Capillary Electrochromatography Coupled to Atmospheric Pressure

Benzo[a]pyrene, one of the most carcinogenic PAHs, has 12 monomethylated positional isomers (MBAPs). A strong correlation between the carcinogenicity ...
0 downloads 0 Views 309KB Size
Download PDF
Anal. Chem. 2006, 78, 6921-6927

Capillary Electrochromatography Coupled to Atmospheric Pressure Photoionization Mass Spectrometry for Methylated Benzo[a]pyrene Isomers Jie Zheng†,‡ and Shahab A. Shamsi*

Department of Chemistry, Center of Biotechnology and Drug Design, Georgia State University, Atlanta, Georgia 30303, and Abbott Laboratories, North Chicago, Illinois 60064

Benzo[a]pyrene, one of the most carcinogenic PAHs, has 12 monomethylated positional isomers (MBAPs). A strong correlation between the carcinogenicity of these isomers and methyl substitution has been reported. In this study, on-line coupling of capillary electrochromatography (CEC) and atmospheric pressure photoionization mass spectrometry (APPI-MS) provides a unique solution to highly selective separation and sensitive detection of MBAP isomers. The studies indicated that APPI provides significantly better sensitivity compared to electrospray ionization and atmospheric pressure chemical ionization modes of MS. A systematic investigation of APPI-MS detection parameters and CEC separation is established. First, several sheath liquid parameters (including type and concentration of volatile buffers, type and content of organic modifiers, use of dopants and inorganic/organic additives, and sheath liquid flow rate) and APPI-MS spray chamber parameters (capillary voltage, vaporizer temperature, nebulizer pressure) were found to have effects on detection sensitivity as well as the profile of mass spectrum. For example, when ammonium acetate was replaced with acetic acid in the sheath liquid, the MS signal was enhanced as much as 90% and the formation of ammonia adduct was effectively suppressed. Next, the separation of MBAP isomers was conducted on internal tapered columns packed with polymeric C18 stationary phase. With the use of a mobile phase consisting of slightly higher acetonitrile content (90%,v/v) and a small amount of tropylium ion, the analysis times were significantly shortened by 20 min without compromising the resolutions between the isomers. Finally, quantitative aspects of the CEC-APPI-MS method were demonstrated using 7-MBAP as the internal standard. The calibration curves of three of the most carcinogenic isomers, namely, 1-MBAP, 3-MBAP, and 11-MBAP, showed good linearity in the range of 2.5-50 µg/mL with a limit of detection at 400 ng/mL.

* Corresponding author. Phone: 404-651-1297. Fax: 404-651-2751. E-mail: [email protected]. † Center for Biotechnology and Drug Design. ‡ Abbott Laboratories. 10.1021/ac061024c CCC: $33.50 Published on Web 08/29/2006

© 2006 American Chemical Society

Benzo[a]pyrene (BaP) is one of the most carcinogenic polycyclic aromatic hydrocarbons (PAHs) and has been listed among the 16 priority pollutants by the Environmental Protection Agency. The monomethylated benzo[a]pyrene (MBAP) has 12 positional isomers since the methyl group can be substituted to 12 different positions around the BaP ring (Figure 1).1 The MBAP isomers have been found not only in the environment2 but also in vivo.3 A strong correlation between the carcinogenicity of these isomers and position of methyl substitution has been reported.4 For example, 1-, 3-, and 11-MBAP isomers are highly carcinogenic while 7-MBAP shows fairly low carcinogenicity.4-6 Hence, the development of chromatographic assays capable of resolving individual MBAP isomers with high sensitivity is desirable. The separations of MBAP isomers based on gas chromatography,7 high-performance liquid chromatography (HPLC),8 and cyclodextrin-modified micellar electrokinetic chromatography (CD-MEKC)9,10 have been reported. Recently, our group demonstrated the feasibility of applying capillary electrochromatography (CEC) to resolve these isomers.11,12 It was found that the CEC separation provided much higher efficiency than HPLC. Using a capillary column packed with 5-µm polymeric C18 stationary phase, 11 out of 12 isomers were resolved. In addition, this CEC method showed good compatibility with high-organic-content mobile phase, which is critical to improve the solubility and detection sensitivity of these hydrophobic PAHs. Although the CEC-UV method has demonstrated good resolving power for the MBAP isomers, several aspects still need to be improved. For example, the UV detection offers good sensitivity, (1) Harvey, R. G. Polycyclic Aromatic Hydrocarbons; Wiley-VCH: New York, 1997. (2) Santella, R.; Kinoshita, T.; Jeffrey, A. M. Mutat. Res. 1982, 104, 209-213. (3) Flesher, J. W.; Myers, S. R.; Stansbury, K. H. Carcinogenesis 1990, 11, 493496. (4) Silverman, B. D. Cancer Biochem. Biophys. 1981, 5, 207-212. (5) Iyer, R. P.; Lyga, J. W.; Secrist, J. A., 3rd; Daub, G. H.; Slaga, T. J. Cancer Res. 1980, 40, 1073-1076. (6) Utesch, D.; Glatt, H.; Oesch, F. Cancer Res. 1987, 47, 1509-1515. (7) Sander, L. C.; Schneider, M.; Wise, S. A.; Woolley, C. J. Microcolumn Sep. 1994, 6, 115-125. (8) Wise, S. A.; Sander, L. C. In Chromatographic Separations Based on Molecular Recognition; Jinno, K., Ed.; Wiley-VCH: New York, 1997; pp 1-64. (9) Copper, C. L.; Sepaniak, M. J. Anal. Chem. 1994, 66, 147-154. (10) Norton, D.; Shamsi, S. A. Anal. Chim. Acta 2003, 496, 165-176. (11) Norton, D.; Zheng, J.; Shamsi, S. A. J. Chromatogr., A 2003, 1008, 205215. (12) Norton, D.; Shamsi, S. A. J. Chromatogr., A 2003, 1008, 217-232.

