-
Journal of the Korean Physical Society, Vol. 59, No. 5, November
2011, pp. 31923200
Infrared and Raman Spectroscopic Studies of
Tris(trimethylsilyl)silaneDerivatives of (CH3)3Si)3Si-X [X = H, Cl,
OH, CH3, OCH3, Si(CH3)3]:Vibrational Assignments by Hartree-Fock
and Density-functional Theory
Calculations
Bong Hyun Boo
Department of Chemistry, Chungnam National University, Daejeon
305-764, Korea
(Received 3 February 2011)
IR and Raman spectra were measured to elucidate the vibrational
structures oftris(trimethylsilyl)silane derivatives
[((CH3)3Si)3Si-X, X = H, Cl, OH, CH3, OCH3, and Si(CH3)3]in the
fundamental state. Hartree-Fock (HF) and density-functional theory
(DFT) calculationswere carried out to study the molecular structure
and the vibrational spectra. In the IR spectra,two scaling factors
of 0.978 and 0.917 were applied to the low (below 2300 cm1) and the
high(above 2300 cm1) energy fundamental frequencies giving rms
deviations of 44.0 and 20.4 cm1,respectively. In the Raman spectra,
however, a uniform scaling factor of 0.900 was applied to thewhole
spectra, yielding a rms deviation 40.0 cm1. Comparison between the
experimental and thesimulated IR and Raman spectra of the
tris(trimethylsilyl)silane derivatives were made. We clearlyobserve
the Si-H stretching fundamental at 2052 and 2048 cm1 in the IR and
the Raman spectra,respectively. The intensities for the vibrations
are relatively intense compared with those of theC-H stretching
fundamental. The skeletal vibrations involving Si-Si are found to
have relativelyweak intensities in the IR and the Raman spectra.
Pure Si-O stretching at 611 cm1 in the IRspectrum is clearly
observed to have an intense absorption both in the IR and the Raman
spectra.The strong IR intensity is due to the electronegativity
difference between the corresponding twoatoms.
PACS numbers: 82.50.FvKeywords: Tris(trimethylsilyl)silane
derivatives, IR, Raman, HF, DFT, Vibrational assignmentDOI:
10.3938/jkps.59.3192
I. INTRODUCTION
Polysilane consisting of Si-Si central bonds and Si-Hand Si-C
terminal bonds has received considerable at-tention due to its
spectroscopic properties [1]. Owingto the higher polarizability of
the Si atom in comparisonwith C and H atoms, the normal modes of
the vibrationsinvolving displacements of Si atoms have been found
tohave significantly enhanced Raman intensity [25]. Cost-effective
density-functional theory (DFT) calculationsare predominantly used
for vibrational assignments ofmolecules [617]. On the other hand,
peak assignment ofRaman spectrum is accomplished by using the
Hartree-Fock (HF) method. Skeletal vibrational spectroscopy
in-volving the Si-Si backbone have been used to researchtargets
because IR and Raman spectroscopies are impor-tant tools for
analyzing polysilanes [18]. Displacementsof electron-rich atoms
give rise to intense Raman scatter-
E-mail: [email protected]; Tel: +82-42-821-6551; Fax:
+82-42-821-8896
ing owing to the high polarizability of the correspondingatoms.
The major goals to achieve in the present studyare (1) to
synthesize various polysilanes, (2) to measurethe IR and Raman
spectra, (3) to reproduce reasonableIR and Raman spectra by using
fundamental vibrationalpeaks and their corresponding intensities
obtained withDFT method and Lorentzian statistics which can be
ap-plied to the simulated vibrational spectrum, (4) to elu-cidate
the effect of the skeletal Si-Si vibrations on theRaman intensity,
and finally (5) to investigate group fre-quencies and their
intensities involving Si-X (X = H,Cl, OH, CH3, OCH3, and Si(CH3)3)
experimentally andtheoretically, where refers to ((CH3)3Si)3.In our
previous work, we measured IR and Raman vi-
brational spectra of (Me3Si)3SiX, where Me = CH3and X =
H,Cl,OH,CH3,OCH3,Si(CH3)3; then, we as-signed the vibrational peaks
with the aid of the AM1semiempirical method [18]. In the present
work, how-ever, we analyzed the vibrational spectra with the aid
ofab-initio and DFT calculations and clarified whether ornot DFT
and HF calculations would reproduce reason-
-3192-
-
Infrared and Raman Spectroscopic Studies of
Tris(trimethylsilyl)silane Bong Hyun Boo -3193-
ably the IR and the Raman spectra. These methods pro-vide
straightforward assignments of the group frequen-cies corresponding
to the terminal Si-H, Si-C, and Si-Obonds, as well as the Si-Si
central bonds. This studymight help to clarify the presence of
impurities formed inthe syntheses with the aid of only IR and Raman
spec-troscopies. In many cases, the presence of
impuritiescontaining Si-O bonds hampers the straightforward
as-signments of polysilanes.
