MALDI Mechanism of Dihydroxybenzoic Acid Isomers: Desorption of

Mar 25, 2013 - Sheng-Ping Liang , I-Chung Lu , Shang-Ting Tsai , Jien-Lian Chen , Yuan Tseh Lee , Chi-Kung Ni. Journal of The American Society for Mas...
2 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/JPCB

MALDI Mechanism of Dihydroxybenzoic Acid Isomers: Desorption of Neutral Matrix and Analyte Chi Wei Liang, Chih Hao Lee, Yu-Jiun Lin, Yuan Tseh Lee,† and Chi Kung Ni*,‡ Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei 10617 Taiwan ABSTRACT: Angular resolved velocity distributions of laser desorbed neutral matrices (dihydroxybenzoic acids, DHB) and analytes (tryptophan) embedded in these matrices were investigated at 322 nm by a modified crossed molecular beam apparatus. Desorbed ions generated from MALDI were measured by a time-of-flight mass spectrometer. Desorptions of neutral matrix and analyte from 2,3-DHB, 2,4-DHB, 2,5DHB, 2,6-DHB, and 3,5--DHB at 322 nm have similar properties, but the ion intensities are in the order 2,3DHB ≅ 2,6-DHB > 2,5-DHB ≅ 2,4-DHB > 3,5-DHB. It indicates that the combination of various parameters related to neutral species, including absorption coefficient, sublimation energy, contact of analyte and matrix in crystal, and plume dynamics of desorbed species are not crucial in the determination of MALDI process for DHB isomers. The difference of matrix activity of DHB isomers at this wavelength must result from the other properties, like the excited state lifetime, proton affinity, gas-phase basicity, acidity, ionization energy, or the other properties related to the primary reactions in ion generation.

I. INTRODUCTION Since the invention of matrix-assisted laser desorption ionization (MALDI) in late 1980s by Karas and Hillenkamp,1,2 it has become an established technique for mass analysis of large biomolecules. The basic principle of MALDI is to mix thermally unstable, nonvolatile analytes into highly photoabsorbing organic compounds which are so-called matrices. Irradiation of the matrix compounds by a pulsed laser beam rapidly increases the surface temperature, resulting in sublimation of matrix. It brings the nonvolatile compounds, especially ionic species, into the gas-phase for mass analysis. In spite of its great success and wide acceptance of this technique, the mechanisms remain debatable.3−5 The selection of matrix is a crucial parameter for successful MALDI mass analysis. 2,5-Dihydroxybenzoic acid (2,5-DHB) is one of the most widely used matrices since the report in 1990.6,7 Among the six positional isomers of dihydroxybenzoic acid (DHB), 2,5-DHB is better than the other isomers regarding analyte detection sensitivity.8−11 The amounts of analyte ions generated from the matrices of other positional isomers are either very small or no analyte ions at all. However, the reason why 2,5-DHB is better than the other remains unclear. A good matrix needs to have a sufficient absorption coefficient at the applied laser wavelength. The absorption spectra have been measured by various research groups.10,12,13 The spectra measured by Horneffer et al.10 are shown in Figure 1. It can explain why 2,5-DHB is superior to the others at 355 nm, since only 2,5-DHB has sufficient absorption coefficient at this wavelength. But it does not explain why 2,5-DHB performs better than the others at shorter wavelengths where other isomers have comparable absorption coefficients.10 © 2013 American Chemical Society

Figure 1. UV spectra of DHB isomers adapted from ref 10. The spectra were obtained in diffusion reflection for solid sample material. R is the intensity of the reflected light from a matrix doped sample normalized to that of the pure BaSO4. No quantitative absorption cross section can be extracted from these diffusion reflection spectra. They are individually scaled to have a value equal to the molar absorption coefficient of the solution spectra. Dashed lines represent 355 and 322 nm.

