The Fourier transform infrared spectra of the complexes of boron trifluoride with dimethyl ether and dimethyl sulphide in cryogenic matrices

Article abstract of DOI:10.1016/S0022-2860(03)00342-9

The infrared spectra of boron trifluoride, co-condensed with dimethyl ether and with dimethyl sulphide in nitrogen and argon matrices at ca. 17 K, have been recorded. The spectra are consistent with the formation of 1:1 electron donor-acceptor molecular c

Full text of DOI:10.1016/S0022-2860(03)00342-9

Journal of Molecular Structure 656 (2003) 303–319
www.elsevier.com/locate/molstruc
The Fourier transform infrared spectra of the complexes
of boron trifluoride with dimethyl ether and dimethyl sulphide
in cryogenic matrices
a
Lawrence M. Nxumalo , Thomas A. Ford
b,*
a
Department of Chemistry, Vista University, Soweto Campus, Private Bag X09, Bertsham 2013, South Africa
b
School of Pure and Applied Chemistry, University of Natal, King George V Avenue, Durban 4041, South Africa
Received 9 January 2003; accepted 31 January 2003
In honour of Prof. Achim M u¨ ller
Abstract
The infrared spectra of boron trifluoride, co-condensed with dimethyl ether and with dimethyl sulphide in nitrogen and argon
matrices at ca. 17 K, have been recorded. The spectra are consistent with the formation of 1:1 electron donor–acceptor
molecular complexes, with donation occurring from the orbitals of the ether or the sulphide dominated by the oxygen or sulphur
lone pairs to the orbital of boron trifluoride correlating with the empty axial 2p atomic orbital of boron. The wavenumber shifts
of the respective monomer absorption bands have been determined in each matrix and correlated with the physical properties of
the bases. The spectra may be rationalized by referring to the results of some molecular orbital calculations on these systems.
q 2003 Elsevier B.V. All rights reserved.
Keywords: FTIR spectroscopy; Matrix isolation; Electron donor–acceptor complexes; Boron trifluoride
1
. Introduction
of the infrared spectra, obtained under similar
conditions, of the complexes of boron trifluoride
with dimethyl ether (DME) [3,9] and dimethyl
sulphide (DMS) [9]. These experimental spectro-
scopic studies have been supported by a parallel series
of theoretical studies, at the ab initio level, of the
dimer [1,10,11] and of the complexes of boron
trifluoride with nitrogen [12,13], carbon monoxide
In several recent publications we have described
our studies of the infrared spectra, in cryogenic
matrices, of the dimer of the Lewis acid boron
trifluoride [1,2] and of a variety of 1:1 molecular
complexes of boron trifluoride with some common
Lewis bases [3–8]. These bases have included carbon
monoxide [3,4], water [5], carbon dioxide [6], nitrous
oxide [7], sulphur dioxide [3] and ammonia [3,8]. In
addition we have published two preliminary reports
[4], hydrogen cyanide and cyanogen fluoride [12,13],
water [5], carbon dioxide [14], nitrous oxide [15],
sulphur dioxide [16], ammonia [8], nitrogen trifluor-
ide [17], cyanogen and cyanoacetylene [12,13].
The BF ·O(CH ) complex is relatively well-
3
3 2
*
Corresponding author. Tel.: þ27-31-260-1096; fax: þ27-31-
60-3091.
E-mail address: ford@nu.ac.za (T.A. Ford).
known, and has been the subject of electron diffraction
[18–20], gas, liquid and solid phase infrared and
2
0022-2860/03/$ - see front matter q 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0022-2860(03)00342-9

304
L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
Raman [21–23] and matrix isolation infrared [3,9,24]
studies. It has also been examined by means of ab
initio molecular orbital theory [25–28]. Complexes
with sulphur bases have not enjoyed such extensive
study; the BF ·H S and BF ·CS adducts have been
Sample annealing was carried out routinely to
35 K. Nitrogen (99.998%), argon (99.9995%), boron
trifluoride (99.57%, natural isotopic abundance),
DME (99.0000%) and DMS were obtained from Air
Products S.A. (Pty.) Ltd, Johannesburg, and were used
without further purification. Other details may be
found in our previous publications [1–9].
3
2
3
2
studied using UV photoelectron and electron energy
loss spectroscopy [29–31], and ab initio computations
have been carried out on BF ·H S [30–32] and
3
2
BF ·CS [30,31]. The complex of BF with DMS has
3
3
2
been investigated by manometric and thermodynamic
techniques [33], and by matrix isolation infrared
spectroscopy [9].
3. Results and discussion
3.1. The BF , (CH ) O and (CH ) S monomers
3
3 2
3 2
In this paper we wish to discuss the spectra of the
analogous pair of complexes BF ·O(CH3)2 and
1
0
11
3
The infrared spectra of the BF3 and BF3
isotopic variants of the BF ·O(CH ) and
BF ·S(CH ) , which were described briefly in our
3 3 2
3
3 2
two earlier reports [3,9]. The BF ·S(CH ) adduct is
3 2
BF ·S(CH ) complexes were computed at the
3 3 2
3
expected to be considerably more weakly bound than
the BF ·O(CH ) species, since BF is regarded as a
restricted Hartree–Fock (RHF) level of theory,
using the 6-31G** basis set [28]. The computed
wavenumbers and infrared band intensities were
presented, along with appropriate assignments of the
calculated wavenumbers to the various complex
normal modes [28]. We are aware of the limitations
of RHF wavefunctions, particularly for the compu-
tation of the vibrational spectra of molecular com-
plexes, where the dispersion contribution can
represent a substantial proportion of the total
interaction energy, and electron correlation is necess-
ary to account for this. Density functional theory
(DFT) [35], like RHF, also fails to account for
electron correlation, but has had some success in
reproducing the properties of a variety of systems with
reasonable accuracy [36]. We compared the exper-
3
3 2
3
hard acid, and hard bases like DME are known to
interact more strongly with less polarizable Lewis
acids than do soft bases like DMS [34]. We also wish
to compare the experimental spectra with those
predicted in our earlier ab initio (RHF/6-31G**)
molecular orbital study [28], in which we concluded
that the relative changes in the monomer geometrical
properties and the complex-monomer wavenumber
shifts on bonding were governed largely by the
differences in the strengths of interaction of the acid
(
BF ) with a hard and with a soft base.
3
2
. Experimental details
imental wavenumbers of the BF , DME and DMS
3
The spectra were recorded using a Bruker IFS 88
Fourier transform infrared interferometer, in the
monomers determined here in nitrogen and argon
matrices with our RHF/6-31G** calculated values
[28], and found mean computed/experimental ratios
2
wavenumber range from 4000 to 400 cm . The
1
2
resolution was set routinely at 1 cm , but for some
1
of 1.088 (BF ), 1.108 (DME) and 1.109 (DMS). In an
3
2
1
spectra it was increased to 0.5 cm . The cryostat used
was an Air Products Displex CS202A closed-cycle
helium refrigerator, coupled with an Air Products
model APD-B temperature controller. Gas samples
were prepared using standard manometric techniques,
on an all-glass vacuum line. Matrix/absorber (M/A)
ratios were typically in the range 1000/1–50/1, with
nitrogen or argon as the matrix gas. Deposition rates
effort to improve on the level of agreement between
the observed and computed wavenumbers we
have now repeated those calculations using DFT
[35] with the B3LYP functional [37–39], and with the
same 6-31G** basis set.
1
0
11
The wavenumbers of the BF and BF mono-
3
3
mers observed in nitrogen and argon matrices are
listed in Table 1. This table also includes the
calculated/experimental ratios, and these ratios for
both isotopic species have mean values of 1.026
(nitrogen) and 1.017 (argon). The corresponding
2
1
were usually about 5 mmol h , and samples were
deposited at 17 K; samples for the co-condensation
experiments were deposited in parallel gas streams.

