Dissociation of energy-selected c-C3H6S+ studied with threshold photoelectron-photolon coincidence experiments and calculations

Dissociation of energy-selected c-C

3

H

6

S

+

studied

with threshold photoelectron-photoion coincidence experiments

and calculations

Su-Yu Chiang

, Yung-Sheng Fang

, Chun-Neng Lin

National Synchrotron Radiation Research Center, Research Division, 101, Hsin Ann Road, Hsinchu Science Park, Hsinchu 30076, Taiwan b

Department of Applied Chemistry, National Chiao Tung University, 1001, Ta Hsueh Road, Hsinchu 30010, Taiwan Received 24 November 2005; in final form 6 March 2006

Available online 10 March 2006

Abstract

The dissociation of energy-selected c-C3H6S+was investigated in a region 10.2–10.9 eV with a threshold photoelectron-photoion

coin-cidence technique. Branching ratios and average releases of kinetic energy for channel c-C3H6S+! H2CS++ C2H4were derived from

coincidence mass spectra. The measured small releases of kinetic energy near the appearance onset agree with statistical calculations; a linearly extrapolated threshold at 10.39 ± 0.01 eV agrees with a predicted energy 10.35 eV with the GAUSSIAN-3 method. We discuss

plau-sible mechanisms for c-C3H6S +

dissociating to CH3CS +

+ CH3based on G3B3 calculations to rationalize the absence of CH3CS +

signal in these experiments.

 2006 Elsevier B.V. All rights reserved.

1. Introduction

The four-membered ring molecule thietane (c-C3H6S) has been the subject of numerous spectral investigations and theoretical calculations as a result of efforts to under-stand the vibrational effects on the ring structure: ring-puckering motions likely alter the planar structure enforced by ring strain[1,2]. The photochemistry of c-C3H6S is also of fundamental importance as photodissociation or photo-ionization of c-C3H6S offers a powerful means to produce reactive H2CS or H2CS+ that plays an important role in atmospheric chemistry and serves as a prototypical system in the chemistry of sulfur-containing organic compounds [3–12].

Several groups reported that photodissociation of c-C3H6S yields exclusively H2CS + C2H4 on excitation at 313 and 254 nm; based on experimental results, those authors proposed that dissociation proceeds through a direct breaking of the C–S bond to form a diradical, with

subsequent rearrangement [3–5]. The channel S + C3H6 was also observed on excitation at 214 and 193 nm; ground-state S(3P) atoms are predominantly produced at 214 nm, but excited S(1D) atoms are formed exclusively at 193 nm[6–8]. With photon energy of excitation increas-ing into the vacuum ultraviolet (VUV) region, dissociative photoionization of c-C3H6S takes place; the dissociation of c-C3H6S+ is expected to become more complicated, as isomerization and rearrangement of ionized polyatomic molecules commonly occur before dissociation [9–11].

Butler and Baer investigated the dissociative photoion-ization of c-C3H6S in a region 8.5–13.5 eV with photo-ionization mass spectroscopy (PIMS); four fragment ions – H2CS+, C2H3S+, HCS+, and C3Hþ5 – were identi-fied with their respective appearance energies (AE) 9.9, 10.0, 10.4, and 10.5 eV at 298 K, determined from the onsets of photoionization efficiency (PIE) curves [9,10]. According to these determined AE, heats of formation of H2CS+, C2H3S+ and HCS+ were derived, but such AE values represent upper limits due to the possible presence of exit channel barriers and kinetic shifts [13,14]. In the present work, we pursue our investigation

0009-2614/$ - see front matter  2006 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2006.03.020

*

Corresponding author. Fax: +886 3 578 3813. E-mail address:schiang@nsrrc.org.tw(S.-Y. Chiang).

of the dissociation of energy-selected c-C3H6S+ in a region 10.2–10.9 eV with a threshold photoelectron-photoion coincidence (TPEPICO) technique to obtain branching ratios and average releases of kinetic energy for channel H2CS++ C2H4. We also predict plausible dissociation mechanisms for channel c-C3H6S+! CH3CS++ CH3 with the G3B3 method to shed light on the lack of observation of CH3CS+ in this energy region.

2. Experiments

The coincidence measurements were conducted on the Seya-Namioka beamline at the NSRRC in Taiwan. Photon energies with resolution 30 meV and photon flux >109 pho-tons s1 in a region 10.2–10.9 eV were selected with a monochromator (1 m; 1200 grooves mm1; slit width 0.15 mm). Absolute photon energies were calibrated within ±0.006 eV on measurement of Rydberg signals in the threshold photoelectron spectrum of Ar.

