Delta34S measurements of sulfur by multicollector inductively coupled plasma mass spectrometry

An accurate and precise method for the determination of delta34S measurements by multicollector inductively coupled plasma mass spectrometry has been developed. Full uncertainty budgets, taking into consideration all the uncertainties of the measurement process, have been calculated. The technique was evaluated by comparing measured values with a range of isotopically enriched sulfur solutions prepared by gravimetric addition of a 34S spike. The gravimetric and measured results exhibited a correlation of R2 >0.999. Repeat measurements were also made after adding Na (up to 420 microg g(-1)) and Ca (up to 400 microg g(-1)) salts to the sulfur standard. No significant deviations in the delta34S values were observed. The Russell correction expression (Ingle, C.; Sharp, B.; Horstwood, M.; Parrish, R.; Lewis, D. J. J. Anal. At. Spectrom. 2003, 18, 219) was used to correct for mass bias on the 34S/32S isotope amount ratio from the mass bias observed for the 30Si/28Si isotope amount ratio. Consistent compensation for instrumental mass bias was achieved. Resolution of the measured delta34S values was better than 1 per thousand after consideration of all uncertainty components. The technique was evaluated for practical applications by measurement of delta34S for a range of mineral waters by pneumatic nebulization sample introduction and the analysis of genuine and counterfeit pharmaceuticals using both laser ablation sample introduction and liquid chromatography. For the former two cases polyatomic interferences were resolved by operating the MC-ICPMS in medium resolution, while for the chromatographic analyses polyatomic interferences were minimized by the use of a membrane desolvator, allowing the instrument to be operated at a resolution of 400.     

Determination of the total oxygen isotopic composition of nitrate and the calibration of a delta 17O nitrate reference material

A thermal decomposition method was developed and tested for the simultaneous determination of delta 18O and delta 17O in nitrate. The thermal decomposition of AgNO3 allows for the rapid and accurate determination of 18O/ 16O and 17O/16O isotopic ratios with a precision of +/- 1.5 per thousand for delta 18O and +/- 0.11 per thousand for delta 17O (delta 17O = delta 17O - 0.52 x delta 18O). The international nitrate isotope reference material IAEA-NO3 yielded a delta 18O value of +23.6 per thousand and delta 17O of -0.2 per thousand, consistent with normal terrestrial mass-dependent isotopic ratios. In contrast, a large sample of NaNO3 from the Atacama Desert, Chile, was found to have delta 17O = 21.56 +/- 0.11 per thousand and delta 18O = 54.9 +/- 1.5 per thousand, demonstrating a substantial mass-independent isotopic composition consistent with the proposed atmospheric origin of the desert nitrate. It is suggested that this sample (designated USGS-35) can be used to generate other gases (CO2, CO, N2O, O2) with the same delta 17O to serve as measurement references for a variety of applications involving mass-independent isotopic compositions in environmental studies.     

Measurement of sulfur isotope compositions by tunable laser spectroscopy of SO2

Sulfur isotope measurements offer comprehensive information on the origin and history of natural materials. Tunable laser spectroscopy is a powerful analytical technique for isotope analysis that has proven itself readily adaptable for in situ terrestrial and planetary measurements. Measurements of delta(34)S in SO2 were made using tunable laser spectroscopy of combusted gas samples from six sulfur-bearing solids with delta(34)S ranging from -34 to +22 per thousand (also measured with mass spectrometry). Standard deviation between laser and mass spectrometer measurements was 3.7 per thousand for sample sizes of 200 +/- 75 nmol SO(2). Although SO(2)(g) decreased 9% over 15 min upon entrainment in the analysis cell from wall uptake, observed fractionation was insignificant (+0.2 +/- 0.6 per thousand). We also describe a strong, distinct (33)SO(2) rovibrational transition in the same spectral region, which may enable simultaneous delta(34)S and Delta(33)S measurements.     

Comparison of delta18O measurements in nitrate by different combustion techniques

Three different KNO3 salts with delta18O values ranging from about -31 to +54 per thousand relative to VSMOW were used to compare three off-line, sealed glass tube combustion methods (widely used for isotope studies) with a more recently developed on-line carbon combustion technique. All methods yielded roughly similar isotope ratios for KNO3 samples with delta18O values in the midpoint of the delta18O scale near that of the nitrate reference material IAEA-NO-3 (around +21 to +25 per thousand). This reference material has been used previously for one-point interlaboratory and intertechnique calibrations. However, the isotope ratio scale factors by all of the off-line combustion techniques are compressed such that they are between 0.3 and 0.7 times that of the on-line combustion technique. The contraction of the 6180 scale in the off-line preparations apparently is caused by O isotope exchange between the sample and the glass combustion tubes. These results reinforce the need for nitrate reference materials with delta18O values far from that of atmospheric O2, to improve interlaboratory comparability.     

