Determination of Halogens (F, Cl, Br, I), Sulfur and Water in Seventeen Geological Reference Materials

Fluorine, chlorine, bromine, iodine and sulfur were determined in seventeen geological reference materials after extraction by pyrohydrolysis. Fluorine, Cl and S (as sulfate ions) were determined in the extraction solution by ion chromatography with detection limits of around 0.2 mg l⁻¹. Bromine and I were measured by ICP-MS with detection limits of 1 μg l⁻¹ for Br and 0.1 μg l⁻¹ for I. For rock samples, using normal extraction conditions (500 mg of sample and 100 ml of final solution) detection limits were 40 mg kg⁻¹ for F and Cl, 15 mg kg⁻¹ for S, 0.2 mg kg⁻¹ for Br and 0.02 mg kg⁻¹ for I. These detection limits may be improved by increasing the amount of sample and hence the concentration of the final solution. Water was also determined using an extraction technique based on H₂O degassing, reduction on zinc at 1000 °C and H₂ manometry. Our results for fluorine, chlorine, sulfur and water are in good agreement with literature data. Very few reference materials have recommended values for bromine and especially for iodine. Among the analysed samples, three are new reference materials: BHVO-2, BCR-2 and AGV-2.     

Flotation Separation of Trace Elements from Alkaline-Earth Matrices by Co(III) Hexamethylenedithiocarbamate before ICP-AES Determination

A method for the selective separation of Ag, Cd, Cr, Cu, Ni, Pb and Zn in traces from solutions of calcite (CaCO₃), dolomite (CaMg(CO₃)₂) and gypsum (CaSO₄.2H₂O) before their determination by inductively coupled plasma-atomic emission spectrometry (ICP-AES) is presented. The expected interferences of Ca and Mg on intensities of trace analytes were removed by collecting the elements of interest with cobalt(III) hexamethylenedithiocar-bamate, Co(HMDTC)₃. The flotation of aqueous solutions (1 l) of calcite, dolomite and gypsum was performed at pH 6.0, by 1.5 mg l⁻¹ Co and 0.6 mmol l⁻¹ HMDTC⁻. To minimise the effect of the reaction between Ca/Mg, which restrains the function of the surfactant, careful selection of the most suitable foaming reagent was necessary. The accuracy of the method was established by analysing natural alkaline-earth minerals by the standard addition method as well as using the dolomite reference materials GBW 07114 and GSJ JDo-1. The ICP-AES limits of detection following flotation on different minerals were found to be 0.080 μg g⁻¹ for Cd, 0.105 μg g⁻¹ for Ag, 0.142 μg g⁻¹ for Cu, 0.195 μg g⁻¹ for Cr, 0.212 μg g⁻¹ for Ni, 0.235 μg g⁻¹ for Zn and 0.450 μg g⁻¹ for Pb.     

Determination of Trace Elements in Calcite Using Solution and Laser Ablation ICP-MS: Calibration to NIST SRM Glass and USGS MACS Carbonate, and Application to Real Landfill Calcite

We report new data on the trace element concentrations of Mg, Cr, Mn, Co, Ni, Cu, Zn, Sr, Cd, Ba, La, Ce, Nd, Pb and U in USGS carbonate reference materials (MACS-1 and MACS-2) and compare solution ICP-MS and LA-ICP-MS trace element determinations on landfill calcites using calibration to different reference materials (MACS-1 and MACS-2 carbonate and NIST SRM 612 glass). Very good agreement (differences below 10% relative) was found between laser ablation and solution ICP-MS data for MACS-1 with higher concentrations of trace elements (values between 100 and 150 μg g⁻¹), with the exception of Cu and Zn. Similarly good agreement was found for MACS-2 with lower trace element concentrations (units to tens of μg g⁻¹), with the exception of Cr, Co and Zn. The MACS-1 reference material for calibration of LA-ICP-MS was found to be extremely useful for in situ determination of trace elements in real-world carbonate samples (landfill calcites), especially those present in calcite in higher concentrations (Mn, Sr, Ba; < 5% RSD). Less accurate determinations were generally obtained for trace elements present at low concentrations (∼ units of μg g⁻¹). In addition, good agreement was observed between the instrument calibration to MACS and NIST SRM 612 glass for in situ measurements of trace elements in landfill calcites K-2, K-3 and K-4 (differences below 15% relative for most elements). Thus, the application of MACS carbonate reference materials is promising and points to the need for the development of new carbonate reference materials for laser ablation ICP-MS.     

