Quantitative Determination of PAHs in Diesel Engine Exhausts by GC

Jul 7, 1999 - In the Laboratory. 962. Journal of Chemical Education • Vol. 76 No. ... of incomplete combustion of organic material. As they are ... in advance, students can complete the experiment in one laboratory ... two lab periods. During ...
79KB taille 2 téléchargements 223 vues
In the Laboratory

Quantitative Determination of PAHs in Diesel Engine Exhausts by GC–MS Paul Fleurat-Lessard* Laboratoire de Chimie Théorique, Université Paris XI, Bâtiment 490, F-91405 Orsay cedex; *[email protected] Karine Pointet and Marie-France Renou-Gonnord DCMR, URA CNRS 1307, Ecole Polytechnique, F-91128 Palaiseau

Polycyclic aromatic compounds (PACs) and their derivatives are toxic compounds (1) generally formed as the result of incomplete combustion of organic material. As they are stable molecules, they are widely distributed in the environment. There is a need to quantitate these molecules because some of them are carcinogenic. GC–MS is one of the most powerful analytical tools available for the chemical analysis of complex mixtures. The use of mass spectrometry enhances the capabilities of gas chromatography: the specific information provided in the mass spectrum makes the mass spectrometer a highly selective detector that can be used for qualitative analysis and structural determination. The mass spectrometer is among the most sensitive chromatographic detectors, having a detection limit below the picogram level, through the use of Selected Ion Monitoring (SIM) mode. Several protocols have already been proposed (2–4 ). However, they are long or usually not consistent with the limitations in time and budget of university laboratories. In this paper, we propose a quantitative analysis for the determination of PAHs (polycyclic aromatic hydrocarbons) in diesel engine exhaust, using a rapid cleanup procedure to reduce sampling time and consumption of adsorbant and solvents. If the instructor can determine the retention times of the molecules studied, prepare a SIM program for the chromatographic system to be used, and establish calibration curves in advance, students can complete the experiment in one laboratory period. This experiment could also be divided into two lab periods. During the first period, students perform the sample extraction and establish the calibration curves. During the second one, they perform the silica gel cleanup and qualitative and quantitative analysis.

Supelco (Bellefonte, PA, USA) as methylene chloride solutions. Internal standards for quantitation are a mixture of 7 deuterated PAHs diluted to 10 ng/µL in methylene chloride. A solution of chrysene d12 at the same concentration is used to quantitate the recovery yields. HPLC-grade dichloromethane is supplied by Burdick & Jackson (Fluka, Buchs, Switzerland). Pentane is provided by SDS (Vitry-sur-Seine, France). Silica gel 60, 0.063–0.100-mm particle size, is purchased from Merck (Darmstadt, Germany).

naphthalene

fluorene

pyrene

fluoranthene

benzo[b]fluoranthene

acenaphthylene

acenaphthene

anthracene

phenanthrene

benzo[a]anthracene

benzo[k]fluoranthene

chrysene

benzo[a]pyrene

Lab Summaries The analytical protocol involving GC–MS described here has proved its efficiency for quantitation of PAHs in diesel exhausts. For students, analyzing real samples is attractive and introduces difficulties of daily analytical work. During the laboratory period, students are introduced to the importance of experimental exactness through analytical criteria of trace analysis: limit of detection, repeatability, and recovery yields. Students should have at least a minimal background in chromatography and mass spectrometry. Experimental Procedure

Materials Analyses are performed on a gas chromatograph–mass spectrometer. Deuterated and native PAHs (U.S. EPA priority pollutants [2]) (see structures below) are purchased from 962

indeno[1,2,3-cd]pyrene

benzo[ghi]perylene

dibenzo[a,h]anthracene

16 U.S. EPA priority PAH pollutants

Gravitational columns are homemade from 5-mL disposable Corning pipets (Prolabo, Paris, France) after cutting the restricted end. Silica gel (1.5 g) is immobilized within the pipet tube on a glass wool plug. Diesel engine particle exhausts are sampled on filters according to the Standard Sampling Procedure (European Directive 94/12/CE). Diesel exhaust particulate matter, on which PAHs are adsorbed, is sampled from the exhaust manifold of a diesel engine vehicle through a dilution tunnel. A pump with a gas meter is used to pass a predetermined volume of exhaust gas through each glass-fiber filter. In this study, filters are supplied by a car company.

