XI B RARY OF THE UNIVERSITY or ILLINOIS 666 ue ne. 16-23 CENTRAL CIRCULATION AND BOOKSTACKS The person borrowing tliis material is re- sponsible for its renewal or return before the Latest Date stamped below. You may be charged a minimum fee of $75.00 for each non-returned or lost item. Thcfl, mutilation, or dafoccinenl of library moterlali can bo (OUMt for ttudcnt ditclplinary action. All mat*rialt ownsd by lh« Unlvarsity of Illinois Library ar« th* property of the Stale of Illinois and ar« protactod by Articl* 16B of f/lfno/i Oi'niinol Law and Proccdur*. TO RENEW. CALL (217) 333-8400. University of lllinoii Library at Urbana-Champaign MAY 2 9 2001 When renewing by phone, write new due date below previous due date. L162 [TIG UNIX KRSri V OF ILLINOIS BL LLETIN ISSUED W E E K I ^ \<)l. XI. Jl'I.^' MK 1414. No. 47 [Kntered as second-class matter Dtceinher 11, 1912, at tlie post office at Urbana, Illinois, under the Act of Aug^ust 24, 1912.] BULLETIN No. 23 DKPART.MENT OK CKRA.MICS R. T. STULL, Acting Director NOTES ON THE DEVELOPMENT OF THE RUBY COLOR IN GLASS BY E."^ WILLIAMS PUBLISHED BY THE UNIVERSITY OF ILLINOIS, URBANA 1913-1914 Authorized Hcpriiit from X'oluinc X\'I, 19U, Transaetioiis American Ceramic Society NOTES ON THE DEVELOPMENT OF THE RUBY COLOR IN GLASS BY A. E. WILLIAMS The term "riil)}^ glass" is applied to red glass colored by the use of copper, gold, selenium and in some cases, flowere of sulphur, the color varying considerably in intensity and shade. In case of copper, the color varies from amber to various shades of reds to brown and to opaque black. With gold the red has a rose tint, and selenium ruby seems to be a brighter red of vary- ing intensities. The red from sulphur is rather unreliable, in that a uniform color is hard to obtain, and therefore only used for lower grades of glass. Copper and gold reds are said to be due to the metals in suspension as colloids. V. PoschP describes the preparation of Purple of Cassius from gold, and shows that the red or the purple gold-hydrosol may be obtained, depending upon the proper electrolyte present. Paal 's- proee.ss for the preparation of colloidal solutions shows that a red or blue hydro-sol of copper is obtained, depend- ing- upon the properties of the solutions. In G. Bredig V method of producing colloids elect rolytically, he obtained finely divided metallic gold, dark purple in color, when the arc takes place under distilled water. If a trace of caustic soda is added, deep red color is obtained. That copper and gold are in the same condition in glass as in solutions is proven by the use of the ultra-microscope. Zsigmondy^ says that ruby gla.ss will become red, or remain colorless upon slow cooling according to its quality. It will al- ways remain colorless on chilling, the normal red color generally being brought out upon reheating to the softening point; (high lead glasses show yellow or brown instead of red). The coloring is due to the gold, which is at first homogeneously dissolved in 1 V. Poschl, Chemistry of Colloids, p. 55. 2 Ibid, p. 6G. 3 Ibid, p. 67. 2 DEVELOPMENT OP RUBY COLOR IN GLASS the g-lass, later separatir.g out in the form of ultra-microscopic particles which reflect green light. He compares this phenomenon with devitrification, and re- fers to Tannnann's'"^ work on de^^tritieation. Tammann shows that the speed of crystallization, and the ability to crystallize in- crease with diminishing temperature from the melting point and then decrease again, while viscosity steadily increases. Zsig- mondy applies Tammann 's results to ruby glass in this manner: "Ruby glass is worked several hundred degrees lower than its melting temperature. At the working temperature, conceive it as a super-saturated crystalloid solution of metallic gold and the smallest amicroscopic particles to be centers of crystallization, it will at once be seen why ruby glass sometimes remains colorless upon simple cooling. In this case tlie optimum temperature for spontaneous crystallization is so low that the glass is very viscous and the speed of crystallization reduced to a minimum. If by reheating, the glass acquires a certain mobility, the gold separates out upon the nuclei present which by growth become sul)-microns, visible in the ultra- apparatus and turning tlic glass red or darker." V. Poschl'' says that gold ruby is obtained by an addition of gold chloride to the glass melt from which particles