A good matrix can transfer the absorbed photon energy into thermal energy efficiently for desorption of matrix and analyte molecules. Low sublimation energy reduces the threshold of laser fluence and helps to bring the matrix and analyte molecules into gas phase. The sublimation enthalpy of DHB isomers have been measured in previous study by Price et al.9 The correlation between sublimation and MALDI performance was investigated by the same group.9 They concluded that sublimation enthalpy was not the principal factor in the MALDI Received: February 12, 2013 Revised: March 22, 2013 Published: March 25, 2013 5058

dx.doi.org/10.1021/jp4015492 | J. Phys. Chem. B 2013, 117, 5058−5064

The Journal of Physical Chemistry B

Article

process of DHB isomers. Some other factor may be more important in determining ion yields in the MALDI process. The early hypothesis stated that only 2,5-DHB can produce p-benzoquinone by photochemical decarboxylation. None of the other isomers can form p-benzoquinone after decarboxylation. Consequently, p-benzoquinone may be particularly efficacious in MALDI reactions.14 However, recent molecular beam study showed that decarboxylation is not important in photochemical reactions of 2,5-DHB. Instead, water elimination is the major photodissociation channel of 2,5DHB.15 As is shown later in this work, water elimination remains as the major dissociation channel for 2,5-DHB molecules in the MALDI process. Close contact between analyte and matrix molecules increases the reaction probability to generate ions if ions are not preformed. It is also helpful to bring analyte molecules into gas phase during sublimation. It has been suggested that the strength of signal could be contributed from the incorporation of the analyte into matrix crystals. Horneffer et al. demonstrated that protein incorporation into the matrix crystals of DHB isomers was helpful but not necessary.10 A related study showed that the MALDI performance was inversely proportional to crystal size, indicating that surface contact between analyte and matrix was more crucial than the protein incorporation into the matrix crystals.16 Nitrogen gas laser (337 nm) and frequency-tripled Nd:YAG solid-state laser (355 nm) are by far the most commonly used ones because of their simplicity, small size, and relatively low cost. In this work, we report the measurement of desorbed neutral matrix and analyte molecules at 322 nm. We demonstrate that at proper wavelength which DHB isomers have similar UV absorption coefficients, the amounts of desorbed matrix and analyte, and the velocity and angular distributions of desorbed matrix and analyte are similar, but the ion generation efficiencies are different. It indicates that the combination of various parameters related to neutral species, including absorption coefficient, sublimation energy, contact of analyte and matrix in crystal, and plume dynamics of desorbed species, is not crucial in the determination of MALDI activity of DHB isomers. The difference of matrix activity of DHB isomers must be due to the other factors related to ion generation.

Figure 2. Schematic of crossed molecular beam apparatus for the measurement of mass and angular resolved velocity distributions of desorbed neutrals.

the reaction center. The mass spectrometer was located inside a differential pumped chamber where the pressure was kept below 1 × 10−9 Torr in order to reduce the background. The mass spectrometer consisted of an electron ionizer located at 350 mm from the reaction center, a quadrupole mass spectrometer, and a Daly ion detector. The time-resolved output from the Daly detector were accumulated in an ion counting mode using a multichannel scalar (Turbo-MCS, EG&G-Ortec, Oak Ridge, TN) with a dwell time of 30 μs per channel and total 1000 channels for each mass-selected time-offlight spectra. The time-of-flight spectrum for each angle was taken from a new sample surface for 5−60 laser shots, depending on laser fluences. Angular resolution determined by the detection solid angle is 2.5°. The 322 nm laser beam was generated from the sum frequency of the second harmonic (532 nm) of a Nd:YAG laser and the fundamental laser beam (816 nm) of a Ti:sapphire laser pumped by the second harmonic of the other Nd:YAG laser (LS-2134, LOTIS TII). The output of the sum frequency was 322 nm (∼6 ns pulse duration). The laser beam was focused by a lens ( f = 300 mm) onto the sample surface with 45° incidence angle. The laser energy was attenuated by a polarizer, a rotatable half-wave plate, and a continuously variable metallic neutral density filter (ThorLABs, Newton, NJ) and was measured by a pyroelectric detector. The beam spot size on the sample (∼0.2 mm2) was determined by projecting the laser onto a CCD camera and taking the usual 1/e2 intensity as the profile boundary. In some cases, the third harmonic (355 nm, 5 ns) of a Nd:YAG laser was used for comparison. Ions generated directly from MALDI were detected by a reflection time-of flight (TOF) mass spectrometer (Autoflex III, Bruker Daltonik). The laser systems used in desorption of neutrals, as described in previous paragraph, were used in this TOF mass spectrometer. All chemicals were used as purchased from Sigma-Aldrich without further purification. A 0.2 M DHB stock solution and a 0.01 M L-tryptophan (Tryp) stock solution were prepared by dissolving the corresponding compounds in a 50% acetonitrile (ACN) aqueous solution separately. For the investigation of desorbed matrix, 0.2 M DHB stock solution was used directly. For the investigation of desorbed analyte, the stock solutions of matrix and analyte were mixed in the molar ratio of 50:1. The solution was then dropped on the sample holder and vacuumdried. The thickness of solid sample after vacuum-dried was about ∼20 μm.

