Polycyclic aromatic hydrocarbons are classified as priority ecotoxicants of the 1st hazard class, having carcinogenic, teratogenic, and mutagenic properties. They are resistant to degradation, including thermal, and can remain in the environment for a long time [1]. The release of PAHs into the environment occurs under the effect of both anthropogenic and natural factors (forest fires, volcanic activity).

Anthropogenic sources of PAHs are emissions from various industries, for example, fossil fuel combustion, vehicles, waste and oil processing plants, and production of coke, asphalt, aluminum, etc. Polyarenes are concentrated near city centers mainly because of the pyrolysis of organic materials [2–4]. Other possible sources of PAHs are worn-out automobile tires, intensively maintained road surfaces [5], and crematoria [6]. The sources of PAHs in the urban atmosphere are oil refineries, power plants, and car exhausts [7, 8]. Tobacco smoke [9] and various heating systems [10] can also increase the concentration of PAHs in indoor air; they occur in significant amounts in sedimentary rocks and oil [11–14].

In recent years, the determination of PAHs has received increased attention because some of these compounds are highly carcinogenic and/or mutagenic [15]. The following PAH representatives are usually considered as possible carcinogens: benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, dibenzo(a,h)anthracene, indeno(1,2,3-cd)pyrene, and benzo(g,h,i)perylene [16]. The US Environmental Protection Agency named 16 unsubstituted PAHs as priority pollutants. In Russia, hygienic standards (GN 2.1.5.1315-03) for water for household, drinking, and cultural use set MPCs only for benzo(a)pyrene (0.00001 mg/L), biphenyl (0.001 mg/L), and naphthalene (0.01 mg/L). In the countries of Western Europe, the Council Directive 98/83/EC on November 3, 1998, establishes the total concentration of polyaromatic compounds in drinking water at the level of 0.1 µg/L. In highly contaminated waters, the concentration of pollutants can reach 10 µg/mL [17]. A direct correlation between the growth of oncological diseases among the population and an increase in the emissions of organic ecotoxicants into the environment was found [18]. In addition to the direct action on the functioning of the biosphere, polyarenes possess properties of accumulation and migration in natural objects.

From the atmosphere, PAHs enter the water bodies by settling on particles and dissolving in precipitations, presenting an undoubted toxicological risk for all living things. The contribution of atmospheric sources of PAHs to surface water pollution is approximately 80%. Rainwater contains significant amounts of organic compounds, including PAHs [19].

The pollution of surface water can also be of an emergency nature in spills of petroleum products or in the dumping of industrial wastes. Polyarenes can come permanently into surface and ground water from such sources as shore fortifications or ship bottoms treated with creosote and bitumen as anticorrosive coatings. For the same reason, traces of PAHs can be detected even in drinking water supplied to consumers through old distribution systems. The probability of the presence of polyaromatic compounds in groundwater is low, the lipophilic character of arenes explains their low concentration (0–5 µg/L) in groundwater in unpolluted areas [20]. The PAH concentrations in the waters vary depending on the type of water source: surface water, groundwater, or drinking water.

Thus, the search for new approaches to the determination of PAHs in waters by various analytical methods remains a problem of current importance. The difficulty in determining polyarenes in water bodies is due to their low concentrations and complex matrix composition. In addition, PAHs are rarely found in samples as individual compounds, most often they are constituents of multicomponent mixtures. To find out not only the total concentration but also the concentration of each component individually, it is necessary to apply analytical methods possessing high sensitivity, specificity, and informational value. Chromatographic methods are the most widely used [21, 22]. High-performance liquid chromatography (HPLC) is conventionally used for these purposes, but the method is not devoid of certain drawbacks. For example, reversed-phase HPLC requires a high consumption of highly pure reagents, and a solvent change operation is used during sample preparation, which leads to the loss of analytes. The selection of chromatographic conditions, in particular, an elution gradient profile and optimal detection wavelengths for different types of PAHs, is also time-consuming.

Most of the procedures for determining PAHs in waters, certified in Russia, are based on HPLC. PND (Environmental Regulatory Document) F 14.1:2:4.70-96 recommends the determination of 15 PAHs with fluorimetric detection in drinking, natural, and waste waters, but the lower limits of the analytical range are at the MPC level. Thus, the current procedure does not ensure the control of the concentration of polyarenes in order to forecast an operative intervention for the elimination of the contamination. Some normative documents presume the determination of only benzo(a)pyrene as an indicator of the presence of the entire group of priority PAHs in drinking and natural waters by voltammetry (MR (Methodological Recommendations) 146-11110) or cryoluminescence (PND F 14.1.2:4.66-96). These procedures are laborious and characterized by low selectivity and accuracy.

