Introduction

When one thinks of the higher analogs of the simple, six-carbon benzene ring, familiar images of multiple-fused rings are often envisaged and the term “polycyclic aromatic hydrocarbons” (PAHs) comes to mind. PAHs, or polyarenes, constitute a large class of organic molecules that likely exhibits the most structural variety in nature relative to any other class of non-halogenated molecules in the eco- and biosphere. PAHs are ubiquitous environmental contaminants derived principally from combustion of fossil fuels in heat and power generation, refuse burning, coke ovens, and motor vehicle operation [1–3]. Although three PAHs—acenaphthene, acenaphthylene, and anthracene—are produced commercially in the United States in quantities greater than research level, commercial production is not regarded as a significant source of PAHs in the environment [4]. Open burning of biomass and fuel pools also emits PAHs, and PAH emissions have been noted in plumes from the combustion of polymers [5, 6]. Natural emission sources include forest fires, volcanoes, and hydrothermal processes [7–9]. Some PAHs are relatively potent carcinogens, and biological and mutagenic effects are well documented [10]. Given the right conditions, PAHs can persist in the environment and accumulate, for example in anaerobic sediments, to the extent that the potential for adverse health effects is high. As a result, characterization of PAHs in environmental compartments has been an important focus for decades. PAHs have been monitored in the marine environment since the 1960s and this continues today in local, regional, and national-scale programs [11, 12]. Major long-term monitoring programs, for example the Arctic Monitoring and Assessment Program (AMAP), the Baltic Marine Environment Protection Commission (Helsinki Commission), and the National Status and Trends Program (US), focus on PAHs in marine biota and sediment as part of their strategies to assess marine environmental quality [12]. PAHs are also frequently measured in the atmosphere for air-quality assessment, in biological tissues for health-effects monitoring, and in foodstuffs for safety reasons.

Measurements of PAHs in environmental matrices tend to require difficult analytical chemistry procedures, largely because of the extreme complexity of environmental samples. This complexity is realized when one considers the general categories of phases into which environmental samples may be categorized; these include aqueous, air (gaseous or condensates/particulate matter), oil or organic liquids, solids or sludges, biological samples, and even multiphase samples. Multiphase samples may be a combination of these, and typically require some sort of phase separation, for example filtration, before the start of the analytical process. In addition, numerous structural isomers are often present in environmental samples, and in some matrices, for example fuel products or emissions, there may be heteroatoms (nitrogen, sulfur, oxygen) [13–16]. Hence, analytical methods must include processes for isolation of compounds that are part of complex phases, and separation and detection techniques for multi-component mixtures that consist of compounds with a wide range of polarities, volatilities, and molecular sizes and shapes. Chromatographic approaches offer avenues to address these elements and play a critical role in the determination of PAHs in environmental samples.

Chromatographic methods for PAH analyses in environmental media have been developed and evaluated extensively over the past few decades, as reviewed in the literature [11, 17–19]. Liquid and/or gas chromatography (LC and GC, respectively) are prominent techniques in US Environmental Protection Agency (EPA) PAH-related test methods, including (although not limited to): drinking water [20, 21]; municipal and industrial discharges [22–25]; soils, sludges, and solid waste [26]; ambient air [27]; and shellfish tissue [28] (Table 1). Many standard analytical methods approved by other US federal agencies, for example the National Institute for Occupational Safety and Health (NIOSH), and non-government organizations, for example the Association of Official Analytical Chemists (AOAC) and the American Public Health Association (APHA), make use of both LC and GC for the determination of PAHs. These methods have been compiled and reviewed by the US Agency for Toxic Substances and Disease Registry in a PAH toxicological profile [4]. Similarly, analytical methods for trace organic contaminants, including PAHs, used by laboratories associated with the National Status and Trends Program National Benthic Surveillance and Mussel-Watch Projects have LC and GC components, and these have been compiled by the National Ocean and Atmospheric Administration [29, 30]. Several international standard methods also stipulate use of LC and GC for determination of PAHs (Table 2), as well as European national-level reference methods for PAHs in air (France, Germany, and Italy [31]).

Table 1 Selected US Environmental Protection Agency test methods for determination of PAHsa in environmental media
Full size table
Table 2 Standard methods for the determination of PAHsa in environmental media
Full size table

Each technique, LC or GC, offers unique information or has unique aspects. For example, the GC procedure in EPA Method 610 does not adequately resolve four of the sixteen PAHs targeted for measurement (Table 3)—anthracene and phenanthrene; chrysene and benz[a]anthracene; benzo[b] and benzo[k]fluoranthene; and dibenz[a,h]anthracene and indeno[1,2,3-cd]pyrene. If, however, the purpose of the analysis is not served by the determination of the sum of one of these unresolved pairs, the LC method in EPA Method 610, which resolves all 16 PAHs, can be used [23]. In contrast, EPA Method 625 [24] only provides gas chromatographic–mass spectrometric (GC/MS) conditions appropriate for qualitative and quantitative confirmation of results for the PAHs listed in Table 3, with many other compounds (base/neutrals and acids), in the extract produced by EPA Method 610.

Table 3 Selected PAHs reported for environmental and health-effects studies
Full size table

The scope and application of EPA Method 610 applies to 16 PAHs, commonly referred to as the EPA 16 priority pollutants (Table 3). This list has evolved somewhat since the development of EPA Method 610, for example the list includes two additional analytes in EPA Method TO-13A for the analysis of air samples (Table 3). Moreover, the 16 PAHs listed for EPA Method 610 are often grouped with a wider range of semivolatile organic compounds or base/neutral extractable compounds in other environmental test methods (e.g. EPA Method 625 [24]), and are often the “common” range of analytes for national and international environmental/health studies. In a review on approaches to consider for the risk assessment of PAHs, the World Health Organization (WHO) lists 17 individual compounds over and above the 16 PAHs in EPA Method 610 [32] (Table 3), a total of 31 parent PAH plus two alkyl derivatives, and AMAP has an additional 23 groups of alkylated PAHs. (See a review of marine monitoring programs [12]). The WHO provides a summary of genotoxicity and carcinogenicity results for the 33 PAHs (reviewed in Ref. [32], Table 3). The International Agency for Research on Cancer (IARC) has identified a subset of these PAHs as probable (group 2A) or possible (group 2B) human carcinogens (Table 3) [33]. The Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR) lists up to 46 PAH compounds (or mixtures such as anthracene oil and different distillates) as Substances for Possible Concern (http://www.ospar.org).

