yellow super- natant liquid Toa; Pincers 5 : 4 precipitate, faintly turbid, yellow supernatant liquid COAGULATION AND PEPTIZATION 163 Pees, 2 c.c. Fe(OH), — fine precipitate "3 eee he est 1 “ni , clear yellow. Expt. 174. Reciprocal flocculation of Congo red ! and night blue (Buxton and Teague)—Prepare the following four solutions of Congo red and the following eight solutions of night blue, starting from I per cent. solutions : Congo red: 0:0125, 0:0100, 0:0083, 0:0063 per cent. Night blue: 0-0333, 0°:0250, 0:0200, 0:0167 per cent. 0:0125, O-0100, 0:0083, 0:0063 per cent. 5 c.c. of Congo red are mixed with 5 c.c. of night blue according to the following scheme :— Congo red. Night blue. O°0125 0:0333 0°0250 0°0200 0O-0167 00125 100 0:0250 0°0200 0-0167 0:0125 0-0I00 83 0°0200 O:0167 O°0125 O:0I00 0:0083 63 0:0167 0°0125 o-0100 0:0083 0:0063 In an example personally conducted, complete floccu- lation resulted and a colourless supernatant liquid appeared. The optimum quantities necessary for floccu- lation usually vary with the salt content of the individual sols. Expt. 175. Reciprocal titration of two dyes (L. Pelet-Jolivet)—The previous experiment may be modi- fied and performed quicker by using Tupfel’s method. Place a drop of both dyes mixed in proportions insuff- cient for complete flocculation upon filter paper. The dye present in excess forms a “mirror.” If the Congo red is in excess, the mirror or its edge is red. If the night blue is in excess, the edge is blue. Two reciprocally 1 Not to be confused with Congo rubin. 164 PRACTICAL COLLOID CHEMIS2Ra flocculating dyes may be titrated by the Ttipfel method described by L. Pelet-Jolivet, The Theory of Dyeing, p. 49 (Dresden, 1910). The following experiment conducted by the author is illustrative : 5 c.c. O-OL per cent. Congo red titrated with 0-033 per cent. night blue. 3-0 c.c. night blue, red mirror. 5000, - same, but weaker. 6-0 C.c. : much weaker. 6-276.) e indifferent mirror. O57e:C :, faint blue mirror. Oe} he decided blue mirror. The proportional amounts in this example are 6-2 parts of night blue to 5 parts Congo red, equal to 1:24. When titrating 5 c.c. of 0-0063 per cent. Congo red with 0:33 per cent. night blue, the proportions amount to 61 parts of night blue to 38 parts Congo red or1-:22. Ina great number of experiments carried out in the laboratory, 5 c.c. of 0-01 per cent. Congo red, titrated with o-or per cent. night blue in a porcelain dish with a glass rod, gave a proportion 1:2 to 1-3. When titrating with more dilute mixtures, a value of about two is obtained. The amounts by weight necessary for complete reciprocal adsorption are independent of the concentration of the reaction mixture. A convenient pair of dyes is methylene blue and crystal ponceau. II. PROTECTIVE ACTION M. Faraday discovered that small amounts of solvated emulsoids bestow a considerably greater stability upon suspensoids toward the flocculating action of electrolytes. The action of such protective colloids was already men- DOASUEATION AND PEPTIZATION 1365 tioned in the preparation of colloidal solutions (Expts. 45 and 46). Their mode of action is not due to an in- crease in the viscosity of the dispersion medium, for in many cases very small amounts prove effective. There appears to be a union between the suspensoid and emul- soid particles, with the result that the relative stability of the protective colloid is decisive for the whole complex. As yet it is not known if there is a ce coating’’ or “ en- veloping ”’ of the suspensoid particle by the liquid drops of the protective colloid. Such protective colloids are gelatin, isinglass, albumin, casein, hemoglobin, tragacanth, acid and _ alkaline hydrolysis products of egg white, lysalbin and protalbin acids, tannin, etc. Freshly prepared stannic acid is a protective inorganic colloid. Related material in regard to organic protective colloids may be obtained from the studies of A. Gutbier and his students in the Kolloid. Zeitschrift, IQlO-1g22. Expt. 176. Gold numbers (R. Zsigmondy)—Use an electrolyte-sensitive red gold sol, prepared in Expt. 3 with alcohol. [R. Zsigmondy, Colloid Chemistry, 2nd edition, 1918, p. 174; Zettschr. f. analyt. Chem., 40, Oooo ty iat ace) O01, 0-1, 1°, etc., c.c. of the protective colloid to be studied in a series of small beakers with just Io c.c. of a red electrolyte-sensitive gold sol. After 3 minutes, pour I c.c. of a ro per cent. solution of NaCl into each beaker, with constant shaking. By systematic decreasing of the limits of concentration determine which concentration of protective colloid is just sufficient to prevent the sudden colour change from red to blue. These numbers expressed according to R. Zsigmondy in mg. of protective colloid, may be more conveniently expressed in per cent. and are known as the “ gold numbers ’”’ of the protective colloids used. The orders of magnitude are: 166 PRACTICAL COLLOID "CHEMiSeta a. Gelatin . : . 