ANALYST, JANUARY 1989, VOL. 114 47 Determination of Volatile Aromatic Hydrocarbons in Estuarine and Coastal Sediments Using Gas Syringe Injection of Headspace Vapours and Gas Chromatography With Flame-ionisation Detection Alexander Bianchi Environmental Laboratory, Exxon Chemical Compan y, Cadland Road, H ythe, Southampton, Hampshire SO4 4WH, UK Mark S. Varney Department of Oceanography, Building 3, University of Southampton, Southampton, Hampshire, UK A simple, low cost flame-ionisation detection method for the determination of volatile aromatic compounds in estuarine and coastal sediments is described. Headspace vapours are drawn from a modified sample vessel at 80 "C and injected by means of a gas-tight syringe with a valved needle. The method eliminates the difficulties norma I ly encou ntered with solvent extraction and dynamic " non-equ i I i bri u m " heads pace methods.The effect of varying the sample preparation parameters is discussed and results giving the optimised values are presented. Relative standard deviations of less than 2% were achieved for a variety of sub-marine and tidal sediments and the results were found to be superior to those given by an existing solvent extraction method. The limits of detection of the method are below 0.5 pg kg-1 (dry mass) for ten key volatile organic aromatic compounds and the response is linear up to at least 200 pg kg-1. Keywords: Volatile aromatics; estuarine sediments; static headspace; gas syringe; flame-ionisation detection gas chromatography The study of the temporal and spatial variation of volatile organic carbon (VOC) in coastal and estuarine sediments has received comparatively little attention in the past.Geochem- ical studies carried out on the coastal marine environment have tended to restrict themselves largely to the study of metals,' total dissolved carbon' and the concentrations and distribution of higher relative molecular mass hydrocarbons.' Numerous studies in the last area have been carried out by the Ministry of Agriculture, Fisheries and Foods (MAFF), principally to widen the understanding of environmental pollution caused by oil from both shipping and industry.4 Consequently, studies of VOC in sub-marine sediments have been conducted almost exclusively as part of large scale programmes associated with petroleum exploration and pet- rogenesis studies.5 This is perhaps not surprising considering the extensive practical difficulties involved in sampling sedi- ments at depths in excess of 10 m from smaller marine vessels.Further, until comparatively recently, few analytical metho- dologies had been developed sufficiently to enable analyses other than simple "oil content" and alkane fingerprint profiles to be performed on recovered sediments. Examples of newer techniques are provided by Readman et a1.6 who developed a complete method for the analysis of sewage, oil and polycyclic aromatic hydrocarbon (PAH) pollution using a single sedi- ment sample. The term VOC, however, covers a group of compounds ranging approximately from methane to dodecane, which originate from both anthropogenic and biogenic sources and contain a variety of organic groups including alkanes, alde- hydes, furans, alcohols, organochlorines and alkylbenzenes.Benzene and substituted benzene compounds, although thought to be derived principally from man-made sources,7 e.g., fuel oils and gasolines, can also be generated selectively as by-products of biological processes via plant metabolic pathways.8 The major VOCs are benzene, toluene, ethylben- zene, m-xylene. o-xylene, cumene, propylbenzene, 13-5- trimethylbenzene, 1,2,4-trimethylbenzene and 1,2,3- trimethylbenzene. Many studies, including those of Laskin and Goldstein," have pointed to the carcinogenicity of benzene in animals. In addition, the United States Environ- mental Protection Agency (EPA) has defined benzene, toluene and ethylbenzene among its published criterion list of 65 priority pollutants, based on factors such as the frequency of occurrence in water, chemical stability and the structure and mass of the pollutant produced.10 The EPA policy states that there is no scientific basis for calculating safe levels of carcinogens, nevertheless they set certain risk levels for various toxic substances, e.g., the risk that 1 in 100000 people will contract cancer is at the 1 .0 pg 1 - 1 level. Levels exceeding 120 pg 1-1 have been reported in the water column of the Southampton estuary. 