The range of isomers to be examined or measured in any environmental survey will ultimately depend on its purpose (e.g. health-related, ecotoxicological, source inventory) [32], and often only specific PAHs are targeted, or PAHs are included as a class/group. For example, only four to six PAHs are typically used as markers of emissions or emission inventory compilations (Table 3). The persistent organic pollutant (POPs) protocol under the United Nations Economic Commission for Europe’s Convention on Long Range Transboundary Air Pollution requires that emissions of only four PAH compounds have to be reported annually by European Union member states [31, 34]. For the assessment of human exposure to air pollutants in European cities, benzene was the only major aromatic compound measured in six cities, although in selected campaigns measurements were extended to include several PAHs, particulate matter, and heavy metals [35]. The European Union regulated pollutant list for the protection of human health includes only one PAH, benzo[a]pyrene (see Table 1 in Ref. [35] and a paper by Menichini et al. [36]). Similarly, the Great Lakes Binational Toxics Strategy [37] between Canada and the US lists only benzo[a]pyrene as the PAH under the category “Level 1 substance” [38] (a primary focus for the virtual elimination of the substance), although PAHs as a group, with selected specific PAHs (anthracene, benz[a]anthracene, benzo[ghi]perylene, perylene, and phenanthrene) are listed in the next category (Level 2 [39]). Level 2 compounds are those that have been identified by one or both countries as having the potential to cause a significant impact on the Great Lakes ecosystem.

At the National Institute of Standards and Technology (NIST), EPA Method 610 PAHs (sixteen, Table 1) plus many other PAHs (Table 4) are measured in environmental reference materials [40], and coupled LC and GC methods have been long standing parts of the analytical methods (Fig. 1) [18, 41–45]. The use of multiple methods of analysis provides examination of a wider range of analytes than can be determined by a single method and also provides confirmation of measurements made by other methods [44]. For example, at low concentrations anthracene and perylene are best measured by use of LC coupled with fluorescence detection (LC–FL) because of their selective and sensitive fluorescence-detection characteristics [41]. This approach is also well suited for the determination of benzo[a]pyrene [46], often the only target analyte in environmental surveys (see discussion above). In contrast, GC/MS provides more accurate results than LC–FL for the determination of benzo[ghi]perylene because of its inherently low fluorescence sensitivity [41]. LC is also a useful fractionation technique for isolation of PAHs for subsequent analysis by other chromatographic and spectroscopic techniques, either in discrete fractions based on the number of aromatic rings or as a total PAH fraction [18, 47, 48]. LC–FL determinations of PAHs in LC fractions based on the number of aromatic rings and in a total PAH fraction were both part of the analytical scheme for value assignment of PAH concentrations in SRM 1941b Organics in Marine Sediment (Fig. 1 [45]). This fractionation approach is particularly suitable for high-molecular-mass PAHs (>300 amu) which are typically present at low concentrations in environmental samples and have many isomers [49, 50].

Table 4 Number of PAHsa recently measured in selected environmental Standard Reference Materials (SRMs)
Full size table
Fig. 1
figure 1

Analytical scheme for the determination of PAHs in SRM 1941b Organics in Marine Sediment [45]

Full size image

Obviously, the use of multiple methods of analysis, for example those shown in Fig. 1, for the determination of PAHs in environmental samples is often not possible for many laboratories, because the approach is not only costly and intense in terms of effort, but samples may be limited or even rare (e.g. an air particulate sample from the Arctic or a shipboard water sample from one cruise). A single method is therefore usually selected on the basis of the matrix (e.g. simple or complex), the target analytes (e.g. individual isomers or mixture), and sensitivity requirements (e.g. screening or quantification needs, expected concentration range). In contemporary analyses of environmental samples for the determination of PAHs, following carefully tailored matrix extraction and clean-up approaches (see examples in references on compilations of analytical methods [4, 29, 30], Tables 1 and 2, and Fig. 1), capillary GC, rather than LC, is usually the preferred analytical technique, for reasons discussed below. Incidentally, the term “capillary” typically refers to columns with inner diameters ranging from 0.18 mm to 0.32 mm whereas microbore inner diameters range from 0.05 mm to 0.10 mm and megabore inner diameters range from 0.45 mm to 0.53 mm. The stationary phase is most often a uniform film affixed to the inner wall of the column (other configurations exist although the uniform thin film is the most widely used) [51], and lengths usually range from 15 m to 60 m for environmental applications. Several active test methods call for the use of packed columns, which are shorter (2 m to 5 m) and wider (1 mm to 5 mm) than capillary columns, and are filled with an inert granular support, each particle which is coated with the stationary phase [51]. For example, EPA Method 610 recommends a 1.8 m × 2 mm glass column packed with 3% OV-17 (50% phenyl methylsilicone; Table 1.9 in Ref. [52]) on Chromosorb W-AW-DCMS (100/120 mesh) or equivalent, although capillary columns can be used if the relative standard deviations of responses for replicate injections are demonstrated to be less than 6% and other quality-control requirements are met [23]. EPA Method 8100 describes the use of both packed and capillary columns for evaluation of solid waste [26]. In contrast, only capillary columns are part of test methods for the determination of PAHs in ambient air (EPA Method TO-13A [27] and ASTM D 6209-98) and drinking water (EPA Method 525.2), and in environmental methods provided by the International Organization for Standardization (ISO) (Tables 1 and 2). Packed columns are rarely used in contemporary GC methods for PAHs. Even more than 20 years ago it was recognized that packed column GC would be rapidly replaced with “open tubular” (now denoted “capillary”) columns [52], on which separations were noted as “vastly superior” (see Refs. in [51]). Nonetheless, ASTM International has active practice guides for both packed [53] and open-tubular capillary [54] columns.

Fig. 2
figure 2

Comparison of GC/MS single-ion chromatograms for PAHs of molecular mass 302 on two different GC stationary phases for SRM 2585 Organic Contaminants in House Dust. GC conditions for the 50% phenyl methylpolysiloxane column (DB-17MS, J&W Scientific, 60 m × 0.25 mm × 0.25 μm): 60 °C for 2 min, heat to 150 °C at 40° min−1, hold for 3 min, heat to 300 °C at 2° min−1, and hold for 120 min; on-column injection with a 5-m retention gap; carrier gas helium in the constant-pressure mode (33.4 psig). GC conditions for the dimethyl 50% liquid-crystalline polysiloxane column (LC-50, J&K Environmental, 15 m × 0.25 mm × 0.25 μm): 60 °C for 1 min, heat to 150 °C at 40° min−1, hold for 2 min, heat to 270 °C at 1.5° min−1, and a final hold for 220 min; on-column injection with a 5 m retention gap; carrier gas helium in the constant flow mode (1.8 mL min−1). Peak identities are: #1 dibenzo[b,e]fluoranthene; #3 naphtho[1,2-b]fluoranthene; #4+5 naphtho[1,2-k]- and naphtho[2,3-j]fluoranthene; #7 naphtho[2,3-b]fluoranthene; #8 dibenzo[b,k]fluoranthene; #9 dibenzo[a,k]fluoranthene; #10 dibenzo[j,l]fluoranthene; #11 naphtho[1,2-e]pyrene; #12 dibenzo[a,l]pyrene; #14+15 naphtho[2,3-k]fluoranthene and naphtho[1,2-a]pyrene; #17 naphtho[2,3-e]pyrene; #18 dibenzo[a,e]pyrene; #20 naphtho[2,1-a]pyrene, #21 dibenzo[e,l]pyrene; #22 naphtho[2,3-a]pyrene; # 23 benzo[b]perylene; # 24 dibenzo[a,i]pyrene; # 25 dibenzo[a,h]pyrene (unknowns: #2, A, #6, #13, #19; Cor=coronene)