II. EXPERIMENTAL AND THEORETICALMETHODOLOGIES
The ground-state equilibrium geometries and the en-ergies were
probed by using the Kohn-Sham DFT [19].Beckes three-parameter
exchange functional [20,21] andthe gradient-corrected Lee-Yang-Parr
correlational func-tional (B3LYP) [22, 23] were used with the 6-31G
andthe 6-31G(d) basis sets. For each structure, completesets of
harmonic vibrational frequencies were evaluatedby using the
analytical second-derivative techniques us-ing the HF/6-31G(d) and
the B3LYP/6-31G(d) meth-ods for the geometries optimized with the
B3LYP/6-31Gmethod. The harmonic vibrational frequencies were
alsoused to determine whether a given structure was a localminimum
on the potential energy surface or not. Notethat inner-shell
electrons were excluded in the electron-correlation calculation.
All the ab-initio and DFT cal-culations were carried out with the
Gaussian 03 packagesuite [24].
III. RESULTS AND DISCUSSION
The simulated IR and Raman spectra mimic the corre-sponding
experimental spectra. A one-to-one correspon-dence was observed
between the theories and the exper-iments. This enabled
straightforward assignments of thevibrational fundamentals of the
tris(trimethylsilyl)silanederivatives. Surprisingly, all the IR and
Raman lineswere assigned unequivocally via these relatively
simpleDFT and HF calculations. Fortunately, all the funda-mentals
were observed without any overlap of neighbor-ing
fundamentals.Figures 1 and 2 present the IR and the Raman spec-
tra of SiH, respectively, together with their simulatedspectra.
We clearly observe the Si-H stretching funda-mental at 2052 and
2048 cm1 in the IR and the Ramanspectra, respectively. The
intensities of the Si-H fun-damental vibration are relatively
strong compared withthose of the C-H stretching fundamentals. This
spectro-scopical phenomenon may arise from the fact that
theelectronegativity of Si is lower than that of C, resultingin a
change of the bond polarity. The silicon atom is
Fig. 1. Infrared spectra of (Me3Si)3SiH: (a) experimentand (b)
B3LYP/6-31G(d)//HF/6-31G.
Fig. 2. Raman spectra of (Me3Si)3SiH: (a) experimentand (b)
HF/6-31G(d)//HF/6-31G.
more electron-rich than the C atom, ending up with in-creased
polarization. Contrary to expectation, the skele-tal vibrations
involving Si-Si are found to have relativelyweak intensities in the
IR and the Raman spectra.
-
-3194- Journal of the Korean Physical Society, Vol. 59, No. 5,
November 2011
Fig. 3. Infrared spectra of (Me3Si)3SiCl: (a) experimentand (b)
B3LYP/6-31G(d)//HF/6-31G.
Fig. 4. Raman spectra of (Me3Si)3SiCl: (a) experimentand (b)
HF/6-31G(d)//HF/6-31G.
Figures 3 and 4 depict the IR and the Raman spec-tra of SiCl,
respectively, together with their simulatedspectra. As expected,
the fundamental vibrations in-volving the main framework are quite
similar to thoseof SiH. The Si-Cl stretching vibration is not
observedin the IR and the Raman spectra because of limitations
Fig. 5. Infrared spectra of (Me3Si)3SiOH: (a) experimentand (b)
B3LYP/6-31G(d)//HF/6-31G.
Fig. 6. Raman spectra of (Me3Si)3SiOH: (a) experimentand (b)
HF/6-31G(d)//HF/6-31G.
of both spectrometers. However, the fundamentals
aretheoretically evaluated. The scaled fundamentals for theSi-Cl
stretching are 368 and 355 cm1 in the IR and theRaman spectra,
respectively.The IR and the Raman spectra of SiOH are pre-
sented in Figs. 5 and 6, respectively, together with their
-
Infrared and Raman Spectroscopic Studies of
Tris(trimethylsilyl)silane Bong Hyun Boo -3195-
Fig. 7. Infrared spectra of (Me3Si)3SiMe: (a) experimentand (b)
B3LYP/6-31G(d)//HF/6-31G.