Determination of the S isotope composition of methanesulfonic acid

Sulfur (S) isotopes have been used to apportion the amount of biogenic and anthropogenic sulfate in remote environments, an important parameter that is used to model the global radiation budget. A key assumption in the apportionment calculations is that there is little isotope selectivity as reduced compounds such as dimethyl sulfide (DMS) are oxidized. This paper describes a method to determine, for the first time, the S isotope composition of methanesulfonic acid (MSA), the product of DMS oxidation. The isotope composition of MSA was measured directly by EA-IRMS and was used as an isotope reference for the method. Synthetic mixtures approximating the conditions expected for aerosol MSA samples were prepared to test this method. First, MSA solutions were measured alone and then in combination with MSA and SO4(2-). In synthetic mixtures, SO4(2-) was separated from MSA by precipitating it as BaSO4 prior to preparation of MSA for isotope analysis. The delta 34S value for MSA solutions was -2.6 per thousand (SD +/- 0.4 per thousand), which is not different from the delta 34S obtained from MSA filtrate after precipitating SO4(2-) from the mixture (-2.7 +/- 0.3 per thousand). However, these values are offset from direct EA-IRMS analysis of MSA used as the isotope reference by -1.1 +/- 0.2 per thousand, and this must be accounted for in reporting MSA measurements. The S isotope measurements using this method approach a limiting value above 300 microg of MSA. This is approximately equal to the MSA found in 20,000 m3 of air, assuming ambient concentrations of approximately 15 ng m(-3). Three samples of MSA from the Pacific Ocean measured using this technique have an average delta 34S value of +17.4 +/- 0.7 per thousand.     

Evaluation of gas chromatographic isotope fractionation and process contamination by carbon in compound-specific radiocarbon analysis

The relevance of both modern and fossil carbon contamination as well as isotope fractionation during preparative gas chromatography for compound-specific radiocarbon analysis (CSRA) was evaluated. Two independent laboratories investigated the influence of modern carbon contamination in the sample cleanup procedure and preparative capillary gas chromatography (pcGC) of a radiocarbon-dead 3,3',4,4',5,5'-hexachlorobiphenyl (PCB 169) reference. The isolated samples were analyzed for their 14C/12C ratio by accelerator mass spectrometry. Sample Delta14C values of -996 +/- 20 and -985 +/- 20 per thousand agreed with a Delta14C of -995 +/- 20 per thousand for the unprocessed PCB 169, suggesting that no significant contamination by nonfossil carbon was introduced during the sample preparation process at either laboratory. A reference compound containing a modern 14C/12C ratio (vanillin) was employed to evaluate process contamination from fossil C. No negative bias due to fossil C was observed (sample Delta14C value of 165 +/- 20 per thousand agreed with Delta14C of 155 +/- 12 per thousand for the unprocessed vanillin). The extent of isotopic fractionation that can be induced during pcGC was evaluated by partially collecting the vanillin model compound of modern 14C/12C abundance. A significant change in the delta13C and delta14C values was observed when only parts of the eluting peak were collected (delta13C values ranged from -15.75 to -49.91 per thousand and delta14C values from -82.4 to +4.71 per thousand). Delta14C values, which are normalized to a delta13C of -25 per thousand, did not deviate significantly (-58.9 to -5.8 per thousand, considering the uncertainty of approximately +/-20 per thousand). This means that normalization of radiocarbon results to a delta13C of -25 per thousand, normally performed to remove effects of environmental isotope fractionation on 14C-based age determinations, also cor-rects sufficiently for putative isotopic fractionation that may occur during pcGC isolation of individual compounds for CSRA.     