Full size image

The popularity of capillary GC for the determination of PAHs is based on a favorable combination of greater selectivity, resolution, and sensitivity compared to that achieved with LC [18, 19, 55]. In capillary GC of complex samples such as carbon black, coal tars, and shale oils it is not uncommon to resolve hundreds of components [17] whereas in LC a practical limit is a few dozen components, because of the limited peak capacity of LC columns. The easy use and compatibility of GC with mass spectrometers (GC/MS) are additional reasons for selection of GC in preference to LC for the determination of PAHs in environmental samples [17]. Other advantages are that PAHs tend to have thermal properties amenable to GC, and MS techniques are fairly sensitive due to the projection of large molecular ion peaks or little fragmentation in common MS sources. The ongoing development of enhancements to GC for the determination of PAHs, for example large-volume injection, thermal desorption, and the coupling of GC to fast detectors, such as time-of-flight MS (TOF-MS), in which full range mass spectral acquisition rates on the order of hundreds of spectra per second enables a significant reduction in chromatographic analysis time, demonstrates the vigor and robustness of the technique. In addition, GC is usually the technique most often recommended in environmental test methods for the determination of PAHs (discussed above; Tables 1 and 2) and LC is typically more often used for clean-up or fractionation (as described previously) [29, 30] before GC analysis. Use of LC, as HPLC, is still stipulated in some test methods (Tables 1 and 2). For example, EPA Method 8310 makes use of HPLC for the determination of PAHs in ground water, soils, sludges, and non-water-miscible waste [26]. However, the decreasing popularity of LC for the determination of PAHs in environmental samples is reflected by actions such as the withdrawal of method ASTM D4657-92 (Standard Test Method for PAHs in Water, Table 1), which described HPLC methods for 16 PAHs in water and wastewater. The method was withdrawn in January 2005, due to lack of use and support to review and update the method.

As GC is so widely used for the determination of PAHs in environmental samples, this paper reviews contemporary state-of-the-art environmental PAH GC applications. An examination of modern GC separations of PAHs is followed by presentation on the properties and uses of selected mass spectrometric techniques. PAH literature on the use of GC with MS, including chemical ionization MS, ion trap MS, TOF-MS, and isotope-ratio MS (IRMS), is discussed. Enhancements to GC (large-volume injection, thermal desorption, fast GC and LC–GC) are reviewed. Purge-and-trap GC is not discussed, because this technique is not widely used for the determination of PAHs—it is more suited to the determination of volatile organic compounds (VOCs) [56, 57]. The determination of VOCs in ambient air is reviewed in this issue by Wang [58]. Relevant discussions pertinent to GC and PAHs, although beyond the scope of this review, are the use two-dimensional GC (GC×GC) and solid-phase micro-extraction (SPME), and the determination of PAH metabolites, nitro-PAHs, and polycyclic aromatic sulfur heterocycles (PASH). These topics are covered in this issue by Gorecki (GC×GC) [59], Ouyang and Pawliszyn (SPME) [60], Budzinski (PAH metabolites) [61], Zielinska (nitro-PAHs) [62], and Andersson et al. (PASH) [63].

Gas chromatography and PAHs

In 2002, TrAC—Trends in Analytical Chemistry released an issue devoted to GC which provides detailed reviews on the history, principles, and applications of GC [64]. Environmental analyses were reviewed by Santos and Galceran, and the use of GC for the determination of selected families of environmental contaminants is discussed, although this is somewhat limited with respect to PAHs [55]. Capillary GC is one of the most widely used and successful chromatographic techniques for the determination of the concentrations of PAHs in environmental matrices, owing mainly to its high resolving power. In the GC process the stationary phase is a thin immobilized film confined to a column that is continuously swept by a stream of mobile phase (carrier gas) [51]. A compound subjected to this process partitions between the stationary and mobile phases and separations occur on the basis of the fraction of time the compound spends in the stationary phase relative to its total transit time through the column. Component separation, i.e. resolution, is strongly related to the optimization of chromatographic conditions [18]. Berezkin and Viktorova have reviewed how important chromatographic conditions, for example column length and diameter, film thickness, stationary liquid phases, and separation temperature mode, have evolved over the period 1970–2000 [65]. The retention and separation of PAHs is affected by such conditions as solvent type and amount, solvent effects, injection conditions (speed, liner size, sample size, temperature), and temperature programming [66–70]. Capillary columns are often used in conjunction with a retention gap—a piece of empty (i.e. uncoated), deactivated, capillary glass tubing (approx. 0.25 mm × 5 m). The retention gap is used to remove non-volatile compounds, minimizing contamination of the column, and to reconcentrate analytes, thus improving peak shape [51]. Cold on-column injection is preferred for PAH analysis, because this improves the resolution of low-molecular-mass PAHs (i.e. early eluting compounds) and reduces discrimination of high-molecular-mass PAHs. Such biases are difficult to avoid when using splitless injection [11].

PAH separations by capillary GC

The most critical aspect of GC method development is selection of an appropriate stationary phase for a specific PAH separation problem [42–45]. Methyl and phenyl-substituted polysiloxanes are the most widely used capillary column stationary phases for separation of PAHs in environmental samples. Columns prepared with polysiloxane stationary phases give relatively low background from column bleed, even at high temperatures (>300 °C) and with nonselective detectors such as the flame ionization detector (FID). Methylpolysiloxane phases containing 0%, 5%, 50%, and even 65% phenyl substitution are commercially available. Method TO-13A (Table 1) for GC/MS determination of PAHs in ambient air calls for use of a 5% phenyl methylpolysiloxane column (30 m × 0.32 mm i.d., 1.0 μm film thickness), although other columns may be used if specific criteria are met. This is the general recommendation in most standard test methods (Tables 1 and 2). Proprietary phases with unique selectivity for selected compounds are also available (DB-XLB). Columns containing liquid-crystalline stationary phases have shape selectivity aspects that are well suited to the separation of PAH isomers.

