Structure, conformations, and internal hydrogen bonding in gaseous 4-fluorobutan-1-ol. Gas-phase electron diffraction and Ab initio study

Article abstract of DOI:10.1021/jp012392h

Excluding enantiomers, there are 14 possible rotomeric conformations of 4-fluorobutan-1-ol, of which two are capable of forming an internal O-H...F hydrogen bond. The composition of the gaseous system is of special interest because it is determined by the energies of the conformers which reflect the energies of the hydrogen bonds. We have investigated the conformational composition of the gaseous system and the molecular structures of the conformers at 83 °C by gas-phase electron diffraction (GED) augmented by molecular orbital calculations. Because of the complexity of the 4-fluorobutan-1-ol system, the parameters of the several models tested were simplified by various constraints taken from the theoretical work. With these constraints, the best agreement with the GED data was obtained with a model consisting of about equal amounts of hydrogen-bonded and non-hydrogen-bonded conformers. Because the curled-up shape of the two forms capable, in principle, of forming internal hydrogen bonds is expected to be energetically unfavorable in the absence of such bonding, the experimental results are interpreted as strong evidence for its existence. Weighted average values, with estimated 2σ uncertainties, of the more important bond distances (r_a/A) and bond angles (∠_a/deg) for the preferred model are r(C_o-C) = 1.529(2), r(C_C-C_C) = 1.537(2), r(C-C_F) = 1.520(2), r(C-O) = 1.430(5), r(C-F) = 1.401(5), ∠(C_o-C-C) = 112.5(33), ∠(C-C-O)_{H bond} = 112.3(55), ∠(C-C-O)_{no H bond} = 108.2(45), and ∠(C-C-C_F) = 109.8(12). The two hydrogen-bonded conformers comprise 48.5% of the mixture with an estimated 2σ uncertainty of 14.0%. A rough estimate of the energy of the O-H...F hydrogen bond is 3 kcal mol^-1. The average O...F separation in this bond for the H-bonded conformers is 2.46(4) A, about 0.3 A less than the sum of the van der Waals radii.

Full text of DOI:10.1021/jp012392h

J. Phys. Chem. A 2001, 105, 9587-9594
9587
Structure, Conformations, and Internal Hydrogen Bonding in Gaseous 4-Fluorobutan-1-ol.
Gas-Phase Electron Diffraction and Ab Initio Study
Marit Trætteberg,*, Alan D. Richardson, Kenneth Hedberg,* Rolf W. Winter, and
†,‡
‡
,‡
§
Gary L. Gard§
Department of Chemistry, Oregon State UniVersity, CorVallis, Oregon 97331, and Department of Chemistry,
Portland State UniVersity, Portland, Oregon 97207
ReceiVed: June 21, 2001
Excluding enantiomers, there are 14 possible rotomeric conformations of 4-fluorobutan-1-ol, of which two
are capable of forming an internal O-H‚‚‚F hydrogen bond. The composition of the gaseous system is of
special interest because it is determined by the energies of the conformers which reflect the energies of the
hydrogen bonds. We have investigated the conformational composition of the gaseous system and the molecular
structures of the conformers at 83 °C by gas-phase electron diffraction (GED) augmented by molecular orbital
calculations. Because of the complexity of the 4-fluorobutan-1-ol system, the parameters of the several models
tested were simplified by various constraints taken from the theoretical work. With these constraints, the best
agreement with the GED data was obtained with a model consisting of about equal amounts of hydrogen-
bonded and non-hydrogen-bonded conformers. Because the curled-up shape of the two forms capable, in
principle, of forming internal hydrogen bonds is expected to be energetically unfavorable in the absence of
such bonding, the experimental results are interpreted as strong evidence for its existence. Weighted average
values, with estimated 2σ uncertainties, of the more important bond distances (r
/deg) for the preferred model are r(C -C) ) 1.529(2), r(C -C ) ) 1.537(2), r(C-C
r(C-O) ) 1.430(5), r(C-F) ) 1.401(5), ∠ (C -C-C) ) 112.5(33), ∠ (C-C-O)H bond ) 112.3(55),
(C-C-O)no H bond ) 108.2(45), and ∠ (C-C-C ) ) 109.8(12). The two hydrogen-bonded conformers
comprise 48.5% of the mixture with an estimated 2σ uncertainty of 14.0%. A rough estimate of the energy
a
/Å) and bond angles
(
∠
a
O
C
C
F
) ) 1.520(2),
O
∠
F
-
1
of the O-H‚‚‚F hydrogen bond is 3 kcal mol . The average O‚‚‚F separation in this bond for the H-bonded
conformers is 2.46(4) Å, about 0.3 Å less than the sum of the van der Waals radii.
Introduction
of systems in which the substituents are farther apart. In previous
work, we have discovered that gaseous butan-1,4-diol consists
The 1,2-disubstituted ethanols consist of a mixture of gauche
and anti conformers in proportions that depend on the nature
of the substituents. In general, the dominant form of these
molecules is anti; however, when the possibility of internal
hydrogen bonding exists, the gauche form may be the more
abundant. An interesting special case is 1,2-difluoroethane;
despite the absence of hydrogen-bonding capability, it is nearly
all gauche in the gas phase, whereas the dichloro and dibromo
compounds are mostly anti. This effect arising from the high
2
of about 40% H-bonded forms at 144 °C, but gaseous 1,4-
3
difluorobutane has scarcely any forms with the fluorine atoms
in close proximity. As expected, the results for the latter show
that the gauche effect plays no role in the stabilization of the
“
curled-up” forms of the 1,4-disubstituted butanes. The set of
4
5
results from the series ethane-1,2-diol (glycol), 2-fluoroethanol,
and 1,2-difluoroethane (all primarily gauche) and the pair of
6
1
2
butanes, butane-1,4-diol and 1,4-difluorobutane, suggests a
similar investigation of 4-fluorobutan-1-ol (hereafter 4-FB;
Figure 1) to shed more light on the strength of the (possible)
internal hydrogen bond in this molecule. This paper is a report
of our results.
electronegativity of the fluorine atoms has been termed the
1
“
gauche effect”.
Because strong hydrogen bonds are formed between donor
and acceptor groups of high electronegativity (e.g., OH and F,
or NH2 and F), where in the ethane derivatives the gauche effect
is known from the example of 1,2-difluoroethane to be present
as well, the question of the relative roles played by the gauche
effect and internal hydrogen bonding presents itself. Because
the gauche effect appears to operate only when the ligands are
separated by three bonds, as in the 1,2-disubstituted ethanes,
one way to investigate the problem is to study the composition
Experimental Section
The sample of 4-FB was prepared at Portland State University
by the transesterification of 4-fluorobutyl acetate, which in turn
had been made from 4-bromobutyl acetate. The NMR spectrum
of the 4-FB sample obtained just after its preparation and again
from the residue after the diffraction experiments showed the
purity of the sample to be very high.
7
*
Corresponding author.
