Anal. Chem. 2007, 79, 4135-4140
Ultrasensitive Determination of Polycyclic Aromatic Compounds with Atmospheric-Pressure Laser Ionization as an Interface for GC/MS R. Schiewek,† M. Schellentra 1 ger,† R. Mo 1 nnikes,† M. Lorenz,‡ R. Giese,‡ K. J. Brockmann,‡ S. Ga 1 b,† ‡ ,† Th. Benter, and O. J. Schmitz*
Division of Analytical Chemistry and Division of Physical Chemistry, University of Wuppertal, Gauss-Strasse 20, 42119 Wuppertal, Germany
Recently we introduced atmospheric pressure laser ionization (APLI) as a complementary ionization method for coupling LC-MS systems (HPLC and CEC), allowing ionization of nonpolar aromatic compounds via nearresonant two-photon excitation. In this paper, we demonstrate that APLI with the same source enclosure as for LC coupling is also suited for hyphenation of GC with atmospheric-pressure ionization mass spectrometry. This technique permits the qualitative and quantitative determination of aromatic compounds in an ultralow concentration range, as we show here with polycyclic aromatic hydrocarbons (PAHs), alkylated PAHs, and hetero-PAHs as examples. The outstanding sensitivity is demonstrated for chrysene, with a detection limit of 22 amol. Polar functional groups reduce the sensitivity, but after methylation or silylation, the analytes can also be determined very sensitively in complex matrixes, as is shown with 1-hydroxypyrene in urine. Discroll and Spaziani used photoionization detection (PID) in gas chromatography as early as 1976.1 A 10.2-eV lamp was employed as the light source for these detectors.2-4 Comparison with the flame ionization detector (FID) revealed that the PID was more sensitive for unsaturated aliphatics and especially for aromatic compounds.4 Nevertheless, to date neither the PID nor ion mobility spectrometry, based on PID development with photoionization,5 has been widely accepted. The reason may be that with PID, unlike FID, the number of carbon atoms in a hydrocarbon molecule is not proportional to the relative molar response, so that quantitative analysis is rendered more complicated.6 This problem can be overcome by use of isotope standards, which is why new methods of photoionization such as atmospheric* Corresponding author. Phone: +49-202-439-2492. Fax: +49-202-439-3915. E-mail:
[email protected]. † Division of Analytical Chemistry. ‡ Division of Physical Chemistry. (1) Driscoll, J. N.; Spaziani, F. F. Res./Dev. 1976, 27, 50-54. (2) Freedman, A. N. J. Chromatogr. 1980, 190, 263-273. (3) Senum, G. I. J. Chromatogr. 1981, 205, 413-418. (4) Cox, R. D.; Earp, R. F. Anal. Chem. 1982, 54, 2265-2270. (5) Baim, M. A.; Eatherton, R. L.; Hill, H. H. Anal. Chem. 1983, 55, 17611766. (6) Cohen, M. J.; Karasek, F. W. J. Chromatogr. Sci. 1970, 8, 330-337. 10.1021/ac0700631 CCC: $37.00 Published on Web 05/02/2007
© 2007 American Chemical Society
pressure photoionization (APPI) and dopant-assisted (DA) APPI, in combination with mass spectrometric detection, are again the focus of intensive research.7,8 However, both APPI and DA-APPI have only been applied with liquid chromatographic methods such as HPLC and CE. Recently, we described atmospheric-pressure laser ionization as a novel method for liquid chromatography-mass spectrometry.9,10 The technique is based on resonantly enhanced multiphoton ionization (REMPI) at atmospheric pressure with two-photon excitation. As an extension of this work, we present here for the first time GC/MS with laser ionization at atmospheric pressure. EXPERIMENTAL SECTION Chemicals and Samples. Acetonitrile, phenanthrene, 2-methylanthracene, 2-ethylanthracene, pyrene, chrysene, benzo[a]pyrene, and 1-hydroxypyrene were obtained