Chinese Journal of Chemical Physics  2016, Vol. 29 Issue (6): 645-649

#### The article information

Meng Li, Xu Shan, Shan-shan Niu, Ya-guo Tang, Fang Wu, Chun-kai Xu, Xiang-jun Chen

High Resolution Electron Momentum Spectroscopy Study on Ethanol: Orbital Electron Momentum Distributions for Individual Conformers

Chinese Journal of Chemical Physics, 2016, 29(6): 645-649

http://dx.doi.org/10.1063/1674-0068/29/cjcp1604080

### Article history

Received on: April 18, 2016
Accepted on: May 12, 2016
High Resolution Electron Momentum Spectroscopy Study on Ethanol: Orbital Electron Momentum Distributions for Individual Conformers
Meng Li, Xu Shan, Shan-shan Niu, Ya-guo Tang, Fang Wu, Chun-kai Xu, Xiang-jun Chen
Dated: Received on April 18, 2016; Accepted on May 12, 2016
Hefei National Laboratory for Physical Sciences at the Microscale and Department of Modern Physics, University of Science and Technology of China, Hefei 230026, China
*Author to whom correspondence should be addressed. Xu Shan, E-mail:xshan@ustc.edu.cn
Abstract: The outer-valence binding energy spectra of ethanol in the energy range of 9-21 eV are measured by a high-resolution electron momentum spectrometer at an impact energy of 2.5 keV plus the binding energy. The electron momentum distributions for the ionization peaks corresponding to the outer-valence orbitals are obtained by deconvoluting a series of azimuthal angular correlated binding energy spectra. Comparison is made with the theoretical calculations for two conformers, trans and gauche, coexisting in the gas phase of ethanol at the level of B3LYP density functional theory with aug-cc-pVTZ basis sets. It is found that the measured electron momentum distributions for the peaks at 14.5 and 15.2 eV are in good agreement with the theoretical electron momentum distributions for the molecular orbitals of individual conformers (i.e., 8a' of trans and 9a of gauche), but not in accordance with the thermally averaged ones. It demonstrates that the high-resolution electron momentum spectrometer, by inspecting the molecular electronic structure, is a promising technique to identify different conformers in a mixed sample.
Key words: Electron momentum spectroscopy    Individual conformer    Ethanol    Density functional theory
Ⅰ. INTRODUCTION

Electron momentum spectroscopy (EMS), also known as binary (e, 2e) spectroscopy, has been a powerful experimental technique in investigating the electron structures of atoms and molecules, which is based on the unique ability to directly obtain the electron density distributions for individual orbitals in momentum space [1-3]. During the past decades, the EMS technique has been extended to explore the influence of molecular conformations on the electron binding energies of molecular orbitals (MOs) and the corresponding electron momentum distributions (EMDs) for a series of structurally flexible molecules such as n-glycine [4, 5], n-butane [6-8], 1-butene [9-11], tetrahydrofuran [12-14], ethylamine [15], ethanethiol [16], ethanol [17-21], and 2-fluoroethanol [22]. However, variations of the electron binding energies induced by the molecular conformations are usually quite small, and make it difficult to resolve the spectral bands ionized from the correlated molecular orbitals of different conformers under the limited energy resolution of EMS. As a result, all the EMDs acquired in previous EMS experiments on the conformational molecules were invariably the compound results from more than one conformers [4-22]. Correspondingly the interpretation of EMS experiments becomes more intricate and requires extensive theoretical studies. Therefore, it is an essential challenging task to obtain the EMDs for individual conformers.