Permanent address: Department of Chemistry, Norwegian University
4-Fluorobutyl Acetate. 4-Bromobutyl acetate (36.12 g, 0.185
mol) was treated with KF (26 g, 0.45 fw, spray dried) at 160
°C in an evacuated 300 mL Carius tube. After 2 weeks of
continuous reaction, a sample was withdrawn and GC-MS
†
of Science and Technology, N-7491 Trondheim, Norway.
‡
§
Oregon State University.
Portland State University.
1
0.1021/jp012392h CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/22/2001

9588 J. Phys. Chem. A, Vol. 105, No. 41, 2001
Trætteberg et al.
Figure 2. Intensity (sI
m
(s)) curves. Experimental curves are aver-
ages of data from three plates at the long camera distance and four
plates at the middle distance. The theoretical curve is for the preferred
model A.
projector slide plates (medium contrast) developed for 10 min
in a D19 developer diluted 1:1. The temperature of the nozzle
tip was 83 °C during the experiments. The nominal accelerating
voltage was 60 kV (electron wavelength, 0.049 Å) calibrated
against CO2 in separate experiments [ra(CdO) ) 1.1646 Å;
ra(O‚‚‚O) ) 2.3244 Å]. The nominal nozzle-to-plate distances
were 750 (long camera, LC) and 300 mm (middle camera, MC).
Three out of four plates were selected for analysis from the
LC/MC experiments. The procedures for obtaining the total scat-
Figure 1. Diagrams of some of the more important conformers of
-fluorobutan-1-ol. These conformers comprise the preferred model A.
4
analysis showed a conversion of ∼60%; the product was
removed by repeated washing with CH2Cl2, and after distillative
removal of the solvent, the partially fluorinated product was
again subjected to fluorination with 16.20 g of KF. After another
4
tered intensities (s It(s)) and the molecular intensities (sIm(s))
10,11
have been described.
To reduce the level of noise in the
2
weeks, the contents of the Carius tube was diluted with 200
data, all of the plates were traced several times. The background
for each trace was then subtracted, and the resulting molecular
intensities were averaged to yield a data set for each plate. These
data sets were averaged in turn to give average intensities for
mL of light petroleum ether and then stirred with 14.2 g of
finely ground CaCl2 for 2 weeks. This was done to remove an
impurity (Rt ) 6.17 min, 25 m DB5 column, 50 f 200 °C at
-1
1
0 °C min ) which was surmised to be 4-hydroxybutyl acetate
each of the two camera distances. Ranges of these average
by GC-MS analysis. After this contaminant had been largely
removed by the above treatment, the liquor was filtered off,
the petroleum ether was distilled at atmospheric pressure, and
the residue was subjected to simple distillation at 29-32 Torr,
-1
modified molecular intensities (sIm(s)) were 2.00 e s/Å
1
0
e
)
-1
-1
6.00 (LC) and 8.00 e s/Å e 40.00 (MC), with ∆s/Å
.25. The experimental intensity curves are shown in Figure 2
and the corresponding radial distribution (RD) curves in Figure
. The complex scattering factors used in these and other
7
1-79 °C, giving a purity of >95% by GC-MS (13.82 g, 56%
3
8
of theory; bp 56-57 °C at 12 Torr).
-Fluorobutan-1-ol. A total of 8.10 g of 4-fluorobutyl
12
calculations were taken from tables. The average intensity data
from each of the camera distances is available as Supporting
Information.
4
acetate, 40 mL of absolute ethanol, and 4 drops of concentrated
H2SO4 were refluxed for 9 h; because the composition had
changed little after 4 h, the solvent was largely distilled off (33
mL), and the residue was refluxed for another 4.5 h with 28
mL of fresh ethanol. At that time, GC-MS analysis showed
Theoretical Calculations
Ab Initio Optimizations. If one assumes 3-fold potentials
for rotation about the C-O and C-C bonds in 4-FB, there are
81 possible conformations for the molecule. However, those
conformations, differing only by rotations around the C-O bond
(i.e., essentially in the position of the hydroxyl hydrogen atom),
are not distinguishable by gas-phase electron diffraction (GED).
Moreover, mirror image structures give identical scattering
patterns in a GED experiment. These circumstances reduce the
number of distinguishable conformers to 14, equal to (81/3 -
1)/2 + 1, where the 1 refers to the unique extended form. These
conformers have similar bond lengths and valence angles which
are not separately measurable by GED. The average values are
4
-FB to be the sole product. After cooling, solid NaHCO3 was
added (≈100 mg), the mixture was stirred vigorously for 1 h
and filtered, and ethyl acetate was distilled away through a 12
cm Vigreux column. The residue was then vacuum distilled with
the same apparatus, and a clear, colorless liquid (2.4 g, 43%),
which was free of any contaminants according to GC-MS
analysis, was collected at 66.5-69 °C (22-26 Torr); the
7
reported boiling points are 57.5-58 °C at 15 Torr and 56 °C
9
at 15 Torr.
Electron Diffraction. The diffraction experiments were made
3
with the Oregon State apparatus using an r sector and Kodak

Hydrogen Bonding in Gaseous 4-Fluorobutan-1-ol
J. Phys. Chem. A, Vol. 105, No. 41, 2001 9589
TABLE 1: Theoretical Free-Energy Differences (kcal mol-1)
and Estimated Conformational Composition for
4-Fluorobutan-1-ol
HF/6-31G* MP2/6-31G* B3LYP/6-31G*
conformer
∆G°
%
∆G°
0.00 35.3 0.00
0.22 26 0.84
%
∆G°
%
FB1a G G G g
-
+
-
+
+
0.67
0.86
0.00 24.5
0.89 7.1
0.62 10.2
0.71
0.70
1.08
1.20
0.66
1.69
1.75
1.72
1.94
9.5
7.3
49.1
15
FB2a G G G g
+
-
-
+
-
FB3
FB4
FB5
FB6
FB7
FB8
FB9
G AG a
GGGa
0.75 12.3 1.11
1.35 5.3 1.94
1.39 5.0 1.75
1.87 2.5 1.90
2.23 0.8 2.13
1.71 3.2 2.13
1.86 2.6 1.99
1.76 3.0 1.87
2.02 2.1 2.46
2.75 0.7 2.72
2.73 0.8 2.66
3.18 0.4 3.09
10.3
3.2
4.2
3.4
1.2
2.4
3.0
3.5
1.6
1.1
1.2
0.6
-
-
G AG a
-
G AAa
9.0
4.6
5.4
4.6
9.6
2.3
2.1
2.2
1.6
AAAa
AG G a
+
-
+
-
G G Aa
-
FB10 AAG a
+
+
-
FB11 G G G a
-
-
FB12 AG G a
FB13 GGAa
FB14 AG Aa
-
∆
G° ) G°(FBn′) - G°(FBn)
-
+
- -
FB1′
FB2′
FB4′
FB7′
G G G g
4.24
0.52
0.08
0.27
4.75
1.14
-0.02
0.07
+
-
-
G G G a
GGGg
AAAg
a
Hydrogen-bonded conformers.