L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
305
Table 1
Observed and calculated wavenumbers of the BF
3
monomers, and calculated/experimental wavenumber ratios
2
1
a
b
Isotopic
species
Symmetry
species
Mode
Approximate
description
Wavenumber (cm
)
Ratio
Ratio
a
b
Calc.
Expt
Expt
10
BF
3
0
c
c
a
a
e
1
0
n
n
n
1
2
3
n
d
n
s
(BF
(BF
3
)
)
889.4
712.8
–
–
–
–
0
2
s
3
684.0
702.9
1.042
1.032
1.035
1.004
1.014
1.034
0
d
a
(BF
3
)
1544.4
1495.8
1493.5
d
1
492.0
n
4
d
a
(BF
3
)
481.0
479.2
479.1
1.004
1
1
0
c
c
BF
3
a
a
e
1
0
n
n
n
1
2
3
n
d
n
s
(BF
(BF
3
)
)
889.4
684.8
–
–
–
–
0
2
s
3
662.3
673.5
1.034
1.030
1.034
1.000
1.017
1.033
0
d
a
(BF
3
)
1489.5
1445.5
1442.3
d
1
440.9
n
4
d
a
(BF
3
)
479.1
479.2
479.1
1.000
1.017
Mean ratio
1.026
a
In nitrogen.
b
c
d
In argon.
Infrared inactive.
Matrix splitting.
results for the (CH ) O and (CH ) S monomers are
3 2
warm-up behaviour and, in the case of the modes
mainly localized in the BF fragment, the similarity of
3
2
presented in Tables 2 and 3, as checks on the
consistency of the monomer assignments. The mean
ratios are 1.031 (nitrogen and argon) for DME and
3
1
the complex BF / BF isotopic shifts with those of
0
11
3
3
the BF monomer [1,2]. These shifts are approxi-
3
2
1
1
.037 (nitrogen) and 1.038 (argon) for DMS. The
mately 0, 20, 50 and 0 cm for n ; n ; n and n ;
1 2 3 4
wavenumber ratios given in Tables 1–3 represent
favourable improvements on those obtained at the
RHF/6-31G** level [28], and give us confidence in
their use in our interpretation of the spectra of the
complexes. The calculated B3LYP/6-31G** intensi-
ties are in qualitative agreement with the RHF/6-
respectively [6,7,40]. Table 6 contains the computed
and experimental complex-monomer wavenumber
shifts. In order to avoid excessive repetition, only
1
1
those experimental shifts for the BF3 isotopic
variant, in nitrogen, are included.
21
3
1G** values, particularly in respect of the sequence
3.2.1. The CH stretching region, 3100–2800 cm
3
of the magnitudes of the individual modes, and are not
reproduced here.
The spectrum of pure (CH ) O suffers virtually no
3 2
changes as a result of annealing. The complex
spectrum in the CH stretching region, surprisingly,
3
3
.2. The BF ·O(CH ) complex
3 3 2
showed no additional bands, even though the shifts
were calculated to be quite large, and substantial shifts
Table 4 shows the computed B3LYP/6-31G**
were observed on going from (CH O in CCl
solution to liquid and solid BF ·O(CH ) [23].
3 2
3
)
2
4
1
0
11
wavenumbers of BF ·O(CH ) and BF ·O(CH ) ,
3
3 2
3
3 2
3
along with our proposed assignments. The wavenum-
bers are in general agreement with those in the liquid
state [21] and in argon matrices [24] reported in the
literature. We will discuss the assignments by
spectroscopic region.
2
1
3.2.2. The CH bending region, 1500–1400 cm
3
In addition to the matrix split BF monomer n
3
3
bands, new bands appeared in this region at 1454.0,
2
1435.9 and 1424.2 cm in nitrogen, and at 1461.3
1
2
and 1446.1 cm in argon, as shown in Fig. 1. These
1
Table 5 lists the observed wavenumbers of the new
bands assigned to the complex. Assignments were
made on the usual bases of concentration studies,
are assigned to three of the six expected CH bending
3
modes of the complex. Four of these bending bands

