VUV Photoionization and Dissociation of Tyramine and Dopamine: the Joint Experimental and Theoretical Studies

Transcription

1 CHINESE JOURNAL OF CHEMICAL PHYSICS VOLUME 5, NUMBER 1 FEBRUARY 7, 01 ARTICLE VUV Photoionization and Dissociation of Tyramine and Dopamine: the Joint Experimental and Theoretical Studies Hui-jun Guo, Li-li Ye, Liang-yuan Jia, Li-dong Zhang, Fei Qi National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei 3009, China (Dated: Received on November 3, 011; Accepted on November, 011) Photon induced dissociation investigations of neutral tyramine and dopamine are carried out with synchrotron vacuum ultraviolet photoionization mass spectrometry and theoretical calculations. At low photon energy, only molecular ions are measured by virtue of nearthreshold photoionization. While increasing photon energy to 11.7 ev or more, four distinct fragment ions are obtained for tyramine and dopamine, respectively. Besides, the ionization energies of tyramine and dopamine are determined to be 7.98±0.05 and 7.67±0.05 ev by measuring the photoionization efficiency curves of corresponding molecular ions. With help of density function theory calculations, the detailed fragmentation pathways are established as well. These two molecular cations have similar aminoethyl group elimination pathways, (m/z=14) and C 7 H 8 O (m/z=108) are supposed to be generated by the McLafferty rearrangement via γ-hydrogen (γ-h) shift inducing β-fission. And CH NH is proposed to derive from the direct fission of C7 C8 bond. Besides, the McLafferty rearrangement and the C7 C8 bond fission are validated to be dominant dissociation pathways for tyramine and dopamine cations. Key words: Tyramine, Dopamine, VUV photoionization, Mass spectrometry, Ionization energy, Dissociation pathway I. INTRODUCTION Neurotransmitters, an important family of endogenous chemicals, serve as messengers to transmit signals from neuron to target receptors. This chemical family contains several types of chemical groups, such as neuropeptides, amino acids, and catecholamines [1]. As members of catecholamines, tyramine and dopamine play importantly physiological roles in vivo. Tyramine could affect vessel constriction and dopamine is related to cognition and learning behaviors [ 4]. However, insufficient or excess supplies of them can cause some neurological disorders, such as Parkinson s disease derived from insufficient biosynthesis of dopamine by dopaminergic neurons [5, 6]. Due to the significances of these molecules, a lot of researches have been carried out. Bourcier et al. used liquid chromatography/mass spectrometry (LC/MS) to detect neurotransmitters and analogous compounds, and multiple reaction monitoring (MRM) was used to study the neutral fragment losses of protonated neurotransmitter molecules in tandem mass spectrometry [7]. Domelsmith et al. reported the vertical ionization energies (IEs) of tyramine and dopamine us- Author to whom correspondence should be addressed. zld@ustc.edu.cn, FAX: ing photoelectron spectroscopy [8, 9]. However to the best of our knowledge, their adiabatic IEs are absent in experiment. Vorsa et al. have reported the femto-second laser ionization of neutral dopamine [10]. In their work, three distinct fragments at m/z=14, 13, and 30 were observed from the dissociation of dopamine molecular ion (m/z=153) under multiphoton excitation. Also some other mass spectrometric methods, including chemical ionization/mass spectrometry (CI/MS) and electrospray ionization/mass spectrometry (ESI/MS), were used to probe the structural information of them [11, 1]. In order to get more structural information, Wu et al. investigated the ion-molecule reaction of dopamine, and a lot of adduct ions were detected, such as [MCH] (m/z=166) and [MC H 3 ] (m/z=180) [13]. However, the fragments and adduct ions obtained with these methods are limitedly generated from the protonated and deprotonated molecular ions ([MH] and [M H] ). Furthermore, the investigations on gaseous dissociation of tyramine and dopamine radical cations, which are majorly generated from electron impact (EI) and photon ionization (PI), were rarely reported. In this work, we present the dissociative photoionization investigations of neutral tyramine and dopamine with synchrotron VUV photoionization mass spectrometry. Four distinct fragment ions are obtained at different photon energies for tyramine and dopamine, respec- 11