ing the conformational composition of the mixture. We had
hoped to be able to form a rough estimate of the magnitude of
the stabilization energy due to the hydrogen bonding by
optimizations of the conformations having the FB1 and FB2
skeletal arrangements but with orientations of the O-H bonds
incompatible with hydrogen-bond formation. These forms are
labeled FB1′ and FB2′ in Table 1. The results from the MP2
calculations seem to indicate that the hydrogen-bond stabiliza-
tion of the two conformers amounts to about 4.8 and 1.1 kcal
Figure 3. Radial distribution curves. Theoretical curves are for the
models of Tables 5 and 6 and in each case show contributions from
the groups of hydrogen-bonded species and non-hydrogen-bonded
species. The vertical bars indicate the positions and the relative weights
of the distances within each group.
measurable with high accuracy, however. In such cases, useful
information about the deviations from the average bond-length,
valence-angle, and torsion-angle parameters for each of the
conformers can be obtained from ab initio calculations. By
introducing the calculated parameter differences, one obtains a
more realistic model for refinement and presumably more
accurate structural results. The calculated energies of the
conformers also give an indication about the composition of
the conformational mixture.
-1
mol , respectively.
The HF/6-31G* energies indicate that the two hydrogen-
bonded conformers together comprise about 17% of the con-
formational mixture. The situation is decidedly different ac-
cording to the results of higher level calculations (Table 1). For
the MP2/6-31G* optimizations of the same set of conformers,
the hydrogen-bonded pair FB1 and FB2 dominate the mixture
to the extent of 61%; for the B3LYP/6-31G* optimizations, the
value is 64%. In the B3LYP/6-31G* calculations, but not the
MP2/6-31G*, the ratio of the amounts of the two hydrogen-
bonded forms is also changed substantially in favor of FB1.
Normal Coordinate Analyses. To obtain estimates of
vibrational amplitudes and possibly other corrections involved
in the formulation of a 4-FB model for refinement, it was
necessary to carry out normal coordinate analyses. This was
Full geometry optimizations at the ab initio HF, MP2, and
B3LYP levels of theory with the 6-31G* basis set were carried
out for the 14 conformers listed in Table 1 using the program
Gaussian98.13 In the following symbolism for individual con-
formers, capital letters indicate (A) anti or (G) gauche tor-
sions of the heavy atom skeleton given in the order -C3-C4-,
14
-C2-C3-, and -C1-C2-, and the small letter (a or g)
done with the program ASYM40 in a new version that permits
the symmetrization of the Cartesian force constants obtained
from ab initio calculations. The force constants for each con-
former were taken from the results of the HF/6-31G* calcula-
tions. Theoretical vibrational amplitudes were calculated for all
of the distances, except for the H‚‚‚H in all of the conformers
included in the model of the system to be refined. Values for
the most prominent distances in conformers comprising one of
the models are seen in Table 2. The force fields are of no special
significance, and no further use was made of them.
symbolizes torsion about the -C1-O bond showing the
orientation of the O-H bond. The sense of the rotation is
positive for the counterclockwise motion of the near bond
looking along the connecting bond to the far bond; the zero of
rotation is with the near and far bonds eclipsed. For these
calculations, the orientation of the O-H bond was in all of the
cases but two taken as anti on the assumption that the energy
+
of this conformer was lower than that of the corresponding g
-
or g forms, an assumption generally consistent with the results
for FB4′ and FB7′ versus those for FB4 and FB7 (seen in Table
Structure Analysis. Model Definitions. Because it is impos-
sible to measure the structures of the complex 4-FB system from
GED data alone, it was necessary to introduce a number of
assumptions and constraints to reduce the large number of
structural parameters to a manageable level for refinement.
(Because these assumptions involved distances and angles, they
were drawn from the theoretical results at the HF/6-31G* level,
1
). The two exceptions are FB1 and FB2, each representing a
molecule with internal hydrogen bonding where the hydrogen
atom is pointed toward the fluorine as seen in Figure 1. For
these conformers, the calculated energies are substantially lower
than those of the other conformers, indicating that the formation
of O-H‚‚‚F hydrogen bonds is an important factor in determin-

9590 J. Phys. Chem. A, Vol. 105, No. 41, 2001
Trætteberg et al.
a
TABLE 2: Calculated (Upper) and Experimental (Lower)
bond distances of the conformers in both groups were tied to
those of FB1 via the theoretical differences. The C-H and O-H
bond lengths in all of the conformers were assumed equal. (4)
The C-C-O, C1-C2-C3, C2-C3-C4, and C-C-F bond
angles in FB2 were constrained to be smaller than those in FB1
by the theoretical amounts and fixed at the theoretical values
for the other conformers. The difference between C1-C2-C3
and C2-C3-C4 in FB1 was constrained to the theoretical value.
All of the C-C-H angles were tied to the FB1 value through
the theoretical differences, and all of the C-O-H bond angles
were assumed equal.
Average Root-Mean-Square Amplitudes (RMS, l/Å) for
Some Important Distances in Conformers of
b,c
4-Fluorobutan-1-ol
conformer
FB3
FB1
FB2
FB5
FB6
FB7
-
+
-
+
+
-
-
+
+
-
-
-
-
G G G g G G G g G AG a G AG a G AAa AAAa
C-H
0.077
0.077
0.081
0.052
0.057
0.074
0.078
0.082
0.050
0.056
0.075
0.078
0.082
0.051
0.056
0.075
0.078
0.082
0.050
0.056
0.078
0.078
0.082
0.050
0.056
0.078
0.084
0.107
A
B
C
D
E
E
E
E
E
E
A
B
A
B
A
B
A
B
C
A
B
C
0
.081
0.051
.057
0.074
.080
0.108
.092
0.142
.155
0.127
.139
0.127
.139
0.198
.210
0.211
.223
0.144
.156
C1-C2
C1‚‚‚C3
C‚‚‚H10
C1‚‚‚C4
C2‚‚‚F
C3‚‚‚O
C1‚‚‚F
C4‚‚‚O
F‚‚‚O
0
0
0.079C
0.081C
0.081C 0.084
The results of these simplifications were several. First, the n
- 1 mole fraction parameters were reduced to just one, which
permitted an experimental evaluation of the relative amounts
of H-bonded and non-H-bonded forms in the system. Second,
the very large number of structural parameters was reduced to
0.108
0.107
0.107
0.107
0.092D
0.091D
0.091D
0.091D
0.091D
0
0.123
0.135
0.123
0.135
0.075
0.073
0.076
0.074
0.077
0.075
0.078
0.076
0.071
0.069
0.072
0.070
0.096
E
F
F
F
F
0
0.147
0.159
0.150
0.162
0.149
0.161
E
E
E
E
F
0
18: five bond lengths, five bond angles, and eight torsion angles.