Gas chromatography–mass spectrometry, under certain conditions, may be an alternative to HPLC. This method is devoid of a number of the drawbacks mentioned above and is by far one of the main methods of water analysis, which makes possible the identification of the majority of pollutants because of the presence of an integrated spectral library and ensures their determination at a level below the MPC.

The purpose of this work was to develop a method for determining 16 priority PAHs in surface water samples by GC–MS.

EXPERIMENTAL

Materials and reagents. To identify and build the calibration curves for the analytes, we used a set of individual reference materials (IRMs) for the composition of PAH solutions in acetonitrile, including naphthalene (IRM 0109-03 ER-PAH 6), biphenyl (IRM 0107-03 ER-PAH 4), 2-methylnaphthalene (IRM 0101-03 ER-PAH 5), fluorene (IRM 0113-03 ER-PAH 9), acenaphthene (IRM 0103-03 ER-PAH 1), phenanthrene (IRM 0111-03 ER-PAH 7), anthracene (IRM 0102-03 ER-PAH 2), fluoranthene (IRM 0112-03 ER-PAH 8), pyrene (IRM 0110-03 ER-PAH 12), benzo(a)anthracene (IRM 0105-03 ER-PAH 15), chrysene (IRM 0114-03 ER-PAH 13), benzo(b)fluoranthene (IRM 0115-03 ER-PAH 14), benzo(k)fluoranthene (IRM 0116-03 ER-PAH 16), benzo(a)pyrene (IRM 0106-03 ER-PAH 3), dibenzo(a,h)anthracene (IRM 0108-03 ER-PAH 11), and benzo(g,h,i)perylene (IRM 0117-03 ER-PAH 17) (Ekros, Russia).

To extract PAHs from aqueous media, n-hexane (high-purity grade; Ekos-1, Russia) was used.

For the chromatographic determination of PAHs in water samples, we used a Shimadzu GC-2010 system equipped with an HP ULTRA 1 quartz capillary column (length 50 m, diameter 0.20 mm, dimethylpolysiloxane phase thickness 0.33 µm) and a GCMS-QP2010 Plus mass spectrometric detector using Wiley8 and NIST-05 mass spectral libraries.

Samples were processed using an ultrasonic oscillator (Sapfir, Russia).

Preparation of calibration solutions. The initial certified solution of 16 priority PAHs with a concentration of 1 µg/mL of each PAH was prepared from individual reference materials (200 µg/mL) as follows: 0.25 mL of each component was placed into a 50-mL flask and diluted with acetonitrile to a mark. Working certified solutions with concentrations ranging from 0.5 to 250 ng/mL were prepared by successive dilutions of the stock solution. Calibration solutions were model mixtures. A sample of water of a particular type (1 L), not containing analytes, was placed in a flat-bottom flask, and 1 mL of a certified working solution of a mixture of individual PAHs was pipetted to the flask.

The linearity of the signal was monitored at five concentrations for each PAH studied in the following ranges, ng/L: 50–250 for naphthalene; 25–250 for 2-methylnaphthalene; 5–250 for benzo(k)fluoranthene, dibenzo(a,h)anthracene, and ben-zo(g,h,i)perylene; 1–250 for biphenyl; and 0.5–250 for acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benzo(a)anthracene, chrysene, benzo(b)fluoranthene, and ben-zo(a)pyrene. The selection of these ranges of PAH concentrations was determined by the need to control their content in surface waters at the MPC level and below. The data [23–25] on the concentrations of PAHs in waters and individual features of ecotoxicants were also taken into account. In each case, no less than six parallel measurements were made for one concentration. All calibration dependencies were also constructed from six replicate measurements; the correlation coefficients were ≥0.98.

RESULTS AND DISCUSSION

Optimization of the conditions of gas chromatography–mass spectrometry. Capillary columns with methylpolysiloxane or phenylmethylpolysiloxane as a stationary liquid phase are conventionally used for the efficiently chromatographic separation of PAHs [26]. Polyakova et al. described the successful use of a GC Rtx®-Dionix 2 column, first developed by RESTEK (United States) for the determination of dioxins and furans, for the separation and determination of PAHs [27]. Based on the analysis of the published data and the structural features of the studied polyaromatic compounds, an HP ULTRA1 column (length 30 m) was selected to optimize the separation conditions for 16 PAHs. However, using this column, the retention times of individual compounds were very close for the pairs of phenanthrene–anthracene, benzo(b)fluoranthene–benzo(a)pyrene, and benzo(a)anthracene–chrysene, and a satisfactory separation of the peaks was not achieved even by adjusting the temperature program. Increasing the length of the column to 50 m made it possible to select the optimal conditions (Table 1) for a satisfactory separation of all substances in a joint presence (Fig. 1) under optimized conditions. The retention times of PAHs under study are listed in Table 2.

