Structure of 4-fluorophenol and barrier to internal -OH rotation in the S1-state

Article abstract of DOI:10.1039/b210188b

The structure and barrier to internal rotation of 4-fluorophenol in the ground state and the electronically excited S₁-state has been examined by resonantly enhanced two photon ionization spectroscopy and by rotationally resolved laser induced fluorescence spectroscopy of 4-fluorophenol and 4-fluorophenol-d₁. The rotationally resolved spectrum of the electronic origin of 4-fluorophenol is comprised of two subbands, which are split by 174.1 ± 0.5 MHz. From the splitting, determined from the HRLIF and several torsional bands observed in the R2PI spectrum an excited state barrier for the internal rotation of the hydroxy group of 1819.0 ± 5 cm^-1 was calculated. The subtorsional splitting of 4-fluorophenol-d₁ could not be resolved. The experimentally determined structural parameters from a fit to the rotational constants and the barrier to internal rotation in both electronic states are compared to the results of ab initio calculations. The molecule shows quinoidal distortion upon electronic excitation, with a shortening of both the C-O and the C-F bonds.

Full text of DOI:10.1039/b210188b

View Article Online / Journal Homepage / Table of Contents for this issue
Structure of 4-fluorophenol and barrier to internal –OH rotation in
the S -state
1
Christian Ratzer, Michael Nispel and Michael Schmitt*
Institut f u¨ r Physikalische Chemie, Universit a¨ tsstraße 26.43.02, D-40225 D u¨ sseldorf, Germany
Received 16th October 2002, Accepted 15th January 2003
First published as an Advance Article on the web 29th January 2003
The structure and barrier to internal rotation of 4-fluorophenol in the ground state and the electronically
excited S -state has been examined by resonantly enhanced two photon ionization spectroscopy and by
1
rotationally resolved laser induced fluorescence spectroscopy of 4-fluorophenol and 4-fluorophenol-d . The
1
rotationally resolved spectrum of the electronic origin of 4-fluorophenol is comprised of two subbands, which
are split by 174.1 ꢀ0.5 MHz. From the splitting, determined from the HRLIF and several torsional bands
observed in the R2PI spectrum an excited state barrier for the internal rotation of the hydroxy group of
ꢁ
1
1
819.0 ꢀ5 cm was calculated. The subtorsional splitting of 4-fluorophenol-d could not be resolved. The
1
experimentally determined structural parameters from a fit to the rotational constants and the barrier to
internal rotation in both electronic states are compared to the results of ab initio calculations. The molecule
shows quinoidal distortion upon electronic excitation, with a shortening of both the C–O and the C–F bonds.
1
Introduction
2 Experimental setup
The determination of the structure of aromatic molecules in
electronically excited states allows insight into electronic
effects upon electronic excitation. These electronic effects
show up in molecular properties like excited state lifetimes,
barriers to internal motions, and geometry changes in the
molecule. Comparison of microscopic data like the barriers
to internal rotation to macroscopic molecular properties like
pK -values of substituted aromatics might provide relations
a
for the straightforward determination of pK -values in differ-
a
ent electronic states. In a recent study we examined the struc-
ture and the barrier to internal rotation of the hydroxy group
of phenol in the first excited S -state. In order to study the
1
The experimental setup for the rotationally resolved LIF is
described elsewhere. Briefly, it consists of a ring dye laser
9
(Coherent 899-21) operated with Rhodamine 110, pumped
with 6 W of the 514 nm line of an Ar -ion laser (Coherent
Innova 100). The light is coupled into an external folded ring
+
1
0
cavity (LAS WaveTrain) for second harmonic generation
(SHG). The molecular beam machine consists of three differen-
tially pumped vacuum chambers that are linearly connected by
ꢁ1
skimmers. The expansion chamber is evacuated by a 8000 l s
oil diffusion pump (Leybold DI 8000), which is backed by a
3
ꢁ1
250 m
3
h
roots blower pump (Saskia RPS 250) and a
65 m h rotary pump (Leybold D65B). The second chamber
ꢁ1
ꢁ
1
serves as a buffer chamber and is pumped by a 400 l s turbo-
electronic nature of the barrier to internal rotation we intro-
duced substituents in the aromatic ring with various elec-
tronic influences. The first of these substances which we
investigated using rotationally resolved LIF spectroscopy
3
molecular pump (Leybold Turbovac 361), backed by a 40 m
h
ꢁ1
rotary pump (Leybold D40B), maintaining a chamber
ꢁ
5
pressure below 1 ꢂ10 mbar. The third chamber is pumped
1
ꢁ1
was 4-cyanophenol. If we divide the substituents into groups
by a 145 l s turbo-molecular pump (Leybold Turbovac
151) through a liquid nitrogen trap and backed by a 16 m
3
as known from the theory of electrophilic aromatic substitu-
tion, the cyanogroup is representative for substituents with an
electron withdrawing inductive (ꢁI) and mesomeric effect
ꢁ1
h
rotary pump (Leybold D16B) resulting in a vacuum better
ꢁ
6
than 1 ꢂ10 mbar. The molecular beam is crossed at right
angles with the laser beam 360 mm downstream of the nozzle.