Fig. 8. Raman spectra of (Me3Si)3SiMe: (a) experimentand (b)
HF/6-31G(d)//HF/6-31G.
simulated spectra. This compound is very interestingbecause the
molecule shows both free and H-bonded O-H vibrations. We presume
that the spectroscopic phe-nomenon arises from the intermolecular
properties thatthe molecule exists in two forms, as a free
molecules and
Fig. 9. Infrared spectra of (Me3Si)3SiOMe: (a) experimentand (b)
B3LYP/6-31G(d)//HF/6-31G.
Fig. 10. Raman spectra of (Me3Si)3SiOMe: (a) experi-ment and (b)
HF/6-31G(d)//HF/6-31G.
a dimer formed via intermolecular hydrogen bonding.Figures 7 and
8 present the IR and the Raman spectra
of SiMe, respectively, together with their simulatedspectra. In
this molecule, two types of Si-Me bondingsexist. However, the
molecule does not exhibit any char-
-
-3196- Journal of the Korean Physical Society, Vol. 59, No. 5,
November 2011
Fig. 11. Infrared spectra of (Me3Si)4Si: (a) experimentand (b)
B3LYP/6-31G(d)//HF/6-31G.
acteristic fundamental vibrational frequencies.The IR and Raman
spectra of SiOMe are presented
in Figs. 9 and 10, respectively, together with their sim-ulated
spectra. As expected, the C-H stretching funda-mental corresponding
to the -OMe group is clearly ob-served at 2892 cm1. Why do we
observe the strongabsorption of the C-H stretching in the methoxy
group?We assume that the oxygen atom donates an electronto the
methoxy carbon atom, resulting in an enhance-ment of the C-H bond
polarity. Pure Si-O stretching at611 cm1 in the IR spectrum is
clearly observed. The in-tense absorption is reflected in the
electronegativity dif-ference between the corresponding two atoms.
Also, weobserve an intense Raman line corresponding to the Si-O
vibration. The intense scattering may be reflected inthe enhanced
polarizability involving the electron cloudof the silicon
atom.Figures 11 and 12 present the IR and the Raman spec-
tra of Si(SiMe3)4, respectively, together with their simu-lated
spectra. The optimized molecular geometry quiteresembles a Td
structure. Therefore, only T2 funda-mental vibrations are expected
to be observed. How-ever, almost all the fundamentals are observed
in the IRand the Raman spectra. This presumably indicates thatthe
Si(SiMe3)4 molecular structure deviates from a Tdframework owing to
the steric hindrance.
Fig. 12. Raman spectra of (Me3Si)4Si: (a) experiment and(b)
HF/6-31G(d)//HF/6-31G.
IV. CONCLUSIONS
The simulated IR and Raman spectra mimic the corre-sponding
experimental spectra. A one-to-one correspon-dence is observed
between the theories and the exper-iments. This enabled unequivocal
and straightforwardassignments. We clearly observe the Si-H
stretching fun-damental at 2052 and 2048 cm1 in the IR and the
Ra-man spectra of SiOH, respectively. The intensities forthe
vibrations are relatively intense compared with thoseof the C-H
stretching fundamental. The skeletal vibra-tions involving Si-Si
are found to have relatively weakintensities in the IR and the
Raman spectra. The scaledfundamentals for Si-Cl stretching are 368
and 355 cm1in the IR and the Raman spectra SiCl, respectively.Pure
Si-O stretching at 611 cm1 is clearly observed tohave the intense
absorption both in the IR and the Ra-man spectra of SiOMe. The
strong IR intensity isdue to the electronegativity difference
between the cor-responding two atoms. The intense Raman line may
bedue to the enhanced polarizability involving the electroncloud of
the silicon atom. The optimized molecular struc-ture of Si(SiMe3)4
resembles the Td structure. However,both the almost all the
fundamentals are observed in theIR and Raman spectra. This
presumably indicates thatthe SiSiMe3 molecular structure deviates
from a Tdgeometry.
-
Infrared and Raman Spectroscopic Studies of
Tris(trimethylsilyl)silane Bong Hyun Boo -3197-
Table 1. IR and Raman frequencies of
tris(trimethylsilyl)silane.
Approximate type IR (Rel. int., %) Raman (Rel. int.,%)
Literatureb
of modea
Experiment B3LYP/ Experiment HF/ IR (Rel. int.) Raman (Rel.
int.)