Ethanol, a common organic molecule in life and industry, exists as two stable conformers named trans and gauche in its electronic ground state due to the internal rotation of hydroxyl group about the C-O axis [23]. It has been determined that the relative abundance of trans conformer is about 39% and those of two equivalent gauche conformers are about 61% at room temperature, according to the theoretical calculations [18] and the various experiments employing microwave spectroscopy [24], infrared spectroscopy [23, 25], and synchrotron radiation photoelectron spectroscopy [26]. In the recent EMS experiments on ethanol [17-21], the measured EMDs, corresponding to the ionization peaks from the cooperative contributions of two conformers, have been explained quantitatively by using the thermally averaged theoretical calculations in which the relative abundances (39.3% trans and 60.7% gauche, or 20% trans and 80% gauche) are taken into account. But for some MOs, the calculated EMDs failed to reproduce the experimental ones quantitatively [17, 18, 21]. Recently, molecular dynamical simulations of Hajgato et al. [19] improved the agreement of theory with experiment in view of the ultrafast nuclear dynamics in the initial and final states. But the discrepancy between t-newpage hspace heory and experiment still remained. On the other hand, Morini et al. [18] suggested that a shoulder at~14.5 eV observed in the He Ⅰ photoelectron spectrum can be specifically ascribed to the trans conformer fraction according to the high-precise theoretical calculation of the ionization energies (IEs) using one-particle Green's function (1p-GF) theory in conjunction with the so-called third-order algebraic diagrammatic construction scheme (ADC (3)) [18]. It is worth noticing that the IEs of the corresponding MOs (8a' for trans and 9a for gauche) for the two conformers are separated about 0.9 eV, indicating the presence of significant conformational fingerprints. Thus it could be possible to extract the EMDs for individual conformers in ethanol by high-resolution EMS measurement.

In this work, we report a high-resolution EMS experiment on ethanol to obtain tentatively the EMDs for individual conformers at the electron impact energy of about 2.5 keV. The outer-valence binding energy spectra (BES) in the range of 9-21 eV have been obtained, as well as the EMDs corresponding to the observed peaks in the BES for ethanol. The experimental EMDs are compared with the theoretical ones calculated by B3LYP/aug-cc-pVTZ. It is found that the EMDs for the peaks at 14.5 and 15.2 eV in the BES are in good agreement with the theoretical EMDs for individual MOs 8a' (trans) and 9a (gauche) of pure conformers, respectively, indicating that it is possible to investigate the electronic structures of individual conformers with high-resolution EMS method.

Ⅱ. EXPERIMENTAL AND THEORETICAL BACKGROUND

EMS is based on the kinematically complete (e, 2e) process in which fast incident electron induces the ionization of target atom or molecule. The present experiment of ethanol molecule is carried out by the high-resolution (e, 2e) spectrometer employed the non-coplanar asymmetric geometry which has been described in detail elsewhere [27]. Briefly, an incident electron beam generated from the electron gun is monochromized by a monochromator, and accelerated by a lens system to the desired energy of 2.5 keV plus the binding energy, and then transferred to the reaction region where the electron beam impacts with the target molecules injected by a nozzle. The scattered electron outgoing along polar angle $\theta_\textrm{a}$ =14 $^{\circ}$ enters a fast electron analyzer and is detected by a two-dimensional position sensitive detector over a large range of both energies and azimuthal angles ( $\phi$ ) of interest. The ionized electron outgoing along polar angle $\theta_\textrm{b}$ =76 $^{\circ}$ enters a slow electron analyzer and is detected by a one-dimensional position sensitive detector. Under such experimental condition, considering conservation of energy and momentum, the binding energy $\varepsilon$ and magnitude of momentum $p$ of the target electron can be expressed by:

 $\varepsilon = {E_0} - {E_{\text{a}}} - {E_{\text{b}}}$ (1)
 $\begin{gathered} p = [p_0^2 + p_{\text{a}}^2 + p_{\text{b}}^2-2{p_0}{p_{\text{a}}}{\text{cos}}{\theta _{\text{a}}}-2{p_0}{p_{\text{b}}}{\text{cos}}{\theta _{\text{b}}} + \hfill \\ 2{p_{\text{a}}}{p_{\text{b}}}({\text{cos}}{\theta _{\text{a}}}{\text{cos}}{\theta _{\text{b}}}-{\text{sin}}{\theta _{\text{a}}}{\text{sin}}{\theta _{\text{b}}}{\text{cos}}\phi ){]^{1/2}} \hfill \\ \end{gathered}$ (2)

where $E_i$ , $p_i$ (i=0, a, b) are the energies and momenta of the incident, scattered and ejected electrons, respectively and $\phi$ is the relative azimuthal angle between the two outgoing electrons. Therefore, the binding energy and the momentum of the target electron can be determined by detecting the two outgoing electrons in coincidence. Before the experiment of ethanol, the energy and momentum resolution of the present EMS spectrometer are determined to be~0.6 eV (full width at half maximum) and~0.1 a.u. respectively by measuring the ionization spectrum and electron momentum distribution of Ar 3p.