The resulting fluorescence is collected perpendicular to the
plane defined by the laser and molecular beam by an imaging
optics setup consisting of a concave mirror and two plano-
convex lenses. The integrated molecular fluorescence is
detected by a photo-multiplier tube (Thorn EMI 9863QB)
whose output is discriminated and digitized by a photon
counter (Stanford Research Systems SR400) and transmitted
to a PC (Pentium II/233). Also the UV laser power, the iodine
absorption spectrum and the interferometer transmission sig-
nal are detected by photo diodes and stored with the spectrum.
The experimental setup for the resonance enhanced two-
(ꢁM).
The present study is concerned with 4-fluorophenol, in
which the fluorine atom exerts an electron withdrawing ꢁI-
effect too, but in contrast to 4-cyanophenol a +M-effect. The
aim of this study is to investigate the influence of these electro-
nic effects on the spectrum and especially on the torsional
barrier in both electronic states. The torsional barrier of 4-
fluorophenol in the electronic ground state has been deter-
ꢁ
1
2
mined by Larsen to be 1006 cm
using microwave and
far-infrared spectroscopy. Up to now no experimental value
for the S barrier has been reported to our knowledge.
3
1
Cvita sˇ et al. examined structural changes of several mono-
and disubstituted benzenes upon electronic excitation from
their rotational band contours in the electronic spectra.
1
1,12
photon ionization (R2PI) is explained in detail elsewhere.
Briefly, the apparatus consists of a source chamber pumped
with a 1000 l s oil diffusion pump (Alcatel) in which the
4
–7
8
ꢁ1
Christoffersen et al. fitted the rotational constants of 4-fluoro-
aniline and 4-fluorophenol to the band contour in the electro-
nic spectrum and reported the rotational constants of ground
and excited electronic states.
molecular beam is formed by expanding a mixture of helium
and 4-fluorophenol through the 500 mm orifice of a pulsed noz-
zle (General Valve, Iota One). The skimmed molecular beam
8
12
Phys. Chem. Chem. Phys., 2003, 5, 812–819
This journal is # The Owner Societies 2003
DOI: 10.1039/b210188b

View Article Online
(
Beam Dynamics Skimmer, 3 mm orifice) crosses the laser
asymmetric rotor in both electronic states as can be seen
from its k-value of ꢁ0.844 in the S -state and ꢁ0.85 S -state,
beams at right angles in the ionization chamber. The ions
are extracted in a gridless Wiley–McLaren type time-of-flight
(
0
1
cf. Table 1. The rotational temperature determined, taking
nuclear spin statistics into account (cf. section 3.2) is 1.3 K.
TOF) spectrometer (Bergmann Meßger a¨ te Entwicklung) per-
pendicular to the molecular beam and laser direction and enter
the third (drift) chamber where they are detected using multi
channel plates (Galileo). Ionization and drift chamber are both
pumped with a 150 l s rotatory pump (Leybold). The
vacuum in the three chambers with the molecular beam on
1
The S state lifetime of 4-fluorophenol was determined from
the Lorentzian contribution to the Voigt line shape to be
p1 .h8 eꢀno0l .. 1 ns. This is close to the lifetime determined for
Fig. 2 shows the rotationally resolved LIF spectrum of
ꢁ
1
ꢁ
3
ꢁ5
was 1 ꢂ10 mbar (source), 5 ꢂ10 mbar (ionization) and
the electronic origin of 4-fluorophenol-d
cm . The origin band of the deuterated isotopomer is
1
at 35 108.87
ꢁ7
ꢁ1
1
ꢂ10 mbar (drift), respectively. The resulting TOF signal
ꢁ
1
red-shifted by 7.4 cm , a value considerably larger than
was digitized by a 500 MHz oscilloscope (TDS 520A, Tektro-
nix) and transferred to a personal computer, where the TOF
spectrum was recorded and stored. The R2PI measurements
were carried out using the frequency doubled output of a
Nd:YAG (Spectra Physics, GCR170) pumped dye laser
ꢁ1
in phenol (1.8 cm ). The lifetime determined from the Lor-
entzian width of 49.7 MHz amounts to 3.2 ꢀ0.1 ns, which is
considerably shorter than the lifetimes obtained for all OD-
isotopomers of phenol. The torsional subbands could not be
resolved for the deuterated isotopomer. Because the tor-
sional splitting in the electronic ground state, determined
(
LAS, LDL205) operated with Fluorescein 27.
-Fluorophenol was purchased from Fluka and was used
without further purification. 4-Fluorophenol-d was prepared
4
2
by microwave spectroscopy amounts only to 1.16 MHz,
1
from 4-fluorophenol by refluxing with an excess of D O three
times with subsequent removal of the solvent. This procedure
resulted in an isotopic purity of > 95%.
2
we were not able to resolve the splitting, with a Gaussian
contribution to the line width of 20 MHz and a Lorentzian
width of 50 MHz.