Journal of Chemical Education • Vol. 76 No. 7 July 1999 • JChemEd.chem.wisc.edu

Fluorene

166–167

6.8 →9.0

188–189

6.8 →9.0

Phenanthrene d10 Phenanthrene

178–179

Anthracene

178–179

Fluoranthene d10 Fluoranthene Pyrene d10

212–213 212–213 202–203

Chrysene

228–229

Benzo[a]anthracene

228–229

Benzo[a]pyrene d12

264–265

Benzo[ b]fluoranthene

252–253

Benzo[ k]fluoranthene

252–253

Benzo[a]pyrene

252–253

Benzo[ghi ]perylene d12

288–289

Benzo[ghi ]perylene

276–277

Indeno[1,2,3-cd ]pyrene

276–277

Dibenzo[ah]anthracene

278–279

chrysene

fluoranthene pyrene

phenanthrene

0 4

6

8

10

12

14

16

18

20

Time / min 9.0 →14.0

202–203

Pyrene

benzo[a]pyrene

152–153 154–155

5

benzo[k]fluoranthene

Acenaphthylene Acenaphthene

y x 20

anthracene

5.5 →6.8

fluorene

160–161

acenapthene

128–129

naphthalene

Naphthalene Acenaphthylene d8

10

Relative abundance

Table 1. Mass and Time Window MS Acquisition Parameters Standard Monitored Ion Time Window/ Analyte (Mass) min Naphthalene d8 136–137 0 →5.5

benzo[a]anthraene

In the Laboratory

9.0 →14.0

14.0 →18.0

18.0 →22.0

Figure 1. Chromatogram obtained in the SIM mode for the aromatic fraction of diesel exhaust.

band corresponds to oxygen- and nitrogen-containing PAHs and is not analyzed in this protocol. Concentrate the solvent under a nitrogen flow and add a last internal standard (40 ng of chrysene d12) to control recovery yields. Analyze the final extract in the scan mode to determine the structure of molecules in the aromatic fraction of the diesel particle exhausts. Quantitation of the 16 U.S. EPA priority pollutant PAHs is performed by analyzing this fraction in the SIM mode. Only molecular peaks of studied PAHs and internal standards are monitored within any given time window (5) (Table 1). An example of a chromatogram obtained in the SIM mode is presented on Figure 1. Results and Discussion

Extract Preparation First, deposit 40 ng of internal standards (4 µL of the internal standard solution) on the filters, which contain about 1 mg of diesel exhaust particles. After letting the filters soak in 350 mL of dichloromethane in a soxhlet apparatus for two hours, extract them in the soxhlet apparatus for 24 h. After extraction of the soluble organic fraction, concentrate the solvent to 1–3 mL in a rotary evaporator and continue the concentration to near dryness under a nitrogen flow. Then add 200 µ L of pentane. This step can be performed by students if the experiment is split into two sessions.

Ten groups of students have carried out this protocol. In Table 2 are reported the recovery yields obtained. In Table 3 are reported quantitative results (expressed in µ g/g of particles) for 4 representative molecules: naphthalene (for the smallest molecules) and phenanthrene, fluoranthene, and pyrene (the most abundant PAHs of the diesel exhaust particles). Within the larger groupings (A, B, C, D) the results of the individual groups can be compared because the students analyzed samples obtained from the same filters. The standard deviation of the recovery yields does not exceed 50% if results of group 8 are discarded (they are much below others). This standard deviation is acceptable, since our students are “analyst learners”. Recover y yields for acenaphthylene d8, phenanthrene d10, fluoranthene d10, and

Silica Gel Cleanup If the experiment is to be performed during a single session, the students’ work starts at this point. Wash the gravitaTable 2. Percentage Recover y Yields for Internal Standards tional column with 6 mL of Student Data, Percentage Recovery Yield pentane before use. Monitor the migration of aromatics along the Compound Group No. Av (SD) column with a UV lamp (λ = 365 1 2 3 4 5 6 7 8 9 10 nm). Elute nonpolar compounds Naphthalene d8 17 22 51 33 30 2 10 7 12 0 25 (14) with 6 mL of pentane (flash chro- Acenaphthylene d 5 5 5 2 1 0 9 1 2 2 9 6 4 9 3 1 1 1 6 1 1 5 7 2 (33) 8 matography). Add dichloro83 82 84 98 116 90 59 16 93 68 86 (16) Phenanthrene d10 methane until the aromatic com105 96 91 109 92 84 51 17 94 71 88 (18) pounds, which correspond to the Fluoranthene d10 6 102 97 111 87 83 47 1 6 94 67 86 (21) Pyrene d10 violet band, reach the bottom of a ] p y r e n e d 5 1 6 2 4 0 2 3 1 6 2 4 1 1 3 4 4 2 2 33 (17) B e n z o [ 12 the column. Then add another 3 g h i ] p e r y l e n e d 2 3 2 7 4 2 1 5 1 2 6 0 0 2 7 1 8 25 (8) B e n z o [ mL of dichloromethane to recover 12 all aromatic compounds. The blue N OTE: Italicized underlined values were discarded in calculating the average and SD.

JChemEd.chem.wisc.edu • Vol. 76 No. 7 July 1999 • Journal of Chemical Education

Instructor Data (%) 62 67 86 – 76 – –

963

In the Laboratory

pyrene d10 all lie between 60 and 120%, which corresponds to the usual acceptable interval for recovery yields. Moreover, students’ results are in good agreement with those obtained by the instructor (also reported in Table 2). Recovery yields for naphthalene d8, benzo[a]pyrene d12, and benzo[ghi]perylene d12 are below 60%. This increases the limit of detection compared to the instructor’s. These internal standards do not allow accurate quantitation but lead to an indicative result. Quantitative results can be compared only when the protocol has been carried out on the same filter by several groups. In this case, results are reproducible except those of group 3, which should be compared with the results of groups 4, 5, and 6. Relative amounts of PAHs are found to depend on the sampling operation cycle and time, which differs from one filter to another.