Within the binary encounter approximation and the plane wave impulse approximation, as well as the target Hartree-Fork or Kohn-Sham (KS) approximation, the triple differential cross section (TDCS) of (e, 2e) reaction process for randomly oriented atom or molecule can be expressed as [1-3].

 $\sigma_{\textrm{EMS}}\infty S_j^{(f)} \int|\psi_j (\textbf{p})|^2 \textrm{d}\Omega$ (3)

where $\psi_j$ ( $\textbf{p}$ ) represents the one-electron canonical HF or KS wave function in momentum space for the $j$ th orbital from which the electron is ejected, and $S_j^{(f)}$ denotes the spectroscopic factor or pole strength which is the possibility of forming an one-hole configuration in the final state $f$ . The integral in Eq.(3) is known as the spherical averaged electron momentum distribution, i.e., electron momentum profile.

Ⅲ. RESULTS AND DISCUSSION

The trans conformer of ethanol has C $_\textrm{s}$ point group symmetry and the electronic configuration for its ground state is:

 $\begin{gathered} {({\text{core}})^6}\underbrace {{{(4{\text{a'}})}^2}{{(5{\text{a'}})}^2}{{(6{\text{a'}})}^2}}_{{\text{inner - valence}}} \hfill \\ \underbrace {{{(7{\text{a'}})}^2}{{(1{\text{a''}})}^2}{{(8{\text{a'}})}^2}{{(9{\text{a'}})}^2}{{(2{\text{a''}})}^2}{{(10{\text{a'}})}^2}{{(3{\text{a''}})}^2}}_{{\text{outer - valence}}} \hfill \\ \end{gathered}$

While, the gauche form of ethanol belongs to the C $_1$ point group and its ground state electronic configuration can be written as:

 $\begin{gathered} {({\text{core}})^6}\underbrace {{{(4{\text{a}})}^2}{{(5{\text{a}})}^2}{{(6{\text{a}})}^2}}_{{\text{inner - valence}}} \hfill \\ \underbrace {{{(7{\text{a}})}^2}{{(8{\text{a}})}^2}{{(9{\text{a}})}^2}{{(10{\text{a}})}^2}{{(11{\text{a}})}^2}{{(12{\text{a}})}^2}{{(13{\text{a}})}^2}}_{{\text{outer - valence}}} \hfill \\ \end{gathered}$

The vertical IEs of the outer-valence MOs of ethanol have been measured by the photoelectron spectroscopy [28, 29] and calculated by the ADC (3) method [18], which are listed in Table Ⅰ. In the present work, the outer-valence BES of ethanol in the energy range from 9 eV to 21 eV is measured simultaneously in the desired range of azimuthal angles by a high-resolution EMS spectrometer, and the summed spectrum over all the measured relative azimuthal angles ( $\phi$ ) is displayed in Fig. 1. Five resolved peaks and some weak structures can be observed in the BES, which has a similar outline with the previous high-resolution photoelectron spectra (PES) measured by He Ⅰ [28] and He Ⅱ [29] ultraviolet radiation sources. A series of Gaussian functions are employed to fit the BES corresponding to the ionization from the seven outer-valence MOs for trans and gauche conformers. The widths of the Gaussian peaks are from the cooperative contributions of the EMS instrumental energy resolution and the Franck-Condon widths of the ionization bands observed in the He Ⅱ PES [29]. Peak positions are mainly referred to the IEs reported by the He Ⅱ PES study [29].

 FIG. 1 The outer-valence binding energy spectrum of ethanol measured by high-resolution EMS. The dash lines represent the fitted Gaussian peaks whose positions are indicated by vertical bars, while the solid line is the summed fit.