11 Therefore, in order to screen sediments for VOC and in particular those compounds classified as priority pollutants, it is desirable to have an analytical method capable of detection at the 1.0 pg kg-1 level (dry mass).The majority of methods developed so far for the determi- nation of hydrocarbons in sediments involve liquid - liquid extraction into a purified colvent, using either sonication or Soxhlet extraction processes.1*,13 However, these approaches are hindered by inadequate and time consuming solvent purification steps,'4 by the need for clean-up stages to remove interfering compounds (i. e., oxygenates) and by Soxhlet extraction times of up to 72 h.15 Once extraction is complete, the analysis is normally conducted using gas - liquid chromato- graphy and high-performance fused silica capillary columns. Accordingly, unless strict quality control procedures are applied throughout the preparation of the sample, contamina- tive effects and volatilisation losses will make the analysis both complicated and difficult.Among the various headspace alternatives, purge and trap (P and T) methods, used largely for aqueous samples,Ih have been applied only rarely to sediment samples, and have been reported to be potentially complex. 17 In view of the difficulties encountered in solvent extraction and P and T analysis, we have investigated a simple manual "static" equilibrium head- space approach using a gas syringe and a re-designed sediment headspace sampling vessel. The developed technique avoids the necessity for solvents and clean-up steps, retains the integrity of the sample, facilitates handling strategies and is an improvement over an existing method for static headspace48 ANALYST, JANUARY 1989, VOL.114 sampling of sediments. 18 Manual headspace sampling by means of a gas syringe has been reported to give poor reproducibility, particularly in its application to the determi- nation of organohalogens.19 However, Croll et af.2" have developed an improved version of this method and demon- strated that relative standard deviations (RSDs) of less than 2% could be attained by strict control of the key method parameters. Equilibrium headspace methods for the analysis of marine sediments have been reported by Burke et af.,21 but a more comprehensive method for the specific determination of VOCs in sediments was subsequently developed by Hunt and Whelan.22 The sampling protocol involves placing 100 g of frozen sediment in a 600-ml capacity metal can fitted with two silicone rubber septa.The can is filled with degassed water, sealed and a headspace created by removing 150 ml of water, which are replaced by an equal volume of helium. The can is allowed to thaw overnight, shaken vigorously for 3 min and heated in boiling water for 30 min. Headspace gas aliquots (1.CL50.0 ml) are then injected either directly with a gas syringe or via gas sampling valves into a gas chromatograph. This method was duplicated in the Department of Oceano- graphy, but was found to present practical difficulties. These included leakage of gas from the can, leakage at the septum - can joints and the adherence of sediment to the roof of the can after shaking, causing subsequent fouling of the sampling syringe needle. The repeatability was found to be inadequate with a typical RSD in excess of 8% for all components (possibly owing to fugitive leaks).Accordingly, the headspace vessel was re-designed in an all-glass construction and a more detailed procedure adopted in order to improve the performance of the method. Good precision can be obtained by using modern gas-tight syringes and by maintaining strict control of the experimental paramet- ers, i.e., temperature and equilibration time. This paper describes a simple and rapid method for the routine determi- nation of volatile aromatic compounds in coastal, estuarine and beach sediments. The method involves the collection of sediment cores, scrapings, etc.. into septum-sealed glass vessels, equilibration at 80 "C in a water-bath and headspace gas sampling with a gas syringe.The test data generated using the proposed method are presented. In addition, the tech- nique has been used successfully in an industrial environmen- tal laboratory for the routine analysis of waste sediments and sludges. Experimental Apparatus A Perkin-Elmer Sigma 3B gas chromatograph with a flame- ionisation detector and an LCT-100 computing integrator - plotter was used together with a fused silica capillary column, SO m X 0.22 mm i.d., of WCOT (BP-1) (0.S-vm film thickness) (SGE, Milton Keynes, UK) under the following conditions: injector