of gold separate out, when the mass is quickly cooled. These particles, however, have the magnitude of amicrons, so that the glass ap- pears colorless. By heating anew until the glass becomes soft, the particles grow until they attain the size of ultra-microns, to which the cause of the red color is traced. The preparation of copper ruby glass is performed by an analogous method. Copper ruby has, in the past, been made by a process known as flashing. Tliis process is described somewhat as follows by Rosenhain :" "Flashing glass is the process of placing a very thin layer of colored glass on the surface of a more or less colorless glass of usual thickness. This is generally accomplished by taking a small gather- ing of the colored glass on the pipe, and the remaining gathering for the piece to be made from the colorless glass pot. When this glass is blown, the ruby glass lies in a thin layer over the inner surface of the cylinder. The special skill required is in blowing this layer to a uniform thickness to obtain a uniform color." * Ztiprmondy, Colloids and the ultia-mirroacopc, p. 16.'). = Taii.nianii. Ziit. for Electro-chcmi, 1904, Vol. 10, p. 532. «Ibid I, p. 103. ' Waltrt Ros^-^^hain, Glass Manufacture. DEVEI-OPMEXT OF RUBY COLOR IX (JLASS 3 The necessity of flashing- is tliic to tJlie density of the coloi-. Cop})er coUns are so dense that many ylasses are opaque when over 3 ni.iii. tliiek, the eolor depending ni)on tlie e-(inipt)sitit)n and rate of eooling. However, it is p()sNil)h' to control tlie density of the ec>h>r somewliat in the Hashed i-uby glass by carefully eon- trolling- the temperature of worldng the glass and rate of cooling in the moldis. These factors nuust be controlled very carefully in praccice to produce uniform results. If these gkisses are cooled very quiekly, as for instance, chilling in water or rolling- very thin (2 ni.m. thick) on an iron plate, the red coloa* will not develop, 01' at least shows only in scattered streaks. By reheating at defi- nite temperatures, the color miay be obtained in varying degrees of intensity from aml)er to opaque black, dei)ending upon the temperature to which the glass is reheated. Thus it will be seen that the temperature and rate of cooling nnist l)c constant, to I^roduce a unifoi-m shade of red when this color is developed during blowing. At the present time, howevei", copper ruby glass is being- made in which the color does not come out in the pressing or working, but is brought out later by reJieatiug. The density of the color in this glass is very much less than the Hashed rul)\- glass, and pieces of greater thickness can be easily made. The eolor range from a light amber through reds to a dense opaque bliaek. with an increasing temperature. Available literature consulted (ui the subject gdve no eom- ])lete or definite methods foi- woi-l DEVELOPMENT OF RUBY COLOR IN GLASS GERMAN COPPER GLASS . 100.0 Sand 25.0 Potasli 17.0 Borax 2.5 CU2O 5.0 SnOo 0.2 Fe.Oa 2.5 MnO. 0.5 Bone ash Calculated Formula^ 0.200 PbO 0.390 K.O 0.120 Na,0 0.095 CuO 0.079 MnO 0.01-4 CaO 0.0060 FeA 0.2500 B063 0.0044 P.,0. 4.3G SiO, '0.09 Slid., 100 SiO_, 50 Pb;X), 25 KXO, 5 NaNO, FRENCH COPPER GLASS 'J'liis hatdi is fiuscd, c'lliilled, dried, ground and iiiixod with 1 CuJ), 1.5 SnO., 5 cream of tartair. This is incited and hhisted one hnuf duriiiii' melt. Calculated Formula 0.534 PbO 0.346 K.O 0.074 Na,0 0.046 CuO 3.900 SiO, 0.034 SiiO., Notes on ruby gbus.s from SpreohsaiaP" give the following by translation : "In the manufacture of ruby glass it is not in the field of the furnace man to control the color. Repeated fusion and cooling makes the best color, and the color does not depend as much upon the per- cent of coloring oxide in the mix as upon the temperature of the glass while working, the rate of fusion and rate of cooling the fin- ished piece." The following batch is given: "The empirical formulae of all fflasses Riven in the follnwint; work were calculated l)v the writer. ^" Sviechsaal, Feb. 6, 1913, p. 92. DEVELOPMENT OF RVBY COLOR IX CLASS D I.KillT HKI) HARK RED Sand 100.0 kg. 100.0 kg. Soda ash IG.O kg. 16.0 kg. Potash IC.o kg. IG.O kg. Borax 4.0 kg. 6.0 kg. Whiting 10.0 kg. 12.0 kg. Witherite 10.0 kg. 10.0 kg. CiuO 2.0 kg. 4.0 kg. SnO. 2.0 kg. 4.0 kg. Fe.O. 0.5 kg. 1.0 kg. Cream of tartar 0.8 kg. 1.15 kg. Calculated Molecular Formula 0.385 X;i,() j 0.210 K,0 0.222 CaO 0.110 BaO 0.0H4 CnO 0.0066 FeX)., )3.63 SiO, 0.0660 B.,().. 