II. EXPERIMENTAL SECTION The mass and angular resolved velocity distributions of desorbed neutral molecules were measured using a modified crossed molecular beam machine. Figure 2 shows the schematic of the modified crossed molecular beam apparatus. It consisted of a main chamber, a rotatable sample holder, and a rotatable detector chamber. Most of the details concerning the main chamber and detector chamber have been given elsewhere.17−19 Only a brief description of the apparatus and the modifications made specifically for the laser desorption studies are given here. A stainless steel cylindrical sample holder (10 mm diameter) was positioned inside the main chamber which was evacuated by two turbomolecular pumps to keep the pressure below 1 × 10−7 Torr. The laser irradiation spot on the sample surface was located at the reaction center of the main chamber. The rotation axis of the sample holder was offset 4 mm from the reaction center of the main chamber. This allows us to change to new sample surface for laser desorption by rotating the sample holder without breaking the vacuum. The angular resolved velocity distributions of desorbed products were detected by a rotating mass spectrometer around 5059

dx.doi.org/10.1021/jp4015492 | J. Phys. Chem. B 2013, 117, 5058−5064

The Journal of Physical Chemistry B

Article

III. RESULTS Desorption of Neutral DHB Molecules from Pure DHB Matrix. Ions m/z 54, 136, 108, 80, 52, 28, and 18 were the major ions observed from DHB isomers containing orthohydroxyl functional group (2,3-DHB, 2,4D-HB, 2,5-DHB, and 2,6-DHB) as the electron energy of electron ionizer was set at 70 eV. Ion m/z 154 represents the DHB parent ion, and ions m/z 136, 108, 80, and 52 represent the loss of H2O, followed by the successive elimination of one, two, three CO molecules, respectively. On the other hand, ions m/z 154, 137, 109, 81, 53, 44, and 17 were observed from DHB isomers without orthohydroxyl functional group (3,4-DHB and 3,5-DHB). Ions m/z 137, 109, 81, and 53 represent the loss of OH, followed by the successive elimination of CO molecules. The velocity distributions measured at 0° for various DHB isomers at laser fluence 40 and 150 J/m2 are shown in Figure 3. In general, the velocity are in the order 2,5-DHB ≅ 2,6-DHB ≅ 2,3-DHB > 3,5-DHB ≅ 2,4-DHB > 3,4-DHB at low laser fluence. However, the velocities are similar at high laser fluence, except for 3,4-DHB which still has slower velocity than the other isomers. The distributions illustrated in Figure 3 have been multiplied by different factors for different ions (the factors are shown in parentheses) so that they have similar heights in the figure and the shapes between them can be compared easily. For DHB isomers containing ortho-hydroxyl functional group, the intensities (after multiplied by the factors) of m/z 136, 108, and 80 are much larger than that of the parent ion m/z 154 at the leading edge. The difference at the leading edge of velocity distribution is enlarged for large laser fluence. The difference represents the portions of parent molecules distributed in the leading edge have larger internal energy. They either dissociated into fragments before they arrived at the detector or they cracked into ionic fragments much easier upon electronic ionization due to large internal energy. In a separate experiment, water in 2,5-DHB solution was replaced by methanol in sample preparation. The velocity distribution of m/z 18 from the sample prepared by anhydrous (methanol and acetonitrile) 2,5-DHB solution shows significant different from the velocity distribution of m/z 18 prepared from normal (water and acetonitrile) 2,5-DHB solution, as illustrated in Figure 3. The velocity distribution of m/z 18 from the sample prepared by anhydrous solution only contains fast component, but the velocity distribution of m/z 18 from the water solution contains both fast and slow components. This suggests that the slow component in the velocity distribution of m/z 18 comes from the residual water in the crystal after vacuum-dried. The fast component in the velocity distribution of m/z 18 results from the dissociation of 2,5-DHB. Since the other fragment after water elimination is m/z 136, the difference of velocity distributions between fragments m/z 136 and 18 suggests that at least part of the DHB molecules dissociated into fragments (m/z 136 and 18) before they arrived at the detector. Almost no difference of velocity distributions between parent ion m/z 154 and fragment ions m/z 137, 109, and 81 were found for the DHB isomers without ortho hydroxyl functional group (3,5-DHB and 3,4-DHB). It indicates that these DHB isomers do not decompose into fragments upon UV irradiation. This is consistent with the quantum chemistry calculations,15 which show that DHB isomers with ortho-hydroxyl functional group have a very low dissociation barrier and they can easily