0:00005—0-0001 per cent. Oxyhzemoglobin . _0°0003-0-0007 per cent. Sodium caseinate O-OOOI per cent. Albumin . ; 0-OOI—0:002 per cent. Staroh wee . ..€a.70°25 pencens The gold numbers give a quantitative estimate of the protective power of various emulsoids. They may only be taken as relative and not as absolute values because their numerical values vary not only according to the nature of the gold sol, such as degree of dispersion, con- centration, mode of preparation, etc., but also with the colloidal nature of the protective colloid. Determine the silver number in a similar manner by using a brown- red sol, prepared in Expt. 10, and choose for the end point its sudden change to grey-violet. Determine the As,.S, number by assuming at the end point the appearance of turbidity upon mixing. Expt. 177. ‘*Congo-rubin numbers ’’—Congo rubin is also suitable for the quantitative study of pro- tective.action and may be used as a gold sol substitute. Start with a 1 per cent. dye solution and pipette I c.c. into small test-tubes or beakers. Add various amounts of the protective colloid solution to the Congo rubin, make up to a volume of 5 c.c. with water and add to each mixture 5 c.c. of 0:-5N KCl. Determine the concen- tration of protective colloid, which produces a difference in colour shade after ten minutes. Compare with a control solution containing KCl of the same concentration. The following ‘‘ Congo rubin numbers””’ are illustrative : Sodium caseinate 0-004 per cent. Hemoglobin . ; 0-008 per cent. Albumin . 0-020 per cent. Gelatin . : 0-025 per cent. Soluble starch . ca, O-L per cent: bec Lames ; ; : ca. 0-2 per cent. COAGULATION AND. PEPTIZATION 167 Expt. 178. Cassius purple—Add a few c.c. of a O-OI percent. solution of stannous chloride to a 0:05 gold chloride solution. A brown to a beautiful purple-red colour first appears and the sol flocculates upon addition of any neutral salt. This so-called Cassius purple is an “ adsorption’? compound of colloidal gold and colloidal stannic acid. Such a composition was predicted by M. Faraday. As in the preparation of tannin gold (Expt. 2), the addition of stannous chloride acts in two ways: (1) It produces colloidal gold by reduction; (2) the colloidal stannic acid, formed by hydrolysis at such a dilution, acts as a protective colloid. The correctness of this assertion is shown by the fact, as pointed out by R. Zsigmondy, that if separately prepared solutions of colloidal gold and stannic acid are mixed, the resulting mixture behaves like Cassius purple. Add to a red gold sol a stannic acid sol prepared from stannous chloride, according to Expt. 39, and then a neutral salt to the solution. Compare with a similar experiment, using a pure gold sol. The gold-stannic acid mixture does not change suddenly to blue-violet. This mixture illustrates the colloidal reaction of stannic acid sols, since a coarse rather than a fine red precipitate is formed. Old preparations of stannic acid sometimes show a rather weak protective action. Analogous adsorption compounds may be prepared with colloidal silver or platinum and stannic acid. Expt. 179. Rubin purple—tThe protective action of stannic acid may be shown with Congo rubin as well as with gold sol. Add 2-3 c.c. of a stannic acid sol, prepared in Expt. 39, to 10 c.c. of a o-or per cent. Congo-rubin solution, freshly prepared with CO, free distilled water. Add the same amount of water to a control solution. 1 The amounts added vary with the concentration of the stannic acid sol and can be determined in advance. 168 PRACTICAL COLLOID (CHENS ii The control suddenly changes to a deep blue or violet in afew seconds. Addition of neutral salts causes the most rapid and complete change, especially upon addition of a few drops of o-‘o1N aluminium sulphate. The sol protected with stannic acid remains red. It is difficult to prepare mixtures containing a strong excess of salt which will remain distinctly red after standing a few hours, for a red precipitate gradually separates out. The precipitate, rubin purple, is an anologue to Cassius purple. D, PEPTIZATION Peptization is the reverse of coagulation. It involves a change of a coarsely disperse precipitate into the colloidal state. Examples of peptization were given in Fxpts. 30-39. Inthesimplest cases the precipitate spon- taneously decomposes to form the colloidal solution. In other cases the precipitate may be changed into a colloid by dilution or washing. This is known as reversible colloidal solubility and has already been mentioned in the paragraph preceding Expt. 30. Such examples represent peptization processes in a restricted sense and generally consist in the treatment of precipitates with electrolyte solutions. Examples were given in Expts. 