Analytical Chemistry, Vol. 78, No. 19, October 1, 2006 6921

Figure 1. Chemical structures of methylated benzo[a]pyrene isomers. The numbers show the position of substituted methyl groups on the benzo[a]pyrene molecule.

but poor specificity especially when other PAHs coelute with MBAPs (because all the PAHs have a strong UV chromophore). In addition, the analysis time, which is often in excess of 2 h/run, needs to be shortened for practical purposes. The CEC coupled to mass spectrometric (MS) detection is one of the recently developed alternative detection techniques for CEC for both achiral and chiral analyses.13-15 Because MS detection offers excellent sensitivity, outstanding specificity, and valuable structural information, this technique is gaining popularity. Among all the atmospheric pressure ionization (API) techniques, electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) have been commonly adopted because of their versatility to ionize a broad range of polar to moderately polar compounds. However, since MBAP isomers are highly nonpolar, ESI has difficulties providing sufficient ionization. Atmospheric pressure photoionization (APPI) is a newly introduced ionization technique, which extends the scope of API-MS application.16,17 In particular, APPI is able to efficiently ionize those compounds that are not readily ionized by ESI and even APCI.18-20 Therefore, APPI-MS coupled to HPLC has attracted growing interest during recent years.18-30 Among many of these applica(13) Shamsi, S. A.; Miller, B. E. Electrophoresis 2004, 25, 3927-3961. (14) Klampfl, C. W. J. Chromatogr., A 2004, 1044, 131-144. (15) Barcelo-Barrachina, E.; Moyano, E.; Galceran, M. T. Electrophoresis 2004, 25, 1927-1948. (16) Raffaelli, A.; Saba, A. Mass Spectrom. Rev. 2003, 22, 318-331. (17) Bos, S. J.; Leeuwen, S. M.; Karst, U. Anal. Bioanal. Chem. 2006, 384, 8599. (18) Hanold, K. A.; Fischer, S. M.; Cormia, P. H.; Miller, C. E.; Syage, J. A. Anal. Chem. 2004, 76, 2842-2851. (19) Straube, E. A.; Dekant, W.; Voelkel, W. J. Am. Soc. Mass. Spectrom. 2004, 15, 1853-1862. (20) Cai, S.-S.; Syage, J. A. Anal. Chem. 2006, 78, 1191-1199. (21) Turnipseed, S. B.; Roybal, J. E.; Andersen, W. C.; Kuck, L. R. Anal. Chim. Acta 2005, 529, 159-165. (22) Wang, G.; Hsieh, Y.; Korfmacher, W. A. Anal. Chem. 2005, 77, 541-548. (23) Kauppila, T. J.; Kostiainen, R.; Bruins, A. P. Rapid Commun. Mass Spectrom. 2004, 18, 808-815. (24) Cai, Y.; Kingery, D.; McConnell, O.; Bach, A. C., II. Rapid Commun. Mass Spectrom. 2005, 19, 1717-1724. (25) Cai, S.-S.; Syage, J. A. J. Chromatogr., A 2006, 1110, 15-26. (26) Sforza, S.; Dall’asta, C.; Marchelli, R. Mass Spectrom. Rev. 2006, 25, 5476. (27) Varga, M.; Bartok, T.; Mesterhazy, A. J. Chromatogr., A 2006, 1103, 278283. (28) Lembcke, J.; Ceglarek, U.; Fiedler, G. M.; Baumann, S.; Leichtle, A.; Thiery, J. J. Lipid Res. 2005, 46, 21-26. (29) Moriwaki, H.; Ishitake, M.; Yoshikawa, S.; Miyakoda, H.; Alary, J.-F. Anal. Sci. 2004, 20, 375-377. (30) Robb, D. B.; Blades, M. W. J. Am. Soc. Mass. Spectrom. 2005, 16, 12751290.

