Condensed-phase effects on the structural properties of C6H 5CN-BF3 and (CH3)3CCN-BF 3: IR spectra, crystallography, and computations

Article abstract of DOI:10.1021/jp052495q

Condensed-phase effects on the structure and bonding of C₆H ₅CN-BF₃ and (CH₃)₃CCN-BF₃ are illustrated by a variety of results, and these are compared to analogous data for the closely related complex CH₃CN-BF₃. For the most part, the structural properties of C₆H₅CN-BF ₃ and (CH₃)₃CCN-BF₃ are quite similar, not only in the gas phase but also in the solid state and in argon matrices. However, the structures do change significantly from medium to medium, and these changes are reflected in the data presented below. Specifically, the measured crystallographic structure of C₆H₅CN-BF ₃ (s) has a B-N distance that is 0.17 A shorter than that in the equilibrium gas-phase structure obtained via B3LYP calculations. Notable differences between calculated gas-phase frequencies and measured solid-state frequencies for both C₆H₅CN-BF₃ and (CH ₃)₃CCN-BF₃ were also observed, and in the case of (CH₃)₃CCN-BF₃, these data implicate a comparable difference between solid-state and gas-phase structure, even in the absence of crystallographic results. Frequencies measured in argon matrices were found to be quite similar for both complexes and also very near those measured previously for CH₃CN-BF₃, suggesting that all three complexes adopt similar structures in solid argon. For C₆H ₅CN-BF₃ and (CH₃)₃CCN-BF ₃, matrix IR frequencies differ only slightly from the computed gas-phase values, but do suggest a slight compression of the B-N bond. Ultimately, it appears that the varying degree to which these systems respond to condensed phases stems from subtle differences in the gas-phase species, which are highlighted through an examination of B-N distance potentials from B3LYP calculations. The larger organic substituents appear to stabilize the potential near 1.8 A, so that the structures are more localized in that region prior to any condensed-phase interactions. As a result, the condensed-phase effects on the structural properties of C₆H₅CN-BF₃ and (CH₃)₃CCN-BF₃ are much less pronounced than those for CH₃CN-BF₃.

Full text of DOI:10.1021/jp052495q

J. Phys. Chem. A 2005, 109, 8199-8208
8199
Condensed-Phase Effects on the Structural Properties of C6H5CN-BF3 and
(CH3)3CCN-BF3: IR Spectra, Crystallography, and Computations
J. A. Phillips,*,† D. J. Giesen,‡ N. P. Wells,† J. A. Halfen,† C. C. Knutson,† and J. P. Wrass†
Department of Chemistry, UniVersity of WisconsinsEau Claire, 105 Garfield AVe.,
Eau Claire, Wisconsin 54701 and Research Laboratories, Eastman Kodak Company,
Rochester, New York 14650-2109
ReceiVed: May 12, 2005; In Final Form: June 16, 2005
Condensed-phase effects on the structure and bonding of C6H5CN-BF3 and (CH3)3CCN-BF3 are illustrated
by a variety of results, and these are compared to analogous data for the closely related complex CH3CN-
BF3. For the most part, the structural properties of C6H5CN-BF3 and (CH3)3CCN-BF3 are quite similar, not
only in the gas phase but also in the solid state and in argon matrices. However, the structures do change
significantly from medium to medium, and these changes are reflected in the data presented below. Specifically,
the measured crystallographic structure of C6H5CN-BF3 (s) has a B-N distance that is 0.17 Å shorter than
that in the equilibrium gas-phase structure obtained via B3LYP calculations. Notable differences between
calculated gas-phase frequencies and measured solid-state frequencies for both C6H5CN-BF3 and (CH3)3CCN-
BF3 were also observed, and in the case of (CH3)3CCN-BF3, these data implicate a comparable difference
between solid-state and gas-phase structure, even in the absence of crystallographic results. Frequencies
measured in argon matrices were found to be quite similar for both complexes and also very near those
measured previously for CH3CN-BF3, suggesting that all three complexes adopt similar structures in solid
argon. For C6H5CN-BF3 and (CH3)3CCN-BF3, matrix IR frequencies differ only slightly from the computed
gas-phase values, but do suggest a slight compression of the B-N bond. Ultimately, it appears that the varying
degree to which these systems respond to condensed phases stems from subtle differences in the gas-phase
species, which are highlighted through an examination of B-N distance potentials from B3LYP calculations.
The larger organic substituents appear to stabilize the potential near 1.8 Å, so that the structures are more
localized in that region prior to any condensed-phase interactions. As a result, the condensed-phase effects
on the structural properties of C6H5CN-BF3 and (CH3)3CCN-BF3 are much less pronounced than those for
CH3CN-BF3.
TABLE 1: Proton Affinities of Various Nitrogen Donors,
and B-N Distances in Their Respective BF3 Complexes
I. Introduction
The structural chemistry of nitrile-boron trifluoride com-
plexes has been the subject of numerous studies over the past
few decades.1-17 Recent interest has stemmed in part from the
observation that gas-phase structures of these systems depict
interactions that defy classification as purely bonding or
nonbonding interactions.1 Furthermore, these structures change
dramatically upon crystallization; the B-N dative bond shortens,
and accordingly, the N-B-F angle increases as well.1,2 For
example, the experimental gas-phase structure of CH3CN-BF3
has a B-N distance of 2.01 Å and an N-B-F angle of 95.6°.3
Clearly, these structural parameters are very much intermediate
between those observed for strong complexes such as H3N-
BF34 and “van der Waals complexes” like N2-BF35 or NCCN-
BF3,6 for which gas-phase B-N distances are listed in Table 1
below (together with proton affinity values for the nitrogen
donors). In the solid state, however, the B-N distance of
CH3CN-BF3 is 1.65 Å and the N-B-F angle is 105.6°,7 much
like H3N-BF3. Thus, crystallization causes the B-N distance
to contract by about 0.4 Å, and the N-B-F angle to open by
molecule
proton affinitya
R(B-N)b
NH3
853.6
811.5
810.9
779.2
745.7
712.9
674.7
493.8
1.673(10)c
C6H5CN
(CH3)3CCN
CH3CN
ClCH2CN
HCN
NCCN
N2
1.798d
1.778d
2.011(7)e
2.473(29)f
2.647(3)g
2.875(2)h
a -∆H values (in kJ/mol) for the reaction B (g) + H+ (g) f HB+
(g). From ref 18. b In units of Å and from experimental data, except
for those obtained in this work. c Ref 4. d This work. e Experimental,
vibrationally-averaged value from ref 3. The calculated equilibrium
distance is 1.868 Å at the B3LYP/aug-cc-pVTZ level, and 2.315 Å at
the B3LYP/aug-cc-pVQZ level. f Ref 8. g Ref 6. h Ref 5.