Table 3. Student Results for Quantitation of Representative PAHs Results for Group No./(µg g{1)a Compound

A 1

Naphthalene

18

B 2

C

3

4

5

22

36

31

27

6

7 9

20

D 8 –

9

10

63 –

Phenanthrene

173

132

209

99

84

85

245



93 9 4

Fluoranthene

138

130

55

11

20

21

35



90 8 1

Pyrene

144

140

77

14

14

14

20



88 6 1

aAbsurd

results are not reported. The 10 groups are further grouped into clusters A–D. Results within a cluster can be compared because students in these groups analyzed samples obtained from the same filters.

Notes for Instructor GC–MS analyses are performed on an HP 5890 gas chromatograph (Hewlett Packard, Avondale, PA, USA) connected to an HP 5989B mass spectrometer operated in the electron impact mode under standard conditions (electron energy 70 eV). A DB5 column 30-m long with 0.25-mm i.d. and 0.25 µm film thickness (J&W, Arcueil, France) is used with helium as a carrier gas at a linear velocity of 1 mL/ min. The injector temperature is set at 280 °C. The mass spectrometer source temperature and pressure are set at 250 °C and 5–6 × 10 {6 torr, respectively. The mass spectrometer quadrupole temperature is set at 100 °C. The column temperature program is monitored from 100 °C (hold 2 min) to 240 °C at 25 °C/min (no hold), and then to 320 °C at 5 °C/min. Seven fully deuterated PAHs are used as internal standards (Table 1) for quantitation. In Table 1, all compounds that follow a given internal standard will be quantitated relative to this internal standard. Quantitative analysis consists in comparing results obtained for unknown samples to results for known amounts of PAHs through calibration curves. These curves link the response ratio (response of a compound divided by the response of its internal standard) to the concentration ratio (concentration of the compound divided by the concentration of the internal standard). The response of a compound corresponds to its peak area. Curves are established by analyzing several solutions containing known amounts of PAHs and internal standards for each molecule to be quantitated (Fig. 2). Then unknown samples are analyzed and a quantitative report for PAHs is generated. These operations are automatically performed by the MS ChemStation software through a calibration database. This database consists of all the PAHs to be quantitated and all the internal standards, each characterized by its retention time (RT) and its molecular weight (MW). The MS ChemStation software allows the transfer of quantitative results from data analysis into the Microsoft Excel spreadsheet program for compilation of students’ results. Internal standard quantitation allows reliable analysis. Unknown amounts of PAHs are lost through the extraction and cleanup procedure. If the structures of internal standard and PAH are similar, it can be assumed that the concentration ratio is conserved throughout the protocol, since the loss of analytes through the extraction and cleanup procedure should be consistent with that of quantitation surrogates. Thus, 964

Figure 2. Calibration curve for phenanthrene vs phenanthrene d 10. Response ratio = 8.63e {001 × amount + 6.62e {002; R = .996; curve fit, linear.

knowing the final ratio and the initial amount of internal standard allows quantitation of PAHs. Toward that goal, the calculation of the recovery yield for quantitation surrogates constitutes a verification that the preparation steps of the whole analytical procedure have been properly carried out. Tips If budget constraints are a problem, only 3 standards can be used for the quantitation: naphthalene d8, phenanthrene d10, and benzo[a]pyrene d12. If filters cannot be obtained from the standard sampling procedure, particles can also be recovered at the end of the exhaust pipe. This less standardized sampling alternative has also been successfully tested, and the same compounds, in different proportions, have been found in both types of samples.

Journal of Chemical Education • Vol. 76 No. 7 July 1999 • JChemEd.chem.wisc.edu

In the Laboratory

Literature Cited 1. Santodonato, J.; Howard, P.; Basu, D. In Health and Ecological Assessment of PAH; Lee, S. D.; Grant, L., Eds.; Pathotox: Park Forest S., IL, 1981; p 177. 2. Winberry, W. T. Jr.; Jungclaus, G. In Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air, 2nd ed.; EPA/625/R-96010b; Center for Environmental Research

Information, U.S. Environmental Protection Agency: Cincinnati, OH, 1999; pp 13A1–13A78. 3. Williams, P. T.; Abbass, M. K.; Andrews, G. E. Combust. Flame 1989, 75, 1–24. 4. Veigl, E.; Posch, W.; Linder, W.; Tritthart, P. Chromatographia 1994, 38, 199–208. 5. Pointet, K.; Renou-Gonnord, M. F.; Milliet, A.; Jaudon, P. Bull. Soc. Chim. Fr. 1997, 134, 133–140

JChemEd.chem.wisc.edu • Vol. 76 No. 7 July 1999 • Journal of Chemical Education

965