A density functional, infrared linear dichroism, and normal coordinate study of phenol and its deuterated derivatives: Revised interpretation of the vibrational spectra

Article abstract of DOI:10.1021/jp972585y

The assignment of the vibrational spectra of phenol has been reexamined on the basis of Raman and new IR measurements and theoretical analysis of the normal modes of vibrations in the electronic ground state. The infrared spectra of C₆H₅OH, C₆D₅OD, and C₆D₅OH have been studied in solution and vapor phases, as well as has the Raman spectra in solutions. New experimental data were obtained from infrared linear dichroism (IR-LD) studies of phenol aligned in uniaxially oriented nematic liquid crystal solution. The measured dichroic ratios and orientation factors indicate an effective C_s symmetry of the molecule with coplanar orientation of OH bond with the benzene ring and supply unique information on the extent of symmetry lowering of benzene normal modes. The fundamental vibrational frequencies, force constants, and dipole derivatives have been calculated by ab initio quantum chemical methods applying the B3P86 density functional approximation with 6-311G** basis set. The force field optimized by means of a least-squares scaling procedure for phenol-d₀ (using six scale factors) was used to calculate the frequencies (with a mean deviation from the observed values less than 1%), normal modes, potential energy distributions, transition moment vectors, and IR intensities for phenol-d₀, -d₁, -d₅, and -d₆ isotopomers. Compared to the deviations between the calculated and observed absorption intensities, a more satisfactory correlation was found between the calculated and experimentally determined vibrational transition moment directions. The results indicate unanimously that the perturbation of the normal modes of benzene by the asymmetric hydroxyl substituent is so great that the previous practice of assigning the normal vibrations of phenol to those of benzene or even to C_2v symmetry species is not justified.

Full text of DOI:10.1021/jp972585y

J. Phys. Chem. A 1998, 102, 1371-1380
1371
A Density Functional, Infrared Linear Dichroism, and Normal Coordinate Study of Phenol
and its Deuterated Derivatives: Revised Interpretation of the Vibrational Spectra
G a´ bor Keresztury*,†
Central Research Institute for Chemistry, Hungarian Academy of Sciences,
H-1525 Budapest, P.O. Box 17, Hungary
Ferenc Billes and Mikl o´ s Kubinyi§
‡
Department of Physical Chemistry, Technical UniVersity of Budapest,
H-1521 Budapest, Budafoki u´ t 8, Hungary
Tom Sundius|
Department of Physics, UniVersity of Helsinki, P.O. Box 9, FIN-00014, Helsinki, Finland
ReceiVed: August 7, 1997; In Final Form: NoVember 6, 1997
The assignment of the vibrational spectra of phenol has been reexamined on the basis of Raman and new IR
measurements and theoretical analysis of the normal modes of vibrations in the electronic ground state. The
infrared spectra of C
as has the Raman spectra in solutions. New experimental data were obtained from infrared linear dichroism
IR-LD) studies of phenol aligned in uniaxially oriented nematic liquid crystal solution. The measured dichroic
ratios and orientation factors indicate an effective C symmetry of the molecule with coplanar orientation of
6 5 6 5 6 5
H OH, C D OD, and C D OH have been studied in solution and vapor phases, as well
(
s
OH bond with the benzene ring and supply unique information on the extent of symmetry lowering of benzene
normal modes. The fundamental vibrational frequencies, force constants, and dipole derivatives have been
calculated by ab initio quantum chemical methods applying the B3P86 density functional approximation
with 6-311G** basis set. The force field optimized by means of a least-squares scaling procedure for phenol-
d
0
(using six scale factors) was used to calculate the frequencies (with a mean deviation from the observed
values less than 1%), normal modes, potential energy distributions, transition moment vectors, and IR intensities
for phenol-d , -d , -d , and -d isotopomers. Compared to the deviations between the calculated and observed
0
1
5
6
absorption intensities, a more satisfactory correlation was found between the calculated and experimentally
determined Vibrational transition moment directions. The results indicate unanimously that the perturbation
of the normal modes of benzene by the asymmetric hydroxyl substituent is so great that the previous practice
of assigning the normal vibrations of phenol to those of benzene or even to C2V symmetry species is not
justified.
I. Introduction
A number of articles were published during the last 40 years
on the molecular structure1-6 and vibrational spectra7-17 of
phenol reporting the results of both experimental and theoretical
studies. In this paper we are going to present a reassessment
of the vibrational assignment of the IR and Raman spectra of
phenol on the basis of new experimental data and high-level
quantum chemical force field and normal coordinate calculations
fully supporting each other. Our interpretation is aimed at
eliminating some of the contradictions encountered in earlier
works.
Figure 1. Numbering of the atoms in phenol molecule.
The structure of phenol molecule in the ground electronic
state (Figure 1) has been studied by several authors. The
experimental geometry of the free molecule has been determined
structure with ab initio HF/STO-3G method, Puebla and Ha4
used the HF/4-31G approximation, while Bock et al. applied
5
1
by Larsen on the basis of the microwave spectra of six
the ab initio HF method with 6-31G and 6-31G* basis sets.
2
6
isotopomers and by Portalone et al. by means of electron
Later Bock et al. calculated the equilibrium geometry using
diffraction. Konschin3 calculated the optimized molecular
MP2 perturbation with 6-31G* basis set. The structural
parameters determined in the papers cited above are collected
in Table 1 (the numbering of atoms used is given in Figure 1).
The vibrational spectra of phenol have also been investigated
quite extensively. Concerning the assignment of fundamental
†
E-mail: kergabor@cric.chemres.hu.
E-mail: billes@ch.bme.hu.
E-mail: kubinyi@ch.bme.hu.
E-mail: sundius@phcu.helsinki.fi.
‡
§
|
S1089-5639(97)02585-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/03/1998

1
372 J. Phys. Chem. A, Vol. 102, No. 8, 1998
TABLE 1: Experimental and Calculated Structural Parameters of Phenol Molecule
experimental Hartree-Fock
Keresztury et al.
