Microchip Atmospheric Pressure Photoionization for Analysis of

Mar 6, 2009 - E-mail: [email protected] (R.K.); [email protected] (A.G.M.). Phone: 358-9-191 59 134 (R.K.); 1-850-644-0529 (A.G.M.)...
0 downloads 0 Views 549KB Size
Anal. Chem. 2009, 81, 2799–2803

Technical Notes Microchip Atmospheric Pressure Photoionization for Analysis of Petroleum by Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Markus Haapala,† Jeremiah M. Purcell,‡ Ville Saarela,| Sami Franssila,| Ryan P. Rodgers,‡,§ Christopher L. Hendrickson,‡,§ Tapio Kotiaho,†,⊥ Alan G. Marshall,*,‡,§ and Risto Kostiainen*,† Division of Pharmaceutical Chemistry, Faculty of Pharmacy, University of Helsinki, P.O. Box 56, FI-00014 Helsinki, Finland, Department of Chemistry and Biochemistry, Florida State University, Tallahassee, Florida 32306-4390, Ion Cyclotron Resonance Program, National High Magnetic Field Laboratory, 1800 East Paul Dirac Drive, Tallahassee, Florida 32310-4005, Microfabrication Group, Department of Micro and Nanosciences, Helsinki University of Technology, P.O. Box 3500, FI-02015 TKK, Finland, and Laboratory of Analytical Chemistry, Department of Chemistry, University of Helsinki, P.O. Box 55, FI-00014 Helsinki, Finland Atmospheric pressure photoionization (APPI) Fourier transform ion cyclotron resonance mass spectrometry (FTICR MS) has significantly contributed to the molecular speciation of petroleum. However, a typical APPI source operates at 50 µL/min flow rate and thus causes a considerable mass load to the mass spectrometer. The recently introduced microchip APPI (µAPPI) operates at much lower flow rates (0.05-10 µL/min) providing decreased mass load and therefore decreased contamination in analysis of petroleum by FT-ICR MS. In spite of the 25 times lower flow rate, the signal response with µAPPI was only 40% lower than with a conventional APPI source. It was also shown that µAPPI provides very efficient vaporization of higher molecular weight components in petroleum analysis. The characterization of petroleum components is highly important in development and selection of refining processes. Petroleum is a complex matrix with over 100 000 different compounds covering polar and nonpolar compounds and mass range from 100 Da to over 1000 Da. Thus, the analysis of petroleum is challenging and requires highly advanced analytical techniques. Gas chromatography/mass spectrometry is a widely used technique in analysis of volatile and semivolatile compounds in petroleum.1 However, gas chromatography is not suitable for higher-boiling point compounds, and other techniques such as high-resolution mass spectrometry combined with various ioniza* To whom correspondence should be addressed. E-mail: risto.kostiainen@ helsinki.fi (R.K.); [email protected] (A.G.M.). Phone: 358-9-191 59 134 (R.K.); 1-850-644-0529 (A.G.M.). Fax: 358-9-191 59 556 (R.K.); 1-850-644-1366 (A.G.M.). † Faculty of Pharmacy, University of Helsinki. ‡ Florida State University. § National High Magnetic Field Laboratory. | Helsinki University of Technology. ⊥ Department of Chemistry, University of Helsinki. (1) Wang, Z. D.; Fingas, M. J. Chromatogr., A 1997, 774, 51–78. 10.1021/ac802427m CCC: $40.75  2009 American Chemical Society Published on Web 03/06/2009