6-31G(d) 6-31G(d)
//HF/6-31G //HF/6-31G
(Scaled with two (Scaled with a uniform
factorsb) factor of 0.900)
Antisym C-H str 2955 (34) 2953 (17) 2948 (34) 2946 (100) 2956
(s) 2951 (vs)
Sym C-H str 2893 (17) 2879 (9.9) 2892 (100) 2884 (97) 2900 (m)
2897 (vs)
Si-H str 2052 (20) 2086 (14) 2048 (13) 1976 (9.8) 2051 (s) 2049
(vs)
C-H bending 1439 (9.1) 1456 (3.3) 1434 (1.3) 1440 (12) 1443 (m)
1443 (vw)
1404 (12) - 1404 (2.6) - 1400 (w) 1402 (m)
Sym C-H bending 1250 (38) 1253 (11) 1258 (2.7) - 1245 (vs) 1254
(w)
- 1242 (3.0) - - 1239 (m)
Antisym C-H bending 841 (100) 831 (100) 868 (2.6) 844 (1.0) 838
(vs) 865 (w)
838 (2.3) - - 833 (m)
Si-H bending, 752 (22) 729 (8.8) 744 (2.6) - 745 (m) 743 (m)
Antisym C-H bending
Antisym Si-C str, 691 (23) 633 (12) 686 (6.7) - 686 (s) 685
(s)
Si-H bending
Sym Si-C str, 613 (34) 566 (6.4) 624 (12) 626 (5.2) 636 (w) 623
(vs)
Si-H bending
Antisym Si-Si str 450 (6.5) - 444 (2.2) 409 (1.1) 445 (s) 444
(m)
aStr refers to stretching. bTwo scaling factors of 0.978 and
0.917 were applied to the low (below 2300 cm1) and the high(above
2300 cm1) energy fundamental frequencies, respectively.
Table 2. IR and Raman frequencies of
tris(trimethylsilyl)chlorosilane.
Approximate type IR (Rel. int., %) Raman (Rel. int.,%)
of modea
Experiment B3LYP/ Experiment HF/
6-31G(d) 6-31G(d)
//HF/6-31G //HF/6-31G
(Scaled with two (Scaled with a uniform
scaled factorsb) factor of 0.900)
Antisym C-H str 2955 (61) 2954 (11) 2946 (36) 2946 (94)
Sym C-H str 2897 (43) 2883 (6.2) 2886 (100) 2887 (100)
Antisym 1400 (34) 1453 (3.6) 1396 (3.6) 1444 (14)
Sym Si-C-H bending 1250 (83) 1255 (11) 1232 (5.4) -
Antisym Si-C-H bending 837 (100) 830 (100) 826 (3.6) 846
(1.3)
Antisym 745 (37) 707 (1.7) 734 (5.5) -
Anitsym Si-C str, 691 (43) 636 (3.1) 678 (15) 634 (5.2)
antisym Si-C-H bending
Sym Si-C str 625 (26) 568 (2.0) 614 (73) 558 (6.6)
Antisym Si-Si str 486 (44) 431 (0.9) 474 (4.4) 435 (1.0)
aStr refers to stretching. bTwo scaling factors of 0.978 and
0.917 were applied to the low (below 2300 cm1) and the high(above
2300 cm1) energy fundamental frequencies, respectively.
-
-3198- Journal of the Korean Physical Society, Vol. 59, No. 5,
November 2011
Table 3. IR and Raman frequencies of
(trimethylsilyl)silanol.
Approximate type of modea IR (Rel. int., %) Raman (Rel.
int.,%)
Experiment B3LYP/6-31G(d) Experiment HF/6-31G(d)
//HF/6-31G //HF/6-31G
(Scaled with two (Scaled with a uniform
factorsb) factor of 0.900)
Pure O-H str 3661 (13) 3729 (5.0) 3664 (8.3) 3706 (11)
H-bonded O-H str 3433 (16, very broad) - - -
Antisym C-H str 2955 (37) 2952 (16) 2936 (49) 2944 (100)
Sym C-H str 2893 (20) 2878 (8.8) 2880 (100) 2884 (98)
Antisym 1400 (13) 1457 (3.2) 1390 (9.0) 1441 (13)
H-C-H bending
Sym Si-C-H bending 1250 (42) 1253 (9.9) 1222 (12) -
Antisym Si-C-H bending 841 (100) 831 (100) 854 (8.5) 842
(1.4)
Antisym 748 (43) 688 (2.8) 727 (10) -
Si-C-H bending
Antisym Si-C str, 691 (24) 633 (3.3) 666 (17) 625 (4.1)
antisym Si-C-H bending
O-H bending 625 (15) 516 (16) 602 (39) 555 (4.9)
Antisym Si-Si str, 471 (7.8) 404 (5.6) 440 (7.6) 445 (0.70)
O-H bending
aStr refers to stretching. bTwo scaling factors of 0.978 and
0.917 were applied to the low (below 2300 cm1) and the high(above
2300 cm1) energy fundamental frequencies, respectively.