10°. For HCN-BF3, the gas-phase B-N distance is 2.47 Å,8
but contracts nearly 0.8 Å upon crystallization to 1.64 Å.9
Such large differences between gas and solid-state structures
raise the question as to how a solvent or other bulk, condensed-
phase medium may affect these systems. While there is no direct
method by which precise structural properties can be measured
in solution, IR frequency shifts have offered some semiquan-
titative insight. A few early studies focused on shifts of the CN
stretching band of CH3CN-BF3 in various solvents,10 but
* Corresponding author. E-mail: phillija@uwec.edu. Telephone: 715-
836-5399. Fax: 715-836-4979.
† University of WisconsinsEau Claire.
‡ Eastman Kodak Company.
10.1021/jp052495q CCC: $30.25 © 2005 American Chemical Society
Published on Web 08/19/2005

8200 J. Phys. Chem. A, Vol. 109, No. 36, 2005
Phillips et al.
a
surprisingly, the observed frequencies (∼2355 cm-1) are largely
solvent independent. Our work has been concerned with the
effects of inert, cryogenic matrices, and we have focused on
vibrational modes involving motions in the BF3 moiety, which
are quite sensitive to structural changes. Frequencies of CH3CN-
BF3 measured in solid argon11 are significantly different from
those measured for pure solid CH3CN-BF3,12 and the shifts
are consistent with the notion that the matrix-isolated complex
is more weakly bonded and has a longer B-N distance than its
crystalline counterpart. Experimental gas-phase frequencies have
yet to be measured, but frequencies were calculated for two
minimum-energy structures with B-N distances of 1.92 and
2.32 Å (see below).13 Ultimately, all matrix frequencies were
found to be bracketed by the solid-state data and those calculated
for the 1.92 Å structure, indicating that the matrix environment
does induce substantial structural changes in CH3CN-BF3,
though not the extent that occurs in the crystal.
TABLE 2: X-ray Crystallographic Data for C6H5CN-BF3
empirical formula
formula weight
crystal system
space group
A (Å)
C7H5BF3N
170.93
orthorhombic
Pnma
17.404(2)
8.678(2)
4.856(1)
733.4(2)
4
B (Å)
c (Å)
V (Å3)
Z
temperature (K)
calculated density (g cm-3
crystal size (mm)
100
1.548
)
0.45 × 0.40 × 0.35
absorption coefficient (mm-1
2θ max (deg)
)
0.144
52.10
764
no. of reflections collected
no. of independent reflections
no. of reflections with I > 2σ(I)
no. of variables
763
668
75
R1 (wR2) [I > 2σ(I)] b
R1 (wR2) [all data]
GOF
0.0378 (0.1036)
0.0432 (0.1079)
1.118
Prior to our matrix work, several computational studies of
CH3CN-BF3 had been conducted, and surprisingly, two
distinctly different types of structures had been obtained, one
with a B-N distance of about 1.8 Å14 and three others with
B-N distances between 2.2 and 2.3 Å.15-17 Moreover, all
compared quite poorly to the experimental structure, although
they were conducted at the MP2 level of theory with reasonably
large basis sets. Regardless, the lack of agreement precluded
any valid comparison between our measured, argon-matrix
frequencies and those calculated for the 1.8 Å structure.14 In
turn, we reexamined CH3CN-BF3 using both MP2 and B3LYP
calculations and larger basis sets, in hopes of resolving the
experiment-theory structure discrepancy and obtaining reliable
gas-phase frequencies.13 Surprisingly, we were also unable to
reproduce the experimental structure, but did obtain structures
very much like both of those reported in the previous compu-
tational studies. The key was the inclusion of diffuse functions
in the basis set (signified by “aug” or “+”). With them, shorter
B-N distances near 1.8 Å were obtained; without them, longer
B-N distances near 2.3 Å were obtained. The exception was
the structure obtained with the largest basis set (B3LYP/aug-
cc-pVQZ), which had an equilibrium distance of 2.315 Å,
despite the diffuse functions. The underlying peculiarity was a
very flat B-N distance potential, and in several cases, including
B3LYP/aug-cc-pVQZ, it exhibited two distinct minima roughly
corresponding to each type of structure obtained previously.
Ultimately, it was inferred that the experiment-theory structure
discrepancy was most likely genuine, and stemmed from a large
amplitude zero-point motion in the B-N stretching mode that
resulted in a bona fide difference between the experimental
(vibrationally averaged) and theoretical (equilibrium) struc-
tures.13
difference peaks (e- Å-3
)
0.234, -0.263
a See Experimental Section for additional data collection, reduction,
b
structure solution, and refinement details. R1 ) Σ ||Fo| - |Fc||/Σ
|Fo|; wR2 ) [Σ[w(Fo2 - Fc2)2]]1/2, where w ) 1/σ2(Fo ) + (aP)2 + bP.
2
ture for C6H5CN-BF3. Together, these results show that
(CH3)3CCN-BF3 and C6H5CN-BF3 are rather sensitive to
condensed-phase media, but the effects are somewhat muted
relative to CH3CN-BF3.
II. Experimental Section
Materials. (CH3)3CCN (98%) and C6H5CCN (99+%) were
obtained from Aldrich and were subjected to a few freeze-
pump-thaw cycles prior to use. BF3 (99.5+%) was also
obtained from Aldrich, and other gases, including argon
(99.9995%) and N2 (standard purity), were obtained from
Praxair. All gases were used without further purification
Preparation of Crystalline Samples. Solid samples of
C6H5CN-BF3 and (CH3)3CN-BF3 were prepared by first
adding about 5 mL of liquid nitrile to a 50- or 100-mL Schlenk
tube. The top of the tube was fitted with a rubber septum. Gases
were introduced through the septum via a small hypodermic
needle and were allowed to vent through a sidearm (and into a
fume hood) by slightly opening a Teflon stopcock. Initially,
the tube was purged with a steady flow of N2 for several minutes
to displace residual air. Then, BF3 was slowly introduced to
the tube in a similar manner. Immediately, a white solid formed,
and the tube became somewhat warm. The BF3 flow was
continued for a few additional minutes to ensure that nearly all
the nitrile had reacted, and it was clear (via white fumes from
the sidearm) that the BF3 had displaced most of the nitrogen.
The excess BF3 was allowed to remain in the tube, and the
sample was sealed and placed in a desiccator for several days
to allow crystal growth. The compounds were apt to decompose
rapidly in air, presumably because of the reaction with ambient
water vapor. Because of this, samples were cooled to 0 °C and
immersed in oil prior to mounting on the diffractometer needle.
Crystallography. A colorless block crystal of C6H5CN-BF3
with dimensions of 0.45 mm × 0.40 mm × 0.35 mm was
cleaved from a larger sample and transferred to a Bruker/Nonius
MACH3S diffractometer equipped with graphite-monochro-
mated Mo KR radiation (R ) 0.71073 Å) for a data collection
at 100 K. Relevant crystallographic information is given in Table
2. Unit cell constants were determined from a least-squares
refinement of the setting angles of 25 machine-centered reflec-
This work represents our first step toward identifying other
systems with remarkable structural features akin to those of
CH3CN-BF3. At this point, we turn to the BF3 complexes of
nitrile donors with larger organic substituents, specifically
trimethyl acetonitrile ((CH3)3CCN) and benzonitrile (C6H5CN).