0.132
0.145
0.146
0.159
0.150
0.072
0.070
0.158
E
E
E
E
E
E
F
F
The first 14 of the following set of parameters are designed to
apply to the reference conformer FB1 and the last four to ap-
ply to FB2: r(C-O), r(C-C), r(C-F), r(O-H), r(C-H),
0
0.144
0.153
0.165
0.162
0.174
0.210
0.223
0.157
0.155
0
0.164G
0.154
0.159G 0.167G 0.375H
0.149
0.096
0.099
∠
∠
(C-O-H), ∠ (C-C-O), ∠ (C1-C2-C3), ∠ (C2-C1-H8),
(C-C-F), ∠ (C-C-C-F), ∠ (C-C-C-C), ∠ (C-C-C-O),
0
0.163G
0.068
0.158G 0.375H 0.364H
0.284
0.204
0.085
0
0.417H [0.284] 0.448H [0.104]
∠ (C-C-O-H), ∠ (C-C-C-F)FB2, ∠ (C-C-C-C)FB2,
(C-C-C-O)FB2, ∠ (C-C-O-H)FB2. If all of the H‚‚‚H
∠
a
b
Force field from HF/6-31G* theory. Quantities in square brackets
were assumed. Same upper case superscript letters indicate experi-
mental amplitudes refined in groups with 2σ uncertainties (least
significant figures of items) as follows: A ) 5, B ) 3, C ) 8, D ) 13,
E ) 25, F ) 27, G ) 95, H ) 585.
c
distances are omitted, there are 62 distances in each 4-FB
conformer leading to a total of 620 different distances in a 10-
conformer-system model, for each of which there is an associ-
ated vibrational amplitude parameter.
We tested several types of trial models of which two were
TABLE 3: Theoretical (HF/6-31G*) Values of Nonbond
Heavy-Atom Distances (Å) in Conformers of
investigated in depth. All of the models were based on ra
1
6
4
-Fluorobutan-1-ol
distances. In all of the models, attempts at the accurate
determination of the dihedral angles in both the hydrogen-
bonded and non-hydrogen-bonded groups of the conformers
failed, and we chose to fix them at the theoretical values shown
in Table 4 for the HF/6-31G* calculations. (The refined values
of these angles were actually not much different from the
calculated ones, but they had very large standard deviations.
Fortunately, the values of the other parameters were only
marginally influenced by reasonable changes in the dihedral
angles.)
conformer C2‚‚‚F5 C3‚‚‚O6 C1‚‚‚O4 C1‚‚‚F5 C4‚‚‚O6 F4‚‚‚O6
a
-
+
-
+
FB1 G G G g
3.06
3.00
2.93
2.91
2.93
2.94
3.75
3.77
3.07
3.75
2.97
3.77
2.92
3.77
3.18
2.91
2.95
2.91
2.95
3.77
3.77
3.07
3.79
3.02
3.02
2.93
3.75
3.78
3.09
2.97
3.90
3.08
3.91
3.90
3.89
3.24
3.21
3.90
3.34
3.17
3.16
3.19
2.89
2.70
4.33
3.46
4.35
4.33
4.92
4.53
2.99
4.94
3.97
4.45
3.63
4.46
3.00
3.37
4.34
3.47
4.35
4.94
4.92
3.01
4.52
4.41
3.10
3.67
4.45
4.47
2.49
2.72
5.14
4.29
4.61
5.17
6.11
4.35
4.37
5.24
3.72
4.69
4.67
5.79
a
+
-
-
+
FB2 G G G g
+
-
FB3 G AG a
FB4 GGGa
FB5 G AG a
FB6 G AAa
FB7 AAAa
FB8 AG G a
FB9 G G Aa
FB10 AAG a
FB11 G G G a
-
-
-
+
-
+
-
-
In the first type of model (model B), the relative amounts of
the conformers within both the hydrogen-bonded (FB1 and FB2)
and the non-hydrogen-bonded (FB3-FB10) groups were kept
at the values determined by the MP2/6-31G* calculations; these
relative amounts are seen in Figure 4. The bond distances, the
heavy-atom bond angles, and the C-C-H angles in both groups
were tied to those of FB1 by the theoretical differences; the
C-O-H bond angle was kept at the theoretical value. The fit
to the experiment given by model B is certainly not bad, but it
has a few small, disquieting features which can be seen in its
RD-difference curve in Figure 3. An example is the small peak
in this curve at about 3.8 Å, which implies an insufficient
contribution from distances near this value, presumably those
between heavy atoms. As is seen from Figure 4, the hydrogen-
bonded group has no distance near 3.8 Å, whereas all of the
+
+
-
-
-
FB12 AG G a
FB13 GGAa
FB14 AG Aa
-
a
Hydrogen bonded.
but when they involved conformational composition, they were
taken from the MP2/6-31G* results.)15 The constraints were the
following: (1) As in the cases of butane-1,4-diol and 1,4-
difluorobutane already cited,1,2 the conformers were divided into
two groups, one of which comprised the (potentially) hydrogen-
bonded pair FB1 and FB2, and the other various conformers
were chosen from the non-hydrogen-bonded group FB3-FB10,
depending on the model of interest. In all of the cases, the forms
FB11-FB14 were omitted on grounds that together they are
predicted to total only a small proportion of the mixture (about
+
+
+
conformers except the G G G a in the non-hydrogen-bonded
group have at least one such distance. Figure 4 also shows that
the 3.8 Å peak is due to X(1)‚‚‚Y(4) skeletal distances arising
from anti dihedral angles.
