Analyst, September 1996, Vol. 121 (85R-89R) 85R Tutorial Review Extract ion of Pol ycycl ic Aromatic Hydrocarbons From Environmental Matrices: Practical Considerations for Supercritical Fluid Extraction J. R. Dean Department of Chemical and Life Sciences, University of Northumbria al Newcastle, Ellison Building, Newcastle upon Tyne, UK NEl 8ST This paper reviews the important practical considerations necessary for the development of successful supercritical fluid extraction using selected examples from the published literature. In particular, the extraction of polycyclic aromatic hydrocarbons from environmental matrices are evaluated. Finally, recommendations are suggested for the implementation of quantitative supercritical fluid extraction. Keywords: Supercritical fluid extraction; polycyclic aromatic hydrocarbons; environmental; selection of operating conditions; review Introduction PAHs are organic compounds containing carbon and hydrogen only, and consisting of fused benzene rings.Their widespread occurrence is due largely to the incomplete combustion of wood, petrol, oil and coal. As we live in a modern society that is continually burning such commodities it is not surprising to find that PAHs are relatively widespread. Recent studies have shown the prevalence of PAHs in the environment,' in urban air,2 in agricultural soils3 and The utilization of supercritical fluid extraction (SFE) as a suitable technique for the extraction of analytes from environ- mental matrices has been reviewed by several groups.5-9 The most common supercritical fluid is carbon dioxide.However, the lack of a dipole moment often precludes its use for the extraction of compounds of even moderate polarity. It is common practice, to discover in the literature therefore, that the addition of a polar modifier (or entrainer) is required; methanol is probably the single most common modifier used. The wealth of literature suggests that SFE should be a suitable replacement for more traditional methods of sample preparation, e.g., liquid- solid, Soxhlet, Soxtec or sonication. However, the uptake of SFE in terms of commercial useage, establishment of robust standard methods and proliferation of instrumentation is relatively slow. In this review the practical issues that are important for the quantitative extraction of PAHs have been reviewed and future recommendations made.No description of the instrumentation required is given in this paper, but can be obtained from any of the reviews mentioned above. Use of Alternatives to Supercritical Carbon Dioxide While supercritical C02 has been the most often applied of supercritical fluids in spite of its limited ability to solubilize moderately polar and polar analytes it is still the most popular. The alteration of polarity, to a more polar supercritical fluid mixture, is often accomplished by the addition of an organic modifier, e.g., methanol. However, a number of groups have approached the limited polarity of supercritical C02 from an alternative approach, by utilizing other supercritical fluids. One of the earliest publications on the extraction of PAHs from river sediment, fly ash and a certified reference sample of urban dust was that of Hawthorne and Miller.lO In this paper, supercritical ethane and nitrous oxide were compared with carbon dioxide.[CAUTION]: The use of nitrous oxide in the presence of organic matter can lead to explosions. In addition, the use of 5% methanol-modified nitrous oxide was also compared to 5% methanol-modified carbon dioxide. The extraction conditions chosen (300 atm; 45 "C for pure solvents and 65 "C for modified solvents) were known to produce less than quantitative recoveries of the PAHs. This allowed direct comparisons between the solvation properties of each supercritical fluid or mixture thereof. Of the pure supercritical solvents, nitrous oxide was the most effective presumably due to its slightly higher dipole moment (0.2 D compared to 0.0 D for C02).However, the addition, of methanol-modifier was the most significant, in that, in each case greater recoveries were obtained. The most effective combination was found to be 5 % methanol-modified nitrous oxide. It was also noted that higher relative molecular mass PAHs were extracted with lower efficiency than lower relative molecular mass PAHs from the same matrix. By increasing the pressure to 350 atm quantitative recoveries could be obtained from the urban