temperature, 300 "C; detector temperature, 300 "C; initial column temperature, 60 "C for 5 min then increased at 10 "C min-1 to give a final column temperature of 200 "C held for 1 min; carrier gas, helium; and column head pressure, 25.3 lb in-2 (pressure control).Further apparatus used included the modified sampling vessels (Hampshire Glassware, Southampton, UK) , nominal capacity 850 ml; aluminium foil coated PTFE septa (Perkin- Elmer HS-6 septa modified for use with the sampling vessels); a gas-tight syringe with a valved needle, Pressure-Lok Type A-2, 2.0-ml capacity (Precision Sampling, Baton Rouge, LA, USA): and plunger-in-needle liquid syringes (SGE), 5- and 25-pI capacity. A commercially available "picnic" insulation box (Geeco Coolbox) was used for storage and transportation of the vessels and was packed with dry-ice (Cardice) for sub-ambient cooling. Plastic safety containers (BDH, Poole, UK) were used to stand the sampling vessels upright in the Alu rn inium-faced PTFE septa (16 mm) ,Schott screw caps \ Glass seal spring I oca t i n g " h o r n s " Nominal 150 ml 700 ml Liquid - solid capacity Fig. 1.Bulk sample headspace equilibration vessel box. A thermostated water-bath (Grant Instruments, Cam- bridge, UK) was also required for high-temperature equilibra- tion of the samples. The modified headspace sampling vessel is shown in Fig. 1. Reagents De-ionised water, containing less than 10 ng 1-1 of total aromatics, was used. It was purged with filtered ultrapure nitrogen 24 h prior to use. Undecane. Redistilled, containing less than 10 ng 1-1 of total aromatics. Aromatic standards. Benzene, toluene, ethylbenzene, m-xylene, o-xylene, cumene, propylbenzene, 1,3,5-trimethyl- benzene, 1,2,4-trimethylbenzene and 1,2,3-trimethylbenzene were of chromatographic grade.Sodium azide. AnalaR grade. Stock standard solution. A combined solution containing 10 mg 1-1 of each of the ten aromatic hydrocarbons (each added with its own syringe) in undecane was prepared. The standard preparation method was based on a published CONCAWE method.23 The solution remained stable for at least 1 week if stored at <4 "C in a glass-stoppered flask. Dilute standard solutions. Prepared using further syringes to dilute the solutions serially to give concentrations of 1,10,20,50,70 and 100 pg 1-1 in undecane. Standard .sediment. The optimum matrix to use has been stated to be that of the matrix itself.17 Sediment was taken from a relatively unpolluted estuarial site on the Southampton estuary.The sediment was subjected to rigorous clean-up procedures in order to remove all the volatile organic compounds. These included solvent extraction methods nor- mally used for sediment clean-up,24 in addition to boiling, stripping with nitrogen and washing with water prior to drying in an oven at 105 "C for 72 h. It was recognised that these procedures could destroy or modify the adsorptive sites within the grain - particle matrix; however, a suitable compromise was sought between the necessity for a representative "blank" sediment and the physico-chemical integrity of the original sediment itself. Method Batches of sediment (100-150 g) were cooled in a desiccator and added gravimetrically to the sample vessels, which were then sealed and purged with ultrapure nitrogen for 3 min.ANALYST, JANUARY 1989, VOL.114 49 Known aliquots of the standard solutions were spiked into the vessels (via the septum inlets) and the sediment was shaken gently to incorporate the spiked material. Blanks were prepared by spiking the sediment with undecane only. Standards were analysed immediately after preparation. The exact concentration (in pg kg-1) of the aromatic hydrocarbons was calculated from the initial concentration of the liquid standards and the exact mass of sediment taken (both known accurately to four significant figures). Sample Collection and Storage The sample vessels were cleaned with detergent, washed with acid and water and stored overnight at 150 "C. New septa were fitted to the vessels for each sample. Sample cores and scoops were taken (with Van-Veen sediment grabs for shipborne sampling) and approximately 100 g were placed in each vessel.The vessels were capped and stored on a dry-ice bed inside the collection box. Sodium azide was added (approximately 0.5 g) Table 1. Analytical procedure Step Experimental procedure Notes 1 Conduct steps 2-14 for duplicate calibration standards. Plot mean integrator counts (or peak heights) against correctcd con- centrations on calibration graphs for