306
L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
Table 2
3 2
Observed and calculated wavenumbers of the (CH ) O monomer, and calculated/experimental wavenumber ratios
2
1
a
b
Symmetry
species
Mode
Approximate
description
Wavenumber (cm
Calc.
)
Ratio
Ratio
a
b
Expt
Expt
a
1
n
1
n
2
n
3
n
4
n
5
n
6
n
7
n
n
d
d
a
(CH
(CH
(CH
(CH
3
)
3128.1
2981.6
1534.9
1503.7
1272.4
954.6
2997.6
2822.1
1476.0
2986.1
2821.4
1478.8
1453.1
1.044
1.057
1.040
–
1.048
1.057
1.038
1.035
–
s
3
)
a
3
)
c
s
3
)
–
c
c
r(CH
(COC)
d(COC)
3
)
–
–
–
n
s
925.3
413.7
925.8
415.4
1.032
1.001
1.032
0.997
414.0
d
d
a
2
n
n
n
n
8
n
d
a
(CH
(CH
3
)
)
3020.0
1493.2
1167.6
222.0
–
–
–
–
–
–
–
–
–
–
d
d
9
a
3
–
–
d
d
10
11
r(CH
t(CH
3
)
)
–
–
d
d
3
–
–
b
1
n
12
n
13
n
14
n
15
n
d
a
(CH
(CH
3
)
)
3016.6
1502.6
1203.1
248.3
2922.4
1456.2
1168.6
2924.6
1456.8
1172.7
1.032
1.032
1.029
–
1.032
1.032
1.026
–
a
3
r(CH
t(CH
3
)
)
c
c
3
–
–
b
2
n
16
n
17
n
18
n
19
n
20
n
21
n
n
d
d
a
(CH
(CH
(CH
(CH
3
)
3126.5
2966.5
1511.2
1472.1
1209.6
1130.9
2994.5
2818.2
1460.6
2986.1
2821.4
1460.5
1.044
1.052
1.035
–
1.047
1.051
1.035
–
s
3
)
a
3
)
c
c
s
3
)
–
–
r(CH
3
)
1245.3
1096.8
1248.9
1096.6
0.972
1.031
0.969
1.031
n
a
(COC)
Mean ratio
1.031
1.031
a
In nitrogen.
b
c
d
In argon.
Not observed.
Infrared inactive.
2
isotopic shifts (73.2 and 113.5 cm (nitrogen) and
1
are predicted to be weak, however [28], and may not
have been observed at the matrix concentrations
which we used. We assign the three bands observed in
21
32.4 and 116.1 cm (argon)) are considerably larger
than those found for the antisymmetric BF stretching
3
0
00
nitrogen to n and n (a ) and n (a ) on the basis of
4
vibrations in a number of BF -containing complexes
3
6
24
their relative computed intensities [28], and of their
small predicted red shifts (Table 6).
[1–9]. However, their clear signature isotopic doublet
patterns make these assignments unambiguous. Their
relatively large values indicate a substantial partici-
pation of the boron atom in the normal modes. In
addition, the complex-monomer wavenumber shifts
for these bands are predicted to be very large and
3
rocking regions, 1300–1050 cm
.2.3. The antisymmetric BF stretching and CH
3 3
2
1
1
Distinct BF / BF isotopic doublets were
0
11
3
3
2
1
observed in this region, at 1240.1/1166.9 and
negative (2155.5 and 2180.9 cm ) (see Table 6).
We therefore assign the higher frequency pair of
2
1
1
1
207.8/1094.3 cm
251.6/1219.2 and 1222.7/1106.6 cm
in
nitrogen
and
in argon.
at
2
1
00
bands in each matrix to the n (a ) and the lower to
2
5
0
the n (a ) modes.
7
These are also shown in Fig. 1. The relative intensities
of this group of bands remained constant in a number
of experiments with different M/A/A ratios and they
clearly belong to the same species. Two of the four
Four bands are expected in the CH rocking region,
3
two of medium and two of low intensity [28]. A new
2
1
band at 1247.3 cm (1260.1 cm in argon), was
21

L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
307
Table 3
3 2
Observed and calculated wavenumbers of the (CH ) S monomer, and calculated/experimental wavenumber ratios
2
1
a
b
Symmetry
species
Mode
Approximate
description
Wavenumber (cm
Calc.
)
Ratio
Ratio
a
b
Expt
Expt
a
1
n
1
n
2
n
3
n
4
n
5
n
6
n
7
n
n
d
d
a
(CH
(CH
(CH
(CH
3
)
3151.4
3047.0
1502.6
1385.2
1063.3
685.4
2975.4
2887.9
1448.7
2973.7
2886.9
1447.1
1.059
1.055
1.037
–
1.060
1.055
1.039
–
s
3
)
a
3
)
c
c
s
3
)
–
–
r(CH
(CSC)
d(CSC)
3
)
1033.2
695.6
1030.3
694.5
1.029
0.985
–
1.032
0.986
–
n
s
c
c
258.7
–
–
d
d
a
2
n
n
n
n
8
n
d
a
(CH
(CH
3
)
)
3132.7
1478.5
960.5
–
–
–
–
–
–
–
–
–
–
d
d
9
a
3
–
–
d
d
10
11
r(CH
t(CH
3
)
)
–
–
d
d
3
184.4
–
–
b
1
n
12
n
13
n
14
n
15
n
d
a
(CH
(CH
3
)
)
3126.8
1487.9
997.5
2926.6
1432.7
978.3
2928.2
1432.2
972.3
1.068
1.039
1.019
–
1.068
1.039
1.025
–
a
3
r(CH
t(CH
3
)
)
c
c
3
182.9
–
–
b
2
n
16
n
17
n
18
n
19
n
20
n
21
n
n
d
d
a
(CH
(CH
(CH
(CH
3
)
3152.3
3049.2
1493.9
1359.5
920.7
2994.6
2862.5
1442.1
1313.6
925.4
2997.7
2858.5
1439.8
1310.9
926.8
1.053
1.065
1.036
1.035
0.995
–
1.051
1.067
1.038
1.037
0.994
–
s
3
)
a
3
)
s
3
)
r(CH
3
)
c
c
n
a
(CSC)
737.8
–
–
Mean ratio
1.037
1.038
a
In nitrogen.
b
c
d
In argon.
Not observed.
Infrared inactive.
observed. This band did not show any isotopic
variation, and it is therefore assigned to the strongest
complex-monomer shift, consistent in sign and
magnitude with the predicted displacement (see
Table 6). No other bands in this region could
conceivably be assigned to this mode.
00
CH rocking band of the complex, n (a ).
3
26
2
1
3
.2.4. The COC stretching region, 1050–900 cm
3
.2.5. The symmetric BF stretching region,
3
The symmetric ðn Þ and antisymmetric ðn Þ COC
1
0
28
2
1
8
50–800 cm
The formally infrared inactive symmetric BF3
stretching bands of the DME monomer absorb near
2
25 and 1100 cm , respectively. A pair of bands was
1
9
detected at 923.2 and 919.8 cm
2
1
in nitrogen
950.2 cm in argon) (see Fig. 2), which correspond
stretching vibration of the monomer is found in this
region, activated in the complex as a result of the
distortion of the BF fragment from planarity. This
2
1
(
with the symmetric COC stretching, and are assigned
3
to n ; having a small isotopic shift, indicating some
1
band has been observed to absorb weakly in other
complexes containing BF [4–9]. A weak isotopic
0
minor involvement of the boron atom in the motion.
The antisymmetric COC stretching band appeared at
3
1
2
pair, at 817.1/812.0 cm in nitrogen and 815.1/
21
811.6 cm in argon, is assigned to the complex n11
21
21
031.3 cm in nitrogen and 999.1 cm in argon. It
1
showed no isotopic shift, and a relatively large red
mode, confirmed by the small isotopic shift of about