The molecular beam/threshold photoelectron photoion coincidence (MB/TPEPICO) system is described in detail elsewhere [15,16]. Briefly, a mixture of c-C3H6S (Aldrich, 98%) and He (>99.999%) at a total stagnation pressure 280 Torr and with a seed ratio 10% was expanded through a nozzle and two skimmers to form a cooled c-C3H6S beam to be ionized in the ionization chamber. The threshold electrons produced were extracted with a dc field 100 V m1 and analyzed with a threshold photoelectron spectrometer; the ions produced were extracted with an electric field 2100 V m1 with pulses of duration 30 ls on detecting a threshold electron and analyzed with a time-of-flight mass spectrometer. A time-to-digital converter (TDC) was used to record the flight durations of ions that were triggered with a threshold electron and a signal gener-ated randomly relative to the preceding threshold electron signal. Subtraction of the randomly generated coincidence spectrum from the electron-triggered coincidence spectrum yielded a true coincidence spectrum. All data acquisition was controlled with a computer via a CAMAC interface and the output of TDC was transferred to a computer for further processing.

3. Calculations

All molecular structures and energies of c-C3H6S and species pertinent to this work were calculated using the GAUSSIAN2003 program[17]. For calculations of G3 ener-gy,E0(G3), the equilibrium structure was fully optimized at the MP2(full)/6-31G(d) level, and single-point calcula-tions were performed at levels MP4/6-31G(d), QCISD(T)/

6-31G(d), MP4/6-31+G(d), MP4/6-31G(2df,p), and

MP2(full)/G3large; additional energies include a spin-orbit correction, higher-level corrections and a zero-point vibra-tional energy (ZPVE) calculated from the HF/6-31G(d) vibrational frequencies on scaling by 0.8929. For a compar-ison, we also calculated G3 energies, E0(G3-B), using

B3LYP/6-31G(d) vibrational frequencies with a scaling fac-tor 0.9613 for ZPVE correction and calculated G3B3 ener-gies, E0(G3B3), with single-point calculations performed at an equilibrium structure fully optimized at the B3LYP/6-31G(d) level. Transition structures were located with geom-etries optimized at the B3LYP/6-31G(d) level because some dissociation paths could not be confirmed with intrinsic reaction coordinate (IRC) calculations at the MP2(full)/6-31G(d) level; all identified transition structures were verified to have only one imaginary vibrational frequency.

4. Results and discussion

4.1. Coincidence mass spectra in a region 10.2–10.9 eV The threshold photoelectron (TPE) spectrum of c-C3H6S measured with an energy step 0.01 eV in a region 8.5– 11.5 eV agrees with the photoelectron (PE) spectrum[18]. One broad band observed in a region10.05 to 11.63 eV with maximum at 10.59 eV corresponds to the removal of a 10a1(rC–S) electron; dissociation of energy-selected c-C3H6S+associated with excitation to this ionic excited state was investigated in this work.

The coincidence mass spectra of c-C3H6S were recorded at selected photon energies in a region 10.2–10.9 eV. Fig. 1a–e show corrected coincidence mass spectra of

14750 15000 15250 18500 18750 19000 Flight time/ ns

Coincidence counts/arb. units

(e)10.53 eV (b)10.44 eV (d)10.49 eV (c)10.46 eV c-C₃H₆S+ H₂CS+ (a)10.43 eV

Fig. 1. Coincidence mass spectra of c-C3H6S excited at photon energies (a) 10.43 eV (b) 10.44 eV (c) 10.46 eV (d) 10.49 eV and (e) 10.53 eV.

c-C3H6S excited at 10.43, 10.44, 10.46, 10.49 and 10.53 eV; solid lines indicate ion time-of-flight (TOF) signals fitted to Gaussian shapes, and all spectra are normalized to 5000 total electron triggers. Ion signals at m/z = 74 and 46, cor-responding to c-C3H6S+ and H2CS+, respectively, were identified according to ion flight durations calculated with an equation T0/ns = 2161.8 (m/z)1/2+ 248 obtained from coincidence mass spectra of He, Ar and Kr.