parametera
MWb
EDc
STO3Gd
4-31Ge
6-31Gf
6-31G*f
MP2 (6-31G*g)
DFT/B3P86 (6-311G**h)
C1-C2
1.391
1.394
1.395
1.395
1.392
1.391
1.086
1.084
1.080
1.084
1.081
1.375
0.957
120.9
119.4
120.5
119.2
120.8
119.2
120.0
119.5
120.3
119.8
121.6
117.0
108.8
1.399
1.399
1.399
1.399
1.399
1.399
1.083
1.083
1.083
1.083
1.083
1.381
0.958
121.6
118.8
120.6
119.7
120.6
118.8
1.397
1.386
1.390
1.384
1.382
1.392
1.082
1.083
1.082
1.083
1.082
1.395
0.989
119.8
119.9
120.5
119.4
120.8
119.6
120.4
120.3
120.4
119.8
121.3
117.7
104.9
1.381
1.385
1.381
1.387
1.379
1.383
1.073
1.072
1.071
1.072
1.069
1.374
0.950
120.4
119.6
120.5
119.4
120.7
119.6
120.2
119.4
120.3
119.9
121.7
117.0
114.8
1.385
1.389
1.385
1.390
1.383
1.386
1.074
1.073
1.072
1.072
1.070
1.377
0.949
120.6
119.4
120.4
119.4
120.6
119.4
120.3
119.5
120.3
119.9
121.9
116.8
114.7
1.385
1.387
1.382
1.388
1.381
1.388
1.077
1.075
1.074
1.075
1.074
1.352
0.947
120.1
119.6
120.5
119.2
120.7
119.5
120.0
119.4
120.4
119.9
121.6
117.3
110.7
1.395
1.395
1.393
1.396
1.392
1.396
1.089
1.087
1.086
1.087
1.086
1.374
0.973
120.3
119.7
120.5
119.4
120.6
119.6
120.1
119.2
120.3
120.1
121.6
116.9
108.4
1.393
1.390
1.389
1.392
1.387
1.393
1.087
1.084
1.083
1.085
1.083
1.360
0.961
119.9
119.9
120.5
119.3
120.8
119.7
119.9
119.3
120.4
120.0
121.6
117.6
108.8
C2-C3
C3-C4
C4-C5
C5-C6
C6-C1
C2-H9
C3-H10
C4-H11
C5-H12
C6-H13
C1-O7
O7-H8
C6C1C2
C1C2C3
C2C3C4
C3C4C5
C4C5C6
C5C6C1
C1C2H9
C2C3H10
C3C4H11
C4C5H12
C5C6H13
C6C1O7
C1O7H8
117.2
106.4
a
Distances in angstroms, angles in degrees. b Reference 1. c Reference 2. d Reference 3. e Reference 4. Reference 5. g Reference 6. h This work;
f
E(RB+HF-VWN+P86) ) -308.445 775 970 hartree.
vibrations, the results of earlier works are collected and
compared in Table 2. An early paper by Mecke and Rossmy
Concerning the description of normal modes and the band
assignments, the following main differences can be noted.
7
reported the infrared spectra of phenol and its OD isotopomer
including the study of solvent effects. Later Evans published
(i) While most of the authors regarded the molecule as planar,
Green accepted a model in which the plane of the COH group
8
9
a study of the infrared (in vapor, solution, neat liquid, and solid
phase) and Raman spectra (including depolarization data) of
C6H5OH and C6H5OD, and proposed a complete vibrational
was perpendicular to the benzene ring.
ii) Bist et al.10 and Wilson et al. used a choice of axes
14
(
8
8
different from that of Evans and Green when adopting C2V
symmetry which led to different notation of the symmetry
species.
9
assignment for these compounds. Green published an inde-
pendent assignment of the vibrational spectra of the same
compounds. Bist et al.10 studied the vibrations of C6H5OH,
C6H5OD, and C6D5OH, and in the assignment of the vibrational
spectra they used also the vibronic vapor spectra of these
(iii) Although the set of frequencies accepted by the different
authors as fundamentals are more or less the same, there are
genuine differences and contradictory assignments as well
reflected by the labels corresponding to the normal modes of
11
isotopomers. In his often-cited book Vars a´ nyi assigned,
among 700 molecules, the IR spectra of phenol and O-deuterated
phenol on the basis of systematic comparison of the spectra of
a great number of “mono-light-substituted” benzenes. Larsen
and Nicolaisen12 and later Hutt and Butcher have investigated
11
benzene. It is particularly problematic to distinguish between
the normal modes of vibrations within certain groups of
vibrations, namely, 2, 20a and 20b; 7a, 7b and 13; 8a, and 8b;
13
1
5, 18a, and 18b; 1, 10a and 12a; and 10b and 11. Since the
1
4
the far-infrared spectrum of phenol, while Wilson et al.
measured and interpreted the Raman vapor spectrum.
normal modes within these groups are very close in character,
it is hard to identify them with benzene normal modes, especially
under the conditions of low symmetry when they are more or
Normal coordinate calculations have been performed on
15
phenol by Kovner et al. using the classical GF-matrix method.
Sarin et al.16 used semiempirical formulas to calculate the PR
separations and relative Q-branch intensities in the vapor
less distorted.
-1
(iv) The band at about 820 cm was assigned by some
8-9,11
authors to an in-plane motion
an out-of-plane mode.
while others assigned it to
The reverse alternative assignments
4
spectrum, while Puebla and Ha applied the ab initio HF/4-
10,17
3
1G method for the calculation of vibrational frequencies but
-1
were given to a nearby band at about 810 cm .
With the exception of our previous paper,17 earlier descrip-
published only a few calculated frequencies. In our earlier
1
7
work, we applied the semiempirical CNDO/2 method with a
subsequent scaling of the calculated force constants to achieve
a better frequency fit. The intensities and polarization properties
of the IR absorption bands have not been studied yet.
tions of the fundamental vibrations of phenol were either limited
to the indication of a single type of internal coordinate
dominating in the normal mode and/or given in terms of the
accepted numbering of the “corresponding” benzene normal
modes. To the best of our knowledge, the only assignment
based on potential energy distribution has been reported by us.17
In that work, however, we applied the semiempirical CNDO/2
method, whereas these days far superior quantum mechanical
approximations are available. A revision of the assignment of
Most of the earlier assignments reviewed in Table 2 were
based on comparison of phenol normal modes to those of
benzene, and the fundamentals were designated using the
accepted notation of benzene normal modes11 or were assigned
to the symmetry species of the hypothetical C2V point group.