from Sigma Aldrich (Deisenhofen, Germany). N-Methyl-N-trimethylsilyltrifluoracetamide (MSTFA), N,O-bistrimethylsilyltrifluoracetamide (BSTFA), and trimethylsulfonium hydroxide (TMSH) (0.2 mol/L in methanol) were delivered by Macherey-Nagel (Dueren, Germany). Toluene was purchased from Merck (Darmstadt, Germany). All solvents were of chromatographic or analytical purity, and all other chemicals were of the highest purity available and were used without further purification. Dibenzofuran, 9H-carbazole, 11Hbenzo[a]carbazole, and 5,11-dihydro-6H-benzo[a]carbazole were provided by Dr. Ryan P. Rodgers (Florida State University) and 1-hydroxypyrene and urine samples were from Prof. Scherer (ABF GmbH, Munich, Germany). Sample Preparation. Stock solutions of the chemicals mentioned above were made by dissolving the compounds in toluene (polycyclic aromatic hydrocarbons (PAHs) and heterocyclic PAHs) or in acetonitrile (1-hydroxypyrene) to obtain a concentration of 100 mg/L. Final working solutions and mixtures were prepared either by stepwise dilution of a stock solution or by (7) Syage, J. A.; Evans, M. D. Spectroscopy 2001, 16, 15-21. (8) Robb, D. B.; Covery, T. R.; Bruins, A. P. Anal. Chem. 2000, 72, 36533659. (9) Constapel, M.; Schellentra¨ger. M.; Schmitz, O. J.; Ga¨b, S.; Brockmann, K.J.; Giese, R.; Benter, Th. Rapid Commun. Mass Spectrom. 2005, 19, 326336. (10) Droste, S.; Schellentra¨ger, M.; Constapel, M.; Ga¨b, S.; Lorenz, M.; Brockmann, K.-J.; Benter, Th.; Lubda, D.; Schmitz, O. J. Electrophoresis 2005, 26, 4098-4103.
Analytical Chemistry, Vol. 79, No. 11, June 1, 2007 4135
mixing stock solutions followed by stepwise dilution (in the case of the PAH mixture) with toluene or acetonitrile. For derivatization, a 1-hydroxypyrene standard was silylated as well as methylated. For silylation, a solution of 10 mg of 1-hydroxypyrene in 400 µL of a 5:1 (v/v) mixture of BSTFA and MSTFA was stored for 1 h at 80 °C in a drying oven. Before GC analysis, the sample was diluted stepwise with toluene to a final concentration of 6.6 µg/L. For methylation, 66 µL of a 1-hydroxypyrene solution (10 µg/L) was mixed with 33 µL of a TMSH solution (0.2 mol/L in methanol) and injected into the GC without further preparation. The urine sample (5 mL) was spiked with 50 µL of 1-hydroxypyrene solution (50 µg/L in acetonitrile) to obtain a final concentration of 500 ng/L. It was then concentrated by solid-phase extraction (SPE). The SPE column (ABS-Elut, Nexus, 3 mL, Varian, Darmstadt, Germany) was conditioned by washing once with 3 mL of acetonitrile and than twice with 3 mL of deionized water. The spiked urine sample was transferred onto the SPE column, and washed with 1 mL of deionized water. Acetonitrile (1.5 mL) was used to elute the analytes. For further sample preparation, the acetonitrile was evaporated in a gentle stream of dry N2, and the residue was dissolved in 50 µL of a 5:1 (v/v) BSTFA/MSTFA mixture. The mixture was stored for 1 h at 80 °C in a drying oven. Prior to the GC analysis, the sample was diluted 1:20 with toluene. Instrumentation. The instrumental setup consisted of the gas chromatograph, a TOF mass spectrometer, a homemade transfer line, and an excimer laser light source. The components are briefly described below. GC-APLI-(TOF)MS. All GC experiments with APLI were performed with a DANI 1000 MPC gas chromatograph from Dani (Monza, Italy). The instrument was equipped with a factorFour VF-5ms column (30 m × 0.25 mm i.d., 0.25-µm film thickness) from Varian (Darmstadt, Germany). Nitrogen (99.999% purity, Air Liquide, Duesseldorf, Germany) was used as carrier gas maintained at a constant pressure of 1.90 