ChemicalGeology, 107 (1993)83-95 83 Elsevier Science Publishers B.V., Amsterdam [AL] The analysis of forms of sulfur in ancient sediments and sedimentary rocks: comments and cautions Cynthia A. Rice, Michele L. Tuttle and Richard L. Reynolds U.S. Geological Surve_v, Box 25046, MS 916, Denver. CO 80225, USA (Received July 7, 1992; revised and accepted January 22, 1993 ) ABSTRACT Assumptions commonly made during analysis of the amount of monosulfides [acid-volatile sulfides (AVS) ] and disul- fides in modern sediments, may not be valid for ancient sedimentary rocks. It is known that ferric Iron can oxidize H2S during AVS analysis unless a reducing agent such as stannous chloride is added to the treatment. In addition, some mo- nosulfides such as greigite and pyrrhotite require heat during the AVS analysis in order to dissolve completely. However, the use of heat and/or stannous chloride in the AVS treatment may partially dissolve disulfides and it is generally recom- mended that stannous chloride not be used in the AVS treatment for modern sediments. Most of the monosulfides are assumed to be recovered as AVS without the addition of stannous chloride_ This study investigates the recovery of mono- sulfides during sulfur speciation analysis with application to ancient sedimentary rocks. Sulfur in samples containing naturally occurring greigite and mackinawite or pyrite was measured using variations of a common sulfur-speciation scheme. The sulfur-speciation scheme analyzes for monosulfide sulfur, disulfide sulfur, elemen- tal sulfur, inorganic sulfate and organically bound sulfur. The effects of heat, stannous chloride and ferric iron on the amounts of acid-volatile sulfide and disulfide recovered during treatment for AVS were investigated_ Isotopic composi- tions of the recovered sulfur species along with yields from an extended sulfur-speciation scheme were used to quantify the effects. Hot 6 N HCI AVS treatment recovers > 60% of the monosulfides as AVS in samples containing pure greigite and mackinawite. The remaining monosulfide sulfur is recovered in a subsequent elemental sulfur extraction. Hot 6 N HC1 plus stannous chloride recovers 100% of the monosulfides as AVS. The add:ition of ferric iron to pure gre~gite and mackinawite samples during AVS treatment without stannous chloride decreased the amount of monosulfides recovered as AVS and, if present in great enough concentration, oxidized some of the AVS to a form not recovered in later treatments. The hot stannous chloride AVS treatments dissolve < 5% of well-crystallized pyrite in this study. The amount of pyrite dissolved depends on grain size and crystallinity. Greigite in ancient sedimentary rocks was quantitatively recovered as AVS only with hot 6 NHC1 plus stannous chloride Hot 6 N HCI AVS treatment of these rocks did not detect any monosulfides in most samples_ A subsequent elemental sulfur extraction did not completely recover the oxidized monosulfides. Therefore, the use of stannous chloride plus heat is recommended in the AVS treatment of ancient sedimentary rocks if monosulfides are present and of interest. All assump- tions about the amount of monosulfides and disulfides recovered with the sulfur-speciation scheme used should be verified by extended sulfur-speciation and/or isotopic analysis of the species recovered. I. Introduction One of the most useful tools for evaluating the depositional conditions and diagenetic his- tory of sediments and sedimentary rocks is the study of sulfur geochemistry. Critical to the application of sulfur geochemistry is the quan- titative separation and collection of sedimen- tary sulfur species. Sulfur in sulfide minerals is operationally characterized as monosulfides or disulfides (Berner, 1964a; Berner et al., 1979; Canfield et al., 1986; Tuttle et al., 1986). The monosulfides are also referred to as acid-vola- tile sulfides (AVS) and analytically separated from the disulfides by their solubility in non- oxidizing acids. The acid-volatile sulfides in- clude mackinawite (Fe~.~LS), pyrrhotite (magnetic Feo.9S), and greigite (magnetic84 C.A. RICE ET AL. Fe3S4), as well as X-ray amorphous monosul- tides. The disulfides include marcasite (ortho- rhombic FeS2 ) and pyrite (cubic FeS2 ) and are commonly recovered in analytical schemes by chromium reduction methods after the AVS has been removed. For the purposes of this pa- per, pyrite is assumed to be the source of the disulfide sulfur. Treatments to recover AVS include the ad- dition of heat to the acid solution, which is necessary to dissolve crystalline pyrrhotite and greigite completely (Berner, 1964b, 1970; Tut- tie et al., 1986). Stannous chloride has also been added to the acid solution to prevent ox- idation of H2S produced in the AVS treatment with HC1 to elemental sulfur by reaction with ferric iron (Pruden and Bloomfield, 1968; Berner, 1974; Zhabina and Volkov, 1978; Ber- ner et al., 1979; Westrich, 1983; Cornwell and Morse, 1987; Tuttle, 1988; Tuttle et al., 1990). Until recently, the analytical procedures for sulfur-species separation were considered to be quantitative and highly selective. However, studies by Chanton and Martens (1985), Cornwell and Morse (1987), and Morse and Cornwell (1987) have shown that the sulfur- species separation is not as complete as had been believed, especially when the more vig- orous (heat and SnCI2) conditions are used to recover the AVS. Their studies indicated that measurable amounts of pyrite can be dissolved and mistakenly included as AVS, and that the proportion of pyrite dissolved depends on the analytical conditions and the grain size of the pyrite. The overestimation of AVS through py- rite dissolution can be significant if the pyrite/ AVS ratio is particularly large, as is reported in some recent marine sediments and salt marshes (Goldhaber and Kaplan, 1974; Howarth, 1984; Oenema, 1990a). Conversely, incomplete re- covery of AVS by milder acid digestions will result in overestimation of pyritic sulfur and underestimation of AVS, a significant error for samples from rapidly deposited sediments and some freshwater lakes in which relatively large amounts ofmonosulfides form (Berner, 1974; Goldhaber and Kaplan, 1974; Kelly and Rudd, 1984; Oenema, 1990b). Although considerable work has been done to assess the recovery of forms of sulfur under varying analytical conditions for recent sedi- ments