3
Results and discussion
3
.1 Rotational constants
Fig. 1 shows part of the rotationally resolved LIF spectrum of
The inertial parameters of 4-fluorophenol and 4-fluorophenol-
d have been determined by a fit of the rovibronic transitions
1
1
~
1
˜
the electronic origin A
3
B
2
X
A
1
of 4-fluorophenol at
5 116.30 cm . The spectrum is comprised of two subbands,
which are of pure b-type and are separated by approximately
70 MHz. This splitting can be traced back to the hindered
ꢁ
1
to a rigid rotor Hamiltonian. The maximum J-value, which
could be detected in our jet-cooled spectrum is J ¼10. There-
fore inclusion of centrifugal distortion constants do not
improve the fit of the inertial parameters to our rovibronic
transitions. The ground state rotational constants have been
1
internal motion of the hydroxy group and will be discussed
in detail in section 3.2. 4-Fluorophenol is a near prolate
Fig. 1 Part of the rotationally resolved spectra of the electronic origin of 4-fluorophenol together with the simulation of the two torsional
+
ꢁ
subbands. The subbands are labeled with 0 for the s ¼0 component of the vibronic ground state and with 0 for the s ¼1 component.
Phys. Chem. Chem. Phys., 2003, 5, 812–819
813

View Article Online
ꢁ
1
Table 1 Molecular constants of 4-fluorophenol and 4-fluorophenol-
obtained from the fit to the experimental spectra. The number of
digits retained for each parameter were calculated according to the
phenol (4710 ꢀ30 cm ), we find a substantial reduction by
d
1
a factor of 2.6, while the S -barrier drops only by a factor
0
of 1.2.
24
scheme of Watson, so that the calculated line frequencies can be
reproduced within 10% of their standard deviation; the ground state
rotational constants have been obtained from a fit of the microwave
transitions from ref. 2 to a rigid rotor Hamiltonian
Using this barrier one can predict higher torsional transi-
tions and by a comparison to the vibronic spectrum further
refine the value for the barrier and include the torsional con-
stant in the fit. Fig. 3 shows the R2PI spectrum of 4-fluorophe-
ꢁ
1
nol in the range of 0–1500 cm , relative to the electronic
origin at 35 116.3 cm . The spectrum is similar to that already
4
-Fluorophenol
4-Fluorophenol-d
1
ꢁ
1
1
4
15
0
0
A /MHz
B /MHz
5625.594(8)
1454.6877(45)
1156.0073(39)
ꢁ0.0734(22)
ꢁ0.86634(3)
5268.0(1)
5584.315(7)
1414.340(3)
1128.758(3)
ꢁ0.0947
ꢁ0.87181
5231.546(42)
1433.3650(74)
1125.7998(47)
ꢁ0.277(2)
ꢁ0.85
published by Tembreull et al. and Fujimaka et al. but we
had to remeasure it in order to get the exact frequencies of
all rovibronic bands, not quoted in the cited literature.
00
0
0
C /MHz
2
ꢀ
ꢀ
˚
DI/u A
The allowed 2
0 transition is calculated between 730
ꢁ1
k
A /MHz
and 750 cm using the uncertainty of the torsional barrier
obtained from the high resolution experiment. In the vibronic
spectrum the only transition in this range is found at 736.6
0
0
B /MHz
0
1473.65(8)
1152.54(4)
ꢁ0.528(22)
ꢁ0.844
ꢁ1
C /MHz
2
cm which therefore can tentatively be assigned to this tor-
sional transition. Using the 0 /0 splitting together with the
˚
+
ꢁ
DI/u A
k
ꢀ
ꢀ
frequency of the 2
0 transition, we predict the frequency
ꢁ1
n~ /cm
1/2/ns
35 116.30(1)
1.8(1)
174.1(5)
44
35 108.87(1)
3.2(1)
+
+
0
ꢁ1
ꢁ
ꢁ
0 to be 1354
cm , in close agreement with experimental transitions at
of 4
to be 1341 cm
and of 4
t
ꢁ1
Dnsub/MHz
Lines
S(y)/MHz
–
ꢁ1
1342.6 and 1361.4 cm . Inclusion of the torsional constant
F into the fit resulted in the torsional transition frequencies
given in Table 2. The value of the barrier height from this fit
101
4.0
2.5
ꢁ
1
is 1819.0 ꢀ5 cm , for the torsional constant F a value of
83 ꢀ1 GHz is found, slightly smaller than in the electronic
ground state.
The G -symmetric 4-fluorophenol exhibits a 10:6 spin statis-
tic for K (even):K (odd) for transitions of the torsional ground
6
fit to experimental microwave transitionsyusing the same rigid
rotor Hamiltonian and have subsequently been kept fixed in
our fit of the rovibronic transitions. Because we want to eluci-
date geometry changes upon electronic excitation the use of
the same Hamiltonian for both electronic states is crucial.
The range of microwave transitions used in the fit of the
ground state rotational constants has been selected to match
the range observed in the jet-cooled LIF-spectrum (Jmax ꢃ10).
The deviations of the fit from a rigid rotor Hamiltonian com-
4
a
a
state s ¼0, and 6:10 for the torsionally excited state. In the fit
of the torsional bands shown in Fig. 1 these spin statistics have
been considered and reproduce the observed intensities at the
rotational temperature of 1.3 K well.
The complete list of observed frequencies and relative inten-
sities in the vibronic spectrum of 4-fluorophenol is given in
Table 3. The transitions were assigned tentatively on the basis
of ab initio calculations described in section 3.4.3.