bar. The sample injections (1 µL) were carried out manually and splitless at 310 °C injector temperature with a microliter syringe (Hamilton, Bonaduz, Switzerland). The temperature program was 60 °C (hold 1 min), 20 °C/min to 265 °C, and 35 °C/min to 340 °C (hold 10 min) in the case of the PAH mixture and 1-hydroxypyrene. For analysis of the urine sample, the program was 60 °C (hold 1 min), 15 °C/ min to 265 °C, and 25 °C/min to 320 °C (hold 10 min) for the urine sample. The homemade transfer line (cf. Figure 1), built by the machine shop of the chemistry department at the University of Wuppertal, was heated by independent resistance heating wires in the two zones, i.e., the probe and the transfer section. The temperatures were controlled individually with a microprocessorbased temperature controller (JumoITRON, M.K. Juchheim GmbH, Fulda, Germany) equipped with K-type thermocouples (Newport Electronics GmbH, Deckenpfronn, Germany). For all experiments, the temperature was maintained at 320 °C in both zones of the transfer line. The gas flow (nitrogen, 15 L/h) through the coaxial holes around the outlet of the GC capillary (see below) at the probe tip was controlled by means of a calibrated 1179A mass-flow controller (MKS Instruments, Wilmington, MA). 4136 Analytical Chemistry, Vol. 79, No. 11, June 1, 2007
Figure 1. Photograph of the entire GC-APLI-(TOF)MS setup used in the present experiments.
The mass spectrometer was a micrOTOF (Bruker Daltonics, Bremen, Germany) equipped with an APLI source built in-house; see below. It was operated in the positive-ion mode. The scan range was 160-310 amu for the PAH mixture and 270-310 amu for the urine sample. The nebulizer gas flow directed through the sprayer was operated at 3 bar corresponding to ∼105 L/h, as measured with a FM-360 mass flow meter (Tylan GmbH, Munich, Germany). GC-EI-(TOF)MS. The GC experiments with EI were performed with an HP 6890N gas chromatograph (Agilent, Bo¨blingen, Germany). The instrument was equipped with the same column used in the GC-APLI-(TOF)MS experiments. Helium (99.999% purity, Air Liquide) was used as carrier gas at a constant flow of 1.4 mL/min. All sample injections (1 µL) were carried out with an autosampler (Agilent 7683 Series) and splitless at 330 °C injector temperature. The temperature program was 40 °C (hold 1 min), 40 °C/min to 100 °C, and 5 °C/min to 320 °C (hold 12 min). The mass spectrometer was a Pegasus IV (TOF)MS (Leco, Moenchengladbach, Germany) with standard electron impact ionization at 70 eV. The scan range of the data acquisition system was set to 50-300 amu. Laser System. The light source employed for APLI experiments was a compact OPTex excimer laser from Lambda Physik (Goettingen, Germany) running at 248 nm (KrF*). Operating conditions in all experiments were a repetition rate of 200 Hz, 8-mJ pulse energy, and 8-ns pulse duration. To provide the required laser safety for operation of a class 4 system, the UV light beam was fully enclosed on its way from the excimer laser to the APLI source. The stage consisted of a laser mount, an optical cube equipped with a tilting steering plate carrying a dielectric mirror with a reflectivity of >99.8% at 248 nm, a beam manipulation stage with a fused-silica lens (nominal focal length 20 mm) for beam positioning and shaping, and the ion source mount. All elements were connected with PVC tubing of 3-mm wall thickness. The irradiated laser beam of ∼1 cm2 area was focused to a final spot size of roughly 0.25 cm2 and was directed in the vicinity of the spray shield of the MS sampling orifice, slightly offset with respect to the main axis of the ion-transfer capillary.