Concentrations of Trace Elements in Carbonate Reference Materials Coral JCp-1 and Giant Clam JCt-1 by Inductively Coupled Plasma-Mass Spectrometry

Trace elements in the Geological Survey of Japan carbonate reference materials Coral JCp-1 and Giant Clam JCt-1 were determined by inductively coupled plasma-mass spectrometry after digestion with 2% v/v HNO₃. A standard addition method was adopted in this determination in order to neutralise the Ca matrix effect. In addition, Sc, Y, In and Bi were used as internal standards to control the matrix effect and correct instrumental drift. Of the eighteen elements measured in JCp-1, precisions for fourteen elements, including Cu, Cd and Ba, were better than 10% RSD and concentrations ranged from 0.002 μg g^-1 (Cs) to 8.02 μg g^-1 (Ba). The concentrations of measured trace elements in JCt-1, except for Cu, were lower than those in JCp-1. Precisions for all elements with concentrations higher than 0.04 μg g^-1 in JCt-1 were also better than 10% RSD and concentrations were found to be between 0.001 μg g^-1 (Cs) and 4.84 μg g^-1 (Ba). The concentrations of more than fifteen trace elements in the aragonite reference materials are reported here for the first time. Both reference materials are suitable for use in geochemical studies of environmental reconstruction based upon biogenic carbonate materials.     

Determination of Selenium Concentrations in NIST SRM 610, 612, 614 and Geological Glass Reference Materials Using the Electron Probe, LA-ICP-MS and SHRIMP II

A combination of EMPA, sensitive high resolution ion microprobe (SHRIMP II) and/or LA-ICP-MS techniques was used to measure the concentration of selenium (Se) in NIST SRM 610, 612, 614 and a range of reference materials. Our new compiled value for the concentration of Se in NIST SRM 610 is 112 ± 2 μg g⁻¹. The concentration of Se in NIST SRM 612, using NIST SRM 610 for calibration, determined using LA-ICP-MS (confirmed using SHRIMP II) was 15.2 ± 0.2 μg g⁻¹. The concentration of Se in NIST SRM 614, using LA-ICP-MS was 0.394 ± 0.012 μg g⁻¹. LA-ICP-MS determination of Se in synthetic geological glasses BCR-2G, BIR-1G, TB-1G and the MPI-DING glasses showed a range in concentrations from 0.062 to 0.168 μg g⁻¹. Selenium in the natural glass, VG2, was 0.204 ± 0.028 μg g⁻¹.     

Preparation and Certification of Re-Os Dating Reference Materials: Molybdenites HLP and JDC

Two Re-Os dating reference material molybdenites were prepared. Molybdenite JDC and molybdenite HLP are from a carbonate vein-type molybdenum-(lead)-uranium deposit in the Jinduicheng-Huanglongpu area of Shaanxi province, China. The samples proved to be homogeneous, based on the coefficient of variation of analytical results and an analysis of variance test. The sampling weight was 0.1 g for JDC and 0.025 g for HLP. An isotope dilution method was used for the determination of Re and Os. Sample decomposition and pre-concentration of Re and Os prior to measurement were accomplished using a variety of methods: acid digestion, alkali fusion, ion exchange and solvent extraction. Negative thermal ionisation mass spectrometry and inductively coupled plasma-mass spectrometry were used for the determination of Re and ¹⁸⁷Os concentration and isotope ratios. The certified values include the contents of Re and Os and the model ages. For HLP, the Re content was 283.8 ± 6.2 μg g⁻¹, ¹⁸⁷Os was 659 ± 14 ng g⁻¹ and the Re-Os model age was 221.4 ± 5.6 Ma. For JDC, the Re content was 17.39 ± 0.32 μg g⁻¹, ¹⁸⁷Os was 25.46 ± 0.60 ng g⁻¹ and the Re-Os model age was 139.6 ± 3.8 Ma. Uncertainties for both certified reference materials are stated at the 95% level of confidence. Three laboratories (from three countries: PR. China, USA, Sweden) joined in the certification programme. These certified reference materials are primarily useful for Re-Os dating of molybdenite, sulfides, black shale, etc.     