Figure 3. Velocity distributions measured at zero degree for various DHB isomers at laser fluence 40 and 150 J/m2. For 2,5-DHB at 150 J/ m2, velocity distributions of m/z 18 from two sample preparation methods (see text for details) are shown together for comparison.

eliminate H2O. This dissociation channel does not occur easily for 3,5-DHB and 3,4-DHB due to the large barrier height. The angular resolved velocity distributions of desorbed neutral DHB are illustrated in Figure 4. It is clear that the intensity as well as velocity decrease as the angle increases. We divide the velocity distribution arbitrarily into three portions. The fastest portion has the velocity larger than the peak value of the distribution at 0°. The slowest portion has the velocity smaller than half of the peak value. The angular distributions of 5060

dx.doi.org/10.1021/jp4015492 | J. Phys. Chem. B 2013, 117, 5058−5064

The Journal of Physical Chemistry B

Article

Figure 5. Total desorption (a) and average kinetic energy (b) of DHB isomers. The desorption and kinetic energy of 2,5-DHB at 355 nm are also shown for comparison.

other isomers at low laser fluence (50 J/m2). However, desorption of 3,4-DHB increases rapidly as the laser fluence increases. It is the same as the other isomers at high laser fluences (300 J/m2). The average kinetic energies of various DHB isomers are shown in Figure 5b. The kinetic energies are in the order 2,5DHB ≅ 2,6-DHB ≅ 2,3-DHB > 3,5-DHB > 2,4-DHB > 3,4 -DHB at low fluence and 3,5-DHB ≅ 2,5-DHB ≅ 2,6-DHB > 2,4-DHB ≅ 2,3-DHB > 3,4-DHB at high fluence. It is interesting to note that the average kinetic energies of DHB isomers containing the ortho-hydroxyl functional group do not increase as rapidly as the 3,5-DHB at large laser fluence. This is because the large laser fluence results in high surface temperature, leading to large internal energy followed by dissociation (H2O elimination). Since part of the energy is used in dissociation, the available energy decreases and the kinetic energy levels off at high laser fluence for these DHB isomers. The kinetic energy of 2,5-DHB at 355 nm is also shown for comparison. The difference in kinetic energy likely results from the different absorption cross section and different photon energies at 322 and 355 nm. No accurate absorption cross sections at these two wavelengths have been reported. The photon energy at 322 nm is larger than that at 355 nm, resulting in higher surface temperature at 322 nm before desorption. Desorption of Neutral Analyte from Mixture of Matrix and Analyte. The velocity distributions of desorbed neutral tryptophan (Tryp) from the mixture of DHB and tryptophan (molar ratio 50:1) are shown in Figure 6. The spectra of analyte were taken from ion m/z 130, which was the dominant fragment ion of Tryp by electron ionization. The signals of the desorbed matrix, m/z 154, 136, 108, and 80, from the same sample are also shown for comparison. Except 3,4-DHB, the velocity distributions of desorbed Tryp from all the other DHB isomers share similar properties. The analyte have smaller intensities than that of matrix at the leading edge (large velocity) of velocity distributions, but they have similar intensities at the rear part of the distributions. It may result from the dissociation of Tryp due to high