2
pared to the fit given by Larsen, which includes four centri-
fugal distortion constants, is smaller than 0.02 MHz. This
value is considerably smaller than the uncertainties of our
excited state rotational constants. The A rotational constant
of 4-fluorophenol decreases by 357.6 MHz upon electronic
excitation, B increases by 18.9 MHz and C decreases by 3.5
3
.3 Determination of the structure
The structure of 4-fluorophenol in the S
0 1
- and S -state has been
8
MHz close to the values determined by Christoffersen et al.
from a rotational band contour analysis (ꢁ374.7 MHz,
obtained from the fit of a model geometry to the rotational
constants of 4-fluorophenol and 4-fluorophenol-d , using
1
1
6
+
21.9 MHz, ꢁ3.3 MHz). The changes of the rotational con-
the program pKrFit. Fig. 4 shows the model used for the
fit and the atomic numbering. The molecule is supposed to be
planar in both electronic states. Because para-disubstituted
stants of 4-fluorophenol-d
MHz, respectively. These constants will be used in section
1
are ꢁ352.8, +19.0, and ꢁ3.0
3
.3 to derive the structural changes upon electronic excitation.
aromatics show a quinoidal distortion upon electronic excita-
7
tion, we applied a model in which the C
1
C
2
bond length is
allowed to differ from C C . In addition the C F distance
2
3
4
3
.2 The vibronic spectrum
has been fit.
The CCC and CCH angles in the aromatic have been set to
+
ꢁ
The torsional splitting (0 /0 ) of 4-fluorophenol in the elec-
tronic ground state has been determined by microwave spec-
troscopy to be 177.121 MHz. Larsen and Nicolaisen
ꢄ
1
20 , the CH bond lengths to the benzene values and the CO
2
3
and OH bond lengths to the values of phenol. The result of
the fit of the r -structure to the rotational constants of both
isotopomers in the S - and S -state is shown in Table 4. In
ꢀ
ꢀ
+
+
determined the torsional transitions
1
0 ,
to be 279.1, 243.3, 247.1,
87.9, and 219.6 cm respectively. From these transitions
2
1 ,
0
ꢁ
ꢁ
+
+
ꢁ
ꢁ
2
2
1
1 , 3
2 , and 3
0
1
ꢁ
1
0
the S -state the geometry of the aromatic is benzenoid, with
+
ꢁ
equal bond lengths between C C and C C although they
1
and the 0 /0 splitting a barrier to internal rotation in the
electronic ground state of 1006 cm was calculated. This is
considerably lower than the ground state barrier in phenol
2
2 3
ꢁ
1
were not constrained equal in the fit. The C
was determined to be 133.5 pm. In the S -state the C C and
4
F bond length
1
1 2
ꢁ
1 13
)
ꢁ1
1
(
1215 cm
The line splitting observed in the LIF spectrum amounts to
and 4-cyanophenol (1420 cm ).
2 3
C C bond lengths are determined to be different, with a
marked quinoidal structure. The C F bond length decreases
4
by 6 pm upon electronic excitation. Both effects are reproduced
by the ab initio calculations, given for comparison in Table 4,
although less pronounced. Certainly the restriction of CO and
OH bond length to the values determined in phenol imposes a
model error on the structural determination, which exceeds the
propagated uncertainties of the rotational constants, given as
uncertainties in Table 4.
1
74.1 ꢀ0.5 MHz, cf. Table 1. Together with the selection rule
Ds ¼ꢀ1, the torsional level splitting in the electronically
excited state can be determined to be 3.0 ꢀ0.5 MHz. Using this
value and a torsional constant F of 690 GHz (as in the electro-
nic ground state), the torsional S -barrier is calculated to be
1
816 ꢀ40 cmꢁ. If we compare this value to the S
1
1
1
-barrier of
In addition, we determined the position of the hydroxylic H-
atom via a Kraitchman analysis. The moments of inertia which
yThe experimental frequencies of the microwave transitions from
ref. 2 have been made available to us by Prof. Larsen.
8
14
Phys. Chem. Chem. Phys., 2003, 5, 812–819

View Article Online
1
Fig. 2 Rotationally resolved spectra of the electronic origin of 4-fluorophenol-d together with a simulation.
are used in the Kraitchman equations of the planar 4-
fluorophenol are calculated from the A and B rotational
constants.
Inspection of Table 5 shows, that the value of the a-coordi-
nate of the hydroxylic H-atom of 4-fluorophenol decreases by
4.3 pm, and the b-coordinate slightly increases by 0.2 pm upon
electronic excitation, while for phenol a and b increase by 0.5
and 1.1 pm, respectively. Because the heavy fluorine atom is
located nearly at the a-axis, the very similar b-coordinate
observed is as expected.
ꢁ1
ꢁ1
Fig. 3 R2PI spectrum 4-fluorophenol in the range of 0–1500 cm , relative to the electronic origin at 35 116.3 cm . The bands marked with an
asterisk are the torsional bands, which have been used in the fit of the barrier (cf. text).