Table 1 shows gas-phase proton affinity values18 for these
compounds and a handful of other nitriles, together with
ammonia and N2 as reference points for fairly strong and very
weak Lewis bases, respectively. These data illustrate that the
species with larger organic substituents are indeed stronger bases
than CH3CN, but they are still weaker than NH3. Below, we
present a myriad of structural data for these complexes,
including gas-phase structures, frequencies, and B-N distance
potentials from B3LYP computations, IR spectra from argon
matrices and thin films of the bulk solids, and an X-ray struc-

Condensed-Phase Effects on C6H5CN-BF3 and (CH3)3CCN-BF3
J. Phys. Chem. A, Vol. 109, No. 36, 2005 8201
tions in the range 9.98° < θ < 25.71°. Intensity data were
collected using the ω/2θ scan technique to a maximum 2θ value
of 52.10°. The intensity data were corrected for Lorentz and
polarization effects and converted to structure factors using the
program XCAD4.19
recorded after each one. Optimal signals were obtained with a
deposition rate of about 5 mmol/h and host concentrations in
the individual gas mixtures of about 0.25%. When two mixtures
were deposited at roughly equal rates, the final composition of
the matrix sample was about 1:1:800 (BF3/nitrile/argon). Most
samples were subsequently annealed to approximately 25-30
K for 30-60 min to facilitate complex formation and aid in
the identification of peaks. A few samples were annealed to
higher temperatures for the purpose of identifying peaks due to
larger aggregates. IR spectra of the bulk, solid complexes were
obtained by codepositing the gas mixtures in a manner similar
to the matrix experiments, except that the cold window was
held at a much warmer temperature, typically 100 K, such that
the host gas did not condense. These spectra closely resemble
low-temperature, solid-state IR spectra of CH3CN-BF3,12
indicating that majority of the sample consists of nitrile-BF3
complexes, though excess, unreacted BF3 was also observed in
some experiments.
The space group Pnma was determined on the basis of
systematic absences and intensity statistics. A successful direct-
methods solution was calculated that provided the positions of
most of the non-hydrogen atoms directly from the E-map. The
remaining non-hydrogen atoms were located after several cycles
of structure expansion and full-matrix least-squares refinement.
Hydrogen atoms were added geometrically. All non-hydrogen
atoms were refined with anisotropic displacement parameters.
Hydrogen atoms were refined as riding atoms with group
isotropic displacement parameters fixed at 1.2 × U(eq) of the
host carbon atom. The final difference map was featureless, with
the largest residual peak (0.23 e- Å-3) located 0.64 Å from
C2. Structure solution and refinement calculations were per-
formed using the SHELXTL suite of programs.20 Complete
crystallographic information in CIF format is provided as
Supporting Information (Table S1).
III. Computational Methods
All calculations were performed using either Gaussian 9821
or Gaussian 03.22 To determine the effect of basis set size, a
variety of basis sets were explored including 3-21G,23 MIDI!,24
6-31G*,25 6-31+G*,25 6-311G*,26,27 6-311+G*,26,27 cc-pVDZ,28
aug-cc-pVDZ,28 cc-pVTZ,28 aug-cc-pVTZ,28 and aug-cc-
pVQZ.28 For aug-cc-pVTZ and aug-cc-pVQZ, a mixed basis
set was used as follows: the large basis set was used on five
atoms, the BF3 group plus the nitrile C and N, and aug-cc-
pVDZ was used on all other atoms. These two mixed basis sets
will be referred to as aug-cc-pV(T|D)Z and aug-cc-pV(Q|D)Z.
When appropriate, symmetry was used, C3V for (CH3)3CCN-
BF3 and Cs for C6H5CN-BF3. B-N distance potentials were
computed by constraining the B-N distance to a specific value
and optimizing all other degrees of freedom. Such points were
calculated every 0.1 Å between 1.6 and 2.5 Å, inclusive.
Because of the shallow B-N potential energy surfaces for these
molecules, minimum-energy conformations were computed by
optimizing all degrees of freedom using the OPT ) TIGHT
option, which tightens the optimization criteria by a factor of
30, and frequencies were calculated to ensure a minimum had
been found. In the case of (CH3)3CCN-BF3, the structure with
the tert-butyl carbons eclipsing the fluorine atoms showed a
Several crystalline samples of (CH3)3CCN-BF3, similar in
size to those of C6H5CN-BF3 were also obtained; however,
these either decomposed or underwent a phase transition en route
to lower temperatures. As such, an insufficient number of
reflections were obtained for (CH3)3CCN-BF3, and a structure
determination was not possible.
Infrared Spectra. All infrared spectra were collected using
a Cryomech ST15 optical cryostat fitted with KBr windows
(both internal and external). Temperature was measured and
controlled using a Scientific Instruments 9600-1 temperature
controller and a single Si diode located at the end of the second
refrigeration stage. As such, the actual sample temperature is
probably a few degrees warmer than that indicated by the
controller. Spectra were recorded using a Nicolet Avatar 360
FTIR at 1.0 cm-1 resolution, and typically, 100 scans were
averaged for both sample and background data. The observed
peak maxima were reproducible to (1 cm-1 for matrix
experiments, and about (5 cm-1 for bulk solid measurements.
The cryostat apparatus is assembled on a movable cart, which
is wheeled into the spectrometer sample compartment for each
spectrum. As such, interference from residual water and CO2
in the unpurged beam path can be problematic on occasion.
Separate gas mixtures containing either the nitrile species or
BF3 were flowed into the cryostat vacuum chamber utilizing a
concentric, dual-deposition flange described previously.11 This
design enables controlled mixing of reactive gas mixtures
immediately prior to the point that they enter the cryostat
vacuum chamber.
very small negative (i.e., imaginary) frequency of -3 cm-1
,
while the structure with the C and F atoms staggered showed a
very small positive frequency of +3 cm-1. Attempts were made
to find a nonsymmetric minimum for C6H5CN-BF3 such as
that seen in the crystal structure, but the tilted geometry
optimized to a structure with a C-N-B angle greater than
179.9° under normal convergence criteria. Attempts to refine
this structure with OPT ) TIGHT to determine whether the
structure was truly linear about the C-N-B angle were
unsuccessful at achieving convergence at the specified criteria,
but the geometry remained very close to 180°. Isotope effects
on the frequencies were computed by comparing the frequencies
calculated with the default 11B isotope to those calculated using
10B. The other isotopes used in the frequency calculations were
1H, 12C, 14N, and 19F.
Matrix IR spectra of (CH3)3CCN-BF3 were obtained by
codepositing separate mixtures of BF3 in argon and (CH3)3CCN
in argon, for which guest-to-host ratios ranged from 1/8 to 1%.
