12 Analysis of Organic Pollutants in Environmental Samples Julian J. C. Dawson, Helena Maciel, and Graeme I. Paton The University of Aberdeen, Aberdeen, Scotland Kirk T. Semple Lancaster University, Lancaster, England I. INTRODUCTION The identification and quantification of organic pollutants in environmental matrices have improved significantly over the past two decades. Fundamen- tally, the determination of organic contaminants requires selective solvent extraction of the determinant(s) from the matrix followed by quantifiable analysis using specialized instrumentation. Often the researcher needs to identify a target compound and/or its metabolites, thus focusing the choice of technique to suit the particular matrix and determinant(s). Significant advances in instrument automation and reliability, precision of flow control, detector development, and competitive instrument pricing have greatly increased the number and range of laboratories capable of fulfilling reliable and routine organic pollutant analysis. This chapter describes the main steps required in analysis of key organic pollutants in environmental samples, concentrating on soil analysis to provide illustrative examples, as soil is one of the more challenging matrices. Citations are made to references that provide specific information about instrumentation and the underpinning principles and scientific rationale. Several widely used methods are described and discussed in detail to exemplify the considerations needed for techniques. TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. A. Why Quantify and Identify Organic Contaminants? The presence of organic pollutants in the environment is ubiquitous. From the high arctic to the tropics (Jones and de Voogt, 1999), recalcitrant and volatile pollutants are detectable in all environmental spheres. Soils and sediments are major sinks for organic pollutants and can retain the highest concentrations of released pollutants (Northcott and Jones, 2000). Drinking water contaminated with biocides from runoff into surface waters or by the leaching of agrochemicals through soil to aquifers is widely acknowledged (Stackelberg et al., 2001). Researchers and regulators need sensitive and routine techniques to identify and quantify these contaminants. Scientists also need to be able to study samples for signs of degradation and the occurrence of metabolites and cocontaminants that may indicate the relative damage or indeed remediation in soil or sediment systems. II. OVERVIEW OF ORGANIC ANALYSIS Once a representative sample has been obtained, there are three further stages that underpin organic analysis: (1) the preparatory (drying) and extraction stage, (2) the cleanup stage(s) and (3) the determination stage. Some determinations may only be performed after derivatization, when the determinant needs to be chemically altered to improve analytical resolution. Organotin determination, for example, requires extensive derivatization because the determinants are not sufficiently volatile for direct gas chromatographic analysis (Abalos et al., 1997). Each of these stages will be dealt with separately, and using illustrative examples, the selection criteria for certain approaches will be justified. A. Sample Preparation and Analysis The type of drying technique carried out is determined by the nature of the determinant(s) and the matrices. It is usually inappropriate to dry a soil or sediment in an oven as may be done for inorganic analysis, as this may cause a substantial loss of the determinants. Instead, a sulfate salt is often used to remove the water (Hess et al., 1995; Guerin, 1999). After drying, the organic determinant present must be brought into an appropriate organic solvent prior to quantification by gas chromatography (GC) or high-pressure liquid chromatography (HPLC). Determinants in water samples can be extracted using sequential volumes of organic solvent, which are then passed through the sulfate salt to remove residual water (Meharg et al., 1999). The extraction techni que also enables the sensitivity of the analysis to be 516 Dawson et al. TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. manipulated through sample concentration. Depending upon the nature of the sample and the target determinant, an appropriate technique can be selected. 1. Liquid/Liquid Phase Extraction When a solvent is immiscible with water and the target determinant is more soluble in the solvent than in water, then this is an ideal technique. The partitioning coefficient of the determinant material is equal to the ratio of its concentration in the solvent divided by that in water. The partitioning coefficient is independent of the volume ratio of solvent : water but constant at any given temperature; thus increasing the amount of solvent increases the amount of determinant extracted. Repeated extractions with the same solvent will also increase the efficiency of determinant extraction. Extraction efficiency can be further improved by heating of the sample-extraction mixture (Dean and Xiong, 2000). 