The first peak (P1 in Fig. 1) at 10.7 eV is well resolved and attributed to the ionization from the highest occupied molecular orbitals (HOMOs) of ethanol, i.e., 3a" for the trans and 13a for the gauche (denoted as 3a"/13a) which have almost the same IEs according to the calculations depicted in Table Ⅰ. Similarly, the next three peaks (P2, P3, P4) at 12.1, 13.2, and 13.9 eV stand for the ionizations from the MOs of 10a'/12a, 2a"/11a and 9a'/10a, respectively. The last two peaks (P7, P8) located at 16.0 and 17.4 eV correspond to the ionizations from 1a"/8a and 7a'/7a, respectively. It is worth to be noticed that the fifth peak (P5) at 14.5 eV which relates to the weak peak observed in the He Ⅰ PES was proposed by Morini et al. [18] to be ascribed to 8a' orbital of trans conformer according to their ADC (3) calculation. Correspondingly, an extra peak, peak six (P6), at 15.2 eV is additionally introduced to represent the ionization from 9a orbital of gauche conformer. This is reasonable because the conformational effect of ethanol makes the IE of gauche 9a orbital well separated (more than 0.7 eV) from the adjacent ones according to the ADC (3) calculation [18].

Table Ⅰ Experimental and theoretical ionization energies (IEs) for the outer-valence orbitals of ethanol (in eV).

The experimental EMDs for the eight peaks are extracted by deconvoluting a series of azimuthal angular correlated BES and plotting the area under the corresponding fitted peaks as a function of electron momentum $p$ (i.e., azimuthal angle $\phi$ ). The theoretical EMDs for the outer-valence MOs of ethanol are calculated using DFT-B3LYP method with aug-cc-pVTZ basis sets by the Gaussian 03 program [30]. In the comparison of the experimental and theoretical EMDs, two different ratios of 39/61 and 20/80 for trans and gauche conformers are taken into account, which were used in the previous studies [17-20]. It is found that for the whole outer-valence MOs the theoretical EMDs employed the ratio of 39/61 are better than those employed the ratio of 20/80 in reproducing the experimental EMDs except for the HOMO, which is similar to the previous findings [19]. So only the conformational ratio of 39/61 will be used in the following discussion. Moreover, the experimental and thermally averaged theoretical EMDs are placed on a common intensity scale using a universal factor acquired by normalizing the experimental and theoretical EMDs for HOMOs (3a"/13a), and all the theoretical EMDs have been convoluted by the present EMS instrumental momentum resolution of 0.1 a.u. using the Gaussian-weighted planar grid method [31]. Figure 2 shows the experimental and thermally averaged theoretical EMDs for the eight ionization peaks from the outer-valence orbitals of ethanol, together with the theoretical EMDs and orbital maps of individual MOs for trans and gauche conformers. In addition, the error bars of experimental data given in Fig. 2 represent the overall error of the statistical and the deconvolution uncertainties.

 FIG. 2 The experimental EMDs for the outer-valence ionization bands (peak 1-8) of ethanol and the thermally averaged theoretical EMDs including 39% trans and 61% gauche conformers, as well as the individual EMDs and the orbital maps calculated by using B3LYP/aug-cc-pVTZ.

Figure 2(a) presents the experimental EMD for the first ionization band (P1), together with the corresponding theoretical one for the HOMOs of which the O 2p electron lone pair predominates. The internal rotation of hydroxyl group about the C-O bond makes the individual EMDs of two conformers showing two remarkably different characteristics, i.e., a p-type profile for trans and a mixed s-and p-type profile for gauche. The measured momentum profile, also having a mixed sp-type character with a maximum intensity at p $\approx$ 0.85 a.u.}, can be reproduced by the thermally averaged calculation (contributed from 39% 3a" and 61% 13a) expect for the discrepancy in the low momentum region. Such discrepancies between theory and experiment are also found in previous EMS studies [17-19]. The ultrafast nuclear dynamics in the initial and final states such as thermal disorder and molecular distortion was invoked to explain it but improvement is very limited [18, 19]. In addition, referring to the previous EMS study [20], the observed noticeable high intensity for the HOMO profile at the zero momentum could be attributed to the hyperconjugative effect between the O 2p lone pair and $\sigma^*$ $_{\textrm{C}-\textrm{C}}$ or $\sigma^*$ $_{\textrm{C}-\textrm{H}}$ bonds.