Figure 4. Angular resolved velocity distributions of desorbed neutral DHB.

these three portions, as well as the fitting to cosn θ, are shown in the right column of Figure 4. In general, the fastest components have narrowest forward peaking distributions (normal to surface). The slowest components are very close to thermal desorption distribution, cos θ. The total desorption of neutral DHB, integrated from all angles, fragments, and velocities (only velocity >300 m/s are taken into account), is shown in Figure 5a. In general, the amount of desorption increases as the laser fluence increases. Although the absorption coefficients of these isomers are not exactly the same, the shapes of desorption curves as a function of laser fluences look very similar, except 3,4-DHB. The desorption of 3,4-DHB is about 50−100 times smaller than the 5061

dx.doi.org/10.1021/jp4015492 | J. Phys. Chem. B 2013, 117, 5058−5064

The Journal of Physical Chemistry B

Article

Tryp also has a large intensity for very slow component (velocity 3,5-DHB > 2,4-DHB > 3,4-DHB at low fluence and in the order 3,5-DHB ≅ 2,5-DHB ≅ 2,6-DHB > 2,4-DHB ≅ 2,3-DHB > 3,4-DHB at high fluence; the ion intensities generated directly from MALDI are in the order 2,3-DHB ≅ 2,6-DHB > 2,5-DHB ≅ 2,4-DHB > 3,5-DHB > 3,4-DHB. The different orders between desorption of neutrals, kinetic energies, and ion intensities suggest that the properties, including absorption coefficient, sublimation energy, contact of analyte and matrix in crystal, and collisions required for possible reactions occurred in expanding plume on desorption process, are not the determining factors for the activity of matrix in DHB MALDI at this wavelength. The matrix activity must be due to the other properties, like the properties of the excited state, proton affinity, gas-phase basicity, acidity, ionization energy, etc. Further comparison of these properties between 2,3-DHB, 2,4-DHB, 2,5-DHB, 2,6-DHB, and 3,5-DHB is helpful to understand the MALDI mechanism. Knochenmuss proposed that the primary ionization in UV MALDI can be described by exciton pooling model.36 In the simulation, desorption is modeled by an adiabatic, isentropic supersonic expansion. When the temperature reaches the experimentally determined sublimation temperature (450 K for 25-DHB),37 the sample is presumed to begin to expand into the vacuum. According to this model, desorption velocity from isentropic supersonic expansion is 1080 m/s for 25-DHB and it is independent of laser fluence. The narrow angular distribution and large desorption velocity obtained in this work confirm the near adiabatic, isentropic supersonic expansion of desorbed molecules. However, our experimental measurement shows that desorption velocity increases as the laser fluence increase. The desorption velocity as large as 3000 m/s was observed at large laser fluence. It indicates that the surface temperature is much higher than sublimation temperature at large laser fluence. The velocity and angular distributions of desorbed neutrals also provide the information about the distributions of

Figure 8. Total desorbed Tryp from the mixture of DHB and tryptophan (50:1) at 322 nm.

Figure 9. Relative ion intensities of (a) protonated matrix ions, (b) protonated Tryp ions, (c) protonated arginine, and (d) protonated bradykinin from MALDI at 322 nm.

