Evaluation of sampling methods for sulfur speciation in volcanic gases

Abstract

Determination of H2S and SO2 in fumarolic gases is necessary to predict volcanic activities. This study evaluated three gas sampling methods (CdS precipitation, Ag2S precipitation, and I2 oxidation) using artificial fumarolic gases with known compositions. H2S was initially fixed as CdS(s) or Ag2S(s) in the sulfide precipitations and S(0) in the I2 oxidation. On the other hand, SO2 was collected as dissolved species (SO32− and SO42−). The Ag2S precipitation, which did not allow HCl determination in fumarolic gases by forming AgCl(s), significantly underestimated H2S and SO2 due to the entrained loss of analytes as a result of water vaporization during H2O2 treatment. The I2 oxidation, which required duplicate sampling for complete analysis of H2S and SO2, tended to overestimate SO2 probably due to difficulties in separating S(0) particles from solutions. The CdS precipitation produced more satisfactory H2S and SO2 recoveries than the other methods. Additionally, the effect of delayed sample treatment was assessed by comparing H2S-to-SO2 molar ratios resulting from immediate and one-week-delayed treatments after gas collection. The delayed treatment significantly lowered the ratios in the CdS precipitation likely due to the susceptibility of CdS(s) to oxidation. In the meanwhile, such an effect was minimal in I2 oxidation. Similar results were also observed in the field evaluation of fumarolic gases at Mt. Hakone. Along with detailed sampling and treatment procedures, the causes for unsatisfactory H2S and SO2 recoveries were discussed in the context of the underlying reaction schemes as well as potential experimental artifacts. Thus, this study helps to select and further modify appropriate gas sampling methods to meet specific needs.