30-39 ; other peptization processes are described in the following experiments. Expt. 180. Peptization phenomena—Flocculate a ferric hydroxide sol prepared by the Graham method with potassium citrate by first preparing a whole series of concentrations in order to determine the flocculation optimum.! Decant or wash the gel by centrifuging and mix the gel with a little ammonium hydroxide to change it into a colloidal solution. 1 The flocculation of ferric hydroxide sol with citrate gives an irregular series, GORGULATION AND PEPTIZATION 169 Flocculate a large amount of silver sol with ammonium nitrate and wash the precipitate as above. Suspend it in distilled water to which a trace of NH,OH has been added. ‘The precipitate regains its colloidal state, giving a clear brownish stable gel. Wash the purple of Cassius obtained by precipitating a red gold sol with potassium chloride and suspend it in water. Upon addition of small amounts of NH,OH, a colloidal solution is obtained. The same experiment may be performed with rubin purple. Colloidally disperse substances are acted on by the same reagents that combine chemically with the sub- stances in a coarser state. Hence, chemical changes in colloid solutions which lead to molecular dispersion are called dissolutions. It appears, however, that such dissolution is modified in some respects by the colloid state. Expt. 181. Dissolution of red gold sols by potas - sium cyanide (C. Paal)—Add a few drops of 2N potassium cyanide solution to a gold sol. The gold, upon gentle warming, instantaneously decolorizes or will do so within five minutes at room temperature. The experiment at the same time furnishes an example of the increased rate of reaction of colloid systems according to the so-called Wenzel law. If a larger piece of gold is left in contact with KCN, a small amount dissolves after some time ; hence the rate of dissolution is slower. Expt. 182. Behaviour of silver sols toward nitric acid—Mix a suitably concentrated brown-red silver sol with a few drops of nitric acid. The sol changes to a grey-violet or black and then flocculates. It gradually dissolves after continued shaking and standing. Expt. 183. Coagulation and dissolution of silver bromide sols by ammonium hydroxide (R. Auerbach) 170 PRACTICAL COLLOID CHEMIS tras —Prepare a fresh silver bromide sol in the following way : Add 12 c.c. of o-IN KBr to 80 c.c. of distilled water and add 8 c.c. of o-IN AgNO;solution. Pour 10 c.c. into four test-tubes, add the following quantities of solution and stir : Tube zr. zo c.c. distilled water. » 2 2:5 0¢.c.2N NH,OH + 25 c.c. distilled water. pyr ti LONER. SINGIN LE le » 4. o.c.c. 4N or stronger NOE. Tube I serves as a control. Immediately after the addition to tube 2 a stronger turbidity appears. The solution in tube 3 increases in turbidity and then becomes clear. Dissolution takes place instantaneously in tube 4. The question whether a sol first flocculates upon addition of a dissolving electrolyte or is directly dissolved is obviously answered by comparing the rates of both processes. The rate of flocculation of silver sol is greater than that of dissolution. Mix an As.S; sol with NaOH or a positive Fe(OH); sol with a little HCl. The dis- solution process proceeds so rapidly that flocculation by this addition of electrolyte is apparently impossible, at least it cannot be observed. The addition of larger amounts of HCl to a Fe(OH), sol first causes a flocculation. The silver bromide experiment illustrates these three possibilities. IX COMMERCIAL COLLOIDS 4 NHERE are numerous commercial ‘“ natural ’’ colloids. Hydrated emulsoids are, as a rule, obtainable as solid resoluble gels. They may be used for colloid chemistry experiments in this state, as well as in the disperse form, in a suitable dispersion medium. Suspensoids are likewise prepared in a solid resoluble form. The sols made by electrical methods are obtainable in solution. A. INORGANIC COMMERCIAL COLLOIDS Me LAL COLLOLDS! GOLD Colloidal gold—Dark red glistening lamelle. Colour of solution : reddish black, metallic ; purple red in trans- mitted light. Au content, about 75 per cent. Electro-colloidal gold solution—cColour of solution : reddish black in reflected light ; violet red in transmitted light. Au content, about 0-03 per cent. Colloidal gold solution—Colour of solution: dark 1 The number of colloid particles is proportional to the strength of the current. The base metals formed in aqueous solution undoubtedly have oxidation products in addition to the metallic element. it 172 PRACTICAL COLLOID CHEM ae red in reflected light ; purple red in transmitted light. Au content, about 0-005 per cent. PLATINUM, PALLADIUM