Microwave-Assisted Extraction of Polycyclic Aromatic Hydrocarbons

---

Anal. Chem. 2001, 73, 3790-3795

Microwave-Assisted Extraction of Polycyclic Aromatic Hydrocarbons from Marine Sediments Using Nonionic Surfactant Solutions Alessandra Bianco Prevot, Monica Gulmini, Vincenzo Zelano, and Edmondo Pramauro*

Dipartimento di Chimica Analitica, Universita´ di Torino, Via Pietro Giuria 5, 10125 Torino, Italy

Microwave-assisted micellar extraction (MAME) has been tested for the recovery of polycyclic aromatic hydrocarbons (PAHs) present in samples of marine sediments. An aqueous solution of the nonionic surfactant polyoxyethylene(23)dodecyl ether (Brij 35) was employed as the extracting medium. The proposed approach showed recovery efficiencies comparable to those afforded by the Soxhlet technique with organic solvents, but a neat reduction of the extraction times and a better reproducibility were observed. A MAME-based protocol was successfully applied for the analysis of a certified sample. Polycyclic aromatic hydrocarbons (PAHs) constitute an important class of ubiquitous environmental pollutants.1,2 Because of the high carcinogenicity, mutagenicity, and toxicity exhibited by most of them,3,4,5 they are considered to be priority pollutants by both the European Environmental Agency (E.E.A.) and the Environmental Protection Agency (E.P.A.).6,7 The anthropogenic contribution to their presence in the environment can be essentially related to pyrolytic and petrogenic factors. PAHs may enter the aquatic compartment through leaching of contaminated soils. In the marine environment, tank washing and accidental oil spillage represent relevant sources of PAHs. Because of their high hydrophobicity, PAHs are normally present at very low concentration levels in water, but sediments are effective PAHs collectors; therefore, monitoring of the sediment contamination level is of primary importance. The PAHs’ extraction from sediment samples is usually achieved by using well-established methods, Soxhlet extraction being the most used one.8-11 However, this method * Corresponding author. Fax: 39-011-6707615. E-mail: pramauro@ ch.unito.it. (1) White, K. L. Environ. Carcinog. Rev. 1986, C4, 163-170. (2) McElroy, A. E.; Farrington, J. W.; Teal, J. M. Metabolism of Polyciclic Aromatic Hydrocarbons in the Aquatic Environment; Varanisi, CRC Press: Boca Raton, 1989; pp 1-40. (3) Harvey, R. G. Polycyclic Hydrocarbons and Carcinogenesis; American Chemical Society: New York, 1985. (4) Lee, M. L.; Novotny, M. V.; Bartle, K. D. Analytical Chemistry of Polyciclic Compounds; Academic Press: New York, 1982; p 462. (5) Grimmer, G. Environmental Carcinogens; Polyciclic Aromatic Hydrocarbons; CRC Press: Boca Raton, 1983. (6) EPA Test Method 610; U.S. Environmental Protection Agency, U.S. Government Printing Office: Washington, DC; 1982. (7) Hennion, M. C.; Pichon, V.; Barcelo`, D. Trends Anal. Chem. 1984, 13, 361-372. (8) Lee, H. B.; Dookhran, G.; Chau, A. S. Y. Analyst 1987, 112, 31-35. (9) Saim, N.; Dean, J. R.; Abdullah, M. P.; Zakaria, Z. J. Chromatogr. A 1997, 791, 361-366.