2. Soxhlet Extraction This is a commonly used technique for quantifying total concentrations of semivolatile and nonvolatile hydrophobic contaminants. A diagram of the main components of the Soxhlet apparatus is shown in Fig. 1. The soil or Figure 1 Soxhlet apparatus for solvent extraction of organic pollutants from soils and sediments. Analysis of Organic Pollutants 517 TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. sediment sample is placed in a porous extraction thimble. Below this thimble is a cup containing the solvent, which is heated and passed through distillation and condensation stages, ensuring that there is a rigorous mixing of the solvent with the sample. Although the procedure is slow, it is a harsh but effective technique. This method continuously re-extracts the sample with the same quantity of solvent: the solvent is refluxed in a Soxhlet apparatus, condenses, and trickles down through the sample back to the bottom of the apparatus, where the determinant collects. This method is used for nonvolatile and semivolatile organics, and it will not collect any compounds with a boiling point lower than, or close to, that of the solvent used. The solvent is typically a mixture of a nonpolar compound such as dichloromethane (DCM) with about 10% of a water-miscible polar solvent such as methanol or acetone added. 3. Supercritical Fluid Extraction (SFE) A supercritical fluid (SCF) is a substance held above its critical temperature and pressure. SCFs have many advantages over liquid solvents for use in extraction of environmental samples (Camel, 2001). Their physical proper- ties include low viscosity, high diffusion coefficients, low toxicity, low flammability, and negligible surface tension. These allow SCFs to penetrate an environmental matrix very quickly, allowing rapid extractions compared to those with conventional solvents. A further advantage is that SFE systems can be interfaced directly with a chromatography column, thus minimizing sample preparation. Supercritical carbon dioxide, possibly modified by the addition of methanol or acetone, is the most common solvent used in environmental analysis; however, a SCF with a dipole moment may be more effective (Alonso et al., 1998). Hawthorne et al. (1992) found that supercritical CHClF 2 (Freon-22) was more effective than CO 2 for the extraction of PAHs and PCBs from soils, consistently extracting 2–10 times more determinant. SFE with CHClF 2 was also fast: 30–45 minutes were required to extract comparable amounts to that obtained by 18 hours ultrasonication in DCM, or 32 hours Soxhlet extraction in hexane/methanol and hexane/acetone mixtures. SCF techniques are not routinely used in soil analysis because they are quite expensive to set up and to run routinely. However, they may be more widely used in future, particularly if they are shown to be applicable to a range of determinants that are not routinely tested for at present. 4. Thermal Desorption This method is used in conjunction with a gas chromatograph (GC) and is suitable for volatile and semivolatile hydrocarbons. The solid sample is 518 Dawson et al. TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. warmed to approximately 85  C in an enclosed system to desorb and vola- tilize the hydrocarbons, which are then purged, trapped, and subsequently transferred onto the column. Volatile organic compounds, such as benzene, toluene, ethylbenzene, xylene (BTEX), methyl tertiary butyl ether (MtBE), and naphthalene from liquid environmental samples, e.g., fresh and marine waters, soil extracts, and wastewater, can also be extracted by purging of the sample using an inert gas and trapping the extracted determinants. The contents of the trap are then injected directly onto the column of the GC. Although slow and costly to set up, the method is the most reliable one for quantifying relatively water-soluble determinants. 5. Solid/Liquid Phase Extraction This technique is also known as solid phase extraction (SPE). The process is simple and fast and may prove cost-effective for some users. A measured volume of the sample is passed through a cartridge tube with a suitable solid material, which sorbs the target determinant. The determinant is then eluted from the cartridge using an appropriate solvent. Semiautomated SPE systems use vacuum pumps to speed up the solvent flow, enabling elution to take place much more quickly. SPE is also extensively used as a cleanup technique to remove material that may damage a chromatography column or slow down the chromatographic procedure. Most of the large chromatographic suppliers sell SPE systems. The selection of the packing material is based upon the polarity of the contaminants to be determined. Nonpolar hydrophobic adsorbents retain the nonpolar determinants and allow polar substances to pass through the column, whereas hydrophilic adsorbents adsorb the polar components, allowing the nonpolar materials to pass through. 