Volatilization of glass components and formation of inhomogeneity in a glass melt during heating often result in striae, devitrification or fracture of the final glass products. Borosilicate glasses were known of their relatively high volatility at the temperatures of melting.
Volatilization loss at 1410°C was measured by thermo-balance method on the ternary Na₂O-B₂O₃-SiO₂ glasses in the composition range of 0-40 Na₂O, 0-40 B₂O₃ and 60-85 SiO₂ in wt%. The glass batches were prepared using purified sand and reagent grade chemicals and were melted in a platinum crucible in electric or gas-fired furnace. For the weight loss measurements, the glasses were held in small platinum crucible suspended in a vertical electric furnace. Results were as follows:
1) The amont of volatilization loss in a fixed time (for example, 1 or 6hr) and with fixed Na₂O/B₂O₃ ratio decreased with increase of the SiO₂ content, and for the constant SiO₂ content the loss had its maximum at Na₂O/B₂O₃ ratio slightly lower than 1. At these maxima, the losses were far greater than those of the corresponding Na₂O- or B₂O₃-SiO₂ binary glasses.
2) In almost all the glasses, the loss increased linearly with time. at the initial periods of heating. Generally, however, volatilization rates decreased with time.
3) Especially in glasses containing larger than 75wt% SiO₂, crystallized surface layeres were formed in the course of the measurement, and in consequence, the volatilization rates decreased distinctly.
4) Under the condition of the measurements, thermal convection was observed in the crucible. Contrary to the case of glasses with relatively high SiO₂ contents, in glasses with low SiO₂ contents thermal convection took place distinctly that the considerable mixing phenomenon was observed in the glass melts.
5) One of the possible mechanisms which explain the results described in 1) and 2) is given in the following: At the initial periods of heating, volatilization rate is controlled by the evaporation of the volatile components from the melt to the atmosphere or by the flowing away of the gas atmosphere near the surface of the melt. Then, diffusion controlled mechanism or the effect of the lowering of the concentration of the volatile components in the melt gradually predominates, and lowers the volatilization rate.
6) Quatitatively, the relation between composition an volatilization loss showed a good agreement with the results of kolykov measuring at 1200°C (ref. 5). Quantitatively, however, the loss was 10-100 times larger than those by kolykov owing to the differences of the mesuring temperatures. Under the assumption that the initial volatilization rate is proportionate to the vapour pressure of glasses, it could be inferred that phase separation in the glass melts occur in the composition range with relatively low Na₂O contents.

View full abstract

Arsenic trisulfide glass is of interest as a material of higher refractive index, transparent in the infra-red and having low melting characteristics.
The structure of arsenic trisulfide glass has been studied by X-ray diffraction technique, and the data indicate a short range order arrangements of the atoms similar to those in the crystalline orpiment structure. However, there have been very few reports of structure determinations of sulfide glasses which have higher content of sulfur than that of As₂S₃. In undertaking to provide the necessary structural information, the infra-red absorption of samples with vorious compositions was examined.
As₂S₃, one of raw materials, is synthesized by passing purified H₂S gas into 6N-HCl solution of purified As₂O₃, and dried at temperatures below 100°C, but it was very difficult to remove the small content of Cl^- included in As₂S₃. This crude As₂S₃ formed black-red glasses. To obtain blight red glasses, As₂S₃ must be purified by vacuum distillation.
In oftaining the glass, at first, the raw materials were mixed up and melted one hour at 360°C by using porcelain crucible, but in this case it was found by infra-red absorption method that the glass contains some oxygen impurities (Fig. 2, Table. 3). Therefore, the method of preparation of the glass samples was changed in the following way.
Sufficiently pure elements that exactly weighted for yielding 5-30 grammes of glass were stored in the thick-walled pyrex-type glass ampoules, which were then evacuated and sealed while still under vacuum. The ampoules were heated up to 600°C or 450°C in a furnace, and then quenched or allowed to cool down to room temperature. Infra-red absorption measurements between 1000-400cm^-1 were made through polished parralle-sided discs, approximately 0.2-2 millimeters thick, or through very thin film sandwitched by KBr plates, using a Hitachi-Perkin Elmer Model 125 and/or Japan Spectroscopic DS-402 Grating Infrared Spectrometer or Hitachi Infracord.
The results obtained were shown in the Table. 4 anb 5.
Kolomites an Pavlov suggested that the absorption band appeared at around 800cm^-1 is due to As₂Se₃. To ascertain their suggestions, the As-S glass samples contained Se (No. 7) were prepared, yet it resulted in no indication of absorption band at around 800cm^-1. According to the study by Tanaka et al. on arsenic-sulfur glasses prepared by evaporation under normal pressure, it was suggested that the absorption band appered at around 800cm^-1 is attributed to the vibration of As=S bond in the As₂S₅ glass, while the absorption band at around 790cm^-1 is due to the inclusion of arsenic disulfide (As₂S₂ or As₄S₄) component dissolved in the As₂S₃ glass. For the experiments of Kolomiets, the use of pure elements did not bring in the absorption band at around 800cm^-1 (from Glass No. 73-83). On the contrary, when unpurified elements were used or arsenic oxide was added to As-S glass, the absorption bands at around 800cm^-1 (in the case of high content of S) and around 790cm^-1 (in the case of low content of S) appeared. And the absorption band shifted gradually from 790cm^-1 to 800cm^-1 with increasing S content (as indicated in Table. 4). When arsenic oxide was added to As-S glass, SO₂ gas produced by chemical reaction between As₂O₃ and S was ascertained by the appearance of special infra-red absorption bands (1330cm^-1 and 1150cm^-1) and by the gas chromatographic analysis. On

View full abstract