In the figures, TOF signals of c-C3H6S+ fitted to a Gaussian profile have a full width at half maximum (fwhm) 42 ns; the transverse temperature of the molecular beam is accordingly calculated to be 22 K[19]. TOF signals of H2CS+were also fitted well to a Gaussian profile, but the bandwidth of TOF signals increases with increasing photon energy, reflecting releases of kinetic energy upon dissocia-tion. Obtained fwhm of TOF signals of H2CS+ and calcu-lated average releases of kinetic energy for channel c-C3H6S

+

! H2CS +

+ C2H4are listed inTable 1. Average release of kinetic energy,ÆKEæ, was calculated from fwhm according to the Maxwellian equation[19,20]

hKEi ¼ 3 16 ln 2ðeeÞ 2 ðfwhmÞ2 Mp MfðMp MfÞ 3 2RT  Mf ðMp MfÞ ð1Þ in which e = 1.602· 1019C is the charge, e = 2100 V m1 is the strength of the pulsed electric field for ion extraction, Mp and Mf are masses of c-C3H6S+ and H2CS+, and T = 22 K is the transverse temperature of the molecular beam.

4.2. Branching ratios and average releases of kinetic energy Fig. 2shows the branching ratios of c-C3H6S+and frag-ment ion H2CS+ in a region 10.2–10.9 eV obtained from the area ratios of their TOF signals in coincidence mass spectra. H2CS

+

appears in a region 10.39–10.41 eV; a cross at 10.44 eV reflects that half of c-C3H6S

+

ions dissociate to form H2CS+at that energy. Formation of H2CS+is likely attributable to increased internal energy of c-C3H6S+as the dissociation threshold 10.39–10.41 eV lies above the onset at10.20 eV of the excited state of c-C3H6S+observed in

our TPE spectrum and a previous PE spectrum[18].Table 2lists calculated G3 energies – E0(G3) and E0(G3B3) – for species pertinent to dissociative photoionization of c-C3H6S, and energy differences – DE(G3) and DE(G3B3) – relative to c-C3H6S+; E0(G3-B) and DE(G3-B) are not listed due to small differences between DE(G3B3) and DE(G3-B). A predicted G3 energy 10.35 eV for dissociative photoionization of c-C3H6S to form H2CS++ C2H4agrees with an observation of a dissociation threshold 10.39– 10.41 eV.

Fig. 3 shows the calculated average releases of kinetic energy for channel c-C3H6S+! H2CS++ C2H4; a solid line indicates a linear fit to data near the dissociation threshold and a dashed curve results from QET calcula-tions, performed according to an equation formulated by Klots [21–23] hm E0¼ ðr þ 1Þ 2 hKEi þ X i hmi exp hmi hKEi    1 ð2Þ

in which hm is the photon energy, E0= 10.39 ± 0.01 eV is the linearly extrapolated dissociation threshold, r is the number of rotational degrees of freedom and miare vibra-tional wave numbers of H2CS+and C2H4; the latter values are 3038, 2929, 1382, 962, 935 and 830 cm1 for H2CS

+ and 2988, 1656, 1336, 1031, 2967, 1437, 981, 3056, 800, 3033, 1207 and 977 cm1 for C2H4, obtained from the HF/6-31G(d) calculations.

A dissociation threshold at 10.39 ± 0.01 eV obtained by linearly extrapolating data at photon energies smaller than 10.6 eV agrees with a G3 prediction 10.35 eV. Also seen in Fig. 3, QET calculations with a dissociation threshold at 10.39 ± 0.01 eV fit satisfactorily the experimental data near the dissociation threshold. The agreements on dissociation threshold and a statistical distribution ofÆKEæ near thresh-old imply that c-C3H6S+dissociates to H2CS++ C2H4 with-out significant exit channel barrier. Moreover, a statistical

Table 1

Calculated average releases of kinetic energy (ÆKEæ) from the full width at half maximum (fwhm) for channel c-C3H6S

+ ! H2CS

+ + C2H4

PE (eV) fwhm (ns) ÆKEæ (eV)

10.88 111 0.076 10.78 105 0.068 10.67 95 0.055 10.57 88 0.047 10.53 79 0.036 10.51 71 0.028 10.49 64 0.022 10.46 58 0.017 10.45 51 0.013 10.44 54 0.015 10.43 50 0.012 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 0 25 50 75 100 10.44 eV

Relative branching ratios /%

C₃H₆S+

H₂CS+

Photon Energy / eV

Fig. 2. Branching ratios of c-C3H6S+ and H2CS+ in a region 10.2– 10.9 eV; fractional abundances of these two ions were obtained from their total TOF signals.

