Technical paper Speciation and Isotopie Composition of Sulphur in Low-rank Coals from Four Slovenian Coal Seams Martina Sturm,1'* Sonja Lojen,1 Miloš Markic2 and Jože Pezdic3 'Department of Environmental Sciences, "Jožef Stefan" Institute, 1000 Ljubljana, Slovenia 2 Geological Survey of Slovenia, Dimiceva ulica 14, 1000 Ljubljana, Slovenia 3 Department of Geology, Faculty of Natural Sciences and Engineering, 1000 Ljubljana Slovenia * Corresponding author: E-mail: martina.sturm@ijs.si Received: 14-01-2009 Abstract The abundance, distribution and isotopie composition of sulphur in low-rank coals from four Slovenian coal seams are presented. Samples were characterised as medium (1-3 wt%) or as high sulphur coals (>3 wt%), with the majority of sulphur occurring in the organic form. The isotopic composition of sulphur in coals in Slovenia was determined for the first time. It was found that 534S varies between different coal seams, as well as between different lithotypes inside particular coal seams. This can be explained by different SO42- and Fe2+ availability and microbial activity inside the coal forming basin. Keywords: Coal, sulphur, stable isotopes, lignite 1. Introduction A knowledge of the abundance and nature of sulphur (S) in coal is important in coal utilization because sulphur oxides released during coal combustion can be a major source of acid rain.1 Sulphur can be present in coal as iron sulphides (mostly as pyrite and marcasite, seldom as sphalerite, galenite, or halcopyrite), as sulphates (mostly as gypsum or barite, sometimes as iron sulphate, formed by oxidation of sulphides, i.e. coquimbite, szomolno-kite), as elemental sulphur or as organically bound sulphur. Variation in the sulphur content of coal is controlled mainly by geological conditions,2 such as the acidity of the peat and by the sulphate content of waters within the peat.3 Coals have been graded according to their sulphur content into low-sulphur coals that contain less than 1 wt% total sulphur, medium-sulphur coals that contain 1-3 wt% total sulphur and high-sulphur coal that contain more than 3 wt% total sulphur.4 It is well known that marine-influenced peats generally have higher sulphur contents than freshwater-influenced ones3 and high sulphur contents in coal seams cannot be accounted for by the original plant sulphur. Therefore, some sulphur must have been incorporated into the peat after the plants died and were buried.2 Sulphur available during plant growth is the principal source of sulphur in low-sulphur coals. In contrast, in medium and high-sulphur coals overlain by a marine roof, most of the sulphur is derived from seawater sulphate.1'2'5 Many researchers5'6'7 emphasized that freshwater coals deposited in calcium-rich environments have similar sulphur contents as brackish-marine coals. Sulphur isotopes can be used to interpret the origin of sulphur and provide important information on processes within peat-forming depositional environments. In general, the 34S/32S ratios in medium and high sulphur coals are variable, suggesting microbiological isotopic fractio-nation,2 which favours the incorporation of 32S into the reaction products. Where sulphate concentrations remain low and limited during early diagenesis, such as in nonmarine environments, near-quantitative bacterial reduction of all the available sulphate results in the addition of comparatively 34S-rich sulphur species to the coal.2 Bacterial reduction of sulphate produces hydrogen sulphide (H2S), which reacts either with Fe2+ or organic matter during early diagenesis to form sulphides or orga-nosulphur compounds.2 The Fe2+ source comprises two pools: detrital Fe2+ within the peat and Fe2+ diffusing from underlying sediments.8 It is well known that Fe2+ has a greater affinity for H2S than organic material does.2 9 Under conditions where SO42- is abundant and bacterial activity is constant, organic matter and H2S can react to produce organic sulphur. However, in closed system conditions under limited 32SO42- availability, both organic matter and Fe2+ incorporate H2S enriched in 34S, producing pyrite and organic sulphur with higher 534S values.2 In the presented work sulphur abundance, specia-tion and for the first time the isotopic composition of low-rank coals from four coal seams in Slovenia were studied. croscopy following the criteria for classification of mace-rals of soft brown coals - lignites after ICCP (1971, 1975).11 Microscopic analysis was performed only qualitatively to estimate the concentration of sulphide (pyrite and/or marcasite) grains. In the laboratory, samples were pulverized to a fine powder using an agate mortar and pestle and homogenized before subsamples were taken for further analyses. The samples were dried at 40 °C to constant weight prior 2. Experimental Eleven low-rank coal samples from four different Slovenian coal seams were taken (with the number of samples and sampling description in parentheses), as follows. The Velenje lignite seam (5; 4 different lithotypes, taken afresh) (Fig. 1); the Kanižarica coal seam (4; samples already taken for previous studies) (Fig. 2); the Tr-bovlje-Hrastnik coal seam, Ojstro coal pit (1; taken afresh) (Fig. 3) and the Senovo coal seam (1; taken from the coal mine collection) (Fig. 4). Samples were stored in sealed plastic vials or plastic bags. The abundance and isotopic composition of sulphur, carbon and nitrogen were determined for all samples. The macroscopic description of the lignite samples in terms of lithotypes was determined following the criteria for lithotype classification for soft brown coals (lignites) of the International Committee for Coal Petrology (ICCP, 1993).10 The microscopic description was determined by optical reflection mi- Figure 1: A sector from the general lithological column of the Velenje (after12). Figure 2: Coal-bearing interval from the V53/89 borehole of the Kanižarica coal seam (after13). Figure 3: Lithological column of the Trbovlje coal-bearing formation of Oligocene age (based