dust matrix using 5% methanol- modified supercritical nitrous oxide. In order to achieve good recoveries for the higher relative molecular mass PAHs investigated it was necessary to extract the certified reference urban dust sample for 60 min.However, this produced seemingly higher extraction efficiencies for the lower relative molecular mass PAHs. It should be borne in mind, however, that the certified reference sample values were obtained using Soxhlet extraction using both dichloromethane and 1 -t 1 benzene-methanol, as solvents. Quantitative recoveries were obtained from the river sediment sample but not the fly ash, with the exception of phenanthrene dlo, which was added as a spike. This would suggest that the fly ash was a potentially more polar matrix than the river sediment. Further work was suggested to optimize the extraction cell geometry and extraction conditions (pressure, temperature, supercritical fluid and modifier addi- tion). The same group" later investigated the extraction of PAHs from a petroleum waste sludge and a railroad bed soil using supercritical CHCIF2, N20 and C02.It was reported that the86R Analyst, September 1996, Vol. 121 more polar fluid, CHCIFz (dipole moment 1.4 D) consistently gave higher recoveries than C02 or N20, and similar or greater recoveries to those obtained using an 18 h dichloromethane sonication. It was suggested that CHC1F2 can efficiently remove matrix water from wet samples thus exposing the target analytes for extraction. The only limitation noted was the potential mechanical breakdown of the fused silica restrictors used in this work, which required replacement on a fairly frequent basis. CHF3 and CHClF2 were evaluated as suitable supercritical fluids for the extraction of PAHs from sodium montmorillonite clay by Howard et a1.12 The higher dipole moments of the two fluorocarbons, as compared to C02, was suggested as the important criteria that should lead to higher extraction recov- eries.In practice, however, this was not the case. Under equivalent extraction conditions the recoveries obtained were as follows: CHCIF? > C02 > CHF3, with CHC1F2 providing similar extraction recoveries to those obtained using a 12 h dichloromethane Soxhlet extraction. It is perhaps unfortunate therefore that the use of CHClF2 has not been investigated further. The reasons for this limited applicability of CHClF2 are significant, however; its high toxicity, ozone depleting proper- ties and high cost, thus negating its future useage as an 'environmentally friendly' solvent.An interesting approach was that reported by Hawthorne et ~ 1 . ~ 3 , who utilized sub- and supercritical water to extract PAHs from contaminated soils. They were able to demonstrate that increasing the temperature (at 350 bar) of the water (sub- critical) provided more efficient recoveries of the PAHs. No additional increase in recoveries was obtained above a tem- perature of 250 "C, even though 300 and 400 "C (supercritical state) were evaluated. In fact it was observed that slightly lower recoveries were obtained using supercritical water, this was explained by the poorer collection efficiencies in the chloroform trap. It should be noted from a safety viewpoint, that supercritical water that has not been purged to remove oxygen may be corrosive to stainless-steel vessels.The effect of pressure at 250 "C was also evaluated. However, changing the pressure from 5, 50 to 650 bar had little impact on extraction recovery. Using optimal conditions (250 OC, 50 bar and a 1 ml min-l flow rate) as determined for the contaminated soil, the conditions (using a 15 min extraction time) were applied to the extraction of PAHs from air particulate matter (a sample containing lower concentrations of the PAHs). A comparison, with the certificate values, obtained on the basis of a 48 h Soxhlet extraction, was in good agreement except for the higher relative molecular mass components. In order to investigate the poorer extraction of the higher relative molecular mass components (indeno [ 1,2,3 -cd] p yrene and benzo [ gh i] perylene) a 14 h sonication was performed in chloroform.No additional material was extracted. Similar lower recoveries were obtained for the same components using supercritical C02 (200 "C and 659 bar) and 10% toluene modified supercritical C02 (80 "C and 405 bar). Upon completion of this recovery data, the time of extraction was then evaluated. No