Because C6H5CN has a vapor pressure below 1 Torr, it was
difficult to make gas mixtures rich enough in C6H5CN at a total
pressure above 1 atm. As such, most of these experiments were
conducted with a sample tube containing the pure liquid attached
directly to the deposition line, while a BF3/argon mixture was
deposited via the other line. Peaks due to the 1:1 complex were
much clearer in these spectra, but the concentration of C6H5CN
in the matrix was somewhat uncertain, though it could be
controlled via fine adjustment of a needle valve. Matrices were
deposited at temperatures between 9 and 14 K, with deposition
rates ranging from 3 to 10 mmol/h. Typically, 2-4 successive
depositions were performed over several hours, and spectra were
IV. Results & Discussion
Structure. Equilibrium gas-phase structures for C6H5CN-
BF3 and (CH3)3CCN-BF3 obtained from B3LYP/aug-cc-
pV(T|D)Z calculations are displayed in Figure 1. As a whole,
these structures are quite similar, with the (CH3)3CCN complex
having a slightly shorter B-N distance. In turn, the (CH3)3CCN-

8202 J. Phys. Chem. A, Vol. 109, No. 36, 2005
Phillips et al.
BF3,13 which has a B-N distance of 1.868 Å, an N-B-F angle
of 95.9°, and B-F distances of 1.328 Å. Thus, the appropriate
CH3CN-BF3 data suggest a complex that is just slightly less
strongly bonded than C6H5CN-BF3 and (CH3)3CCN-BF3,
which is consistent with both intuition and proton affinity data.19
The solid-state structure of C6H5CN-BF3 is shown in Figure
2, and despite the differences in proton affinity values among
HCN, CH3CN, and C6H5CN, it very much resembles the crystal
structures of HCN-BF39 and CH3CN-BF3.7 Clearly, however,
it does differ markedly from the calculated gas-phase structure,
as the B-N distance is 0.17 Å shorter, and the N-B-F angle
is about 3° larger. While these differences are less extreme than
those observed previously for HCN-BF3 (∆R(B-N) ) 0.8 Å)9
and CH3CN-BF3 (∆R(B-N) ) 0.4 Å),3 they are still consider-
ably larger than those observed in most instances (typically 0.01
Å).2 One other somewhat peculiar aspect of the structure is that
it is bent slightly about the C-N-B linkage. No evidence of a
bent minimum-energy structure was found in the computations,
which seems to rule out any intramolecular effect, although it
is worth pointing out that the πCN-type orbitals are not
degenerate in benzonitrile (Cs), as they are in the case of the
higher-symmetry (C3V) donors such as acetonitrile. Thus, in
principle, a slight bend could enhance overlap with the higher-
energy πCN orbital in some cases.
On the other hand, it is probable that the slight bend of the
C-N-B unit is due to an intermolecular B-F‚‚‚H interaction.
Each of the three BF3 fluorine atoms participates as an acceptor
in a B-F‚‚‚H “hydrogen bond”, ranging in length from 2.41 to
2.44 Å. The latter of these two interactions, between F(1) and
a para-hydrogen atom of a neighboring C6H5CN-BF3 unit, is
the likely origin of the bend in the C-N-B unit (Figure 3), as
establishment of this nearly linear B-F‚‚‚H interaction would
displace the boron atom from the N(1)≡C(1)-C(2) vector. A
search of the Cambridge Structural Database reveals 39 previous
instances of B-F‚‚‚H interactions in which a benzonitrile
hydrogen atom serves as the “hydrogen bond” donor, ranging
in length from 2.36 to 2.59 Å.29 Weak B-F‚‚‚H interactions in
which BF3 serves as the “hydrogen bond” acceptor are less
Figure 1. B3LYP/aug-cc-pV(T|D)Za Structures of C6H5CN-BF3 and
(CH3)3CCN-BF3.
BF3 structure has a somewhat larger N-B-F angle, and the
B-F bonds are slightly longer as well. It is worth noting that
these differences, though very minor, are opposite of what would
be expected on the basis of the (perhaps negligible) difference
in proton affinity values,18 but are consistent with the intuitive
notion that the tert-butyl substituent is slightly more electron-
releasing than a phenyl group. Regardless, both structures do
resemble those obtained for CH3CN-BF3 when diffuse func-
tions were included in the basis set (except in the case of the
largest basis set: aug-cc-pVQZ).13 It is worth noting again,
however, that the experimental, vibrationally averaged structure
of CH3CN-BF3 has a B-N distance of 2.01 Å,3 and the
equilibrium structure at the B3LYP/aug-cc-pVQZ level has a
B-N distance of 2.315 Å.13 The most reasonable structure for
comparison is the B3LYP/aug-cc-pVTZ result for CH3CN-
Figure 2. Thermal ellipsoid representations (50% probability boundaries) of the X-ray crystal structure of C6H5CN-BF3, with significant interatomic
distances (top) and angles (bottom) noted. Atom pairs F(2)-F(2A), C(3)-C(3A), and C(4)-C(4A) are related by a crystallographic mirror plane.

Condensed-Phase Effects on C6H5CN-BF3 and (CH3)3CCN-BF3
J. Phys. Chem. A, Vol. 109, No. 36, 2005 8203
Figure 3. Intermolecular interaction of neighboring C6H5CN-BF3
complexes in the solid state.
common. One particularly relevant example is observed in the
BF3 adduct of a cyanide ligand in a transition metal complex,
[HRu(DPPE)2CN-BF3]‚CH2Cl2.30 In this complex, the BF3 unit
participates in a slightly shorter intermolecular B-F‚‚‚H interac-
tion (2.38 Å) with a nearby phenyl ring, resulting in bending
of the CtN-B unit to 172.5°.
Frequencies. The identification and assignment of vibrational
bands for matrix-isolated C6H5CN-BF3 and (CH3)3CCN-BF3
was facilitated greatly by our previous work on CH3CN-BF3
in solid argon.11 In general, all bands assigned to the 1:1
complexes were initially identified because they were observed
only when both BF3 and nitrile were present, grew slightly upon
annealing, and were observed at frequencies very near those of
the corresponding modes of CH3CN-BF3. Mode assignments
were ultimately based on 10B-11B isotope shifts and consistent
sets of relative intensities across all experimental conditions
studied. Frequencies for the experimentally observed bands for
both C6H5CN-BF3 and (CH3)3CCN-BF3 are listed in Tables
3 and 4 (below), respectively, together with those calculated
from the gas-phase complexes.