tion methods have been successfully applied to analysis of petroleum components. High-field Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) is the most efficient and commonly used mass spectrometric method for detailed compositional studies of complex petroleum samples without sample pretreatment or chromatographic separation. This is achieved due to the high mass resolution and accuracy of FTICR MS, which allows confident assignment of elemental composition to thousands of peaks.2 During the past few years the analysis of crude oil and its components by FT-ICR MS has emerged as a new area of high-performance chemical analysis named “petroleomics”.3-10 Several ionization techniques have been applied for ionization of petroleum components. Commonly used electrospray ionization (ESI) is selective toward the ionization of polar heteroatomic molecules (mostly basic and acidic compounds). Field desorption/ field ionization,11,12 matrix-assisted laser desorption ionization,13,14 (2) Hughey, C. A.; Rodgers, R. P.; Marshall, A. G. Anal. Chem. 2002, 74, 4145– 4149. (3) Marshall, A. G.; Rodgers, R. P. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 18090–18095. (4) Miyabayashi, K.; Naito, Y.; Tsujimoto, K.; Miyake, M. Int. J. Mass Spectrom. 2002, 221, 93–105. (5) Hughey, C. A.; Galasso, S. A.; Zumberge, J. E. Fuel 2007, 86, 758–768. (6) Pakarinen, J. M. H.; Teravainen, M. J.; Pirskanen, A.; Wickstrom, K.; Vainiotalo, P. Energy Fuels 2007, 21, 3369–3374. (7) Schrader, W.; Panda, S. K.; Brockmann, K. J.; Benter, T. Analyst 2008, 133, 867–869. (8) Panda, S. K.; Schrader, W.; Andersson, J. T. Anal. Bioanal. Chem. 2008, 392, 839–848. (9) Fuchser, J.; Witt, M. LC · GC Eur. 2007, 44–46. (10) Lu, X. Q.; Shi, Q.; Zhao, S. Q.; Gao, J. S.; Mang, Y. H.; He, J. H. Chin. J. Anal. Chem. 2008, 36, 614–618. (11) Schaub, T. M.; Hendrickson, C. L.; Qian, K. N.; Quinn, J. P.; Marshall, A. G. Anal. Chem. 2003, 75, 2172–2176. (12) Qian, K. N.; Edwards, K. E.; Siskin, M.; Olmstead, W. N.; Mennito, A. S.; Dechert, G. J.; Hoosain, N. E. Energy Fuels 2007, 21, 1042–1047. (13) Robins, C.; Limbach, P. A. Rapid Commun. Mass Spectrom. 2003, 17, 2839– 2845. (14) Rizzi, A.; Cosmina, P.; Flego, C.; Montanari, L.; Smaniotto, A.; Seraglial, R.; Traldi, P. J. Mass Spectrom. 2007, 42, 874–880.