Colloidal platinum—Black glistening lamelle. Colour of solution: deep black in reflected light ; deep brown in transmitted light. Pt content, about 60 per cent. Electro-colloidal platinum solution—Colour of solution: black in reflected light ; dark brown in trans- mitted light. Pt content, about 0-04 per cent. Electro-colloidal palladium—Colour of solution : greenish brown. Pd content, about 0-08 per cent. SILVER Collargol—Metallic glistening, brown green lamelle. Colour of solution: black brown in reflected light ; dark brown in transmitted light. Ag content, about 75 per Gert. ; Electro-collargol—Colour of solution: black brown in reflected hght ; dark brown in transmitted light. Ag content, 0:06 per cent. Electro - collargol, © concentrated — Ten times stronger than the previous solution. Colour of solution : deep black in reflected light ; dark brown in transmitted light. Ag content, 0-6 per cent. Skiargan—A ro per cent. sterile solution of a stable go per cent. colloidal silver. It is used for R6ntgen diagnosis, especially in pyelography. Ag content, 9 per cent. Choleval is a colloidal silver with gallic acid salts as a protective colloid. COMMERCIAL COLLOIDS ses MERCURY Colloidal mercury—Heavy grey black; external surface of particles show metallic lustre. Colour of solution: grey black in reflected light ; deep brown in transmitted light. Hg content, about 7 per cent. Electro-colloidal mercury—Solid, grey _ black, shining, heavy particles. Colour of solution: grey in reflected light; light grey brown in transmitted light. Hg content, about 55 per cent. Electro-colloidal mercury solution—Colour of solution: grey in reflected light ; brown in transmitted light. Hg content, about 0-09 per cent. COPPER Electro-colloidal copper—Colour of solution: black in reflected light ; dark reddish brown in transmitted leh eweiecOntctt, about 0°22 per cent. ARSENIC Colloidal arsenic—Blue black glistening lamelle. Colour of solution: reddish brown to grey in reflected light ; dark reddish brown in transmitted light. As content, about 33 per cent. ANTIMONY Colloidal antimony—Black — glistening lamelle. Colour of solution: grey black in reflected light ; dark reddish brown in transmitted light. Sb content, about 20, per cent, 174 PRACTICAL COLLOID CHEMTsa as VANADIUM Colloidal vanadium—Colour of solution: grey black in reflected light; greenish grey in transmitted light. V content, about 0-07 per cent. TIN Electro-colloidal tin—Colour of solution: grey black in reflected light ; grey brown in transmitted light. Sn content, about 0-3 per cent. TITANIUM Electro-colloidal titanium solution—Colour of solution: grey green in reflected light ; brown green in transmitted light. Ti content, about 0-6 per cent. LEAD Electro-colloidal lead solution—Colour of solution : grey black in reflected light ; dark brown in transmitted light. —Pb content, about 0-11 percene NICKEL Electro-colloidal nickel solution—Colour of solu- tion: black in reflected lght ; brownish green in trans- mitted light. Ni content, about 0-05 per cent. COBALT Electro-colloidal cobalt solution—Colour of solu- tion: deep black brown in reflected ight ; dark brown. in transmitted light. Co content, about 0:03 per cent. GOMMERCIAT ‘COLLOIDS 175 CADMIUM Electro-colloidal cadmium solution—Colour of solution: grey in reflected light; dark brown in trans- mitted light. Cd content, about 0-03 per cent. IRON 1 Colloidal iron—Dark red lamelle or red powder. Colour of solution: red. Fe content, about 12-13 per Cent. Electro-colloidal iron solution—Colour of solution : black in reflected light; dark reddish brown in trans- mitted light. Fe content, 0-5 per cent. CHROMIUM Electro-colloidal chromium —Colour of solution: reddish grey in reflected light ; dark reddish brown in transmitted light. MANGANESE Colloidal manganese—Black glistening lamelle or grey powder. Colour of solution: dark red brown in reflected light; light grey in transmitted light. Mn content, about 12 per cent. MOLYBDENUM Electro-colloidal molybdenum—Colour of solution : black brown in reflected light ; reddish brown in trans- mitted light. Mo content, about 0-04 per cent. 1 Compare Note 1, p. 171. 