In C6H5CN-BF3 experiments, bands requiring both BF3 and
C6H5CN were observed at 2320, 1253, 1295, 848, 602, and 590
cm-1. All of these features grew slightly upon annealing to 25
K. An additional band was also observed at 874 cm-1, but only
when matrices were deposited above 12 K, or after those
deposited at lower temperatures were annealed. Representative
spectra showing the regions of these absorption features are
displayed Figure 4, and peaks definitively assigned to the 1:1,
C6H5CN-BF3 complex are noted with asterisks and are included
in Table 4. The pair at 1253 and 1290 cm-1 were immediately
recognized as a 10B-11B isotopic pair due to the characteristic
intensity ratio, and a splitting (47 cm-1) that compared favorably
with the computed isotope shift (48 cm-1, below), as well as
Figure 4. IR spectra showing the absorption bands of C6H5CN-BF3
in solid argon at 10 K. Bottom traces: BF3 in argon, 1:400. Middle
traces: C6H5CN in solid argon, approximately 1:400. Spectrum was
obtained by mixing C6H5CN vapor into a stream of pure argon
immediately prior to deposition. Top traces: BF3 and C6H5CN in argon,
approximately 1:1:400, obtained by mixing C6H5CN vapor into a stream
of BF3 and argon. Peaks assigned to C6H5CN-BF3 are marked with
asterisks.
to the asymmetric stretching bands, the peaks at 590 and 602
cm-1 were assigned to the umbrella modes of the 11B and 10B
isotopomers of C6H5CN-BF3, respectively. The peak at 2320
cm-1 occurs about 80 cm-1 higher in frequency than the CN
stretching bands observed for C5H5CN, which compares favor-
ably with blue-shifts observed for various other CH3CN
complexes.10 No 11B-10B isotope shift is predicted for the CN
stretching band of C6H5CN-BF3 (below), and none was
observed previously for CH3CN-BF3. The 2320 cm-1 band did
exhibit reasonably consistent intensities relative to those assigned
above, despite being partially obscured by background CO2 in
many experiments. Thus, we ultimately assigned the peak at
2320 to the CN stretching band (νCN) of the complex. The band
those measured for gas-phase BF3 (51 cm-1 19
and matrix-
)
isolated CH3CN-BF3 (45 cm-1).11 Thus, the peaks at 1253 and
1290 cm-1 were assigned to the BF asymmetric stretching
modes (νaBF) of C6H5CN-11BF3 and C6H5CN-10BF3, respec-
tively. The peaks at 590 and 602 cm-1 also exhibited an intensity
ratio characteristic of a 11B-10B isotopic pair and a splitting
(12 cm-1) consistent with that of the BF symmetric bend or
“umbrella” mode (δaBF) as predicted by computations (13 cm-1
,
below) and in previous measurements of this band for CH3CN-
BF3 (16 cm-1).11 Given this, and consistent intensities relative

8204 J. Phys. Chem. A, Vol. 109, No. 36, 2005
Phillips et al.
Figure 5. IR spectrum of a 1:1:400 sample of BF3 and C6H5CN in
argon at 10 K before (bottom) and after (top) annealing to 35 K for 30
min.
at 848 cm-1 lies in the region of the BF3 symmetric stretch,
31
which is forbidden in free BF3 but can become rather strong
when the BF3 moiety is distorted to a pyramidal geometry.14
This band also has consistent intensities relative to those of the
other bands discussed, and a very small 10B-11B isotope shift
is predicted (below), so the band at 848 cm-1 was assigned to
the B-F symmetric stretch of C6H5CN-BF3.
A group of broad features appeared near those discussed
above when matrices were annealed to fairly high temperatures,
about 35 K, and as such, these are presumably due to larger
aggregates of C6H5CN and BF3 (i.e., either (C6H5CN-BF3)n
or (C6H5CN)m(BF3)n). This claim is supported by the fact that
these features appear in the regions between the bands assigned
to the 1:1 complexes and those observed for the bulk solids in
“matrix-free” experiments (below). Furthermore, they are shifted
relative to those assigned to the 1:1 complex in a manner that
is consistent with species having B-N dative bonds compressed
relative to isolated C6H5CN-BF3,13 which would be expected
on the basis of the medium-induced structural changes noted
above. A spectrum displaying these broad absorption features
is shown in Figure 5, and it should be noted that no clearly
resolved peaks could be distinguished within these regions, nor
could any peaks that could be associated with specific clusters.
Peak maxima did occur on the broad feature near the asymmetric
stretch at 1288, 1240, and 1226 cm-1 near the umbrella band
at 649, 628, 607, and 600 cm-1, and another band also grows
in just to the blue of the CN stretch at 2330 cm-1. The band at
874 cm-1, initially observed when matrices were deposited
above 12 K, also grew very rapidly upon annealing to 35 K,
and thus is likely due to larger aggregates as well.
The rationale by which the IR bands were assigned to
(CH3)3CCN-BF3 in argon matrix experiments was essentially
identical to that for C6H5CN-BF3. Bands requiring both BF3
and (CH3)3CCN were observed at 2323, 1247, 1290, 840, 617,
and 603 cm-1. Secondary peaks were observed for nearly all
of these bands, presumably due to distinct trapping sites in the
matrix. The spectral regions in which these bands occur are
displayed in Figure 6, and bands definitively assigned to the
1:1 complex are marked with asterisks. Frequencies, including
both peak maxima observed for bands that were split, are listed
in Table 4. The pair at 1290 and 1247 cm-1 was assigned to
the asymmetric stretching bands of the 10B and 11B isotopomers,
respectively, on the basis of the observed splitting (43 cm-1).
It is worth noting that a relatively strong band of (CH3)3CCN
almost completely obscures the 11B component, and this
prevented any reliable measurement of the peak area. The bands
Figure 6. Representative IR spectra showing the absorption bands of
(CH3)3CCN-BF3 in solid argon at 13 K. Bottom trace: BF3 in argon,
1:400. Middle trace: (CH3)3CCN in solid argon, 1:400. Top trace: BF3
and C6H5CN in argon, approximately 1:1:800. Peaks assigned to
C6H5CN-BF3 are marked with asterisks.
at 617 and 603 cm-1 were assigned to the umbrella modes of
the 10B and 11B isotopomers, respectively, on the basis of the
isotope splitting (14 cm-1) and relative intensities. The band at
2323 cm-1 was assigned to the CN stretch on the basis of
relative intensities and the characteristic blue shift relative to
the CN stretching band of (CH3)3CCN. An additional band was
also measured at 875 cm-1, but it was only observed upon
annealing, or in warmer depositions (above 12 K), much like
the band observed near this frequency in the benzonitrile/BF3
experiments.