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

2799

laser-induced acoustic desorption,15 atmospheric pressure laser ionization (APLI),16 atmospheric pressure chemical ionization (APCI),17,18 and atmospheric pressure photoionization (APPI)19,20 are more universal ionization techniques capable of ionizing nonpolar and aromatic molecules as well as polar heteroatomic compounds. Among these ionization techniques APPI seems to cover the widest range of compounds in petroleum.21 In dopantassisted APPI 10 eV photons emitted by an UV lamp produce radical cations of the dopant. The dopant radicals may ionize analytes through charge exchange or proton transfer or react with solvent producing protonated solvent molecules, which can transfer protons to analytes. As a result protonated molecules, radical cations, or both are produced.22,23 In negative ion APPI deprotonated molecules can be formed by proton transfer or negatively charged molecular ions by charge exchange or electron capture reactions.24,25 The optimal flow rate with commercial APPI sources is usually from 50 to 200 µL/min. This means that the mass load is relatively high and may cause rapid contamination of the ion source and mass analyzer especially when complex mixtures of nonvolatile compounds, such as petroleum components, are analyzed. The use of a low 2 µL/min flow rate in crude oil asphaltene analysis has been reported,20 which suggests that commercial APPI sources have differences in terms of low flow rate performance. The recently introduced microchip APPI source (µAPPI) provides significantly lower flow rates (0.05-10 µL/min) and therefore decreased mass load to the mass spectrometer compared to commercial sources.26 Thus far µAPPI has been applied in analysis of polycyclic aromatic hydrocarbons, steroids, polychlorinated biphenyls, and illicit drugs.27-29 In these studies µAPPI has been shown to provide high sensitivity and good quantitative performance indicating its high potential in several analytical applications. In this report, we evaluate the feasibility of µAPPI FT-ICR MS for the molecular speciation of petroleum (15) Crawford, K. E.; Campbell, J. L.; Fiddler, M. N.; Duan, P.; Qian, K.; Gorbaty, M. L.; Kenttamaa, H. I. Anal. Chem. 2005, 77, 7916–7923. (16) Schmitt-Kopplin, P.; Englmann, M.; Rossello-Mora, R.; Schiewek, R.; Brockmann, K. J.; Benter, T.; Schmitz, O. J. Anal. Bioanal. Chem. 2008, 391, 2803–2809. (17) Hsu, C. S.; Dechert, G. J.; Robbins, W. K.; Fukuda, E. K. Energy Fuels 2000, 14, 217–223. (18) Rudzinski, W. E.; Aminabhavi, T. M.; Sassman, S.; Watkins, L. M. Energy Fuels 2000, 14, 839–844. (19) Purcell, J. M.; Hendrickson, C. L.; Rodgers, R. P.; Marshall, A. G. Anal. Chem. 2006, 78, 5906–5912. (20) Qian, K.; Mennito, A. S.; Edwards, K. E.; Ferrughelli, D. T. Rapid Commun. Mass Spectrom. 2008, 22, 2153–2160. (21) Panda, S. K.; Andersson, J. T.; Schrader, W. Anal. Bioanal. Chem. 2007, 389, 1329–1339. (22) Robb, D. B.; Covey, T. R.; Bruins, A. P. Anal. Chem. 2000, 72, 3653–3659. (23) Kauppila, T. J.; Kuuranne, T.; Meurer, E. C.; Eberlin, M. N.; Kotiaho, T.; Kostiainen, R. Anal. Chem. 2002, 74, 5470–5479. (24) Basso, E.; Marotta, E.; Seraglia, R.; Tubaro, M.; Traldi, P. J. Mass Spectrom. 2003, 38, 1113–1115. (25) Kauppila, T. J.; Kotiaho, T.; Kostiainen, R.; Bruins, A. P. J. Am. Soc. Mass Spectrom. 2004, 15, 203–211. ¨ stman, P.; Marttila, S.; Ketola, R. A.; Kotiaho, T.; Franssila, (26) Kauppila, T. J.; O S.; Kostiainen, R. Anal. Chem. 2004, 76, 6797–6801. (27) Haapala, M.; Luosujarvi, L.; Saarela, V.; Kotiaho, T.; Ketola, R. A.; Franssila, S.; Kostiainen, R. Anal. Chem. 2007, 79, 4994–4999. (28) Luosujarvi, L.; Karikko, M. M.; Haapala, M.; Saarela, V.; Huhtala, S.; Franssila, S.; Kostiainen, R.; Kotiaho, T.; Kauppila, T. J. Rapid Commun. Mass Spectrom. 2008, 22, 425–431. (29) Kauppila, T. J.; Arvola, V.; Haapala, M.; Pol, J.; Aalberg, L.; Saarela, V.; Franssila, S.; Kotiaho, T.; Kostiainen, R. Rapid Commun. Mass Spectrom. 2008, 22, 979–985.

2800

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

Figure 1. Microfabricated heated nebulizer chip (25 mm length by 10 mm width).