Phenol and Its Deuterated Derivatives
TABLE 2: Earlier Assignments of Fundamental Vibrations of the Phenol Moleculea
Evansb Greenc
J. Phys. Chem. A, Vol. 102, No. 8, 1998 1373
Bistd
obs
obs
obs
ν
i
assignment
ν
i
assignment
ν
i
assignment
νOH
νCH, A1, 20a
νCH, B2, 20b
νCH, A1, 2
νCH, B2, 7b
νCH, A1, 13
νCC, B2, 8b
νCC, A1, 8a
νCC, A1, 19a
νCC, B2, 19b
νCC, B2, 14
3
3
3
3
3
3
1
1
1
1
610
074
061
052
046
021
608
600
502
473
νOH
3656
3087
3070
3063
3049
3027
1610
1603
1501
1472
1343
νCH, B1
νCH, A1
νCH, B1
νCH, B1
νCH, A1
νrg, A1
νrg, B1
νrg, A1
νrg, B1
3091
3085
3076
3044
3030
1604
1596
1497
1465
νCH, A′′, B1, 20b
νCH, A′, A1, 20a
νCH, A′, A1, 2
νCH, A′, A1, 13
νCH, A′′, B1, 7b
νCC, A′, A1, 8a
νCC, A′′, B1, 8b
νCC, A′, A1, 19a
νCC, A′′, B1, 19b
νCC, A′′, B1, 14
âCH, A′′, B1, 3
1
1
333
313
1
1
1
1
1
1
1
1
290
259
228
167
151
070
026
000
âCH, B1
1277j
1261
1176
1168
1150
1070
1025
999
âCH, B2, 3
νCO, A1, 7a
âOH
âCH, A1, 9a
âCH, B2, 9b
âCH, B2, 15
âCH, A1, 18a
rg, A1, 1
γCH, A2, 17a
γCH, B1, 5
γCH, B1, 17b
X-sensor, A1, 12
γCH, A2, 10a
γCH, B1, 10b
γCC, B1, 4
rg, B2, 6b
X-sensor, A1
âOH + νrg
âCH, A1
âCH, B1
1259
X-sensor, A′, A1, 7a
1167
1145
1071
1026
999
978
958
881
825
810
749
688
617
526
500
408
408
âCH, A′, A1, 9a
âCH, A′′, B1, 9b
âCH, A′′, B1, 15
âCH, A′, A1, 18a
rg, A′, A1, 1
âCH, B1
âCH, A1
ârg, A1
j
972
958
883
829
814
753
691
619
530
508
415
404
γCH, B2
γCH, A′′, B2, 5
995
γCH, A2
γCH, B2
γCH, A′′, A2, 17a
γCH, A′, B2, 17b
γCH, A′′, A2, 10a
X-sensor, A′, A1, 12a
γCH, A′, B2, 10b
γCC, A′, B2, 4
973
881
823
γCH, A2
X-sensor, A1
γCH, B2
j
817
751
686
619
γrg, B2
ârg, B1
γCC, A′′, B1,6b
X-sensor, A1
X-sensor, B2
γrg, A2
X-sensor, A′, A1, 6a
X-sensor, A′, B2, 16b
γCC, A′′, A2, 16a
X-sensor, A′′, A2, 18b
503
409
403
X-sensor, B1, 16b
γCC, A2, 16a
X-sensor, B2, 18b
τOH
j
X-sensor, B1
3
09
2
42
X-sensor, B2
241
X-sensor, A′, B2, 11
244
X-sensor, B1, 11
Vars a´ nyie
Wilsonf
assignment
Saring
Kovnerh
obs
obs
obs
obs
calcd
ν
i
assignment
ν
i
ν
i
type
Ak
ν
i
ν
i
3
3
3
3
3
3
1
1
1
1
1
1
623
091
076
048
044
030
604
596
497
465
νOH
20a
2
20b
7b
7a
8a
3656
3073
νOH
3656
3650
3071
3061
3052
3046
3021
1603
1597
1502
1474
1370
1292
1252
1230
3653
3075
3070
3061
3054
3045
1624
1610
1504
1477
1374
1312
1239
1233
A1, 2
1612
1602
1504
1470
1388
B2, 8b
A1, 8a
A1, 19a
B2, 19b
B2, 14
8b
19a
19b
14
1501
1471
1344
A
k
B
B
k
333-1376
313
3
1
285
B2, 3
A1, 7a
1
1
1
1
1
1
259
180-1235
13
âOH
9a
1264
1196
1168
1149
167
145
071
026
A1, 9a
B2, 9b
1170
1155
1072
1025
1000
1174
1154
1070
1017
1001
9b
18b
18a
12
1070
1026
B
A
k
1025
999
A1, 18a
A1, 1
9
9
9
8
8
8
7
6
6
5
5
4
4
3
2
99
78
58
81
25
10
49
88
17
26
00
10
08
00
41
5
17a
17b
10a
1
11
4
6b
6a
16b
16a
15
881
C
824
528
235
A1, 12
A1, 6a
B1, 11
814
773
751
686
C
C
620
530
608
498
403
310
Bk
C
γOH
10b
398
415

1374 J. Phys. Chem. A, Vol. 102, No. 8, 1998
Keresztury et al.
TABLE 2 (continued)
TABLE 3: Optimized Scale Factors for the Phenol Force
Field Obtained from the DFT/B3P86/6-311G** Calculation
Kubinyii
assignment (PED %)
type of coordinatea
scale factor
obs
calcd
ν
i
ν
i
CH, CC, and CO stretching
CCH and CCO bending
CCC bending
OH stretching
COH bending
0.9135
0.9860
0.9675
0.8903
0.9996
0.5930
3
3
3
3
3
3
1
1
1
1
1
1
1
610
074
061
052
046
021
608
600
502
473
344
290
259
3600
3063
3061
3054
3045
3038
1628
1613
1492
1469
1363
1281
1269
A′, νOH(100)
A′, νCH(99)
A′, νCH(99)
A′, νCH(99)
A′, νCH(100)
C-OH torsion
A′, νCH(100)
a
A′, νCC(58), âCH(20), âCC((13)
A′, νCC(63), âCH(13)
A′, âCH(55), νCC(34)
A′, âCH(48), νCC(30), âOH(15)
A′, âCH(72), âOH(21)
A′, âCH(39), νCC(36), âOH(11)
A′, νCO(37), νCC(23), âCH(23)
Force constants corresponding to coordinates not listed here were
kept unscaled.
Au/AgCl wire grid polarizer placed in front of the sample was
used to generate linearly polarized radiation. The sample
consisted of a 5% (m/m) solution of phenol in liquid crystal
ZLI-1695 (Merck) filled into a 0.024 mm KBr liquid cell with
1
8
specially pretreated windows. The two absorbance spectra
recorded with the electric vector of light oriented parallel and
perpendicular to the director of the liquid crystal were used in
1
1
1
1
1
1
179
167
151
070
026
000
1179
1162
1155
1054
1007
985
A′, νCC(38), âOH(31), âCH(29)
A′, âCH(74), νCC(25)
A′, âCH(58), νCC(38)
A′, νCC(58), âCH(35)
A′, νCC(72), âCH(21)
A′, âCC(61), νCC(38)
A′′, γCH(122), γCC(-22)
A′′, γCH(114), γCC(-14)
A′′, γCH(102), γCC(-10)
A′′, γCH(100)
1
9
a stepwise reduction process to obtain the so-called reduced
spectra and the dichroic ratios di of every band by successive
interactive subtractions. The orientation factors and transition
moment directions were calculated from the observed dichroic
ratios using the formulas given by Michl and Thulstrup.19
In order to determine the positions and intensities of spectral
bands more precisely, the overlapping bands in both IR and
Raman spectra were resolved by curve fitting.
9
9
72
58
965
954
8
85
8
8
7
6
6
5
5
4
3
4
2
29
14
53
91
19
30
08
15
10
04
42
831
800
750
701
634
519
492
421
309
402
229
A′, νCC(40), âCC(34), νCO(23)
A′′, γCC(53),γCH(27)
A′′, γCC(54), γCH(47)
A′, âCC(80), νCC(15)
A′, âCC(73), νCO(13), νCC(11)
A′′, γCC(53), γCO(43)
A′′, γCC(112), γCH(-14)
A′′, τOH(98)
III. Computational Details
Ab initio QM calculations were performed applying the
density functional theory with Becke’s three-parameter func-
2
0
tional, B3P86 and a large basis set, 6-311G**, by means of
A′, âCO(82)
A′′, γCC(71), γCO(33)
21
the Gaussian 92/DFT program package.
First the vibrational force constants, harmonic vibrational
frequencies, net atomic charges, the dipole moment, and its
derivatives and the thermodynamic properties of undeuterated
phenol were calculated. Our preliminary geometry optimiza-
tions done with simultaneous relaxation of all structural
parameters led to a practically planar structure; thus, in the final
optimization, the geometry was constrained to Cs symmetry and
the perfectly planar optimized structure was used as an equi-
librium reference geometry. The reference geometry and the
corresponding Cartesian force constant matrix (Fx) were taken
as starting data in all further calculations. The Fx matrix was
transformed into a nonredundant set of 33 suitable internal
coordinates (the “natural coordinates”) constructed according
a
Frequencies in cm-1: ν, stretch; â, bend; γ, oop. bend; τ, torsion;
b
c
d
X, substituent; rg, ring. Reference 8. Referrence 9. Reference 10.