6922 Analytical Chemistry, Vol. 78, No. 19, October 1, 2006

tions, HPLC-APPI showed high sensitivity and low noise level along with good linear dynamic range. In addition, APPI requires less heat for desolvation as compared to APCI. Hence, the decomposition of thermally labile compounds could be effectively reduced, which in turn improves the detection sensitivity.31 Recent literature evidence indicates that APPI provides significantly better sensitivity than APCI at lower flow rates.18 This suggests that APPI has a great potential for its on-line coupling with capillary electrophoresis (CE). However, surprisingly, to our knowledge, only four papers on CE-APPI-MS have published so far18,32-34 and none on CEC-APPI-MS. In this work, we present the hyphenation of CEC separation to APPI-MS detection for MBAP isomers. To the best of our knowledge, this is the first attempt to couple these two techniques. We believe that the use of the internally tapered CEC column is the key to the development of a stable and rugged on-line CECAPPI-MS. The goal of the present study is to improve the suitability of the developed CEC method for analysis of MBAP isomers. Thus, in this study, efforts are mainly focused on the enhancement of both the detection specificity and throughput. Initially, to achieve maximum detection sensitivity, sheath liquid composition and flow rate as well as APPI-MS spray chamber parameters were optimized through direct infusion. Next, the composition of CEC mobile phase was studied by varying acetonitrile content and adding a small amount of planar organic cation to improve the throughput. Finally, the quantitation aspects of the CEC-APPI-MS method were validated. These include setting up the calibration curves for several MBAP isomers and determination of limit of detection (LOD). EXPERIMENTAL SECTION Reagents and Chemicals. The 5-µm CEC Reliasil polymeric C18-PAH stationary phase was purchased from Column Engineering Inc. (Ontario, CA). The 12 MBAP isomers (Figure 1) were obtained from the Division of Cancer Prevention, National Cancer Institute (Rockville, MD). Acetone, acetonitrile (ACN), ammonium hydroxide (NH3‚H2O), methanol (MeOH), acetic acid (HOAc), cobalt nitrate, lanthanum nitrate, lead nitrate, lithium bromide, nickel nitrate, silver nitrate, sodium chloride, and toluene were supplied by Fisher (Springfield, NJ). Tropylium tetrafluoroborate ([C7H7]+[BF4]-) and 7.5 M ammonium acetate solution (NH4OAc) were purchased from Sigma-Aldrich (St. Louis, MO). Water was purified by a Barnstead Nanopure II Water system (Dubuque, IA). CEC-MS Instrumentation, Parameters, and Conditions. All CEC-MS experiments were carried out with an Agilent 3D capillary electrophoresis instrument interfaced to a single quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA). Three different MS ionization sources, namely, ESI, APCI, and APPI, were utilized. An Agilent 1100 series HPLC pump equipped with 1:100 splitter was used to deliver the sheath liquid. The Agilent Chemstation software (Version 10.02) was used for instrumental control and data processing. (31) Greig, M. J.; Bolanos, B.; Quenzer, T.; Bylund, J. M. R. Rapid Commun. Mass Spectrom. 2003, 17, 2763-2768. (32) Nilsson, S. L.; Andersson, C.; Sjoeberg, P. J. R.; Bylund, D.; Petersson, P.; Joernten-Karlsson, M.; Markides, K. E. Rapid Commun. Mass Spectrom. 2003, 17, 2267-2272. (33) Mol, R.; de Jong, G. J.; Somsen, G. W. Electrophoresis 2005, 26, 146-154. (34) Mol, R.; de Jong, G. J.; Somsen, G. W. Anal. Chem. 2005, 77, 5277-5282.