176 PRACTICAL COLLOID -GHEM iS. TUNGSTEN Electro-colloidal tungsten—Colour of solution: black in reflected ight ; dark red brown in transmitted light. Wo content, about 0-033 per cent. URANIUM Electro-colloidal uranium—Colour of solution: grey black in reflected ight ; dark brown in transmitted light.. U content, about o:I per-cent: SULPHUR Colloidal sulphur—Grey white powder. Colour of solution: milky white in reflected light; bluish in transmitted light. S content, about 75 per cent. Colloidal sulphur used for injections—Grey white lamellz. Colour of solution: milky in reflected light ; reddish blue in transmitted light. S content, about Osperecent. SELENIUM Colloidal selenium—Dark reddish brown heavy lamella. Colour of solution : brick red, turbid in reflected light ; blood red in transmitted light. Se content, about Sen pelecciite Electro-colloidal selenium—Colour of solution: brick red in reflected light ; blood red in transmitted light. Se content, about o-or per cent. CARBON Colloidal graphite with tannin as a protective colloid or with mineral oil as a dispersion medium. Electro- colloidal carbon is a brownish black liquid. COMMERCIAL COLLOIDS ies COLLOIDAL COMPOUNDS MERCURIC SULPHIDE Colloidal mercuric sulphide—Glistening lamelle. Colour of solution: black in reflected light; brown black in transmitted light. HgS content, about 65 per cent. ANTIMONY TRISULPHIDE Colloidal antimony trisulphide—Red brown to grey green iridescent glistening lamelle. Colour of solution: green and red in reflected light ; blood red in transmitted light. Sb.S; content, about 75-77 per cent. ARSENIC TRISULPHIDE Colloidal arsenic trisulphide—Yellowish brown lamellze. Colour of solution: bright yellow in reflected light ; dark yellow in transmitted light. As,.S,; content, about 66 per cent. SILVER SULPHIDE Colloidal silver sulphide—Black and yellow lamelle. Colour of solution: grey black in reflected light ; brown black in transmitted light. Ag.S content, about 35 per cent: ZINC SULPHIDE Colloidal zinc sulphide—Brown glistening lamelle. Colour of solution: yellowish grey in reflected light ; brownish yellow in transmitted light. ZnS _ content, about 20 per cent. I2 L7o PRACTICAL COLLOID CHEMISiis SILVER CHLORIDE Colloidal silver chloride—Grey white glistening lamellz. Colour of solution: milky white in reflected light ; light brown in transmitted hght. AgCl content, about 77 per cent. SILVER BROMIDE Colloidal silver bromide— Yellow glistening lamelle. Colour of solution: grey yellow in reflected light ; reddish brown in transmitted light. AgBr content, ADOULZO9 Percent. SILVER IODIDE Colloidal silver iodide—Yellow lamellz. Colour of solution: milky yellow in reflected light ; reddish yellow in transmitted light. Ag content, 31-7) penscemiee content, about 37-3 per cent. MERCUROUS CHLORIDE (CALOMEL) Colloidal mercurous’ chloride—Greyish yellow powder. Colour of solution: milky grey in reflected light ; brownish in transmitted light. Hg.Cl, content, about 75 per cent. MERCUROUS BROMIDE Colloidal mercurous bromide—Yellow brown lamellae. Colour of solution: milky grey in reflected light ; brownish yellow in transmitted light. Hg,.Br, content, about 80 per cent. COMMERCIAL COLLOIDS 179 MERCUROUS IODIDE Colloidal mercurous iodide—Yellow brown lamelle. Colour of solution: intense yellow in reflected light ; orange yellow in transmitted light. Hg.1, content, about 87-88 per cent. FERRIC IODIDE Colloidal ferric iodide—Colour of solution: black in reflected light; red brown in transmitted light. Fe.], content, about o-5 per cent. SILVER CHROMATE Colloidal silver chromate—RKeddish black glisten- ing lamelle. Colour of solution: brick red. Ag.Cr,O, Conte, 2 D0uUL 70 per cent. MERCUROUS CHROMATE Colloidal mercurous chromate—Black, faintly glistening lamelle. Colour of solution: grey green in reflected light ; brown in transmitted light. Hg,CrO, content, about 64 per cent. FERRIC ARSENITE Colloidal ferric arsenite—Ruby red, glistening lamelle. Colour of solution: red in reflected light ; orange red in transmitted light. Fe,O, content, about 30-6 per cent.; As,O;, about 35-39 per cent. 180 PRACTICAL COLLOID CHEMISE Tiss MERCURIC SALICYLATE Colloidal mercuric salicylate—Grey yellow glisten- ing lamelle. Colour of solution: grey. Mercuric salicy- late content, about 60 per cent. FERRIC HYDROXIDE The pharmaceutical dialysed ferric oxide in concen- trations of 5 and 10 per cent. usually contains consider- able amounts of chloride. Colour of solution: reddish black in reflected light ; red in transmitted light. Fe,O; content, about 0-55 per cent. ALUMINIUM HYDROXIDE Colloidal aluminium hydroxide—Colour of solu- tion: turbid to bluish. Al,O,; content, about I per cent. An interesting gel of aluminium is the so-called “ native ” alumina, according to H. Wislicenus. SILicic ACID Very pure, neutral to litmus, silicic acid is produced commercially, such as the 2-6 per cent water-white solution used as a toxin adsorbent. The gel “‘ osmosil”’ is a pre- paration which has a definite solubility in cold water. Another preparation is W. A. Patrick’s silica gel. Colour of solution : aclear liquid. SiO, content, about 2 per cent. ZINC OXIDE Colloidal zinc oxide—Colour of solution: grey yellow in reflected light ; brownish yellow in transmitted hght. ZnO content, 0-66) pemeemm COMMERCIAL COLLOIDS 181 MANGANESE PEROXIDE Colloidal manganese peroxide—Black glistening lamellz. Colour