3790 Analytical Chemistry, Vol. 73, No. 15, August 1, 2001

has some drawbacks, namely, it is generally time-consuming, and it needs the use of a relatively large amount of toxic and hazardous organic solvents. To overcome the above-mentioned problems, microwave-assisted solvent extraction (MAE), accelerated solvent extraction (ASE), and supercritical fluid extraction (SFE) have been developed in recent years and proposed as suitable new procedures for the extraction of polycyclic aromatic hydrocarbons from solid matrixes.12-15 Following the general research trend devoted to the development of extraction methods that would be alternatives to Soxhlet, we combined the MAE approach with the use of an aqueous surfactant solution as extracting phase. We applied the proposed MAME approach to the extraction of PAHs from natural sediments. This procedure is based on the well-known solubilization capacity of aqueous micellar solutions toward water-insoluble or sparingly soluble organic compounds. As a general rule, nonionic surfactants are usually the most effective, showing larger solubilization capacities which rapidly increase, together with the solubilization kinetics, as the cloud-point temperature of the solution is raised.16,17 Applications of nonionic surfactant solutions in various analytical18-20 and environmental21 procedures have been reported in the literature and, in particular, the effective removal of PAHs from solid environmental phases has been described.22-24 The possible treatment of aqueous micellar wastes containing aromatic (10) EPA Test Method 8100; U.S. Environmental Protection Agency, U.S. Government Printing Office: Washington, DC; 1986. (11) Simpson, C. D.; Cullen, W. R.; Quinlan, K. B.; Reimer, K. J. Chemosphere 1995, 31, 4143-4155. (12) Camel, V. Trends Anal. Chem. 2000, 19, 229-248. (13) Shu, Y. Y.; Lao, R. C.; Chiu, C. H.; Turle, R. Chemosphere 2000, 41, 17091716. (14) Lopez-Avila, V.; Young, R.; Beckert, W. F. Anal. Chem. 1994, 66, 10971106. (15) Saim, N.; Dean, J. R.; Abdullah, P. M.; Zakaria, Z. Anal. Chem. 1998, 70, 420-424. (16) Saito, H.; Shinoda, K. J. Colloid Interface Sci. 1967, 24, 10-15. (17) O’Rourke, B. G. C.; Ward, A. J. I.; Carroll, B. J. J. Pharm. Pharmacol. 1987, 39, 865-870. (18) Pelizzetti, E.; Pramauro, E. Anal. Chim. Acta 1985, 169, 1-35. (19) Hinze, W. L.; Armstrong, D. W. Ordered Media in Chemical Separations; Am. Chem. Soc. Symp. Ser. No. 342; Washington, D. C., 1987. (20) McIntire, G. L. Crit. Rev. Anal. Chem. 1990, 21, 257-278. (21) Edwards, D. A.; Luthy, R. G.; Liu, Z. Environ. Sci. Technol. 1991, 25, 127133. (22) Yeom, I. T.; Ghosh, M. M.; Cox, C. D.; Robinson, K. G. Environ. Sci. Technol. 1995, 29, 3015-3021. (23) Paterson, I. F.; Chowdhry, B. Z.; Leharne, S. A. Chemosphere 1999, 13, 3095-3107. 10.1021/ac010302z CCC: $20.00

© 2001 American Chemical Society Published on Web 06/30/2001

Table 1. Composition of the HS-6 Certified Sedimenta.

a

The reported standard deviations from the mean values were calculated from three replications

solutes by applying recently developed methods, such as photocatalysis,25 could also extend the use of these alternative solvents in the frame of safe laboratory practice. In the present study, the MAME procedure was applied to a certified reference material (marine sediment). The performances of two nonionic surfactants were examined, and the effects of various experimental parameters on the extraction efficiency were evaluated. Application of MAME to the analysis of river sediments (24) Hinze, W. L.; Singh, H. N.; Fu, Z.; Williams, R. W.; Kippenberger, D. J.; Morris, M. D.; Sadek, F. S. Chemical Analysis of Polyciclic Aromatic Compounds; Wiley: New York, 1989; p 151. (25) Pramauro, E.; Bianco Prevot, A.; Vincenti, M.; Gamberini, R. Chemosphere 1998, 36, 1523-1542.

is also being studied.26 EXPERIMENTAL SECTION Reagents. All of the reagents were of analytical grade and were used as received. Milli-Q-grade water was used throughout the work. Polyoxyethylene(23)dodecyl ether (Brij 35) and polyoxyethylene(10)dodecyl ether (C12E10) were purchased from Sigma. H3PO4, HCl, Na3PO4‚12H2O, Na2HPO4‚2H2O, CH2Cl2 (Suprasolv), and CH3CN (Lichrosolv) were purchased from Merck. Lab Service Analytica (Italy) provided a standard mixture containing 1 g L-1 of all of the investigated PAHs (mix 62). (26) Gulmini, M. Unpublished results.