Phys. Chem. Chem. Phys., 2003, 5, 812–819
815

View Article Online
ꢁ
1
Table 2 Experimental and calculated torsional transitions (in cm
)
of 4-fluorophenol using a V barrier of 1819.0 cm and a torsional
ꢁ1
2
ꢁ1
constant of 22.81 cm (683 GHz)
ꢁ1
ꢁ1
Observed/cm
Calculated/cm
+
ꢁ
ꢀ
ꢀ
ꢀ
ꢀ
+
ꢁ
+
ꢁ
ꢀ
ꢀ
ꢀ
ꢀ
+
ꢁ
0
0
1
2
3
3
4
4
0
0
0
0
1
0
0
0
0
0
ꢁ4
ꢁ4
1.001 ꢂ10
1.001 ꢂ10
383.1
390.6
736.6
786.0
739.2
785.4
1066.0
1342.6
1361.4
1064.0
1341.6
1360.9
Fig. 4 Atomic numbering of 4-fluorophenol and model used for the
fit of the geometry.
as the difference of CIS and HF calculations due to a favorable
compensation of errors.
The position of the hydroxylic H-atom in the inertial system,
obtained from the MP2/6-311++G(d,p) calculation is a ¼
315.7 pm and b ¼83.6 pm, close to the values determined
from the Kraitchman analysis (312.8 and 79.38 pm). The posi-
3
.4 Comparison to the results of ab initio calculations
.4.1 Structure. The structure of 4-fluorophenol in the
electronic ground state has been calculated at the HF/6-
3
3
3
11++G(d,p), MP2/6-
11++G(d,p) level and at the CIS/6-311++G(d,p) level for
B3LYP/6-311++G(d,p),
and
the electronically excited S -state with the Gaussian 98 pro-
1
tion in the S state calculated at the CIS/6-311++G(d,p) level
1
1
gram package. The SCF convergence criterion used for our
7
is a ¼312.1 pm and b ¼81.9 pm, reflecting the decrease of the
ꢁ
8
calculations was an energy change below 10 E , while the
h
a-coordinate found in the experiment.
convergence criterion for the gradient optimization of the
ꢁ
5
molecular geometry was @E/@r > 1.5 ꢂ10 Eh a0 and @E/
ꢁ
5
ꢁ1
@
j > 1.5 ꢂ10
E
Table 6 gives the geometry parameters, obtained at various
h
degree , respectively.
3.4.2 Barrier to internal rotation. The transition state for
the internal rotation, in which the OH-group is perpendicular
relative to the aromatic ring, was optimized by the synchro-
levels of theory. All structures have been proven to be minima
via a normal mode analysis based on the analytical gradients.
Table 7 compares the experimental rotational constants with
the results of the ab initio calculations. The agreement of the
experimental ground state constants with the results of the
MP2 and B3LYP calculations is very good, while HF fails to
describe the geometry correctly. The excited state rotational
constants are reproduced poorly at the CIS level, as could be
expected, regarding the deviations of the HF calculations for
the ground state. Nevertheless, the change of the rotational
constants upon electronic excitation is reproduced very well
1
8,19
nous transit-guided quasi-Newton method (STQN),
imple-
1
7
mented in the Gaussian 98 program package. A normal
mode analysis has been performed utilizing the analytical sec-
ond derivatives of the potential energy surface both at the
ground state and at the transition state. Table 8 gives the cal-
culated barrier to internal rotation as the difference of the fully
optimized equilibrium and transition state energies for the S0
0
Table 4 Experimental r -geometry parameters (distances in pm and
angles in degrees) of 4-fluorophenol, obtained from a fit of the geome-
try to the experimental inertial parameters. Only the parameters with
an asterisk have been optimized in the fit procedure; the ab initio geo-
metry parameters, obtained with the 6-311G++(d,p) basis are given
for comparison
Table 3 Experimental vibronic frequencies and relative intensities of
ꢁ1
4-fluorophenol relative to the electronic origin at 35 116.3 cm ; for the
description of the modes see section 3.4.3
pKrFit
MP2
pKrFit
CIS
ꢁ1
Transition
Frequency/cm
Relative intensity
S
0
S
1
Origin
2
0
100
11
25
47
6
C
C C *
1
C
2
*
139.89(9)
139.89(7)
133.56(13)
133.15
96.6
139.74
139.83
135.88
137.38
96.52
108.1
120
145.18(3)
139.5(50)
127.6(90)
128.15
96.6
142.54
141.55
131.47
131.74
94.7
107.1
120
16a
242.2
340.4
422.2
0
2
3
d(C–F)
C
C O
4
F*
1
0
6a
1
7
7
08.8
20.8
OH
20
7
C
C
ar
H
108.1
120
120
107.1
120
120
ꢀ
ꢀ
2
1
1
1
6
0
736.6
768.2
818.6
827.6
834.8
ar ar ar
C C
1
0
8a
20
93
89
12
31
21
10
10
16
8
ar ar
C C H
C C F
120
120
120
120
1
0
3
4
120
120
120
120
1
2
0
C C O
2 1
120
110
120
111
2
0
a
C OH
1
110
110
8
8
49.8
62.0
hobs ꢁcalci/MHz
0.26
–
0.47
–
1
1
140.0
162.0
1
a
9
1170.4
209.2
0
Table 5 Substitution coordinates (in pm) of the H-atom in 4-fluoro-
phenol and phenol for both electronic states
1
3
a
6
1
1231.6
1240.6
1249.0
37