8
% from HF and less than 5% for MP2 and B3LYP) and that,
in any event, their distance distributions are similar (Table 3)
to those of other species that are already included. (2) For most
models, the relatiVe amounts of the conformers in both the
hydrogen-bonded and non-hydrogen-bonded groups were pegged
at values obtained from an assumed Boltzmann distribution
based on the theoretical energies, but variations of these ratios
were also tested. (3) The values of the C-O, C-C, and C-F
The problems with model B led us to formulate a second
type of model (model A) intended to remedy the defects of the
first by inclusion of more of these forms than is suggested by
theory. Conformers FB3, FB5, FB8, and FB9 have one anti
dihedral skeletal angle (Figure 4), FB6 and FB10 have two such

Hydrogen Bonding in Gaseous 4-Fluorobutan-1-ol
J. Phys. Chem. A, Vol. 105, No. 41, 2001 9591
TABLE 4: Theoretical Values of Bond Lengths (Å) and Valence and Dihedral Angles (deg) in the Conformers of
a
4
-Fluorobutan-1-ol
r(C
1
-C
6
)
r(C
1
-C
2
)
r(C
2
-C
3
)
r(C
3
sC
4
)
r(C
4 5
sF )
conformer
HF
MP2 DFT
HF
MP2
DFT
HF
MP2 DFT
HF
MP2 DFT
HF MP2 DFT
-
+
-
FB1
G G G g
1.340 1.425 1.418 1.529 1.526 1.536 1.535 1.531 1.539 1.516 1.513 1.520 1.386 1.418 1.413
1.401 1.425 1.420 1.529 1.528 1.538 1.534 1.530 1.539 1.515 1.512 1.519 1.384 1.416 1.410
1.405 1.430 1.426 1.518 1.515 1.522 1.529 1.525 1.533 1.514 1.511 1.518 1.376 1.404 1.397
1.406 1.432 1.427 1.520 1.518 1.525 1.532 1.530 1.537 1.515 1.512 1.519 1.377 1.406 1.399
1.404 1.429 1.425 1.519 1.516 1.524 1.529 1.525 1.533 1.515 1.512 1.519 1.374 1.402 1.396
1.403 1.428 1.423 1.518 1.516 1.523 1.529 1.526 1.532 1.515 1.513 1.520 1.373 1.402 1.395
1.403 1.428 1.423 1.518 1.516 1.524 1.529 1.527 1.533 1.519 1.513 1.520 1.372 1.401 1.393
1.406 1.431 1.427 1.520 1.51 7 1.525 1.533 1.530 1.537 1.518 1.517 1.523 1.377 1.406 1.398
1.406 1.431 1.426 1.521 1.519 1.526 1.532 1.528 1.536 1.515 1.513 1.520 1.376 1.405 1.399
1.403 1.429 1.424 1.525 1.524 1.532 1.529 1.526 1.533 1.515 1.513 1.519 1.372 1.400 1.393
+
-
- +
FB2
FB3
FB4
FB5
FB6
FB7
FB8
FB9
G G G g
+
-
G AG a
GGGa
-
-
G AG a
-
G AAa
AAAa
+
-
AG G a
+
-
G G Aa
-
FB10 AAG a
∠
(O
6
-C
1
-C
2
)
∠ (C
1
-C
2
-C
3
)
∠ (C
2
-C
3
-C
4
)
∠ (C
3
sC
4
sF
5
)
conformer
HF
MP2
DFT
HF
MP2
DFT
HF
MP2
DFT
HF
MP2
DFT
-
+
-
FB1
FB2
FB3
FB4
FB5
FB6
FB7
FB8
FB9
FB10
G G G g
113.7
112.9
108.6
108.8
108.7
108.0
107.9
109.0
107.1
113.0
113.0
112.2
107.4
107.6
107.4
107.2
107.1
107.8
106.2
112.6
113.9
113.0
108.2
108.5
108.3
107.9
107.8
108.8
107.0
113.3
116.8
115.3
113.1
114.7
113.2
112.6
112.6
115.6
115.1
113.0
116.1
114.0
112.5
113.4
112.7
112.4
112.3
115.0
114.5
112.4
116.6
114.7
113.2
114.4
113.4
112.9
112.7
115.6
114.9
113.1
116.6
115.5
112.7
114.3
112.8
112.9
112.3
115.3
115.4
112.3
116.0
115.0
112.2
113.0
112.3
112.3
111.9
114.9
114.7
112.2
116.8
115.7
112.8
113.9
112.9
112.9
112.4
115.1
115.3
112.7
110.4
110.1
110.0
110.1
110.0
109.9
109.5
108.3
110.5
109.6
109.9
110.0
109.4
109.6
109.4
109.2
109.4
108.0
109.9
109.5
110.8
110.5
110.0
110.2
110.1
109.9
109.8
108.6
110.7
109.9
+
-
- +
G G G g
+
-
G AG a
GGGa
-
-
G AG a
-
G AAa
AAAa
+
-
AG G a
+
-
G G Aa
-
AAG a
∠
(C
2
-C
3
-C
4
-F)
DFT
∠ (C
HF
63.2
1
-C
2
-C
3
-C
4
)
∠ (O
HF
6
-C
1
-C
2
-C
3
)
∠ (C
HF
75.7
81.2
180.2
178.9
179.7
178.7
180.0
179.1
2
sC
MP2
78.4
80.1
180.0
179.3
179.3
178.7
180.0
177.7
1
sO
6
sH
7
)
conformer
HF
MP2
MP2
DFT
MP2
DFT
DFT
-
+
-
FB1
FB2
FB3
FB4
FB5
FB6
FB7
FB8
FB9
G G G g
-74.0 -75.7 -73.5
58.5
63.0
-76.1
-52.0
-62.2
54.4
-61.7
179.0
180.0
-68.8
-74.6
-51.4
-61.0
53.1
-60.4
179.1
180.0
-69.4
-74.6
-49.4
-61.9
54.5
-61.8
179.2
180.0
-68.9
71.3
+
-
- +
G G G g
71.1
60.5
54.0
71.7
59.4
52.1
70.0 -55.7 -55.5 -55.2
59.6
53.4
77.8
177.7
180.4
178.4
178.8
180.0
177.0
+
-
G AG a
GGGa
180.2
55.7
181.5
178.6
180.0
68.5
180.6
51.0
180.9
178.1
180.0
65.6
180.3
54.8
179.0
177.7
180.0
68.9
-
-
G AG a
-60.1 -59.1 -59.6
-61.7 -60.6 -61.0
180.0
171.5
68.6
-
G AAa
AAAa
180.0
170.8
68.7
180.0
173.8
+
-
AG G a
+
-
G G Aa
67.4 -65.9 -62.6 -66.0 -172.6 -171.1 -173.6 -176.7 -176.1 -176.3
180.0 178.8 178.0 177.4 -60.4 -58.5 -59.1 180.0 176.8 176.1
-
FB10 AAG a
179.8
180.1
a
Basis set was 6-31G* in all cases.
angles, and FB7 has three. Although it is likely that all of the
conformers are present, a study of the torsion-dependent
distances suggested that the makeup of the second model could
be simplified by including only one or two forms to represent
each of the three angle groups. Grounds for such an assumption
are that conformers with the same number of anti dihedral angles
have roughly similar distance distributions. Conformers FB3
and FB5 were taken to represent the group with one anti dihedral
angle when tests showed that the combination was better than
either one alone. FB6 and FB7 were chosen to represent the
groups with two and three anti angles, respectively. FB3 and
FB5 differ mainly in the positions of the OH and F groups that
are either on the same side (FB5) or on opposite sides (FB3) of
the plane of the carbon skeleton. Because these two func-
tional groups are too far away to interact strongly, it seemed
reasonable to incorporate FB3 and FB5 in equal amounts. The
hydrogen-bonded group of conformers in this second model
type was represented by FB1 and FB2 as before in relative
amounts determined by test refinements. As Figure 3 shows,
the fit provided by model A is considerably better than that
from model B.
Refinement Results. The parameter values obtained for model
B are found in Table 5. Models of type A are much more
flexible than the model B types, and it was first necessary to
establish the relative amounts of FB3 + FB5, FB6, and FB7 in
the non-hydrogen-bonded group and those of FB1 and FB2 in
the hydrogen-bonded group that gave the best fit. These relative
amounts were studied systematically in trial refinements. The
best results were obtained with about 75% FB1 and 25% FB2
in the hydrogen-bonded group and 20% FB3 + FB5, 50% FB6,
and 30% FB7 in the non-hydrogen-bonded group. Refinement
results for this model are listed in Table 5 as model A.