each standard. Inte- grator counts should not differ by more than 2-3% of each other Begin with low concentration standards. A linear plot should be obtained. Intermediate stan- dards may be deleted later if linearity is reproducible. A computing integrator can be programmed to identify and calibrate all peaks 2 Samples are removed from the insulation box.Standards and blanks are removed from ref- rigerated storage. Ensure the securing springs are attached to the glass mountings Samples and blanks should be analysed immediately after removal from the refrigerator or from coolbox storage 3 Remove one septum cap and decant 500 ml of de-ionised water into the vessel This step should be executed promptly. A pre-cleaned glass funnel can be used 4 Purge ultrapure helium into the A flow-rate of approximately 200 ml min-1 should be suffi cient to achieve a helium atmosphere headspace. Use only pre-cleaned stainless- headspace above the water level through a stainless-steel tube for 1 min 2 5 s. Refit the septum cap and switch off the helium steel tubing 5 Allow frozen sediment samples This step was unnecessary with to thaw at room temperature for 1 h k 2 min standards and blanks, unless they were re-frozen 6 Agitate the headspace vessel for approximately 5 rnin to achieve dispersion of the solid phase into the water phase Vigorous shaking is not neces- sary 7 Immerse the headspace vessel containing the blank, standard or sample in a hot water-bath pre-set at 80 "C.The vessels can be clamped to avoid instability in the bath. An aluminium lid with apertures cut to support up to eight vessels is a recommen- ded accessory The water line should extend to at least 75% of the height of the vessel. The bath temperature will drop by about 5 "Con entry of the vessel depending on the geometry of the bath and the mass of water 8 Allow the headspace vessel(s) to equilibrate in the bath for 45 k 1 min Agitate the vessel(s) every 10 rnin for approximately 1 min to poison any biological processes occurring in the sediment and principally to minimise the degradation of organic compounds by bacteria.On returning to the laboratory, the box can be replenished with dry-ice if short-term storage (i. e., 2-3 h) is intended prior to analysis. Overnight storage in the deep-freeze compartment of a refrigerator is possible, although same-day analysis is recommended whenever prac- tical. Analytical Procedure The analytical procedure is given in Table 1. Results and Discussion A headspace gas sample volume of 2.0 ml was found to provide a satisfactory chromatographic peak area response (>lo00 pV s) at the 1.0 pg kg-1 level for a range of sub-marine, estuarial and beach sediments.Calibration graphs were linear from 0.1 to 200 pg kg-1. Step Experimental procedure 9(a) During equilibration of the headspace check the syringe by injecting a 2.0-ml aliquot of nitrogen into the gas chromato- graph. Draw 2.0 ml of nitrogen slowly into the syringe, close the syringe valve and slide the needle into the injection port of the gas chromatograph ( h ) Compress the gas against the valve to the 1.0-ml mark, open the valve and inject the gas to the 2.0-ml limit. Switch on the integrator - plotter. Use this technique for all injections. Repeat step 9(a) if contami- nant peaks are found 10 Conduct steps 3-14 for succes- sive blanks. Sample the head- space gas directly from the vessel immersed in the bath.It should not be necessary to remove the bath 11 Equilibrate and analyse each sample (steps 2-14). Include a quality control standard every two analyses. Calculate the absolute concentration of each aromatic hydrocarbon 12 Allow the vessel to cool after removal from the bath. Remove the headspace vessel cap and filter the sediment sample into a pre-weighed Pyrex glass drying dish (300-ml capacity) 13 Dry the sediment in an oven at 105 "C for 8-12 h. Place in a desiccator and allow to cool for 1-2 h, then re-weigh 14 Calculate the mass of aromatic hydrocarbon relative to the mass of sediment (dry mass) recovered for each sample. Express the concentration as pg kg (dry mass) Notes This ensures that the syringe is free from organic contami- nants which would generate "ghost" peaks on the chro- matogram. This procedure should be accomplished in one smooth step Set plotter - chart recorder at 10 mm min-1.An integrator delay can be used to ignore both pressure and air peaks which may appear at the start of the chromatogram. The integrator can be set up to start the GC run automat- ically A satisfactory blank should