Journal of Non-Crystalline Solids 38 & 39 (1980) 251-256 O North-Holland Publishing Company A STUDY ON Cr3+/Cr ~+ EOUILIBRIA IN INDUSTRIAL EMERALD GREEN GLASSES O. GUldal and C. ADak Research and Development Department TUrkiye $ise ve Cam Fabrikalarl A.S. Istanbul TURKEY A study is made to determine the redox equilibria of chromium ions and the factors affecting these equilibria in a soda- lime-silica-glass. The concentrations of individual oxidation states of chromium are determined by both physical and chemi- cal methods. The results of the two methods are found to be in good agreement. Sulfate added as saltcake is seen to shift the equilibrium towards higher hexavalent chromium concentration, whereas iron present in the batch hinders this transformation and stabilizes trivalent chromium until all iron present in the glass is converted to the ferric state. Cobalt, on the other hand, has no effect on this equilibrium. Provided that the sulfate and iron contents of the batch are well adjusted, the redox of the batch and the furnace atmosphere during melting have little effect on the oxidation state of the chromium atoms in the final glass. Consequently, it has been possible to apply the results of the laboratory scale melts directly to the industrial furnaces. INTRODUCTION Chromium has long been used as a glass coloring agent because of the deep green color it imparts to conventional olasses. This use is continued in the production of glass containers for soft drinks. Recently, soft drink manufacturers have not only imposed stringent color specifications on the containers they use, but they also require the containers to possess special radiation-protecting features, especially against UV. A good understanding of the chromium equilibria in glass will help the glass maker in fulfilling the requirements of the market. In soda-lime-silica glasses, chromium can exist either in the trivalent or the hexavalent form [I]. The more stable trivalent chromium has two absorption bands at 450 and 650 nm. The overall effect of these absorption bands is that it gives the glass a deep emerald green color. Hexavalent chromium on the other hand has a very strong absorption band centered at 370 nm. A part of this band extends into the visible spectrum and gives the glass a bright yellow color. Much work has been done to determine the factors affectinq the chromium equilibrium in glasses. Moore [2] has investigated the effects of various concentrations of copper, iron, cobalt and chromium oxides on the tristimulus color values and the spectral trans- mission curves of an industrial glass. He has concluded that, upto 0.2% Fe203 251
`
`O-I Glass, Inc.
`Exhibit 1031
`Page 001
`
`

`

`252 U. G~idal, C. Apak / Industrial Emerald Green Glasses added into a glass containing 0.II% Cr203 (added as K2Cr207) had no effect on the color or the UV cut-off. In addition, he has found nitrate to be equally ineffect- ive in oxidizing furnace atmospheres. CuO and Co304 added in glass decreased the dominant wavelength, brightness and purity. CuO shifted the UV cut-off to lower wavelengths, whereas Co304 pushed it to higher wavelengths. Chromium, as expected, only increased the purity and decreased the brightness of the color. Knupp and Berger [3], too, have investigated the effects of ferric oxide on the chromium equilibria in the UV absorbing green glasses and have found that an addition of 0.097 to 1.60% Fe203 to have a reducing effect thus altering the equilibria in favor of trivalent chromium. In an other work by the same authors [4] the effects of atmospheric oxygen and the sulfate in the batch are discussed. The atmospheric oxygen is claimed to favor the formation of hexavalent chromium~ whereas the decomposition products of sulfates (S02 and S03) hinder this formation. The effect of copper oxide on the chromium equilibria has also been studied by Henning [5] who has found that the additions of CuO shift the equilibrium towards trivalent chromium. While working on chemical analysis of some elements in glass Close and Tillman [6] have found that when two elements, each of which can exist in two oxidation states with sufficiently different redox potentials, are present in a glass, then only one of the elements can be present in both of its possible oxidation forms. They have also observed that the Fe-Cr system obeyed this rule. In this work, information on the chromium equilibria is verified and extended to give the technologist a guideline for designing glass colors with the use of chromium and other additives. DETERMINATIONS OF Cr 3+ AND Cr 6+ IN GLASS Chemical determinations performed on glass are time consuming. Therefore, an alternative method was sought for the determination of Cr 3÷ and Cr6+ concentrations. A sufficiently reliable physical technique requiring minimal sample preparation and applicable under factory conditions would therefore be very valuable. For this purpose, the Cr3+ and Cr G+ contents of a set of laboratory samples were determined by wet chemical methods as well as spectrophotometrically and the results were compared. In the chemical method, the total chromium in glass is taken into solution by dissolving the parent glass in a mixture of hydrofluoric and perchloric acids. The optical density of the solution at 540 nm is measured against a blank after complexing with diphenyl carbazide and the total chromium content of the sample is read off a calibration curve. Hexavalent chromium is separately determined from a sample prepared by the cold decomposition of the parent glass in a mixture of hydrochloric and hydrofluoric acids into which a known amount of arsenous oxide solution is added to reduce the hexavalent chromium. Subsequently, the excess of the arsenous oxide is back titrated with potassium iodate. In the physical method, the optical density of a 2-3 mm thick, polished sample is determined at 370 and 650 nm and the respective concentrations of Cr3+ and Cr6+ are calculated by using tabulated extinction coefficients for each species. Referring to the figures 1 and 2 the agreement between the results of the two methods is satisfactory. Therefore, the physical method was adopted for routine determi nations.
`
`O-I Glass, Inc.
`Exhibit 1031
`Page 002
`
`