Reference 11. Reference 14, Raman vapor spectrum. Reference 15.
Reference 14. Reference 16, CNDO/2 calculations. From UV vapor
spectrum. Hybrid band.
e
f
g
h
i
j
k
the vibrational spectra of phenol is therefore desirable utilizing
a force field derived from a modern post-Hartree-Fock calcula-
tion done with sufficiently large basis set.
In what follows we present the results of such a calculation
together with new experimental evidence obtained from linear
dichroism (LD) measurements that has prompted us to discard
the C2V symmetry species for classification of phenol vibrations
and give a revised assignment more consistent with all available
data.
2
2
to the recommendations. These coordinates include 13 bond
stretches, 5 CH, 1 CO, 1 OH, and 3 ring in-plane bending
coordinates; 1 OH and 3 ring out-of-plane torsions, and 5 CH
and 1 CO out-of-plane bending coordinates.
II. Experimental Section
The F matrix obtained was scaled using six empirical scaling
factors to ensure a better fit of observed and calculated
frequencies. The multiple scaling method of Fogarasi and
C6H5OH and C6D5OD were purchased from Aldrich and used
without further purification, while C6D5OH was obtained by
recrystallization of C6D5OD from water.
The FT-IR spectra of phenol, C6D5OH, and hexadeuteriophe-
nol were measured in vapor phase and in CCl4 and CS2
solutions. The spectra were recorded on a Nicolet Magna 750
FT-IR spectrometer at 1 and 2 cm-1 resolution using a 0.016
mm KBr liquid cell for the solutions and a 1 m gas cell (Carl
Zeiss Jena) for phenol vapors.
Raman spectra of phenol isotopomers were measured in CCl4
and CS2 solutions on a Nicolet FT-Raman Model 950 spec-
trometer at 2 cm resolution using the 1064 nm line of a Nd:
YAG laser for excitation.
2
3
Pulay was used with a least-squares refinement of the scale
factors, whereas the calculated vibrational frequencies of phenol-
d0 were fitted to the experimentally assigned normal frequencies.
Six scale factors were used in all; their optimized values are
given in Table 3. The same optimized (scaled) general valence
force field was applied also for the frequency calculations of
the three deuterated phenol isotopomers. In addition to the
frequencies and vibrational eigenvectors, the potential energy
distribution (PED) matrices, transition moment vectors, and
relative absorption intensities were also calculated for charac-
terization of the normal modes. These final calculations were
also performed for C6H5OD, C6D5OH, and C6D5OD isoto-
pomers.
-
1
Polarized IR spectra were measured with the sample aligned
in uniaxially oriented nematic liquid crystal. A Perkin-Elmer

Phenol and Its Deuterated Derivatives
J. Phys. Chem. A, Vol. 102, No. 8, 1998 1375
TABLE 4: Measured and Calculated (DFT/B3P86/6-311G**, Scaled Force Field) Vibrational Frequencies and Relative
Absorption Intensities of C OH Fundamentals
6 5
H
frequency (cm-1)
intensity (arbitrary units)
i
measdb
calcd
measdb
calcd
potential energy distribution (%)
1
3655
3074
3061
3052
3046
3021
1609
1604
1501
1472
1361
1344
1261
1197
1176
1150
1070
1026
999
973
956
881
(823)
810
752
687
618
526
3655
3073
3066
3053
3045
3025
1613
1601
1502
1476
1359
1328
1260
1192
1170
1159
1069
1015
986
977
955
878
811
808
755
687
620
104.22
0.35
1.75
4.29
11.83
1.93
36.09
62.90
69.12
42.92
0.96
17.52
51.33
106.26
10.34
13.87
12.18
4.73
60.7
4.2
15.5
15.7
0.2
13.7
40.0
47.3
48.1
39.5
43.8
6.0
77.6
135.5
7.4
10.8
12.7
3.8
4.0
0.01
0.1
νOH(100)
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
νCH(99)
νCH(99)
νCH(99)
νCH(99)
νCH(99)
νCC(64), âCH(18), âCC(10)
νCC(66), âCH(16), âCC(9)
âCH(56), νCC(33), νCO(6)
âCH(52), νCC(33), âOH(8)
âCH(63), âOH(26), νCC(7)
νCC(63), âCH(33)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
νCO(50) , νCC(19), âCH(17), âCC(10)
âOH(45), νCC(26), âCH(23)
âCH(75), νCC(24)
νCC(23), âCH(74)
νCC(55), âCH(38)
νCC(69), âCH(23), âCC(6)
âCC(61), νCC(37)
3.68
0.26
0.18
8.94
γCH(83), γCC(17)
γCH(88), γCC(12)
6.9
γCH(88), γCC(12)
0.04
18.7
67.4
26.6
0.4
1.7
13.8
0.4
γCH(99)
17.78
74.81
42.92
0.61
1.93
20.5
0
νCC(42), âCC(30), νCO(24)
γCH(65), γCO(27), γCC(7)
γCC(91), âCH(5)
âCC(82), νCC(11), νCO(5)
âCC(78), νCO(11), νCC(9)
γCC(49), γCO(45), γCH(6)
γCC(88), γCH(12)
âCO(79), νCC(8), âCC(8)
τOH(97)
γCC(87), γCO(13)
525
512
410
402
307
228
503
420
410
8.23
9.9
111.2
1.8
d
310
d
242
a
Mean deviation in frequencies: 4.85 cm-1 (0.70%). b Vapor. c CCl
solution. d From ref 17.
4
-5
The dipole derivatives (derivatives of the electric dipole
moment with respect to the atomic coordinates) obtained from
the DFT calculation were used to calculate the vibrational
transition moments (i.e., the dipole derivatives with respect to
the normal coordinates (corresponding to the scaled force field)).
The directions of the calculated transition moment vectors were
used, via comparison to the values determined experimentally
from the IR linear dichroism measurements, to correct for the
sign ambiguity of the latter values and to evaluate the extent of
symmetry lowering of the normal modes.24
The other HF calculations done with larger basis sets4
underestimate the C-C and C-H bond lengths but give good
or acceptable values for the C-O and O-H distances; their
optimum valence angles are also acceptable with the exception
of the C-O-H angle that is overestimated.
Our DFT calculation shows the best overall agreement with
the experimentally determined values, the results of the MP2
6
calculation being almost equally good. Both methods produced
slightly different values for the individual C-C and C-H bond
lengths; the values are similar to those of the rs structure, but
their distribution around the ring is different. It is probably
well founded to state that in the equilibrium structure the shape
of the phenyl ring deviates slightly from a regular hexagon and
it does not have even C2V symmetry. The C1-O7 bond makes
an angle of nearly 3° with the axis passing through the C1 and
C4 atoms (see Figure 1). The C-O bond length is a little
underestimated in the present calculation, while the O-H
distance is in good accordance with the rs one. There is an
excellent agreement between our calculated values and the
experimental rs data for all valence angles.
IV. Results and Discussion
A. Structural Parameters. The optimized structural pa-
rameters of phenol obtained from the DFT calculation using
B3P86/6-311G** method are compared with the experimental
data and earlier calculated values in Table 1 (last column) where
the numbering of atoms is as given in Figure 1. Of the two
experimental structure determinations the microwave measure-
1
ment of Larsen yielded more detailed results: he allowed
individual values for all bond lengths and valence angles in the
2
In summary, one may conclude that our calculated structure
provides a reliable starting point for the DFT/B3P86/6-311G**
force field and frequency calculations.
determination of the rs structure. In contrast, Portalone et al.
in their electron diffraction work assumed equal lengths for the
six C-C bonds as well as for the five C-H bonds and even
the bond angles were constrained to C2V symmetry of the ring.