Total Sulfur Concentration in Geological Reference Materials by Elemental Infrared Analyser

Total sulfur is an analyte for which there are few determinations published, despite the fact that it is a very important element (e.g., a major element in most ores, an important gas constituent in global warming, an active participant in acid drainage). Most geological reference materials have very poor quality sulfur results, that is with relative standard deviations (RSD) in the range of 30–50%, even for concentrations over 100 μg g⁻¹ S, which compromises their use as calibrators. In order to provide modern results with low RSD, sulfur was determined in twenty-nine geological reference materials with a state-of-the-art elemental S/C analyser using metal chips (certified reference materials with a traceability link) and analytical grade sulfur for high concentration samples. Analytical parameters (sample mass, crucible degassing, calibration strategy, etc.) were optimised by testing. Our results agreed with reference material values provided by issuing bodies. Results for CCRMP SY-2 (129 ± 13 μg g⁻¹ S), which has been proposed as a sulfur reference material, were in agreement with the proposed modern value of 122 ± 3.7 μg g⁻¹ S.     

Speciation Analysis for Iodine in Groundwater Using High Performance Liquid Chromatography-Inductively Coupled Plasma-Mass Spectrometry (HPLC-ICP-MS)

Methods based on high performance liquid chromatography coupled with inductively coupled plasma-mass spectrometry (HPLC-ICP-MS) for iodine speciation analysis are described. Experiments were performed relating to iodine species, preservative medium and time. The concentration and pH of the buffer are also discussed in relation to separation efficiency. The developed methods allowed the fast and sensitive determination of iodine species with detection limits of 0.025 μg l⁻¹ (as 1) for both iodide and iodate with < 5% relative standard deviation (RSD) for 50 nmol l⁻¹. Attempts were made to quantify total organo-iodine by size-exclusion chromatography (SEC). By way of example, a number of groundwater samples were analysed by these methods, revealing that iodide is the main iodine species in the sampled waters, but that high concentrations of organo-iodine compounds were observed in some samples as well. In summary, the total concentration and the dominant species of iodine in aquatic freshwater environments is easily and accurately measured using this new method.     

An Ion Chromatography Method for the Determination of Major Anions In Geothermal Water Samples

A single-column suppressed ion chromatography technique was employed for the simultaneous determination of major and trace anions in sulfaterich groundwater samples. An analytical column, a self regenerating suppressor and sodium carbonate as the eluent were used to separate the anions. Method detection limits for the anions of interest were 10.4, 15.9, 36.8, 62, 60, 61 and 67 μg l⁻¹ for F⁻, Cl⁻, NO₂⁻, Br⁻, NO₃⁻, PO₄³⁻ and SO₄²⁻ respectively. The precision of the method was tested at five different concentration levels for each anion reference sample to evaluate the effectiveness of the method for groundwater analysis. Recovery studies were performed between two successive months by adding reference samples to the geothermal groundwater and drinking water samples. Precision was also assessed as the relative standard deviation of both repeatability (within-day) and reproducibility (between-day and different concentrations) for groundwater samples. Standard deviation and RSD values of 220 groundwater samples acquired over 8 months were evaluated. The suppressed ion chromatography technique was found to be a suitable method for determining major anions in sulfate-rich geothermal water samples.     

Three Secondary Reference Materials for Lithium Isotope Measurements: Li7-N, Li6-N and LiCl-N Solutions

The CRPG (Nancy, France) has prepared secondary reference materials for Li isotope measurements by mixing ⁷Li or ⁶Li spikes and either L-SVEC or IRMM-016 certified reference materials to produce solutions having a known Li concentration and isotopic composition. The Li7-N and Li6-N solution samples (1.5 mol l⁻¹ HNO₃) have nominal δ⁷Li isotopic compositions of 30.1‰ and -9.7‰ respectively relative to L-SVEC and concentrations of 100 mg l⁻¹. Repeated measurement of these samples using the QUAD-ICP-MS at the CRPG yielded δ⁷Li of 30.4 ± 1.1‰ (n = 13) and -8.9 ± 0.9‰ (n = 9) at the 2s level of confidence. An additional LiCl-N solution was measured and yielded a delta value of 9.5 ± 0.6‰ (n = 3). Identical results were obtained at the BRGM (Orléans, France) from determinations performed with a Neptune MC-ICP-MS (30.2 ± 0.3‰, n = 89 for the Li7-N, -8.0 ± 0.3‰, n = 38 for the Li6-N and 10.1 ± 0.2‰, n = 46 for LiCl-N at the 2s level of confidence). The deviation of measured composition relative to the nominal value for the Li6-N solution might be explained by either contamination during preparation or an error during sample weighing. These secondary reference materials, previously passed through ion exchange resin or directly analysed, may be used for checking the accuracy of Li isotopic measurements over a range of almost 40‰ and will be available to the scientific community upon request to J. Carignan or N. Vigier, CRPG.     