The sheath liquid and MS spray chamber parameters were optimized by direct infusion, i.e., continuously flushing the ACN solution containing 0.1 mg/mL of 7-MBAP through a 60 cm-long, 50 µm (i.d.) open tubular capillary into the MS spray chamber and scanning the corresponding mass range (typically m/z 50500). The general parameters for each ionization mode were set as follows: (1) ESI. sheath liquid flow rate at 5.0 µL/min; capillary voltage, 3000 V; fragmentor voltage, 80 V; drying gas flow rate, 8.0 L/min; drying gas temperature, 250 °C; nebulizer pressure, 4 psi. (2) APCI. sheath liquid flow rate at 50 µL/min; corona discharge needle current, 4.0 µA; vaporizer temperature, 200 °C; capillary voltage, 4000 V; fragmentor voltage, 80 V; drying gas flow rate, 8.0 L/min; drying gas temperature, 250 °C; nebulizer pressure, 10 psi. (3) APPI. sheath liquid flow rate at 5.0 µL/min; krypton lamp (photon energy at 10.1 eV); vaporizer temperature, 200 °C; capillary voltage, 1000 V; fragmentor voltage, 80 V; drying gas flow rate, 8.0 L/min; drying gas temperature, 250 °C; nebulizer pressure, 4 psi. CEC-APPI-MS Conditions. The 75-µm-i.d. capillaries provided by Polymicro Technologies (Phoenix, AZ) were utilized to fabricate internally tapered tips with an opening of 10-13 µm at the outlet.35 Briefly, the precut capillary end is slowly heated in the methane-O2 flame (∼800 °C). After the polyimide coating of the column end is burned off, the capillary end is moved quickly in and out of the flame center until the internal taper is gradually formed in 30-60 s. The slurry containing ∼15 mg of the polymeric C18-PAH stationary phase and 400 µL of ACN was allowed to sonicate for 30 min. Then, the internally tapered capillary was packed with the resulting slurry using the same procedure reported elsewhere.36 A typical CEC-MS column has an open segment of ∼45 cm followed by a 25-cm-long packed bed. The mobile phase or sheath liquid was degassed for 30 min before use. Before CECMS run, the column was preconditioned with the desired mobile phase using a HPLC pump (Lab Alliance, State College, PA) at ∼2000 psi for 2 h. Typically, the separation voltage was set at 20 kV, employing a voltage ramp of 3 kV/s. During the on-line CECAPPI-MS, a 12-bar external pressure was applied to the inlet buffer vial. The selective-ion monitoring (SIM) mode was utilized to monitor positive ions of MBAPs at m/z 266 (i.e., [M•]+). Preparation of Standard Analytes. Stock solutions of individual MBAP isomer were prepared by dissolving 2 mg of each isomer in 2 mL of ACN and stored under -20 °C. The mixture of MBAP isomers was prepared by sonicating the individual isomers for 5 min and then taking 100-µL aliquots from each stock solution of the isomers. A typical injection aliquot (40 µL) was prepared by taking 25 µL of MBAP stock mixture and then diluting with 15 µL of triply deionized water. Linearity was checked by three replicate injections of mixtures containing three MBAP isomers, namely, 1-MBAP, 3-MBAP, and 11-MBAP, over the range of 2.550 µg/mL, each containing 10 µg/mL 7-MBAP as the internal standard. Safety Precautions. Since the MBAP isomers are suspected carcinogens, caution was exercised with these compounds. All handling of MBAPs were performed in a ventilated hood with appropriate clothing, mask, and gloves to avoid inhalation or skin (35) Zheng, J.; Norton, D.; Shamsi, S. A. Anal. Chem. 2006, 78, 1323-1330. (36) Zheng, J.; Shamsi, S. A. Anal. Chem. 2003, 75, 6295-6305.

Figure 2. Mass spectra of 7-MBAP using (a) APCI and (b) APPI. For the APCI-MS and APPI-MS conditions, see Experimental Section.