interpretation of the variation ofÆKEæ with photon energy cannot be valid if unimolecular dissociation occurs from the four-membered ring parent ion; isomerization and cleavage of the C–S bond will, however, give the predicted linear variation near threshold. Lee et al. predicted CH2CH2SCHþ2 ð1þÞ to be less stable than c-C3H6S+ by 0.80 eV with the G3 method and located a transition state connecting CH2CH2SCHþ2 ð1þÞ to c-C3H6S+ with a G3 barrier at 0.64 eV; they also predicted a dissociation energy 0.93 eV for channel CH2CH2SCHþ2 ð1þÞ ! H2CSþþ C2H4, but located no transition state for this dissociation

[24,25]. Accordingly, a dissociation mechanism that c-C3H6S

+

isomerizes to CH2CH2SCHþ2 ð1þÞ before dissoci-ating to H2CS++ C2H4 supports our experimental observation.

4.3. Theoretical predictions

We detected none of the three fragment ions – C2H3S+, HCS+ and C3Hþ5 – identified in a previous PIMS experi-ment in a region 10.2–10.9 eV[9]. The absence of HCS+ signal in this region is consistent with our G3 prediction of 10.99 eV for formation of HCS++ C2H5. The AE of 10.5 eV at 0 K for HCS+in the PIMS experiment is under-estimated according to our G3 prediction and that the resultant DHof;0ðHCS

þ_{Þ ¼ 233  2 kcal mol}1

derived from the AE of 10.5 eV is smaller than the established experi-mental values of 243.9 and 243.5 ± 2 kcal mol1 [12,26]. The AE of 10.6 eV at 0 K for C3Hþ5 in the PIMS experi-ment agrees coincidentally with our G3 prediction of 10.62 eV for formation of C3Hþ5 þ HS despite the possible presence of kinetic shifts and activation barriers. More-over, as the AE of 10.0 eV at 0 K for H2CS+in the PIMS experiment is also significantly smaller than our G3 predic-tion of 10.35 eV and the linearly extrapolated threshold at 10.39 ± 0.01 eV, an underestimation of the AE values for C3Hþ5 and C2H3S+in the PIMS experiment could be possi-ble; discussion on the absence of these two ions in this work based on the PIMS results is thus discarded.

The predicted G3 energies for formation of CH3CS++ CH3, CH2CSH

+

+ CH3, c-C2H3S +

+ CH3 are 9.86, 10.96 and 11.09 eV, respectively; formation of CH3CS

+ + CH3 Table 2

Calculated G3 energies, E0(G3) and E0(G3B3), for species pertinent to dissociative photoionization of c-C3H6S, and energy differences, DE(G3) and DE(G3B3), relative to c-C3H6S+

Species Sym. E0(G3) (hartree) E0(G3B3) (hartree) DE(G3) (eV) DE(G3B3) (eV)

c-C3H6S Cs 515.82967 515.83323 8.68 8.67 c-C3H6S+ Cs 515.51058 515.51474 0.00 0.00 CH2CH2SCHþ2 ð1þÞ Cs 515.48267 0.76 cis-CH3CHSCHþ2 ð2þÞ Cs 515.50562 515.50855 0.13 0.17 trans-CH3CHSCHþ2 ð3þÞ Cs 515.50632 515.50930 0.12 0.15 cis-CH3CSCHþ3 ð4þÞ Cs 515.48718 515.49115 0.64 0.64 trans-CH3CSCHþ3 ð4 0 þÞ Cs 515.48937 515.49290 0.58 0.59 CH2CHSCHþ3 ð5þÞ Cs 515.52267 515.52565 0.33 0.30 TS1 C1 515.45757 515.46128 1.44 1.45 TS2 C1 515.45048 1.75 TS3 C1 515.48706 0.75 TS4 C1 515.43070 515.43420 2.17 2.19 TS5 C1 515.45949 515.46277 1.39 1.41 TS6 C1 515.44007 515.44349 1.92 1.94 C2H4 D2h 78.50741 78.50930 H2CS + C2v 436.94173 436.94368 CH3 D3h 39.79329 39.79362 CH3CS+ C3v 475.67411 475.67700 CH2CSH+ Cs 475.63377 475.63672 c-C2H3S+ Cs 475.62884 475.63155 HCS+ _C 1v 436.36196 436.36414 C2H5 Cs 79.06398 79.06569 C3Hþ5 C2v 116.84411 116.84657 HS C1v 398.59531 398.59664 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 11.0 0 50 100 coincidence data QET calculations linear fits E threshold = 10.39 eV