Trace Element Analysis of Fused Whole-Rock Glasses by Laser Ablation-ICP-MS and PIXE

The concentrations of fifty trace elements, including relatively volatile elements and transition metal elements, in fused glasses of Geological Survey of Japan rock reference materials GSJ JR-2, JA-1, JA-2, JB-1a, JB-3, JGb-1 and JF-1 were determined by particle (proton) induced X-ray emission (PIXE) and laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS). The fused glasses were prepared by rapid fusion and subsequent quenching in welded platinum capsules and were found to be homogeneous for major elements and for trace elements with concentrations of more than 1 μg g^-1 within the observed precision (± 10% mean) on a 70 μm sampling scale. The values obtained by PIXE and LA-ICP-MS for the transition elements (Cr, Mn, Fe, Ni and Cu), the relatively volatile elements (Zn, Ga, Rb and Pb) and the refractory elements (Y, Zr, Nb and Th) with concentrations greater than a few μg g^-1 showed good agreement (within 10 % relative difference). The values for almost all the elements detected at concentrations higher than 1 μg g^-1 as determined by LA-ICP-MS also agreed well with the reference values (mean relative difference < ± 10%), except for B and Cu. The good agreement confirmed the appropriateness of the NIST SRM 600 series glass calibration reference material for LA-ICP-MS analysis of glasses with variable major-element compositions for almost all elements. The concentrations of Cu in all the samples were lower than the reference values, which was attributed to adsorption of the transition metals onto the platinum capsule during preparation.     

Development of a Field Preconcentration/Elution Unit for Routine Determination of Dissolved Metal Concentrations by ICP-OES in Marine Waters: Application for Monitoring of the New Caledonia Lagoon

An iminodiacetate chelating resin was optimised for the rapid determination of Co, Cu, Fe, Mn and Ni in seawater. Using inexpensive, high-capacity, reusable cartridges allowed high flow rates of up to 25 ml min⁻¹. High preconcentration factors, of up to 500, were obtained in order to analyse samples using an ICP-OES. The requirement for a buffer was eliminated due to the high tolerance of the ICP-OES to interfering matrix elements, thereby further reducing the potential for contamination. Quantification limits in seawater were: Co = 6 ng l⁻¹, Cu = 8 ng l⁻¹, Fe = 6 ng l⁻¹, Mn = 5 ng l⁻¹ and Ni = 6 ng l⁻¹. The method was verified by the analysis of near shore seawater (CASS-4) and open ocean seawater (NASS-5) reference materials. In order to satisfy the high sampling demands using the iminodiacetate cartridges, a portable off-line preconcentration unit was developed for routine analysis. The multi-channel preconcentration unit, was capable of treating up to eight samples simultaneously with concentrating times as little as 30 minutes. The technique was also used to determine dissolved metals in fresh and interstitial waters. The technique has been successfully used in a number of environmental studies and impact assessments to evaluate the effects of mining on the New Caledonian lagoon.     

Trace Element Determination of Single Fluid Inclusions in Quartz by Laser Ablation ICP-MS

Single fluid inclusions in quartz from a Pb-Zn-Ag carbonate replacement deposit were selected for trace element determination by laser ablation ICP-MS. Spikes in element intensities were noted between first breached fluids versus subsequent analyses, suggesting that accurate element concentrations may not be determined in smaller fluid inclusions when only one analysis is obtained before the fluid is exhausted. Elemental concentrations in the fluid inclusions were determined by external standardisation using solutions sealed in microcapillary tubes. Standards and single natural inclusion analyses give repeatabilities (%RSD) of ˜ 20% for Rb and Sr. Rubidium and strontium concentrations range from 0.56-5.07 μg ml^-1 and 1.12-27.4 μg ml^-1, respectively, whereas Zn and Ag are below detection limits (< 10 ng ml^-1). The results suggest that nearly all Zn and Ag are removed by the time hydrothermal fluids precipitate gangue minerals.     