10.03 SnO., "Tlic manufacture of ruljv glass demands great care and practice in working. This is especially so with pressed glass. The raw l)atch should be put into a preheated pot and melted six hours. The melt is blasted several times and poured into cold water for remelting and refining. If the pressed pieces are not colored enough they can be reheated. The mold must not be too hot to allow the glass to cool too slowly, or too cold to chill and cause the pieces to crack. The following batch is also given:"" Sand 100.0 kg. Potash 25.0 kg. Red lead 25 . kg. Borax 10.0 kg. Soda 5.0 kg. Cu.O 3 . 5 kg. SnOc 2.0 kg. Fe.O., . 5 kg. MnO. 0.5 kg. Gullet 25.0 kg. Cream of tartar 0.5 kg. Ibid 1(1, l>. '.)■!. DEVELOPilEXT OF RUBY COLOR IX GLASS Calculated Molecular Formula 0.2020 K,0 0.6230 PbO 0.1010 Xa.O 0.0668 Cn'o ! 0.0079 MnO 0.00418 Fe,0, I 2.310 8i0, 0.07250 B.,6, 0.018 SnO, Rudolf Ih)lill)aum^- says that red colors may be obtained by tlie use of CuoO, selenium, sulphur and gold, but is most often obtained from Cu.O. He gives the following hatch for a copper ruby : 100.0 SiOo 34.0 KXO3 16.0 CaCO, K,CO3=80 to 85 percent pure 0.6 Cu.O 2.0 Snb, Calculated Formula 0.536 K.,0 1 . -r-f. c?.^ ll()hll)aum says : "Concerning the mixing of the Cu^O, I wish to remark that it is possible to obtain the ruby color with 0.4 percent Cu;0, also with 0.8 percent. However, with 0.8 percent of the batch as CueO the color is so dense that large masses are not workable. As such a small quan- tity of Cu:0 is needed to make ruby, it is mixed best by using 0.8 per- cent CuoO and SnO with half the batch of glass. When the glass is ready to blast then mix the batch containing 0.8 percent Cu:0 with an equal batch of crystal glass, and a 0.4 percent CU2O batch is obtained which gives a weaker color. It is best to employ SnO as a reducing agent to insure the obtaining of a ruby color, and one finds from practical experience that the mix must contain less than double the quantity of Cui:0 as SnO. If this is not sufficient reducing agent, cream of tartar may be used in quantities to satisfy all conditions. Iron scale may also be used as a reducing agent but the pure ruby color is then changed." '- R. liohlbaimi, Siitffeirassi Ilerstellunff lieaibdtung vi.d Verziervng des Felnern liulglasfs, p. 125. DKVKLOI'.MKXT OF RUBY COLOR IN GLASS 7 llolilhauiii'- gives the foll<»\viii-i' hiitrli for a gold ruby: Rose Color Sand 100.0 kg. Potash .'54.0 kg. Calcium carimnatc \7 .0 kg. Gold l''-0 yms. Gold must ht' l)l•ou^ht into tlie mix in ii very finely separ- ted form, best in solution or as Purple of Cassius. To i:et the gold in solution, it must be cut into small pieces and dis.solved with acjua regia. The gold solution is poured on part of the mix. and this mixed with the l)alaiiee of the batch. In the heat of the oven, the decomposition of the go:d chloride takes place so rapidly, that a portion of the gold chlor- ide is carried away nndeeomposed. There is, therefore, not so much gold dissolved in the glass as is introduced, and the color is much weaker than it would be, if all the ,^old were dissolved. It is, of course, reasonable for one to try and reduce the vapori- zation of the gold chloride as much as possible. This may be done by pouring the gold chloride on 1 kguL of sand and evap- orating to dryness. Then mix this well with half of the batch, or iLse gold purple in the same nuinner. According to Hohlbaum's experience, either phosphoric acid or ])ari\nn work favorably in the making of gold ruby, causing the gold to separate out more rapidly. Without either, the ruby is too light. A batch for making a rose glass wdth a violet tinge with the nse of barium is given. Rose Glass with Barium Sand 100.0 kgm. BaCOa 16.0 kgm. 95 percent soda, Na-CO. 4.'?.0kgm. Gold 12.0 gms. Selenium Ruby, Light and Rose Colored Arsenic 200.0 gms. Sand 100.0 kgm. Potash. 80-85 percent ,'54.0 kgm. CaCO. 17.0 kgm. Selenium nitrate 120.0 gms. " ll.id 12. |). 