in solution: dark brown in reflected light ; black in transmitted light. MnO, content, about a0-per cent. Colloidal manganese peroxide solution—Colour of solution : black in reflected light ; dark reddish brown in transmitted light. MnO, content, 2 per cent. solution of a 50 per cent. colloidal MnO,. Peeters Ce COLLOIDS WITH SOLID . DISPERSION MEDIA Gold ruby-glass—tThis is almost colourless or faint yellow. The gold cannot be recognized ultramicroscopic- ally and therefore exists in a molecular disperse state. Other preparations are red to violet, blue by transmitted light and yellowish brown by reflected ight. The latter is strongly turbid, containing aggregated gold. Copper glass—Colloidal metallic copper is the colour- ing component according to R. Zsigmondy (Kolloidchemie, 2nd Edition, p. 109). Silver glass—Yellow, red, violet, greenish, etc., colours, corresponding to increasing size of particle of the colloidal silver. Selenium glass—Yellow, red, violet, etc., colours, corresponding to the size of the particles. Colloidal colour media are present in other coloured glasses such as calcium fluoride in milk glass, chromium and iron compounds in green and violet glasses. Colloidal sodium in rock salt—The coloration of blue rock salt is in all probability due to colloidal metallic sodium. Blue rock salt is prepared synthetically by heating colourless rock salt with metallic sodium. 182 PRACTICAL COLLOID CHEMISTRY B. ORGANIC COMMERCIAL COLLOIDS The great abundance of organic gels may be brought into colloidal solution by treatment with a suitable dis- persing medium. Albuminous bodies and related compounds— Glue, gelatin, isinglass, dried egg and serum albumin, hemoglobin, casein, plant albumins, such as crystalline edestine wet. Carbohydrates—Agar (d-galactose), starch, gum arabic, cherry gum, tragacanth, vegetable glues, such as carrageen (Irish moss), Iceland moss, quince seed glue, etc. Soluble starch and dextrin form transition solutions between colloid and molecular disperse systems. Soaps are colloidal in aqueous solutions and molecular disperse in dilute alcoholic solutions. Rubber is colloidal as a gel in benzene solution. Cellulose and its deriva- tives are colloidal as collodion ; as viscose, which is an alka- line cellulose plus CS, in water ; as filter paper in a solu- tion of copper ammonium hydroxide and concentrated ZnCl, ; and as celluloid, which is a solid solution of cam- phor and cellulose derivatives. Resins and resin soaps are colloidal in mineral oils, etc. Tannin in water forms a colloid transition system. Dyes—Typical colloid dyes in aqueous solution are: night blue, diamine blue, immedial blue, aniline blue, indigo, indulin, Congo red, benzopurpurin. Transition systems are Congo rubin and azoblue. Molecular disperse dyes in water are: methyl violet, — acid fuchsine, safranine, methylene blue, brilliant green, etc. (Sees Expt. 453) Chlorophyll is colloidal in aqueous solution. Colloidal indigo—Black particles. Colour of solu- tion: blue black in reflected light ; indigo blue in trans- mitted light. Indigo content, about 50 per cent. COMMERCIAL COLLOIDS 183 Colloidal cholesterol—Amber yellow _ lamelle. Colour of solution: milky in reflected light ; reddish in transmitted light. Cholesterol content, about 20 per cent. Colloidal. phenolphthalein— Brownish yellow, glistening lamelle. Colour of solution: milky white in reflected light ; reddish in transmitted light. Phenol- phthalein content, 50 per cent. Colloidal tar—Dark brown glistening lamellz. Colour of solution : sooty grey in reflected light ; reddish grey in transmitted light. Tar content, about 20 per cent. DISPERSOIDS OF VARYING DEGREES OF DISPER- | SION SULPHUR . Large sulphur crystals. . Roll sulphur, microcrystalline. . Sulphur flowers, microscopic sulphur globules. 4. Milk of sulphur is in a transitional state between coarsely disperse and colloidal sulphur. The aqueous suspension partially passes through an ordinary filter paper. 5. Aqueous colloidal sulphur, prepared according to Expt. 11, or the commercial preparations. 6. Dissolution of sulphur in paraffin oil, partially colloidal, according to J. Amann. 7. Molecularly disperse sulphur solution in C5,, OW N H SODIUM CHLORIDE 1. Large rock salt crystals. 2. Crystalline common salt. 3. Ground table salt. 184 PRACTICAL COLLOID CHEMI ii. 4. Sodium chloride gel. (Expt. 27.) 5. Sodium chloride-benzene sol. (Expt. 26.) 