6. Other Methods Ultrasonication. An ultrasonic probe may be used to agitate soil or sediment samples in a solvent such as DCM. It is used for non- and semivolatile organics at various concentrations (Guerin, 1999). Accelerated Solvent Extraction (ASE). This uses traditional solvent mixtures as for Soxhlet extraction, but the sample is held at increased temperature and pressure, thus reducing extraction time and solvent volume required (Fisher et al., 1997; Hubert et al., 2001). Microwave-Assisted Solvent Extraction (MSE). This method is similar in function to ASE in that it enables reduced extraction times and solvent volumes compared to traditional techniques. The equipment is quite Analysis of Organic Pollutants 519 TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. costly to purchase, and the technique is not widely reviewed in publications; hence comparative evaluation may prove to be problematic. B. Cleanup Techniques Sample cleanup of organic extracts is used to prolong the life of the instrument column, injector, and detector. A purified sample will also produce clearer peaks with improved resolution that will prove easier to quantify. Sample purification tends to be based on one of the following principles: (1) partitioning between immiscible solvents; (2) adsorption chromatography; (3) gel permeation chromatography; (4) chemical destruc- tion of interfering substances; or (5) distillation. The simplest of the above is the acid/base partitioning, in which a solvent extract is shaken with dilute alkali that enables acidic organics to partition into the aqueous layer while the basic and neutral fractions remain in the organic solvent. The aqueous layer can then be acidified and extracted using DCM so that the organic layer will now contain the acid fraction. This technique is widely used in cleanup procedures for determining phenols and associated herbicides from soils and sediments (Patnaik, 1997). Cleanup columns, either as premanufactured SPE systems or as laboratory-produced columns, are the most common routine technique for cleanup. For example, highly porous and granular aluminum oxide (alumina) can be used and is readily available in acidic, neutral, or basic forms (Polese et al., 1996). Target determinants can be differentiated by chemical polarity. After the column is packed with the granular material it is covered in anhydrous sodium sulfate and the sample is placed on the column. By using the appropriate solvent, this enables the determinants to be separated from impurities that are present. Basic alumina is used in purification of steroids, alcohols, and pigments (Cho et al., 1997); the neutral form is used for esters and ketones (Polese and Ribeiro, 1998), while the acidic form separates strong acids and acidic pigments. Alumina is also ideal for the cleanup of hydrocarbons (Cho et al., 1997; Shen and Jaffe, 2000). Amorphous silica gel is suitable for the removal of interfering compounds of differing polarities (Shamsipur et al., 2000). Activated silica gel is heated for several hours at 150  C prior to use and is also well suited for the cleanup of hydrocarbons (Miege et al., 1999). Deactivated silica gel has significantly more water present and is used to separate plasticizers, lipids, esters, and some organometallic compounds (Shamsipur et al., 2000). If used appropriately, high specificity for target herbicides can be achieved. In addition, the selection of different solvents (Supelco, 2001) can be used to manipulate adsorbent activity of the SPE system. 520 Dawson et al. TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. Florisil is a form of magnesium silicate with acidic properties. A packed column of Florisil is used the same way as silica and alumina columns. The material is ideal for the separation of aliphatic compounds from aromatics (Abdallah, 1994) and is used for a wide range of pesticides (Smeds and Saukko, 2001) and halogenated hydrocarbons (Schenck and Donoghue, 2000). Gel permeation is able to differentiate material on the basis of pore size using hydrophobic gels (Knothe, 2001). As with SPE, this system is capable of performing to a high level of specificity, though equipment and consumable costs will reflect this. In some solid environmental samples, the presence of specific materials may impose analytical problems. For example, sulfur may reduce the resolution of