additional components were extracted even after 90 min. Additionally, the potential for selective extraction was discussed. It was shown that water can extract more polar organics in preference to the less polar organics. However, it was also suggested that normal organic liquid extractions are typically non-selective. This is surely absurd, in that the choice of organic solvents is not as restricted as the choice of supercritical fluid.Therefore, for liquid extractions, it is possible to use a non-polar organic solvent, e.g., hexane, through to a polar organic solvent, e.g., methanol, with a whole range of polarities being achievable in between by the use of single or mixed organic solvents for liquid extractions. The benefits of supercritical fluids are surely related to the use of more environmentally friendly solvents with associated lower costs for disposal. Use of Alternative Organic Modifiers to Methanol for Improved Extraction Efficiency The most frequently encountered organic modifier in the literature for supercritical carbon dioxide is methanol. As the purpose is to increase the polarity of the supercritical mixture the choice of methanol is perhaps not a surprise.So, the reader will often find that the use of methanol-modified supercritical carbon dioxide will lead to more efficient recovery of environmental analytes, e.g., PAHs from soils. The actual percentage of modifier chosen, is often open to debate, with most commonly 10% being used. No particularly scientific approach has led to this conclusion, other than it is effective. However, some workers have focused particularly on the role and purpose of the organic modifier in more detail. Therefore, Hills and Hi1114 evaluated the role of a reactive solvent modifier in supercritical C02 to extract PAHs from two certified reference samples, a harbour sediment, HS-3 and an urban dust, SRM 1649. The reactive modifier used was hexamethyldisilane and trimethylchlorosilane in a 2 + 1 mixture.This reactive modifier mixture was found to be six times more efficient than supercritical C02 only and two times more efficient than 10% methanol-modified supercritical CO2. It was suggested that the derivatizing agent mixture displaces the analytes from the sample matrix. Results using the reactive modifier were in agreement with those obtained using Soxhlet extraction. The use of mixed modifiers may well hold advantages for the extraction of analytes from environmental samples. Tt has previously been reported's that the use of a mixed modifier may well lead to quantitative recoveries from montmorillonite clay. In this situation, it was suggested that the presence of water will act to swell the clay, while the addition of methanol, as a modifier, will desorb the target analytes.Lee et al. l6 have shown that the use of a mixed modifier, dichloromethane and methanol (1 + 4), together with a wet sample (50% moisture content) can lead to efficient recovery of PAHs from certified reference sediment samples (EC-1 and HS-3) and other contaminated sediment samples. In each case the use of a wet sample, together with the mixed modifier, resulted in compar- able results to those obtained by Soxhlet extraction. The SFE conditions were 120 "C and 36 MPa. The use of a second pump for modifier addition was considered to be essential. Dankers et a1.17 reported the use of dichloromethane as an effective modifier for the extraction of PAHs from soils. Tn this situation, the modifier (2 ml) was added as a spot spike to the chemically dried and cryogenically ground soil sample prior to commencement of extraction.Comparable results were ob- tained between this method and liquid-liquid extraction using 2 X 100 ml of light petroleum. It was suggested that the use of a static modifier would allow penetration within the soil matrix. A range of spiked and contaminated soils were evaluated. The lightest PAH, naphthalene, was found to give greater recoveries using SFE than the liquid-liquid extraction method. It was reported that evaporation after liquid-liquid extraction is known to lead to loss of naphthalene. A systematic study of the influence of modifiers on the SFE of PAHs from spiked and native contaminated soils has been reported by Tena et ~ 1 .1 8 In this paper, the modifiers were chosen to reflect their potential influence on analyte/matrix interactions. The following modifiers were evaluated: hexane (non-polar); dichloromethane (polar, but with no hydrogen bonding); acetone (hydrogen bonding acceptor); methanol (hydrogen bonding acceptor and donor, miscible with super- critical C02); and, water (scarcely soluble in supercritical CO2). Each modifier was added as a spot spike (0-400 pl) to the sample (0.2-1.0 g) and allowed to interact for up to 15 h. No effect on the extraction recovery was noted by the addition of hexane or dichloromethane to the supercri tical C02, whereas the positive influence of acetone was unclear due to the highAnalyst, September 1996, Vol. 121 87R RSD values reported.Significant improvements (in relation to the pure supercritical fluid) in extraction efficiency were noted for water and methanol. The use of a modifier was found to be essential for the quantitative extraction of native analytes. The reduction in particle size of the soil resulted in greater extraction efficiency. Methods to Prevent Restrictor Plugging A particular problem when using linear restrictors (silica or stainless-steel) is the occurrence of a reduced flow due to temporary plugging (may also be permanent requiring physical or chemical removal of the material) caused by sample matrices that contain large quantities of extractable material or water. While considerable effort in terms of resources have gone into preventing this problem it does not seem to be possible to invent a restrictor that does not block.Our experience is not quite the same as that reported in the literature. Using the Jasco back- pressure regulator and Hewlett-Packard nozzle, blockage problems are rarely encountered using the equipment, as purchased without any additional modification. In this labo- ratory, the free-standing back-pressure regulator has more commonly been used. It is always dangerous to say that we have never experienced blockages and indeed it would be untrue, but the phrase infrequent blockage problems or with minimal intervention would seem to be more appropriate. Recently, another manufacturer has produced a 'regulator' that should be more robust in terms of operation, compared to fixed silica restrictors. A novel approach to restrictor plugging when extracting analytes from soils was proposed by Pyle and Setty.19 In this work the addition of a copper scavenger removed any potential interference from soils with a high sulfur content, presumably forming copper sulfide.While this original paper proposed the addition of the copper scavenger in the form of a column subsequent more routine applications of the method have featured incorporation of the copper in the extraction cell itself. For the linear capillary flow restrictor, heating has been suggested as a method to prevent the frequently encountered intermittent or irreversible plugging.20 However, heating the entire restrictor to 50 "C resulted in poor collection efficiencies (30-6596) for PAHs. An improvement was made by nebulizing an organic solvent with the restrictor effluent.The collection efficiency improved to 80-90%. A simpler method, was to heat (50-200 "C depending on the sample matrix) all but the final 3 cm of the restrictor and depressurize directly into the collection solvent resulting in quantitative collection efficiencies (90- 100%). Collection of Analytes After Supercritical Fluid Depressurization The rapid adiabatic expansion that occurs upon depressurization of the supercritical fluid results in the production of a sample containing aerosol that must be efficiently trapped for quantita- tive recovery of analytes. In order to do this various potential methods are available and include the use of various collection solvents, solid phase traps, a combination of both or solventless collection.Langenfeld et af.21 have evaluated a range of collection solvents for the efficient collection of a range of compounds including PAHs. They concluded that the most important parameters for the collection solvent after depressur- ization were solvent polarity and temperature rather than solvent volume and height. Efficient trapping of most analytes could be achieved by maintaining the collection solvent at 5 "C. The same group22 evaluated solventless collection after static SFE for the extraction of PAHs from spiked sand. In this case, however, non-quantitative extraction of chrysene and benzo- [blfluoroanthene resulted. They concluded that the higher relative molecular mass PAHs required more supercritical C02 or a longer static extraction