The results of our investigation of the 4-FB system are found
in Tables 5 and 6. The values of Table 5 are weighted averages
(i.e., the contributions of the conformers were weighted ac-
cording to the conformational compositions). The entries of this
table are limited to bond distances and bond angles because
the averaging of gauche and anti dihedral angles is meaningless.
The data of Table 6 deals only with the important H-bonded
conformer FB1 and is thus more detailed. Values of the
parameters in the other conformers may be deduced from those
in Table 6 by applying the differences available from the values
of Table 4. Model A is our preferred model because of the better
fit to the experimental data that it provides; this is evident from
the value of the quality-of-fit factor R (Table 5) which is signif-
icantly smaller for model A than for model B and from the
difference curves of Figure 3 which show a better match be-
tween the experiment and the theory for model A. Figure 3 also
The vibrational amplitudes in all of the refinements were
refined in groups defined largely by regions or peaks of the
RD curve in which the associated interatomic distances were
found. Most of the vibrational amplitudes within such groups
have calculated amplitudes of similar magnitude. The groups
may be identified from Tables 3 and 6.

9592 J. Phys. Chem. A, Vol. 105, No. 41, 2001
Trætteberg et al.
formers should be present only to the extent of about 17%.
Clearly, the conformational stabilization of 4-FB resulting from
O-H‚‚‚F hydrogen bonding is underestimated by the latter with
the 6-31G* basis. Although the matter has not been checked,
results from other work suggest that these predictive failures
likely are due to the small basis set rather than to the theoretical
levels.
The theoretical results for the relative amounts of each
hydrogen-bonded conformer are also interesting. All three sets
of calculations led to a lower energy for FB1 than for the
alternative FB2. The calculated energy difference between the
two forms is modest at the HF and MP2 levels of theory, but
as Table 1 shows, the B3LYP calculations predict about 3 times
as much FB1. This theoretical prediction of the FB1/FB2 ratio
at the B3LYP level received support on the experimental side
when a systematic variation of the ratio led to a best fit that
agreed with the theoretical prediction. The reason for the greater
stability of FB1 is reflected by the structures of the O-H‚‚‚F
groups: the O‚‚‚F distance in FB1 is shorter than that in FB2
(2.46 vs 2.71 Å; the van der Waals distance is about 2.75 Å),
and the O-H‚‚‚F angle is larger (151° vs 133°). The dominant
-
+
- +
role played by FB1 (i.e., the G G G g conformer) in the
hydrogen-bonded group of 4-FB is consistent with the observa-
tion that the hydrogen-bonded conformers in butane-1,4-diol2
are best represented by the same form (or its mirror image) for
the -OB(CH2)4-O-H moiety.
Bond Lengths and Vibrational Amplitudes. Table 8 shows
the values of the bond distances and bond angles for 4-FB and
the related compounds 1,4-difluorobutane and 1,4-butanediol.
The average C-C bond in the two compounds with internal
hydrogen bonding, 4-FB and 1,4-butanediol, is found to be about
Figure 4. Distribution of torsion-sensitive distances in the 10 most
abundant conformers. Distance identification: (1) r(C ‚‚‚F), (2)
r(C ‚‚‚O), (3) r(C ‚‚‚C ), (4) r(C ‚‚‚F), (5) r(C ‚‚‚O), (6) r(O‚‚‚F). For
each conformer the lengths of the vertical bars are proportional to the
weights of the distances Z /rij. Starred conformers comprise model
A; all of the conformers comprise model B (see text). The relatiVe
abundances in the hydrogen-bonded group (FB1 and FB2) and non-
hydrogen-bonded group (FB3-FB10) are listed by the conformer labels.
2
3
1
4
1
4
0
.01 Å longer than that in 1,4-difluorobutane where hydrogen
bonding is absent. The average length of the C-F bond in 4-FB,
like the average C-C bond, is again slightly greater than that
in non-hydrogen-bonded 1,4-difluorobutane. These observations
may be partly understood by an analysis of the theoretical
results. Although ab initio bond-length values are not available
for 1,4-butanediol, they have been calculated for 1,4-difluo-
i j
Z
shows the relative contributions to each model from the hydro-
gen-bonded and non-hydrogen-bonded groups. The correlation
matrix for the parameters of model A is given in Table 7.
2
robutane. It is seen that, for 4-FB itself (Table 4), the average
C-C bond length is increased mainly by contributions from
the central bond (C2-C3) in all of the non-hydrogen-bonded
conformers and by contributions from the bond nearest the OH
group (C1-C2) in the hydrogen-bonded conformers FB1 and
FB2. In 1,4-difluorobutane, the central C2-C3 bond is also the
longest (by about 0.015 Å) throughout the list of conformers,
but the C1-C2 bond is not elongated as in 4-FB. The average
C-C bond in 4-FB is thus longer than it is in 1,4-difluorobutane
because about 50% of the conformational weight accrues to the
hydrogen-bonded group of its conformers. A similar picture
applies to the C-F bonds in that the theoretical values of these
bonds in the hydrogen-bonded conformers of 4-FB are slightly
larger than those in conformers of the non-hydrogen-bonded
group, which are like those in 1,4-difluorobutane. It is not
surprising that the C-O bonds, which occur only in hydrogen-
bonded 4-FB and 1,4-butanediol, are experimentally the same
because both the structures of the molecules and the strengths
of the O-H‚‚‚O and O-H‚‚‚F hydrogen bonds are similar.
Discussion
The most important result of our study of gaseous 4-FB is
the finding of substantial amounts of conformers with internal
O-H‚‚‚F hydrogen bonds. Although there is no doubt that non-
hydrogen-bonded conformers also contribute significantly to the
conformational mixture, neither is there doubt that the hydrogen-
bonded conformers play a very important role; we were unable
to obtain a good fit to the experiment with the models that did
not contain substantial amounts of each. For a large number of
models, the refined values of the total contribution of the
hydrogen-bonded conformers was found to lie in the range of
4
8-62%. At 48(14)% in our preferred model A, the contribution
of the hydrogen-bonded forms is larger than that found for
1
butane-1,4-diol of 40(4)% at 144 °C. The difference seems to
suggest that the energy of the O-H‚‚‚F hydrogen bond may be
larger than that of the O-H‚‚‚O one, but the uncertainties on
the amounts and the host of (different) approximations applied
in each analysis make such a conclusion very risky.