chromatograms. Sulfur has a solubility that is similar to a range of organochlorine and organophosphate pesticides and cannot be resolved using Florisil (Patnaik, 1997). Commonly, copper turnings are shaken with the sample to remove sulfur from the solvent extract (Schulz et al., 1989). Mercury or tetrabutyl ammonium sulfite (Duinker et al., 1991) are also used. Table 1 describes the materials typically chosen for cleanup procedures of selected contaminants extracted from soils and sediments. C. Methods of Determination Chromatography is a simple concept in that analyte components become separated as they either move in the mobile phase or become sorbed in another phase. The characteristics of the sorption phases determine the extent to which analyte components become separated. The resolution can be manipulated by using appropriate columns in consideration of the determinants sought. The major factors to ensure high quality chromatog- raphy are (1) purity of the mobile phase, (2) a reliable flow rate, (3) an Table 1 Suggested Cleanup Techniques for a Number of Common Contaminant Groups Determinant Cleanup technique Nitrosamines Gel permeation, alumina, Florisil Organochlorines Gel permeation, Florisil Organophosphates Gel permeation, Florisil Petroleum Alumina, acid–base Phenols Gel permeation, acid–base, silica gel PAHs Gel permeation, alumina, silica gel Source: Patnaik, 1997. Analysis of Organic Pollutants 521 TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. appropriate column, and (4) a suitably sensitive detector. Regardless of the type of chromatography, these rules must be adhered to. The most commonly used chromatographic techniques in environmental analysis are GC and HPLC, and these methods will be described briefly and then consi- dered in more detail, using representative case studies, later in this chapter. For routine analysis it is important to consider the value of an autosampler. Current microrobotic technology provides high precision and reproducibility. In many instruments, sample vials can undergo heatin g and mixing (with slight modifications to the sampler), thus enabling some automated derivatization. Automated dilution systems where available, are also very useful, as the system is capable of operating with small volumes. The automated injection system resolves problems associated with manual techniques, which may cause excessive and variable peak broadening on the column. Most significantly, the autosampler allows hundreds of samples to be systematically analyzed. This is ideal, because of the long retention times associated with some determinations. 1. Gas Chromatography Traditionally this has been called gas–liquid chromatography because samples being carried through a column undergo partition between a gas phase (mobile) and a sorbed liquid phase (stationary). For the purpose of this chapter, only capillary GC will be considered, but further details on packed columns can be found in Bruno (2000), and in Chap. 10. The main components of a GC are The Injector. After sample preparation and cleanup, the sample is ready for injection. Most GC analysis will be carried out using split or splitless injection. This means that the sample is injected into a chamber where, under heating, it expands and then moves in the gas flow onto the column. The selection of the solvent used for injection is therefore very important, as different solvents have different expansion characteristics. In the case of split injection, a proportion of the sample is discarded, as it may overload the column and detector and cause a reduction in resolution. Common split ratios are betw een 15 : 1 and 40 : 1, and thus a large proportion of the sample is discarded. Splitless analysis, on the other hand, enables expansion of the solvent vapor within a glass liner, but the entire sample is presented to the column. On-column injection is required for trace analysis and has no pre-expansion stage for the sample. The injector systems are usually tailor-made to suit the style of analysis. Gas Flow Selection and Rate. Many laboratories using gas chroma- tography do not have access to high-purity gases and thus have to use 522 Dawson et al. TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. supplies containing small amounts of impurities, e.g., oxygen, moisture, and carbon compounds. In such circumstances, filters should be used to remove these impurities, to avoid damaging the column and affecting the response of the detector. A carrier gas ensures steady flow of sample through the column while often an additional ‘‘make-up’’ gas is required for the detector. Any new GC will have highly sensitive electronic or manual gas controls, which can be altered according to column-specific requirements. The Capillary Column. The nature