contact. The stock solutions were stored in a closed container at -20 °C. In addition, care was taken to dispose of the MBAP waste appropriately. RESULTS AND DISCUSSION Comparison of API-MS Ionization Sources. Because MBAPs are extremely nonpolar and lack ionizable group, ESI-MS is not considered as a suitable ionization source for MS detection of these compounds; instead, APCI and APPI are much better choices. Nevertheless, for comparison purposes, the MS detection of MBAPs using ESI, APCI, and APPI sources was investigated through direct infusion of an ACN solution containing 0.1 mg/ mL 7-MBAP. As expected, ESI provided very poor MS abundance for this extremely nonpolar analyte (data not shown). Figure 2 compares the mass spectra in the m/z 50-500 range for 7-MBAP by APCI and APPI. In the APCI mass spectrum, the protonated precursor ion for 7-MBAP, [M + H]+ at m/z 267, has the highest abundance. The demethylated ion [M + H - CH3]+ at m/z 252 shows the second highest abundance, which accounts for almost half of the molecular ion abundance. In contrast, the APPI mass spectrum shows a predominant molecular ion [M•]+ at m/z 266, which is of much stronger intensity (almost 4-fold) than the [M + H]+ ion in the APCI mass spectrum. This is probably because APPI provides highly selective and less destructive ionization for the analyte. In addition, other ions, including ammonia adduct ion [M + NH3]+ at m/z 283 and demethylated ion [M + H - CH3]+ at m/z 252, show much lower abundances. Optimization of Sheath Liquid Composition and Flow Rate in APPI-MS. Since CEC flow rate (typically 10-100 nL/min) is generally too small to complete the electrical circuit and maintain Analytical Chemistry, Vol. 78, No. 19, October 1, 2006

6923

Figure 3. Mass spectra of 7-MBAP and plots of APPI-MS abundance at m/z 266 (M•+) using various sheath liquid compositions and flow rates. Conditions: (a) plot of MS abundance using sheath liquid containing 0.05-1% HOAc (v/v), (b) mass spectrum using sheath liquid MeOHH2O (90:10, v/v) containing 0.5% HOAc at flow rate of 5 µL/min, (c) plot of MS abundance using sheath liquid containing 70-100% MeOH (v/v), (d) mass spectrum using sheath liquid ACN-H2O (90:10, v/v) containing 0.5% HOAc, (e) plot of MS abundance using sheath liquid MeOHH2O (90:10, v/v) containing 0-10% (v/v) toluene or acetone delivered at flow rate of 5 µL/min, and (f) plot of MS abundance using sheath liquid MeOH-H2O-HOAc (90:10:0.5, v/v/v) at flow rate of 1-15 µL/min. For the other APPI-MS conditions, see Experimental Section.

stable spray for MS, a postcolumn addition in the form of sheath liquid is essential. Our previous CEC-ESI-MS studies have shown that sheath liquid is not only critical for keeping electrical contact but also plays an important role in the API ionization. Therefore, to enhance the APPI-MS detection sensitivity, we first investigated the effects of sheath liquid composition and sheath liquid flow rate. As displayed in Figure 2b, the relative abundance of [M + NH3]+ ion accounted for almost 20% of [M•]+ ion when a sheath liquid containing 5 mM NH4OAc was used. In addition, the relative abundance of [M + NH3]+ seemed to slightly increase when the NH4OAc concentration in the sheath liquid increased from 1 to 50 mM (data not shown). This could be the reason [M•]+ ion abundance dropped slightly as NH4OAc concentration in the sheath liquid increased. Therefore, NH4OAc in the sheath liquid was replaced with HOAc. When adding 0.05-0.5% (v/v) HOAc to the sheath liquid, an enhancement of APPI-MS signal of 7-MBAP was obtained (Figure 3a). Using a sheath liquid containing 0.5% HOAc (v/v), the relative abundance of [M + NH3]+ ion over [M•]+ ion dropped to ∼5% (Figure 3b). As a result, the abundance of [M•]+ ion almost doubled using HOAc (Figure 3b) as compared to the sheath liquid containing NH4OAc (Figure 2b). However, no significant suppression for the demethylated adduct ion of 7-MBAP [M + H - CH3]+ was observed with the use of HOAc in the sheath liquid. 6924

Analytical Chemistry, Vol. 78, No. 19, October 1, 2006

The second sheath liquid parameter investigated was organic modifier content. Initially, a sheath liquid based on a MeOHH2O mixture was utilized. It was observed that the abundance of [M•]+ ion consistently increased as the volume fraction of MeOH in the sheath liquid was increased from 70 to 90% (v/v) (Figure 3c). This is due to the fact that higher MeOH content in the sheath liquid generally enhances the ionization efficiency of MBAP under APPI-MS. However, the sheath liquid containing 100% (v/ v) MeOH led to significant deviations in the mass abundance of [M•]+ ion, which could be associated with unstable spray (Figure 3c). Furthermore, the effects of using a sheath liquid containing a different organic modifier were investigated by replacing 90% (v/v) MeOH with the same volume fraction of ACN. As shown in Figure 3d, the abundance of molecular ion ([M•]+) using ACN is significantly lower than the MeOH-based sheath liquid (Figure 3c). Also, the mass spectrum is more complicated due to the formation of ACN adduct ion [M + H + ACN]+ at m/z 308. The postcolumn addition of dopant has been well recognized as an effective way to improve the detection sensitivity for HPLCAPPI-MS18-30 as well as CE-APPI-MS.18,32-34 For CE-APPI-MS, the dopant is conventionally introduced by mixing with sheath liquid. In this study, two commonly used dopants, toluene and acetone, are investigated.16 The photoionized acetone prefers proton transfer to the compounds with good proton affinity while toluene favors charge transfer to nonpolar analytes. However, severe signal