2 1 Chin. J. Chem. Phys., Vol. 5, No. 1 Hui-jun Guo et al. tively. The photodissociation pathways of tyramine and dopamine molecular ions were proposed and DFT calculations were carried out to validate. Besides, γ-h shift inducing β-fission processes to form C 7 H 8 O and, and C7 C8 bond direct fissions to produce CH NH are validated to be dominant dissociation pathways of these two molecular ions. Additionally, the adiabatic IEs of these two analytes are determined by measuring the photoionization efficiency (PIE) curves of these molecular ions. II. EXPERIMENTAL AND THEORETICAL METHODS A. Experimental methods Tyramine and dopamine with purity of over 98% were obtained commercially and used directly without any further preparation. All experiments were carried out at National Synchrotron Radiation Laboratory (Hefei, China). The experimental setup has been described in detail in previous publications [14, 15]. Briefly, a beam of 1064 nm output of a pulsed Nd:YAG laser (Surelite I-0, Continuum, USA) with duration of 7 ns and repetition rate of 10 Hz was utilized to generate gaseous molecule plumes of analytes. The samples were coated onto the surface of stainless steel substrate without any matrix. In order to generate intact neutral molecules, laser power for desorption was controlled below 9 mj/pulse. The nascent neutral molecules in gas phase near the substrate surface were ionized by the crossed synchrotron VUV light, and the generated ions were detected by a home-made reflection time-offlight (TOF) mass spectrometer. Ion signals were amplified by a preamplifier (VT10C, ORTEC, USA) and recorded by a multiscaler (FAST Comtec P7888, Germany). Time delay between laser and the pulse of TOF repeller field was about 150 µs, which was controlled by a home-made pulse/delay generator. Synchrotron VUV light from a bending magnet of 800 MeV electron storage ring was monochromatized by a 1 m Seya-Namioka monochromator coupled with a laminar grating (100 grooves/mm, Horiba Jobin Yvon, France) with wavelength accuracy of 1 Å. The grating covered photon energy range of nm with energy resolving power (E/ E) of about 500. Average photon flux was measured to be photons/s at ionization region. A silicon photodiode (SXUV-100, International Radiation Detectors Inc., USA) was used to monitor the change of photon flux for normalizing ion signals. And the higher order harmonic radiation was eliminated by a LiF window with thickness of 0.5 mm, whose cut-off wavelength is 105 nm. B. Theoretical methods All calculations were carried out with Gaussian 03 program package [16]. Geometries for the species FIG. 1 Photoionization mass spectra of tyramine and dopamine at different photon energies. involved in this work were optimized at B3LYP/6-311G(d,p) level [17, 18]. And harmonic frequencies were computed at the same level to verify the minima and transition states, and to obtain zero-point energies (ZPEs). The relative energies of neutral precursors are defined as reference zero. The appearance energy (AE) of ionic fragment i is defined as E AE(i) =E max E 0, in which E max refers to the maximum energy of species involved in the formation pathway of fragment i, and E 0 is the energy of corresponding neutral precursor. In this work, we select the representative conformers of tyramine and dopamine to study the photon induced dissociation processes. III. RESULTS AND DISCUSSION A. Photoionization mass spectra of tyramine and dopamine Figure 1 displays the photoionization mass spectra of tyramine and dopamine at different photon energies. At 8.00 and 8.50 ev, only molecular ions at m/z=137 and 153 are observed via near-threshold photoionization. While increasing photon energy to 9.00 ev, the strong signal of m/z=108 and weak signal of m/z=30 from tyramine are obtained, which implies that the dis-