of the column will determine the success or failure of the separation. Users should be aware of the range of columns on the market and the relative merits of inexpensive and expensive purchases. The selection of a column is governed by what is referred to as the ‘‘theoretical plates per meter’’ concept. This parameter describes the chromatographic performance of a column. There is a wide range of texts that consider the principles that underpin this parameter, and for more information, Marr and Cresser (1983) is a good source. All the major capillary column suppliers have catalogs either available in paper format or from the internet. These should be consulted prior to purchase, as they will enable the most appropriate column to be selected. The columns are composed of fused silica, and a narrow-bore inside diameter (i.d.) (usually 0.20, 0.25, or 0.32 mm) will provide the best separation for closely eluting components and isomers. In general, the smaller the i.d., the greater is the level of resolution that can be achieved. Conversely, to avoid sample overload for analytes in high concentrations, a larger i.d. may be more appropriate. The characteristics of some typical columns are shown in Table 2. The Oven Control. The column will have been selec ted to favor the particular determinant(s) and analytes. However, it is possible to alter the analysis most effectively by the manipulation of temperature. For determinants to be separated, they are differentially partitioned between the mobile and stationary phases: the proportion in the gas phase depends Table 2 Examples of Some Available Columns and Their Characteristics Column type Temperature Polarity Dimethyl silicone oil 0–350  C Very low Phenyl methyl silicone oil 0–350  C Low Phenyl cyanopropyl methyl silicone 0–275  C Medium Carbowax 1540 0–200  C High Analysis of Organic Pollutants 523 TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. on temperature. When analyzing a broad suite of hydrocarbons, for example, it is not possible to select a column that is capable of discriminating between all determinants. The most volatile fraction will move rapidly along the column, while the larger molecules will trail significantly behind. To speed up this process, the oven can be adjusted to produce a temperature ‘‘ramp’’ during which the column temperature changes across a predetermined regime. This simply means that as the analysis is progressing, the oven temperature is progressively raised, which means that the sample begins to reach ‘‘vapor pressure’’ and elutes more readily through the column. Without the use of this ramp, retention time would rise significantly if the temperature were set too low, whereas if the temperature were initially set too high, all the determinants would elute together. The Detector. Thermal conductivity detectors (TCDs) and flame ionization detectors (FIDs) are the most commonly used types. Because of its lack of specificity, the TCD is more appropriate for gas analysis (see Chap. 10), and it will not be considered in more detail here. The FID, however, is an excellent detector for a wide range of determinants because it responds to the presence of organic carbon compounds (but not to CO, CO 2 ,orCS 2 ). In the FID, the passage of the organic compounds through a hydrogen-rich flame results in the creation of ions and a corresponding electrical response. The FID is sensitive at the mgL À1 level to a plethora of compounds (Marr and Cresser, 1983). It is also a very forgiving detector, as it has a linear response to concentration over seven orders of magnitude and is resistant to overload and damage. Flame photometric detectors (FPDs) can be used to measure determinants containing specific groups, including organic S, P, and Sn compounds (Singh et al., 1996). The FPD has a range of filters to suit the optical emission characteristics of the target determinants. The halogen-specific detector or the electron capture detector (ECD) is an essential detector for the measurement of trace levels of organochlorine compounds (Schulz et al., 1989). The most significant detector used for routine analysis now is the mass spectrometer. This is an excellent tool for identifying a range of unknown determinants in the target matrix. Over the last decade the application of this detector in water, soil, and sediment analysis has grown enormously, and as a consequence the cost has dropped. After separation of components in an appropriate column, the eluted fractions are subjected to electron impact or chemical ionization. The fragmented and molecular ions are resolved from characteristic mass spectra and determinants identified from their distinctive primary and secondary ions. Quantification is achieved by peak height, representing the total ion count, at each specific mass : charge 524 Dawson et al. TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. [...]