We also compare the experimental EMDs for the other peaks (P2-P8) with the thermally averaged ones for the corresponding MOs (MO12-MO7) in Fig. 2(b)-(h), respectively. It can be seen that the measured EMDs of peaks 2, 3, 4, 7 and 8 are basically consistent with the summed momentum distributions for the related orbitals of two conformers.

For the peak at 14.5 eV (P5) which was proposed to the trans conformer fraction [18], comparison is made between the experimental EMD with the thermally averaged calculation for 8a'/9a and the individual result for 8a' as shown in Fig. 2(e). As can be seen, the individual theoretical EMD for 8a' reproduces the experimental one much better than the thermally averaged result for 8a'/9a. Similarly, for the peak 6 at 15.2 eV, the individual theoretical EMD for 9a is in agreement with the experiment much better than the thermally averaged calculation for 8a'/9a as shown in Fig. 2(f). These results indicate that the orbital EMDs of individual conformers of ethanol have successfully been obtained, which has not been reported before. It is proved that the fingerprint of the molecular conformation for the electronic structures, especially for the orbital EMDs for individual conformers, could possibly be investigated by the high-resolution EMS technique.

Ⅳ. CONCLUSION

The high-resolution EMS measurement on the outer-valence BES of ethanol in the range of 9-21 eV and the corresponding orbital EMDs have been reported in this work. Taking the relative abundance of 39% for trans and 61% for gauche conformers into account, the thermally averaged EMDs calculated using B3LYP method with aug-cc-pVTZ basis sets can reproduce the experimental ones for most of the MOs well. Meanwhile, the individual orbital EMDs for the pure trans (8a') or gauche (9a) conformers in ethanol have been extracted assuredly for the first time, indicating that it is feasible to investigate the electronic structures of pure conformers of molecules with high-resolution EMS technique.

Ⅴ. ACKNOWLEDGMENTS

This work was partially supported by the National Natural Science Foundation of China (No.11534011 and No.11327404).