Figure 2. Photographs of the constructed transfer line (A), the capillary inlet stage (B), the APLI source (C), and the optical device including laser-beam enclosure (D).
APLI Source Enclosure. The source enclosure was made from a solid aluminum piece matching the outer dimensions of the original Bruker Apollo source on the MS side. Since the Apollo source is mounted on two hinges to the MS body and locked into position with a simple spring-loaded latch, the two source types can be rapidly changed. Two fused-silica windows serve as laser beam entry and exit; for safety reasons, the exit port was covered on the outside with a metal plate for routine operation. For ease of laser beam alignment, this cover was removed. Additional ports at the front and top of the enclosure can be used for other purposes; in the present experiments, the LC sprayer (see Results and Discussion section) was mounted on a linear motion stage on the top port. The exit port was connected to an exhaust line installed in the laboratory. RESULTS AND DISCUSSION APLI is based on two-photon REMPI at atmospheric pressure and allows the sensitive and selective ionization of numerous lipophilic aromatic compounds:
M + m hν f M*
(1)
M* + n hν f M•+ + e-
(1b)
Reactions 1a and b represent a classical (m + n) REMPI process, which is used very favorably with n ) m ) 1 for the ionization of aromatic hydrocarbons (e.g., PAHs; for a recent review see ref 11). At room temperature, the absorption features of aromatic hydrocarbons are relatively broad,12 and because of the high molecular absorptions coefficients in the near-UV and the relatively long lifetimes of the first electronically excited S1 and S2 states,13 fixed-wavelength light sources such as excimer lasers can be used (11) Boesl, U. J. Mass Spectrom. 2000, 35, 289-304. (12) Haeflinger, P. O.; Zenobi, R. Anal. Chem. 1998, 70, 2660-2665.
for excitation. The compounds investigated in the present work were favorably ionized with 248-nm radiation. A great advantage of APLI versus APPI is that neither oxygen nor the solvents typically used in LC (e.g., methanol or acetonitrile) absorb significantly at the applied wavelength. As a result, the photon density of the beam essentially remains constant in the propagation direction. This is in stark contrast to APPI, where the 10-eV VUV photons (λ ) 124 nm) have typical penetration depths of only a few millimeters.14-16 A significant reduction of the VUV photon density will most probably also be observed in GC-APPI applications because of the high molecular absorption coefficient of oxygen at 10 eV of ∼2 × 10-17 cm2/molecule.17 O2 is always present in the makeup gases, and even trace amounts will severely attenuate the VUV photon density. Moreover, upon absorption in this wavelength range, O2 is photolyzed with quantum yields approaching unity, leading to highly excited reactive oxygen atoms. Neither the photolysis of oxygen molecules nor a noticeable uptake of energy by other matrix compounds is observed by APLI with the gases and solvents mentioned above. In the case of polar compounds dissolved in CH3CN/H2O mixtures, formation of quasi-molecular ions [M + H]+ is frequently observed with APPI, while rather nonpolar compounds such as naphthalene predominantly form the radical cations M•+.18 A detailed mechanism for the formation of [M + H]+ has been (13) Handbook of Photochemistry, 2nd ed.; Murov, S. L., Carmichael, I., Hug, G. L., Eds.; Marcel Dekker: New York, 1999. (14) Robb, D. B.; Blades, M. W. J. Am. Soc. Mass Spectrom. 2005, 16, 12751290. (15) Robb, D. B.; Blades, M. W. J. Am. Soc. Mass Spectrom. 2006, 17, 130-38. (16) Lorenz, M.; Brockmann, K.-J.; Schiewek, R.; Constapel, M.; Schmitz, O. J.; Ga¨b, S.; Benter, Th. On the ionization mechanisms in APPI vs. APLI. In Proceedings of the 54th ASMS Conference on Mass Spectrometry and Allied Topics, Seattle, WA, May 28-June 1, 2006. (17) Chan, W. F.; Cooper, G.; Brion, C. E. Chem. Phys. 1993, 170, 99-109. (18) Raffaelli, A.; Saba, A. Mass Spectrom. Rev. 2003, 22, 318-331.