308
L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
Table 4
Computed wavenumbers of the BF
3 3 2
·O(CH ) complexes
2
Wavenumber (cm )
1
Symmetry
species
Mode
Approximate
description
10
11
BF ·O(CH
3
3
)
2
3 3 2
BF ·O(CH )
0
a
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
1
n
n
n
d
d
d
n
a
(CH
(CH
3
)
)
3207.3
3140.8
3061.9
1524.2
1511.7
1486.7
1356.2
1282.4
1212.8
936.2
3207.3
3140.8
3061.9
1524.1
1511.3
1486.3
1308.6
1281.5
1212.8
932.5
2
a
3
3
s 3
(CH )
4
a
a
(CH
(CH
3
)
)
5
3
6
s
3
(CH )
7
a
3
(BF )
8
r(CH
r(CH
3
)
)
9
3
10
11
12
13
14
15
16
17
18
n
n
d
d
s
s
s
(COC)
(BF
(BF
3
)
)
827.7
824.3
3
623.2
606.3
a
(BF
3
)
495.8
493.6
d(COC)
t(CH
451.2
450.9
3
)
363.0
362.9
n(B· · ·O)
v(COC)
d(FB· · ·O)
253.0
251.6
230.4
230.3
134.5
134.4
0
0
a
n
19
n
20
n
21
n
22
n
23
n
24
n
25
n
26
n
27
n
28
n
29
n
30
n
31
n
32
n
33
n
n
n
d
d
d
n
a
(CH
(CH
3
)
)
3207.2
3141.8
3055.8
1509.0
1491.9
1460.1
1381.6
1184.6
1161.3
1077.0
498.6
3207.2
3141.8
3055.8
1508.9
1491.4
1457.8
1334.0
1184.4
1161.3
1076.9
496.4
a
3
s
3
(CH )
a
a
(CH
(CH
3
)
)
3
s
3
(CH )
a
3
(BF )
r(CH
r(CH
3
)
)
3
n
d
a
(COC)
(BF
a
3
)
t(CH
r(COC)
t(COC)
3
)
321.8
321.8
221.0
221.0
188.8
188.7
v(BF
3
)
52.6
52.6
2
cm (see Fig. 2); the gas phase isotopic shift in the
1
5
Raman spectrum of the monomer is zero [40].
in magnitude, with the predicted complex-monomer
shift (Table 6).
3
.2.6. The symmetric BF bending region,
3
3.2.7. The antisymmetric BF and COC bending
3
2
1
21
7
00–600 cm
An isotopic pair of bands was again detected in this
region. In nitrogen they were observed at
regions, 500–400 cm
The antisymmetric BF bending modes of com-
3
plexes containing BF have been found to be rather
3
2
1
6
appeared at 637.8/624.1 cm in argon. The isotopic
2
shifts, 15.4 and 13.7 cm , respectively, are consist-
52.5/637.1 cm , as shown in Fig. 2, while they
1
insensitive to the nature of the bonding partner [3–9].
We thus expect these bands in BF ·O(CH ) to absorb
2
3
3 2
1
very close to the corresponding modes in the
monomer. Moreover, they are found to be virtually
isotopically invariant [3–9]. We therefore assign the
two features observed in this region, at 480.9 and
ent with their assignment to the symmetric BF
0
3
bending mode, n12 (a ), and this assignment is
confirmed by the agreement in sign, and qualitatively

L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
309
Table 5
Observed infrared bands of the BF
ratios
3 3 2
·O(CH ) complexes isolated in nitrogen and argon matrices, and calculated/experimental wavenumber
a
Complex
Symmetry
species
Mode
Nitrogen
Argon
2
1
)
21
Wavenumber (cm
Ratio
Wavenumber (cm
)
Ratio
1
0
0
BF ·O(CH
3
3
)
2
a
n
n
n
n
n
n
n
n
n
n
n
n
4
1454.0
1435.9
1207.8
923.2
1.048
1.035
1.123
1.014
1.013
0.955
1.046
1.025
1.114
0.950
1.044
1.037
1461.3
1446.1
1222.7
950.2
1.043
1.028
1.109
0.985
1.015
0.977
–
6
7
10
11
12
13
24
25
26
28
29
817.1
815.1
652.5
637.8
b
474.2
–
0
0
0
b
a
a
1424.2
1240.1
1247.3
1031.3
480.9
–
–
1251.6
1260.1
999.1
1.104
0.940
1.078
–
b
–
11
BF ·O(CH
3
3
)
2
n
n
n
n
n
n
n
n
n
n
n
4
1454.0
1435.9
1094.3
919.8
1.048
1.035
1.196
1.014
1.015
0.952
1.041
1.143
0.950
1.044
1.032
1461.3
1446.1
1106.6
1.043
1.028
1.183
–
6
7
b
10
11
12
13
25
26
28
29
–
812.0
811.6
1.016
0.971
–
637.1
624.1
b
–
474.2
0
0
a
1166.9
1247.3
1031.3
480.9
1219.2
1260.1
999.1
1.094
0.940
1.078
–
b
–
a
See Table 4 for assignments.
Not observed.
b
2
1
00
0
21
4
Fig. 2), which yields complex-monomer shifts of 1.7
74.2 cm in nitrogen, to n29 (a ) and n (a ) (see
work), 637.8/624.1 cm
1
(argon, this work), 638/
21
13
2
624 cm (argon [24]) and 672/661 cm
[21), attributed to the B· · ·O stretching mode. Our
(liquid
2
and 25.0 cm for the antisymmetric BF bending
1
3
modes. These modes were not observed in argon. The
COC bending vibration of the DME monomer is
predicted to be shifted significantly to higher
wavenumber on complexation. This vibration could
not be detected, however.
calculations indicate that this mode should be assigned
0
to the a BF
located in the BF
bending vibration, with most of the motion
group straddling the symmetry plane
3
2
(Table 4). Our calculations place the B· · ·O stretching
mode outside the range of our spectrometer.
The CH3 torsional (n15 and n ) and the
30
intermolecular modes (n ; n ; n ; n ; n and
17
16
18
31
32
3.3. The BF ·S(CH ) complex
3 3 2
n ) are predicted to lie outside the available
3
3
wavenumber range of our instrument.
10 11
The computed spectra of the BF and BF3
3
Our assignments are in essential agreement with
those of Begun et al. in the liquid state [21] and Hunt
and Ault in argon [24], allowing for the differences in
the environments. The major point of disagreement
isotopomers of the BF ·S(CH ) complex are listed in
3
3 2
Table 7, together with their assignments.
As was the case for the corresponding DME
monomer, the DMS monomer spectrum was practi-
cally invariant to annealing. The positions of the new
2
1
relates to the bands at 652.5/637.1 cm (nitrogen, this