and compare the results to those obtained with a conventional APPI ion source. EXPERIMENTAL METHODS The heated nebulizer microchips (Figure 1) used in µAPPI experiments were fabricated at the Helsinki University of Technology, Finland, Espoo at the Micronova center. The fabrication30 and operation27 of the chips have been described in detail previously. The nozzle of the nebulizer chip was positioned inline with the mass spectrometer heated metal capillary inlet port. The distance between the microchip nozzle and the heated metal capillary inlet was 4 mm. The rf-excited krypton lamp (Thermo Fisher Scientific, Waltham, MA) was positioned orthogonal to the microchip at about 1 mm distance from the microchip edge and about 3 mm from the sample vapor jet. The nebulizer gas (carbon dioxide) was regulated (115 mL/min) with a flow meter (Aalborg, Orangeburg, NY) and fed into the chip through the Nanoport connector. Sample dissolved in HPLC grade toluene (Fisher Scientific, Pittsburgh, PA) was infused through the fused-silica sample capillary of the chip with a syringe pump at flow rates from 0.5 to 4 µL/min. The heating power was adjusted in the range of 2-4 W, corresponding to an approximate vaporizer channel temperature of 220-380 °C, with a variable dc power supply (American Reliance Inc., El Monte, CA). The temperature of the chip is not uniform since the heat is localized near the heater wires (platinum lines) due to the low thermal conductivity of glass. Thus, the highest temperature in the vaporizer channel exceeds the average temperature. The conventional APPI source was from Thermo Fisher Scientific (Waltham, MA). In the source, the vaporized analyte gas stream flow is orthogonal to the mass spectrometer heated metal capillary inlet and the krypton UV lamp. The source was mounted to a custom-built adapter, which interfaced to the first differentially pumped stage of the mass spectrometer. For conventional APPI, the solvent flow rate was 50 µL/min, the temperature of the heated nebulizer was set to 350 °C, carbon dioxide was used as a nebulizing gas at 550 kPa, and the auxiliary gas port was plugged. The lamp was the same as for µAPPI experiments. Naphtho[2,3-a]pyrene (NAP) was purchased from SigmaAldrich (St. Louis, MO) and dissolved in HPLC grade toluene to produce a 400 µM stock solution. All other concentrations were (30) Saarela, V.; Haapala, M.; Kostiainen, R.; Kotiaho, T.; Franssila, S. Lab Chip 2007, 7, 644–646.

achieved by serial dilution with toluene from the stock solution. The petroleum samples were dissolved in toluene (500 µg/mL). A previously described 9.4 T FT-ICR MS31,32 and experimental conditions3 were utilized. Multiple (100) time-domain acquisitions were averaged for the petroleum samples, Hanning-apodized, and zero-filled once before fast Fourier transform and magnitude calculation.33 All observed ions were singly charged, based on the unit m/z separation between 12Cn and 13C112Cn-1 isotopic variants of the same elemental composition.34 Therefore, mass spectral peak positions are reported in Da rather than as m/z. RESULTS AND DISCUSSION NAP was used to optimize the µAPPI parameters. The optimal distance between the nozzle of the heated nebulizer microchip and the heated metal capillary inlet of the mass spectrometer was about 4 mm. However, that distance was not very critical. The photoionization lamp was positioned close (3 mm) to the sample vapor jet to ensure maximum intensity of the photons at the surface of the jet. The optimal nebulizer gas flow rate was 115 mL/min. The sample flow rate had a clear effect on the NAP signal magnitude: the response increased with increasing flow rate from 0.5 to 4 µL/min, confirming the earlier results that µAPPI is mass flow sensitive.26 The signal stability was good with flow rates above 1 µL/min, but decreased at the flow rate of 0.5 µL/min, perhaps because the number of reagent ions (i.e., radical cations of toluene) is insufficient for efficient ionization at lower flow rates. The same effect has been observed in APPI experiments with too low dopant (toluene) flow rates, i.e., flow rates below a few microliters per minute.35 The feasibility of µAPPI FT-ICR MS was evaluated by infusing 40, 400, and 4000 nM NAP standard solutions at a flow rate of 1 µL/min. A factor of 10 increase in concentration (40-400 nM) resulted in a factor of 10 signal increase with the same 10 s accumulation period. Moreover, an even more concentrated sample (4000 nM) was externally accumulated for 1 s, i.e., 10% as long as for the 400 nM sample. Although the signal was noisier, the 4000 nM sample produced similar signal response as the 400 nM sample. These results show that the signal response was linearly dependent on concentration. Figure 2 shows repeatability for different external ion accumulation periods for 400 nM NAP standard infused to µAPPI FT-ICR MS at a flow rate of 1 µL/min. The relative standard deviations with accumulation periods of 1, 2, 4, and 10 s were 13%, 10%, 10%, and 5%, respectively, indicating good signal repeatability. The repeatability improved with longer accumulation periods likely because fast signal variations were filtered out from the output signal. Figure 2 also shows that the signal response scales linearly with ion accumulation period. All in all, these results (31) Senko, M. W.; Hendrickson, C. L.; PasaTolic, L.; Marto, J. A.; White, F. M.; Guan, S. H.; Marshall, A. G. Rapid Commun. Mass Spectrom. 1996, 10, 1824–1828. (32) Hakansson, K.; Chalmers, M. J.; Quinn, J. P.; McFarland, M. A.; Hendrickson, C. L.; Marshall, A. G. Anal. Chem. 2003, 75, 3256–3262. (33) Marshall, A. G.; Verdun, R. F. Fourier Transforms in NMR, Optical, and Mass Spectrometry: A User’s Handbook; Elsevier: Amsterdam, The Netherlands, 1990. (34) Senko, M. W.; Beu, S. C.; McLafferty, F. W. J. Am. Soc. Mass Spectrom. 1995, 6, 52–56. (35) Robb, D. B.; Blades, M. W. J. Am. Soc. Mass. Spectrom. 2005, 16, 1275– 1290.