Nonetheless, the numerical values obtained by the two experi-
mental methods are rather close; yet the microwave results do
not justify the assumptions made in the electron diffraction work.
B. Fundamental Vibrational Frequencies and Intermo-
lecular Association. The primary aim of our infrared and
Raman survey measurements was to check the correctness of
the experimental frequency values that were used in earlier
studies, with high-precision, up-to-date instrumentation. The
results obtained for the four phenol isotopomers are in good
Regarding the ab initio HF calculated values, the STO-3G
3
results show the greatest deviations from the experimental ones.

1
376 J. Phys. Chem. A, Vol. 102, No. 8, 1998
Keresztury et al.
TABLE 6: Measured and Calculated (DFT/B3P86/
TABLE 5: Measured and Calculated (DFT/B3P86/
6-311G**, Scaled Force Field) Vibrational Fundamentals of
C
ODa
6-311G**, Scaled Force Field) Vibrational Fundamentals of
H
5
C
6
D
5
OHa
frequencies (cm-1
6
frequencies (cm-1)
)
measdb
calcd
potential energy distribution (%)
measd
calcd
potential energy distribution (%)
3
3
3
3
3
2
1
1
1
1
1
1
1
1
1
1
1
1
075
062
046
040
014
670
602
597
499
463
333
305
251
169
156
078
027
001
70
62
32
80
27
08
53
90
17
28
11
16
83
41
3073
3067
3054
3045
3026
2661
1609
1592
1498
1464
1334
1311
1255
1169
1160
1077
1013
989
978
956
942
878
811
802
755
686
νCH(99)
3656
2294
2281
2262
2258
2249
1579
1567
1404
1372
1301
1204
1179
1021
960
3655
2279
2271
2258
2248
2235
1581
1566
1403
1374
1318
1210
1184
1023
948
868
843
830
812
801
775
758
748
νOH(100)
νCH(99)
νCD(96)
νCH(99)
νCD(96)
νCH(99)
νCD(96)
νCH(99)
νCD(97)
νOD(99)
νCD(96)
νCC(64), âCH(21), âCC(10)
νCC(68), âCH(16), âCC(9)
âCH(59), νCC(33), νCO(6)
âCH(56), νCC(35), âOD(5)
âCH(80), νCC(14)
νCC(74), âCC(9), âCD(9)
νCC(73), âCC(8), âCD(7)
νCC(46), âCD(17), âOH(16), νCO(13)
νCC(47), âCD(19), âOH(16), νCO(8)
âCC(85), âCD(7), âOH(6)
νCC(40), âOH(38), âCD(17)
νCO(38), νCC(26), âCD(18), âCC(12)
âCD(84), νCC(5), âCO(5), âOH(5)
âCC(62), νCC(33)
νCC(72), âCH(19)
νCO(52), νCC(18), âCH(16), âCC(12)
âCH(75), νCC(24)
âCH(76), νCC(23)
νCC(48), âCH(44)
869
837
831
813
âCD(52), νCC(34), âCC(9)
âCD(82), νCC(15)
νCC(66), âCH(23), âCC(9)
âCC(57), νCC(40)
âCD(63), νCC(24), âCC(7)
âCD(78), νCC(9)
9
9
9
8
8
8
7
6
6
5
5
4
3
2
γCH(84), γCC(16)
γCH(88), γCC(12)
γCD(92), γCC(5)
âOD(78), νCC(16)
776
756
754
636
624
594b
550
513
430c
386
357
307c
232
γCD(86), γCC(12)
γCH(78), γCC(11), γCO(9)
γCH(99)
γCD(60), γCO(36)
νCC(39), âCD(29), âCC(16), νCO(14)
γCD(95)
νCC(41), νCO(30), âCC(30)
γCH(66), γCO(27), γCC(7)
γCC(92), γCH(4)
634
628
597
552
514
435
385
359
γCD(71), γCC(17), γCO(12)
âCC(79), νCC(10)
γCC(64), γCD(33)
âCC(77), νCO(10), νCC(8)
γCD(39), γCC(31), γCO(29)
âCO(76), νCC(9), âCC(8)
γCC(88), γCD(12)
τOH(96)
γCC(85), γCO(10), γCD(5)
622
522
511
406
383
241
211
âCC(83), νCC(11), νCO(4)
âCC(76), νCC(9), νCO(11)
γCC(49), γCO(45), γCH(6)
γCC(88), γCH(12)
âCO(78), νCC(8), âCC(7)
τOD(47), γCC(44), γCO(10)
τOD(55), γCC(45)
305
214
(
232)
a
Mean deviation: 5.84 cm-1 (0.78%). b From ref 8.
a
Mean deviation: 5.00 cm-1 (0.66%). b Condensed phase value.
c
From ref 9b.
general agreement with the previously reported experimental
data, with just a few minor deviations concerning the choice of
fundamentals.
-
1
shifted, broad band (3130-3540 cm ) belonging to the
hydrogen-bonded OH groups increases with the concentration
rapidly.
In order to be consistent with the DFT calculations where
only isolated molecules were studied, we selected the vapor
phase values of the fundamental frequencies for comparison
wherever they could be located with certainty. They were
The influence of the environment on the position of the
monomer ν(OH) band is shown in Figure 3. The νOH
frequency of a monomeric phenol molecule is shifted toward
lower frequencies in the order: vapor phase f CCl4 solution
f CS2 solution. The relative permittivities of CCl4 and CS2 at
-
1
8
typically 1-2 cm higher than those reported by Evans. The
frequencies we assigned to fundamental vibrational transitions
of the free molecules are listed in the first columns of Tables
2
0 °C are 2.238 and 2.641, respectively, which suggests that
the lowering of frequencies of the monomer νOH band in
solution is due to the solvent polarity effect of the environment.
A similar trend is observed for most of the other vibrations as
well, although the shifts are usually much smaller: the CCl4
solution frequencies are in general closer to the vapor phase
values, while the CS2 solution frequencies are 1-3 cm lower.
At variance with previous investigators, we assign the OH
4
-7 for phenol-d0, -d1 (OD), -d5 (OH), and -d6, respectively.
The solution data were used to check the number of bands in
crowded spectral regions (where the interpretation of the
rotational band structure in the vapor spectra posed greater
difficulties) or as substitutes if the bands in the vapor spectrum
were too weak to be observed (some γCH bands).
-1
Phenol is prone to self-association which must be kept in
mind when interpreting its condensed phase vibrational spectra.
The spectral features significantly affected by H-bonding
interactions are those of the OH-group: the OH stretching
bending mode of free phenol to a clearly observed Q-branch at
-1
1
198 cm , a feature which may not have been resolved in
-1
earlier measurements. The broader band at 1218-1220 cm
in the solution spectra corresponds to the same mode of
H-bonded OH-groups.