There are significant differences between the matrix frequen-
cies and those observed in the spectra of matrix-free, bulk solid
samples. IR spectra of the bulk solids resulting from codepo-
sition of C6H5CN or (CH3)3CCN with BF3 are displayed in
Figures 7 and 8, and observed frequencies of the major bands
are listed Tables 3 and 4, respectively. In general, the spectra

Condensed-Phase Effects on C6H5CN-BF3 and (CH3)3CCN-BF3
J. Phys. Chem. A, Vol. 109, No. 36, 2005 8205
isotopomer, which comprises 81% of the sample. In the solid-
state C6H5CN-BF3 spectrum, the BF asymmetric stretch is
observed at 1177 cm-1, the BF symmetric stretch at 868 cm-1
,
and the umbrella mode at 654 cm-1. These differ from the
corresponding band in the matrix by -76, +20, and +64 cm-1
,
respectively. The CN stretch is observed as a doublet, just to
the blue of the matrix frequencies, at 2328 and 2374 cm-1. The
underlying reasons for the doubling were not pursued, but we
note that the 2328 peak is essentially coincident with the matrix
band, while the latter is shifted about 50 cm-1 to the blue. All
of these shifts, with the possible exception of the one CN
stretching doublet, are consistent with the idea that the matrix-
isolated complex has a weaker intramolecular bond with a longer
B-N distance than its crystalline counterpart.13
Figure 7. IR spectrum of a bulk solid sample of BF3 and C6H5CN
In the solid-state (CH3)3CCN-BF3 spectrum, the BF3 asym-
metric stretch is observed at 1172 cm-1, the BF3 symmetric
stretch at 865 cm-1, the umbrella mode at 654 cm-1, and the
CN stretch is observed at 2340 cm-1. These differ from the
corresponding bands in the matrix by -75, +25, +34, and +17
cm-1, respectively. The solid-state frequencies, as well as the
relative band intensities are quite similar to those observed for
C6H5CN-BF3 (s) and CH3CN-BF3 (s), and this suggests that
the solid-state structure of (CH3)3CCN-BF3 is also similar to
that of those species. This is significant in this case, because
no X-ray structure for (CH3)3CCN-BF3 could be obtained.
Again, the shifts observed relative to the matrix data are
consistent with the idea that the matrix-isolated complex is more
weakly bonded and has a longer B-N distance than its solid-
state counterpart.
codeposited at 100 K.
Computed (B3LYP/aug-cc-pV(T|D)Z) frequencies of the
experimentally observed bands are included in Tables 3 and 4,
while the complete lists, including intensities, symmetry labels,
and approximate mode descriptions are included as Supporting
Information (Tables S2 and S3). A somewhat unusual scaling
procedure for these frequencies, used previously for CH3CN-
BF3,13 was adopted in order to facilitate agreement between
experiment and theory for CH3CN and BF3. That is, all
frequencies above 2000 cm-1 are scaled by a factor of 0.96,
but all others are left unscaled. Thus, the CN stretch is the only
mode of interest affected by this procedure. In general, the
calculated frequencies differ only slightly from the measured
matrix bands, by less than 20 cm-1 in almost every instance.
However, the differences are systematic, in that gas-to-matrix
shifts are consistent with the idea that the matrix does offer
stabilization for the complexes and, in turn, causes a slight
Figure 8. IR spectrum of a bulk solid sample of BF3 and (CH3)3CCN
codeposited at 100 K.
very closely resemble the solid-state IR spectrum of CH3CN-
BF3,12 which is dominated by three strong bands near 1200,
880, and 650 cm-1 that correspond to the BF3 asymmetric
stretch, the BF3 symmetric stretch, and the BF3 umbrella mode,
respectively. The CN stretch is also reasonably strong in those
spectra,12 and it is observed at 2376 cm-1, but all other bands
of CH3CN-BF3 (s) are quite weak in comparison to these.
Given the similarities between the spectra shown in Figures 7
and 8 and the CH3CN-BF3 data, not only in terms of
frequencies but also relative band intensities, it is presumed that
the dominant bands can be assigned by comparison to CH3CN-
BF3 and that peak frequencies correspond to those of the 11B
TABLE 3: Observed and CalculatedaVibrational Frequencies of C6H5CN-BF3
frequency (cm-1
)
mode
description
phase
C6H5CN-11BF3
C6H5CN-10BF3
A
δBF
B-F symmetric deformation or
“umbrella” mode
gas (B3LYP)
Ar matrix
solid
585
590
654
598
602
S
νBF
BF3 symmetric stretch
BF3 asymmetric stretchb
C-N stretch
gas (B3LYP)
Ar matrix
solid
843
848
868
848
848
A
νBF
gas (B3LYP)
Ar matrix
solid
1266
1253
1177
1314
1295
νCN
gas (B3LYP)
Ar matrix
solid
2300
2320
2328/2374
2300
2320
a Frequencies were calculated using the B3LYP/aug-cc-pVT|D)Z method, and frequencies above 2000 cm-1 were scaled by a factor of 0.96. See
text for discussion. b This mode is not formally degenerate in the complex, as it is in free BF3, because of lower symmetry, but the calculated
splitting is only 0.1 cm-1
.

8206 J. Phys. Chem. A, Vol. 109, No. 36, 2005
Phillips et al.
TABLE 4: Observed and Calculateda Vibrational Frequencies of (CH3)3CCN-BF3
frequency (cm-1
)
mode
description
phase
(CH3)3CCN-11BF3
(CH3)3CCN-10BF3
A
δBF
B-F symmetric deformation or
“umbrella” mode
gas (B3LYP)
Ar matrix
solid
585
603/607
637
601
617/621
S
νBF
BF3 symmetric stretch
BF3 asymmetric stretch
C-N stretch
gas (B3LYP)
Ar matrix
solid
834
840
865
840
840
A
νBF
gas (B3LYP)
Ar matrix
solid
1258
1247
1172
1304
1290
νCN
gas (B3LYP)
Ar matrix
solid
2313
2323/2327
2340
2313
2323
a Frequencies were calculated using the B3LYP/aug-cc-pV(T|D)Z method, and frequencies above 2000 cm-1 were scaled by a factor of 0.96.
See text for discussion.