The shape selectivity of stationary phases used for gas chromatography of PAHs has been extensively studied [71, 72], and PAH separations using liquid-crystalline columns have been reviewed [18]. Liquid-crystalline columns have been part of the NIST analytical scheme for the determination of PAHs in environmental SRMs [73–75] including selected mussel tissue and marine sediment SRMs [44, 45]. Three important pairs of PAHs that typically coelute on traditional 5% phenyl methylpolysiloxane columns (chrysene and triphenylene; benzo[b]- and benzo[j]fluoranthene; and dibenz[a,c]- and dibenz[a,h]anthracene) are separated by liquid-crystalline columns [18, 76]. The greater selectivity of liquid-crystalline columns has also been used for the separation of methyl-substituted PAHs [18]. Liquid-crystalline columns have some limitations, including variations in selectivity, changes in the order of elution of PAHs among different columns, and a limited temperature range [72]. The relatively low temperature limit of early developed liquid-crystalline columns resulted in a limited mass range for the determination of PAHs. More recently, a liquid-crystalline column in the form of a 50% dimethyl 50% (mole fraction) liquid-crystalline polysiloxane phase has become commercially available and has a greater operating temperature range than previous smectic versions [77]. Even the higher-molecular-mass PAHs, for example dibenz[a,c]- and dibenz[a,h]anthracene, coronene, and the six-ring C24H14 PAHs (molecular mass 302) can be determined by use of this new liquid-crystalline phase [78]. Many higher-molecular-mass PAHs (molecular mass >300) give a positive mutagenic response when isolated from environmental and combustion-related samples [78], and several have been identified by IARC as possible human carcinogens (Table 3). It is, therefore, of great interest to develop analytical methods enabling accurate isolation, separation, and quantification of this group of PAHs. This novel liquid-crystalline column, which was recently used as part of the analytical scheme for the determination of PAHs in a house dust SRM (SRM 2585 Organic Contaminants in House Dust) and a diesel particulate SRM (SRM 1650b Diesel Particulate Matter), has some advantages for separation of the C24H14 PAHs compared with use of a 50% phenyl methylpolysiloxane column (Fig. 2). For example, dibenzo[e,l]pyrene (#21) and naphtho[2,1-a]pyrene (#20) in a house dust sample are completely separated on the liquid-crystalline column, as are benzo[b]perylene (#23) and dibenzo[a,i]pyrene (#24). These compounds are poorly separated on the 50% phenyl methylpolysiloxane column (Fig. 2). It should be noted that long retention times are usually necessary to obtain the resolution needed for the accurate quantification of the higher-molecular-mass PAHs, although use of fast GC (discussed below) may lead to shorter analysis times [79].

Ultimately, the use of different GC columns permits the determination of PAHs that may typically coelute on one GC column. For example, chrysene and triphenylene, which coelute on a 50% phenyl methylpolysiloxane column, separate on a DB-XLB column and a liquid-crystalline column, although the liquid-crystalline column has the greater selectivity for the two isomers (Fig. 3). Other PAHs, for example benzo[b]- and benzo[j]fluoranthene, and dibenz[a,c]- and dibenz[a,h]anthracene, also separate on a liquid crystalline column [76], although benzo[a]- and benzo[j]fluoranthene coelute (Fig. 4), illustrating the advantage of using multiple columns for quantification of PAHs in environmental samples.

Fig. 3
figure 3

Comparison of GC/MS single ion chromatograms for PAHs of molecular mass 228 on three different GC stationary phases for SRM 2585 Organic Contaminants in House Dust (representative GC conditions are given in Fig. 2)

Full size image
Fig. 4
figure 4

Comparison of GC/MS single ion chromatograms for PAHs of molecular mass 252 on two different GC stationary phases for SRM 2585 Organic Contaminants in House Dust (representative GC conditions are given in Fig. 2)

Full size image

Multiple columns may not be an option in environmental PAH analysis because of cost or time constraints, and one column must be selected. The 50% phenyl methylpolysiloxane column usually enables separation of environmental PAHs that coelute on a 5% phenyl methylpolysiloxane column, including benzo[b]- and benzo[j]fluoranthene (Fig. 4), and dibenz[a,c]- and dibenz[a,h]anthracene, making this column the best choice for PAH separations. Benzo[c]chrysene, if present, may coelute with dibenz[a,j]-anthracene, however, and chrysene and triphenylene coelute (Fig. 3). As discussed in the Introduction, many of these isomers (and others) do not need to be determined individually. In such circumstances, the 5% phenyl methylpolysiloxane column (as suggested in Method TO-13A and many other test methods; Tables 1 and 2) may be used, although it is important to consider possible coelution (for example chrysene and triphenylene) when calculating concentrations. In addition, although 60 m columns and very slow temperature programming (Fig. 2) typically enable the best PAH separations, shorter columns and faster temperature programs may be warranted when high throughput is required or site surveys are being conducted (e.g. field GC–FID), for example for only one compound, such as benzo[a]pyrene. Fast GC is discussed later in this paper.

Although GC is a mature technique, the development of novel stationary phases for GC analysis of PAHs continues. For example, liquid crystalline stationary phases containing crown ether groups have been developed which separate some three-ring PAHs [80]. Two high-purity all-hydrocarbon side-chain liquid-crystalline polysiloxane polymers have been synthesized by grafting all-hydrocarbon liquid-crystal monomers onto a polymethylhydrosiloxane backbone. The two polysiloxane polymers enabled enhanced separation of 21 PAHs compared with the 5% phenyl methylpolysiloxane stationary phase, although triphenylene was not included in the analysis [81]. Triphenylene coelutes with chrysene on a 5% phenyl methylpolysiloxane column (see above). Other new stationary phases have been developed for PAH separations. A mesogenic polymer was prepared from the metal complexes of 4-(dec-9′-en-1′-oxy)dithiobenzoate and polysiloxane and used as the stationary phase for ligand-exchange gas chromatographic separation of low-molecular-mass PAHs [82]. Phenanthrene and anthracene, two key low-molecular-mass environmental PAHs, were, however, only partially separated. More recently, columns containing synthetic polymer filaments have been introduced as the support material in packed capillary GC. Filaments of the heat-resistant polymer Zylon have been coated with polydimethylsiloxane and shown to be useful for the separation of some PAHs (naphthalene, fluorene, phenanthrene, pyrene, and triphenylene) [83].

Retention data

A common way of monitoring gas–liquid interactions in capillary columns and of comparing stationary phases is to examine retention data of isomers as retention indices. Retention indices are a way of normalizing solute retention relative to the retention of specific standards [84]. Different retention index systems have been based on the use of alkanes [85] or PAHs [86] as retention standards. In general, retention indices are less dependent upon chromatographic conditions than are retention times, and are a way of comparing relative behaviors on different stationary phases. In this way, the order of elution and potential for separation of specific compounds may be predicted, facilitating column selection. Relationships between gas chromatographic retention indices of PAHs have been investigated since the late 1950s, and Lee et al. [86] introduced a system based on PAHs with an increasing number of aromatic rings as retention markers. Values may be measured, although many prediction methods have been developed (discussed below). Andersson and Weis [87] have investigated the retention indices and orders of retention of a range of PAHs and fluorinated PAHs on five capillary columns of different selectivity (PB (polymethylsiloxane), DB-5 (95:5 methyl phenylsilicone), OV-1701 (5:7:88 cyanopropyl–phenyl methylsilicone), DB-17 (50:50 methyl phenylsilicone), and DB-Wax 20M (polyethylene glycol). Similarly, Escrivá et al. [88] compared separations of PAH mixtures in airborne particulate extracts on five different capillary columns. Retention indices for PAHs with molecular mass 300 and 302 have been determined using picene and benzo[c]picene as retention markers with a 50% phenyl methylpolysiloxane column, and picene and coronene as retention markers with a liquid-crystalline column [78]. Correlations of retention and PAH shape on the basis of length-to-breadth ratios [89, 90] were also investigated [78]. More recently, retention indices for 13 dibenzopyrene homologues (molecular mass 302) have been determined on four different stationary phases (5%, 35%, and 50% phenyl methylpolysiloxane columns and a 35% trifluoropropyl methylpolysiloxane column). Correlations of retention on each phase with eight molecular descriptors, including PAH volume and length-to-breadth ratio, were reported [91].