(
(
νOH), bending (âOH), and out-of-plane bending, or torsional
τOH) bands. Our observations confirm the coexistence of free
It is of interest to compare the IR spectra of phenol and
hexadeuteriophenol. The fingerprint region of the CCl4 solution
spectra is shown in Figure 4 and the high frequency region in
Figure 5. Note that simultaneously with the expected band shifts
significant changes of band intensities occur, especially in the
fingerprint region, that may be explained by altered intermo-
lecular vibrational couplings.
and different associated forms, their equilibrium being dependent
on concentration. In Figure 2, the infrared spectrum of a 5%
(m/m) solution of phenol in carbon tetrachloride is compared
with that of a 1% solution (for easier comparison the intensity
of the latter was multiplied by 5). The intensity of the monomer
-
1
ν(OH) band at 3611 cm decreases, while that of the down-

Phenol and Its Deuterated Derivatives
J. Phys. Chem. A, Vol. 102, No. 8, 1998 1377
TABLE 7: Measured and calculated (B/DFT3P86/6-311G**,
6 5
scaled force field) vibrational fundamentals of C D
ODa
frequencies (cm-1)
measd
calcd
potential energy distribution (%)
2
2
2
2
2
2
1
1
1
1
1
1
1
9
9
8
8
8
8
8
7
7
7
6
6
5
5
5
4
3
3
2
2
655
294
286
277
256
246
585
567
391
351
292
178
056
57
05
74
55
34
11
01
86
53
50
55
23
93
51
09
32
67
58
35
03
2661 νOD(99)
2279 νCD(96)
2271 νCD(96)
2258 νCD(96)
2248 νCD(97)
2235 νCD(96)
1573 νCC(74), âCC(9), âCD(9)
1557 νCC(76), âCC(7), âCD(6)
1390 νCC(52), âCD(21), νCO(21)
1345 νCC(61), âCD(24), âCO(7)
1303 âCC(91)
1184 νCO(40), νCC(26), âCD(18), âCC(13)
1066 âCD(78), âOD(16)
948 âCC(60), νCC(34)
919 âOD(65), âCD(17), νCC(13)
868 âCD(55), νCC(33), âCC(8)
843 âCD(82), CC(15)
828 âCD(64), νCC(24), âCC(6)
812 âCD(77), νCC(20)
801 γCD(91), γCC(5)
775 γCD(88), γCC(10)
758 γCD(60), γCO(36)
744 νCC(38), âCD(28), âCC(17), νCO(14)
634 γCD(95)
627 γCD(72), γCC(16), γCO(12)
597 âCC(80), νCC(9)
552 γCC(64), γCD(34)
508 âCC(75), νCO(10), νCC(8)
435 γCD(39), γCC(31), γCO(29)
367 âCO(76), νCC(8), âCC(7)
358 γCC(76), γCD(24)
235 τOD(68), γCC(27), γCO(5)
203 γCC(64), τOD(32)
Figure 3. Solvent effect in the IR spectrum of phenol in the 3800-
-1
2
4 2
900 cm region: A, vapor spectrum; B, CCl solution; C, CS
solution.
Figure 4. IR spectra of phenol (A) and hexadeuteriophenol (B) in the
-
1
1
4
7000-400 cm region in CCl solution.
a
Mean deviation: 7.05 cm-1 (0.64%)
Figure 5. IR spectra of phenol (A) and hexadeuteriophenol (B) in the
-
1
3
4
750-2200 cm region in CCl solution.
energy distributions for each normal mode in Tables 4, 5, 6,
and 7, respectively. In case of the parent molecule the observed
and calculated relative absorption intensities are also given in
Table 4.
In spite of the fact that the optimization of the scale factors
was based only on the C6H5OH frequencies, the resulting
absolute and relative mean deviations between the measured
and calculated frequencies of all the four isotopomers are similar
and very small indeed, which is a good proof of the reliability
of the force field.
Obviously, a shift of certain fundamental frequencies towards
lower wavenumbers was expected on deuteration. The analysis
of the eigenvectors and PEDs reveals some interesting details.
As expected, the CH, CD, OH, and OD stretching vibrations
have highly characteristic frequencies, since they are pure group
vibrations. The CO stretching and OH bending coordinates,
on the other hand, contribute considerably to more than one
vibrations, and these contributions change substantially with the
Figure 2. Effect of dilution on the infrared spectrum of the CCl
solution of phenol in the 3800-2900 cm-1 region (CCl
spectrum
subtracted): A, 1% (m/m) solution (5x ordinate expansion); B, 5%
4
4
(m/m) solution.
Comparing the spectra in the νCD-νOD and νCH-νOH
region (Figure 5) a similar association pattern of the OD- and
-1
OH-groups is apparent (see the 2500-2600 cm and the 3200-
-
1
3
600 cm regions). When interpreting the spectra we had to
keep in mind that our phenol-d6 spectra contained bands of
phenol-d5 as well: the deuteration of the ring was practically
complete, but only about 70% of the hydroxyl hydrogens were
exchanged by deuterium.
C. Normal Coordinate and Intensity Calculations. The
observed and calculated frequencies of C6H5OH, C6H5OD,
C6D5OH, and C6D5OD are presented together with the potential

1378 J. Phys. Chem. A, Vol. 102, No. 8, 1998
Keresztury et al.
Figure 7. Calculated Vs. measured relative absorption intensities of
fundamental vibrational transitions of phenol (from data in Table 4).
Figure 6. Comparison of the calculated (A) and measured (B, in
solution) relative absorption intensity patterns of phenol (from data in
Table 4).
extent of deuteration. The energy distribution of the CO
stretching motion is not sensitive to O-deuteration but changes
on deuteration of the ring. When the phenyl group is not
deuterated, the ν13 normal mode with a frequency of about
-
1
1
250-1260 cm contains about 50% of νCO (according to
the PED), while only 6% of νCO is found in higher frequency
-1
modes, which is the 1500 cm vibration in phenol and phenol-
OD. If, however, the phenyl group is fully deuterated (C6D5
group instead of C6H5), both frequencies concerned decrease
-
1
by about 70 cm and the contribution of the νCO coordinate
increases in the higher frequency mode and decreases in the
lower frequency mode.
In contrast, the OD bending vibration is more characteristic
than is the OH bending vibration. There are rather pure OD
in-plane bending characteristic frequencies with âOD participa-
tion of 65 and 78% in the spectra of C6D5OD and C6H5OD,
respectively, while the highest contribution of the âOH coor-
dinate to a normal coordinate is 45 and 38% in the case of C6H5-
OH and C6D5OH, respectively.
Figure 8. Polarized IR spectra of phenol aligned in a nematic solution
(5% m/m) in ZLI-1695; uniaxial alignment, electric vector parallel (A),
and perpendicular (B), with the director of the liquid crystalline sample.
asymmetry for different vibrations. Besides, it always remains
a question how reliable our calculated eigenvectors are.
We consider it a more effective way to estimate the deviation
of the dynamic properties of the molecule from symmetric
behavior by examining the vibrational transition moment
directions. This can be done both experimentally and theoreti-
cally, which provides a cross-checking possibility. When the
symmetry is high enough, C2V or higher, the transition moments
are aligned along three mutually orthogonal axes within the
molecule. When the symmetry is lower than that (i.e., C2h, Cs,
Ci, or C1), the transition moments assume more than three
different directions. In practice the simplest way may be to
measure the linear dichroism of uniaxially oriented samples
The OH torsional vibrations are very characteristic: their PED
is 97 and 96% in one normal coordinate of C6H5OH and C6D5-
OH, respectively. In contrast, the OD torsional coordinate is
distributed between two normal modes with the two lowest
frequencies; it gets strongly coupled with an out-of-plane ring
deformation vibration.