6. Molecular disperse aqueous sodium chloride solution. Another colloid series consists of a variously disperse gold ruby-glass in the three states described in a preceding paragraph: (1) Colourless to bright yellow when mole- cularly disperse ; (2) red to violet when colloidal; (3) blue and turbid with yellow brown colorations when coarsely disperse. Steel is a solid dispersoid, in which numerous structural constituents, such as the pure iron or ferrite, the iron carbide or troostite, the carbon or temper carbon, arein a state of colloidal dispersion. A coarsely disperse as well as a molecularly disperse state of the same constituents is found in other iron alloys. Colloidal carbon occurs besides the coarsely disperse graphite as the molecularly disperse hardening carbon. Specimens of iron of various grain sizes are likewise suitable for demonstration of a colloid series possessing different degrees of dispersion. DISPERSOID SERIES ACCORDING TO THEIR PHYSICAL STATE The following substances illustrate separate classes of disperse systems, the dispersion medium being given first. (a) Liquid-solid 1—Aqueous suspensoids of quartz, animal charcoal, kaolin, etc., are coarsely disperse. Colloidal gold, silver, etc., are suspensoids. Aqueous sodium chloride solution is a molecular disperse system.? (0) Liquid -liquid—Coarse emulsions of oil in water, such as commercial cod-liver oil emulsion, are coarsely disperse : 1 The dispersion medium is always given first. 2 It should be pointed out that the concept of degree of aggrega- tion no longer holds when the systems are molecularly disperse. COMMERCIAL COLLOIDS 185 Non-hydrated emulsoids. Colloidal emulsions of mineral oil in water, prepared in Expt. 1, or colloidal sulphur, prepared in Expt. IT. Hydrated emulsoids. Aqueous solutions of gelatin, starch pastes, benzol-rubber solutions, collodion solutions, eit. Solutions of alcohol in water are molecular disperse systems.! (c) Liquid-gas—Foams, prepared by shaking soaps or saponin solutions, albumins, etc., are coarsely disperse. Colloidal foams, as yet little investigated, are seen as critical phenomena during the liquefaction of gases when the opalescence in the fluid phase occurs. Carbon dioxide- water is a molecular disperse system.1 (ad) Solid-solid—Coagulated gold ruby-glass, metallic alloys, minerals such as granite, are coarsely disperse. Glass with colloidal colouring materials, steel, blue rock salt, smoky quartz and other coloured minerals, are colloids. Solid solutions, such as mixed crystals, alum, ammonium chloride and ferric chloride, etc., are mole- cular disperse systems. (e) Solid-liquid—Minerals with microscopic liquid inclusions, such as milky quartz, crystals with occlusions of mother liquor or water, are coarsely disperse colloids. Solid systems with colloidal liquid occlusions are as yet unknown. The water contained within zeolites may be removed without affecting their form, probably because it is in a highly disperse state, existing both as colloidal drops and as a continuous phase. Crystals contain water of crystallization in molecular state. (f) Solid-gas—Lava, meerschaum, pumice are coarse- ly disperse systems. Colloids of this nature have not yet been studied. Solutions of gases in solid substances, such as hydrogen in palladium, are molecular disperse systems. 4 See previous note. 186 PRACTICAL COLLOID CHEMiSi (g) Gas-solid—Smoke, such as soot, produced by burning benzene in a spirit lamp, or ammonium chloride fumes, produced by pouring together a few drops of con- centrated HCl and NH,OH into an empty litre flask. The degree of dispersion of such systems is variable. The combustion products of a faintly luminous Bunsen flame are colloidally disperse (H. Senftleben). (h) Gas-liquid—Liquid fogs, such as water vapour, clouds, etc., or fuming HCl, are examples of typical cloud formations. DISPERSE SYSTEMS Coarse dispersion. Colloidal. | Molecular dispersion. Increasing degree of dispersion | i Particle sizes | Y Oo'lu to Tuy Particles larger than | pass through filter | Particles smaller o-Iu, do not pass| paper, are held by | than IML, pass through filter paper, cans = be.) observed microscopically, not diffusible and non- dialysable. Dispersion————_~>_ <- Coagulation <——_—_ an ultrafilter, cannot be represented mi- croscopically, may be recognized micro- scopically sometimes, not diffused and dia- lysable or only very slowly. through both ordin- ary filter paper and ultrafilter, cannot be recognized ultra- . microscopically, dif- fuse and dialyse with remarkable rapidity. Condensation —> Dissolution Pas DISPERSOID ANALYSIS FREOUENT question is whether an unknown system has colloidal properties. The colloidal procedures outlined in this manual may be used to answer such questions. The following table gives a systematic scheme of analyses : A. GENERAL DETERMINATION OF DEGREES OF DISPERSION I. CHEMICAL ANALYSIS OF A HOMOGENEOUS SUBSTANCE 1. According to Expts. 