Saura-Calixto and Garcia-Raso presented one of the early studies on the prediction of retention indices [92]. Temperature-programmed retention indices of unsubstituted PAHs were examined in relation to different molecular descriptors (first-order valence molecular connectivity, ionization potential, length, height, and moment) for compounds of molecular mass ranging from 178 to 350 amu [93]. Temperature-programmed retention indices of 70 PAHs on a methylsilicone capillary column have been reported and applied to the analysis of fuel and fuel deposits [94]. Quantitative structure–retention relationships for predicting capillary GC retention indices of PAHs using pseudo-conjugated pi-system surface and quasi-length of carbon chain have been examined for 100 PAHs on SE-52 columns [95]. The retention indices of 17 PAHs of molecular mass 302 have been predicted using a unique model that calculates the interaction energy between solutes and a dielectric medium; this is the first report on the relationship between retention indices and the interaction between the solute and the GC column [96].

Detection of PAHs: GC–MS

Detection of PAHs in environmental samples following GC separation is most commonly accomplished by use of quadrupole electron impact (EI) mass spectrometry. In contrast with many other organic contaminants, in this impact mode most environmental PAHs yield intense molecular ions with little fragmentation [11]. GC/EI MS operated in the selective ion monitoring (SIM; selected ions in Table 5) mode has advantages over full-scan mode, for example low detection limits and discrete monitoring capabilities. In general, ease of operation, and now reasonable purchase and operating costs have enabled GC/EI MS to become a popular, probably the most used, method for the determination of PAHs in laboratories worldwide. The use of multiple columns (see discussion above) that exhibit dissimilar PAH selectivity with GC/EI MS enhances this approach and addresses possible isomer coelution issues.

Table 5 Ions reported for sixteen priority pollutant PAHs for GC/MS and GC–ion-trap MSa
Full size table

GC/MS is often more accurate than conventional universal detection methods, for example GC–FID, for the quantification of PAHs, because interferences from coeluting compounds are minimized by the selective nature of the detector. Removal of the more polar analytes from the extract by column chromatography (solid-phase extraction or liquid chromatography, Fig. 1) before either GC–FID or GC/MS assists in obtaining the best chromatographic conditions. GC–FID is still a useful method, and a wealth of literature is available on the use of this method for the determination of PAHs in all types of environmental sample. For example, GC–FID is a component of national standard methods for the determination of PAHs in air (Fig. 4 in Ref. [31]) and in solid waste [26]. One advantage of FID is that the response is proportional to the number of carbons so one can quantify compounds in the same isomer group even if a matched calibrant is not available. Recent examples of the use of GC–FID for determination of PAHs in environmental samples include the analysis of air particulates [97], water [98], and sediment [99, 100]. A review of analytical methods for determination of PAHs in air (gaseous and particle-associated) documents the use of GC–FID and/or GC/MS to quantify compounds collected from a variety of sampling environments [101]. In the analysis of complex matrices, such as sediment or oil, GC/MS is often used for the determination of PAHs, and GC–FID is reserved for the determination of alkanes or aliphatic compounds [102, 103]. The usefulness of GC/MS compared with GC–FID for the analysis of complex matrices such as biological organisms has been presented in terms of sensitivity and selectivity [104]. Baumard and Budzinski concluded that the observed reduction in the number of interfering peaks in GC/MS chromatograms in comparison with those obtained by use of GC–FID makes GC/MS “the preferred analytical system for PAH environmental analyses” [104]. A recent overview of developments and applications of GC/MS in the analysis of environmental samples for the determination of a range of persistent pollutants has been presented by Santos and Galceran [105]. Different mass analyzers and recent developments in field-portable GC/MS have also been examined [105].

Alternative MS modes to EI for the determination of PAHs include positive and negative chemical ionization (PCI and NCI, respectively). Conventional chemical ionization methods using methane as the reagent gas typically produce mass spectra similar to those produced by EI. NCI often produces a low yield of negative ions, and a decrease in sensitivity in NCI may be observed [11]. However, NCI with methane has been shown to be highly sensitive and selective for some isomeric PAHs. Selective detection of the molecular anions of fluoranthene over pyrene and of benzo[a]pyrene over benzo[e]pyrene has been demonstrated, and the approach has been used to determine selected PAHs and alkyl-PAHs in diesel and air particulate matter [106]. GC/MS strategies for both PCI and NCI modes with a range of reagent gases, for example methane, dimethyl ether, tetramethylsilane, ammonia, carbon dioxide, or mixtures of gases, have been reported for the determination of PAHs. Selected priority pollutant PAHs (Table 5, excluding naphthalene, acenaphthylene, acenaphthene, and fluorene) have been separated and identified on the basis of the identities and relative abundances of the ions produced by PCI using dimethyl ether as reagent gas [107]. When ammonia was used as the reagent gas, however, benzo[ghi]perylene and indeno[1,2,3-cd]pyrene were not distinguishable [107]. Many GC PCI and NCI applications for PAHs are mentioned by Yurchenko and Mölder, who used both PCI with ammonia and methane and NCI with methane to determine PAHs in smoked fish and oil samples [108].

GC–ion-trap MS

A method for the separation and determination of PAHs using an ion-trap MS has been developed and compared with alternative GC and LC methods, including GC/MS using a quadrupole and SIM techniques and LC–fluorimetric methods. Use of the ion trap detector resulted in greater sensitivity than using GC/MS or LC–fluorimetry. This ion-trap method was designed to run simultaneously with GC–FID, enabling simultaneous qualitative identification and quantitative determination [109]. This method, because of its increased selectivity, is receiving increased attention for PAH analyses. In contrast to quadrupole instruments, there is no sensitivity advantage to be gained by limiting the number of ions detected in the ion-trap. Hence, all samples can be scanned in full-scan mode with little or no extra effort. In addition, archived sample files can be viewed later for compounds that may not have been a target at the time of analysis [11]. However, GC–ion-trap MS is not without limitations. Linearity may be a problem, because of the capacity of the ion trap. Despite this, GC–ion-trap MS has recently been used for determination of PAHs in marine environmental matrices (seawater, sediment, and mussel homogenate samples [110], estuarine sediments [111], and sewage sludge [112]). Ions reported for GC–ion-trap MS analysis of sewage sludge samples are listed in Table 5 [112]. Low-pressure GC–ion-trap MS (LPGC–ITMS) has been used for the determination of PAHs in aerosol samples from sugar cane burning, and 16 PAHs were separated in less than 13 min [113]. LPGC–ITMS has also been used for the analysis of gas and aerosol phase samples collected in the ambient air of Hasselt, Belgium [114]. In low-pressure GC, a gain in speed is obtained by operating a column under vacuum-outlet conditions rather than at atmospheric pressure. This gain in speed becomes substantial for short and/or wide-bore thick-film columns [115]. PAHs may also be determined by GC–ion-trap MS using collision-induced dissociation (CID) and electron impact (EI) [116, 117]. CID analysis has been shown to be more sensitive than EI in the analysis of used motor oil [117].