Analytical Chemistry, Vol. 79, No. 11, June 1, 2007
4137
Figure 3. (A) Mass-resolved chromatograms with GC-EI-(TOF)MS as observed in the analysis of a test mixture containing dibenzofuran (1), phenanthrene (2), 9H-carbazole (3), 2-methylanthracene (4), 2-ethylanthracene (5), pyrene (6), 5,11-dihydro-6H-benzo[a]carbazole (7), 11Hbenzo[a]carbazole (8), and benzo[a]pyrene (9). All traces result from monitoring the parent ions. Sample concentration: PAHs (each 100 µg/L), alkyl-PAHs (100 µg/L), and heterocyclic PAHs (1000 µg/L); Injection, 2 µL. (B) Chromatogram (TIC) with GC-APLI-(TOF)MS of the same mixture but at one-tenth of the concentration. In all cases, the radical cation was monitored. Injection: 1 µL
discussed by Syage,19 who used a combined thermodynamic and kinetic approach. In that model, the equilibrium (2) is shifted toward the product side by the large excess of solvent molecules present, facilitating H-atom abstraction from matrix compounds by the initially formed radical cations:
M•+ + S f [M + H]+ + [S - H]•
(2)
In APLI, the laser beam is positioned close to the cone, and 4138
Analytical Chemistry, Vol. 79, No. 11, June 1, 2007
H-abstraction reactions are kinetically suppressed, since the transition time of initially generated charge carriers to the sampling orifice is very short (well below 100 µs). Only analytes with rather high gas-phase basicities are converted measurably to quasi-molecular ions. In contrast hereto, in APPI the transition time of initially formed ions in the vicinity of the lamp exit window (19) Syage, J. A. J. Am. Soc. Mass Spectrom. 2004, 15, 1521-1533.
to the sampling orifice is at least 1 or 2 orders of magnitude longer, so that extensive ion-molecule chemistry may occur.16,20 Therefore, it is an inestimable advantage that the laser beam can be positioned precisely at each desired point in the ionization source. This is hardly possible with the krypton lamp used in APPI. Through variation of the laser beam position, the optimal ionization region for maximum sensitivity is easily determined. This position is not, as anticipated, directly in front of the MS sampling orifice, but slightly offset with respect to the main axis of the ion-transfer capillary of the micrOTOF. This surprising result has been further investigated in separate experiments, and the results from the study will be presented in an upcoming paper. These considerations reveal the advantages and possibilities of APLI over APPI as an ionization method for GC/MS applications. Construction of a GC-APLI-MS Interface. APLI as an ionization interface has been coupled with HPLC and CE.9,10 In order to also use this interface in GC, we have constructed a flexible, temperature-controlled transfer line of ∼1-m length to couple the separation capillary from the GC oven to the APLI source without condensation of the analytes. Because of the different heat emission characteristics in the two transfer line stages (supply of the capillary to the source and the inlet inside the source enclosure), these regions have to be heated separately with two independent heating wires. Figure 1 shows the complete setup of the GC-APLI-(TOF)MS including the excimer laser. To realize the flexible transfer line, the downstream terminal of the separation capillary, i.e., the last 10 cm inside the GC oven, was inserted into a steel capillary with an outer diameter of 2 mm and inner diameter of 1 mm, respectively. The steel capillary was covered with ceramic fiber tape to ensure electrical isolation from the heating wire, which is wound around it. This assembly was covered with a second layer of ceramic fiber tape and inserted into a corrugated metal hose, whose downstream end was connected with a flange on a custom stainless steel T-piece (cf. Figure 2A). The whole assembly was clamped to a second flange on the T-piece (cf. Figure 2B). While the first heating wire heats the steel capillary from the GC oven to the T-piece, the second wire heats the T-piece to the downstream end. The third port of the T-piece serves as electrical feed through for the heating wires and the two thermocouples (cf. Figure 2A). The GC carrier gas flow is encompassed by an additional N2 gas flow, which leaves the GC probe tip coaxially around the GC capillary. The capillary protrudes ∼2 mm from the centric bore of the probe tip and is located ∼5 mm upstream of the MS sampling orifice. The LC sprayer (see Experimental Section) mounted on top of the source enclosure is used as an additional N2 supply to provide a makeup gas flow, which is adjusted to a value that, including all other flows, matches the pumping speed at the MS inlet capillary. Comparison of GC-EI-MS with GC-APLI-MS: PAH Mixture. To obtain an estimate of the sensitivity of the gas chromatographic PAH analysis with APLI- (TOF)MS detection, a test mixture of PAHs, alkylated PAHs, and heterocyclic PAHs was first analyzed by GC-(TOF)MS with electron impact ionization (EI). Figure 3A shows the mass traces of each molecular ion. The TIC of the whole analysis is not presented because of the interfering (20) Lorenz, M.; Schiewek, R.; Brockmann, K.-J.; Schmitz, O. J.; Ga¨b, S.; Benter, Th. Submitted to J. Am. Soc. Mass Spectrom.