The contribution of the hydrogen-bonded conformers to the
mixture comprising 4-FB, predicted at 61% from the MP2/
The experimental vibrational amplitudes (Table 3) of the bond
distances in 4-FB are in good agreement with the calculated
ones, but those for conformationally similar 1,4-butanediol were
found to be 0.01-0.02 Å larger than calculated. The differing
levels of agreement for these molecules probably reflect the
differing assumptions applied to the structural analyses. For the
1,4-butanediol study, the various bond lengths were assumed
6-31G* optimizations, is in good agreement with the experi-
mental result for our model A. This is not the case for the results
from the HF theory, which predicts that hydrogen-bonded con-

Hydrogen Bonding in Gaseous 4-Fluorobutan-1-ol
J. Phys. Chem. A, Vol. 105, No. 41, 2001 9593
TABLE 5: 4-Fluorobutan-1-ol. Some Average Bond Lengths and Bond Anglesa
bond lengths (r
model Ab
a
/Å)
model Bc
bond angles (∠
model Ab
a
/deg)
model Bc
parameter
parameter
C-O
1.430(5)
1.529(2)
1.537(2)
1.520(2)
1.401(5)
[0.947]
1.430(6)
1.528(2)
1.537(2)
1.521(2)
1.402(7)
[0.947]
C-O-H
[106.1]
[106.1]
C
1
C
2
C
3
-C
-C
-C
2
3
4
C-C-O (H-bond)
112.3(55)
108.2(45)
112.5(33)
112.4(32)
109.8(12)
111.8(45)
108.9(45)
112.9(39)
112.7(37)
111.2(59)
C-C-O (non-H-bond)
C
C
1
-C
-C
2
-C
-C
3
C-F
O-H
C-H
2
3
4
C-C-F
1.102(4)
1.103(5)
compositions and agreement factors
%
%
R
(H-bond)
(non-H-bond)
48.5
51.6
0.072
22.7
77.4
0.086
(140)
(196)
}
}
d
a
Values are weighted averages with weights according to conformational composition. Quantities in parentheses are estimated 2σ uncertainties;
those in square brackets were assumed. Preferred model containing six conformers. c Contains 10 conformers. d Goodness of fit factor. R ) [Σ
b
w
2
i
i
i
∆ /
2
1/2
Σ
i
I
i
(obsd) ] where ∆
i
) I
i
(obsd) - I
i
(calcd) with I
i
i i
) s I .
TABLE 6: 4-Fluorobutan-1-ol. Some Interatomic Distances (r/Å), Bond Angles (∠ /deg), and RMS Amplitudes of Vibration (l/Å)
for the Principal Hydrogen-Bonded Conformer, G G G g (FB1)a
-
+
- +
model Ab
model B
model Ab
model B
parametersc
r
a
; ∠
l
r
; ∠
l
parameters
r
; ∠
l
r
; ∠
l
a
a
a
a
a
a
a
C-O
1.428(5) 0.052
1.533(3) 0.057
1.540(3) 0.057 (3)
1.520(3) 0.056
1.408(6) 0.052
1.425(6) 0.035
1.535(3) 0.058
1.541(3) 0.058 (4)
1.522(3) 0.057
1.410(7) 0.053
C‚‚‚H
2.14-2.20(20) 0.090(13)
2.15-2.22(23) 0.090(18)
C
1
C
2
C
3
-C
-C
-C
2
3
4
C
C
C
C
C
C
C
C
C
1
2
3
2
1
2
3
1
4
‚‚‚C
‚‚‚C
‚‚‚F
3
2.556(12)
2.542(12)
2.397(27)
2.463(29)
3.083(29)
3.067(33)
3.159(31)
2.882(51)
2.974(43)
2.463(49)
0.080
0.080
0.076
0.077
0.155
0.139
0.139
0.210
0.223
0.156
2.535(64)
0.074
0.074
0.070
0.070
0.187
0.172
0.171
0.242
0.255
0.189
4
2.521(64)
(
8)
(8)
2.488(115)
2.457(173)
3.026(164)
3.146(111)
3.126(193)
2.925(188)
2.889(277)
2.416(238)
}
}
}
}
C-F
O-H
C-H
C-O-H
C-C-O
‚‚‚O
[0.947]
1.102(4) 0.081
[106.1]
0.072
[0.947]
1.103(5) 0.081
[106.1]
0.072
‚‚‚C
‚‚‚F
‚‚‚O
‚‚‚F
‚‚‚O
4
} (5)
} (5)
(25)
(39)
112.5(20)
112.5(8)
109.5(16)
109.8(18)
112.1(120)
111.0(43)
110.5(18)
116.1(85)
C
C
1
-C
-C
2
-C
-H
3
2
1
8
O‚‚‚F
C-C-F
a
Quantities in parentheses are estimated 2σ uncertainties; those in square brackets were assumed. All dihedral angles are set equal to those from
optimized HF/6-31G* calculations (Table 4). Preferred model. c For atom numbering, see Figure 1.
b
a
TABLE 7: Correlation Matrix (×100) for Parameters of 4-Fluorobutan-1-ol
b
1
00σLS
r
1
r
2
r
3
r
4
∠
5
∠
6
∠
7
∠
8
l
9
l
10
l
11
l
12
l
13
l
14
l
15
l
16
R
H bond
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
r(C-O)
r(C -C
r(C-F)
r(C-H)
0.17
100
1
2
)
0.07 -14 100
0.19 -76 42 100
0.14
71.0
27.7
57.1
64.1
0.13
0.09 -31
0.27
0.86
0.43
0.95
3.35
20.7
5.0
-4
-17
2
-6
5
5
16
-16
1
16
19
7
3
25
-7
11 -13
<1
8
100
3
∠ C-C-O
100
∠ C
∠ C
1
-C
-C
2
-C
-H
3
14 -37 100
2
1
8
12
33 100
∠ C-C-F
15 -38
-1
11
-1 -28 -32 -48
-5 -11
19
5
7
-5
-1 100
4
7
l(C-H)
<1
-3
6
6
-1 100
1
1
1
1
1
1
1
1
l(C-C)
30
-1
12
3
13 100
l(C
l(C
1
‚‚‚C
‚‚‚C
3
)
)
-1
2
3
-8
-6
-1
3
49
4
2
5
10 100
<1 -33 100
8
10
8
1
4
-18 -12
1
30 -14
l(C‚‚‚H)
12
31
15
3
-1
25
14
3
-4 -26 -85
13
54 -26 100
<1 -25
l(C
l(C
1
‚‚‚C
‚‚‚F)
4
)c
28
14
4
-8
9
-3
6
4
20
-2
23 <1
3
100
8
1
c
1
10
2
3
-1
-9 -15 -15
100
17 100
c
l(F‚‚‚O)
R
<1
<1
1
3
-2
42
2
d
H bond
12 -49 -39 -11 -39
-11 -38 <1 -13
<1 -64 -36 -7
100
a
For preferred model A, Table 6. Standard deviations from least-squares refinement. Distances (Å), angles (deg). c Conformers containing
b
d
A-type distances. Mole fraction of H-bonded conformers (FB1 + FB2).
to be the same in all of the conformers, while in 4-FB, the bond-
length differences from ab initio calculations have been
incorporated into the GED model. In any case, the accuracy of
the measured amplitudes must be viewed with caution because
the results depend heavily on the many assumptions made to
allow their refinement. These include the amplitude differences
imposed on the members of the amplitude groups and the
arbitrary makeup of the groups themselves.
the RD peak at 5.1 Å are in satisfactory agreement with the
theoretical prediction, but the experimental amplitudes for
distances under this peak are nominally much larger. The
distances contributing to the area of the RD peak at about 5.1
Å from several conformers are the following: FB3, F5‚‚‚O; FB6,
C4‚‚‚O and F5‚‚‚O; FB7, C1‚‚‚F and C4‚‚‚O. When the theoretical
amplitudes for these distances, which range between 0.0676 and
0.2043 Å, are used in the calculations of theoretical RD curves,
the peak at 5.1 Å is considerably sharper than that seen
experimentally. The large experimental amplitudes for the
Matters are only slightly different for the amplitudes of the
nonbond distances. All of the values except those connected to

9594 J. Phys. Chem. A, Vol. 105, No. 41, 2001
Trætteberg et al.
