The dissolution mechanism of sulphur in hydrous silicate melts. I : assessment of analytical techniques in determining the sulphur speciation in iron-free to iron-poor glasses

Klimm, Kevin and Kohn, Simon C. and O'Dell, Luke A. and Botcharnikov, Roman E. and Smith, Mark E. (2012) The dissolution mechanism of sulphur in hydrous silicate melts. I : assessment of analytical techniques in determining the sulphur speciation in iron-free to iron-poor glasses. Chemical Geology, 322-32. pp. 237-249. ISSN 0009-2541

Abstract

S K-edge XANES, λ (SKα) wavelength shift, 33S MAS NMR and Raman spectroscopy have been applied to a series of experimentally synthesised hydrous sulphur bearing silicate glasses to determine the oxidation state of sulphur dissolved in the glass. Glasses investigated include soda–lime glass (SLG), K2Si4O9 (KSG) and albitic (Albite) and trondhjemitic (TROND) glass compositions. The four spectroscopic techniques are compared with each other to investigate the applicability of each technique as a method to determine the sulphur oxidation state and structural aspects of sulphur dissolution in silicate melts. XANES and Raman spectroscopy provide the most comprehensive information on the S-speciation and S6 + and S2 − (plus S4 + in the case of XANES) can be detected. XANES allows detection of S-species at lower concentrations (< 50 ppm) than all the other methods. NMR also allows direct determination of S6 + and S2 − in the glasses. However, signal loss in NMR prevents both S6 + and S2 − to be detected simultaneously even in samples where the presence of both species can be confirmed using XANES and Raman. Measuring the λ (SKα) wavelength shift does not provide direct constraints on the S-species in silicate glasses but can be used to determine the S6 +/ΣS when corrected for beam damage. λ (SKα) wavelength shift determination works best for S and Fe-rich glasses because in Fe-free hydrous glasses and at low S contents significant beam damage is observed, requiring an intensive correction procedure to account for changes of S6 +/ΣS during electron beam exposure. The other techniques show either no (NMR and Raman) or minimal (XANES) beam damage that can be easily avoided by moving the sample during spectra acquisition. The linear combination of XANES spectra of glasses containing either only S2 − or S6 + provides direct constraints on the S6 +/ΣS of glasses containing a mixture of S2 − and S6 +. In addition the Raman intensities of bands related to S6 + and S2 − can also be used to quantify S6 +/ΣS and determined values are in good agreement with those obtained by XANES at least for S contents of > 150 ppm. XANES and Raman both show that the dissolution of S2 − is different in Fe-free and Fe-bearing hydrous glasses. In Fe-free glasses S2 − is present as SH− and H2S whereas Fe–S complexes are observed in the presence of Fe.