contraction of the B-N bonds. Given that these observed shifts
are only slightly larger than the differences usually observed
between matrix and gas-phase frequencies, typically 10 cm-1 32
,
this cannot be considered definitive evidence for a matrix effect
on the structure of these complexes. In any event, the data are
consistent with some slight matrix-induced structural changes,
though certainly much less extreme than those depicted by the
analogous data for CH3CN-BF3.11,13
In comparing frequencies for C6H5CN-BF3, (CH3)3CCN-
BF3 and CH3CN-BF3 within a given medium, it is clear that
variations in the degree to which these species are affected by
various condensed-phase environments stem mainly from dif-
ferences in structural properties of the gas-phase species. Indeed,
matrix frequencies for the BF3-localized modes among all three
complexes are quite consistent, suggesting that the structures
of the complexes are similar in solid argon, at least in regard to
the effects on the BF3 moiety. For example, the BF3 asymmetric
stretching frequencies for the 11B isotopomers of C6H5CN-
BF3, (CH3)3CCN-BF3, and CH3CN-BF3 in solid argon are
1253, 1247, and 1249 cm-1, and those of the umbrella modes
are 590, 603, and 601 cm-1, respectively. The frequencies of
the solid-state complexes are nearly as consistent, and further-
more, all of the known crystal structures of RCN-BF3-type
complexes7,9 have nearly identical structures about the nitrile
and BF3 subunits (aside from the slight bend in C6H5CN-BF3
(s)). Very slight gas-phase structure differences were noted
above, and indeed, the calculated frequencies of C6H5CN-BF3
and (CH3)3CCN-BF3 do reflect shifts that parallel these subtle
differences, as expected on the basis of how calculated CH3CN-
BF3 frequencies shift with B-N distance.13 For CH3CN-BF3,
there are two minimum-energy structures at the B3LYP/aug-
cc-pVQZ level of theory, with B-N distances of 1.919 and
2.319 Å, the latter being lower in energy by only 0.11 kcal/
mol. For the shorter structure, the frequencies of the BF3
asymmetric stretching and BF3 umbrella bands are 1306 and
564 cm-1, respectively (for the 11B isotopomer). These are
significantly different from the analogous frequencies of
C6H5CN-BF3 and (CH3)3CCN-BF3 listed in Tables 3 and 4,
and the shifts parallel the differences in predicted structure
between CH3CN-BF3 and the two complexes with larger
organic substituents. However, intermolecular interactions exert
a stabilizing effect that apparently nullifies the gas-phase
structural differences, and in turn, frequencies measured in the
matrix and crystalline environments reflect structures that are
quite similar.
Figure 9. Electronic energy of C6H5CN-BF3 versus B-N distance
using the B3LYP method and various basis sets.
Figure 10. Electronic energy of (CH3)3CCN-BF3 versus B-N distance
using the B3LYP method and various basis sets.
bonding of CH3CN-BF3, C6H5CN-BF3, and (CH3)3CCN-BF3
are further illustrated by a comparison of B-N distance potential
curves. Curves obtained using the B3LYP method with a variety
of basis sets are shown in Figures 9 and 10 for C6H5CN-BF3
and (CH3)3CCN-BF3, respectively. The zero of energy for these
plots were chosen to be the lowest point on the curve, as the
overall shape relative to the minimum is the main concern. In
general, they are all fairly flat, especially the C6H5CN-BF3
curves, and both exhibit the key features observed previously
for CH3CN-BF3, including signs of two distinct minima near
1.8 and 2.3 Å, and a peculiar sensitivity to diffuse functions in
B-N Potential Energy Curves. The differences in the extent
to which condensed-phase environments affect the structure and

Condensed-Phase Effects on C6H5CN-BF3 and (CH3)3CCN-BF3
J. Phys. Chem. A, Vol. 109, No. 36, 2005 8207
V. Summary and Conclusions
Condensed-phase effects on the structure and bonding of
C6H5CN-BF3 and (CH3)3CCN-BF3 have been illustrated by
a myriad of data, and these have been compared and contrasted
to analogous results for the closely related complex CH3CN-
BF3. For the most part, the structural properties of C6H5CN-
BF3 and (CH3)3CCN-BF3 are similar, not only in the gas phase
but also in the solid state and in argon matrices. There is clear
evidence, however, that these structures do change significantly
from medium to medium. The measured crystallographic
structure of C6H5CN-BF3 (s) differs markedly from the
equilibrium gas-phase structure obtained via B3LYP calcula-
tions, though these differences are somewhat less extreme than
those observed for HCN-BF39 and CH3CN-BF3.7 Furthermore,
the IR spectrum of (CH3)3CCN-BF3 (s) suggests that its solid-
state structure resembles those of CH3CN-BF3 and C6H5CN-
BF3, and in turn, that there is a significant gas-solid structure
difference for this complex as well. Notable differences between
calculated gas-phase frequencies and measured solid-state
frequencies for both C6H5CN-BF3 and (CH3)3CCN-BF3 have
been also observed, and the shifts are consistent with those
expected on the basis of the structural changes.13 Frequencies
measured in argon matrices were found to be quite similar for
both complexes and also very near those measured previously
for CH3CN-BF3, suggesting that all three complexes adopt
similar structures, at least about the BF3 moiety, in the inert
matrix environment. This is although there are significant
differences between the calculated gas-phase structures of
CH3CN-BF3 and the two complexes with larger organic
substituents. For C6H5CN-BF3 and (CH3)3CCN-BF3, matrix
IR frequencies differ only slightly from the computed gas-phase
values, but they are shifted systematically in a manner that
suggests a slight compression of the B-N bond relative to the
gas phase. By contrast, there are large differences between
matrix and calculated gas-phase frequencies for CH3CN-
BF3.11,13 Since both IR spectra and available crystallographic
data indicate that the structures of CH3CN-BF3, C6H5CN-
BF3, and (CH3)3CCN-BF3 are similar in the solid-state and
matrix environments, it appears that the varying degree to which
these systems are affected by condensed phases stems ultimately
from subtle differences in gas-phase structural properties. These
differences were highlighted through an examination of B-N
distance potentials obtained from B3LYP calculations using a
variety of basis sets. In general, curves for all three species are
somewhat flat, and indeed, those for both C6H5CN-BF3 and
(CH3)3CCN-BF3 show signs of the two distinct minima
exhibited in the B-N potential of CH3CN-BF3. However, the
larger organic substituents seem to cause a greater stabilization
of the inner region of the potential, such that the structures are
more localized in the 1.8 Å region, even in the absence of
interactions with a bulk, condensed-phase medium. Thus, while
C6H5CN-BF3 and (CH3)3CCN-BF3 do share some of the
peculiarities exhibited by CH3CN-BF3, the structural changes
induced by condensed-phase environments are less extreme.
Figure 11. The B-N distance potentials of CH3CN-BF3, C6H5CN-
BF3, and (CH3)3CCN-BF3, calculated at the B3LYP/aug-cc-pVTZ
(CH3CN-BF3) or B3LYP/aug-cc-pV(T|D)Z (others) level of theory.
the basis set. For CH3CN-BF3, the minima occur near 1.8 Å,
and there is a shelflike region near 2.3 Å when diffuse functions
are included in the basis set. In the absence of diffuse functions,
as well as with the largest basis set used (aug-cc-pVQZ), the
global minimum resides in the 2.3 Å region, and a shelflike
feature occurs near 1.8 Å.13 The C6H5CN-BF3 curves follow
the same general trend, but the basis set effect is not as distinct
for (CH3)3CCN-BF3. It is perhaps most apparent upon com-
parison of the 6-311G* and 6-311+G* data. For the former,
the potentials are quite flat, and for C6H5CN-BF3, the global
minimum occurs in the outer region near 2.3 Å. For (CH3)3CCN-
BF3, the 6-311G* curve is still quite flat, but the global
minimum is still near 1.8 Å, like all others. In contrast, the
6-311+G* curves are steeper and depict a structure that would
be more localized in the 1.8 Å region for both complexes. For
the larger basis sets, the curves lie between the 6-311G* and
6-311+G* data, but those that lack diffuse functions (Figure
10 only) are indeed flatter and systematically lower in energy
in the 2.3 Å region.