IR laser extraction technique applied to oxygen isotope analysis of small biogenic silica samples

An IR-laser fluorination technique is reported here for analyzing the oxygen isotope composition (delta18O) of microscopic biogenic silica grains (phytoliths and diatoms). Performed after a controlled isotopic exchanged (CIE) procedure, the laser fluorination technique that allows one to visually check the success of the fluorination reaction is faster than the conventional fluorination technique and allows analyzing delta18O of small to minute samples (1.6-0.3 mg) as required for high-resolution paleoenvironmental reconstructions. The long-term reproducibility achieved with the IR laser-heating fluorination/O2 delta18O analysis is lower than or equal to +/-0.26 per thousand (1 SD; n = 99) for phytoliths and +/-0.17 per thousand (1 SD; n = 47) for diatoms. When several CIE are taken into account in the SD calculation, the resulting reproducibility is lower than or equal to +/-0.51 per thousand for phytoliths (1 SD; n = 99; CIE > 5) and +/-0.54 per thousand (1 SD; n = 47; CIE = 13) for diatoms. A minimum reproducibility of +/-0.5 per thousand leads to an estimated uncertainty on delta18Osilica close to +/-0.5 per thousand. Resulting uncertainties on reconstructed temperature and delta18Oforming water are, respectively, +/-2 degrees C and +/-0.5 per thousand and fit in the precisions required for intertropical paleoenvironmental reconstructions. Several methodological points such as optimal extraction protocols and the necessity or not of performing two CIE prior to oxygen extraction are assessed.     

Isotopic analysis of N and o in nitrite and nitrate by sequential selective bacterial reduction to N2O

Nitrite is an important intermediate species in the biogeochemical cycling of nitrogen, but its role in natural aquatic systems is poorly understood. Isotopic data can be used to study the sources and transformations of NO2- in the environment, but methods for independent isotopic analyses of NO2- in the presence of other N species are still new and evolving. This study demonstrates that isotopic analyses of N and O in NO2- can be done by treating whole freshwater or saltwater samples with the denitrifying bacterium Stenotrophomonas nitritireducens, which selectively reduces NO2- to N2O for isotope ratio mass spectrometry. When calibrated with solutions containing NO2- with known isotopic compositions determined independently, reproducible delta15N and delta18O values were obtained at both natural-abundance levels (+/-0.2-0.5 per thousand for delta15N and +/-0.4-1.0 per thousand for delta18O) and moderately enriched 15N tracer levels (+/-20-50 per thousand for delta15N near 5000 per thousand) for 5-20 nmol of NO2- (1-20 micromol/L in 1-5 mL aliquots). This method is highly selective for NO2- and was used for mixed samples containing both NO2- and NO3- with little or no measurable cross-contamination. In addition, mixed samples that were analyzed with S. nitritireducens were treated subsequently with Pseudomonas aureofaciens to reduce the NO3- in the absence of NO2-, providing isotopic analyses of NO2- and NO3- separately in the same aliquot. Sequential bacterial reduction methods like this one should be useful for a variety of isotopic studies aimed at understanding nitrogen cycling in aquatic environments. A test of these methods in an agricultural watershed in Indiana provides isotopic evidence for both nitrification and denitrification as sources of NO2- in a small stream.     

Sulfur and oxygen isotope analysis of sulfate at micromole levels using a pyrolysis technique in a continuous flow system

The discovery of a mass-independent isotopic composition (delta17O = (delta17O - 0.512 * delta18O) no equal to 0) in aerosol sulfate and the identification of its origin (aqueous-phase oxidation by 03 and H2O2) have renewed interest in measuring the oxygen isotopic content of sulfate. In this paper, we present a new method to measure both delta17O and delta18O in SO4, with the possibility of sulfur isotope analysis on the same sample. The technique takes advantage of the easy pyrolysis of Ag2SO4 to SO2, O2, and Ag metal in a continuous flow system. Because the technique is not quantitative in oxygen (yield approximately 45% for O2), a calibration is needed. Correction factors of +0.87 and +0.44% were obtained for delta18O and delta17O, respectively. A technique to convert micromole levels of sulfate in any form to silver sulfate is described. To reach this goal, a solid electrolyte (Nafion membrane) is used in an electrolysis apparatus. Reproducibilities for micromole sample sizes are (1sigma) 0.5, 0.3, and 0.1% for delta18O, delta17O, and delta17O, respectively. No memory effects or isotopic exchange during the treatment of the sample is observed. The main advantages of this new method over the existing ones are no fluorinating agent is needed, both oxygen and sulfur isotopes can be measured on the same sample, only very small amounts of sulfate are needed (down to 100 microg (1 micromol)), it is relatively fast and inexpensive, and the possibility exists to couple this technique to an on-line analysis.     