12G. 8 DEVELOPMENT OF RUBY COLOR IX GLASS In the reds with sulphur, one should not use the alkali sul- phates, but only sulphur with charcoal as a reducing agent. The charcoal keeps the sulphur from conil)ining with the soda and potash. In sulphur ruby, a great part of the sulphur vaporizes in the working. The melting glass foams vigorously, and there- fore one should fill the pot only half full at first, and after the batch reaches quiet fusion, put in the second half. Sulphur ruby is hard to make in uniform coloi*s. and dark- ens in the nuiffle. It is not used for nuikiug higher grades of glass. Two batches for sulphur i-uby are given : No. 1 No. 2 Sand 100 . kgm. 100 . kgm. Soda 45.0 kgm. 45.0 kgm. CaCOs 20.0 kgm. 20.0 kgm. Flowers of sulphur T.okgm. 10.0 kgm. Antimony sulphate 5.0 kgm. Charcoal 2.0 kgm. EXPERIMENTAL DATA BY WRITER The foregoing typical batches for ruby glass are but a few of a large numl)er given in the literature pertaining to glass making. An examination of these shows a wide variation in com- position, but all agree in that they ai-e high in silica and contain tin. In copper ruby, the amounts of c()pi)ei' and tin vary widely in their ratios to each other. These copper rubies are probably u.'-ed in the manufacture of tlashe<^l glass. In the beginning of the following experimental work, sam- ples of conniiercial copper ruby, both the quick-cooled colorless and ruby colored were obtained. The uncolored sample was broken into fragments, and different fragments were heated to different temperatures for various lengths of time. A small Iloskins electric furnace was used, and temiperatures were read with a Leeds Xorthiiip potentiometer, using a platinum, plati- num-rhodium thermocouple. DEVELOPMENT OF RI'BY COLOR IX GLASS The followinoj results were obtained : PIECE MAXIMUM TEMPERA- TIME HELD AT MAX. REMARKS NO. TURE TEMP. °c minutes 1 500 30 No change in color 2 500 60 No change in color 3 550 30 No change in color 4 550 60 No change in color 5 575 1 No change in color 6 575 30 No change in color 7 600 1 \'ery light amber 8 600 15 \'ery lig"ht amber 9 600 30 Brig-ht amber, sHg-htly darker than No. 8 10 600 60 Bright amber, same as No. 9 11 650 1 Bright amber, same as No. 9 12 650 30 Deep ruby, edges slightly soft- ened 13 650 60 Same as No. 12, edges slightly softened 14 675 15 Same as No. 12, edges slightly softened 15 675 30 Same as No. 12, edges slightly softened 16 675 60 Darker than No. 15, edges slig-htly softened 17 700 1 Same as No. 10, edges slightly softened 18 700 30 Dark red, edges slightly softened 19 900 30 Grayish purple, opaque, softened out of shape The rate of increase of temperature was a constant factor in all of these tests, as follows : ten minutes from room tempera- ture to 300° C; 300° C to 500° C at rate of 50° per minute; 500° C to maximum temperature at a rate of 25° per minute. The results seem to show that the color at any definite tem- perature is practically constant, and that the color chang-e at that temperature is apparently instantaneous. However, time k 10 DEVELOPMENT OF RUBY COLOR IN GLASS required for the temperature to even up through-out the thick- ness of the piece. It will be noticed that the glass shows signs of softening at that temperature at which the strong color develops. This is probably the softening point Zsigmondy^^ refers to in the article pre\dously quoted. It is observed that there is little or no ap- parent change in color brought out between 650° and 675°, giv- ing a safe range for an annealing oven. Most of the glass fornnilas observed were high in lead and in silica. Accordingly the following formuia was selected, it being the upper silica limit for most glasses : 0.5 PbO I ., ^.^ 0.5 Xa.O } ' ^^^^ In order to determine a suitable iiietliod of working, several small batches of this glad's were fused. The inetliod adopted was as follows : The glass was fused in Battersea crucibles in a small pot furnace using gas and compressed air. The temperature range required for firing and to make the glass liquid enough for pour- ing, was between 1480 C and 1520 C. One-half hour was taken for complete fusion of the lead glasses and one ht)ur for the lead- less glasses. Not much trouble was experienced in reducing the copper oxide ami preventing oxidation. Although a slight reducing flame was used, the presence of cream of tartar ^al)r)ut '^ per- cent) seemed to make reduction certain, if the time of heating was not too long. When fusion was complete the gla.ss was poured on a heavy cast iron plate 1 in. thick, and then rolled to a thickness varying from 2 to 5 m.ni. The thinner portions usually cooled colorless, and the color developed in the thicker, .slower cooled portions, i. e. turning red or opaque brown or black. "Ibid 4. DEVEI.Ol'MKXT OF HrHV COLOR IX GLASS 11 c c s o cJ t^ o o c o o o o o in or CJ O O O '" o o o c^ o c c; Q O C O CJ C: — o o c o o ^ o >o ro CI o o o '~ O O O C^ o o o O O O O C^ o ~ O o o o c •- c: >n re 01 o o c '" o c o r- o o s O O O O »!