3 Chin. J. Chem. Phys., Vol. 5, No. 1 VUV Photoionization and Dissociation of Tyramine and Dopamine 13 FIG. Photoionization efficiency (PIE) curves of tyramine and dopamine molecular ions. The experimental uncertainty is within ±0.05 ev. Note: the ion intensity of dopamine cation is added by 10 to avoid curve overlaping. sociation of tyramine cation do not require too much internal energy. Similarly, the signals of m/z=14 and 30 are obtained from dopamine at 9.00 ev. When photon energy increases to ev, fragment ions at m/z=10 and 107 from tyramine, and fragment ions at m/z=136 and 13 from dopamine are measured. The vertical IEs of tyramine and dopamine have been reported to be 8.41 and 8.18 ev by Domelsmith et al. [8, 9]. However as we know, the accurate adiabatic IEs have not been reported in experiment. In this work, the adiabatic IEs of tyramine and dopamine are determined to be 7.98±0.05 and 7.67±0.05 ev, respectively (see Fig.). The theoretical IEs of tyramine and dopamine are calculated to be 7.50 and 7.30 ev at the B3LYP/6-311G(d,p) level, which are close to the experimental ones, taking the computational uncertainty into account. Compared to tyramine, 3-hydroxyl group substituent increases the electron density of dopamine molecule, which makes it easier to be ionized. Actually, the IE of dopamine is lower than that of tyramine by about 0.3 ev. The laser heating effect could result in thermal tails in the PIE curves of molecular ions, which makes it difficult to determine the accurate ionization threshold. Though some methods have been employed to determine the accurate ionization thresholds of polyatomic molecules [19, 0], it s still difficult to determine the precise IEs of molecules. Also it is noteworthy that the internal energy excess of gaseous molecules derived from the photon absorption is proved to be common in the photoionization process, which could cause kinetic energy shift [1]. Though the accurate IE values of neutral molecules are difficult to get, the measured IEs in this work are assumed to be credible as references. B. Dissociation pathways of tyramine and dopamine The proposed dissociation pathways of tyramine and dopamine cations are displayed in Scheme 1, including the atomic numbering convention in structures. In this work, the fragment ions at m/z=14 ( ) from dopamine and m/z=108 (C 7 H 8 O ) from tyramine are supposed to be generated from the β-fission induced by the γ-h migration accompanied by the loss of CH NH fragment, which is known as McLafferty rearrangement []. Also the further dissociations of C 7 H 8 O (m/z=108) and (m/z=14) to lose one hydrogen atom are employed to generate C 7 H 7 O (m/z=107) and C 7 H 7 O (m/z=13). The fragment ion at m/z=30 (CH NH ) is proposed to derive from the direct fission of C7 C8 bond from tyramine and dopamine cations (CH NH formation from dopamine cation is not shown in Scheme 1), respectively. And the fragment ions at m/z=136 (C 8 H 8 O ) and 10 (C 8 H 8 O ) are supposed to be generated from NH 3 eliminations from tyramine and dopamine cations. It is noteworthy that tyramine and dopamine cations might have two charge-retaining centers in the photoionization, NH group and benzene ring. Unfortunately, it is difficult to determine the primary ionization position. Significantly, the fragment ion CH NH is supposed to be formed from tyramine and dopamine cations with NH group charging via the direct fission of C7 C8 bond. Reversely, other fragment ions are supposed to be formed starting with the ionization at benzene ring. The calculated potential energy profiles for these fragmentation pathways of tyramine and dopamine cations are provided in Schemes 4, including optimized geometries for involved species as inserts. 