Table 1.   Conditions of gas chromatography–mass spectrometric analysis
Full size table
Fig. 1.
figure 1

Chromatogram of a model mixture of 16 polycyclic aromatic hydrocarbons.

Full size image
Table 2.   Retention times of polycyclic aromatic hydrocarbons under optimized chromatographic conditions
Full size table

The use of the selected ion monitoring mode in the spectrometric detection of the analytes not only decreased their detection limits but also significantly eliminated the interfering effect of the matrix of composite samples.

Optimization of sample preparation conditions. An essential stage in any analysis is sample preparation, which allows not only the isolation of the substance under study but also its preconcentration. The main difficulties at the stage of sample preparation in the determination of PAHs in waters are the difficulty of maximizing the extraction of target components and the need to eliminate the matrix effect. In addition, in analyzing water objects by gas chromatographic methods after quantitative extraction of PAHs into an organic solvent, the extract must be dried to exclude a possibility of hydrolysis of the stationary liquid phase of the capillary column. The extract is usually passed through anhydrous sodium sulfate, which leads to additional losses of the analyte.

One of promising ways to extracting PAHs from aqueous media is solid-phase extraction (SPE); both reversed-phase [28] and porous polymeric adsorbents [29] are successfully used. Using Amberlite XAD-2 for the extraction and prefractionation of 14 PAHs from aqueous samples, the limits of detection for the analytes ranged from 0.1 to 4 ng/L [29]. Amberlite XAD-2 is one of the most popular polymeric adsorbents for trapping impurities of toxic substances; however, the duration of the procedure limits its use. Song et al. described the application molecularly imprinted polymers (MIPs) for the extraction of PAHs from natural waters by SPE [30]; however, due to the complex matrix of the samples, it is necessary to wash the adsorbent, which can also lead to the loss of target components. In addition, calcium and magnesium cations of natural waters can interact with carboxyl groups of MIPs, and some joint compounds can bind to the polymer matrix, leading to the loss of the ability of the adsorbent to recognize the identified components [31].

There are examples of a combination of SPE and liquid–liquid extraction (LLE). A rapid determination of 16 polyarenes in natural water samples was proposed with successive use of dispersive liquid–liquid microextraction and dispersive solid-phase microextraction [32]. In this case, 1-octanol was used as the extractant together with derivatized magnetic nanoparticles, and a vortex mixer was used for the intensification of the process. The particles were separated from the solution by a magnet; PAHs were desorbed with acetonitrile under sonication. The solution was analyzed by GC–MS in the selected ion monitoring mode. The limits of detection for the analytes ranged from 0.05 to 0.21 µg/L. This method, compared with the conventional version of SPE, enables excluding the stage of centrifugation, inconvenient and long-lasting for routine studies. It has a number of advantages such as simplicity, rapidity, and a lower level of losses at the stage of sample preparation. An increase in the area of phase contact because of the uniform distribution of the magnetic adsorbent in the solution ensures fast mass transfer and the isolation of analytes from complex samples without preliminary separation of the matrix components. At the same time, the excess free energy of magnetic nanoparticles determines their tendency to aggregate in solution with the formation of agglomerates and the weakening or loss of some physicochemical properties [33].

Despite the diversity of innovative methods for extracting PAH from water bodies, liquid–liquid extraction remains one of the most common ways to isolating and preconcentrating PAHs. Analytes are extracted, as a rule, by individual organic solvents; the use of one extractant avoids the introduction of interfering components. For example, in [27], 16 PAHs were extracted from 1 mL of the water sample with dichloromethane, followed by drying the extract with anhydrous sodium sulfate. This method, despite its rapidity and simplicity, does not offer low detection limits (10 µg/mL). Loss of analytes is due to the use of a significant amount of sodium sulfate, which binds not only water but also partially organic extract.

A useful tool for accelerating the extraction of organic and inorganic compounds from aqueous samples is ultrasonic treatment [34]. For example, 10 PAHs were extracted by ultrasound-assisted emulsification liquid–liquid microextraction from a 12-mL sample of natural water into 11 µL of toluene; the aqueous and organic phases were separated by centrifugation [35]. The extract was analyzed by gas chromatography with flame ionization detection. The detection limits for various PAHs ranged from 0.02 to 0.05 µg/L.