< o o C O O O O iH o in re 01 o o o o o C o oe o o = o o o ?- o o o o o o in lO o o o c • o c: J- o o c • c c; o o mo . c; o c '" o o • re c c o C ^ J- C: o o o o o >n C 01 I- o O C: O O o o o o ce c: o o o o o o o •+ o o o o iH o o o o o o o • o -t< -f o 1 o c; ■ c o o o in o O O O O o o • CO o o o o o • o -f -t* ^ '"' 0.5 0.5 :i.o 0.0 0.0 0.5 1 o u 1- o J t-" "a. H 03 ■< -*-• ^ "c ■ S L. \00A06t 1 fs^ rt^ = n u 1 ^ ^ ^ h'_ ; Si L )'j :U 1 C C c c c 01 o 1 o • c c c J- '.e c re • ce .-e c c c c in • -f c o c c ^ fO n T^ C-. o . o c o I- Le o re • CO C C c; I- O i.O . Tl< C O C C T-i ■n • ■* o o o c iH re • ce 1-1 Ci o o • • c; o 01 c o o • ■ o o o o re o • • c o o c i.e t^ • • C: O O T-H re >n (T-. O O • • C 00 01 o o o • . c; th m o ce o • • coco in i^ • • c o o ^ re in c: o o • ■ o m oi o o o • O re m o rO' c • o c c c i.e i^ • C C C --I re m ~ o o • • c o Ol o o o • c I- in o ?; CO o • coco i m f- • COO" '■ re in ::; o o • ■ o -* m o re o • O l-H I- O m t- • c o c --^ re m • ■ c-. o o . ■ o Th o o re o • o th ce c m t^ • c O C 1-^ re in . • ci ce o • . c Tti o o mi- • • o ^ re iH re m • • cj u rt ;- -o : rt C O - uc •id. 'C ,r ' ^ rt^r-rt— 3ci- 4 , ^ < 1^ ^uu. ''- )'y (_ ) 12 DEVELOPMENT OF RUBY COLOR IN GLASS SERIES A Glass batches were then made corresponding to the for- mulas given in Series A. The following results were obtained: Number 1 — Colored out vei*y dense opaque grayish-brown color. Number 2— (Decreasing the coloring agent.) This poured well, and cooled practically coloiless at 5 m.m. thick. Softened out of shape at 675°C. ar.d colored out, streaked with reddish color. At 700°C. it beeime dnrk l)rown. (>pa(iue and still streaked, very soft. Xumber 8— (Irc:e:!sii g tin to harden.). This poui'ed well and was colorless except i'nr a pnle giH'enish-yellow coh)r at 5 iii.ni. thick. Heated to 480 "C, gives amber coloi'. Heated t'> 525° C, gives deep red color. Heated to 700° C, softened out of shape giving a dense, brown opaque glass. Color change very rapid. Xumh.'r 4— (Uecren.sing- the Cu.O to reduce intensity of color). Color developed darker !han No. :} in ])nnring, having a greenish cast. Heated to 600-C. its culo!' was deep opaque, and neai-ly black, aiiihei- at 550°C., and hrown at 575°C. Xuiiiher 5— Developed a ralhei- intense bi'own color while pouring. Thin colorless sections gave a deep greenish l)i'nwn at 550" C. and a dense opaque black at 600° C. Xumber 6— (Still reducing airount of coloring matter). This glass poured clear and colorless. On reheating it changed to opaque black from 550° C. to 600° C. Color change very rapid. Xuiid)er 7— (Coloring matter left out to test purity of ma- terials for iron). This glass on reheating at various temi)era- tures gave no change in color. The conclusions from this series of glasses, (excluding No. 1)1^' are: (1) Low amounts of copper seemed to increa.se the density '■■^ Tliis slass \v;is not nioltrd well enough to jiKlge results. DE\'EI.OPMKNT OF RUBY COLOR IX GLASS 13 or oi)aeity of the colof, ar.d decreas-e the signs of red, giviiio m in lO lO i.O "^ ^ "^ . ^o o o o o o c o o o o o o O o O ^ o ■Ci. >n o o o c o o o o o o o c o o o M .'J 0.1 i"^.' C^i CO CO CO CO ly^ CO r- CO CO i,-— t^ f- (^ f^ r^ y^ (~ f^ ?^ ?^ !^ t- t^ O o o o o o o o o c o c C o o o o o o o o o c o o o o C o c o" (M c> c.> ry> ■* -f -f "* O o c o o o c: c o o c 3 OS 05 cr. t> r^ f- a o o o a o o o • o o o • 00 00 00 OC' oc on cr or nr r? t~ ^ o T— * 1-H 1-H ^- .— 1-H ^« ^^ ^- C! to CO o o c c <= o c c c C o o c c c o c c o o c o CO o 00 ^^ c o X c o o © o c o o o to f^ o f^ (^ i^ <^ o o o f- r~ t- t~ to o o o o o w LO -* c? t^ )~ o to to o o c o a c o !== c c c o o c o o o o c o o c o o c o o c c c 1-1 CO CO CO CO rH CO CO CO CO CO CO CO CO o CO CO M CO CO CO CO CO CO CO CO CO cr. CO CO m (M (M w w M (M IM (M PJ CI CJ CI CJ C} CI o N N N Ol w M O) w w CI CI CI c> CJ CJ o o o o o o o o o o o o o o o ^^ w C! w w w CJ CI CJ a Ci CJ CJ o o o o o o o o o o o o c c o o fO t- cv: t^ r- f^ CO r^ r- r^ r- r- r- «> t- N O-l Oi 0-1 w OJ w w O) (^> c? Cl c> C! CJ o •-0 •o •o o o o o o «o to to to to — o o = o o c o o o c c c c o o tH w ^_l w (M w ^ OT o m 00 c> c> CJ CJ O o o o o o o r^ o> CO m o c o ^ ^. '■; ^. " ^. '^ O o o o o o o -^ o ■^ — o - ~ c o c w tH w CO ^ lO o t- GO Ol o 1-1 CJ CO ^ »c < o 1— t tH 1—1 1— ( 1—" 1— ( J » o DEVELOPMENT OF RUBY COLOR IX OLASS 15 The results showed this very inarkciUy. The fusions, luade at the same temperature range 1480 (" and l.l'JO (', were more fluid and poured easier. Numbers 1. 