Figure 4. Electropherograms (a-c) showing effects of mobile phase ACN (v/v) for separation of MBAP isomers. The plot shown in (d) represents signal-to-noise ratio of 9-MBAP. Conditions: mobile phase (a) 85, (b) 90, and (c) 92% ACN (v/v) containing 5 mM NH4OAc at pH 8; sheath liquid MeOH-H2O-HOAc (90:10:0.5, v/v/v) at flow rate of 10 µL/min. For the other conditions, see Experimental Section.

suppression of 7-MBAP was observed as both of the dopant concentrations increased from 1 to 10% (v/v) (Figure 3e). Furthermore, high background noise, severe arcing, and abnormally high chamber current (8-10 µA) were observed at higher toluene content. The signal suppression by acetone is due to poor proton affinity of MBAP isomers. On the other hand, the signal suppression by toluene is somehow unexpected. A similar observation was reported by Hanold et al.18 in their HPLC-APPI-MS study for vitamins using toluene as dopant. Besides dopants, metal ions (usually Ag+) or planar organic ions (e.g., tropylium) were often utilized as postcolumn additions to improve the detection sensitivity of HPLC-ESI-MS for PAHs or their derivatives.37-41 The group of von Brocke42 also observed a marked improvement of CEC-ESI-MS detection sensitivity of several neutral drugs through a similar approach. Therefore, in this study, we also investigated the effect of sheath liquids containing metal ions (including Co3+, La3+, Li+, Pb2+, Ni2+, and Ag+ at a concentration of 100 µg/mL) as well as planar organic cations (37) Ng, K. M.; Ma, N. L.; Tsang, C. W. Rapid Commun. Mass Spectrom. 2003, 17, 2082-2088. (38) Ng, K. M.; Ma, N. L.; Tsang, C. W. Rapid Commun. Mass Spectrom. 1998, 12, 1679-1684. (39) Takino, M.; Daishima, S.; Yamaguchi, K.; Nakahara, T. J. Chromatogr., A 2001, 928, 53-61. (40) Moriwaki, H.; Imaeda, A.; Arakawa, R. Anal. Commun. 1999, 36, 53-56. (41) Moriwaki, H. Analyst 2000, 125, 417-420. (42) von Brocke, A.; Wistuba, D.; Gfrorer, P.; Stahl, M.; Schurig, V.; Bayer, E. Electrophoresis 2002, 23, 2963-2972.

(i.e., tropylium and 2,4,6-triphenylpylium at concentrations of 10200 µg/mL) for APPI-MS. However, no adduct ion was observed in the mass spectrum, which indicated that the direct photoionization of MBAPs is likely dominant in the case of APPI. As a result, no significant enhancement for MS signal was obtained (data not shown). The effects of sheath liquid flow rate were also investigated by varying it from 1 to 15 µL/min. The low sheath liquid flow rate (1-2.5 µL/min) provided a ∼40% lower abundance of molecular ion ([M•]+) as compared to the medium and high flow rate (5-15 µL/min, Figure 3f). In addition, high deviations on the observed MS signal indicated unstable spray under this low sheath liquid flow rate. With the increase of sheath liquid flow, the spray is stabilized, which in turn lowers the fluctuation of MS signal. However, the signal intensity was essentially constant at both medium and high flow rates. This is probably because that the APPI-MS is a mass-sensitive detector43 (instead of a concentrationsensitive detector like ESI-MS). Consequently, a medium sheath liquid flow rate at 10 µL/min was considered as optimum since it provided good sensitivity and also provided a sufficient electrical contact between the CEC capillary and MS nebulizer. Optimization of Spray Chamber Setting in APPI-MS The APPI-MS spray chamber parameters were also optimized using the same direct infusion approach (data not shown). Among these (43) Takino, M.; Daishima, S.; Nakahara, T. J. Chromatogr., A 2003, 1011, 6775.