Table 4. IR and Raman frequencies of
(trimethylsilyl)methylsilane.
Approximate type of modea IR (Rel. int., %) Raman (Rel.
int.,%)
Experiment B3LYP/6-31G(d) Experiment HF/6-31G(d)
//HF/6-31G //HF/6-31G
(Scaled with two (Scaled with a uniform
factorsb) factor of 0.900)
Antisym C-H str 2951 (45) 2953 (22) 2944 (56) 2944 (100)
Sym C-H str 2893 (31) 2878 (13) 2886 (100) 2884 (94)
Antisym 1400 (19) 1456 (4.0) 1400 (6.4) 1440 (13)
H-C-H bending
Sym Si-C-H bending 1250 (43) 1254 (12) 1234 (7.6) -
Antisym Si-C-H bending 837 (100) 833 (100) 832 (6.3) -
Antisym 783 (41) 753 (21) 736 (7.4) -
Si-C-H bending
Antisym 741 (21) 690 (3.1) - -
Si-C-H bending
Anitsym Si-C str, 6914 (30) 633 (5.2) 678 (21) 626 (4.3)
antisym Si-C-H bending
Sym Si-C str 621 (21) 568 (4.2) 612 (31) 556 (3.9)
Antisym Si-Si str - - 444 (4.3) 416 (1.0)
aStr refers to stretching. bTwo scaling factors of 0.978 and
0.917 were applied to the low (below 2300 cm1) and the high(above
2300 cm1) energy fundamental frequencies, respectively.
-
Infrared and Raman Spectroscopic Studies of
Tris(trimethylsilyl)silane Bong Hyun Boo -3199-
Table 5. IR and Raman frequencies of
(trimethylsilyl)methoxysilane.
Approximate type of modea IR (Rel. int., %) Raman (Rel.
int.,%)
Experiment B3LYP/6-31G(d) Experiment HF/6-31G(d)
//HF/6-31G //HF/6-31G
(Scaled with two (Scaled with a uniform
factorsb) factor of 0.900)
Antisym C-H str in 2955 (45) 2952 (16) 2942 (49) 2946 (86)
the silyl groups
Antisym C-H str in 2897 (29) 2927 (6.8) 2884 (100) 2885
(100)
the -OMe group
Sym C-H str in 2820 (18) 2878 (11) 2812 (14) -
the silyl groups
Antisym 1400 (14) 1455 (3.0) 1398 (7.6) 1440 (12)
H-C-H bending
Sym Si-C-H bending 1250 (40) 1253 (10) 1234 (8.8) -
Antisym O-C-H bending 1169 (11) 1144 (3.9) - -
C-O-Si str 1080 (53) 1076 (28) 1072 (5.1) 1029 (0.84)
Antisym Si-C-H bending 841 (100) 831 (100) 824 (6.3) 843
(1.2)
Antisym Si-C str 702 (35) 635 (4.6) 674 (17) 628 (4.2)
Pure Si-O str 625 (14) 600 (4.6) 608 (32) 579 (2.7)
Antisym Si-Si str 475 (8.6) 435 (2.1) 452 (4.0) -
aStr refers to stretching. bTwo scaling factors of 0.978 and
0.917 were applied to the low (below 2300 cm1) and the high(above
2300 cm1) energy fundamental frequencies, respectively
Table 6. IR and Raman frequencies of
tetrakis(trimethylsilyl)silane.