79
87
14
66
4
0
1
0
9a
S0
S1
d(CO–H)
1
271.4
a
b
a
b
n(C–O)
1285.4
1342.6
1361.4
ꢁ
+
ꢁ
+
4
4
0
0
4-Fluorophenol
Phenol
312.84(9)
258.6(1)
79.38(12)
85.3(2)
308.51(10)
259.1(2)
79.60(12)
86.4(3)
4
8
16
Phys. Chem. Chem. Phys., 2003, 5, 812–819

View Article Online
Table 6 Ab initio structural parameters of 4-fluorophenol; rotational constants are given in MHz, distances in pm, angles and dihedral angles
in degrees
6
-31G(d,p)
6-311++G(d,p)
HF
HF
B3LYP
MP2
CIS
B3LYP
MP2
CIS
A/MHz
B/MHz
C/MHz
5736
1475
5627
1447
5623
1447
5396
1499
5745
1477
5646
1449
5607
1446
5397
1505
1173
1151
1151
1173
1175
1153
1149
1177
C
C
C
C
C
C
C
C
C
C
C
1
2
3
4
5
2
3
5
6
1
4
C
C
C
C
C
2
3
4
5
6
138.43
138.78
137.35
138.00
138.10
107.69
107.41
107.42
107.41
135.32
133.40
94.25
139.86
139.60
138.80
139.11
139.23
108.76
108.44
108.45
108.45
136.88
135.30
96.59
139.74
139.64
138.78
139.07
139.26
108.41
108.10
108.10
108.14
137.38
135.88
96.52
141.85
140.77
140.47
140.12
140.92
107.42
107.15
107.12
107.16
132.62
131.81
94.45
138.32
138.81
137.20
137.86
138.13
107.67
107.41
107.42
107.40
133.12
135.29
94.04
139.54
139.48
138.38
138.70
139.10
108.56
108.27
108.28
108.29
137.07
135.91
96.26
139.99
140.04
138.98
139.24
139.69
108.80
108.51
108.52
108.55
137.14
135.29
96.24
142.23
140.37
140.43
140.22
140.50
107.40
107.17
107.14
107.18
131.98
131.05
94.30
H
H
H
H
O
F
2
3
5
6
OH
C
C
C
C
C
C
C
C
C
1
C
2
C
3
C
4
C
1
C
2
C
3
C
4
C
5
C
2
C
C
C
C
3
4
5
6
120.1896
119.0720
121.5544
119.2368
120.2133
121.1419
119.3262
119.5786
120.9520
120.1994
119.0209
121.5997
119.2213
120.0986
121.2501
119.2218
119.5463
121.0110
120.0631
118.8672
121.8908
119.0029
120.1923
121.4084
119.0667
119.5963
121.1353
118.2500
117.7106
124.8907
117.4762
120.2286
122.7545
117.3608
119.7497
122.5505
120.1936
119.0556
121.5687
119.2413
120.2685
121.0189
119.3197
119.6962
120.7987
120.1558
118.8458
121.9024
119.0585
119.6325
121.2039
119.0646
119.7684
120.8678
120.1893
118.7677
121.9114
118.9800
120.1512
121.3014
119.0487
119.7334
120.8518
118.205
117.8879
124.7542
117.5073
120.1345
122.6564
117.4726
119.7471
122.4492
3
4
5
2
3
4
5
6
H
H
F
2
3
H
H
5
6
(
HF, B3LYP, MP2). Barriers for phenol and 4-cyanophenol
energy between equilibrium geometry and transition state
ꢁ
1
are given for comparison. The value of the torsional barrier
depends critically on the truncation of the basis set and the
order of the perturbation correction to the SCF energy as
has been shown by Kim and Jordan for phenol. As one
can see from Table 8, the more flexible the basis, the lower
is the barrier given one method. Therefore we also employed
amounts to 15–20 cm the zero-point energy of the transition
state being smaller than of the equilibrium structure. Therefore
this effect tends to raise the calculated barrier relative to the
experimental one. A third effect on the value of the barrier,
2
0
2
0
which was considered by Kim and Jordan is the neglect of
higher terms in the Fourier expansion:
2
1
a complete basis set method (CBS) combined with higher
order corrections for the perturbation series, as described by
1
X
V2n
VðfÞ¼
½1 ꢁcosð2nfÞꢅ
ð1Þ
2
2
Ochterski et al. (CBS-Q) to calculate the ground state barrier
of 4-fluorophenol. The best agreement between experiment
and theory can be found for B3LYP/6-311++G(d,p) and
the CBS-Q model chemistry. Interestingly, the theoretical
value is larger than the calculated barrier using CBS-Q. In
the case of phenol Kim and Jordan found a theoretical bar-
rier at the QCISD(T)/6-31++G(2df,p) level of theory which is
1
this difference mainly to different definitions of the barrier.
While the experimental barrier is determined assuming a rigid
frame and rotor model, both geometries are allowed to relax in
the determination of the theoretical barrier. This should lower
the calculated barrier relative to the experimental value. Kim
2
n¼1
in the determination of the experimental barrier. If we intro-
duce a V -term into the fit of the vibronic transitions, we
4
ꢁ
1
obtain a V -barrier only 10 cm lower and a V -term of 26
2
4
2
0
ꢁ1
cm . Thus, assessing the deviations between the experimental
and the theoretical barrier, by far the largest effect is the
relaxation of all molecular coordinates other than the torsion
in the course of the calculation. Zero-point effects, and the
inclusion of higher terms in the Fourier expansion (1) lead to
errors of about one order of magnitude less.