contain <1 pg kg of each aromatic hydrocarbon. If levels in excess of this are recovered, check the syringe and, if necessary, check with a third blank sample Plot - integrate as described in step 1 Weigh the drying vessel to within 0.001 g. Spread the sediment evenly across the vessel surface. Agitate and stir during drying Continue to agitate the sedi- ment to assist the removal of water This calculation should be programmed into the comput- ing integrator if possible50 ANALYST, JANUARY 1989, VOL.114 Table 2. Results of replicate analyses of standard sediment(s) expressed as relative standard deviation (YO) Component Benzene . . . . . . Toluene . . . . . . Ethylbenzene . . . . m-Xylene . . . . . . o-Xylene . . . . . . Cumene . . . . . . Propylbenzene . . . . 1.3.5-Trimethylbenzene 1.2,4-Trimethylbenzene 1.2,3-TrimethyIbenzene No. of samples: Concentration of standard in sediment(s)/pg kg-1 (dry mass) 1 10 20 50 75 100 . . . . 1.8 1.8 1.9 2.0 1.9 1.8 . . . . 1.8 2.0 2.0 2.1 2.1 2.0 . . . . 2.0 2.2 2.2 2.2 2.3 1.9 . . . . 1.5 1.8 2.2 1.8 1.7 1.7 . . . . 1.5 1.7 2.3 1.8 1.6 1.7 . . . . 2.8 2.5 2.5 2.7 2.0 1.9 . . . . 1.9 2.2 2.2 2.3 2.2 2.2 .. . . 1.5 1.9 1.9 1.4 1.5 1.4 . . . . 1.6 1.8 1.8 2.0 2.0 2.1 . . . . 1.7 1.8 1.8 1.9 2.2 2.0 9 9 5 8 5 9 c a, - I- A C ID II 0 5 10 Timeimin 15 Fig. 2. Headspace analysis chromatogram for 10 pg kg-1 of each aromatic compound. (A) Benzene; (B) toluene; (C) ethylbenzene: (D) m-xylene; (E) o-xylcnc; (F) cumene; (G) propylbenzene; (14) 1.3.5-trimethylbenzene; (I) 1,2,4-trimcthylbenzene; and (J) 1.2,3- trime t hylbenzenc The precision of the method was evaluated and the data obtained are presented in Table 2. A specimen chromatogram of the 10 pg kg-1 calibration standard is shown in Fig. 2. The total GC run-time is just under 15 min per sample; however, the analysis is run up to 200 "C in order to remove higher boiling compounds from the column. Note that undecane elutes at approximately 17.0 min under the standard condi- tions described under Apparatus.Standards were spiked with known aliquots of higher relative molecular mass organic compounds, e.g., naph- thalene, to determine if there were any quantitative interfer- ence effects on the calibration values. Subsequent analyses yielded data falling within the RSD values given in Table 3. Comparison of the headspace method with an existing dichloromethane extraction technique13 yielded an average 260% increase in efficiency, expressed as actual peak area, for Table 3. Replicate analyses of a mid-estuarial sediment sample Aromatic hydrocarbon Foundipg kg- 1 RSD, Yo Benzene . . . . . . Toluene . . . . . . Ethylbenzene . . . . m-Xylene . . . . . . o-Xylene .. . . . . Cumene . . . . . . Propylbenzene . . . . 1,3 ,S-Trimethylbenzene 1,2,4-TrimethyIbenzene 1,2,3-Trimethylbenzene . . . . . . 8.4 . . . . . . 49.3 . . . . . . 7.1 . . . . . . 4.7 . . . . . . 2.2 . . . . . . 1.2 . . . . . . 0.6 . . . . . . Not detected . . . . . . 0.5 . . . . . . 0.5 1.9 2.0 1 .5 1 . 1 1.2 1.1 1.9 1.8 1.8 - benzene, toluene and ethylbenzene and an increase of >I%% for the remaining seven aromatic compounds. The limit of detection of the method was defined as recommended by Grob and Kaiser14 and Kolb et aZ.,25 i.e., as the smallest amount of sample that will cause a measurable signal (e.g., twice the noise) over the noise signal. This is also known as the minimum detectable level (MDL). The detector specificity is the ratio of the detector response for a contami- nant to that for the desired component.Using this definition, the MDL -1alues were as follows: benzene, 0.009; toluene, 0.01: ethylbenzene, 0.07; 0- and m-xylene, 0.08; cumene, propylbenzene and 1,3,5-trirnethylbenzene, 0.17-0.23; and 1,2,4-trimethylbenzene and 1,2,3-trimethylbenzene, 0.33- 0.36 ug kg-1. Experiments were carried out to determine whether the natural salt content in the sediments would contribute to an unquantifiable "salting-out"effect. Sea water salinity values overlying the sediments varied from 10 to 33.6 parts per thousand at the seaward end of the Southampton water estuary. Although it was necessary to employ the salting-out effect in the method (which leads to an increase in the concentration of non-polar or low polarity compounds in the vapour phase by the addition of a soluble electrolyte to the liquid phase), interstitial water in the pores of the sediment will contain salt.Consequently, one of two identical specimen samples was analysed for its volatile aromatic content. A prepared blank sediment was spiked to