Figure 2. µAPPI FT-ICR MS signal repeatability for naphtho[2,3a]pyrene, measured for each of four different accumulation periods.

Figure 3. Stability of the µAPPI signal for crude oil analysis. Heavy crude oil dissolved in toluene (500 µg/mL) was infused at a flow rate of 2 µL/min continuously for 4 h before data was collected.

show that flow rates of 1-2 µL/min and signal accumulation periods between 2 and 10 s are optimal for µAPPI FT-ICR MS analysis of model compounds. The stability of µAPPI for crude oil analysis was tested by infusing heavy crude oil dissolved in toluene (500 µg/mL) at a flow rate of 2 µL/min continuously for 4 h before data was collected. The total ion chromatogram over the 1 h period shows good stability (Figure 3) indicating that the chip is suitable for crude oil analysis. The use of low flow rates of 1-2 µL/min provided by µAPPI is beneficial because contamination of the ion source is significantly less than at the higher flow rates (50-200 µL/min) used in conventional APPI, an especially important advantage for the analysis of heavy petroleum samples with a large fraction of high molecular weight nonvolatile components. Petroleum Analysis. Crude oil is an extremely complex mixture containing hydrocarbons with multiple aromatic rings, which may also contain heteroatoms such as nitrogen, sulfur, and oxygen. Variations in crude oil composition affect oil refinery processes, and thus knowledge of oil composition is important. For example, sulfur-containing compounds are detrimental to refinery processes and harmful to the environment on combustion. Moreover, boiling point increases with increasing heteroatom number, double-bond equivalent (DBE, equal to the number of rings plus double bonds to carbon), and number of carbon atoms. Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

2801

Figure 4. FT-ICR mass spectra of crude oil sample with µAPPI (bottom) and conventional (top) APPI sources. The flow rate was 2 µL/min for µAPPI and 50 µL/min for conventional APPI. The temperature of the conventional APPI source was 350 °C, and 3 W of heating power, corresponding to roughly 300 °C in the vaporizer channel, was used for µAPPI. The accumulation period was 3 s, and 100 acquisitions were summed for each spectrum.

Figure 6. Isoabundance-contoured plots of double-bond equivalents vs carbon number for the S1 and O1S1 class ions in FT-ICR mass spectra produced by µAPPI and conventional APPI of crude oil. The highest abundance species are colored red, with lower abundance species changing color as indicated.

Figure 5. Heteroatom class distributions derived from the mass spectra of Figure 4.