Approximate type of modea IR (Rel. int., %) Raman (Rel.
int.,%)
Experiment B3LYP/6-31G(d) Experiment HF/6-31G(d)
//HF/6-31G //HF/6-31G
(Scaled with two (Scaled with a uniform
factorsb) factor of 0.900)
Antisym C-H str 2955 (38) 2953 (16) 2942 (66) 2945 (100)
Sym C-H str 2893 (28) 2879 (9.7) 2886 (100) 2885 (99)
Antisym 1400 (16) 1458 (2.2) 1394 (13) 1444 (9.8)
H-C-H bending
Sym Si-C-H bending 1246 (36) 1252 (7.9) 1232 (16) -
Antisym Si-C-H bending 833 (100) 830 (100) 826 (14) 844
(1.2)
Antisym 750 (13) 715 (0.81) 734 (18) 745 (0.67)
Si-C-H bending
Anitsym Si-C str, 687 (24) 629 (3.5) 674 (34) 625 (4.9)
antisym Si-C-H bending
Sym Si-C str 621 (20) 571 (3.2) 616 (35) 561 (4.6)
Antisym Si-Si str 460 (3.8) 383 (0.0) 442 (13) 414 (1.1)
aStr refers to stretching. bTwo scaling factors of 0.978 and
0.917 were applied to the low (below 2300 cm1) and the high(above
2300 cm1) energy fundamental frequencies, respectively.
-
-3200- Journal of the Korean Physical Society, Vol. 59, No. 5,
November 2011
ACKNOWLEDGMENTS
This study was financially supported by the researchfund (for
the Development of Distinguished Scientists ofRegional Universities
in Republic of Korea) of NationalResearch Foundation in 2011 (Grant
Number: 2011-0011564), which is gratefully acknowledged.
REFERENCES
[1] C. X. Cui and M. Kertesz, Macromolecules 25, 1103(1992), and
references therein.
[2] S. S. Bukalov, L. A. Leites, G. I. Magdanurov and R.West, J.
Organomet. Chem. 685, 51 (2003), and refer-ences therein.
[3] H. Kuzmany, J. F. Rabolt, B. L. Farmer and R. D. Miller,J.
Chem. Phys. 85, 7413 (1986).
[4] J. F. Rabolt, D. Hofer, R. D. Miller and G. N.
Fickes,Macromolecules 19, 611 (1986).
[5] L. A. Leites, S. S. Bukalov, T. S. Yadritzeva, M. K.Mokhov,
B. A. Antipova, T. M. Frunze and V. V. De-mentev, Macromolecules
25, 2991 (1992).
[6] P. J. Stephens, F. J. Devlin, C. F. Chabalowski and M.J.
Frisch, J. Phys. Chem. 98, 11623 (1994).
[7] F. J. Devlin, J. W. Finley, P. J. Stephens and M. J.Frisch,
J. Phys. Chem. 99, 16883 (1995).
[8] N. C. Handy, P. E. Maslen, R. D. Amos, J. S. Andrews,C. W.
Murray and G. J. Laming, Chem. Phys. Lett. 197,506 (1992).
[9] N. C. Handy, C. W. Murray and R. D. Amos, J. Phys.Chem. 97,
4392 (1993).
[10] A. A. El-Azhary and H. U. Suter, J. Phys. Chem. 99,12751
(1995).
[11] S. Y. Lee and B. H. Boo, J. Phys. Chem. 100,
15073(1996).
[12] C. J. M. Wheeless, X. Zhou and R. Liu, J. Phys. Chem.99,
12488 (1995).
[13] B. G. Johnson, P. M. W. Gill and J. A. Pople, J. Chem.Phys.
98, 5612 (1993).
[14] S. Y. Lee and B. H. Boo, J. Phys. Chem. 100,
8782(1996).
[15] S. R. Langhoff, J. Phys. Chem. 100, 2819 (1996).[16] G.
Rauhut and P. Pulay, J. Phys. Chem. 99, 3093 (1995).[17] B. H. Boo,
S. Y. Lee and H.-K. Na, J. Phys. Chem. 102,
2679 (1998).[18] B. H. Boo, H. K. Kang, S. K. Kang, S. S. Lee,
D. Kim
and M. H. Lee, J. Organomet. Chem. 436, 1 (1992).[19] W. Kohn
and L. J. Sham, Phys. Rev. A 140, 1133 (1965).[20] A. D. Becke,
Phys. Rev. A 38, 3098 (1988).[21] A. D. Becke, J. Chem. Phys. 98,
5648 (1993).[22] C. Lee, W. Yang and R. G. Parr, Phys. Rev. B 37,
785
(1988).[23] B. Miehlich, A. Savin, H. Stoll and H. Preuss,
Chem.
Phys. Lett. 157, 200 (1989).[24] M. J. Frisch et al., Gaussian
03, Revision D. 01 (Gaus-
sian, Inc., Wallingford, CT, 2004).