310
L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
Table 6
Observed and calculated complex-monomer wavenumber shifts for the BF
11
3 3 2
·O(CH ) complex
11
Complex mode
(CH
3
)
2
O monomer
3
BF monomer
2
1
)
21
Wavenumber shift (cm )
Mode
Wavenumber shift (cm
Mode
a
a
Calc.
Expt
Calc.
Expt
0
n
1
n
2
n
3
n
4
n
5
n
6
n
7
(a )
n
n
n
n
n
n
1
(a
1
)
79.2
0
(a )
12(b
1
)
124.2
80.3
0
(a )
2
3
(a
(a
1
)
)
0
(a )
1
210.8
8.7
222.0
0
(a )
13(b
1
)
0
(a )
4
(a
1
)
217.4
0
0
b
b
(a )
n
3
(e )
2180.9
2351.2
2
346.6
0
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
8
(a )
n
n
n
5
(a
1
)
9.1
9.7
0
9
(a )
0
14(b
1
)
10(a )
6
(a
1
)
222.1
22.1
0
0
0
0
0
11(a )
n
1
n
2
n
4
(a
1
)
)
265.1
278.5
14.5
–
00
12(a )
(a
2
225.2
25.0
0
13(a )
(e )
14(a )
n
n
n
n
n
n
n
n
7
(a
1
)
36.9
114.6
80.7
0
15(a )
15(b
16(b
1
)
)
00
19(a )
2
0
0
20(a )
8
(a
2
)
121.8
89.3
00
21(a )
17(b
18(b
2
)
)
0
0
22(a )
2
22.3
21.8
214.3
00
23(a )
9 2
(a )
0
0
24(a )
19(b
2
)
236.4
00
0
b
b
25(a )
n
3
(e )
2155.5
2278.6
2
270.0
0
0
n
n
n
n
n
26(a )
n
n
n
20(b
2
)
225.2
26.3
254.0
2.0
0
0
27(a )
2
10(a )
0
0
28(a )
21(b
2
)
265.5
0
0
0
0
29(a )
n
4
(e )
17.3
1.7
0
30(a )
a
n11(a
2
)
99.8
In nitrogen.
b
Matrix splitting.
bands observed for the BF ·S(CH3)2 complex are
3
3.3.2. The CH bending and the antisymmetric BF
3 3
2
1
given in Table 8.
stretching regions, 1500–1350 cm
Product bands were located in this region at
1498.7, 1478.2, 1459.1, 1424.1, 1410.1, 1393.2 and
2
1
3
.3.1. The CH stretching region, 3100–2800 cm
3
2
With the exception of a band at 2918.9 cm in
1
21
1363.8 cm
,
with
2
a
matrix split pair at
1380.4/1378.4 cm in nitrogen, and with counter-
1
nitrogen, the CH3 stretching region showed no
additional bands resulting from complexation. Evi-
dently the shifts of these modes on formation of the
complex are much smaller than predicted. It is not
21
parts at 1498.3, 1443.1, 1395.1 and 1363.9 cm in
argon (see Fig. 3). Table 8 indicates that we have
2
1
assigned the 1498.7, 1478.2 and 1363.8 cm bands,
which are invariant to isotopic substitution and
virtually independent of the matrix, to the n ; n and
clear to which of the six CH stretching modes of the
3
complex to assign this band.
4
5

L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
311
2
1
Fig. 1. The infrared spectra (1600–1000 cm ) of boron trifluoride and dimethyl ether in nitrogen (M/A/A ¼ 250/1/1), deposited at 17 K, (a)
before, (b) after annealing at 35 K.
2
1
n24 CH bending vibrations. The remaining absorp-
3
these modes are 29.7/31.7 cm
(nitrogen) and
2
1
21
31.2 cm (argon) for n and 60.3 cm (nitrogen)
tions in this region, the isotopic pairs at 1410.1 and
6
2
1
1
1
380.4/1378.4 cm
in nitrogen (1395.1 and
21
in argon) and at 1424.1/1363.8 cm
for n22=n24:
21
363.9 cm
2
1
in nitrogen, which has no counterpart in argon, to the
3.3.3. The CH rocking region, 1100–1000 cm
3
1
n6 and n22 ( BF ) or n ( BF ) vibrations, which
0
11
21
3
24
3
Three bands, at 1096.8, 1028.0 and 925.0 cm
2
(nitrogen) and 1040.1 and 1034.1 cm (argon) are
located in the CH3 rocking region, as shown in
1
correlate with the antisymmetric BF stretching mode
3
1
0
11
of the monomer. The BF / BF isotopic shifts of
3
3

312
L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
2
1
Fig. 2. The infrared spectra (1000–450 cm ) of boron trifluoride and dimethyl ether in nitrogen (M/A/A ¼ 250/1/1), deposited at 17 K, (a)
before, (b) after annealing at 35 K.
Fig. 4. Since they are not expected to be affected
1
3.3.4. The symmetric BF stretching region,
3
0
11
21
by
assigned to n ; n and n of both isotopic variants
BF / BF isotopic substitution, they are
900–800 cm
3
3
2
A weak, broad band, at 853.5 cm in nitrogen and
1
8
9
27
2
1
of the complex, on the basis that these three modes
are calculated to have the highest intensities of the
four CH rocking vibrations found in BF ·S(CH )
3 2
at 837.7 cm in argon, appears in the spectra of some
of the more concentrated matrices in the region of
the inactive symmetric BF3 stretching mode of
the monomer (see Fig. 5). It is expected to be
3
3
[
28].