1. β-fission pathways induced by γ-h shift Formation pathway for C 7 H 8 O from tyramine. The C 7 H 8 O (m/z=108) generation from tyramine is depicted briefly in Reaction (1). In this process, a γ-h atom from NH group is transferred intramolecularly to C (or C6) via a six-membered ring transition state (TS1-1) to produce an intermediate (INT1-1) with two hydrogen atoms at C (or C6). The energy barrier for this process is.06 ev, which is the relative energy of TS1-1 to RC1. Then the fragment ion C 7 H 8 O (product 1, m/z=108) is generated by the elongated C7 C8 bond cleavage driven by the formation of C8=N double bond to lose CH NH. The calculated potential energy profiles for the dissociative reactions of tyramine cation are provided in Scheme. And the AE of C 7 H 8 O is calculated to be 9.70 ev. C 8 H 11 NO (RC1 ) CH NH C 7 H 8 O (m/z = 108) (1) E AE (C 7 H 8 O ) = E(TS1-) E(RC1) = 9.70 ev Due to the unstable structure of product 1 (C 7 H 8 O ) with two hydrogen atoms at C (or C6), it could be isomerized into more stable para-methylphenol cation (C 7 H 8 O, product ). As shown in Scheme,

4 14 Chin. J. Chem. Phys., Vol. 5, No. 1 Hui-jun Guo et al. Scheme 1 Proposed dissociation pathways of (A) tyramine and (B) dopamine cations. this process involves two-step H[1,] shifts. The AE of product is calculated to be ev, which is the relative energy of (TS1-3CH NH) to RC1. Concerning the energy barrier and the structural stability of C 7 H 8 O, it is proposed that the product 1 contributes more to signal of m/z=108 at low photon energy than the product. However at high photon energy, excessive internal energy induced by photoionization could make this isomerization process easy to happen. Actually in Fig.1, strong signal of m/z=108 is observed at photon energy of 9.00 ev, indicating the experimental AE of fragment ion at m/z=108 is lower than 9.00 ev, which is much lower than theoretical AE of 9.70 ev. The internal energy excess of gaseous tyramine cation induced by laser heating effect and photon absorption could make the AE potential drifting to low energy. Also concerning the computational uncertainty, the difference between experimental AE value of m/z=108 and theoretical one is acceptable. Formation pathway for from dopamine. In previous work, Vorsa et al. suggested the fragment ion at m/z=14 was generated from dopamine cation by McLafferty rearrangement with loss of CH NH under multiphoton excitation [10]. Actually due to the 3,4- dihydroxyl substituent of benzene ring in dopamine, the C and C6 are alternative to accept one γ-h atom transferred from the terminal NH group. Thus two corresponding pathways for formation via McLafferty rearrangement are proposed and briefly depicted in reactions () and (3). Along reaction (), initially one hydrogen atom from NH group is transferred to C via a six-membered ring transition state (TS-1). Then C7 C8 bond cleaves to lose CH NH and to generate product 5 (, m/z=14), which is driven by the formation of C8=N double bond. Along this route, the AE of is computed to be 9.39 ev. In reaction (3), the similar dissociation pathway of dopamine cation induced by intramolecular hydrogen transfer occurs to give product 7 (, m/z=14), of which the AE is calculated to be 9.36 ev. The calculated potential energy profiles for these formation pathways of (products 5 and 7) are shown in Scheme 3. C 8 H 11 NO (RC ) CH NH (5, m/z = 14) () E AE ( ) = E(TS-) E(RC) = 9.39 ev C 8 H 11 NO (RC ) CH NH (7, m/z = 14) (3) E AE ( ) = E(TS-5) E(RC) = 9.36 ev

5 Chin. J. Chem. Phys., Vol. 5, No. 1 VUV Photoionization and Dissociation of Tyramine and Dopamine CH NHH (10.84) Energy / ev TS1-4 INT1- TS1-3 (10.7) (10.03).117 (9.70) (9.56) CHNH C7H7O CHNH CHNH TS1- TS CHNH (10.36) (9.0) (9.14) 1CH NH. IN1-1 (8.86) (8.41) CH NH (7.50) 1.6 RC RC (0.00) Scheme The calculated potential energy profiles for the formation pathways of fragment ions at m/z=108 and 107 from tyramine at B3LYP/6-311G(d,p) level. Numbers in brackets are relative energies, which are calculated relatively to neutral RC1. The energy unit is ev. Bond length is in A CH NHH CH NHH (10.74) (10.74) Energy / ev TS- (8.80) (9.39) 5CH NH TS-1 (8.85) (8.87) INT CHNH C7H7O (8.84) TS-4 (8.77).035 TS-5 (9.36) (8.73) (8.91) 7CH NH INT (7.30) RC (7.36) TS-3 (7.3) INT RC (0.00) Scheme 3 The calculated potential energy profiles of the formation pathways of fragment ions at m/z=14 and 13 from dopamine at B3LYP/6-311G(d,p) level. Numbers in brackets are relative energies, which are calculated relatively to neutral RC. The energy unit is ev. Bond length is in A. c 01 Chinese Physical Society