Stable carbon and oxygen isotopic analysis of atmospheric carbon monoxide using continuous-flow isotope ratio MS by isotope ratio monitoring of CO

We have developed a rapid and simple measurement system for both content and stable isotopic compositions (13C and 18O) of atmospheric CO, using continuous-flow isotope ratio mass spectrometry by simultaneously monitoring the CO+ ion currents at masses 28, 29, and 30. The analytical system consisted sequentially of a sample trapping port (liquid nitrogen temperature silica gel and molecular sieve 5A), a gas dryer, a CO purification column (molecular sieve 5A), a cryofocusing unit, and a final purification column using a GC capillary. Analytical precision of 0.2 per thousand for 13C and 0.4 per thousand for 18O can be realized for samples that contain as little as 300 pmol of CO within 40 min for one sample analysis. Analytical blanks associated with the method are less than 1 pmol. The extent of analytical error in delta13C due to mass-independent fractionation of oxygen in natural CO is estimated to be less than 0.3 per thousand. Based on this system, we report herein a kinetic isotopic effect during CO consumption in soil.     

A method for the extraction of ammonium from freshwaters for nitrogen isotope analysis

The measurement of delta15N values of inorganic nitrogen species is an important analytical tool to trace nitrogen species in order to understand nitrogen cycling in aquatic systems. Nitrogen isotope analysis of freshwater ammonium has, however, been hindered by the lack of a simple and reliable technique to measure delta15N values at natural abundance levels. We present a simple and rapid method to concentrate ammonium from freshwater samples for on-line N-isotope ratio determination. Ammonium is collected by adsorption on N-free cation exchange resins. The dried N-loaded exchange resin is then directly combusted to produce N2 gas for subsequent delta15N analysis. The method was evaluated with simulated freshwater solutions containing varying amounts of standard NH4+-N (delta15N = 2.1 per thousand) and potentially interfering inorganic and organic compounds. In general, the cation exchange resin method gives accurate and reproducible delta15N values (sigma1 < 0.3 per thousand; n = 10). Because of adsorption interference, high concentrations of cations in solution may cause ammonium loss but do not result in measurable isotope fractionation. Replicate extractions of the ammonium standard added to water collected from four Swiss lakes demonstrate the good performance of this method when applied to low ionic strength natural water samples with modest concentrations of dissolved organic nitrogen.     

Isotope (18)O/(16)O Ratio Measurements of Water Vapor by use of Photoacoustic Spectroscopy

We applied a photoacoustic spectroscopy technique to isotope ratio measurements of (16)O and (18)O in water-vapor samples, using a pulsed tunable dye laser pumped by a Nd:YAG laser. The fourth overtone bands (4nu(OH)) of water molecules near 720 nm were investigated. We identified the absorption lines of H(2)(16)O and H(2)(18)O in the photoacoustic spectra that we measured by using an (18)O-enriched water sample and the HITRAN database. We measured the difference in the (18)O/(16)O isotope ratios for normal distilled water and Antarctic ice, using the photoacoustic method. The value obtained for the difference between the two samples is delta(18)O = -32 ? 16 per thousand, where the indicated deviation was a 1varsigma value among 240-s measurements, whereas the value measured with a conventional isotope mass spectrometer was delta(18)O = -28 ? 2 per thousand. This method is demonstrated to have the potential of a transportable system for in situ and quick measurements of the H(2)(18)O/H(2)(16)O ratio in the environment.     

Measurements of the 17O excess in water with the equilibration method

The equilibration method is the present-day standard method for measuring delta18O in water samples. The mass-to-charge ratio of 45 is measured at the same time but generally not used for further analysis. We show that an improved equilibration method can be used for precise determination of delta17O in addition to that of delta18O, and therefore can estimate 17O excess values to a precision of better than 0.1 per thousand. To control the masking effect of the 14 times more abundant 13C on mass 45, we propose to use a chemical buffer in the water samples to keep the pH value and therefore the fractionation during the equilibration process of the 13C constant. With this improved method, the precision for the delta18O value could also be slightly improved from 0.05 to 0.03 per thousand. Furthermore, we discuss the influences of the amount of water, the temperature, the CO2 gas pressures, and changes in the pH during the measuring procedure on oxygen and carbon isotopes. We noticed that measured delta45 values are a good control for delta18O measurements. This study tries to fathom the possibilities and limitations of the equilibration method for measuring 17O excess values of water samples.     