Analytical Chemistry, Vol. 73, No. 15, August 1, 2001

3791

Table 2. Excitation and Emission Conditions Used in the Work for PAHs Detection analytes phenanthrene anthracene fluoranthene pyrene benzo[b]fluoranthene benzo[k]fluoranthene benzo[a]pyrene dibenzo[a,h]anthracene benzo[g,h,i]perylene indeno[1,2,3-c,d]pyrene

excitation nm

emission nm

wavelength change min

246 250 280 270 290

370 406 450 390 430

0.00 8.80 10.30 11.40 17.50

290

410

23.00

300

500

27.00

Sediment Sample. The examined reference material chosen for the determination of PAHs was a certified harbor marine sediment (HS-6) collected from four harbors in Nova Scotia (Canada), provided by The National Research Council of Canada, Institute for Marine Biosciences (Halifax, Nova Scotia). The structures of the investigated analytes, together with their certified amounts in the sample, are reported in Table 1. All of the experiments were carried out on 30-g subsamples. Soxhlet and MAME extractions were run on both the dried sediment and the wet sediment, previously spiked with PAHs standards according to the procedure described below. MAME Procedure and Extracts Treatment. A CEM MDS2000, 950 W microwave sample preparation system was used in all of the reported experiments. Different amounts of the dried (1.0 and 0.6 g) or of the wet sediment (1.2 g) were heated in the presence of 10.0 mL of 0.020 M surfactant solution (C12E10 and Brij 35). The experiments were performed in closed PTFE vessels, and the working temperature was 90 °C, a value below the reported cloud point of the nonionic surfactant chosen for most experiments. Six liners at a time were irradiated, applying 20% of the maximum microwave irradiation power during the first 5 min of the treatment and 15% power for the remaining time. This parameter is particularly significant, because the irradiation stops when the predefined temperature value has been reached, and it restarts only when the temperature decreases below this value. The optimization of the irradiation power allows the achievement of the desired T with the maximum number of microwave activations. After the irradiation, the liners were cooled to 25 °C, then opened and centrifuged at 1500 rpm for 20 min using a Hermle Z380 centrifuge mounting a swing-out rotor with an operating radius of 17.2 cm. The solution was transferred to a 25.0-mL graduated flask, and two further irradiation cycles of 5 min each were run on the solid phase to which 5.0 mL of surfactant solution were added (sediment washing steps). The collected extracts were filtered through a cellulose acetate membrane (HA 0.45 µm, Millipore) prior to analysis, and a glass syringe was employed in order to reduce the PAHs’ adsorption. Soxhlet Extraction. Approximately 1.0 g of sediment was directly weighted in the cellulose thimble and then introduced into the Soxhlet extraction chamber. Dichloromethane was chosen as the solvent on the basis of previous studies10 performed on similar sediment samples; 100.0 mL of this solvent were added and brought to boiling for 16 h. The extract was cooled at room temperature and kept at 4 °C. 3792

Analytical Chemistry, Vol. 73, No. 15, August 1, 2001

A solvent switch from CH2Cl2 to CH3CN was then performed according to the following procedure: 4.0 mL extract was concentrated to 1.0 mL under a nitrogen stream; 1.0 mL of CH3CN was added, and the total volume of the mix was again reduced to 0.2 mL under a nitrogen stream; last, the volume was brought to 1.0 mL with CH3CN. These final solutions were filtered through a 0.45-µm membrane (Durapore HV, Millipore). PAHs Determination. PAHs were determined by HPLC with fluorescence detection, using a Kontron instrument equipped with model 420 and model 422 pumps and a spectrofluorimetric detector (model Spectra-System FL-200, Thermo Separation Products). Table 2 reports the optimized detection program employed. A Lichro CART 250-4 Lichrospher PAHs (Merck) column (25-cm length, 4-mm i.d., and 4-µm-diameter particles) was used together with a Lichro CART 4-4 Lichrospher PAHs precolumn. A gradient elution program using H2O (A) and CH3CN (B) was applied according to the following scheme:

Two procedures were adopted to prepare the standards for the calibration curve, starting from a stock solution containing the PAHs mixture. Dilutions of the stock mixture were made using aqueous surfactant solution prior to the analysis of samples coming from MAME treatments; whereas, CH2Cl2 was employed to dilute the stock mixture (followed by the previously described solvent switching) for the samples coming from Soxhlet extractions. In all of the investigated cases, the calibration curves were performed in the PAHs concentration range of 0.01 ÷ 0.1 ppm. A comparison of the chromatograms of PAHs obtained from standard solutions with those of PAHs contained in a surfactant extract is shown in Figure 1. No effects attributable to the surfactant were observed on peak resolution, elution order and elution times. Samples Conditioning and Spiking. To investigate the matrix effect on the extraction efficiency, some experiments were run on aliquots of sediment enriched using PAHs standards. The chromatographic signals (peak areas) vs the concentrations of PAHs added to the solid samples were plotted, and the PAHs content in the sediment was calculated according to the standard addition procedure. The spiking procedure is a crucial point, because it should reproduce as well as possible the solutesediment adsorption taking place in the natural environment. In our case, aliquots of ∼6.0 g of sediment were placed in centrifuge glass tubes with 6.0 mL of deionized water. Proper volumes of standard PAHs solution (20 mg L-1 in CH3CN) were added to each sample, and the resulting suspensions were kept 24 h under stirring at 4 °C. Afterwards, the tubes were centrifuged at 1000 rpm for 15 min. After the separation of the upper layer, the spiked