L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
313
Table 7
Computed wavenumbers of the BF
3
·S(CH
·S(CH
Approximate description
3
)
2
complexes
10
BF
3
11
BF
Symmetry
species
Mode
3
)
2
3 3 2
·S(CH )
a
21
a
21
Wavenumber (cm )
Wavenumber (cm
)
Approximate description
0
a
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
1
n
n
n
d
d
n
d
a
(CH
(CH
3
)
)
3166.5
3157.0
3064.1
1501.5
1488.5
1482.5
1387.1
1067.6
1006.7
850.0
683.8
584.2
475.8
263.2
183.4
165.9
125.6
58.5
n
n
n
d
d
n
d
a
(CH
(CH
3
)
)
3166.5
3157.0
3064.0
1495.7
1488.5
1435.8
1386.3
1067.6
1006.7
849.3
683.8
564.7
473.9
263.2
183.3
165.8
125.5
58.0
2
a
3
a
3
3
s
(CH
3
)
s 3
(CH )
4
a
a
a
(CH
(CH
3
)
)
a
a
a
(CH
(CH
3
)
)
5
3
3
6
(BF
(CH
3
)
3
(BF )
7
s
3
)
s
(CH
3
)
8
r(CH
r(CH
3
3
)
)
r(CH
r(CH
3
3
)
)
9
10
11
12
13
14
15
16
17
18
n
n
d
d
s
s
s
(BF
(CSC)
(BF
(BF
3
)
n
n
d
d
s
s
s
(BF
(CSC)
(BF
(BF
3
)
3
)
3
)
a
3
)
a
3
)
d(CSC)
t(CH
d(CSC)
t(CH
3
)
3
)
n(B· · ·S)
v(CSC)
d(FB· · ·S)
n(B· · ·S)
v(CSC)
d(FB· · ·S)
0
0
a
n
19
n
20
n
21
n
22
n
23
n
24
n
25
n
26
n
27
n
28
n
29
n
30
n
31
n
32
n
33
n
n
n
n
d
d
d
a
(CH
(CH
3
)
)
3167.7
3160.7
3066.2
1490.3
1479.9
1455.0
1362.0
973.8
928.1
738.0
475.0
179.0
172.3
103.9
4.1
n
n
n
d
d
n
d
a
(CH
(CH
3
)
)
3167.7
3160.7
3066.2
1486.6
1479.1
1408.8
1360.6
973.8
928.1
738.0
473.1
179.0
172.2
103.9
4.1
a
3
a
3
s
(CH
3
)
s 3
(CH )
a
a
a
(BF
(CH
(CH
3
)
a
a
a
(CH
(CH
3
)
)
3
3
)
)
3
3
(BF )
s
(CH
3
)
s
(CH
3
)
r(CH
r(CH
3
)
)
r(CH
r(CH
3
)
)
3
3
n
d
a
(CSC)
(BF
n
d
a
(CSC)
(BF
a
3
)
a
3
)
t(CH
r(CSC)
t(CSC)
3
)
3
t(CH )
r(CSC)
t(CSC)
v(BF
3
)
v(BF
3
)
a
11
Note that the assignments of n22 and n24 of BF
10
are transposed relative to those of BF
3
·S(CH
3
)
2
3
·S(CH
3
)
2
.
1
0
11
insensitive to BF / BF substitution, although in
either in nitrogen or in argon; it is expected to
absorb only very weakly [28]. No new bands
attributable to n ðCSCÞ or n ðCSCÞ of BF ·S(CH )
3
3
argon a slight, poorly resolved splitting is observed.
This is assigned to n : (In Table 2 of Ref. [9] the
position of this band in argon is erroneously reported
2
at 873.7 cm .)
1
0
s
a
3
3 2
could be detected.
1
However, an intense pair of isotopic bands, at
2
1
21
690.5/670.2 cm
(argon) was observed in the region of the out-of-plane
(nitrogen) and 688.2/668.2 cm
3.3.5. The CSC stretching and the symmetric BF
2
bending regions, 800–650 cm
3
1
BF monomer bending absorption (see Figs. 4 and 5).
3
In the DMS monomer, the symmetric CSC
2
This feature was assigned to the n12 mode of the
21
complex, having isotopic shifts of 20.3 cm (nitro-
1
stretching vibration appears at about 695 cm
The antisymmetric counterpart was not observed,
.
2
gen) and 20.0 cm (argon).
1

314
L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
Table 8
Observed infrared bands of the BF
ratios
3
·S(CH
3
)
2
complexes isolated in nitrogen and argon matrices, and calculated/experimental wavenumber
a
Complex
Symmetry
species
Mode
Nitrogen
Argon
2
1
21
)
Wavenumber (cm
)
Ratio
Wavenumber (cm
Ratio
1
0
0
BF ·S(CH
3
3
)
2
a
n
n
n
n
n
n
n
n
n
n
n
n
4
1498.7
1478.2
1410.1
1096.8
1028.0
853.5
1.002
1.007
1.051
0.973
0.979
0.996
0.846
0.989
1.022
1.021
1.003
1.013
1498.3
b
1.002
–
5
–
6
1395.1
1040.1
1034.1
837.7
1.063
1.026
0.974
1.015
0.849
0.986
–
8
9
10
12
13
22
24
27
29
690.5
688.2
481.1
482.8
b
0
0
0
a
1424.1
1393.2
925.0
–
b
–
–
b
–
–
474.8
473.9
1.002
11
BF ·S(CH
3
3
)
2
a
n
4
n
5
n
6
1498.7
1478.2
0.998
1.007
1.040
1.042
0.973
0.979
0.995
1.020
0.985
1.033
1.019
1.003
0.996
1498.3
b
–
0.998
–
c
1380.4
1363.9
1.053
c
1
378.4
b
n
n
n
n
n
n
n
n
n
8
1096.8
1028.0
853.5
670.2
481.1
1459.1
1363.8
925.0
474.8
–
–
b
9
–
–
10
12
13
23
24
27
29
837.7
668.2
1.014
0.845
0.982
1.025
–
482.8
0
0
a
1443.1
b
–
b
–
–
473.9
0.998
a
b
c
See Table 7 for assignments.
Not observed.
Matrix splitting.
3
.3.6. The antisymmetric BF bending region,
3
3.4. Complex-monomer wavenumber shifts
2
1
5
00–450 cm
The absorptions at 481.1 and 474.8 cm (nitro-
2
1
The wavenumber shifts experienced by each of
the monomer units on forming the complexes are
presented in Tables 6 and 9. For consistency, only
2
1
gen) and 482.8 and 473.9 cm
assigned to the n13 and n29 modes of the complex
respectively. By analogy with the corresponding
mode of the BF monomer, no isotopic dependence
(argon) were
1
1
those shifts for the BF -containing species in
3
nitrogen matrices are listed. With regard to the
modes of the BF fragments, the predicted signs of
3
is expected to be observed, therefore these bands are
1
assigned to both the BF3 and BF3 isotopic
variants of the complex.
3
0
11
the antisymmetric stretching vibration shifts are in
agreement. While for BF ·S(CH ) the agreement is
3
3 2
The CSC bending ðn Þ; the CH torsional (n and
also quantitatively accurate, in the case of the
BF ·O(CH ) complex the experimental values are
severely underestimated by our calculations.
1
4
3
15
n ) and the intermolecular modes (n ; n ; n ; n ;
31
3
0
16
17
18
3
3 2
2
1
n32 and n ) are all predicted to lie below 400 cm
33
.