Analysis of Re, Au, Pd, Pt and Rh in NIST Glass Certified Reference Materials and Natural Basalt Glasses by Laser Ablation ICP-MS

A new technique for the in situ analysis of Re, Au, Pd, Pt and Rh in natural basalt glass by laser ablation (LA)-ICP-MS is described. The method involves external calibration against NIST SRM 612/613 or 614/615 glass certified reference materials, internal standardisation using Ca, and ablation with a 200 μm wide beam spot and a pulsed laser repetition rate of 50 Hz. Under these conditions, sensitivities for Re, Au, Pd, Pt and Rh analyte ions are ˜ 5000 to 100,000 cps/μg g^-1. This is sufficient to make measurements precise to ˜ 10% at the 2-10 μg g^-1 level, which is well within the range of concentrations expected in many basalts. For LA-ICP-MS calibration and a demonstration of the accuracy of the technique, concentrations of Re, Au, Pd, Pt and Rh in the NIST SRM 610/611 (˜ 1 to 50 μg g^-1), 612/613 (˜ 1 to 7 μg g^-1), 614/615 (˜ 0.2 to 2 μg g^-1) and 616/617 (˜ 0.004 to 2 μg g^-1) glasses were determined by solution-nebulisation (SN)-ICP-MS. Using the 612/613 or 614/615 glasses as calibration standards, LA-ICP-MS measurements of these elements in the other NIST glasses fell within ˜ 15% of those determined by SN-ICP-MS. Replicate LA-ICP-MS analyses of the 612/613 and 614/615 glasses indicate that, apart from certain anomalous domains, the glasses are homogeneous for Re, Au, Pd, Pt and Rh to better than 3.5%. Two LA-ICP-MS analyses of natural, island-arc basalt glasses exhibit large fractionations of Re, Au and Pd relative to Pt and Rh, compared to the relative abundances in the primitive mantle.     

High Sensitivity Analysis of Trace Element-Poor Geological Reference Glasses by Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry (LA-ICP-MS)

Fifty-two trace elements in NIST SRM 614, 616 and MPI-DING BM90/21-G glass reference materials as well as in NIST SRM 612, USGS BCR2-G and other MPI-DING reference glasses (KL2-G, GOR132-G, GOR128-G, ATHO-G, Tl-G, StHs6/80-G and ML3B-G) were determined by laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS). Accurate ultra-low trace element abundances in the NIST SRM 614, 616 and BM90/21-G reference glasses down to lower ng g⁻¹ levels were determined with relative standard deviations (RSD) of less than 10%. Limits of detection using He as carrier gas were up to two times lower than with Ar and were 0.004 to 0.12 μg g⁻¹ for elements of lower mass numbers (amu < 85) and 0.002 to 0.06 μg g⁻¹ for elements having amu < 85. The measured concentrations generally agree within 15% with previous studies except for B in NIST SRM 614 and 616, which appears to be heterogeneously distributed, and Co, Zn, Ga and Ag in NIST SRM 616 for which the existing data set is too small to evaluate the discrepancies. New values for As (0.593 μg g⁻¹), Ag (0.361 μg g⁻¹) and Cd (0.566 μg g⁻¹) in NIST SRM 614 and new values for Na (94864 μg g⁻¹) and As (0.276 μg g⁻¹) in NIST SRM 616 are reported.     

Trace Element Analysis of NIST SRM 614 and 616 Glass Reference Materials by Laser Ablation Microprobe-Inductively Coupled Plasma-Mass Spectrometry

Fifty elements in NIST SRM 614 and 616 glass reference materials were determined by laser ablation microprobe-inductively coupled plasma-mass spectrometry (LAM-ICP-MS). The values determined for NIST SRM 614 agreed well with the NIST-certified and information values (mean relative difference ± 3.6%), except for B, Sc and Sb. The values determined for NIST SRM 616 agreed with the NIST-certified and information values within a mean relative difference of ± 1.5%, except for B, Sc and Ga. In addition, at an 80 μm sampling scale, NIST SRM 614 and 616 glass discs were homogeneous for trace elements within the observed precisions of 5 and 15% (mean), respectively. Detection limits were in the range 0.01 - 0.3 μg g⁻¹ for elements of lower mass numbers (amu < 80) and 1 - 10 ng g⁻¹ for heavy elements (amu > 80). Detection at the sub ng g⁻¹ level is possible for most of the heavy elements by using an ablation pit size larger than 10 0 μm.     