on14; from15). Figure 4: Geological profile of the Senovo coal deposit in N-S direction (from16). to the sequential analysis of sulphate (SULPH), mono-sulphides estimated as acid volatile sulphides (AVS), sulphides estimated as total chromium reducible sulphur (CRS), and bulk insoluble organic (ORG) sulphur which was than determined gravimetrically with a limit of detection of 0.01 mg. Total (TOT) sulphur content was determined as the sum of the separate species. The above sequential extraction of sulphur species was made following the procedure described in references.1718 The reproducibility and yield of the method were determined during optimisation of the procedure on organic-rich sediment samples (1) by replicate analyses of the same samples, (2) by comparing the concentrations of separate sulphur species to the total S as determined by elemental analysis, and (3) by the analysis of pure calcite and silica powder mixtures with known concentrations of added S-species (sphalerite, pyrite and Ca sulphate as substitutes for AVS, CRS, and sulphate, respectively). The average yield was better than 96%, and repeatable wit- hin 10%. Most important, differences in reaction yield did not significantly affect the isotopic composition of the precipitates, as the measured 5 values of parallel samples were within 1 %o, which is negligible compared to the isotope separation between different S species. The concentration and isotopic composition of sulphur was determined for each sulphur fraction separately, as well as for total nitrogen (NTOT) and organic carbon (CORG). Prior to sulphur isotope analyses, Ag2S was mixed with an equal amount of V2O5, whereas sulphate was mixed with equal amounts of V2O5 and SiO2 to facilitate the combustion of samples in the elemental analyser. Homogenised sample mixtures were put into tin capsules. Samples for carbon analysis were pulverized, homogenized and soaked with 3 M HCl overnight to remove carbonates, rinsed with deionised water and dried prior to analysis, whereas pulverized samples untreated with HCl were analysed for determination of the N isotope composition. The concentration (C, N) and isotopic composition (C, N, S) of the samples were determined simultaneously using an Europa 20-20 continuous flow isotope ratio mass spectrometer with an ANCA-SL preparation module (PDZ Europa Ltd., U.K.), connected to an elemental analyser. The 34S/32S, 13C/12C and 15N/14N ratios are reported in delta notation (5) indicating parts per thousand (%«) as deviations from default reference materials: Vienna Canyon Diablo Troilite (VCDT), Vienna Pee Dee Belem-nite (VPDB) and AIR, respectively, as follows: (1) where R is the ratio between the heavier and lighter stable isotope of a particular element, S is sample and RM is the reference material. Positive 5 values mean that the sample contains more heavier isotope than the reference material, and negative 5 values that it contains less.19 The analytical precision of the determination of carbon and nitrogen contents was estimated to be ±6%, and for isotopic measurements ±0.2%0 for 513C and 515Ntot, and ±0.4%0 for 534S, based on replicate measurements of reference materials and samples. The following reference materials were used: NBS 23, NBS-127, IAEA-S-1, IAEA-S-2, IAEA S-3, IAEA-S-4, IAEA-SO-5 for sulphur, IAEA-CH-7, USG-24 for carbon and IAEA-N-1, IAEA-N-2 for nitrogen isotope measurements. 3. Results and Discussion 3. 1. Elemental Composition The total sulphur, nitrogen and total organic carbon contents of the samples analysed (Tab. 1) varied among different coal seams, as well as among different lithotypes inside particular coal seams. With regard to total S, two samples from the Velenje lignite coal seam were characterized as high-sulphur coal with more than 3 wt% total sulphur (fine detritial lignite and gelified fine detritial lignite), and three as medium-sulphur coal, with 1-3 wt% total sulphur (xylite, fusite and gelified fine detritial lignite taken ahead of a working face). All coal samples from the Kanižarica lignite coal seam, as well as coal samples from the Trbovlje-Hrastnik and the Senovo coal seams were characterized as medium-sulphur coals. Acid volatile sulphides (AVS) were below the detection limit. In all samples, with the exception of the sample from the Trbovlje-Hrastnik coal seam and Kanižarica 3 (organic-rich clay), organic sulphur predominated (Tab. 1, Fig. 5). Considering the fact that Fe2+ has a higher affinity to H2S than organic material,2,9 the higher organic sulphur contents compared to CRS could be explained by a lack of reactive iron. This, however, was not the case in the Tr-bovlje-Hrastnik coal seam and Kanižarica 3, where CRS was more abundant than organic sulphur. In the samples analysed, sulphate sulphur was mostly present in very low abundances (<0.1 wt%), with the exception of two samples from Kanižarica with 0.7 and 1.0 wt%, where the first sample was coal in alteration with clay and the second sample not coal but organic-rich clay. Since the isotopic composition of sulphate S was lower compared to sulphides (see Section 3.2), the presence of sulphate sulphur could be explained by the reoxidation of sulphides, where lighter isotopes are preferentially oxidized, leaving the remaining sulphide enriched with the heavier isotope. Total nitrogen and total organic carbon in samples from the Velenje lignite coal seam varied from 0.4 to 1.8 wt% and from 42 to 59 wt%, respectively. Microbial degradation of organic matter during early diagenesis affects the variation in C/N ratio,20,21 where higher micro-bial activity results in lower C/N ratio.22 This phenomenon was also observed in Velenje lignite samples (Tab. 1), with the C/N ratio decreasing from xylite, with a C/N Table 1: Elemental composition and C/N weight ratios of the coals studied. Coal minefv Macroscopic Elemental composition (wt%; dry basis) description CORG ntot STOT SCRS SORG SSULPH C/N Velenje 1 Xylite 50 0.4 2.7 0.1 2.6 0.1 125 Velenje 2 Fusite 59 0.6 1.4 0.3 1.2