`

`U. G~idal, C. Apak / Industrial Emerald Green Glasses 253 0.16 • A 0.1/.. >., #. 0,1~ o ~ U o~ 0.10 ,/~ i i i i 0.10 012 01z. 0.'16 % C~03(Chcmicol) Figure 1 Comparison of the physical and chemical determinations of total chromium in glass (expressed as Cr203) 0.012 8 O010 aE n C ,j" 0008 OOOE 0.014 0006 0.008 0.010 0.612 ' "/o Cr203 ( Chcmicel ) Figure 2 Comparison of the physical and chemical determinations of hexavalent chromium in glass (expressed as Cr203) 0.014 LABORATORY MELTS The laboratory melts were made using industrial raw materials. Chromium was added as chromite ore containing about 15% iron oxides. Additional iron was introduced as pure Fe203. Batches prepared to give I00 g glass were placed in high alumina pots and melted in an electric furnace at 1430oc for a period of 6 hours. The melt was cast into 15x20x40 mm blocks,annealed to room temperature. By subsequent grinding and polishing 3 mm thick spectrophotometric samples were obtained. The main batch composition for I00 g glass consisted of the following ingredients. Sand : 68.1 g Feldspar : 6.4 g Limestone: 7.4 g Dolomite : 14.9 g Soda ash : 23.2 g Necessary corrections in the batch composition were made depending on the quantity of the minor additives. The concentrations of the ingredients that are effective on color are given in Table I. In the table, Fe203 represents the total iron in glass contributed by iron bearing chromite, raw material impurities, and that added as Fe203. DISCUSSION The transmission curves of the samples in the visible part of the spectrum correct- ed to 4 mm glass thickness are given in Figures 3 to 7. From Figure 3, it is clear that an addition of 0.4% SO 3 into the batch as Na2S04 favors the formation of hexavalent chromium. This is an expected consequence of the oxidizing nature of Na2SO 4. However, introduction of increasing amounts of carbon into a sulfate-containing batch does not have any effect on the chromium
`
`O-I Glass, Inc.
`Exhibit 1031
`Page 003
`
`

`

`254 U. Guldal, C. Apak / Industrial Emerald Green Glasses Table 1. Concentration of minor ingredients effective on color (%) Melt Cr203 Fe203 CoO S03 C A1 0.232 0.107 B1 0.356 0.143 0.40 C1 0,270 0,118 0.40 C2 0.270 0.118 0.40 C3 0,270 0.118 0.40 D1 0.230 0.400 0.30 D2 0.230 0,400 0.30 D3 0.230 0.109 0.30 D4 0.230 0.109 0.40 El 0.230 0.400 0.30 F1 0.164 0.220 0.30 F2 0.164 0,132 0.30 F3 0.164 0.132 0.40 F4 0.130 0,122 0.30 F5 0.130 0.122 0,004 0.30 G1 0.120 0.119 0.004 0.30 G2 0,115 0.118 0.006 0.40 G3 0.II0 0,116 0.008 0.40 0.025 0.089 0.I00 0.150 0.250 0.018 0.081 0,007 0.026 0.048 0.070 equilibrium. This is evident from Figure 4. In the figure, samples C1 to C3 contain increasing amounts of carbon without any effect on their short wavelenqth transmissions. Figure 5 shows the transmission curves of samples with varying amounts of iron oxide. Here, the marked effect of the iron oxide on the formation of hexavalent chromium is seen. Samples D3 and D4 with low Fe203 content show a strong Cr 6+ absorption, whereas a corresponding absorption band is seen to be absent from the spectra of samples D1 and D2 containinq 0.4% Fe203, Referring to Figure 3, and comparing the spectra it is seen that sample El containing 0.4% Fe203 and 0.3% SO 3 has only slightly more hexavalent chromium than sample AI, containing no sulfate and only 0.10% Fe203. Figure 6 shows the combined effects of various factors on the transmission curves of the glasses. SampleF2 is obtained by the removal of free Fe203 and from the batch of FI. Consequently formation of hexavalent chromium becomes possible. In addition to this change, increasinq the S03 to 0.4% in F3, further increases the hexavalent chromium concentration. Because chromium is introduced to the glass as iron chromate a decrease in chromium as in the sample F4 is accompanied by a proportionate decrease in the iron content of the glass, resulting in a further increase of hexavalent chromium. F5 is obtained by an addition of 0.004% CoO on F4. This addition does not have a significant effect on the chromium equilibria. Figure 7 illustrates the transmission curves of the samples G1 to G3 which are prepared with increasing amounts of cobalt oxide. Analysis of the transmission curves reveals that cobalt oxide does not affect the chromium equilibria and the absorption of cobalt is quantitatively reflected in the transmission curve. CONCLUSION The available data show that the ratio of the hexavalent to trivalent chromium in glass is sensitive to the other oxidizing and reducing materials in the batch. But the effectiveness of these agents vary from one material to another. Iron oxide is the most effective of all the additives. It reduces the chromium to trivalent state and keeps it at this state in spite of a strong oxidizing agent
`
`O-I Glass, Inc.
`Exhibit 1031
`Page 004
`
`