Figure 1. HPLC profile of a PAHs standard mixture (concentration of each analyte, 0.05 mg L-1) in 0.020 M Brij 35. Inset: HPLC profile of PAHs in the surfactant extract. Analyte numbers as in Table 1.

sediments were kept at 4.0 °C. RESULTS AND DISCUSSION Selection of the Extracting Micellar Phase. Before examining the influence of parameters such as the pH of the extracting solution, solution/sediment ratio, and matrix effect on the extraction efficiency, two nonionic surfactants were tested. Both amphiphiles possess the same hydrophobic moiety and a comparable critical micellar concentration value, but have a different number of oxyethylene units. To speed up the analytical procedure, a

reduced detection program was adopted and applied to analyze a group of eight components. The amount of the recovered analytes (expressed as PAHs concentration in the sediment) reported in Figure 2 shows that Brij 35 is more efficient than C12E10 for the examined sample. The reported values correspond to the mean of three independent experiments. Extraction times were also varied, and the best results were obtained irradiating for 25 min. Effect of pH. The pH can usually influence the extraction of organic molecules from solid matrixes, even if the considered substrates do not participate in acid-base equilibrium. In fact, at definite pH values, the bonds between organic and inorganic components of the sediment could be broken, leading to an easier release of the analytes in the extracting solution.27 Experiments were thus run, in addition to those performed at pH 6.5 (pH of the surfactant solution) by adjusting the pH to 11.0 and 3.0 using NaOH and HCl, respectively. The obtained results showed no significant modifications of the extraction yields; only in two cases (dibenzo[a,h]anthracene and benzo[g,h,i]perylene) was the analyte recovery significantly lower when operating at pH 6.5; however, the calculated concentrations still fall in the certified range for the above-mentioned compounds. Sediment/Solvent Ratio. Two extractions were performed on different amounts of weighted sediment (0.6 and 1.0 g) keeping the volume of the surfactant solution constant (20 mL). The results are reported (Figure 3) as the mean values of 6 independent extractions ( standard deviation. It is possible to observe that the extraction efficiency slightly increases when working with the higher solvent/sediment ratio; a PAHs recovery ranging between 44 and 100% was estimated, with a recovered amount falling in the certified concentration range for five compounds. It is noteworthy that the MAME procedure gave rise to more precise results than those reported for the certified sample. However, because the certified concentration values were not attained using direct extraction, further investigations were necessary in order to understand the reason and overcome the problem. Soxhlet extractions were performed for comparison purposes. Also in this case, the PAHs analysis yielded results (mean values of 5 independent extractions ( standard deviation) below the certified interval for four of the investigated compounds.

Figure 2. Comparison of extraction efficiency using C12E10 and Brij 35.

Analytical Chemistry, Vol. 73, No. 15, August 1, 2001

3793

Figure 3. Comparison between certified and found data: (A) 1.0 g of sediment, (B) 0.6 g of sediment, (C) certified PAHs content, and (D) Soxhlet extraction. Table 3. Comparison between Calibration Curves with External Standards and Standard Additions on the Extracts analyte

standard additions

R2

external standards

R2

phenanthrene anthracene benzo[b]fluoranthene benzo[k]fluoranthene benzo[a]pyrene

y ) 429.29 x + 15.15 y ) 4789.60 x + 53.00 y ) 277.33 x + 13.19 y ) 1163.60 x + 29.41 y ) 449.54 x + 11.71

0.995 0.999 0.998 0.999 0.989

y ) 425.49 x + 0.18 y ) 4751.20 x + 4.26 y ) 279.96 x + 0.28 y ) 1204.90 x + 1.06 y ) 452.37 x + 0.11