Analytical Chemistry, Vol. 78, No. 19, October 1, 2006

6925

Figure 5. Electropherograms (a-c) showing effects of addition of tropylium ion (C7H7+) into mobile phase for separation of MBAP isomers. Conditions: mobile phase 90% (v/v) ACN containing 5 mM NH4OAc at pH 8 and (a) 0, (b) 0.1, (c) 0.5, and (d) 2.0 mM C7H7+; vaporizer temperature, 300 °C; capillary voltage, 1000 V; fragmentor voltage, 80 V; drying gas flow rate, 5.0 L/min; drying gas temperature, 250 °C; nebulizer pressure, 10 psi. The other conditions are same as Figure 4 or described in Experimental Section.

parameters, capillary voltage showed the most significant effects. For example, a low capillary voltage (1000 V) provided almost 4 times higher abundance than a high voltage setting (3500 V). This setting, which is similar to the optimum setting of Mol et al.33 in their MEKC-APPI-MS studies, is much lower than the normal capillary voltage for ESI. The other spray chamber parameters showed moderate effects for the MS signal intensity. For example, when increasing vaporizer temperature from 200 to 300 °C, the abundance of [M•]+ improved ∼20%. However, no further improvement of the signal intensity was observed with increase of the vaporizer temperature to 350 °C. To avoid overheating the CEC-MS capillary, vaporizer temperature at 300 °C was selected as the optimum. Moreover, a nebulizer pressure at 10 psi was found to provide a reasonably high and stable signal for 7-MBAP. The optimum spray chamber conditions are summarized as follows: capillary voltage, 1000 V; fragmentor voltage, 80 V; vaporizer temperature, 300 °C; nebulizer pressure, 10 psi; drying gas flow rate, 5.0 L/min; drying gas temperature, 250 °C. CEC-APPI-MS. In our previous CEC-UV study, the polymeric C18-PAH stationary phase demonstrated high efficiency and superior substitution pattern selectivity for the MBAP isomers as compared to the other stationary phases.11,12 Therefore, the internally tapered capillaries35 were packed with the same stationary phase for the following CEC-APPI-MS experiments. To improve the throughput and detection sensitivity and ensure reasonable resolutions between the isomers, the mobile-phase composition and the use of mobile-phase additive were investigated. 6926 Analytical Chemistry, Vol. 78, No. 19, October 1, 2006

The effects of ACN content in the mobile phase was studied from 85 to 92% (v/v). Initially, the ACN content was set at 85% (v/v) because the stationary phase was unable to maintain stable current and EOF when the volume fraction of ACN was lower than 80%. As displayed in Figure 4a, the mobile phase containing 85% ACN (v/v) provided the best resolution; i.e., 9 out of 12 isomers were resolved. However, due to long analysis time, the throughput and detection sensitivity (Figure 4d) were very poor. The increase of ACN content to 90% (v/v) shortened the analysis time by 25% and improved the detection sensitivity although a loss of partial resolution between 10-MBAP and 6,7-MBAP was observed (Figure 4b). A further increase of ACN content to 92% (v/v) caused a significant loss in the isomeric resolutions despite the best throughput being achieved (Figure 4c). Therefore, 90% (v/v) ACN was considered as a good compromise between resolution and analysis time or detection sensitivity. The feasibility of improving the throughput through the addition of planar organic cations (e.g., tropylium ion, C7H7+) into the CEC mobile phase was also investigated. This approach was purposed by Miller et al.44 for the separation of PAHs with nonaqueous capillary electrophoresis. As displayed in Figure 5a-c, the retentions of MBAPs are significantly reduced due to the formation of charge-transfer complexes between the MBAP and C7H7+ (Figure 5 inset).44 For example, with the addition of 0.5 mM C7H7+ in the mobile phase, the analysis time was shortened (44) Miller, J. L.; Khaledi, M. G.; Shea, D. Anal. Chem. 1997, 69, 1223-1229.