Similar orders of ion intensities were found for these two analytes. Although the ion intensities obtained from 2,3-DHB and 2,6DHB at 322 nm are larger than that of 2,5-DHB at 322 nm, it does not necessary mean that the ion intensities obtained from 2,3-DHB and 2,6-DHB at 322 nm are larger than that of 2,5DHB at 355 nm. We did not make this comparison because the laser modes, including spatial distribution and time-resolved profile, at these two wavelengths are different. Comparison of the ion intensities at these two wavelengths with the similar laser mode is planned in future work.

IV. DISCUSSION The narrow angular distribution for the fast component and the near thermal desorption in angular distribution for the slow component in velocity distribution suggest that they result from different mechanisms. These two mechanisms control the velocity distribution at the beginning of desorption and before the end of desorption, separately. The fast temperature raise following laser irradiation results in the rapid phase transition. The sudden increase of gas phase molecules above the surface produces a plume in which the pressure is high before expansion. The expansion of the plume is similar to the isentropic expansion of a molecular jet. It resulted in large velocity normal to the surface and very narrow angular distribution. The velocity can be much larger than average velocity of the plume temperature, since part of the molecular rotational energy and vibrational energy can be transferred into translational energy during the isentropic expansion. As 5063

dx.doi.org/10.1021/jp4015492 | J. Phys. Chem. B 2013, 117, 5058−5064

The Journal of Physical Chemistry B

Article

desorbed ions. Since the number of desorbed neutrals is 106− 109 times larger than that of ions,38 the velocity and angular distributions of desorbed ions must be similar to that of neutrals due to the large number of collisions between ions and neutrals right after desorption. However, the distributions of ions may change from the distributions of neutrals due to the space charge effect and the electric field in mass spectrometer when the number of collisions between ions and neutrals decreases (i.e., after further expansion of these desorbed molecules). The velocity and angular distributions of desorbed neutrals measured in this work can be used to imitate the distributions of ions if the number of charge particles is small (i.e., space charge effect is small) and no external electric field (e.g., in delay extraction conditions). This information is useful in the design of a mass spectrometer.

■

(22) Juhasz, P.; Vestal, M. L.; Martin, S. A. J. Am. Soc. Mass Spectrom. 1997, 8, 209. (23) Gluckmann, M.; Karas, M. J. Mass Spectrom. 1999, 34, 467. (24) Berkenkamp, S.; Menzel, C.; Hillenkamp, F.; Dreisewerd, K. J. Am. Soc. Mass Spectrom. 2002, 13, 209. (25) Ermer, E. R.; Baltz-Knorr, M.; Haglund, R. F. J. Mass Spectrom. 2001, 36, 538. (26) Spengler, B.; Bökelmann, V. Nucl. Instrum. Methods Phys. Res., Sect. B 1993, 82, 379. (27) Beavis, R. C.; Chait, B. T. Chem. Phys. Lett. 1991, 181, 479. (28) Zhang, W.; Chait, B. T. Int. J. Mass Spectrom. Ion Processes 1997, 160, 259. (29) Pan, Y.; Cotter, R. J. Org. Mass Spectrom. 1992, 27, 3. (30) Bökelmann, V.; Spengler, B.; Kaufmann, R. Eur. Mass Spectrom. 1995, 1, 81. (31) Huth-Fehre, T.; Becker, C. H. Rapid Commun. Mass Spectrom. 1991, 5, 378. (32) Karas, M.; Dreisewerd, K.; Schürenberg, M.; Wang, B.; Hillenkamp, F. AIP Conf. Proc. 1995, 329, 53. (33) Tsai, S. T.; Chen, C. H.; Lee, Y. T.; Wang, Y. S. Mol. Phys. 2008, 106, 239. (34) Puretzky, A. A.; Geohegan, D. B.; Hurst, G. B.; Buchanan, M. V. Phys. Rev. Lett. 1999, 83, 444. (35) Puretzky, A. A.; Geohegan, D. B. Chem. Phys. Lett. 1998, 286, 425. (36) Knochenmuss, R. J. Mass Spectrom 2002, 37, 867. (37) Stevenson, E.; Breuker, K.; Zenobi, R. J. Mass Spectrom 2002, 35, 1035. (38) Tsai, M. T.; Lee, S.; Lu, I. C.; Chu, K. Y.; Liang, C. W.; Lee, C. H.; Lee, Y. T.; Ni, C. K. Rapid Commun. Mass Spectrom. 2013, 27, 955−963.

AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected]. Present Addresses †

Y.T.L.: Also at Department of Chemistry, National Taiwan University, Taipei 10617, Taiwan. ‡ C.K.N.: Also at Department of Chemistry, National Tsing Hua University, Hsinchu 30013, Taiwan. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Karas, M.; Bachmann, D. Anal. Chem. 1985, 57, 2935. (2) Karas, M.; Bachmann, D.; Bahr, U.; Hillenkamp, F. Int. J. Mass Spectrom. Ion Processes 1987, 78, 53. (3) Karas, M.; Kruger, R. Chem. Rev. 2003, 103, 427. (4) Knochenmuss, R.; Zenobi, R. Chem. Rev. 2003, 103, 441. (5) Knochenmuss, R. Analyst 2003, 131, 966. (6) Karas, M.; Bahr, U.; Ingendoh, A.; Nordhoff, E.; Stahl, B.; Strupat, K.; Hillenkamp, F. Anal. Chim. Acta 1990, 241, 175. (7) Strupat, K.; Karas, M.; Hillenkamp, F. Int. J. Mass Spectrom. Ion Processes 1991, 111, 89. (8) Krause, J.; Stoeckli, M.; Schlunegger, U. P. Rapid Commun. Mass Spectrom. 1996, 10, 1927. (9) Price, D. M.; Bashir, S.; Derrick, P. R. Thermochim. Acta 1999, 327, 67. (10) Horneffer, V.; Dreisewerd, K.; Ludemann, H. C.; Hillenkamp, F.; Lage, M.; Strupat, K. Int. J. Mass Spectrom. 1999, 185/186/187, 859. (11) Jessome, L.; Hsu, N. Y.; Wang, Y. S.; Chen, C. H. Rapid Commun. Mass Spectrom. 2008, 22, 130. (12) Heise, T. W.; Yeung, E. S. Anal. Chim. Acta 1995, 299, 377. (13) Allwood, D. A.; Dyer, P. E. Chem. Phys. 2000, 261, 457. (14) (a) Harvey, D. J. Rapid Commun. Mass Spectrom. 1993, 7, 614. (b) Karbach, V.; Knochenmuss, R. Rapid Commun. Mass Spectrom. 1998, 12, 968−974. (15) Bagchi, A.; Dyakov, Y. A.; Ni, C. K. J. Chem. Phys. 2010, 133, 244309. (16) Trimpin, S.; Rader, H. J.; Mullen, K. Int. J. Mass Spectrom. 2006, 253, 13. (17) Lee, Y. T.; McDonald, J. D.; LeBreton, P. R.; Herschbach, D. R. Rev. Sci. Instrum. 1969, 40, 1402. (18) Sparks, R. K. Ph.D. Thesis, University of California, Berkeley, 1980. (19) Liang, C. W.; Lee, C. H.; Lee, Y. T.; Ni, C. K. Chem.Asian J. 2011, 6, 2986. (20) Schürenberg, M.; Schulz, T.; Dreisewerd, K.; Hillenkamp, F. Rapid Commun. Mass Spectrom. 1996, 10, 1873. (21) Fournier, I.; Brunot, A.; Tabet, J. C.; Bolbach, G. Int. J. Mass Spectrom. 2002, 213, 203. 5064

dx.doi.org/10.1021/jp4015492 | J. Phys. Chem. B 2013, 117, 5058−5064