Thermal Separation of Arsenic and Antimony Oxides G.A. BROOKS, W.J. RANKIN, and N.B. GRAY Experiments were carried out to remove arsenic from antimony trioxide by two techniques: first, by selectively volatilizing the more volatile arsenic trioxide from a mixed oxide sample; and, second, by selectively condensing the less volatile antimony tetroxide at high temperatures, leaving arsenic trioxide to condense out at a lower temperature. Thermodynamic analysis of the As-Sb-O system indicated that if arsenic and antimony oxides behave like pure solid phases, then there is no limitation to producing pure antimony oxide by these proposed techniques. The selective volatilization experiments were carried out at 379 ~ to 587 ~ using both nitrogen and air as carrier gases and an industrial antimony trioxide fume containing 13.8 wt pct As. The results showed it is difficult to achieve an arsenic content below 5.0 wt pct using either air or nitrogen, and formation of solid solutions between arsenic and antimony trioxides appears to be the main barrier to the removal of arsenic. Selective condensation experiments were carried out in which a mixed oxide vapor was progressively cooled through a series of condensers over a controlled temperature profile. Injection of oxygen into the vapor improved separation, and antimony tetroxide containing as low as 0.23 wt pct As was obtained. The recovery of antimony tetroxide in the experiments seems to have been limited by kinetic factors, and the results sug- gest that high conversions to antimony tetroxide are likely to be achieved only in antimony- rich vapors. I. INTRODUCTION THE problem of separating arsenic from antimony oxides arises as one of the purification steps in the pro- duction of commercial-grade antimony oxide (<0.5 pct As) or arsenic oxide from fume produced in metallur- gical processes. A simple technique for separating ar- senic oxide from antimony oxide would allow commercial exploitation of otherwise hazardous by-products. The separation of antimony oxide from arsenic oxide has been barely covered in any theoretical or experimental detail in the literature. In particular, there has been a lack of attention to the chemistry of the mixed oxide system. Only two industrial processes are known that separate arsenic oxide from antimony oxide. These are arsenic- trioxide refining by selective condensation, which pro- duces commercial-grade arsenic trioxide from a mixed oxide feed r~,2] and the WR Metal Industries (WRMI) pressure-leaching process, which produces commercial- grade arsenic acid from a mixed oxide feed. p] In both processes, antimony oxide is recovered in an impure form. Other industrial techniques separate arsenic and anti- mony in their sulfide state, as metal or as chlorides in ~olution. Two pyrometallurgical techniques for separating ar- senic oxide from antimony oxide have been proposed. [4] 1. Selective volatilization of the more volatile arsenic rioxide from a mixed oxide sample to leave high-purity mtimony oxide. ~.. Selective condensation of antimony oxide from ar- ;enic oxide in the vapor phase, by condensing the less rolatile antimony oxide at a higher temperature. G.A. BROOKS, formerly with the G.K. Williams Cooperative ',esearch Centre for Extractive Metallurgy, is Lecturer, Department f Materials Engineering, The University of Wollongong, Wollongong 500, Australia. W.J. RANKIN, Director, and N.B. GRAY, Reader, re with the G.K. Williams Cooperative Research Centre for Extractive 4etallurgy, Department of Chemical Engineering, The University of lelbourne, Parkville 3052, Australia. Manuscript submitted December 17, 1993. In the case of selective volatilization, no separation process producing commercial-grade product has been reported. In the case of selective condensation, the ex- perimental work of Read I~] showed that arsenic trioxide with less than 0.3 pct Sb can be condensed selectively from a mixed oxide vapor. The vapor forms of antimony and arsenic oxides are Sb40 6 and As40 6. Mixed oxides also form in the gaseous state when arsenic and anti- mony are both present. According to Norman and StaleyfSl and Mauser, trJ the mixed oxides of arsenic and antimony that can exist in the vapor state are As3SbOr, As2Sb2Or, and AsSb306. Hager and Li tTj proposed improving the separation of antimony oxide from mixed oxide vapors by injecting oxygen to form nonvolatile antimony tetroxide, accord- ing to the following reactions: Sb406 (g) + Oz (g) ~ 2 Sb20 4 (s) [1] AsaSbO6 (g) + 0.25 02 (g) 0.5 Sb204 (s) q- 0.75 As406 (g) [2] As2SbzO6 (g) + 0.50 02 (g) --~ Sb204 (s) + 0.50 As406 (g) [3] AsZb306 (g) + 0.25 O~ (g) 1.5 5b204 (S) q- 0.25 As406 (g) [4] Hager and Li carded out selective condensation exper- iments using an air-cooled tube at the exit of a transpi- ration apparatus. They observed that in an inert atmosphere, decomposition of the vapor complexes re- suited in deposition of predominantly cubic As203 in the cold zone and a mixture of cubic and orlhorhombic 5b203 in the hot zone of the tube. With oxygen injection at the entrance of the condenser tube, decomposition of the complexes resulted in deposition of cubic As203 in the cold zone and a mix of cubic 5b203, orthorhombic 5b203, and 5h204 in the hot zone. The best separation [ETALLURGICAL AND MATERIALS TRANSACTIONSB VOLUME '25B, DECEMBER 1994--873