The curves are not as flat as those obtained previously for
CH3CN-BF3,13 however. For C6H5CN-BF3 at the B3LYP/aug-
cc-pV(Q|D)Z level, the potential is at or below 0.3 kcal/mol
for 0.7A, from just beyond 1.7 Å to 2.4 Å. For (CH3)3CCN-
BF3 the curves are much steeper. At the B3LYP/aug-cc-
pV(Q|D)Z level, the curve remains under 0.3 kcal/mol for only
about 0.3 Å. A comparison of the B3LYP/aug-cc-pVQZ and
B3LYP/aug-cc-pV(Q|D)Z curves for CH3CN-BF3, C6H5CN-
BF3, and (CH3)3CCN-BF3 is shown in Figure 11, and it
illustrates the very subtle effects that the larger organic
substituents exert on the structures of these systems. Clearly,
the (CH3)3CCN-BF3 curve is steepest and depicts a structure
that is very much localized in the 1.8 Å region. The CH3CN-
BF3 curve is quite flat, and in fact, the experimental, vibra-
tionally averaged gas-phase B-N distance lies near the middle
of the flat region at 2.01 Å.3 The C6H5CN-BF3 curve lies
between the other two but is also quite flat in the 2.1-2.3 Å
region. Thus, with larger organic substituents, the inner region
of the potential is stabilized relative to the 2.3 Å region, and in
turn, the structures are more localized near 1.8 Å. Ultimately,
the medium effects on the structure of the two larger complexes
are more muted because the inner region is more stable even in
the absence of any interactions with some condensed-phase
environment. By contrast, condensed-phase interactions must
stabilize the inner region of the very flat CH3CN-BF3 potential
relative to the outer 2.3 Å region because the complex is more
polar at shorter B-N distances,13 which would drive the
medium-induced structural changes.
Acknowledgment. J.A.P. Gratefully acknowledges the Na-
tional Science Foundation (CHE-0216058 and CHE-0407824)
for financial support of this work, as well as the donors of the
ACS Petroleum Research Fund (37834-B6), and Research
Corporation (CC5622). Additional support was provided by
Office of Research and Sponsored Programs at the University
of WisconsinsEau Claire.
Supporting Information Available: Crystallographic in-
formation file (Table S1), a stereoscopic projection of the

8208 J. Phys. Chem. A, Vol. 109, No. 36, 2005
Phillips et al.
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A9; Gaussian, Inc.:
Pittsburgh, PA, 1998.
C6H5CN-BF3 unit cell (Figure S1), and computed frequencies
of C6H5CN-BF3 (Table S2) and (CH3)3CCN-BF3 (Table S3).
This material is available free of charge via the Internet at http://
pubs.acs.org.
References and Notes
(22) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A.
D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A.
G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian
03, Revision B.05; Gaussian, Inc., Wallingford CT, 2004.
(23) Binkley, J. S.; Pople, J. A.; Hehre, W. J. J. Am. Chem. Soc. 1980,
102, 939.
(1) Leopold, K. R.; Canagaratna, M.; Phillips, J. A. Acc. Chem. Res.
1997, 30, 57.
(2) Leopold, K. R. In AdVances in Molecular Structure Research;
Hargittai, M., Hargittai, I., Eds.; JAI Press: Greenwich, CT 1996; Vol. 2.
(3) Dvorak, M. A.; Ford, R. S.; Suenram, R. D.; Lovas, F. J.; Leopold,
K. R. J. Am. Chem. Soc. 1992, 114, 108.
(4) Fujiang, D.; Fowler, P. W.; Legon, A. C. J. Chem. Soc., Chem.
Commun. 1995, 113.
(5) Janda, K. C.; Bernstien, L. S.; Steed, J. M.; Novick, S. E.;
Klemperer, W. J. Am. Chem. Soc. 1978, 100, 8074.
(6) Leopold, K. R.; Fraser, G. T.; Klemperer, W. J. Am. Chem. Soc.
1978, 100, 8074.
(7) Swanson, B.; Shiver, D. F.; Ibers, J. A. Inorg. Chem. 1969, 8, 2183.
(8) Reeve, S. W.; Burns, W. A.; Lovas, F. J.; Suenram, R. D.; Leopold,
K. R. J. Phys. Chem. 1993, 97, 10630.
(9) Burns. W. A.; Leopold, K. R. J. Am. Chem. Soc. 1993, 115, 11622.
(10) (a) Purcell, K. F.; Drago, R. S. J. Am. Chem. Soc. 1966, 88, 919.
(b) Coerver, H. J.; Curran, C. J. Am. Chem. Soc. 1958, 80, 3522.
(11) Wells, N. P.; Phillips, J. A. J. Phys. Chem. A 2002, 106, 1518.
(12) Swanson, B.; Shriver, D. F. Inorg. Chem. 1970, 9, 1406.
(13) Giesen, D. J.; Phillips, J. A. J. Phys Chem. A 2003, 107, 4009.
(14) Cho, H-.G.; Cheong, B.-S. THEOCHEM 2000, 496, 185.
(15) Jurgens, R.; Almlo¨f, J. Chem. Phys. Lett. 1991, 176, 263.
(16) Jiao, H. J.; Schleyer, P. v. R. J. Am. Chem. Soc. 1994, 116, 7429.
(17) Jonas, V.; Frenking, G.; Reetz, M. T. J. Am. Chem. Soc. 1994,
116, 8741.
(24) Easton, R. E.; Giesen, D. J.; Welch, A.; Cramer, C. J.; Truhlar, D.
G. Theor. Chim. Acta 1996, 93, 281.
(25) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56,
2257.
(26) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.
1980, 72, 650.
(27) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639.
(28) Woon, D. E.; Dunning, T. H., Jr. J. Chem. Phys. 1993, 98, 1358.
(29) Allen, F. H. Acta Crystallogr. 2002, B58, 380-388.
(30) Rocchini, E.; Rigo, P.; Mezzetti, A.; Stephan, T.; Morris, R. H.;
Lough, A. J.; Forde, C. E.; Fong, T. P.; Drouin, S. D. J. Chem. Soc., Dalton
Trans. 2000, 3591-3602.
(31) Herzberg, G. Molecular Spectra and Molecular Structure, Volume
II, Infrared and Raman Spectra of Polyatomic Molecules; Krieger Publishing
Co.: Malabar, FL, 1991.
(18) The NIST Chemistry Webbook; http://webbook.nist.gov/chemistry
(accessed December 2004).
(19) Harms, K. XCAD4: Program for the Lp-correction of CAD4/Mach3
Diffractometer Data; University of Marburg: Marburg, Germany, 1996.
(20) SHELXTL NT v. 6.12; Bruker AXS: Madison WI 2001.
(21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
(32) See for example: Dunkin, I. R. Matrix Isolation Techniques: A
Practical Approach; Oxford University Press: Oxford, 1998.