ꢁ1
31 cm smaller than the experimental value. They attributed
The electronically excited S state has been calculated at the
1
CIS and time dependent (TD) DFT level. The excited state
barrier is determined from the results of the CIS and TD-
DFT calculations via:
2
0
ꢁ1
and Jordan estimated this difference to be about 70 cm .
Another difference is the inclusion of zero-point vibrational
effects from coordinates other than the torsion in the experi-
mentally determined value. We performed a normal mode ana-
lysis for each level of theory given in Table 8. Independent of
method and basis set applied, the difference in zero-point
½
EequilibriumðRHFÞþEequilibriumðexcitationÞꢅ
ꢁ½Etrans stateðRHFÞþEtrans stateðexcitationÞꢅ ð2Þ
where E(excitation) is the excitation energy of the lowest
excited singlet state for the equilibrium structure and the tran-
sition state, respectively and E(RHF) are the respective Har-
tree–Fock energies. Regarding the fact that HF reproduces
the ground state barriers very poorly, it cannot be expected,
Table 7 Comparison of ab initio rotational constants of 4-fluorophe-
nol, obtained with the 6-311++G(d,p) basis set with the experiment
Exp.
)
Exp.
HF B3LYP MP2 (S
Exp.
CIS (S
ꢁS
1
that CIS gives much better values for the S state. Neverthe-
(
S
0
1
)
1
0
)
CIS ꢁHF
less, an increase of the barrier height upon electronic excitation
is found in the calculation, although the absolute value shows
very large deviations. Also the barrier calculated at the time
dependent DFT level shows unacceptable large deviation from
the experiment.
A
B
C
5625.5 5745 5646
1454.7 1477 1449
1156.0 1175 1153
5623 5268.0 5397 ꢁ357.6 ꢁ348
1447 1473.6 1505 +18.9
1151 1152.5 1177
ꢁ3.5
+28
+2
Phys. Chem. Chem. Phys., 2003, 5, 812–819
817

View Article Online
Table 8 Experimental and calculated barriers to internal rotation of 4-fluorophenol, phenol, and 4-cyanophenol
-Fluorophenol Phenol
4
4-Cyanophenol
State
S
Method
V
2,e
V
2,0
V
2,e
V
2,0
V
2,e
V
2,0
0
Experiment
1006
652
–
1215
905
–
1420
1192
1127
1682
1508
1393
1205
–
HF/6-31G(d,p)
HF/6-311++G(d,p)
B3LYP/6-31G(d,p)
B3LYP/6-311++G(d,p)
MP2/6-31G(d,p)
MP2/6-311++G(d,p)
CBS-Q
486
463
969
870
869
443
881
716
691
1169
1054
1028
965
–
963
931
1426
1297
1187
1080
–
596
846
1176
1030
1036
843
1387
1224
1209
1033
–
1054
1820
2337
2180
3490
–
S
1
Experiment
–
4710
2061
2248
3211
–
ꢆ5000
2133
2248
2714
–
2226
CIS/6-31G(d,p)
2684
B3LYP-TD/6-311++G(d,p)
2178
2354
–
1978
2140
–
CIS/6-311++G(d,p)
–
3.4.3 Vibrations. Table 9 shows the harmonic frequencies,
calculated at the CIS/6-311++G(d,p) level of theory. The des-
ignation of the modes follows the nomenclature of Varsanyi.
The torsional mode t mixes strongly to several modes, in par-
ticular to the low frequency out of plane modes 16a, 16b, and
fluorescence data, using the propensity rule, as an additional
guidance in the assignment.
2
3
4
Conclusion
1
1. The Varsanyi modes 15 and 9b are indistinguishable due to
the similar masses of the fluorine and the hydroxy substituent,
one being the C–F bending mode, the other the C–O bending
mode. Comparison of the calculated harmonic modes with the
experimental frequencies leads to the assignment of the vibro-
nic transitions given in Table 3. This very tentative assignment
for the vibronic bands has to be checked against dispersed
The change of the structure of 4-fluorophenol upon electro-
nic excitation was investigated by high resolution LIF spec-
troscopy of two isotopomers of the compound. Both
experiment and ab initio theory predict a benzenoid structure
in the electronic ground state, but a quinoidal distortion
along the long molecular axis in the S1 state. The CF and
CO bond lengths were both found to decrease upon electro-
nic excitation. The change of the inertial parameters upon
electronic excitation can be predicted with high accuracy
from the difference of the CIS and the HF rotational con-
stants. Although the close agreement between experiment
and theory is mainly due to a cancelation of errors, it proves
very helpful to get a first guess for the rotational constants in
Table 9 Calculated vibrational frequencies of 4-fluorophenol in the
state
S
0
and the S
1
Number
Mode
CIS 6-311++G(d,p)
1
2
3
4
5
6
7
8
9
16a
11
o.o.p
o.o.p
o.o.p
o.o.p.
i.p.
95
143
the S state.
1
The barrier to internal rotation of the hydroxy group in
-fluorophenol increases from 1006 cm in the S -state to
16b [t(OH)]
t(OH) [16b]
d(C–F) [15]
d(C–OH) [9b]
6a
299
315
ꢁ1
4
1
0
ꢁ1
819 cm
1
in the electronically excited S -state, while the
387
449
barriers in phenol increase from 1215 to 4710 and in 4-fluoro-
ꢁ1
i.p.
i.p.
phenol from 1420 to ꢆ5000 cm . The increase in barrier
466
545
height can be attributed to a transfer of electron density
from the hydroxy group into the aromatic ring upon electro-
nic excitation, which is reflected in the shorter bond length
17b
10a
o.o.p.
o.o.p.
o.o.p.
i.p.