GC–time-of-flight (TOF) MS (GC–TOF-MS)

In recent years TOF-MS with GC and LC sample-introduction has become increasingly prevalent in environmental analysis [118, 119]. The unique aspect of this detection method is that accurate mass measurements and full-scan data can be acquired on short time scales. Quantitative data are also obtainable. Because of its higher mass resolving power, TOF-MS can enable better structural conformations and better signal-to-noise ratios than single quadrupole analyses, especially with complex samples [119]. When conducting surveys or discovery-based analysis for new contaminants the full spectral sensitivity of TOF-MS is an advantage over traditional scanning MS instruments where limits of sensitivity are often experienced even when scans are conducted over a narrow mass range [119]. Vreuls et al. provide a brief summary of TOF-MS and recently evaluated the spectrum storage rate, linearity of response, and detection limits of GC–TOF-MS for a range of organic contaminants, including the 16 priority pollutant PAHs, in a sediment and boiled tea [120]. Other studies of PAHs using GC–TOF-MS include the rapid analysis of PAHs in fly ash using thermal desorption directly coupled to the GC [121, 122]. Thermal desorption techniques with GC are discussed below, although the approach is also reported in combination with GC–TOF-MS. For example, thermal desorption coupled to GC–TOF-MS has recently been used for analysis of air particulate matter (≤2.5 μm) for the determination of PAHs, oxidized PAHs, n-alkanes, hopanes, and long-chain linear alkylbenzenes [123]. In a related study, two hundred compounds were quantified and “semi-quantified” on a daily basis by thermal desorption GC–TOF-MS with data thought to be suitable for source receptor modeling and epidemiological time series studies on the health effects of air particulate matter [124]. The authors note, however, that use of comprehensive two-dimensional GC (GC×GC) with TOF-MS enhanced the chromatographic resolution of the components of ambient air particulate matter compared with GC–TOF-MS and significantly improved the peak-identification capabilities of TOF-MS [124]. Methods that make use of multidimensional GC and GC×GC are gaining popularity, because of the benefits of enhanced separation to quantitative analysis [125, 126]. A collection of papers from the 2003 First International Symposium on Comprehensive Multidimensional Gas Chromatography reveals the wide range of multidimensional GC and GC×GC approaches for the determination of analytes, including PAHs, in environmental samples, petrochemical-related samples, cigarette smoke, fish, food, and essential oils [127]. The analysis of petrochemical and related samples by GC×GC using FID and TOF-MS as detection methods has recently been reviewed [128]. This issue contains a review of GC×GC by Gorecki [59].

GC–isotope-ratio mass spectrometry (IRMS)

Approximately a decade ago it became clear that source inputs of PAHs to the environment could be determined using carbon isotope measurements of individual PAHs [129–131]. Compound-specific carbon-isotope (13C/12C) ratios enable quantitative assessment of PAH sources in natural environments and have been applied to source apportionment studies of PAHs in soils and sediments from marine, lacustrine, and terrestrial environments [129, 132–134]. These specialized measurements are conducted by use of GC coupled with isotope-ratio mass spectrometry (IRMS), as recently reviewed by Schmidt et al. [135] and Lima et al. [3]. In this technique, a GC is coupled to a combustion furnace and the isotopic composition of the resulting CO2 is continuously analyzed using a magnetic sector MS. McRae et al. described the determination of the isotope ratios of specific aromatic and aliphatic hydrocarbons from coal-conversion processes [136]. GC–IRMS instruments for the determination of the isotope ratios of four elements (H, C, N, and O) are commercially available, and coupling of LC with IRMS has been attempted, although no commercial instrument is available because of the difficulties associated with the presence of a solvent mobile phase in the combustion unit [135]. Kim et al. have described a unique approach for purification of environmental extracts targeted for specific isotope analysis using column chromatography, LC, and thin-layer chromatography [134]. The GC–IRMS technique is a potential environmental forensic method at sites formerly used for gas manufacture [137].

Quantification of PAHs

Prior to using GC for the separation of PAHs in environmental samples, samples are usually extracted with solvent, and the extracts are concentrated and cleaned up by use of normal-phase LC or solid-phase extraction procedures to remove potential interfering polar constituents (Tables 1 and 2 and Fig. 1) [4, 29, 30, 42–45]. Modern methods of sample preparation have been reviewed by Smith [138]. Schantz et al. [139] have evaluated pressurized liquid extraction (PLE) for the determination of PAHs in environmental matrices, and this issue includes a review of PLE in environmental samples [140].

Identification and quantification of individual isomers is dependent on the use of authentic standards. These may be characterized by spectroscopic methods (UV, MS, or high-resolution nuclear magnetic resonance (NMR)) or by melting point [141]. Many PAH standards are commercially available and several PAH-related solutions are available from NIST (Table 6). Schantz et al. have recently reviewed NIST SRM solutions [142].

Table 6 Solution SRMs for PAHs and substituted-PAHsa
Full size table

Two solution SRMs are available that contain PAHs in either acetonitrile (SRM 1647e, 16 PAHs) or toluene (SRM 2260a, 36 aromatic hydrocarbons, primarily PAHs, Table 6). The concentrations of the 16 priority pollutant PAHs in SRM 1647e range from 1 mg kg−1 for anthracene to 26 mg kg−1 for acenaphthene. With acetonitrile as the solvent, this SRM is intended for the calibration of LC systems for PAH analyses but it can also be used to fortify aqueous samples with known concentrations of PAHs. SRM 2260a is the replacement for two previously available materials, SRM 1491 and SRM 2260, which contained the same 24 PAHs at nominal concentrations of 10 mg kg−1 and 70 mg kg−1, respectively. SRM 2260a contains 36 aromatic hydrocarbons at concentrations that mimic the relative concentrations of PAHs found in environmental matrices. Concentrations range from 2 mg kg−1 for cyclopenta[cd]-pyrene to 12 mg kg−1 for phenanthrene. SRM 1491a contains 18 methyl-substituted PAHs in toluene at concentrations ranging from 1 mg kg−1 to 2.5 mg kg−1. Several solutions are also available for the determination of nitro-substituted PAHs (Table 6).

The solutions listed in Table 6 are intended for use as calibration solutions and are typically processed and analyzed in parallel with environmental samples to generate individual response factors relative to internal standards for quantification purposes [18]. The internal standard(s) added to samples should have similar or matching chemical and physical properties relative to the target analytes, must not be present in the environment, and should elute from the GC column at a time near to that of the target analyte(s). Stable isotope-labeled internal standards are preferred to others. When labeled reagents are not available, a compound as similar as possible to the analyte should be selected (i.e. an isomer, a methyl analog, or other related compound not present in the sample). Internal standards in PAH environmental analyses are most often perdeuterated PAHs (2H-labeled PAHs) [42–45, 104, 143, 144] (Table 6). Perdeuterated PAHs are resolved from parent PAHs by both LC and GC. They are components in most standard test methods (US EPA [24, 27], also see Tables 1 and 2). Fluorinated PAHs [145, 146] have also been shown to be useful. Both of these analogs may be used with either GC/MS or GC–FID. Carbon-13 labeled PAHs [147] may be used as internal standards, although only in GC/MS analysis in which the selective nature of the mass spectrometer affords the opportunity for detection of these compounds, which typically coelute with the unlabeled compounds. Use of selected 13C-labeled PAHs was part of the analytical scheme for the determination of PAHs in a mussel tissue SRM [44]. Eighteen monohydroxy PAHs (OH-PAHs) containing up to four rings have been measured in human urine using 13C-labeled internal standards and GC/MS (high resolution) [148]. Although many of these compounds are commercially available, 13C-labeled compounds are usually more expensive than perdeuterated compounds. NIST has produced two SRM solutions containing perdeuterated PAHs (Table 6). These may be used as internal standard or surrogate solutions. Both solutions are prepared in a 96:4 (v/v) hexane–toluene, which can be gravimetrically added to samples and/or calibration solutions.