give an equivalent concentration of benzene, toluene, o-xylene and 1,2,3- trimethylbenzene. The second specimen sample and the spiked blank were spiked further by "known additions'' of these four key aromatic hydrocarbons. On analysis, the resulting data, corrected for the mass of sediment taken, yielded results that were within 2.5% for all components, indicating that the original specimen samples were not contributing a salting-out effect at levels that were sufficient to interfere with the calibration of the method.Further analyses on samples taken from high- to low-salinity regions provided similar results. It was concluded that the comparatively small contribution of "salt" from the sediment relative to the large volume of water added, i.e., 500 ml, effectively negates any salting-out effect that may begin to influence the analysis at signficantly greater sediment volumes or smaller water volumes. Investigations were conducted into the effect of equilibra- tion-bath temperatures on peak area recovery data. Standards were equilibrated at 10 "C intervals from 30 to 90 "C for 45 k 1 min before analysis. The recovery was optimised for benzene, as this is the component of greatest environmental concern. The maximum recovery of benzene and toluene was achieved at 80 "C, although the recovery of the remaining compounds levelled off between 55 and 65 "C.The temperature was then varied at 2 "C intervals, i.e., 78, 80, 82 and 84 "C, respectively. From the plots of log(peak area) versus the reciprocal of the absolute temperature, changes in peak area of between 2 and 3% resulting from a 2 "C change in temperature were observed (2.6% for benzene and <2% for the other aromatic hydrocar- bons). These results are in broad agreement with those ofANALYST, JANUARY 1989, VOL. 114 Croll et ~ 1 . 2 ~ ) and suggest that temperature stability to within +2 “C is acceptable. It was calculated that the water-bath thermostat unit was capable of operating within these limits. Owing to the relatively high combined mass of the sample and vessel. stable equilibration was not fully achieved until a minimum of 30 min had expired. Equilibration times of between 35 and 60 min were found to provide repeatable recoveries.To achieve the best compromise between stable equilibration and total analysis time, an equilibration time of 45 min was selected as the optimum. Two further aspects of the technique were investigated, viz., the effect of varying the sample to water volume ratio with respect to mass peak area recovery and the effect of performing repeated injections from the same vessel, known as multiple headspace extraction (MHE). It was found that an increase in the mass of sediment relative to the volume of water added enriched the concentra- tion of organic vapour in the headspace, mainly owing to the high mass and a secondary, smaller, salting-out effect.However, estuarine samples actually contain a wide range of organics and this is reflected in the more complex chromato- grams obtained. Therefore, the resulting peak area is much larger than that necessary for accurate quantification at the 1 .0 pg kg-1 level. Frequently, complex chromatograms require the use of GC - MS to elucidate the identities of the organic components. Although we have used GC - MS in conjunction with headspace analysis, this is not necessary for complete quantification and identification using the given method parameters. The use of MHE, as described by McAuliffe,26 allows the analyst to define the distribution coefficient of a particular compound or range of compounds.It provides a qualitative mechanism €or the identification of trace amounts of com- pounds as part of a primary analysis. The nature of the distribution coefficient indicated by successive equilibrations provides information on the behaviour of the compound. Volatile aromatic hydrocarbons have a low distribution coefficient compared with alkanes and cycloalkanes, enabling them to be recovered quantitatively after three equilibrations. The application of MHE to a variety of sample matrices, e.g., in the analysis of halogenates and aromas has been discussed by Kolb et al.27 The use of MHE in this method is therefore limited to a theoretical appreciation. Our results indicated that a second 2.0-ml aliquot could be withdrawn from the headspace with no major alteration in the peak recovery within 1 standard deviation.However, removal of a third aliquot resulted in a reduction in the benzene peak area of at least 5%. We would not therefore