TABLE 8: Bond Lengths (Å) and Bond Angles (deg) for
University Computer Center. We are grateful to Lise Hedberg
for help with the normal coordinate calculations.
4
-Fluorobutan-1-ol and Some Related Compoundsa,b
parameter 1,4-difluorobutane 4-fluorobutan-1-ol 1,4-butanediol
Supporting Information Available: Table of the averaged
experimental molecular intensities from each camera distance.
This material is available free of charge via the Internet at http://
pubs.acs.org.
r(C-H)
1.106(3)
1.516
1.108(4)
1.531
1.116(12)
1.531
c
r(C
r(C
r(C
r(C-F)
r(C-O)
1
2
3
-C
-C
-C
2
3
4
)
)
)
b
b
1.532 (2)
1.539 (3)
1.531 (2)
}
}
}
1.516
1.398(2)
C
1.522
1.40110(5)
1.432(5)
1.531
C
1.429(2)
110.9(10)
C
114.3(22)
1.530(2)
1
References and Notes
∠
∠
〈
〈
C-C-O
C-C-F
C
110.1(43)
109.7(12)
112.4(31)
1.531(3)
this work
110.1(19)
113.8(22)
1.521(2)
(1) Wolfe, S. Acc. Chem. Res. 1972, 5, 102.
∠ C-C-C〉
r(C-C)〉
(2) Trætteberg, M.; Hedberg, K. J. Am. Chem. Soc. 1994, 116, 1382.
(3) Krosley, K.; Hagen, K.; Hedberg, K. J. Mol. Struct. 1995, 352/
53, 87
d
d
3
ref
2
(4) Kazerouni, M. R.; Hedberg, L.; Hedberg, K. J. Am. Chem. Soc.
a
. b The
Values are conformational averages. Distances, r
g
; angles, ∠
g
1997, 119, 8324.
magnitudes of the C-C distances were held at the theoretical
(5) Huang, J.; Hedberg, K. J. Am. Chem. Soc. 1989, 111, 6909; 1973,
95, 8263.
c
d
differences. Distances were assumedequal. Uncertainties do not
reflect the error in calculated differences.
(6) Friesen, D.; Hedberg, K. J. Am. Chem. Soc. 1980, 102, 3987.
Fernholt, L.; Kveseth, K. Acta Chem. Scand. 1980, A34, 163. van Schaick,
E. J. M.; Geise, H. J.; Mijlhoff, F. C.; Renes, G. J. Mol. Struct. 1973, 16,
23. Brunvoll, J. Thesis, Norges Tekniske Høgskole, Trondheim, Norway,
distances in this region that are required to give good agreement
are not easily understood, but we believe they are due to the
inadequacies in our models of the system structure. Fortunately,
the geometrical parameter values of the models were not
appreciably influenced by the changes in these amplitude values,
but the use of refined values for distances in the 5.1 Å region
led to a change, still not significant as judged by the associated
uncertainties, in the percentage of hydrogen-bonded conformers
from 58.3 to 48.5%.
1
962.
(7) Kovalev, G.; Matveeva, E. D.; Stan, V. V.; Vovk, G. G.; Yudin,
L. G.; Kost, A. N. Zh. Org. Khim. 1980, 16, 1728.
(8) Pattison, F. L. M.; Howell, W. C.; McNamara, A. C.; Schneider,
J. C.; Walker, J. F. J. Org. Chem. 1956, 21, 739.
(
9) Peterson, P. E.; Coffey, J. C. J. Am. Chem. Soc. 1971, 93, 5208.
(
(
10) Gundersen, G.; Hedberg, K. J. Chem. Phys. 1969, 51, 2500.
11) Hedberg, L. Abstracts, 5th Austin Symposium on Gas-Phase
Molecular Structure, Austin, TX, 1974; p 37.
12) Ross, A. W.; Fink, M.; Hilderbrandt, R. International Tables of
Crystallography; Kluwer Academic Publishers: Dordrecht, The Netherlands,
(
Bond Angles. As Table 8 shows, the experimental value of
the average C-C-C angle in 4-FB is about the same as it is in
1
992; Vol. 4, p 245.
(13) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
1,4-difluorobutane, where hydrogen bonding does not exist, and
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.;
Ochterski, J.; Peterson, 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.; Baboul, A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Gomperts, R. L.; Martin, D. J.; Fox, T.; Keith, M. A.;
Al-Laham, C. Y.; Peng, A.; Nanayakkara, C.; Challocombe, M.; Gill, P.
M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.;
Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian98, Revision A.9;
Gaussian, Inc.: Pittsburgh, PA, 1998.
somewhat smaller than that in butane-1,4-diol, where it does.
These comparative values are mildly surprising. It seems likely
that the reason the 4-FB value is smaller than that in butane-
1
,4-diol lies in the structural constraints applied to our model
A, where the angles in the non-hydrogen-bonded conformers
were fixed at theoretical values instead of being tied to those
in conformer FB1. However, any error in this angle value would
likely lie within the range of the attached uncertainty.
(
14) Hedberg, L.; Mills, I. M. J. Mol. Spectrosc. 2000, 203, 82.
Acknowledgment. Support for this work was provided by
the National Science Foundation under Grants CHE95-23581
and CHE99-87359 to Oregon State University and Grant
CHE96-32815 to Portland State University and by the Petroleum
Research Fund under Grant ACS-PRF 31099-AC1 to Portland
State University. Other support in the form of large grants of
computer time was given by the Research Council of Norway
(15) This arrangement was one of timing and convenience. Much of
the work in setting up the models was completed before it was decided to
carry out the calculations at higher theoretical levels. In any case, the bond-
length and bond-angle differences are not much different for calculations
at different theoretical levels.
(16) Because of the complexity of the system and the many approxima-
tions being applied, it was deemed of no consequence whether the geometry
of the conformers was calculated on the basis of ra or rR parameters; it was
therefore decided to use the ra approach, ignoring corrections for shrinkage
effects.
(Program for Supercomputing, TRU) and by the Oregon State