CONCLUSIONS

Figure 6. Calibration plot (a) for 1-, 3-, and 11-MBAP isomers. Electropherograms (b) at LOD of MBAP isomers. The conditions are same as Figure 5 or described in Experimental Section.

from 100 to 80 min (Figure 5c). Although the formation of a charge-transfer complex in the solution seems to be supported by the reduced retention of MBAP isomer, this complex cannot be detected by APPI-MS. As suggested by Airiau et al.,45 this is probably because [M + C7H7+] complex in the solution could undergo a charge transfer to generate [M•]+ in the gas phase. In addition, no loss of resolution of MBAP isomers is observed with the addition of C7H7+. However, further increase of the C7H7+ concentration to 2.0 mM led to slightly longer retention times (Figure 5d) and decrease of EOF caused by the increase of ionic strength. Quantitative Analysis. In this set of experiments, we examined the quantitative applicability of the developed CEC-APPI-MS method for several MBAP isomers, including setup of calibration curves and determination of LOD. For the setup of calibration curves, the 1-, 3-, and 11-MBAP were selected due to their strong carcinogenicity using 7-MBAP as the internal standard. As shown in Figure 6a, all the calibration curves showed good linearity (R2 ) 0.9995-0.9997) over the concentration range of 2.5-50 µg/ mL. In addition, the electropherograms shown in Figure 6b illustrate that the MBAP isomers are detectable at 400 ng/mL (LOD, at S/N ∼2). Compared with our previous CEC-UV analysis for MBAPs (estimated LOD ∼40 µg/mL for 1-MBAP),12 the detection sensitivity is enhanced ∼100 times with APPI-MS detection. Utilizing a triple quadrupole mass spectrometer, further improvements on the detection sensitivity, which could be comparable to laser induced fluorescence (LIF),46 are expected. (45) Airiau, C. Y.; Brereton, R. G.; Crosby, J. Rapid Commun. Mass Spectrom. 2001, 15, 135-140. (46) Yan, C.; Dadoo, R.; Zhao, H.; Zare, R. N.; Rakestraw, D. J. Anal. Chem. 1995, 67, 2026-2029.

The combination of CEC and MS detection via an APPI ionization interface for determination of MBAP isomers was demonstrated. The use of the APPI source successfully broadened the range of API-MS to these highly nonpolar isomers by providing highly abundant molecular ion [M•]+ through direct photoionization. In terms of molecular ion abundance of MBAP, the sensitivity of APPI is ∼4 times higher than APCI. The replacement of NH4OAc in the sheath liquid with HOAc enhanced the MS signal up to ∼90% and effectively suppressed the formation of the ammonia adduct. Furthermore, the MeOH-based sheath liquid provided less adduct ions and higher abundance of [M•]+ than the ACN-based one. Unlike ESI, the response of APPI-MS showed less dependence on the sheath liquid flow rate in the range of 5-15 µL/min. No significant signal enhancement was observed with the addition of metal ions, planar organic cations, and dopants in sheath liquid. The use of toluene as the dopant caused significant suppression for photoionization efficiency and abnormally high noise level, which is just the opposite observation of other groups. Thus, it could be concluded that the sheath liquid composition and sheath liquid flow rate as well as APPI-MS spray chamber settings were the key parameters for the detection sensitivity of MBAP. Further investigation including exploration of lower photon energy lamp (e.g., xenon lamp, 8.4 eV) and alternative dopant is under investigation. The use of internally tapered columns packed with polymeric C18 stationary phase showed high efficiency and resolution for difficult separation of MBAP isomers. Besides increasing the ACN content in the mobile phase, the addition of small amounts of tropylium ion was also able to effectively shorten the analysis time. Although the charge-transfer adduct ion ([M + C7H7]+) was not directly observed in the APPI mass spectrum, the formation of such a complex ion was supported by reduced retention for MBAPs. Finally, quantitative aspects of the CEC-APPI-MS method were evaluated. Good linearity in the range of 2.5-50 µg/mL for 1-MBAP, 3-MBAP, and 11-MBAP and LOD at 400 ng/mL (which is ∼100 times more sensitive than our CEC-UV result) for each isomer were obtained. The use of SIM mode in APPI-MS shows excellent specificity, which in turn provides high tolerance to the presence of C7H7+ ions in the mobile phase. In contrast, the use of C7H7+ ions for either CEC-UV or CEC-LIF could be problematic due to strong UV absorbance in the range of 200-300 nm.

ACKNOWLEDGMENT Financial support for this project was provided by the National Institutes of Health (Grant GM 62314). The authors thank Dr. Christine Miller of Agilent Technologies for providing the spacer for CE-APPI-MS. J.Z. thanks Dean Norton and William Bragg for helpful discussion and suggestions on the manuscript, and the support of the Molecular Basis of Disease fellowship.

Received for review June 3, 2006. Accepted July 27, 2006. AC061024C Analytical Chemistry, Vol. 78, No. 19, October 1, 2006

6927