recommend the use of MHE beyond a second injection from the same headspace sample. However, the analysis of headspace aliquots (single head- space extraction) from replicate samples generated reprodu- cible data to within an RSD of 2% (four replicates). Each aliquot was injected by separate analysts in both a University and an external environmental laboratory, including an operator with no prior experience of gas chromatography or syringe handling. The data obtained are shown in Table 3. Conclusions The comparative simplicity of the headspace vessel and the sampling preparation technique minimise the problems asso- ciated with handling bulk sediment samples.The technique is designed to circumvent contamination and component losses by sampling up to the moment of headspace injection. Samples can be taken. prepared and analysed by relatively inexperienced staff with the minimum of training. Any sources of error are negated by avoiding the use of solvents or ancillary sample handling steps. The sensitivity and precision of the method are superior to those of an existing standard solvent extraction method. 13 Temperature control, equilibration time and syringe sampling techniques are the nucleus of the method 51 and accurate reproduction of the method parameters will maintain the integrity of the method. The proposed method is now in use in the Exxon Chemical Company Environmental Laboratory, where it is applied to the rapid sampling and analysis of sludges, muds and solid wastes. Over 150 coastal and estuarine samples have also been analysed successfully using the method, and valuable environ- mental data on the flux of key aromatic compounds have been obtained.Additional experiments have also led to a broaden- ing of the scope of the method and to its use in the simultaneous determination of other organic compounds including alkanes and cycloalkanes. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. References Clark, R. B., “Marine Pollution,” Clanden Press. Oxford, 1986, p. 173. Froelich, P. N., Limrzol. Oceunogr., 1980, 25, 564.Page, D . S., Foster, J. C.. Fickett, P. M., and Gilfillan. E . S., Mar. Pollut. Bull., 1988, 19, 107. Law, R. J . , Mar. Pollut. Bull., 1981, 12, 153. Hunt, J . M., Huc, A. Y., and Whelan, J . K., Nature (London), 1980, 288, 688. Readman, J . W., Preston, M. R., and Mantoura, R . F. C . , Mar. Pollut. Bull., 1986, 17, 298. “Benzene: Control of Toxic Substances in the Atmosphere,” OECD Environment Monograph No. 5 , OECD, Paris, 1986. “Evaluation of Benzene Toxicity in Man and Animals,” DGMK Project 1 7 6 6 , German Society for Petroleum Sciences and Coal Chemistry, Hamburg, 1980. Laskin, S., and Goldstein, B. D., “Benzene Toxicity,” Ameri- can Petroleum Institute Report, Hemisphere Publishing, Washington, DC. 1977. Grob, R . L., in Grob. R . L., Editor, “Modern Practice o f Gas Chromatography,” Second Edition, Wiley, New York, 1985, Chapter 10. Knap. A. H., Le B . Williams, P. J . , and Tyler, I.. Nature (London), 1979. 279. 517. Murray, D. A. J . , J. Chromatogr., 1979, 177. 135. Venkaste, M. I., Ruth, E., and Kaplan, I. R . , Mar. Pollict. Bull., 1986, 17, 554. Grob, R. L., and Kaiser, M. A., “Environmental Problem Solving Using Gas and Liquid Chromatography,” Journal of Chromatography Librury , Elsevier, Amsterdam, 1982, Vol- ume 21, p. 91. “Determination of Very Low Concentrations of Hydrocarbons and Halogen Hydrocarbons in Water 1984-1985,” HM Stationery Office, London, 1985. Grob, K., and Zurcher, F., J. Chromatogr., 1976, 117, 285. Pizzie, R . , PhD Thesis, University of Southampton, 1984. Hunt, J. H., and Whelan. J . K., Org. Ceochem., 1979, 1,219. Otson, R., Williams, D. T., and Bothwell. P. D . , Environ. Sci. Technol., 1979, 13, 936. Croll, B. T., Sumner, M. E.. and Leathard, D. A , , Analyst, 1986, 111, 73. Burke, R. A , , Jr., Brookes, J . M., and Sackett, W. M.. Geochim. Cosmochim. Acta, 1981, 45, 627. Hunt, J . H . , and Whelan, J . K., Geochim. Cosmochim. Acta, 1980.44, 1767. CONCAWE Report No. 8/86. CONCAWE (“Oil Companies International Study Group for the Conservation o f Clean Air and Water”), The Hague, The Netherlands, 1986. Lee, H.-B., Hong-You, R. L.. and Chau, A.S.Y.. Analyst, 1986, 111, 81. Kolb. B.. Kraub, H . , and Aucr. M., “Pcrkin-Elmer Applica- tions Paper 21/1978/I-ISA-21,** Perkin-Elmer, Buckingham- shire, 1978. McAuliffe, C.. Chemrech, 1971, January, 4. Kolb, B., Pospisil, P., and Auer, M., Clzromatogruphia, 1984, 19. 113. Paper 8103 181 I Received August 3rd, I988 Accepted September 22nd, 1988