Response factors relative to internal standards are typically calculated using either a single concentration or multiple concentrations to generate calibration plots. Single concentrations are useful when determining an expected or known concentration (e.g. samples subjected to preliminary characterization or solutions for which the mass of the analyte is known from the gravimetric preparation of the solution). Calibration plots are useful when a range of concentrations is expected. Although zero or non-zero intercept models may be used, a non-zero intercept should usually be small in relation to the concentration of the analyte. Standard test methods, for example those provided by the US EPA [24, 27], typically provide approaches for quantification (Tables 1 and 2).

An essential part of excellent analytical chemistry is the appropriate use of reference materials. Reference materials are stable, homogeneous materials that are well-characterized for one or more properties and enable instrument calibration, method evaluation, or characterization of other materials. Reference materials may also be used to obtain estimates of intermethod and/or interlaboratory comparability [149]. A certified reference material (CRM) is one with properties certified by validated procedures; and is accompanied by documentation from the certifying organization that assesses the uncertainty of the property values (or its properties are traceable to such documentation) [150]. An SRM [150] is a CRM issued by NIST (http://www.nist.gov/srm). Reviews on environmental reference materials and their use are provided in this issue by Wise et al. [40], Zeisler [151], and Ulberth [152]. Many reference materials are available for the determination of PAHs (Table 4).

Enhancements to GC for PAH analysis

Large-volume injection GC

The growing demand in trace analysis for lower detection limits has led to the development of innovative ways of introducing larger sample volumes into LC and GC systems [153]. Large-volume injection techniques have advantages in addition to improved sensitivity. For example, use of large-volume injection often reduces or eliminates the need for sample concentration (i.e. solvent evaporation), thus saving time and effort. Accuracy may also be improved, because potential losses of analytes in the concentration step are minimized [51]. If the sample extract is sufficiently clean and/or the detector selectivity high, detection limits improve in proportion to the volume injected [154]. Methods for, and applications of, large-volume injection in GC have been reviewed by Teske and Engewald [155]. Large-volume samples (30 μL to approx. 100 μL and larger) can be injected by use of several techniques, including on-column [156] and temperature-programmable approaches [154]. In the latter mode, referred to as a programmed-temperature vaporizing (PTV) injection, the inlet temperature at the time of injection is usually below the boiling point of all the analytes and the solvent and can be time-programmed during the GC runtime. Large-volume injections are usually performed in split mode, so most of the solvent vapor exits the system via the split vent [155]. During this time, the higher-boiling compounds are enriched in the insert and the analytes are subsequently transferred onto the analytical column in the splitless mode by rapid heating of the injector [155]. Retention gaps (or precolumns) are recommended to assist with sample recondensation and desolvation, and to maintain system performance by preventing contamination of the GC column with non-volatile materials or traces of water [154, 157]. Norlock et al. recently reported results obtained from the optimization of PTV-large-volume injections (up to 150 μL) for the determination of PAHs by MS [158]. Aspects investigated included detection limits, maximum total volume of injection, dependence of linearity of MS response on injection volume, and analytical repeatability [158]. Norlock et al. also investigated the determination of PAHs in air and sediment samples using their optimized method [158]. Other examples of applications in which large-volume injection methods were used include analysis of extracts from aqueous solutions [153], river water [154], soils and sediment [154, 156, 159, 160], and air particulate samples [161]. In the air particulate study, the authors note that a ten to fiftyfold increase in sensitivity was achieved with the large-volume injection technique compared with use of a 2-μL splitless injection [161]. Large-volume injection has also been used for the determination of PAHs by dual column GC [157]. Large-volume injection techniques will probably become more prevalent in environmental analysis now that commercial injectors are available in a variety of configurations (PTV, cool on-column) [154, 156, 158]. Večeřa et al. [156] constructed a large-volume injection system from common materials found in a GC laboratory; a wide-bore retention gap was placed between an on-column injector and the analytical column to enable injection of up to 80 μL [156]. Use of large-volume injection as an interface for coupling GC with other separation techniques (e.g. LC) or with sample-preparation methods (e.g. extraction) has been recently reviewed by Teske and Engewald [155].

Thermal desorption techniques with GC

Thermal desorption is a powerful technique that eliminates the need for sample extraction. Sample processing is accelerated, and potential contamination from solvents is eliminated. Initial studies of adsorption/thermal desorption have been directed toward the analysis of organic contaminants in environmental samples and volatile organic compounds in air [101]. In the early 1980s the sorbent Tenax was used for direct sampling and analysis of water for sample volumes up to several liters, and recoveries near 100% were obtained for a variety of phenols, monocyclic aromatic compounds, and selected PAHs, pesticides, phthalates, and alkanes [162, 163]. This technique was later applied to air sampling, and gaseous organic contaminants were directly sampled and analyzed by thermal-desorption GC/MS [164]. The use of adsorption/thermal desorption for PAHs is not reported extensively in the literature as a routine approach. Instead, the focus in recent years has been more on direct sample introduction via thermal desorption [123, 165–168]. In this approach, solid or liquid samples are introduced into the GC inlet with no extensive modification of the inlet. The sample is usually stored in a sample holder that is inserted into the GC inlet. The compounds are thermally desorbed directly in the liner and focused onto a cool GC column for separation and detection, usually by MS [166]. Several air particulate SRMs have been used to study the effectiveness of the desorption process [101, 166, 167]. Thermal desorption has also been coupled to GC–TOF-MS (see above) for the determination of PAHs in fly ash [121, 122] and air particulate matter [123, 124]. Direct thermal desorption of samples via Curie point pyrolysis coupled to GC/MS has been demonstrated for determination of PAHs in soil [169], and this pyrolysis technique has been directly compared with thermal desorption GC/MS for the determination of PAHs in river sediments [170]. EPA Method 8275 [26] makes use of thermal extraction–GC/MS (Table 1).