Determination of total sulfur at microgram per gram levels in geological materials by oxidation of sulfur into sulfate with in situ generation of bromine using isotope dilution high-resolution ICPMS

We have developed a new, simple, and accurate method for the determination of total sulfur at microgram per gram levels in milligram-sized silicate materials with isotope dilution high-resolution inductively coupled plasma mass spectrometry equipped with a flow injection system. In this method, sulfur can be quantitatively oxidized by bromine into sulfate with achievement of isotope equilibrium between the sample and spike. Detection limits for 32S+ and 34S+ in the ideal solution and silicate samples were 1 and 6 ng mL(-1) and 0.07 and 0.3 microg g(-1), respectively. The total blank was 46 ng, so that a 40-mg silicate sample containing 10 mirog g(-1) sulfur can be measured with a blank correction of < 10%. This total blank can be lowered to 8 ng if a low-blank air system is used for evaporations. To evaluate the applicability of this method, we analyzed not only silicate reference materials with sulfur content of 5.25-489 microg g(-1) and sample sizes of 13-40 mg but also the Allende meteorite with a sulfur content of 2%. The reproducibility for various rock types was < 9%, even though blank corrections in some samples of low sulfur content were up to 24%. This method is suitable for analyzing geological samples as well environmental samples such as soils, sediments, and water samples.     

Isotopic variations of Zn in biological materials

Variations in the isotopic composition of Zn present in various biological materials were determined using high-resolution multicollector inductively coupled plasma mass spectrometry (MC-ICPMS), following digestion and purification by anion exchange chromatography. To correct for differences in instrumental mass discrimination effects between samples and standards, Cu was employed as an elemental spike. Complementary analyses of Zn separates by sector field ICPMS instruments revealed that the concentrations of the majority of potentially interfering elements were reduced to negligible levels. Residual spectral interferences resulting from (35)Cl(16)O(2)(+), (40)Ar(14)N(2)(+), and (40)Ar(14)N(16)O(+) could be instrumentally resolved from the (67)Zn, (68)Zn, and (70)Zn ion beams, respectively, during measurement by MC-ICPMS. The only other observed interference in the Cu and Zn mass range that could not be effectively eliminated by high-resolution multicollection resulted from (35)Cl(2)(+), necessitating modification of the sample preparation procedure to allow accurate (70)Zn detection. Complete duplication of the entire analytical procedure for human whole blood and hair, as well as bovine liver and muscle, provided an external reproducibility of 0.05-0.12 per thousand (2sigma) for measured delta(66/64)Zn, delta(67/64)Zn, and delta(68/64)Zn values, demonstrating the utility of the method for the precise isotopic analysis of Zn in biological materials. Relative to the selected Zn isotopic standard, delta(66/64)Zn values for biological samples varied from -0.60 per thousand in human hair to +0.56 per thousand in human whole blood, identifying the former material as the isotopically lightest Zn source found in nature to date.     

New guidelines for delta13C measurements

Consistency of delta13C measurements can be improved 39-47% by anchoring the delta13C scale with two isotopic reference materials differing substantially in 13C/12C. It is recommended that delta13C values of both organic and inorganic materials be measured and expressed relative to VPDB (Vienna Peedee belemnite) on a scale normalized by assigning consensus values of -46.6 per thousand to L-SVEC lithium carbonate and +1.95 per thousand to NBS 19 calcium carbonate. Uncertainties of other reference material values on this scale are improved by factors up to two or more, and the values of some have been notably shifted: the delta13C of NBS 22 oil is -30.03 per thousand.     

Sensitive measurement of NH4+ 15N/14N (delta 15NH4+) at natural abundance levels in fresh and saltwaters

We report a new method for determining the 15N/14N of NH4+ at natural abundance level in both freshwater and seawater. NH4+ is first quantitatively oxidized to NO2- by hypobromite (BrO-) at pH approximately 12. After the addition of sodium arsenite to consume excess BrO-, yield is verified by colorimetric NO2- determination. NO2- is further reduced to N2O using a 1:1 sodium azide and acetic acid buffer solution using previously established procedures. The product N2O is then analyzed for isotopic composition using a continuous flow purge and cryogenic trap system coupled to an isotope ratio mass spectrometer. Reliable delta 15N values (standard deviation is 0.3 per thousand or better) are obtained over an NH4+ concentration range of 0.5-10 microM using 20 mL volumes of either freshwater or seawater samples. Higher concentration samples are readily diluted to lower concentration. Preexisting NO2- is removed by treatment with sulfanilic acid. There is no interference from any of the nitrogen-containing compounds tested except short-chain aliphatic amino acids (i.e., glycine) which typically are present at very low environmental concentrations. As compared to published methods, our approach is more robust, readily applicable at low concentrations and small sample volumes, and requires less time for preparation and analysis.     