2, 3 and 4 developed deej) oijatpie glasses when poured 4 to -■) m.m. tliiek. The thinner portions, however, in- creased in degree of transparency to about 2 iilul at which thickness the glasses cooled colorless, but of course very l)rittle. Upon reheating, the colorless pieces of these four glasses colored to about the same color density \vhn »ra • • • i-o o 00---0C o' oooo ■ • -co coo in oo OOOO • • •OOOOO'O'tC-. N o «.o "f; "C in CO 1-1 >c i> o -t Tfi Tti (rcr-:cor^c^ror^(MiMr-icOTtirocoro MW .MW(MeJ (M N (M (N X 3 ooooooooooocooo oooooooooocooco rHNCO'+l>rttO!>COO>Oi-IWfC'*<0 DEVELOPMENT OF RIBV COI.OH IN GLASS 17 The hii.sis ol" this sci'ics obtaiiitnl from 1 Inhlliauiii' ' is eii- tii-ely dilVri'i'iit than sri'ii's \i. It is a liiiie-potfish. hijjch silica, leadk-.»s ghuss, with lii^h tin, tiuTt'l'ore, a coniparatively refrac- tory and viscous jrlass at l!)\v tc:i:peratiii-i's. One hour was taken for fusion. Iloldbaum's hatcli calls for SiiO as thi^ I'eilucing aiz;ent, cream of tartar hein«r atlded as a prccaiiiioii to insure sufficient reduction. Nuiiihci- C-l was first made hy suhstitution of SnO^, for SnO, antl leavinji out the ci'eam of tai'tar. An oxidized clear colorless p'lass was the result, uivin^' no color change when reheated beyond tlie softeiiin*;- point. Xunil)er C'-I was again made using SnO._. and 0.5 percent cream of tartar. This glass was exceedingly visciuis and quickly cooled belo'W the jioint of easy jiouring. Upon ixtui'ing and roll- ing, (although taking a little more time), no coloi- change took place, the glass remaining clear and coloi-less. Upon reheating, no color change took place until 8(X) ('. was reached, when a light amhci- color was ohtaiiu'd, 850°C. gave a pale reddish brown, 900° C. g'ave a light brown, lOOO'C. softened with an opacpic bi'own color. The red color was not good in this glass and it seemed to be entirely too refractory. Series C. Xo. 2. (Reducing SiO. to soften). This showed an improvement in the working qualities \nth no tendency to color out on pouring. Reheating this glass gave the following residts : 800° C. a distinct light red, 850° C. a good ruby color. 900°C. a deep dark red nearly opa(|ue when -i m.m. thick. Series C, No. 3 (reducing SiO._, still further) gave a fusion which poured colorless and flowed freely. Reheated to 850^ it showed a reddish brown, slightly streaked. 900° showed a dis- tinct deep brown. 1' Il>i(l n. p. 121. 18 DEVELOPMENT OF RUBY COLOR IN GLASS (rt < -] o D to ^ nr lU o It) Ll oooooooooooo c: ':^ C: coo C! C^J ^ OOOOOOOOOOOO oooooooooooo PI C! C! C-i IT! Cf CI C! CI Ci CI C! OOOOOOOOOOOO oooooooooooo OOOOOOOOOOOO Tti-<*<-*-*-rc}Cjrccocv:fcc^ OOOOOOOOOOOO IC O O »0 *0 i.'t ».0 »0 »0 *'3 *0 O oooooooooooo ■*'*c • • u^ . C! Ci >- • • ^ • d' c CO OtC5Cf5iom f^CCCCC-:cO<-lr-(Cl(MMNW o OOOOOOOOOOOO occocccccooocoocoooooooo 5 OOOOOOOOOOOO o>nroi-iooo'nio>noo Ot-OCCWOt-l^-l^t^-OO CJ>-<'HlHTHClrH^T-lrtClC* i-tcicc'*'-':^t-occ. O'-iCJ DEVELOPMENT OF RLBY COLOR L\ GLASS 19 Series (', Xo. 4 (less SiO. than L'-i). Poured clear and col- orless l)ut when reheated to 850' beeame more streaked and showed a more decided brown. Series C, Xo. 5, ])oured clear and colorless as tiie others, but showed brown streaks. When reheated to 801)" it showed a very streaked brown color. AVlien the glass was remelted and re- poured it gave a very clear glass. Upon reheating this to 750^C. the color came out a clouded black, increasing in intensity with the reheating temperature. The foregoing five glasses in group C show that : (1) Reducing the SiO. from 4.57 to -i.U molecules improved the color in this series. Further reduction, however, changed the color to browns and then blacks, giving about the same range of brown and blacks with o SiO. as series A gave, having 3 SiO, and small amounts of copper. (2) High silica seems necessary in order to develop a good red color. The color change takes place at rather high tempera- tures for a reheating furnace, and the glass appears to be too viscous for good working properties. Glasses C6, C7 and C8 were made by introducing PbO in place of part of the CaO with the idea of softening and, if possible, still retaining the property of not coloring out