period for quantitative extraction.As this work was performed on spiked sand it has no relationship to native, contaminated soils. Meyer and Kleibohmer23 have investigated the role of liquid- solid sorbent trapping as a method for efficient collection of PAHs from a hydrochloric acid treated marine sediment. They found that 10% toluene-modified supercritical carbon dioxide at 140 "C and 400 atm resulted in efficient extraction of PAHs. Quantitative recovery was established using liquid-solid traps to minimize analyte losses during depressurization of the supercritical fluid. Higher recoveries were obtained using this method rather than the previously employed organic solvent collection solvent. The solid sorbents evaluated for sample retention and in situ cleanup were silica gel, Florisil and silica gel/alumina.In each case, quantitative recoveries were obtained for all the heavier PAHs with the exception of the lighter ones (naphthalene, acenaphthalene, acenaphthene and fluorene). Optimization of Operating Conditions One of the first reported papers on the investigation of operating conditions for the extraction of PAHs from a contaminated soil (SRS 103-100) was that reported by Lopez-Avila et aZ.24 In the paper seven variables were considered: pressure (150 and 300 atm), temperature (50 and 70 "C), percentage moisture of the soil (0 and lo%), cell volume (2.0 and 4.7 ml), sample size (1 .O and 2.5 g), extraction time (30 and 60 min) and modifier spike volume (50 and 250 pl). Owing to the limited number of experiments carried out (eight experiments), only main effects could be evaluated.The results showed that the recovery of PAHs was most affected by the time of extraction with pressure and percentage moisture, being second and third most im- portant, respectively. Quantitative recovery, relative to the certified values, was not obtained in most cases. A univariate approach to method optimization was performed by Levy et al.25 for the extraction of PAHs from soils and sediments. The variables investigated were pressure, temper- ature and flow rate. Using a fixed temperature of 75 "C for 40 min, it was found that the highest pressure investigated 450 atm gave acceptable results, as compared to the EPA method 8270, from the contaminated soil. Then, extraction temperature was investigated at a fixed pressure of 475 atm for 55 min from a river sediment.The highest temperature investigated, 150 "C, gave the closest values to the Soxhlet method values. Increasing the restrictor flow from 2.6 to 4.7 ml min-1 also provided concurrent recoveries of the PAHs studied, but in a shorter time period, 25 min, instead of 55 min. A more comprehensive study of temperature and pressure on the extraction of PAHs from an urban particulate matter and a highly contaminated soil was reported by Langenfeld et ~ 1 . 2 6 At 50 "C increasing the pressure from 350 to 650 atm has no effect on extraction efficiencies. High recoveries were obtained from the contaminated soil using a 40 min extraction period irrespective of the temperature. However, extraction from the urban particulate matter required both a high temperature (200 "C) and pressure (650 atm). It would appear therefore that extraction temperature is a more important variable than pressure for ensuring high extraction efficiencies.A further univariate study of the variables that affect extraction efficiency was undertaken by Reindl and Hofler.27 In this work extraction pressure, temperature, amount of C02, cell volume, collection solvent, collection solvent temperature, restrictor temperature, flow rate and addition of cosolvent were evaluated from soils spiked with PAHs. It was found that the optimum parameters were: 40 MPa, 80 "C, 31 1 (gas), 10 ml, 15 ml of acetone, 5 "C, 200 "C, 500 ml min-l (gaseous C02) and 5% methanol, respectively. The experimental conditions were88R Analyst, September 1996, Vol.121 then evaluated with respect to the extraction of PAHs from a real world loam soil sample and compared to an 8 h acetone Soxhlet extraction. However, prior to extraction various other modifications/alterations were made. The soil was mixed with a drying agent (pure magnesium chloride or magnesium sulfate) prior to placing in the extraction cell. In addition, 0.5 g of magnesium chloride or magnesium sulfate and copper granules were placed between the sample and the restrictor to eliminate any residual water or sulfur present