Figure 4. GC-APLI-(TOF)MS analysis of a solution of 5 ng/L chrysene dissolved in toluene. The mass trace representing the radical cation recorded at 228 amu is shown. The chromatograms show the solvent (toluene, black line) and the sample (5 ng/L chrysene, dotted line).
signals from column bleeding and contaminants in the toluene solvent. The TIC of the GC-APLI-(TOF)MS analysis of the same mixture, but diluted 10-fold, is shown in Figure 3B. In addition to the clear increase in sensitivity of the APLI, a further advantage in comparison to EI is identified: GC-typical signals of column bleeding at elevated temperatures, which often complicate the analysis of high-boiling compounds, are strongly suppressed with APLI because the compounds bleeding out lack detectable groups. Sensitivity and Linearity. In one-dimensional GC-(TOF)MS with electron impact ionization, the limit of detection (signal greater than three times the standard deviation of the noise level21) for PAHs and alkylated PAHs is in the range of 10-100 µg/L, whereas for heterocyclic PAHs the detection limit is ∼10 times as high. In comprehensive two-dimensional gas chromatography (GC×GC), the peak capacity and the efficiency are higher than in one-dimensional GC, which leads to a 10-fold increase in sensitivity with an identical TOF-MS detector and the same ionization mode (data not shown). With one-dimensional GC-APLI(TOF)MS, a further improvement in sensitivity, i.e., a factor of 100 as compared to GC×GC-EI-(TOF)MS and a factor of 1000 as compared to GC-EI-(TOF)MS, was realized. The limit of detection21 of chrysene was determined in Figure 4 as ∼5 ng/L (22 pM), which translates to a detectable amount of 22 amol. The chrysene that is determined in the control sample may have resulted from contamination of the solvent or the GC system. Repeated analyses result in 4419 ( 414 area units for toluene and 7825 ( 608 area units for the sample (n ) 3). The linear range for this method (for chyrsene) was investigated with seven calibration mixtures between 0 and 10 µg/L (n ) 3). The area units determined for the solvent (toluene) were substracted from the total area units for the sample at each concentration. The data were subjected to a linear least-squares analysis and gave a correlation coefficient of 0.9998. Derivatization. Many PAHs are genotoxic carcinogens. One of these, pyrene, undergoes simple metabolism to 1-hydroxypyrene. 1-Hydroxypyrene and its glucuronide are excreted in (21) Kaiser, H.; Specker, H. Fresenius Z. Anal. Chem. 1956, 149, 46-56.