... Doliva, and J Kirchhoff 1999 Vanillin content and several ratios of components—influence of different extraction parameters of the analysis results Deut Lebensm.-Rundsch 5 :123 129 Eriksson, M., A Swartling, and G Dalhammar 1998 Biological degradation of diesel fuel in water and soil monitored with solid-phase micro-extraction and GC-MS Appl Microbiol Biotechnol 50 :129 –134 Fisher, J.A., M.J Scarlett, and A.D... Shamsipur, M., F Raoufi, and H Sharghi 2000 Solid phase extraction and determination of lead in soil and water samples using octadecyl silica membrane disks modified by bis[1-hydroxy-9,10-anthraquinone-2-methyl]sulfide and flame atomic absorption spectrometry Talanta 52:637–643 Shen, L., and R Jaffe 2000 Interactions between dissolved petroleum hydrocarbons and pure and humic acid-coated mineral surfaces... are in soils Physical effects, such as migration through soil, could also alter isotope ratios B Determination of Readily Extractable Chlorophenols and Total Chlorophenols in Soil Using HPLC 1 Readily Extractable Chlorophenols In this study, 2,4-dichlorophenol (2,4-DCP), 2,4,6-trichlorophenol (2,4,6TCP), pentachlorophenol (PCP), and 2,4-dibromophenol (2,4-DBP) are the target determinants, and 2,4-dibromophenol... extraction and microwave-assisted extraction: their potential and pitfalls Analyst 126 :1182–1193 Cho, B.H., H Chino, H Tsuji, T Kunito, H Makishima, H Uchida, S Matsumoto, and Oyaizu, H 1997 Analysis of oil components and hydrocarbon-utilizing microorganisms during laboratory-scale bioremediation of oil-contaminated soil of Kuwait Chemosphere 35:1613–1621 Choudhary, G., A Apffel, H.F Yin, and W Hancock... residues in soil and sediment Environ Pollut 108:19–43 Patnaik, P 1997 Handbook of Environmental Analysis: Chemical Pollutants in Air, Water, Soil, and Solid Wastes Lewis Publishers, CRC Press, New York Polese, L., E.V Minelli, E.F.G Jardim, and M.L Ribeiro 1996 Small scale method for the determination of selected organochlorine pesticides in soil Fresen J Anal Chem 354:474–476 Polese, L., and M.L Ribeiro... Harvey, D 2000 Modern Analytical Chemistry McGraw-Hill, Boston Hess, P., J de Boer, W.P Cofino, P.E.G Leonards, and D.E Wells 1995 Critical review of the analysis of non- and mono-ortho-chlorobiphenyls J Chromatogr A 703:417–465 Hawthorne, S.B., J.J Langfield, D.J Miller, and M.D Burford 1992 Comparison of supercritical CHClF2, N2O, and CO2 for the extraction of polychlorinatedbiphenyls and polycyclic... rapid rise in temperature makes possible an on-column injection of the VOCs, and standard chromatographic analysis For soil analysis, instead of using a water extract, a soil sample can be placed in a glass liner and purged directly on the instrument The PTI-FID-GC system requires H2 (fuel gas), air (combustion/ oxidizing gas), N2 (makeup/preflush/backflush gas) and He (purging/carrier gas) The PTI has a... target determinants, and 2,4-dibromophenol is used as an internal standard Soil samples are sieved through a 2 mm sieve and stored at 4 C A 5 g soil subsample (based on dry soil) is weighed and extracted with water or a mixture of methanol and water with a ratio of 1 : 1 The soil extract is cleaned up by SPE in order to cleanup the sample and preconcentrate the chlorophenols 100 mg Bond Elut C18 (1 mL... then passed through a column composed of Florosil and acid-rinsed copper powder At this stage, an internal standard such as p-chlorobiphenyl or tetrachloro-m-xylene can be added to the sample 2 Analysis The instrument used is a GC equipped with an ECD and a selective column The columns used are either a CP-Sil5 CB fused silica WCOT or a Phenomenex ZB-5 (5% phenyl polysiloxane) capillary column (60 m... and A.L Baehr 2001 Frequently co-occurring pesticides and volatile organic compounds in public supply and monitoring wells, southern New Jersey, USA Environ Toxicol Chem 20:853–865 Supelco 2001 Technical and Application Notes http://www.sigma-aldrich.com Whittaker, M., S.J.T Pollard, and T.E Fallick 1995 Characterisation of refractory wastes at heavy oil-contaminated sites: review of conventional and . and 2,4-dibromophenol (2,4-DBP) are the target determinants, and 2,4-dibromophenol is used as an internal standard. Soil samples are sieved through a 2 mm sieve and stored at 4  C. A 5 g soil subsample. Chlorophenols and Total Chlorophenols in Soil Using HPLC 1. Readily Extractable Chlorophenols In this study, 2,4-dichlorophenol (2,4-DCP), 2,4,6-trichlorophenol (2,4, 6- TCP), pentachlorophenol (PCP), and. (MtBE), and naphthalene from liquid environmental samples, e.g., fresh and marine waters, soil extracts, and wastewater, can also be extracted by purging of the sample using an inert gas and trapping