L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
315
2
1
Fig. 3. The infrared spectra (1600–1100 cm ) of boron trifluoride and dimethyl sulphide in nitrogen (M/A/A ¼ 250/1/1), deposited at 17 K,
(a) before, (b) after annealing at 35 K.
The signs of the shifts for n12 and n29 of
BF ·O(CH ) and n29 of BF ·S(CH ) are also
DME and DMS fragments, with the exceptions of
n26 of BF ·O(CH ) and n and n of BF ·S(CH ) ,
3
3 2
3
3 2
3
3 2
4
27
3
3 2
correct, but for the smaller shifts of either complex,
the predicted directions are in error. For n12 of
BF ·S(CH ) , moreover, the very large negative
the signs are correctly predicted, although for the
n ; n ; n and n9 modes of BF ·S(CH ) the
computed shifts are an order of magnitude smaller
than those measured.
4
5
8
3
3 2
3
3 2
predicted shift is not realised. In the case of the

316
L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
2
1
Fig. 4. The infrared spectra (1100–450 cm ) of boron trifluoride and dimethyl sulphide in nitrogen (M/A/A ¼ 250/1/1), deposited at 17 K, (a)
before, (b) after annealing at 35 K.
Comparing the experimental shifts of the BF3
modes in the DME and the DMS complexes, it may be
seen that, with the exception of the antisymmetric BF3
bending modes, for which the shifts are negligible,
The directions of the shifts, with the exception of
those of the BF bending modes, are consistent with
3
one another, and are almost exclusively to the red.
1
1
In the case of BF ·O(CH ) two bands associated
3
3 2
1
1
those of BF ·O(CH ) are a factor of four larger than
3
with the COC fragment were detected. The two
stretching bands moved to lower wavenumbers. No
3 2
1
the corresponding displacements for BF ·S(CH ) .
1
3
3 2

L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
317
21
Fig. 5. The infrared spectra (1100–450 cm ) of boron trifluoride and dimethyl sulphide in argon (M/A/A ¼ 250/1/1), deposited at 17 K, (a)
before, (b) after annealing at 35 K.
absorptions could be reasonably assigned to the CSC
between BF3 and DME, on the one hand, is
substantially greater than that with DMS on the
other. This is entirely in agreement with the principle
1
1
fragment in BF ·S(CH ) , mainly on account of
3
3 2
their low predicted intensities [28].
All the experimental and theoretical spectroscopic
indications are that the strength of the interaction
that hard Lewis acids like BF undergo far stronger
3
interactions with hard bases like DME than with softer

318
L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
Table 9
HF·S(CH ) it takes place through the p-type S lone
3 2
Observed and calculated complex-monomer wavenumber shifts for
pair orbital. The factor-of-two difference in the
calculated interaction energies [41] is consistent
with these structural differences. In a similar ab initio
study of the hydrogen-bonded complexes of HF with
H O and H S, Platts et al. [43] also detected the
11
the BF ·S(CH
3
3
)
2
complex
11
Complex (CH
mode
3
)
2
S monomer
BF monomer
3
Mode
Wavenumber
1
Mode Wavenumber shift
21
(cm )
2
2
2
shift (cm
)
preference of the sulphur base to form more
perpendicular hydrogen bonds with HF than the
oxygen base, where the interaction was more linear.
This observation correlates with the known limiting
nucleophilicities of H O (10.0) and H S (4.8) [44].
a
a
Calc.
Expt
Calc.
Expt
0
n
1
n
2
n
3
n
4
n
5
n
6
(a )
n
n
n
n
n
1
(a
1
)
15.1
30.2
17.0
26.9
0.6
0
(a )
12(b
1
)
)
2
2
0
(a )
2
3
(a
(a
1
)
)
Although the interaction between BF and DME and
3
0
(a )
1
50.0
45.5
0
DMS, which is the subject of this study, is not of the
hydrogen-bonded type, similar principles apply with
regard to the properties of the bases involved.
(a )
13(b
1
0
0
b
b
(a )
n
3
(e )
253.7 265.1
2
67.5
0
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
7
8
9
(a )
n
n
n
4
(a
(a
1
)
)
1.1
4.3
9.2
It is apparent that the level of theory used in these
calculations, coupled with an insufficiently extensive
basis set, is inadequate for reproducing the exper-
imental features of the complex spectra with reliable
accuracy. Notwithstanding this deficiency, the useful-
ness of good quality ab initio predictions of the
infrared spectra of molecular complexes as aids to the
interpretation of their experimental matrix isolation
spectra cannot be denied. In particular, the computed
normal coordinates provide valuable information
which can be used to distinguish those complex
0
(a )
5
1
63.6
49.7
0
(a )
0
14(b
1
)
0
10(a )
n
1
(a
1
)
240.1
–
0
0
0
0
11(a )
n
6
(a
1
)
21.6
00
12(a )
n
n
2
(a
2
)
2120.1
25.2
7.9
1.9
0
13(a )
4
(e )
14(a )
n
n
n
n
n
n
n
7
(a
1
)
4.5
0.4
0
15(a )
15(b
16(b
1
)
)
00
19(a )
2
15.4
28.0
17.0
8.1
0
0
20(a )
8
(a
17(b
(a
2
)
00
21(a )
2
)
)
0
0
22(a )
9
2
)
00
23(a )
18(b
2
214.8
modes derived from the BF fragment and those from
3
00
0
b
b
24(a )
n
3
(e )
280.7 281.7
77.1
the bases.
2
0
0
n
n
n
n
n
n
25(a )
n
n
n
n
19(b
2
)
1.1
0
0
26(a )
10(a
20(b
21(b
2
)
13.3
0
0
27(a )
2
)
)
7.4 20.4
0.2
Acknowledgements
0
0
28(a )
2
0
0
0
0
29(a )
n
4
(e )
26.0
24.4
0
TAF wishes to acknowledge the financial support
of the National Research Foundation and the Univer-
sity of Natal Research Fund, while LMN thanks Sasol
30(a )
a
n11(b
2
)
25.4
In nitrogen.
b
Matrix splitting.
(Pty.) Ltd for the award of a Postgraduate Research
Bursary.
ones such as DMS [34]. In a recent ab initio and DFT
study, Ammal and Venuvanalingam [41] examined
the hydrogen-bonded complexes formed between HF
and DME and DMS. Applying the natural bond orbital
References
(
NBO) approach [42], they concluded that the
[
1] F.M.M. O’Neill, G.A. Yeo, T.A. Ford, J. Mol. Struct. 173
(1988) 337.
significant differences in the geometries of the
complexes (in HF·O(CH3)2 the FH· · ·O angle is
[2] L.M. Nxumalo, T.A. Ford, Vib. Spectrosc. 6 (1994) 333.
[
3] L.M. Nxumalo, T.A. Ford, in: J.E. Bertie, H. Wieser (Eds.),
Ninth International Conference on Fourier Transform Spec-
troscopy, Proc. SPIE Int. Soc. Opt. Engng, vol. 2089, 1993, pp.
200,201.
1
808, while the FH· · ·S angle in HF·S(CH ) is about
3 2
1008) could be attributed to the fact that in
HF·O(CH ) the hydrogen bonding interaction occurs
through the s-type non-bonding O orbital, while in
3
2
[4] L.M. Nxumalo, T.A. Ford, S. Afr. J. Chem. 48 (1995) 30.