The calculation of IR intensities is known to be less successful
than that of the frequencies. At this level of theory, however,
there is a recognizable similarity between the calculated and
observed intensity pattern of phenol-d0 (see Figure 6), although
the correlation shown in Figure7 is still not as good as would
be desirable.
(solute molecules aligned in nematic liquid crystal or in stretched
polymer) to determine the dichroic ratios of absorption bands,
parameters that are dependent on the transition moment direc-
tions.
Out-of-plane Vibrations. The two basic polarized IR spectra
of phenol in LC solution are shown in Figure 8. The dichroic
ratios di determined as di ) Ip/Is (i.e., as ratios of the band
intensities measured at p (parallel) polarization to those at s
D. Linear Dichroism and Asymmetry of Molecular
Dynamics. As we have seen above, even disregarding the
H-atom of the O-H-group, the C2V symmetry of the molecular
skeleton in phenol is ruined. Concerning the vibrational
assignment, the key question is whether these small distortions
and the asymmetric influence of the OH bond are negligible
with respect to the phenyl group vibrations so that they can be
classified under the C2V symmetry species, or are significant
enough to defy such a description.
(perpendicular) polarization) are given in the third column of
Table 8. Note that the di values range from 0.325 to 3.71 but
they assume more than three different values. The transition
moments of the out-of-plane vibrations, being perpendicular to
the plane of the molecule, are expected to be the least aligned
along the direction of sample orientation (the director of the
liquid crystal); thus, the smallest di values are expected to belong
One way of answering this question is by inspecting the
calculated normal modes to see whether they are symmetric or
not. This, however, may be a tedious procedure; in addition, it
would be hard to define a common limit of tolerance of

Phenol and Its Deuterated Derivatives
i
TABLE 8: Dichroic Ratios (d ), Orientation Factors (K ),
J. Phys. Chem. A, Vol. 102, No. 8, 1998 1379
i
and Observed and Calculated Transition Moment Directions
exp
calc
(O
i
and O
i
, Respectively) of Phenol-d
0
exp
calc
i
ν
i
d
i
K
i
φ
i
φ
i
1
2
3
4
5
6
7
8
9
3403
3093
3074
3050
3043
3021
1607
1595
1502
1472
1358
1335
1268
1219
1165
1151
1069
1024
999
971
952
883
(829)
814
754
692
620
619
3.71
3.70
0.92
1.27
1.27
1.86
2.00
1.01
1.18
2.03
1.57
3.30
1.40
1.40
1.46
0.86
2.41
1.58
1.38
nmV
nmV
0.325
1.13
1.13
0.325
0.325
0.325
3.70
0.79
0.325
1.05
0.65
0.65
0.32
0.39
0.39
0.48
0.50
0.34
0.37
0.50
0.44
0.62
0.41
0.41
0.42
0.30
0.55
0.44
0.41
nmV
nmV
0.14
0.36
0.36
0.14
0.14
0.14
0.65
0.28
0.14
0.34
0.50
0.0
0.0
-0.6
0.7
60.7
-50.5
-50.5
38.2
35.7
57.7
52.8
35.2
43.7
-14.4
47.4
47.4
46.0
63.0
-29.0
43.5
47.8
nmV
nmV
x
54.1
54.1
x
x
x
0.0
65.9
x
56.5
35.7
73.8
-30.0
-21.1
15.1
26.1
60.2
47.9
26.7
38.7
-6.2
38.3
44.2
26.6
67.2
-12.7
41.5
46.2
Figure 9. Comparison of measured (A) and computed (B) vibrational
transition moment directions of phenol fundamentals (see data in Table
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
2
7
8
9
0
1
8
).
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
3
2
2
2
3
3
out-of-plane vibration in question is masked and too week to
be observed (the ab initio calculated IR intensity of the ν23
vibration confirms this assumption, see Table 4), and on the
other hand, the two bands with the same in-plane polarization
are due to Fermi resonance of ν23 with 2ν31 (2 × 410) ) 820
-1
cm ).
In-Plane Vibrations. The bands corresponding to the in-plane
vibrations have individual dichroic ratios ranging from 0.79 to
3.71, the highest di value being that of the OH stretching band
a
b
x
x
x
x
40.4
x
x
x
-78.3
55.7
x
x
38.4
-
1
(the very broad band centered at 3403 cm ). This indicates
that the OH bonds of phenol molecules are oriented preferen-
tially along the director of the liquid crystal, probably due to a
nearly linear hydrogen bond formed with the nitril groups of
the rodlike molecules of the LC material. For this reason the
transition moment direction of the O-H stretching band was
accepted as the axis of preferential orientation (the “long” axis,
z) of phenol molecules.
529
509
(424)
415
2.00
To obtain the transition moment directions of other in-plane
bands from the measured dichroic ratios, the method described
a
nmV ) no measured value. b x: out-of-plane direction.
1
9
by Michl and Thulstrup was used. According to this, the
to the out-of-plane vibrations. This means that the observed
dichroic ratios can be used, without any further manipulations,
to check the assignment of the out-of-plane vibrations.
Indeed, data in Table 8 show that the strong bands at 754
orientation factors K were first calculated for each band as
i
K ) d /(d + 2)
(1)
i
i
i
-
1
and 692 cm and the medium intensity bands at 883 and 509
then the angles φi that the transition dipole moment of an in-
plane vibration makes with the axis of preferential orientation
z were obtained from the relation:
-
1
cm have this low dichroic ratio (di ) 0.325) confirming that
they belong to out-of-plane vibrations. Note that the broad
-1
absorption band at 620 cm originating from τOH, the torsional
vibration of the H-bonded OH-group, shows the same dichroic
behavior as all out-of-plane vibrations. This can be regarded
as a direct experimental evidence of the planar structure of
phenol (at least in this solution), which agrees with the results
of the quantum chemical calculations.
2
tan φ ) (K - K )/(K - K )
(2)
i
z
i
i
y
where K ) K
z
νOH
) 0.65, and K ) 0.21, the orientation factor
y
of the “short” in-plane axes, y; the latter was obtained from the
condition K + K + K ) 1, where K ) 0.14 is the orientation
x
y
z
x
Our frequency calculations indicate that three more out-of-
plane γCH bands should be found in the spectrum, two in the
factor of the out-of-plane x axis of the molecule.
Surprisingly, the computed transition moment direction of
the νOH vibration makes an angle of nearly 38° with the OH
bond direction according to our DFT/B3P86 calculation (see
the z-axis in Figure 9). Nevertheless, this direction (rather than
the direction of the O-H-bond) was accepted in the molecule-
fixed coordinate system as the axis of preferential orientation
of phenol molecules, and the angles φi were measured from
this direction in anticlockwise sense within the plane of the
molecule (Figure 9).
-
1
-1
9
50-980 cm region and another one near 810-820 cm .