Average kinetic energy releases / meV

Photon energy / eV

Fig. 3. Average kinetic energies released in the channel c-C3H6S+! H2CS++ C2H4 with excitation at photon energies in a region 10.4– 10.9 eV.

is the most likely channel on energetic grounds and requires the least energy relative to 10.35 eV for formation of H2CS

+

+ C2H4. If this is the case, the absence of CH3CS+ signal in this work indicates that greater activa-tion energies during dissociaactiva-tion are applicable for H migrations and structural alterations. We explored dissoci-ation mechanisms for channel c-C3H6S+! CH3CS++ CH3with the G3B3 method.

Fig. 4a,b show relative energies DE(G3B3) of two

feasi-ble dissociation mechanisms except DE(G3-B) for

CH2CH2SCHþ2 ð1þÞ due to the failure of G3B3 calculations at the QCISD(T)/6-31G(d) level. Relative energies DE(G3) are not adopted in the figure because reaction paths for TS2 and TS5 optimized at the MP2(full)/6-31G(d) level could not be confirmed with IRC calculations. Nevertheless, according to Table 2, the difference between DE(G3B3) and DE(G3) is only 0.04 eV despite DE(G3B3) greater than

DE(G3). InFig. 4a, c-C3H6S+first breaks the C–C bond via TS1 to form intermediate CH2CH2SCHþ2 ð1þÞ with a G3B3 barrier at 1.45 eV. Next, CH2CH2SCHþ2 ð1þÞ undergoes H migration from the central CH2group to the terminal CH2 group via TS2 with a barrier at 1.75 eV to form intermediate CH3CHSCHþ2 ð2þÞ. Finally, cis-CH3CHSCHþ2 ð2þÞ rotates dihedral angle \CCSC via TS3 with a barrier at 0.75 eV to form trans-CH3CHSCHþ2 ð3þÞ, which subse-quently proceeds through H migration via a four-membered ring TS4 with a barrier at 2.19 eV to form a dissociation pre-cursor trans-CH3CSCHþ3 ð4

0_{þÞ. The other dissociation} pre-cursor cis-CH3CSCHþ3 ð4þÞ might be formed from intermediate cis-CH3CHSCHþ2 ð2þÞ. In Fig. 4b, cis-CH3CHSCHþ2 ð2þÞ proceeds through H migrations from the CH3group to the CH2 group and from the CH group to the CH2group via TS5 and TS6 with their pre-dicted barriers at 1.41 and 1.94 eV, respectively, to form a

C C _S C H H H H H H TS3 C C S C H H H H H H 3+ C C S C H H H H H H TS4 C C C S H H H H H H 4'+ + + + + 0.75 0.15 2.19 0.59 C C C _S H H HH H H C C C S H H _H H H H C C H H H H C S H H C _C C S H H H H H H 1+ TS2 2+ + + + + 0.00 0.81 1.45 0.17 C C S C H H H H H H TS1 1.75 + 1.20 CH₃CS++ CH₃ C _C C _S H H H H H H 2+ TS5 5+ TS6 4+ + + + + 0.17 C _C C S H H H_H H H C C _S C H H H H H H C C C S H H H H H H C C C S H H H H H H 1.41 1.94 -0.30 0.64 1.20 CH₃CS++ CH₃ + (a) (b)

Fig. 4. Theoretical predictions of relative energies DE(G3B3) of two feasible dissociation mechanisms for channel c-C3H6S + ! CH3CS + + CH3 (a) isomerization of c-C3H6S + into trans-CH3CSCHþ3 ð4 0_{þÞ (b) isomerization of cis-CH} 3CHSCHþ2 ð2þÞ into cis-CH3CSCHþ3 ð4þÞ.

more stable isomer CH2CHSCHþ3 ð5þÞ and a dissociation precursor cis-CH3CSCHþ3 ð4þÞ. Finally, a direct cleavage of the C–S bond in trans-CH3CSCHþ3 ð4

0_{þÞ or in} cis-CH3CSCHþ3 ð4þÞ to form CH3CS++ CH3requires no exit barrier.