Trace Element Data for Gold, Iridium and Silver in Seventy Geochemical Reference Materials

New concentrations for Au, Ir and Ag obtained by instrumental neutron activation analysis are presented for seventy geochemical reference materials. Results in agreement with literature values for Au and Ir down to concentrations of a few ng g⁻¹ were obtained. For Au and Ir concentrations above 10 ng g⁻¹, the repeatability of replicate analyses of reference materials was mostly better than 10%. For concentrations between 1 and 10 ng g⁻¹ the RSD for Ir was 10–30%, whereas for Au it was higher and more variable (20–50%). In addition, concentrations for Cd and Hg are presented for some of the same reference materials. The high RSD at relatively high concentrations seen in gold for some RMs (e.g., WMG-1, WMS-1) did not exist for Ir and suggests homogeneity for this platinum-group element at the sub-sample size used in this study. For the following eight RMs, mostly ultramafic rocks (CHR-Pt+, OREAS-13P, OREAS-14P, PCC-1, UMT-1, WMG-1, WMS-1, WPR-1), Ir measurements agreed within ± 10% of mostly certified or recommended concentrations, which ranged from 2 ng g⁻¹ to 6 μg g⁻¹. For the reference material UB-N, iridium concentration compared favourably to published results obtained by isotope dilution ICP-MS methods and a previously unrecognised heterogeneity is inferred for Au, Hg and Sb, but not for the other measured elements.     

Sampling, Sample Preparation and ICP-MS Chemical Analysis of Lake Baikal Sponges

Procedures for sampling, sample preparation and ICP-MS analysis of endemic sponges from Lake Baikal have been developed. Sample decomposition was carried out using an open acid decomposition with ultrasound treatment. The distribution of nineteen elements (Mg, Al, P, Ca, Ti, Mn, Co, Ni, Cu, Rb, Sr, Y, Cd, Ba, La, Ce, Pb, Th and U) in different parts of a sponge's body (outer and inner layers and layers adjacent to the substratum) was studied. Detection limits were determined; these ranged from 0.013 to 4.12 μg g^-1 for trace elements and from 23 to 130 μg g^-1 for biogenic elements. The degree of elemental uptake by living substances is discussed with regard to the environment.     

Determination of Chlorine in Nine Rock Reference Materials by Isotope Dilution Mass Spectrometry

A procedure for the determination of chlorine by the isotope dilution technique (ID) using negative thermal ionisation mass spectrometry (N-TIMS) is described. Silicate samples of about 10 mg were spiked and decomposed with hydrofluoric acid, and chlorine was isolated by precipitation of silver chloride after neutralisation with Ca(OH)₂. The ammonical solution of AgCl was then subjected to N-TIMS. Replicate analyses of rock reference materials, typically of JB-1 and JR-1, demonstrated the high quality of the analyses (precision for Cl was ± 1-2%). We present here the most precise data sets of chlorine concentrations in nine igneous rock reference materials, three basalts (JB-1, JB-2, JB-3), two andesites (JA-3, AGV-1), two rhyolites (JR-1, JR-2) and two granodiorites (JG-3, GSP-1). The chlorine concentrations found ranged from 152 μg g-1 in AGV-1 to 1008 μg g^-1 in JR-1. Our results presented here are partly (but not completely) in agreement with recommended values, where they are available. The N-TIMS ID technique can thus be used as a means of determining low chlorine contents in silicate materials to high precision.     

DETERMINATION OF SULPHUR AT LOW LEVELS IN STANDARD REFERENCE MATERIALS BY PYROHYDROL YSIS/LON CHROMATOGRAPHY

The determination of sulphur in rocks, soils and sediments by pyrohydrolysis/ion chromatography has been modified to reach lower detection limits of the order of 2 μg g⁻¹. Thirty-one standard reference materials containing sulphur at levels ranging from 3 to about 900 μg g⁻¹ have been analysed in triplicate. In general, the comparison of the results with those published is favorable. The method is also applicable to organic-rich samples as demonstrated by the analysis of NBS samples (citrus leaves, tomato leaves, bovine liver).