Thermal desorption coupled with stir-bar sorptive extraction is gaining popularity. Stir-bar sorptive extraction is an extraction technique similar to SPME but larger in scale than the fibers in SPME (SPME is reviewed in this issue by Ouyang and Pawliszyn [60]) and is based on the solid phase polydimethylsiloxane. Stir-bar sorptive extraction is relatively novel, and is often used for the determination of PAHs in water (reviewed by Rawa-Adkonis et al. [171]). This extraction technique may be coupled with thermal desorption GC. The theory and practice of this technique for a wide range of organic contaminants in aqueous samples, including PAHs, has been described [172]. A recent paper describes the optimization of this method coupled with GC/MS for 35 semi-volatile organic contaminants, including PAHs, in water [173]. Method sensitivity, linearity, repeatability and reproducibility, accuracy, matrix effects, and overall uncertainty have been studied by the same group, using ISO guidelines, for validation of the method, and the accuracy of the method was also evaluated by participation in a proficiency interlaboratory test with ground, tap, and surface-water samples [174]. This method has also been applied to the analysis of seawater for the determination of PAHs [175]. Other extraction techniques coupled on-line with GC include supercritical-fluid extraction (SFE), liquid–liquid extraction, and membrane-based techniques. These approaches have recently been reviewed in several articles on coupled extraction–GC techniques [176–181].

Fast GC

Fast GC is designed to minimize analysis time without compromising chromatographic resolution; it is usually accomplished by reducing the characteristic diameter of the GC column. A major advantage of this technique is that peak width is small so the signal-to-noise ratio is larger. A disadvantage is that the stationary phase may be substantially overloaded if analytes are highly concentrated. The concepts of fast GC with packed and capillary columns has been reviewed by Cramers et al. [182]. In addition to reduced column diameter, nearly all fast GC instruments utilize short columns. A recent review by Snow [183] describes the practical implications of the use of short columns on fast GC method development, optimization, and resolution. Short, microbore columns have been shown to work well for separation of the priority pollutant PAHs [184]. Microbore (0.1 mm i.d.) columns are much more efficient than conventional 0.25 mm i.d. columns; this enables separations to be performed on shorter columns (10–20 m) and with faster analysis times [184]. Analysis times for PAHs on 20 m and 10 m columns coated with 5% phenyl methylpolysiloxane have been reduced by approximately 45% and 60%, respectively, compared with 30 m columns, with no compromise on data quality (precision and accuracy), although smaller injection volumes (0.2 μL to 0.5 μL) and injection liners (1 mm to 2 mm) were necessary to achieve optimum, reproducible chromatography [184].

The main limitation of fast GC is the speed of detector response. Fast GC is therefore limited mainly to GC–FID, GC–ECD, and GC–TOF-MS. An ion-trap MS GC method (LPGC–ITMS, see above) that is considered “fast” has also been reported; 16 PAHs are separated in approximately 13 min [113, 114]. However, much shorter analysis times are possible. Up to 16 PAHs have been shown to elute from 1 m and 5 m 5% phenyl methylpolysiloxane columns in approximately 3 min with fast oven-temperature programming and GC–FID [185]. While pairs of peaks were not completely resolved, chromatography coupled to MS in the selected-ion-monitoring mode enabled spectrometric resolution of the isomers if enough points per peak were acquired. The best resolution is, in fact, obtained by combining fast GC with either selective sampling techniques or selective detectors [182, 183]. The use of MS is often necessary for positive identification of solutes. This is particularly important in ultra-fast GC, in which very-high-speed temperature or pressure programming is used, because this results in reduced reproducibility of retention data [182]. In addition, because peak widths decrease substantially with increasing analysis speed, TOF-MS or spatial-array detection may be necessary to complete fast scans over large mass ranges [182].

Fast GC/MS has recently been reviewed [186], and five main approaches to the technique were discussed:

  1. 1.

    short, microbore capillary GC columns;

  2. 2.

    fast temperature programming;

  3. 3.

    low-pressure GC/MS;

  4. 4.

    use of a supersonic molecular beam for MS at high GC carrier gas flow; and

  5. 5.

    pressure-tunable GC×GC.

Topic 4, supersonic GC/MS, has been reviewed by Fialkov et al., and the use of this approach for rapid elution of pyrene and higher-molecular-mass PAHs has been described [79]. Fast GC/MS coupled with on-line thermal desorption has been used by Münchmeyer et al. for analysis of particle-associated PAHs formed in combustion processes [187]. This group designed a unique system that made use of a dilution method and filter tape sampling for the collection of particles in combustion effluents. Results were obtained in minutes at the ppb level for particle-associated PAHs. The analytical system was tested at oil combustion plants, coke ovens, and with diesel engine emissions [187]. A membrane inlet system was used for the MS [187]. In this technique, which is applied mainly to the analysis of VOCs [56], organic compounds are separated from water or air by a thin membrane (typically polydimethylsiloxane) installed between the sample and the ion source of a mass spectrometer. Organic compounds dissolve in and diffuse through the membrane then flow, via evaporation, into the ion source. Ketola et al. have written comprehensive review of environmental applications of membrane-introduction MS [188]. An advantage of membrane-inlet MS is that the method is fast, because the separation step is eliminated. However, its use for the determination of PAHs has not been widely reported.

As manufacturers respond to the need of the analytical community to reduce analysis times, techniques such as fast GC and membrane-introduction MS will probably become more prevalent in the PAH literature. The development of computer programs to predict temperature and pressure programs to enhance fast GC results will also foster growth in this area [51].

LC–GC approaches

Coupled liquid chromatography–gas chromatography (LC–GC) is a powerful technique that combines the best features of both techniques and is ideal for analysis of complex samples [179, 189, 190]. When using LC–GC, the whole sample is typically injected into the liquid chromatograph and subsequently analyzed by GC. In this way, several milliliters of eluent are introduced into the GC, affording excellent sensitivity and low detection limits. LC–GC methods have been reported for analysis of PAHs in atmospheric particles. Shimmo et al. have reviewed the use of LC–GC/MS coupled with on-line SFE as the extraction technique [191, 192]. Christensen et al. have reported an LC–GC/MS method for determination of PAHs in particulate matter from ambient air and diesel exhaust emissions with a limit of detection of approximately 1 pg per sample for individual compounds [193]. The method was examined using SRMs 1649a Urban Dust and 2975 Diesel Particulate Matter (Industrial Forklift). More recently, an LC–GC/MS method has been described for determination of dibenzopyrenes in diesel and air particulate materials; air and diesel particulate matter SRMs were used as control materials [194].

Concluding remarks

Interest in the determination of PAHs in environmental samples is motivated by their biological activity and the possible adverse health effects of this class of compounds. The oil and gas industries are also interested in the determination of PAHs, as also the space research community for analysis of meteorites and planetary materials. Gas chromatographic techniques have enabled researchers to measure and monitor PAHs in environmental, biological, and food-related samples. The development of new or enhanced GC techniques, for example GC–TOF-MS, has afforded scientists the opportunity to meet the ever increasing demands for the determination of PAHs with greater sensitivity, selectivity, and speed. As environmental forensic applications increase, sophisticated GC techniques, for example GC–IRMS, will probably be reported more often for the determination of the stable isotope ratios of individual PAHs in environmental matrices. Enhancements to GC for determination of PAHs can also be expected to be a continuing topic of research, with the objective of meeting the evolutionary measurement and monitoring needs associated with this class of compound in relation to ecological and human health effects.