6 16 Chin. J. Chem. Phys., Vol. 5, No. 1 Hui-jun Guo et al Energy / ev TS-6CH NH (10.35) TS-7CH NH TS-7CH NH TS-8CH NH (9.93) (10.18) (9.93) (10.18) (10.) INT-4CH NH INT-4CH NH (8.80) 5CH NH 1.5 (8.91) 7CH NH (8.09) 6CH NH Scheme 4 The calculated potential energy profiles for isomerization processes of isomeric fragment ions at m/z=14 at B3LYP/6-311G(d,p) level. Numbers in brackets are relative energies, which are calculated relatively to neutral RC. The energy unit is ev. Bond length is in Å. In experiment, the strong signal of m/z=14 is observed at photon energy of 9.0 ev, indicating the experimental E AE value of is below 9.0 ev, which is not in good accordance with the calculated AEs of. As we discussed above, the internal energy excess and computational uncertainty are responsible for this difference. Additionally, the 3-hydroxyl substituent in dopamine increases the p-π conjugation of benzene ring and hydroxyl groups, which enhances the electron delocalization of benzene ring and makes C (or C6) easier to accept one hydrogen atom transferred from the NH group. Thus comparatively to the formation pathway of C 7 H 8 O (m/z=108) from tyramine, the formation route of (products 5 or 7) should have lower energy barrier. Actually, the calculated AE value of is lower than that of C 7 H 8 O by about 0.3 ev. It is noteworthy that the products 5 and 7 are not structurally stable, which could be isomerized into more stable 3,4-dihydroxytoluene radical cation (, product 6) via two-step H[1,] shifts. The calculated potential energy profiles for these isomerization processes are given in Scheme 4. The AE of (6) is computed to be 10. ev, which is higher than those of (products 5 and 7). So it is proposed that products 5 and 7 contribute more to signal of m/z=14 at low photon energy than product 6. However they could be isomerized into the more stable product 6 with high photon energy excitation.. Further dissociation pathways The C 7 H 7 O (m/z=107) and C 7 H 7 O (m/z=13) formations are proposed to derive from the further dissociations of C 7 H 8 O (product 1, m/z=108) and (products 5 or 7, m/z=14) to lose one hydrogen atom, as depicted in reactions (4) and (5), respectively. The unstable product 1 (m/z=108) could lose one hydrogen atom from C (or C6) to generate C 7 H 7 O (product 3, m/z=107). Similarly, along reaction (5) the products 5 and 7 can dissociate to generate C 7 H 7 O (m/z=13). The AEs of C 7 H 7 O (m/z=107) and C 7 H 7 O are calculated to be and ev, which are in fair agreement with the experimental measurements of m/z=107 and 13 signal (see Fig.1). No any signals of m/z=107 and 13 are observed at photon energy of 11.0 ev, which are not shown in Fig.1. Due to the weak π-δ conjugation of benzene ring and the side chain aminoethyl group, the decomposition of aminoethyl group in tyramine and dopamine is slightly affected by the hydroxyl groups of benzene ring. Thus, the energy barrier for the C 7 H 7 O formation should be close to that for the C 7 H 7 O formation. Actually, the calculated AE of C 7 H 7 O is close to that of C 7 H 7 O (only 0.1 ev difference), also indicating the C7 C8 bonds have close dissociation energy in tyramine and dopamine cations. C 7 H 8 O (1) C 7 H 7 O (m/z = 107) H (4) E AE (C 7 H 7 O ) = E(C 7 H 7 O ) E(CH NH) E(H) E(RC1) = ev (5 or 7) C 7 H 7 O (m/z = 13) H (5) E AE (C 7 H 7 O ) = E(C 7 H 7 O ) E(CH NH) E(H) E(RC) = ev 3. CH NH formation pathway As we mentioned above, the ionic fragment CH NH is proposed to derive from the direct fission of C7 C8 bonds in tyramine and dopamine cations with the ionization at NH groups accompa-