`

`U. Gu'Idal, C. Apak / Industrial Emerald Green Glasses 255 70. 60. ~~~ 50 40 o "~ ~ E .m 30 E 10, 0 4oo t.io ~o sio s~o 6~ 6~.o 68o Wovelzn9 th (nm) Figure 3 Transmission curves of glasses Al, Bl and El 70. 60 ¸ 50, "G .~ 30 20. 10" O. ~[~ 4~ ~0 4SO 52o 56o 6o0 640 Wavelength (nm) Figure 5 Transmission curves of glasses D1 to D4 680 like sodium sulfate. '/0. 6oi • .-~ 50. 40' 30. F: ~20- o 1- 10- 70- 04 ., m d.o ~so s~o s~o 660 ~o 68o Wavelength (nrn) Figure 4 Transmission curves of glasses C1 to C3 60. o ° 50 ,~40 ~30 c O ~ 2o 10 70 A // ~ \ < e ~ ~./~3 ./ ~oo ~o ~8o s2o stm 6OO 62,0 68o Wovelencjth (nm) Figure 6 Transmission curves of glasses F1 to F5 ~60 Sulfate in the absence of iron oxide "~ so has a very strong oxidizing effect. This effect can not be neutralized m m 40 even by carbon. Therefore, it can be '~ concluded that oxidizing and reducing ~ 30 agents which decompose during the earlier stages of melting do not have ~ 2o. a lasting effect on the chromium equilibrium. As an extension of this I0, conclusion, one would expect alkali nitrates alone to be ineffective whereas a mixture of alkali nitrates 4i with arsenic oxide to be a very effect- ive oxidizing agents for chromium° ,/v , Gt 4Zo ,-80 s~o s60 660 6Zo 6~0 Wev¢len 9th (nm) Figure 7 Transmission curves of glasses G1 to G3
`
`O-I Glass, Inc.
`Exhibit 1031
`Page 005
`
`

`

`256 U. GJldal, C. Apak / Industrial Emerald Green Glasses Owing to the stability of iron oxides at high temperatures, furnace atmosphere will not affect the reduced chromium glasses containing iron oxide. This fact is also verified experimentally. The composition given as sample F1 is melted both in an electrically heated laboratory furnace and in a I00 ton/day oil fired industrial tank. The transmission curves of the two glasses obtained were identical. Cobalt oxide has no detectable effect on the redox equilibrium in these glasses. This permits the glass technologist to design a color first by obtaining the right concentrations of hexavalent and trivalent chromium ions and subsequently calculating the effects that different concentrations of cobalt would have on the color. REFERENCES [I] Bamford, A.C. Colour generation and control in glasses, (Elsevier Scientific Pub. Co., N.Y. 1974) p.45 [2] Moore, H., Ultraviolet absorbing green glass, The Glass Ind., 5(1964) 244 [3] Knupp, R.C. and Berger, D.F. Effects of iron in ultraviolet absorbing green glass, The Glass Ind., 5(1966) 253 [43 Knupp, R.C. and Berger, D.F., Colour characteristics of ultraviolet absorbing green glass. Ceram. Bull. 3(1969) 244 [5] Henning, H., Ober die gegenseitige Beeinfl~ssung der Oxide von Kupfer und Chrom in einem N-K-S-Glas und die Auswirkung auf Farbe und Spektrale Durchl~ssigkeit, Glass-Email-Keramo-Tech., 3(1969) 77-85 [6] Close, W.P., and Tillman, J.F., Chemical analysis of some elements in oxidation reduction equilibria in silicate glasses, Glass Tech., 5(1969) 134-146.
`
`O-I Glass, Inc.
`Exhibit 1031
`Page 006
`
`