Reference
 [1] E. Weigold and I. E. McCarthy, Electron Momentum Spectroscopy, New York:Kluwer Academic/Plenum Publishers, (1999). [2] I. E. McCarthy, and E. Weigold, Rep. Prog. Phys. 54 , 789 (1991). DOI:10.1088/0034-4885/54/6/001 [3] M. A. Coplan, J. H. Moore, and J. P. Doering, Rev. Mod. Phys. 66 , 985 (1994). DOI:10.1103/RevModPhys.66.985 [4] Y. Zheng, J. J. Neville, and C. E. Brion, Science 270 , 786 (1995). DOI:10.1126/science.270.5237.786 [5] J. J. Neville, Y. Zheng, and C. E. Brion, J. Am. Chem. Soc. 118 , 10533 (1996). DOI:10.1021/ja9613015 [6] W. N. Pang, J. F. Gao, C. J. Ruan, R. C. Shang, A. B. Trofimov, and M. S. Deleuze, J. Chem. Phys. 112 , 8043 (2000). DOI:10.1063/1.481403 [7] M. S. Deleuze, W. N. Pang, A. Salam, and R. C. Shang, J. Am. Chem. Soc. 123 , 4049 (2001). DOI:10.1021/ja0039886 [8] F. Wang, and M. Downton, J. Phys. B 37 , 557 (2004). DOI:10.1088/0953-4075/37/3/003 [9] X. J. Chen, F. Wu, X. Shan, and K. Z. Xu, J. Phys:Conf. Ser. 80 , 012003 (2007). DOI:10.1088/1742-6596/80/1/012003 [10] F. Wu, X. J. Chen, X. Shan, S. X. Tian, Z. J. Li, and K. Z. Xu, J. Phys. Chem. A 112 , 4360 (2008). DOI:10.1021/jp710757y [11] S. H. R. Shojaei, J. Vandenbussche, M. S. Deleuze, and P. Bultinck, J. Phys. Chem. A 117 , 8388 (2013). DOI:10.1021/jp405122p [12] T. Yang, G. L. Su, C. G. Ning, J. K. Deng, F. Wang, S. F. Zhang, X. G. Ren, and Y. R. Huang, J. Phys. Chem. A 111 , 4927 (2007). DOI:10.1021/jp066299a [13] C. G. Ning, Y. R. Huang, S. F. Zhang, J. K. Deng, K. Liu, Z. H. Luo, and F. Wang, J. Phys. Chem. A 112 , 11078 (2008). DOI:10.1021/jp8038658 [14] S. H. R. Shojaei, F. Morini, and M. S. Deleuze, J. Phys. Chem. A 117 , 1918 (2013). DOI:10.1021/jp310722a [15] M. Yan, X. Shan, F. Wu, X. X. Xia, K. D. Wang, K. Z. Xu, and X. J. Chen, J. Phys. Chem. A 113 , 507 (2009). DOI:10.1021/jp808281w [16] X. X. Xue, M. Yan, F. Wu, X. Shan, K. Z. Xu, and X. J. Chen, Chin. J. Chem. Phys. 21 , 515 (2008). DOI:10.1088/1674-0068/21/06/515-520 [17] C. G. Ning, Z. H. Luo, Y. R. Huang, B. Hajgat, F Morini, K. Liu, S. F. Zhang, J. K. Deng, and M. Deleuze, J. Phys. B 41 , 175103 (2008). DOI:10.1088/0953-4075/41/17/175103 [18] F. Morini, B. Hajgato, M. S. Deleuze, C. G. Ning, and J. K. Deng, J. Phys. Chem. A 112 , 9083 (2008). DOI:10.1021/jp804284p [19] B. Hajgato, M. S. Deleuze, and F. Morini, J. Phys. Chem. A 113 , 7138 (2009). [20] X. J. Chen, F. Wu, M. Yan, H. B. Li, S. X. Tian, X. Shan, K. D. Wang, Z. J. Li, and K. Z. Xu, Chem. Phys. Lett. 472 , 19 (2009). DOI:10.1016/j.cplett.2009.02.064 [21] F. Wu, Ph.D Dissertation, Hefei:University of Science and Technology of China (2007). [22] Y. F. Shi, X. Shan, E. L. Wang, H. J. Yang, W. Zhang, and X. J. Chen, Chin. J. Chem. Phys. 28 , 35 (2015). DOI:10.1063/1674-0068/28/cjcp1410175 [23] M. L. Senent, Y. G. Smeyers, R. Domingues-Gomez, and M. Villa, J. Chem. Phys. 112 , 5809 (2000). DOI:10.1063/1.481155 [24] J. C. Pearson, K. V. L. N. Sarsty, and E. Herbst, F. C. de Lucia, J. Mol. Spectrosc. 175 , 146 (1996). [25] S. Coussan, Y. Bouteiller, J. P. Perchard, and W. Q. Zheng, J. Phys. Chem. A 102 , 5789 (1998). DOI:10.1021/jp9805961 [26] M. Abu-samha, and K. J. Borve, Phys. Rev. A 74 , 042508 (2006). DOI:10.1103/PhysRevA.74.042508 [27] X. Shan, X. J. Chen, L. X. Zhou, Z. J. Li, T. Liu, X. X. Xue, and K. Z. Xu, J. Chem. Phys. 125 , 154307 (2006). DOI:10.1063/1.2358981 [28] K. Kimura, S. Katsuwata, Y. Achiba, T. Yamazaki, and S. Iwata, Handbook of HeI Photoelectron Spectra of Fundamental Organic Molecules, New York:Halsted Press, (1981). [29] W. V. Niessen, G. Bieri, and L. Asbrink, J. Electron Spectrosc. Relat. Phenom. 21 , 175 (1980). DOI:10.1016/0368-2048(80)85046-8 [30] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Jr. Montgomery, T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople, Gaussian 03, Revision B 04, Pittsburgh PA:Gaussian, Inc., (2003). [31] P. Duffy, M. E. Casida, C. E. Brion, and D. P. Chong, Chem. Phys. 159 , 347 (1992). DOI:10.1016/0301-0104(92)87062-E