0.999 1 0.999 0.999 1

Evaluation of Matrix Effects. Because of the nature of the organic fraction present in the sediments, the release of fluorescence quenchers from the matrix during the extraction step cannot be excluded. In this case, PAHs quantification based on calibration curves with external standards would not be suitable. To check the presence of interfering species, the standard addition method was adopted to analyze the extracts, focusing attention on the compounds for which the calculated recovery did not attain the certified amount. Different amounts of PAHs standard solution were added to aliquots of the extracts prior to the sample filtration. Table 3 reports the obtained results in terms of regression equations and related regression coefficients, each of them calculated on five points; because the standard addition line and the calibration line show practically the same slope for each analyte, the matrix effects on the analytical step can be neglected. On the other hand, when micellar solutions are used, the extraction step can be considered the result of a competition between adsorption-desorption of analytes from the surfactantmodified solid phases and micellar solubilization. Because physical and structural properties of analytes and sediment components can play different roles in these processes, the variation of the extraction yields is expected on the basis of aging of the samples, (27) Buffle, J. Complexation Reactions in Aquatic System: an Analytical Approach; Ellis Horwood: Chichester, 1988; Chapter 3.

3794 Analytical Chemistry, Vol. 73, No. 15, August 1, 2001

Table 4. MAME-based Analysis of PAHs Using the Standard Addition Method on Spiked Sediment Samples

analyte

determined value, MAME mg/Kga

certified value mg/Kg

determined value, Soxhlet mg/Kg

phenanthrene anthracene fluoranthene pyrene benzo[b]fluoranthene benzo[k]fluoranthene benzo[a]pyrene dibenzo(a,h)anthracene benzo(g,h,i)perylene indeno(1,2,3-c,d)pyrene

2.50 ( 0.10 1.08 ( 0.12 3.89 ( 0.18 2.85 ( 0.15 2.35 ( 0.13 1.26 ( 0.17 1.86 ( 0.13 0.39 ( 0.07 2.30 ( 0.36 1.40 ( 0.23

3.0 ( 0.6 1.1 ( 0.4 3.54 ( 0.65 3,0 ( 0.6 2.8 ( 0.6 1.43 ( 0.15 2.2 ( 0.4 0.49 ( 0.16 1.78 ( 0.72 1.95 ( 0.58

3,60 1,16 3,37 3,40 2,66 1,20 1,57 0,49 2,31 1,97

a

mean values of 4 independent extractions ( the standard deviation.

matrix composition (organic carbon percent, clay and oxides content, etc.) changes, and laboratory conditioning treatments. To overcome these effects, experiments were performed on aliquots of 1.2 g of wet sediment, which corresponds to 0.70 g of dry sediment, spiked with PAHs. The analytes quantification was performed according to the standard addition method, as previously described. Figure 4 shows the results obtained for benzo[b]fluoranthene and phenantrene; the significant difference be-

the standard addition procedure yielded higher PAHs recoveries (see Table 4); only two replicate experiments were performed for comparison purposes, and the reported results correspond to the mean values.

Figure 4. Determination of benzo[b]fluoranthene (A) and phenantrene (B). Matrix effect on the extracting step.

tween the slopes of the lines obtained from a calibration with external standards and from the standard additions on sediment sub-samples, confirms the existence of relevant matrix effects. Therefore, the standard addition method was applied to quantify all the analytes. The results summarized in Table 4, falling in the certified range for all of the examined PAHs, indicate that the chosen procedure is suitable. In addition, for Soxhlet extraction,

CONCLUSIONS The above examined results indicate that the proposed MAME approach is suitable for the quantitative determination of PAHs present in marine sediments as a result of the times required for the analysis being shorter than those necessary using Soxhletbased methods. Because of the observed efficiency and good reproducibility, the procedure could also be proposed for the analysis of other types of environmental solid matrixes, such as soils or river sediments. Moreover, the use of aqueous surfactant solutions instead of organic solvents is generally safer and cheaper (the estimated cost is about 90× lower than extraction with CH2Cl2 in the examined case). Another potential advantage is related to the possibly easier disposal of the laboratory wastes that are produced during the analysis. In fact, the removal of the undesirable toxic residues could be further performed by applying certain degradation procedures, such as photocatalysis, which cannot be used for the treatment of organic solvent wastes. This last aspect is of particular interest in the light of development of environmentally friendly analytical protocols. ACKNOWLEDGMENT Financial support from MURST (Rome) is gratefully acknowledged. Received for review March 14, 2001. Accepted May 3, 2001. AC010302Z

Analytical Chemistry, Vol. 73, No. 15, August 1, 2001

3795