L.M. Nxumalo, T.A. Ford / Journal of Molecular Structure 656 (2003) 303–319
319
[
5] D.G. Evans, G.A. Yeo, T.A. Ford, Faraday Disc. Chem. Soc.
[25] F. Hirota, Y. Koyama, S. Shibata, J. Mol. Struct. 70 (1981)
305.
86 (1988) 55.
[
[
6] L.M. Nxumalo, T.A. Ford, J. Mol. Struct. 436/437 (1997) 69.
7] L.M. Nxumalo, T.A. Ford, Spectrochim. Acta 53A (1997)
[26] T.J. Le Page, K. Wiberg, J. Am. Chem. Soc. 110 (1988) 6642.
[27] M. Sana, G. Leroy, C. Wilante, Organometallics 11 (1992)
781.
2
511.
8] L.M. Nxumalo, M. Andrzejak, T.A. Ford, Vib. Spectrosc. 12
1996) 221.
9] L.M. Nxumalo, T.A. Ford, Mikrochim. Acta [Suppl.] 14
1997) 383.
[
[
[28] L.M. Nxumalo, T.A. Ford, J. Mol. Struct. (Theochem) 369
(
1996) 115.
(
[
[
[
[
[
29] T. Pradeep, C.S. Sreekanth, M.S. Hegde, C.N.R. Rao, Chem.
Phys. Lett. 151 (1988) 499.
(
30] T. Pradeep, C.N.R. Rao, J. Mol. Struct. (Theochem) 200
[
[
10] L.M. Nxumalo, T.A. Ford, J. Mol. Struct. 300 (1993) 325.
11] L.M. Nxumalo, T.A. Ford, J. Mol. Struct. (Theochem) 357
(
1989) 339.
31] T. Pradeep, C.S. Sreekanth, C.N.R. Rao, J. Chem. Phys. 90
1989) 4704.
(1995) 59.
(
[
[
[
[
[
12] L.M. Nxumalo, M. Andrzejak, T.A. Ford, J. Chem. Inf.
Comput. Sci. 36 (1996) 377.
32] R.M. Archibald, D.R. Armstrong, P.G. Perkins, J. Chem. Soc.,
Faraday Trans. II 69 (1973) 1793.
13] L.M. Nxumalo, M. Andrzejak, T.A. Ford, J. Mol. Struct. 509
33] H.L. Morris, N.L. Kulevsky, M. Tamres, S. Searles, Inorg.
Chem. 5 (1966) 124.
(1999) 287.
14] L.M. Nxumalo, G.A. Yeo, T.A. Ford, Theoret. Chem.
Accounts 96 (1997) 157.
[
[
34] R.G. Pearson, J. Org. Chem. 54 (1989) 1423.
35] R.G. Parr, W. Yang, Density Functional Theory of Atoms and
Molecules, Oxford University Press, Oxford, 1989.
36] R. Janoschek, Pure Appl. Chem. 73 (2001) 1521.
37] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
38] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
39] B. Miehlich, A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett.
15] L.M. Nxumalo, G.A. Yeo, T.A. Ford, S. Afr. J. Chem. 51
(1998) 25.
[
[
[
[
16] S.A. Peebles, L. Sun, R.L. Kuczkowski, L.M. Nxumalo, T.A.
Ford, J. Mol. Struct. 471 (1998) 235.
[
[
17] T.A. Ford, D. Steele, J. Phys. Chem. 100 (1996) 19336.
18] S.H. Bauer, G.R. Finlay, A.N. Laubengayer, J. Am. Chem.
Soc. 65 (1943) 889.
1
57 (1989) 200.
[
[
40] T. Shimanouchi, Tables of Molecular Vibrational Frequen-
cies, Consolidated Volume 1, NSRDS-NBS 39, National
Bureau of Standards, Washington, DC, 1972.
[
19] S.H. Bauer, G.R. Finlay, A.W. Laubengayer, J. Am. Chem.
Soc. 67 (1945) 339.
[
[
20] S. Shibata, K. Iijima, Chem. Lett. (1977) 29.
21] G.M. Begun, W.H. Fletcher, A.A. Palko, Spectrochim. Acta
41] S.S.C. Ammal, P. Venuvanalingam, J. Chem. Soc., Faraday
Trans. 94 (1998) 2669.
1
8 (1962) 655.
[42] A.E. Reed, L.A. Curtiss, F. Weinhold, Chem. Rev. 88 (1988)
899.
[
[
22] E. Taillandier, M. Taillandier, Compt. Rend. 257 (1963) 1522.
23] J. Lascombe, J. Le Calv e´ , M.-T. Forel, Compt. Rend. 258
[43] J.A. Platts, S.T. Howard, B.R.F. Bracke, J. Am. Chem. Soc.
118 (1996) 2726.
(1964) 5611.
[24] R.L. Hunt, B.S. Ault, Spectrosc. Int. J. 1 (1982) 45.
[44] A.C. Legon, D.J. Millen, J. Am. Chem. Soc. 109 (1987) 356.