Anal. Chem. 1997, 69, 944-950
Carbon Isotope Analyses of Semivolatile Organic Compounds in Aqueous Media Using Solid-Phase Microextraction and Isotope Ratio Monitoring GC/MS Robert F. Dias and Katherine H. Freeman*
Department of Geosciences, The Pennsylvania State University, University Park, Pennsylvania 16802
Solid-phase microextraction (SPME) was used to facilitate the measurement of stable carbon isotope compositions (at natural abundance) of six organic compounds representing four compound classes in aqueous solution. Toluene, methylcyclohexane, hexanol, and acetic, propionic, and valeric acids were extracted from aqueous solutions with appropriate SPME phases and thermally desorbed into the split/splitless inlet of an isotope ratio monitoring gas chromatograph/mass spectrometer (irmGC/MS). Hydrophobic compounds (toluene, methylcyclohexane, hexanol) extracted by a nonpolar SPME phase were slightly (e0.5‰) enriched in 13C while organic acids extracted with a polar phase were depleted in 13C to a somewhat greater degree (e1.5‰) relative to material remaining in the aqueous phase. Isotopic fractionation was not observed to vary systematically as a function of equilibration time or solute concentration. Further, isotope fractionation did not vary consistently with the partition coefficient (Kfw). However, both salinity and cosolvent effects, which altered the partition coefficients of the solutes, also yielded a reduction in the magnitude of isotopic fractionation (to e0.4‰ for the hydrocarbons, e0.5‰ for the organic acids). We conclude that fractionations are most likely associated with the interactions of organic compounds with the organic phase coating SPME fibers and are specifically due to mass-dependent energy shifts upon solution of each analyte into the organic phase. In addition, fractionations are also influenced by energy shifts associated with electrostatic forces acting on the analyte in the water phase during the partitioning process. The magnitude of isotopic fractionations can be minimized under conditions appropriate for the analysis of natural waters, and with careful calibration, SPME and irmGC/MS should be a valuable means for isotopic analyses for a wide range of organic constituents in aqueous samples. Solid-phase microextraction (SPME) enables the rapid extraction of organic compounds from an aqueous sample.1,2 Designed for the quantitative analysis of trace constituents in water samples, this method is highly sensitive and is considerably simpler than liquid-liquid extraction, solid-phase extraction, or purge and trap (1) Belardi, R. G.; Pawliszyn, J. Water Pollut. Chem. 1990, 24, 179. (2) Arthur, C. L.; Pawliszyn, J. Anal. Chem. 1990, 62, 2145.
944 Analytical Chemistry, Vol. 69, No. 5, March 1, 1997
methods.2-5 SPME is based on the equilibrium partitioning of organic compounds between an aqueous sample phase and the organic polymer extraction phase and has been used successfully in the quantitative analyses of a wide variety of groundwater contaminants.3,5 Isotope ratio monitoring GC/MS is used routinely to determine the carbon isotope abundances of individual organic compounds in natural samples. Compound-specific isotope abundance data allow the geochemist to make a genetic link between compounds and to identify, quantify, and monitor biologically and diagenetically induced molecular transformations.6-8 Organic transformations in aqueous systems is a topic of considerable interest, especially for understanding the fates of crude oil and refined petroleum hydrocarbons in near-surface aquifers. The concentrations of these hydrocarbons are affected by processes that include groundwater flow, volatilization, dispersion, sorption, and biodegradation, including oxic and anoxic biogeochemical processes.9,10 To study these processes, previous work has focused on the quantitation of dissolved organic species to identify contaminant degradation products as a function of time and distance from a contamination source. Molecular products from natural biological and chemical degradation reactions often demonstrate a systematic 13C depletion, reflecting the sequential imposition of primary kinetic effects involved with bond breakage and formation.11-15 It is likely this is also true for the degradation of organic contaminants in water, and isotopic analysis of dissolved organic species will allow further study of reaction mechanisms and associated isotopic fractionation in such systems. (3) Arthur, C. L.; Killam, L.; Motlagh, S.; Lom, M.; Potter, D.; Pawliszyn, J. Enivorn. Sci. Technol. 