Reply to “Comment on ‘Influence of Fe 3+ /Fe 2+ Ratio on the Crystallization of Iron‐Rich Glasses Made with Industrial Wastes'”

OUR previous paper 1 described studies on modification of the crystallization ability and spinel/pyroxene ratio, as induced by thermal treatment and carbon addition, in high-iron-content glasses obtained from jarosite (J) and electric arc furnace dust (EAFD). The composition of these wastes is variable and complex. As an example, J-glass generated from one of the main zinc producers in the world (and a partner in the Brite-Euram CT94-018 project) has been reported in a simplified form; all the elements are not shown (the glass also has fractional amounts of cadmium, indium, silver, arsenic, antimony, mercury, germanium, and sulfur, or traces of nickel, cobalt, and thallium). The loss on ignition is very high, usually 35%–40%, because of the presence of SO4 2 and NH4 species; fractions of sulfur remain in the glass. This aspect introduces some uncertainties in the evaluation and tries to explain the results as a trend of behavior of the system under the investigated experimental conditions, i.e., powder (P) and bulk (B) glasses heated in air and nitrogen. It is also evident that these silicate glasses from industrial waste cannot be considered as model glasses, nor should their behavior be considered similar or compared to a phosphate model glass. The first comment notes that the weight gain due to the Fe oxidation, obtained via thermogravimetry (TG), is higher than that expected, according to the reported Fe /Fe ratio in the parent glass. The associated TG experimental error has been estimated to be 0.1% and 0.2% in the temperature intervals of 20°–500°C and 20°–1000°C, respectively. Therefore, the measured weight gain in the 500°–1000°C interval for the P-air, J-0 glass (jarosite glass without carbon)—0.44%, reported as 0.5% of the total weight gain—is in fair agreement with the theoretical value of 0.22%. The uncertainty on the determination of Fe ions and total iron, which influences the Fe /Fe ratio by 5%, also must be considered. A great difference is observed in the P-air, J-2 sample (jarosite glass with 2% carbon addition), where the theoretical value should be 0.82% but a value of 1.62% (reported as 2%) was measured in the 500°–1000°C interval. This anomaly is explained by the presence of carbon in the glass batch, which might have reduced some of the Fe , Zn , and Pb ions, as well as the other minor elements, to the elemental form. Previous research has assumed that, in silicate melts obtained in a reducing atmosphere, iron exists simultaneously in the following forms: Fe , Fe , and Fe. This assumption is confirmed by the large crystallization field of iron in the phase diagram of the FeO–Fe2O3–SiO2 system. 5 Metallic iron has not been observed via X-ray diffractometry (XRD), probably because of the amorphous structure. Metallic drops have been observed in EAFD glass with 5% carbon addition. In this case, an oxidation weight gain of 6.5 wt% is observed in the 700°– 1100°C temperature range and is accompanied by a very intensive and large exothermic effect that overlaps the crystallization peak, at 760°C, and the endothermic melting effects in the 1100°– 1300°C range. In regard to the two exothermic effects in the differential thermal analysis (DTA) trace of the J-0 sample, the first—large and with low intensity, corresponding to a weight gain of 0.4%—was attributed to a diffusion-controlled process with low reaction order (i.e., surface oxidation). In the temperature range where this exothermic effect was noticed (the first onset at 620°C is close to the glass transition (Tg 580°C)), the viscosity is very high ( 10–10 dPa s); the crystallization processes are very slow and cannot be identified via DTA. On the other hand, several studies were dedicated to the identification of the crystallization peaks in the TG-DTA of iron-rich jarosite glasses, presented in literature referenced in our previous paper. In particular, Karamanov et al. investigated a glass composition similar to that of the J-0 sample. The crystallization peaks, at a heating rate of 10°C/min, were observed at 780°, 805°, and 875°C for the P-N2, B-air, and P-air samples, respectively. A clear oxidation exothermic effect in the P-air sample was observed at 720°C, with a weight gain of 0.6%. In that work, the degree of phase transformation ( ) at 680°C was measured using density variation and XRD methods for the P and B samples. For the B sample, a consistent variation of density, which was related to magnetite and pyroxene formation, was observed, whereas for the P sample, the crystallization was negligible, even after 300 min. This result was attributed to the oxidation of the P sample with consequent inhibition of the crystallization process. Therefore, the weight gain was detected at 620° and 680°C and the oxidation was related to a simple diffusion process (square-root-of-time dependence). If we report the value of the B sample and the weight gain for the P sample on the same timescale, the plot in Fig. 1 is obtained. Thus, after 20 min, the crystallization is negligible, whereas oxidation comprises 30% of the total weight gain. Therefore, we concluded that, in these powder glasses, oxidation occurs at low temperature and inhibits the crystallization of magnetite and pyroxene. Figure 2 shows the isothermal crystallization “S” curve at 620° and 660°C