77, 78, and 86, homo- geneous appearing liquids (unless hyloteopic | Ree ally convertible) possess definite boiling and > ; liquids. freezing temperatures, normal molecular surface meh: tensions, etc. 2. Physical mixtures of materials of similar analytical composition but of different Iso- physico-chemical properties, such as melting| dispersoids point, boiling point, density, solubility, etc. -f eventually mixtures of isomers, polymers, allotropic sub-| isocolloids. stances and strongly associated liquids, etc., 187 188 PRACTICAL COLLOID CHEMIST II. CHEMICAL ANALYSIS OF HETEROGENEOUS SUBSTANCES Experiments of hylotropic transformations, such as vaporization, distillation, freezing, give two or more constituents of different chemical composition. geneous, according to Expts. 77, 78 and 86.} Molecular Rapid diffusibility, according to Expt. 48.) disperse Rapid dialysis, according to Expts. 52, 53, or| solutions. 24: 2. Fluids appearing heterogeneous optically upon microscopic and ultramicroscopic examination. I. Substances appearing optically nas (a) Macroscopic and microscopic hetero- geneity ; separation of components by Meas ; dispersions ordinary filtration or by spontaneous sett- ; Meek (Suspensions ling, etc. Separation into two layers by eon moderate centrifuging ; spontaneous separ- isons ation (usually redispersable. ) ee (>) Macroscopic, often turbid, opalescent (Expt. 92) ; positive Tyndall cone (Expts. 77 Colloidal and 78), for differentiation of fluorescence). ”s solutions. Slow diffusion (Expt. 48); non-dialysable, according to Expts. 52, 53, or 54. B. SPECIAL COLLOIDRAN 1. Viscosity not essentially greater than that of the dispersion medium ; easily coagulated by electrolytes (Expts. 153-} Suspensoids. 158) ; spontaneously ultrafiltered spt 57). Pisce RSOID. ANALYSIS 2. Viscosity greater than that of the dis- persing medium; more difficult to coagu \ \ late by neutral salts ; decomposable by ultra- 189 Non- hydrated filtration (Expt. 57) ; resolvable eal emulsoids. scopically. 3. Viscosity essentially greater than the dispersing medium, especially at small con- centrations ; greater temperature coefficient of viscosity (Expt. 70). Difficult to coagu- late by neutral salts (Expt. 162). Dzisper- sion medium and disperse phase not com- pletely separable by spontaneous ultrafiltra- tion. Separate particles not recognizable ultramicroscopically, but only by Tyndall cone. Hydrated emulsoids. TABLE OF NORMAL SOLUTIONS HE concentrations in grams per litre refer to the hydrated salts of the composition given. A molar solution of BaCl, contains, for example, 208:3 g. of anhydrous salt per litre. Since the usual commercial preparation has two mols of water of crystal- lization, the following table gives 244-3 g. dissolved in one litre of water, etc. The bracketed numbers denote that the molar or normal solution cannot be prepared on account of small solubility, which is given in column 3. The data of saturated concentrations are given for 15° C., if not otherwise stated. The concentration data in column 3 also refer to the hydrated salts. MSine Saturated olar Normal solution Substance. solution solution erams per ey oie 100 grams per litre. per litre. of Sblmtians ARNO. = o).- Eta 169°9 169°9 64:9 AICI, - C2 ent sy ae 133°5 44°5 4I°I AL (DOE Olt. eget 666-7 TL TL 50°4 BaCl. 25. O sere 244°3 1222 31°0 CaCl iene. (cate Rone Tilo 53°5 -41°0 Ca@i 661,07 anne 219°1 I109°6 80°90 CasOfi! 1 sagas (136-1 68-T) 0°20 CasO 2H OR es (72:2 86:1) = 0°25 CdSO7* HO aan 256°5 1283 92:6 CoC) tg eee 134°5 6752 43°0 CusOr 5 Osan ae 249°7 124'9 25°3 FieCle ahve cere aae 198°8 99°4 63°1 HeChs ete Ue aes 162°2 54°1 46°4 HéeGl.- OH (seen 270°3 gorl 77°3 FeSO (77 rsO) ee ee 278:0 139°0 35°4 Urea (CO(NH 2). i252: 60°1 — 42:0 190 feoee OF NORMAL SOLUTIONS I9I Molaz | Normal | Saturated Srhstance: solution | solution grams per grams per grams per 00 grams litre. litre: Breather Citric acid UcIAG4 SORA! a] 210°1 70:0 64°8 HgCl, (271°5 135'8) 6°54 Hg(CN), va ar (252°6 126:3) 74 KAI(SO,),.12H,O (474°5 1188) 8°75 KBr ee IIg:0 II9g:0 38-9 moN. 651 65°1 a KCNS . 97°2 97°2 67°5 Tichol) a's 138:2 69°1 52°5 KCl 74°6 74:6 24°4 KCIO, . (122-6 122°6) 5°79 K,-citrate [K,C,H,O,.H,O] 324°3 108-1 64°7 (30°) K iFe(CN)e aH 0. (422:6) 105°7 20°6 BSL eins 166-0 166-0 58°4 KNO, TOI-I LOtt 20°7 K.SO,. (174°3) 87°1 9°25 Pinos ~ I109°9 55:0 25°7 MgCl, . . 95'2 47°6 55a MgCl,.6H,O 203°3 IOI‘7 75° MgsO,.7H,O . 246:5 123°2 51:0 NH,CNS . 76°1 76° 60°7 NET CI . 53°5 53°5 26:0 (NE ,50% 1321 66:1 42°6 ELS bee 58°5 58°5 26°4 Na-salicylate [NaC,H,;O3] 160:0 160-0 52-0 (at 20°) Na,SO, eee 142°1 71-0 BIT Na,5,03.5H,O 248-0 124°0 62-0 Na,5O,.10H,O 322°2 I6I°I 26°5 Oxalic acid feet OO 2ri.0| (126-06) 63°03 10+2 Pb-acetate.3H,O . 379°3 189°7 ca. 30 PbCl, : (2 Fond 139°I) 09 Pb(NOs;)» 331-2 1656 e353 J Ne) eg d's We Oe 28755 143°8 60-0 ved : e- g = Gee ; Bo 5 & Bu 3 ky fea) ye . . ha ry am eA ot « t 5 : oY - : m . ¥ 3 = A ie a eet A on = if As Coe \ ‘ ike \ os . i, aa Se gh