on pouring. C6 and C7 in which 0.2 PbO replaced 0.2 CaO showed a dis- tinct improvement in the working cpialities and uniformity of color, although these glasses colored out in the thicker portions during the pouring; C6 to a light red and C7 to a deep ruby. Thase glasses, however, were transparent to a thickness of 8 m.m. in comparison with series B, which were not transparent in pieces over 2V2 m.m. in thickness. Reheating clear portions of C6 gave a good, deep, ruby color at 650°C., a considerable lowering of the temperature over the leadless glasses for developing color. This g'ass also has a fairly constant color over a temperature range of 25^ C (625° C. to 650°C). Series C, No. 7 colored out at 570° to the same shade as C6. Serias C, No. 8 (Reducing PbO to 0.1 with 4.00 SiOJ. This glass gave evidences of being harder than the ])rcvious glass 20 DEVELOPMENT OF RUBY COLOR IN GLASS (C7) as the fusion colored out a veiy little clearer at 6 m.m. thick (similar to C6), and the color. ess portions gave a deep clear ruby on reheating to 570'', the same as C7 and about 60° lower than C6. This glass gave the clearest and best red ob- tained in the foregoing work. Series C, No. 9 (in which 0.1 FbO was replaced in C8 by 0.1 XaoO as borax) gave a glass considerably more fusible, and flowed well in pouring. A very streaked, nearly black, color de- veloped in portions over 3 m.m. in thickness on pouring. Thin transparent pieces heated to 750^" gave a red color, streaked with opacjue black lines. This fusion, therefore, did not give good results. The po.ssibility of spoiling the color by over-heating is ever present. It is possible that less BoO.j would give better re- sults, though this was not tried. Series C, No. 10 (0.1 Xa,,0 replacing 0.1 L'aO). The result- ing glass was clear and colorless, sliowing a very few light red streaks. The working properties of the glass were very good, especially in i)ouring and cooling. On heating to 700 C the gla.ss tui-ned a clear light red. 625 ^C. showed a clear light red. 725^0. showed a clear light red. The color range of this glass is therefore good. Series C, No. 11 (0.1 R)0 replacing 0.1 CaO and with 4.57 SiOJ. Results from this gla.ss were a failure as the fusion was incomplete and very viscous and colored out a dense opaque black on ])()uring. If i)rop(M'ly fiisiMl, liettcr results would no .l()ul)t have been obtained. Series C. No. 12 (0.1 Xa,0 replacing 0.1 CaO and with 4.57 SiOo). This glass gave a very good fusion, but was rather vis- cous and showed no color on pouring. Heating this glass to 700' C gave an amber colored glass streaked with dark red lines. At 800° C it showed a good even ruby color. The conclusions from this last series of glasses (C6 to C12) are (1), that soda replacing lime softened the glass without causing the color to come out in cooling. (2) Lead on the other hand caused these glasses to color out rapidly on cooling, but did not make them opaque. DEVELOPMENT OF RUBY COLOR IN GLASS 21 General Conclusions. The following are general conclu- sions one may drav; from this work regarding the composition of a workable ruby gla.ss. A workable rul)y glass is one which will not color out when cooled at the rate obtained in the press- ing process, and yet will gi\e a workable range of temperature for reheating to a uniform color at temperatures below TOO". 1st, Highly Huid glasses will color out rapidly, viscous glasses slowly, 2nd. Keplacing iime with either lead or soda, increiises the rapidity of color development, lead more so than soda. 3d. High SiO. is necessary for good color, low SiO. gives a tendency towards brown or black, and opacity. •4th. High SiOo (4.0 to 4.5 moL), is necessary to give suffi- cient viscosity, 5th, With high silica, lime-potash glasses the tendency to streakiness increases. Small amounts of lead reduce streaki- ness. 6th. The glass giving the bast color in series B is No. 4. Glasses Xos. 1, 2, 10 and 12 of Series C, mo.st nearly approached the requirements of a good ruby glass. They could all be poured without the color developing, and on reheating, the color devel- oped at favorable temperatures. Glasses Xos. 6 and 8, Series C gave the most transparent colore. 7th. Iron and manganese are detrimental to a good red color. 8th. Remelting improves the uniformity of the color which indicates that streakiness is due to lack of homogeneity. 