In situ infrared spectroscopy and mass spectrometry were used to investigate the gas and liquid phases in lithiumsulfur oxyhalide cells driven into anode-limited reversal at 1-5 mA/cm ~. In the lithium-thionyl chloride system the species HC1, CS2, SO~, S~O, SCI~, and SO..,C1._, were identified in the gas phase and HC1, A1CI:,OH-, SOs, SO=,CI.,, and SOC1 + A1C14-in the liquid phase. A species giving rise to three absorption bands at 1337, 1070, and 665 cm-' was observed in the liquid phase of that system during anode-limited reversal only, and in the lithium sulfuryl chloride system during normal discharge and during reversal; this compound was tentatively identified as Li(SO~, SO._,CI~) § AIC14-and is analogous to the well-known complexes involving LiA1C14, SO2, and SOCI.,. The lithium-sulfuryl chloride cell behaved similarly to the thionyl chloride cell, specifically with respect to formation of SO2-and SOCl+-like species--the latter tentatively identified as SO2C1 § Indirect evidence suggests that chlorine may accumulate in both systems at -20~ but at 25~ itsaccumulation in the cells is prevented by its reaction with SO2 to form SO2C12.In the ten years since the feasibility of lithium-sulfur oxyhalide cells was first recognized (1), remarkable progress has been made in hardware development; however, their widespread use has been impeded, owing to the safety hazards associated with their relatively high energy density. Conflicting reports exist as to the cause of several explosions involving the lithium-thionyl chloride system. Explosions have been observed in anode-limited cells driven into reversal (2); a 'hazard also exists in cathode-limited cells since they contain both lithium and sulfur at the end of discharge, and the highly exothermic formation of Li2S may be possible at elevated temperatures (3, 4). In addition, poor mechanical design is suspected in some cases.The lithium-sulfuryl chloride cell which does not form any sulfur during discharge is potentially safer than Li/SOC12 (5). However, the relatively high corrosion rate of lithium in that system limits its use to special high rate reserve applications.Resolution of the unpredictable safety hazards associated with these systems may only be possible by analytical determination of their causes, and several investigations have been conducted along these lines. Istone and Brodd (6) have performed an in situ infrared study of the electrolyte of Li/SOC12 cells and observed that only two absorption bands change in intensity during normal discharge, one at 1335 cm-' due to SO~ formation, the other at 689 cm -1 which they assigned to S~O. Elemental sulfur was also seen to deposit on the spectroscopic cell walls. These authors proposed a mechanism involving formation of SO2 and $20 with subsequent decomposition of the $20 to sulfur and SO2. This mechanism had been previously rejected on the basis of stoichiometry by Schlaikjer et al. (7), who postulated instead the formation of a sulfur monoxide polymer according to the reaction 2nLi + nSOCI~ = 2nLiC1 + (SO)...