7 Chin. J. Chem. Phys., Vol. 5, No. 1 VUV Photoionization and Dissociation of Tyramine and Dopamine 17 nied by the losses of 4-hydroxylbenzyl (C 7 H 7 O ) and 3,4-dihydroxylbenzyl (C 7 H 7 O ) radicals, respectively. These processes are described in reactions (6) and (7), along which the AEs of CH NH are calculated to be 8.86 and 8.84 ev, respectively. As shown in Fig.1, weak signal of m/z=30 is observed at photon energy of 9.00 ev, which is in consistence with theoretical AEs of CH NH. As we discussed above, the C7 C8 bond fission to form CH NH might be slightly affected by the substituted hydroxyl groups of benzene ring. Actually, the calculated AEs of CH NH formed from dopamine and tyramine cations are close to each other. Additionally, the C7 C8 bond cleavage in these two molecular ions could competitively generate fragments (C 7 H 7 O (m/z=107) CH NH ) and (C 7 H 7 O (m/z=13) CH NH ), which are supposed to start with ionization at benzene ring. In experiment, we could observe higher intense signal of m/z=30 than those of m/z=107 and 13, which indicates the formation pathways for fragment ions at m/z=30 are more dominant than those of C 7 H 7 O and C 7 H 7 O. Additionally, due to the high conjugated stabilities of C 7 H 7 O and C 7 H 7 O radicals, the formations of CH NH with losses of C 7 H 7 O and C 7 H 7 O radicals dominantly govern the C7 C8 bond fission. That is why we could not obtain the signals of m/z=107 and 13 at low photon energy (below 11.0 ev). C 8 H 11 NO (RC1 ) C 7 H 7 O CH NH (m/z = 30) (6) E AE (CH NH ) = E(CH NH ) E(C 7 H 7 O ) E(RC1) = 8.86 ev C 8 H 11 NO (RC ) C 7 H 7 O CH NH (m/z = 30) (7) E AE (CH NH ) = E(CH NH ) E(C 7 H 7 O ) E(RC) = 8.84 ev IV. CONCLUSION The photoionization and photodissociation of gaseous tyramine and dopamine are investigated with synchrotron VUV photoionization mass spectrometry and theoretical calculations. By virtue of near-threshold photoionization, only molecular ions were observed by choosing the appropriate photon energy. At higher photon energy, four distinct fragment ions are obtained for tyramine and dopamine, respectively. Also, the IEs of tyramine and dopamine are determined experimentally. With help of theoretical calculations, we establish the detailed photon induced dissociation pathways of tyramine and dopamine cations. These two molecular cations have similar dissociation pathways. They could lose NH 3 to generate the corresponding fragment ions at m/z=10 and 136. Also the product ions C 7 H 8 O (m/z=108) and (m/z=14) are generated similarly via intramolecular McLafferty-type rearrangement, which is validated to be the primary dissociation pathway of these molecular ions. The C7 C8 bond fission could generate the same product of CH NH with losses of C 7 H 7 O and C 7 H 7 O radicals. Additionally, the further dissociation of C 7 H 8 O (m/z=108) and (m/z=14) could produce the fragment ions C 7 H 7 O (m/z=107) and C 7 H 7 O (m/z=13) via loss of one hydrogen atom. With the ionization at benzene ring, the 3-hydroxyl substitution in dopamine could increase the electron delocalization of benzene ring, which makes the benzene ring in dopamine cation easier to accept one γ- H atom than that in tyramine cation. Thus the energy barriers for the formation pathways of fragments at m/z=14 ( ) and 13 (C 7 H 7 O ) are lower than that for the generations of fragments at m/z=108 (C 7 H 8 O ) and 107 (C 7 H 7 O ). However, with the ionization at NH group, the hydroxyl substitutions in benzene ring slightly affect the C7 C8 bond fission, due to the weak σ-π conjugated effect. Actually, the formation pathways of CH NH (m/z=30) from tyramine and dopamine have nearly same energy barrier. 4. Other dissociation pathways The ionic fragments at m/z=10 (C 8 H 8 O ) and 136 (C 8 H 8 O ) are supposed to be generated by the NH 3 losses from tyramine and dopamine cations, respectively. However in these processes, intramolecular hydrogen transfers to NH group from C7 and C8 which are involved possibly. After that, the neutral NH 3 fragments are eliminated to form fragment ions at m/z=10 and 136. Due to the conjugated effect of benzene ring, the intramolecular hydrogen transfers to NH group from C7 which is proposed to be more favorable than that from C8. V. ACKNOWLEDGMENTS Authors thank Dr. Yang Pan for useful discussions. This work was supported by the National Natural Science Foundation of China (No ). [1] R. A. Webster, Neurotransmitters, Drugs and Brain Function, Chichester, New York: Wiley, (001). [] O. Arias-Carrion and E. Poppel, Acta Neurobio. Exp. 67, 481 (007). [3] P. R. Bieck and K. H. Antonin, J. Clin. Psychopharm. 8, 37 (1988).