Determination of bromine stable isotopes using continuous-flow isotope ratio mass spectrometry

A new methodology for bromine stable isotope determination by continuous-flow isotope ratio mass spectrometry (CF-IRMS) was developed. The technique was tested on inorganic samples. Inorganic bromide was precipitated in the form of silver bromide by using silver nitrate in a standard methodology. Bromine stable isotope analysis was carried out on methyl bromide (CH3Br) after converting silver bromide to methyl bromide by reacting it with methyl iodide (CH3I). The system used in this study is an IsoPrime IRMS, with analytical capabilities of both dual-inlet and continuous-flow modes coupled with an Agilent 6890 GC equipped with a CTC Analytics CombiPAL autosampler. This new technique measures samples as small as 0.2 mg of AgBr (1 micromol of Br-). The bromine stable isotope analysis using continuous flow technology showed excellent precision and accuracy. The internal precision using pure methyl bromide gas is better than +/-0.03 per thousand (+/-SD); the external precision using seawater standard is better than +/-0.06 per thousand (+/-SD) for n = 12. Moreover, the sample analysis time is 16 min, as compared to 75 min needed in previous techniques. This allows for 50 samples to be analyzed in 1 day, as compared to 8 samples using the conventional techniques. A series of natural saline formation waters and brines from sedimentary and crystalline rock environments was measured by this new methodology to test the potential natural range of delta81Br. The bromine isotopic composition of the samples ranged from 0.00 to +1.80 per thousand relative to standard mean ocean bromide (SMOB). Initial trends and distinctive isotopic difference were noticed between crystalline shield brines and sedimentary formation brines.     

High-precision laser spectroscopy D/H and 18O/16O measurements of microliter natural water samples

Newly available gas analyzers based on off-axis integrated cavity output spectroscopy (OA-ICOS) lasers have been advocated as an alternative to conventional isotope-ratio mass spectrometers (IRMS) for the stable isotopic analysis of water samples. In the case of H2O, OA-ICOS is attractive because it has comparatively low capital and maintenance costs, the instrument is small and field laboratory portable, and provides simultaneous D/H and 16O/18O ratio measurements directly on H2O molecules with no conversion of H2O to H2, CO, or H2/CO2-water equilibration required. Here we present a detailed assessment of the performance of a liquid-water isotope analyzer, including instrument precision, estimates of sample memory and sample mass effects, and instrumental drift. We provide a recommended analysis procedure to achieve optimum results using OA-ICOS. Our results show that, by using a systematic sample analysis and data normalization procedure routine, measurement accuracies of +/-0.8 per thousand for deltaD and +/-0.1 per thousand delta18O are achievable on nanoliter water samples. This is equivalent or better than current IRMS-based methods and at a comparable sample throughput rate.     

On-line technique for the determination of the delta37Cl of inorganic and total organic Cl in environmental samples

Here we describe an on-line method for measuring delta(37)Cl values of chloride bearing salts, waters, and organic materials using multicollector continuous-flow isotope ratio mass spectrometry (CF-IRMS). Pure AgCl quantitatively derived from total Cl in water, inorganic Cl salts, and biological samples was reacted with iodomethane in evacuated 10-mL stopper sealed glass vials to produce methyl chloride gas. A GV Instruments Multicollector CF-IRMS with CH(3)Cl optimized collector geometry was modified to accommodate a headspace single-sample gas injection port prior to a GC column. The GC column was a 2-m Porapak-Q packed column held at 160 degrees C. The resolved sample CH(3)Cl was introduced to the IRMS source in a helium stream via an open split. delta(37)Cl values were calculated by measurement of CH(3)Cl at m/z 52/50 and by comparison to a reference pulse of CH(3)Cl calibrated to standard mean ocean chloride. Sample CH(3)Cl analysis time was approximately 6 min. Injections of 40 microL of pure CH(3)Cl gas yielded a repeatability (+/-SD) of +/-0.06 per thousand for delta(37)Cl (n = 10). Combined GC and IRMS source linearity for CH(3)Cl was <0.2 per thousand/nA (V) peak height. External repeatability, based on processing of seawater and NaCl reference solutions, was better than +/-0.08 per thousand. The smallest sample for delta(37)Cl analysis by this method was approximately 0.2 micromol of Cl. Selected results from a river basin and biological samples study illustrate the potential of on-line chlorine isotope assays in environmental pollution studies.