1992, 26, 979. (4) Arthur, C. L.; Potter, D.; Bucholz, K.; Motlagh, S.; Pawliszyn, J. LC-GC 1992, 10, 9. (5) Arthur, C. L.; Pratt, K.; Bucholz, K.; Pawliszyn, J.; Belardi, R. J. High Resolut. Chromatogr. 1992, 15, 254. (6) Freeman, K. H.; Hayes, J. M.; Trendel, J.-M.; Albrecht, P. Nature 1990, 343, 254. (7) Hayes, J. M.; Freeman, K. H.; Popp, B. N.; Hoham, C. Org. Geochem. 1990, 16, 1115. (8) Freeman, K. H.; Wakeham, S. G. Org. Geochem. 1993, 19, 277. (9) Baedecker, M. J.; Cozzarelli, I. M.; Eaganhouse, R. P. Appl. Geochem. 1993, 8, 569. (10) Eganhouse, R. P.; Baedecker, M. J.; Cozzarelli, I. M.; Aiken, G. R.; Thorn, K. A.; Dempsey, T. F. Appl. Geochem. 1993, 8, 551. (11) Monson, K. D.; Hayes, J. M. Geochim. Cosmochim. Acta 1982, 46, 139. (12) Clayton, C. J. Org. Geochem. 1991, 17, 887. (13) Hayes, J. M. Mar. Geol. 1993, 113, 111. (14) Popp, B. N.; Hayes, J. M. Energy Fuels 1993, 7, 185. (15) Collister, J. W.; Reiley, G.; Stern, B.; Eglinton, G.; Fry, B. Org. Geochem. 1994, 6/7, 619. S0003-2700(96)00635-X CCC: $14.00
© 1997 American Chemical Society
This study assesses the combination of the two analytical techniques as a potential new tool for the determination of stable carbon isotope ratios of low molecular weight organic compounds in dilute aqueous solutions. Organic compound classes considered here include alphatics, aromatics, alcohols, and organic acids. Representative compounds selected for this study include methylcyclohexane, toluene, hexanol, and acetic, propionic, and valeric acids. For each compound, we determined the SPME partition coefficient (Kfw) and equilibration time, minimum solution concentration necessary for precise isotopic analysis, and carbon isotopic fractionation associated with SPME. EXPERIMENTAL SECTION Materials and Instrumentation. The manual SPME syringe employed in this study was purchased from Supelco Inc. (Bellefonte, PA). Aqueous samples (typically 3-5 mL) were contained in glass vials with Telfon-lined, Thermogreen septum-lined caps and were rapidly stirred with a Teflon-coated small magnetic stir bar in order to minimize the boundary layer around the organic phase.2 The coated fiber was exposed within the stirring sample until equilibrium was established for the partitioning of the analyte between the aqueous and organic phases. Diffusive equilibrium is achieved when the amount of analyte extracted remains constant regardless of increasing exposure time. The SPME syringe was introduced into a gas chromatograph inlet where the adsorbed compounds were thermally desorbed from the organic phase. Once the analytes are introduced onto an appropriate GC column, quantitative, qualitative or isotopic analyses are possible. In this study, the GC (Varian 3400) is connected to a high-precision stable isotope ratio mass spectrometer (Finnigan MAT 252) through a microvolume combustion interface.16,17 Detailed treatment of theoretical and practical aspects of irmGC/MS are presented elsewhere.7,16-20 Equilibration Times. For toluene (289 ng/mL in distilled and deionized water) a 100-µm poly(dimethylsiloxane)-coated fiber was exposed to stirring solutions for a period of 1, 3, 5, 10, 20, and 30 min prior to irmGC/MS analysis. Toluene was desorbed from the SPME fiber at 220 °C (5 min) in the GC inlet under splitless conditions. The GC oven was temperature programmed as follows: 60 °C to 100 °C at 5 °C/min; isothermal at 100 °C for 10 min. The column was a Supelcowax-10 (30-m × 0.25-mm × 0.5-mm phase). Similar extraction and chromatographic conditions were used for methylcyclohexane (80 ng/mL). Hexanol (6.7 mg/mL) was extracted using a 75-µm β-cyclodextrin-coated fiber, under otherwise similar experimental conditions. Organic acids were analyzed as a multicomponent mixture of acetic (477 mg/mL), propionic (239 mg/mL), and valeric (99.7 mg/mL) acids in deionized-distilled water. Multiple solutions were acidified (pH