9th. Density of color is apparently increased with an in- crease in temperature. Time is evidently not an important fac- tor in this case. DISCUSSION Pfof. Silverman: There are a number of points in ^Mr. Williams' paper about which T wish to inquire. In the first place, he speaks of the coloring out in the high-silica copper rub- ies. I should like to ask whether ]\lr. Williams found any direct bearing by the alkali content of the glass. There is a claim 22 DEVELOPMENT OF RUBY COLOR IN GLASS made at i)resent that a copper, ruby eaii be mauufactiired, which is a ruby, out of the pot. I believe his views correspond with mine iu that the red color produced is due to high alkali in the glass. In other words, the glass colors out while cooling in the mold, OT even earlier.. Then as to tin as a reducing agent, I can corroborate these statements also, haidng had the experience that tin alone in connection with copper gives a rich color, while with juanganese and iron the color is olf. Tin has to be con- trolled very carefully. If you get below a certain point you obtain a glass which does not color sufficiently ; and if you go above you get what is caLed clouding or a livery color. I would like to ask, to what 'Sir. AVilliams attributes lack of uniformity of. color; and wiiether he feels that a melt over a short duration, like thirty minutes could give a homogeneous glass. Mr. \Villia))is: To answer the last (piestion first: the uni- formity of color in my gla.sses was not obtained in the first melt. There were signs of streakiiiess at first, l)ut u])()ii remelting, good clear colors wore obtained. It is probably tlie mechanical handling of the glass, or the duration of the melt which has a tendency to make the glass cloudy oi- ch'ai-. The first ({uestiou you asked, regarding the high alkali con- tent, I did not quite understaml, however I will make this point, that when I used lead, replacing the alkali, it caiLsed the colors to coine out move (luickly in the handling. The color was just as good, in fact a little better, but density of color could not be controlled. Lead impi-oved the uniformity of the color but gave a tendency toward opacity. If you do not want the color to come out during pressing, it is necessary to keep away from lead. Frof. Silverman: I should like to know further, what the object is in trying to prevent the coloi- from coming out during pressing. Mr. Williams: If you do not prevent it, the different var- iations in the cooling of the mold wouhl not uive the same shad- ing of red in the finished pieces. Prof. Silverman: But do you not get, the same effect by heating to a certain temperature afterwards? DEVELOPMENT OK KIHV COLOH IX GLASS 23 Mr. Williaiii^: W's: hut (•;in ynu cunt ml I lie i-;iti' ol' codliiii;' of glass ill th(' iiidlil .siit'liciciili}- ;ii'(Mii'iitcly as to L^-ivc iinit'dniiity of color from piece to piece .' Prof. Silvcn)i(ni : T cannot (iiiilc sec how that has a beariiiu' on the rate of cooliiiii'. Suppose your i^'lass does not color out below 7(XV\ You uiiuht have a mold nuywlici'c fi-om 400 1o (iOO. and the fact lliat ndu have no coloi' wouhl he no indication that your mold tempcratui'e is correct. In othei- words, you have such a large range below the coloring-out temperature that it does not seem any better indication as to mold temperature, than if you had a glass that colored out, except possibly to tel! you that the mold is too hot. Mr. Williarns: ]\Iy experience with glass that colored out was that glass of various thicknesses was different in shade. The difference in temperature of a mold would influence the coh)r. The coloring out at a definite temperature also depends upon the speed at which a glass cools through the small temperature range of color development. If the glass cools at a high rate of speed through this temperature, the colloidal copper would not come out in large enough particles to show color. If the cooling rate is slower the particles grow of sufficient size to give color. Mr. GeUiharp: I should like to ask whether that was not sub-oxide of copper. Mr. Williams: I used cuprous oxide. Prof. StuII: Perhaps I mig-ht throw a little light on Prof. Silverman's question by stating, that airoug the things "Sir. Wil- liams is investigating is a study of the temperatures at which the copper ruby comes out, and the effect of length of time as well as temperature in bringing it out. That is why he is trying to secure colorless glass to liegin with.