in the soil, respectively. The results showed a decrease in percentage recovery for the heavier PAHs which was remedied by increasing the modifier content to 8%. In this situation, greater recoveries were noted for the lighter PAHs; this was explained by the losses that may occur in the preconcentration step that follows Soxhlet extraction.A combined temperature/modifier effects study on the recovery of PAHs from marine sediment, diesel soot and air particulate matter was reported by Yang et a1.28 In this work the pressure was fixed at 400 atm while the temperature was varied from 80 to 200 "C and the type of modifier varied (methanol, diethylamine or toluene) while being maintained at 10%. Temperature was found to be significant, in all cases and for all sample types studied, leading to improved recoveries when raised to 200 "C. The nature of the modifier was dependent upon type and the matrix. Perhaps surprisingly, methanol was found to have the least effect.Diethylamine was found to be the best giving increased recoveries for all sample types. The results obtained using 10% diethylamine-modified supercritical C02 compared favourably with those obtained using 14-48 h Soxhlet extractions. An experimental design approach for the extraction of PAHs from contaminated soil was done by Barnabas et al.29 In this paper the following variables were investigated using a central composite design: pressure (100, 150, 200, 250 and 300 kg cm-2), temperature (40, 55, 70, 85 and 100 "C), extraction time (10, 22.5, 35, 47.5 and 60 min) and percentage methanol (0,5, 10, 15 and 20%). Statistical treatment of the data identified extraction time and percentage modifier as the significant variables. Using the maximum values of these two variables and the midpoint values for pressure (200 kg cm-2) and temperature (70 "C) gave values which agreed with extraction recoveries using microwave-assisted extraction using acetone as the solvent, but considerably greater than those obtained using a 6 h Soxhlet extraction with dichloromethane.The use of an experimental design approach allows a rapid approach to method development. Spiking Versus native PAH extraction The effectiveness of SFE is often judged by its reproducibility and repeatability. However, all too often data is reported that indicates the superiority of a particular extraction technique or solvent system above another using data based on spiked recoveries of analytes from Celite, sand or soil. This is unfortunate as spiked samples do not reflect the potential of the extraction technique or solvent system to extract analytes from 'real' matrices.For example, Tena et a1.18 reported that the use of a modifier was essential for quantitative PAH extraction from a native soil, but was unnecessary for extraction from silica. There is also a potential problem, in the use of 'real' samples for comparison purposes, because the actual levels of the PAHs in the sample are unknown. The nature of the comparison means that a particular extraction system is being compared with SFE. Frequently this comparison is being done against an approved method, such as, the EPA method for Soxhlet extraction using a particular solvent system. However, while every attempt is made to optimize the SFE procedure, in terms of collection efficiency and operating conditions, no attempt is usually made to optimize the conditions, e.g., choice of solvent and extraction time, for Soxhlet extraction.So, it is possible to observe widely different recoveries for SFE.16927>28,30 For example, recoveries of PAHs from marine sediment (SRM 1941), diesel soot and air particulate matter (SRM 1649) were compared recently by SFE and either certificate values (based on 16 or 48 h Soxhlet extractions) and 14 h Soxhlet extraction.28 The recoveries for the lighter PAHs, e.g., fluoranthene, ranged from 59 to 121% while a heavier PAH, e.g., indeno[ 1,2,3-cd]pyrene, ranged from 10 to 90%. Indeno[1,2,3-cdjpyrene was not determined in the diesel soot. In these siutations, it is perhaps acceptable to use spiked samples to aid method development.However, the nature of the spiking process is important with slurry spiking being the preferred option (as compared to spot spiking).3' In slurry spiking, PAHs are added to a known mass of the sample in excess solvent. During evaporation of the solvent, the PAHs can migrate and interact not only with the sample surface but also within the