The FT-ICR MS signals for individual components of crude oil were compared for µAPPI and conventional APPI sources (Figure 4). The flow rate of the crude oil sample was 2 µL/min with µAPPI and 50 µL/min with the conventional APPI source. The temperature of the conventional APPI source was set to 350 °C (maximum allowable temperature based on source components, i.e., fused-silica coating and gas connections), and 3 W of heating power, which corresponds to roughly 300 °C in the vaporizer, was used for µAPPI. The ion accumulation period in both experiments was 3 s, and 100 acquisitions were summed for each spectrum. About 60% of the signal response of conventional APPI was obtained with µAPPI at 1/25 of the flow rate for conventional APPI source, indicating significantly about 15-fold greater mass flow sensitivity for µAPPI. Elemental composition assignments were accomplished by converting the mass spectral data from the IUPAC mass scale to the Kendrick mass scale.36,37 Figure 5 shows the heteroatom class distributions derived from the spectra in Figure 4. The same classes were detected by both macro- and µAPPI. The relative abundances for the various classes are roughly comparable for both ion sources. Ergo, no significant information is lost at the much lower flow rate for µAPPI. Figure 6 shows color-coded isoabundance-contoured plots of DBEs versus carbon number for the S1 and O1S1 classes (36) Kendrick, E. Anal. Chem. 1963, 35, 2146–2154. (37) Hughey, C. A.; Hendrickson, C. L.; Rodgers, R. P.; Marshall, A. G.; Qian, K. N. Anal. Chem. 2001, 73, 4676–4681.

2802

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

Figure 7. FT-ICR mass spectra for high-mass ends of heavy crude oil, ionized by µAPPI at 3.0, 3.5, and 4.0 W of heating power. Crude oil concentration was 500 µg/mL in toluene, and flow rate was 2 µL/ min.

measured by µAPPI and conventional APPI FT-ICR MS. The S1 class plots are very similar, whereas the O1S1 class plots show that somewhat higher DBE species are observed with µAPPI than with conventional APPI, possibly because the µAPPI at 3 W is at a higher temperature than the heated nebulizer of the conventional source and therefore higher DBE species are vaporized and ionized more efficiently. A higher µAPPI source temperature may also explain the higher relative abundances of compounds including oxygen (O1S1, O1S2, O2S1,

and O1) with µAPPI (Figure 5). All in all, Figure 6 shows that the performance of µAPPI in terms of DBE and carbon number distribution is very similar to that of conventional APPI. As Figure 4 shows, µAPPI at a heating power of 3 W produced the same kind of mass spectral profile of petroleum as from the conventional APPI source at 350 °C. However, the heating power of the chip can be easily adjusted from 0 to 4 W. By increasing the heating power from 3 to 4 W clearly more components can be detected at the high-mass end of the crude oil mass spectrum indicating high vaporizing efficiency of µAPPI (Figure 7). However, an increased amount of residue accumulated inside the microchip nozzle and vaporizer channels at higher heating powers. It was possible to remove the residue by flushing toluene through the chip. The residue buildup mechanism may be coke formation due to the high temperature in the vaporizer channel. This needs to be studied and the microchip developed further by improving heater configuration and nozzle design to minimize residue adsorption. CONCLUSIONS We showed that µAPPI offers significant improvements over the conventional APPI source for petroleum analysis by FT-ICR MS. Most important, the optimal flow rate for µAPPI was about 2

µL/minssignificantly lower than the 50 µL/min for conventional APPI. In spite of the 25-fold reduction in flow rate, the signal response with µAPPI was only 40% lower compared to conventional APPI. The use of low flow rates without significant loss in sensitivity is especially important to avoid contamination of the ion source in analysis of nonvolatile petroleum samples. We also showed that µAPPI provides very efficient vaporization of higher molecular weight components. Finally, the heated nebulizer microchip can be easily replaced, facilitating a one-chip/onesample analysis to eliminate sample carryover to the next spectrum. Conceivably, the production cost of a chip could be made small compared to other costs of the analysis. ACKNOWLEDGMENT This work was supported by NSF Division of Materials Research through DMR-0654118, the State of Florida, the Finnish Funding Agency for Technology and Innovation (Tekes), and the Graduate School of Chemical Sensors and Microanalytical Systems.

Received for review November 17, 2008. Accepted February 12, 2009. AC802427M

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

2803