Analytical Chemistry, Vol. 79, No. 11, June 1, 2007
4139
respect to spectral position and absorption strength. Second, and much more important with respect to the very high collision rates prevailing at atmospheric pressure, is the enhanced reactivity of functionalized compounds in the ionic state. This is particularly the case for phenolic species in the presence of polar matrix compounds such as water and acetonitrile.23 Such observations are well-known from APPI where charge transfer from the primary analyte ions to solvent molecules frequently occurs.24 With GCAPLI-(TOF)MS, this appears far less likely but cannot be excluded, since trace amounts of water are always present in the makeup gases and thus in the source enclosure. In the present case of 1-hydroxypyrene, this loss of sensitivity as compared to pyrene measurements can be minimized by lowering the polarity of the functional group, either by methylation with TMSH in the injector or by silylation with a mixture of BSTFA and MSTFA at 80 °C (data not shown). As an application example of GC-APLI-(TOF)MS, the urine of smokers was analyzed with respect to the presence or absence of 1-hydroxypyrene. To this end, urine was spiked with 1-hydroxypyrene to a realistic concentration of 500 ng/L. SPE, enrichment, and silylation were carried out on 5 mL of this spiked urine (see Experimental Section) and resulted in a concentration of silylated 1-hydroxypyrene in the sample of 2.5 µg/L, assuming 100% recovery and quantitative silylation. The upper panel in Figure 5 shows the TIC of the gas chromatographic analysis. The middle panel shows a chromatogram of the mass trace of M•+ of hydroxypyrene (290 Da), and the lower panel shows a mass spectrum of the silylated 1-hydroxypyrene in a narrow mass range bracketing the parent ion region. An excellent signal-to-noise ratio is observed for the mass-resolved analysis. In contrast to EI-GC/ MS, the TIC observed with APLI-GC/MS shows well-resolved chromatographic signals well suited for in-depth analysis of the sample.
Figure 5. Panel A: TIC of the GC analysis of 1-hydroxypyrene in urine after SPE and silylation. Panel B: mass-resolved chromatogram (290 Da) of the parent ion of the silylated 1-hydroxypyrene. Panel C: APLI mass spectrum of the silylated 1-hydroxypyrene in the region of the parent mass.
urine. Biological monitoring of exposure to PAH has expanded rapidly since urinary 1-hydroxypyrene was suggested as a biological index of pyrene dosage. Since pyrene is always present in PAH mixtures, it is also used as an indirect indicator of all PAHs.22 However, the excellent sensitivity of the GC-APLI-(TOF)MS as demonstrated above deteriorates when the analytes of interest have polar functional groups attached to the aromatic ring system. Two main reasons are discussed to be responsible for this effect: First, the electronic properties of the undisturbed conjugated aromatic system may significantly change upon introduction of polar functional groups, affecting the absorption features with (22) Jongeneelen, F. J. Ann. Occup. Hyg. 2001, 45, 3-13.
4140
Analytical Chemistry, Vol. 79, No. 11, June 1, 2007
CONCLUSIONS GC-APLI-(TOF)MS allows a qualitative and quantitative analysis of PAHs in the ultratrace range, which is several orders of magnitude more sensitive than GC-EI-(TOF)MS. Furthermore, in many applications the sample preparation is simplified, and ion discrimination is significantly reduced because of the selective two-photon ionization used in APLI. It is emphasized that APLI is an ionization method which allows the coupling of electrophoretic, liquid and gas chromatographic separation systems with the same ion source enclosure. ACKNOWLEDGMENT We thank Gerhard Scherer (ABF GmbH Munich) for the 1-hydroxypyrene standard, the urine sample, and the enzymatic pretreatment of this. We also thank Ryan P. Rodgers for several hetero-PAH standard compounds. Received for review January 11, 2007. Accepted March 30, 2007. AC0700631 (23) An Index of the Literature for Bimolecular Gas Phase Cation-Molecule Reaction Kinetics; Anicich, V. G., Ed.; JPL Publication 03-19;; Jet Propulsion Laboratory, California Institute of Technology: Pasadena, CA, 2003. (24) Kauppila, T. J.; Kuuranne, T.; Meurer, E. C.; Eberlin, M. N.; Kotiaho, T.; Kostiainen, R. Anal. Chem. 2002, 74, 5470-5479.