sample matrix itself. Therefore, specific analyte- matrix interactions may occur allowing some degree of adsorption. Additionally, the sample may be aged by storage under suitable conditions (e.g., 4 "C and in the dark). This approach may be more appropriate for a direct comparison of extraction techniques. Recommendations for Quantitative SFE of PAHs From Environmental Matrices For the quantitative SFE of PAHs from environmental matrices it is therefore necessary to consider both instrumental parame- ters and the nature of the sample.The following points are suggested for practical guidance, with respect to instrumental parameters, most appropriate for quantitative SFE. (i) Use an SFE system with two pumps. (ii) Consider the use of more polar supercritical fluids other (iii) Choose the most appropriate modifier, combined (iv) Smaller particle size important for increasing recovery. (v) Pack the extraction cell appropriately, i.e., use drying (vi) Use a variable restrictor, not a fixed linear restrictor. (vii) Choose an appropriate collection system, i.e., liquid- (viii) Maintain collection solvent at 5 "C. (ix) Extraction time an important variable.(x) Extraction temperature (200 "C) an important variable. (xi) Consider the use of experimental design/chemometrics than COZ. modifier mixture or reactive modifier. agents and copper (for soils with high sulfur content). solid trap and/or collection solvent. for method development. References 1 2 3 4 5 6 7 8 9 10 11 Wild, S. R., and Jones, K. C., Environ. Pollut., 1995, 88, 91. Halsall, C. J., Coleman, P. J., Davis, B. J., Burnett, V., Waterhouse, K. S., Harding-Jones, P., and Jones, K. C., Environ. Sci. Technol., 1994,28, 2380. Wild, S. R., and Jones, K. C., Waste Manag. Res., 1994, 12, 49. Dennis, A. J., Massey, R. C., McWeeny, D. J., and Watson, D. H., Polynuclear Aromatic Hydrocarbons: Seventh International Sympo- sium on Formation, Metabolism and Measurement, ed.Cooke, M. W., and Dennis, A. J., Battelle Press, Columbus, OH, 1982. Bowadt, S., and Hawthome, S. B., J. Chromatogr., 1995, 703, 549. Barnabas, I. J., Dean, J. R., and Owen, S. P., Analyst, 1994, 119, 238 1. Janda, V., Bartle, K. D., and Clifford, A. A., J . Chromatogr., 1993, 642, 283. Greibrokk, T., J. Chromatogr., 1995, 703, 523. Applications of Supercritical Fluids in Industrial Analysis, ed. Dean, J. R., Blackie, Glasgow, 1993. Hawthorne, S. B., and Miller, D. J., Anal. Chem., 1987, 59, 1705. Hawthorne, S. B., Langenfeld, J. J., Miller, D. J., and Burford, M. D., Anal. Chem., 1992, 64, 1614.Analyst, September 1996, Vol. 121 89R 12 13 14 15 16 17 18 19 20 21 22 Howard, A. L., Yoo, W. J., Taylor, L. T., Schweighardt, F. K., Emery, A. P., Chester, S. N., and MacCrehan, W. A., J . Chromatogr. Sci., 1993, 31, 401. Hawthorne, S. B., Yang, Y., and Miller, D. J., Anal. Chem., 1994,66, 2912. Hills, J. W., and Hill, H. H., J . Chromatogr. Sci., 1993, 31, 6. Fahmy, T. M., Paulaitis, M. E., Johnson, D. M., and McNally, M. E. P., Anal. Chem., 1993, 65, 1462. Lee, H. B., Peart, T. E., Hong-You, R. L., and Gere, D. R., J . Chromatogr., 1993, 653, 83. Dankers, J., Groenenboom, M., Scholtis, L. H. A., and van der Heiden, C., J . Chromatogr., 1993, 641, 357. Tena, M. T., Luque de Castro, M. D., and Valcarcel, M., Chromatographia, 1994, 38,43 I . Pyle, S. M., and Setty, M. M., Talanta, 1991, 38, 1125. Burford, M. D., Hawthorne, S.B., Miller, D. J., and Braggins, T., J . Chromatogr., 1992, 609, 321. Langenfeld, J. J., Burford, M. D., Hawthorne, S. B., and Miller, D. J., J . Chromatogr., 1992, 594, 297. Miller, D. J. Hawthorne, S. B., and McNally, M. E. P., Anal. Chern., 1993, 65, 1038. 23 24 25 26 27 28 29 30 31 Meyer, A., and Kleibohmer, W., J. Chromatogr., 1993, 657, 327. Lopez-Avila, V., Dodhiwaia, N. S., and Beckert, W. F., J . Chromatogr. Sci., 1990, 28, 468. Levy, J. M., Dolata, L. A., and Ravey, R. M., J . Chromatogr. Sci., 1993, 31, 349. Langenfeld, J. J., Hawthorne, S. B., Miller, D. J., and Pawliszyn, J., Anal. Chem., 1993, 65, 338. Reindl, S., and Hofler, F., Anal. Chem., 1994, 66, 1808. Yang, Y., Gharaibeh, A., Hawthorne, S. B., and Miller, D. J., Anal. Chem., 1995, 67, 641. Barnabas, I. J., Dean, J. R., Tomlinson, W. R., and Owen, S. P., Anal. Chem., 1995,67, 2064. Dean, J. R., Barnabas, I. J., and Fowlis, I. A., Anal. Proc., 1995, 32, 305. Camel, V., Claude, M., and Tambute, A., J . Chromatogr. Sci., 1995, 33, 123. Paper 6103477B Accepted May 20,1996