8 18 Chin. J. Chem. Phys., Vol. 5, No. 1 Hui-jun Guo et al. [4] S. Gal, Z. A. Abassi, and M. B. H. Youdim, Neurotox. Res. 18, 143 (010). [5] P. Remy, M. Doder, A. Lees, N. Turjanski, and D. Brooks, Brain 18, 1314 (005). [6] D. Weintraub, A. D. Siderowf, M. N. Potenza, J. Goveas, K. H. Morales, J. E. Duda, P. J. Moberg, and M. B. Stern, Arch. Neurol. 63, 969 (006). [7] S. Bourcier, J. F. Benoist, F. Clerc, O. Rigal, M. Taghi, and Y. Hoppilliard, Rapid Commun. Mass Spectrom. 0, 1405 (006). [8] L. N. Domelsmith, L. L. Munchausen, and K. N. Houk, J. Am. Chem. Soc. 99, 4311 (1977). [9] L. N. Domelsmith and K. N. Houk, Int. J. Quantum Chem. Quantum Biol. Symp. 5, 57 (1978). [10] V. Vorsa, K. F. Willey, and N. Winograd, Anal. Chem. 71, 574 (1999). [11] C. Y. Hao, R. E. March, T. R. Croley, S. Chen, M. G. Legault, and P. Yang, Rapid Commun. Mass Spectrom. 16, 591 (00). [1] T. J. Kauppila, T. Nikkola, R. A. Ketola, and R. Kostiainen, J. Mass Spectrom. 41, 781 (006). [13] H. F. Wu, C. H. Chen, and L. C. Lu, Rapid Commun. Mass Spectrom. 17, 1479 (003). [14] Y. Pan, H. Yin, T. C. Zhang, H. J. Guo, L. S. Sheng, and F. Qi, Rapid Commun. Mass Spectrom., 515 (008). [15] Y. Pan, T. C. Zhang, X. Hong, Y. W. Zhang, L. S. Sheng, and F. Qi, Rapid Commun. Mass Spectrom., 1619 (008). [16] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Munnucci, C. Pomelli, C. Adamo, S. Clliford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, N. Rega, P. Salvador, J. J. Dannenberg, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Pikorz, I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nakayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head- Gordon, E. S. Replogle, and J. A. Pople, Gaussian 03, Revision C.0 Ed., Pittsburgh, PA: Gaussian Inc., (004). [17] A. D. Becke, J. Chem. Phys. 97, 9173 (199). [18] R. G. Parr and W. Yang, Density-functional Theory of Atoms and Molecules, Oxford: Oxford Universty Press, (1989). [19] W. A. Chupka, J. Chem. Phys. 54, 1936 (1971). [0] P. M. Guyon and J. Berkowit, J. Chem. Phys. 54, 1814 (1971). [1] C. Lifshitz, Eur. J. Mass Spectrom. 8, 85 (00). [] F. W. McLafferty, Anal. Chem. 31, 8 (1959).