A comparative study of the extraction of PAHs from aqueous samples at their low concentrations by SPE using a cartridge with modified C18 silica, liquid–liquid microextraction (LLME) with n-hexane under stirring with a magnetic stirrer, and ultrasound-assisted emulsification liquid–liquid extraction (UA-ELLE) under frequency of 35 kHz was carried out. PAHs were determined by GC–MS. The results are presented in Table 3. The extraction of an analyte is maximized using UA-ELLE. This, apparently, is due to a more substantial increase in the contact area of the extractant with the analyzed sample than in LLME. We encountered a number of difficulties in the course of SPE, associated with the breakthrough of some analytes at concentrations at the upper limit of the analytical range and the duration of the extraction procedure (on the order of several hours) because of the need to pump a large volume of sample through the cartridge while preconcentrating the trace amounts of contaminants.

Table 3.   Extraction rates (%) of polycyclic aromatic hydrocarbons for different methods of sample preparation
Full size table

The conditions of UA-ELLE were optimized by varying the volume of the extractant (5 mL, then 10 to 60 mL with 10-mL increments) and the duration of the ultrasonic treatment (5 to 45 min with 5-min increments). The extracts were analyzed by GC–MS. It is seen from the data for naphthalene, benzo(a)pyrene, and phenanthrene (Fig. 2) that the maximum extraction of analytes was achieved with an extractant volume of 30 mL. Similar relationships were observed for all PAHs being determined. The maximum extraction for all PAHs is achieved after 30 min, and after 20 min, 80% of the maximum possible amount passes into the organic phase. Further increase in the extraction time did not lead to a significant increase in the recovery rates (Fig. 3).

Fig. 2.
figure 2

Dependence of the analytical signal of (▲) naphthalene, (◆) phenanthrene, and (⚫) benzo(a)pyrene on the volume of the extractant.

Full size image
Fig. 3.
figure 3

Dependence of the analytical signal of (▲) naphthalene, (◆) phenanthrene, and (⚫) benzo(a)pyrene on the time of ultrasonic treatment.

Full size image

Thus, PAHs from a 1-L water sample should be extracted by UA-ELLE using 30 mL of n-hexane for 30 min. Comparison of the results of the two methods of drying the extracts obtained—using anhydrous sodium sulfate and freezing—showed that in the latter case, the loss of analytes is smaller, which leads to a decrease in the limits of detection for pollutants.

To preconcentrate the analytes, dried extracts were evaporated under nitrogen to a volume of 1–2 mL; the residue was quantitatively transferred to a vial, carefully evaporated in air to barely noticeable traces of n-hexane, and kept in air to evaporate the solvent completely. After the evaporation, 100 µL of n-hexane was added to the vial with a microsyringe, and 2 µL of the sample was injected into the chromatograph. To account for the loss of the analytes at various stages of analysis, each calibration solution was passed through all stages of sample preparation. The calibration curves were plotted as the dependence of the peak area of the molecular ion of an individual PAH on its concentration in water (ng/L).

The individual PAHs were identified by the m/z values of their molecular ions in the mass spectrum (the selected ion monitoring mode) and also by the retention times of the peaks of individual standards (Fig. 1). The quantitative determination of individual PAHs was carried out by the ratio of the areas of their chromatographic peaks in the chromatograms of the calibration solution and the extract of the water sample.

The performance characteristics of the procedure for determining polycyclic aromatic hydrocarbons in surface waters were estimated by the standard addition method using certified reference materials (Table 4). The accuracy of the developed procedure for determining polyarenes in waters was examined using the metrologically certified HPLC procedure for determining PAHs (FR (Federal Register) 1.31.2007.03947) using the water sample of the Temryuk Bay (the Azov Sea) (Table 5). The results testify to satisfactory convergence and the possibility of using the procedure proposed to determine the concentration of PAHs in surface waters.

Table 4.   Performance of the procedure for the determination of polycyclic aromatic hydrocarbons in surface waters by gas chromatography–mass spectrometry (n = 6, P = 0.95)
Full size table
Table 5.   Results (ng/L) of analysis of a sample of Azov Sea water (n = 3, P = 0.95)
Full size table

The procedure was tested on water samples collected in the surface layer and at a depth of 1 m of the Karasun Lake (Krasnodar, Russia). The mass chromatogram of the surface water extract is shown in Fig. 4; the quantification results are given in Table 6.

Fig. 4.
figure 4

Chromatogram of the extract of the surface water sample of the lake Karasun (Krasnodar, Russia).

Full size image
Table 6.   Results (ng/L) of determination of polycyclic aromatic hydrocarbons in water samples from the lake Karasun (Krasnodar, Russia) (n = 3, P = 0.95)
Full size table

CONCLUSIONS

Thus, a unified procedure is developed for the simultaneous determination of 16 priority PAHs in surface waters, tested using real samples.