The first two bands can be seen in the CCl4 solution spectrum
at 974 and 956 cm ) and even in the vapor spectrum (973
and probably 951 cm ), but they are too week to be studied in
-
1
(
-
1
the LC solution spectrum (the weak features seen in the 900-
-
1
9
90 cm region in Figure 8 are artifacts due to imperfect
compensation of the LC absorption bands). In the 800-830
-
1
cm region there are two close-lying bands clearly observed
in all phenol spectra; one of these has been assigned by all
previous investigators to an in-plane vibration while the other
to an out-of-plane vibration. According to our LD measure-
ments, however, both bands have the same dichroic behavior
with di ) 1.13, differing very significantly from that of the out-
of-plane bands. At the same time, we do not expect to have
two in-plane fundamentals here either. The only interpretation
consistent with all available data is that, on the one hand, the
The frequencies observed in the LC solution, dichroic ratios,
orientation factors, and transition moment directions determined
exp
calc
both experimentally (φi ) and from the DFT calculation (φi
)
are listed in Table 8.
Unfortunately, the sign of φi cannot be determined from eq
2, whereas there is no sign ambiguity in the theoretical
calc
calculation. Thus the signs of the calculated φi values were
exp
accepted for those of the measured φi values as well. For

1380 J. Phys. Chem. A, Vol. 102, No. 8, 1998
Keresztury et al.
The B3P86 method with the 6-311G** basis set gave us (after
fitting a few scale factors) a reliable force field that proved to
be transferable among the isotopomers for vibrational frequency
calculations and assignments. This way we proposed a more
reliable revised assignment of the vibrational spectra of phenol
molecule, based on PED and supported by measured and
calculated infrared absorption intensities and transition dipole
moment directions. On the basis of the latter, we advocate that
the symmetry of in-plane vibrations in phenol is reduced to such
an extent that there are no real grounds to classify the normal
modes under C2V symmetry species.
Similar work to solve the assignment problems of polyhy-
droxy benzenes such as hydroquinone are in progress.25 The
application of the scale factors obtained for phenol seems to be
feasible in these calculations.
Figure 10. Plot of measured vs DFT calculated vibrational transition
moment directions of phenol fundamentals (see data in Table 8).
the sake of easier comparison, the experimental and theoretical
transition moment directions are presented also in graphical form
in Figure 9A and B, respectively. The mean deviation between
the pairs of corresponding values is not very small (10.2°), but
at this stage we are unable to evaluate the relative errors of the
two sets of values (i.e. it is not known which of them is closer
to reality). Nevertheless, there are definite similarities between
the two patterns.
Acknowledgment. This work has been supported by re-
search grants T014479 (to G.K.), T014064 (to F.B.), and
T015756 (to M.K.) from the Hungarian National Research Fund
(OTKA) for which the authors are greatly indebted.
References and Notes
(1) Larsen, N. W. J. Mol. Struct. 1979, 51, 175.
(
2) Portalone, G.; Schultz, Gy.; Domenicano, A.; Hargittai, I. Chem.
Phys Lett. 1992, 197, 482.
3) Konschin, H. J. Mol. Struct. 1983, 92, 173.
(4) Puebla, C.; Ha, T.-K. J. Mol. Struct. (THEOCHEM) 1990, 204,
37.
(i) The transition moments are not grouped around two
(
mutually perpendicular directions (as should be in a molecule
having C2V symmetry) but are scattered very strongly in a wide
range of angles. The experimental and theoretical results also
agree in that there is a rather large angle interval (more than
3
(
5) Bock, C. W.; Trachtman, M.; George, P. J. Mol. Struct.
(
THEOCHEM) 1986, 139, 63.
(6) Bock, C. W.; Hargittai, I. Struct. Chem. 1994, 5, 307.
(7) Mecke, R.; Rossmy, G. Z. Elektrochemie 1955, 59, 866.
6
0°) where no transition moments have been found.
ii) The sequence of vibrations indicated by the serial numbers
(
(
8) Evans, J. C. Spectrochim. Acta 1960, 16, 1382.
9) Green, J. H. S. J. Chem. Soc. 1961, 2236.
(
of normal modes around the two circles in Figure 9 shows great
similarities, too (i.e. there seems to be a definite correlation
between the observed and calculated transition moment direc-
tions, which is depicted in Figure 10).
Unexpectedly, the data points in Figure 10 are best fitted with
a second-order polynomial instead of a straight line. The
reasons of the deviation of this correlation from linearity is the
subject of further studies involving IR-LD studies of phenol-d5
as well.24
All the above data and observations indicate that the overall
picture of the strongly asymmetric distribution of transition
moment directions must be essentially right. It allows to make
the conclusion that the symmetry of molecular vibrations of
phenol is Cs at most, and classification of the normal modes
under C2V symmetry species, even approximately, is not justified.
(
10) Bist, H. D.; Brand, J. D. C.; Williams, D. R. J. Mol. Spectrosc.
1966, 21, 766; J. Mol. Spectrosc. 1967, 24, 402.
11) Vars a´ nyi, G. The Assignment of Vibrational Spectra of 700 Benzene
DeriVatiVes; Academic Press: Budapest, 1973; p 77.
12) Larsen, N. W.; Nicolaisen, F. M. J. Mol. Struct. 1967, 22, 29.
(13) Hutt, K. W.; Butcher, R. J. J. Phys. C: Solid State Phys. 1988, 21,
013.
14) Wilson, H. W.; MacNamee, R. W.; Durig, J.R. J. Raman Spectrosc.
(
(
6
(
1
981, 11, 252.
(15) Kovner, M. A.; Davydova, N. I.; Zhigunova, I. A. Opt. Spektrosk.
1965,. 18, 152.
16) Sarin, V. N.; Rai, M. M.; Bist, H. D.; Khandelwal, D. P. Chem.
Phys. Lett. 1970, 6, 473.
17) Kubinyi, M.; Billes, F.; Grofcsik, A.; Keresztury, G. J. Mol. Struct.
1992, 266, 339.
(
(
(
(
18) Belhakem, M.; Jordanov, B. J. Mol. Struct. 1990, 218, 309.
19) Michl, J.; Thulstrup, E. W. Spectroscopy with Polarized Light.
Solute Alignment by Photoselection, in Liquid Crystals, Polymers, and
Membranes, VCH Publishers: New York, 1986; Chapter 5, pp 222-268.
(20) Becke, A.D. J. Chem. Phys. 1993, 98, 5648.
(
21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
V. Conclusion
Johnson, B. G.; Wong, M. W.; Foresman, J. B.; Robb, M. A.; Head-Gordon,
M.; Replogle, E. S.; Gomperts, R.; Andres, J. L.; Raghavachari, K.; Binkley,
J. S.; Gonzalez, C.; Martin, R. L.; Fox, D. J.; DeFrees, D. J.; Baker, J.;
Stewart, J. J. P.; Pople, J. A. Gaussian 92/DFT, Revision F.3; Gaussian,
Inc.: Pittsburgh, PA, 1993.
(22) Pulay, P.; Fogarasi, G.; Pang, F.; Boggs, J. E. J. Am. Chem. Soc.
979, 101, 2550.
The molecular structure and vibrational assignment of phenol
and its three main deuterated isotopomers have been re-
examined with the use of new experimental data and DFT
calculations. A planar structure with a distorted benzene ring
has been obtained from both the quantum chemical calculations
1
(
(
23) Fogarasi, G.; Pulay, P. J. Mol. Struct. 1977, 39, 275.
24) Keresztury, G.; Sundius, T. To be published.
(for the free molecule) and infrared linear dichroism measure-
ments (in LC solution).
(25) Billes, F.; Kubinyi, M. To be published.