Experimentally, we observed no CH3CS+ signal in a region 10.2–10.9 eV. c-C3H6S excited at 10.2–10.9 eV cor-responds to 1.53–2.23 eV of internal energy in the parent ion, and the two barriers TS4 and TS6 are predicted to be greater in energy, 2.19 and 1.94 eV above the ground state of c-C3H6S+. Accordingly, if reactions via TS4 and TS6 are the rate-determining steps and the barriers must be surmounted, then CH3CS+ signal is absence in this region.

5. Conclusion

We have investigated the dissociation of energy-selected c-C3H6S+to form H2CS++ C2H4in a region 10.2–10.9 eV with a TPEPICO technique. We obtained branching ratios and average releases of kinetic energy for channel c-C3H6S+! H2CS++ C2H4 from the coincidence mass spectra. Small releases of kinetic energy near the appear-ance threshold agree with QET calculations with a linearly extrapolated dissociation threshold at 10.39 ± 0.01 eV. This statistical energy distribution is supported by a pre-dicted dissociation mechanism that proceeds without an exit barrier. We discuss plausible dissociation mechanisms for channel c-C3H6S+! CH3CS++ CH3 of the least energy to rationalize the lack of observation of CH3CS+ based on G3B3 calculations.

Acknowledgement

We thank the NSRRC and the NSC of Taiwan (Con-tract No. NSC94-2113-M-213-002) for financial support

and the NCHC for providing computing time for calculations.

References

[1] D.M. Turnbull, M.G. Sowa, B.R. Henry, J. Phys. Chem. 100 (1996) 13433.

[2] D.O. Harris, H.W. Harrington, A.C. Luntz, W.D. Gwinn, J. Chem. Phys. 44 (1966) 3467.

[3] D.R. Dice, R.P. Steer, J. Phys. Chem. 77 (1973) 434. [4] D.R. Dice, R.P. Steer, Can. J. Chem. 53 (1975) 1744. [5] S. Braslavsky, J. Heicklen, Chem. Rev. 77 (1977) 473.

[6] F.H. Dorer, M.E. Okazaki, K.E. Salomon, J. Phys. Chem. 85 (1981) 2671.

[7] H.A. Wiebe, J. Heicklen, J. Am. Chem. Soc. 92 (1970) 7031. [8] F. Qi, L. Sheng, M. Ahmed, D.S. Peterka, T. Baer, Chem. Phys. Lett.

357 (2002) 204.

[9] J.J. Butler, T. Baer, Org. Mass Spectrom. 18 (1983) 248. [10] J.J. Butler, T. Baer, J. Am. Chem. Soc. 104 (1982) 5016.

[11] M.J. Polce, C. Wesdemiotis, Rapid. Commun. Mass Spectrom. 10 (1996) 235.

[12] B. Ruscic, J. Berkowitz, J. Chem. Phys. 98 (1993) 2568.

[13] B.P. Tsai, T. Baer, A.S. Werner, S.F. Lin, J. Phys. Chem. 79 (1975) 570.

[14] H. Shiromaru, Y. Achiba, Y.T. Lee, J. Phys. Chem. 91 (1987) 17. [15] Y.-S. Fang, I.-F. Lin, Y.-C. Lee, S.-Y. Chiang, J. Chem. Phys. 123

(2005) 054312.

[16] S.-Y. Chiang, C.-I. Ma, J. Phys. Chem. A 104 (2000) 1991. [17] M.J. Frisch, et al., Gaussian Inc., Pittsburgh, PA, 2003. [18] P.D. Mollere, K.N. Houk, J. Am. Chem. Soc. 99 (1977) 3226. [19] R. Stockbauer, Int. J. Mass Spectrom. Ion Phys. 25 (1977) 89. [20] T. Baer, G.D. Whillett, D. Smith, J.S. Phillips, J. Chem. Phys. 70

(1979) 4076.

[21] C.E. Klots, J. Chem. Phys. 58 (1973) 5364. [22] C.E. Klots, Adv. Mass Spectrom. 6 (1974) 969. [23] C.E. Klots, J. Chem. Phys. 64 (1976) 4269.

[24] H.-L. Lee, W.-K. Li, S.-W. Chiu, J. Mol. Struct. (THEOCHEM) 625 (2003) 237.

[25] H.-L. Lee, W.-K. Li, S.-W. Chiu, J. Mol. Struct. (THEOCHEM) 620 (2003) 107.

[26] S.-Y. Chiang, Y.-S. Fang, J. Electron Spectrosc. Relat. Phenom. 144 (2005) 223.