563
613
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
17a
6b
1
and the expansion of the aromatic ring in the S state. A
621
623
partially quinoidal structure is formed in the S1 state with
an increased bond order between the hydroxylic oxygen
and the ring carbon atom. The same holds true for phenol
and for 4-cyanophenol, but the resulting negative partial
charge in 4-fluorophenol cannot be efficiently delocalized
as in the case of 4-cyanophenol due to the +M effect of
the fluorine substituent. Therefore the increase in barrier
height upon electronic excitation in 4-fluorophenol is the
smallest of the three phenols.
4
10b
o.o.p.
o.o.p.
o.o.p.
i.p.
760
768
5
18a
802
910
1
12
i.p.
i.p.
1067
1137
1224
1239
1379
1404
1436
1487
1551
1625
1769
1831
3365
3400
3409
3417
4172
18b
d(CO–H)
9a
i.p.
i.p.
i.p.
i.p.
n(C–F)
19b
i.p.
i.p.
n(C–O)
3
8b
i.p.
i.p.
Acknowledgements
19a
8a
i.p.
i.p.
This work was made possible by the support of the Deutsche
Forschungsgemeinschaft (SCHM 1043/9-2). We gratefully
acknowledge the support of Prof. Dr. Karl Kleinermanns.
We thank Prof. N. W. Larsen for placing the frequencies of
the microwave transitions of 4-fluorophenol at our disposal.
The help of Petra Imhof in recognizing the Varsanyi modes
and of Isabel H u¨ nig for experimental help with the R2PI
spectrum are gratefully acknowledged.
14
i.p.
i.p.
n(C–H)
n(C–H)
n(C–H)
n(C–H)
n(O–H)
i.p.
i.p.
i.p.
i.p.
8
18
Phys. Chem. Chem. Phys., 2003, 5, 812–819

View Article Online
1
6
C. Ratzer, J. K u¨ pper, D. Spangenberg and M. Schmitt, Chem.
Phys., 2002, 283, 153.
References
1
J. K u¨ pper, M. Schmitt and K. Kleinermanns, Phys. Chem. Chem.
Phys., 2002, 4, 4634.
N. W. Larsen, J. Mol. Struct., 1986, 144, 83.
N. W. Larsen and F. M. Nicolaisen, J. Mol. Struct., 1974,
17 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery Jr.,
R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D.
Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V.
Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C.
Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala,
Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G.
Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S.
Replogle and J. A. Pople, Gaussian 98 Revision A.7, Gaussian,
Inc., Pittsburgh, PA, 1998.
2
3
22, 29.
4
5
6
7
8
T. Cvita sˇ and J. Hollas, Mol. Phys., 1970, 18, 101.
T. Cvita sˇ and J. Hollas, Mol. Phys., 1970, 18, 793.
T. Cvita sˇ and J. Hollas, Mol. Phys., 1970, 18, 801.
T. Cvita sˇ , J. Hollas and G. Kirby, Mol. Phys., 1970, 19, 305.
J. Christoffersen, J. M. Hollas and G. H. Kirby, Mol. Phys., 1970,
18, 451.
9
0
1
2
3
4
5
M. Schmitt, J. K u¨ pper, D. Spangenberg and A. Westphal, Chem.
Phys., 2000, 254, 349.
M. Okruss, R. M u¨ ller and A. Hese, J. Mol. Spectrosc., 1999, 193,
1
1
1
1
1
1
2
93.
M. Schmitt, C. Jacoby and K. Kleinermanns, J. Chem. Phys.,
998, 108, 4486.
18 C. Peng, P. Y. Ayala, H. B. Schlegel and M. J. Frisch, J. Comput.
Chem., 1996, 17, 49.
1
W. Roth, C. Jacoby, A. Westphal and M. Schmitt, J. Phys. Chem.
A, 1998, 102, 3048.
G. Berden, W. L. Meerts, M. Schmitt and K. Kleinermanns,
J. Chem. Phys., 1996, 104, 972.
R. Tembreull, T. M. Dunn and D. M. Lubman, Spectrochim. Acta
A, 1986, 42, 899.
E. Fujimaki, A. Fujii, T. Ebata and N. Mikami, J. Chem. Phys.,
19 C. Peng and H. B. Schlegel, Isr. J. Chem., 1994, 33, 449.
20 K. Kim and K. D. Jordan, Chem. Phys. Lett., 1994, 218, 261.
21 M. R. Nyden and G. A. Petersson, J. Chem. Phys., 1981, 75, 1843.
22 J. W. Ochterski, G. A. Petersson and J. A. Montgomery, J. Chem.
Phys., 1996, 104, 2598.
23 G. Varsanyi, Assignments for Vibrational Spectra of 700 Benzene
Derivatives, Wiley, New York, 1974.
24 J. K. G. Watson, J. Mol. Spectrosc., 1977, 66, 500.
1999, 110, 4238.
Phys. Chem. Chem. Phys., 2003, 5, 812–819
819