Specific solvent effects on the structure and reaction dynamics of benzophenone ketyl radical

Article abstract of DOI:10.1021/jp0121320

Complex formation reaction between benzophenone ketyl (BPK) radical and triethylamine was investigated by means of transient absorption and time-resolved ESR methods. The formation constants (K_f's) in benzene, toluene, o-xylene, and anisole were measured by a transient absorption method. The enthalpy and entropy changes upon the complex formation were determined by van't Hoff plots of K_f values around 300 K. The ΔG values in the complex formation reaction at 298 K were calculated from these thermodynamic parameters, which show notable difference from those in aliphatic solvents and chlorotoluenes previously reported. The BPK conformation in these solvents was investigated by analyzing the time-resolved ESR spectra of BPK and by the DFT calculation. The hyperfine coupling constant of hydroxy-proton (a_OH) was found to depend on solvent as well and there exists a clear correlation between the a_OH and K_f (or ΔG) values. It was concluded that the conformation of BPK, especially the conformation between OH and two phenyl groups, depends on the solvation and is an important factor to control the complex formation reaction.

Full text of DOI:10.1021/jp0121320

9628
J. Phys. Chem. A 2001, 105, 9628-9636
ARTICLES
Specific Solvent Effects on the Structure and Reaction Dynamics of Benzophenone Ketyl
Radical
†
Akio Kawai, Makoto Hirakawa, Toyohiko Abe, Kinichi Obi, and Kazuhiko Shibuya*
Department of Chemistry, Graduate School of Science and Engineering, Tokyo Institute of Technology,
2
-12-1 Ohokayama, Meguro-ku, Tokyo 152-8551, Japan
ReceiVed: June 5, 2001
Complex formation reaction between benzophenone ketyl (BPK) radical and triethylamine was investigated
by means of transient absorption and time-resolved ESR methods. The formation constants (K
toluene, o-xylene, and anisole were measured by a transient absorption method. The enthalpy and entropy
changes upon the complex formation were determined by van’t Hoff plots of K values around 300 K. The
G values in the complex formation reaction at 298 K were calculated from these thermodynamic parameters,
f
’s) in benzene,
f
∆
which show notable difference from those in aliphatic solvents and chlorotoluenes previously reported. The
BPK conformation in these solvents was investigated by analyzing the time-resolved ESR spectra of BPK
and by the DFT calculation. The hyperfine coupling constant of hydroxy-proton (aOH) was found to depend
on solvent as well and there exists a clear correlation between the aOH and K (or ∆G) values. It was concluded
f
that the conformation of BPK, especially the conformation between OH and two phenyl groups, depends on
the solvation and is an important factor to control the complex formation reaction.
Introduction
and the complex obtained by analysis of their ESR spectra.10,11
In the present study, we have focused our attention on the
Benzophenone ketyl (BPK) radical is one of the important
intermediates in the well-known photoreduction of benzophe-
none (BP) and its spectroscopic property and photochemistry
were widely studied.1-6 BPK has the D1-D0 absorption starting
at 545 nm and the D1 state at room-temperature fluoresces with
the quantum yield of 0.16 which is higher than those of other
aromatic ketyl radicals. The ESR spectrum of BPK shows
significant dependence on the solvent, which was understood
by considering slightly flexible internal motions of two phenyl
conformation of BPK in various solvents and revealed the
solvent dependence of Kf and ∆G values for the complex
formation reaction. From the analysis of ESR spectra, we
propose a few different conformations of uncoupled BPK
stabilized by the solvent cage. The different conformations of
OH and two phenyl rings of BPK are essential to distinguish
them. The conformations of OH and two phenyl rings were
derived from the hfcc of hydroxy-proton (aOH) of uncoupled
BPK. In relatively nonpolar solvents, the hydroxy-proton sits
3
7
and OH groups in BPK. The complex formation reaction of
1
2
almost on the molecular plane of BPK. In alcohol, the
hydroxy-proton does not seem to locate on the molecular plane.
The structure of the BPK-TEA complex is also derived from
ESR spectra and it is found that the hydroxy-proton locates far
over the molecular plane. A good correlation between the ∆G
value for the complex formation reaction and the aOH value of
uncoupled BPK is recognized. This correlation suggests that
the conformation of uncoupled BPK is strongly related to the
BPK with TEA has recently been studied by a transient
absorption method in various solvents and the thermodynamic
analysis was carried out to reveal the complex formation
dynamics.8 According to the previous reports,8-10 the BPK-
TEA complex is formed through hydrogen-bonding interaction
between the OH and amino groups of BPK and TEA, respec-
tively, and the formation constant (Kf) significantly depends on
the dielectric constant of solvent. The solvent dependence of
Kf values was explained by introducing solvation energy which
was calculated based on the reaction field model for solvation.
Since BPK and the BPK-TEA complex are open shell species,
ESR spectroscopy is a powerful method to study these radicals
selectively. Hyperfine coupling constants (hfcc) of ketyl-type
radicals are known to be sensitive to the conformation of radical
and we could describe the complex formation of ketyl radical
through the conformational difference between uncoupled ketyl
,9
∆
G value. The complex formation in the BPK-TEA system will
be widely discussed on the basis of the solvent dependence of
G and the conformation of uncoupled BPK and the complex
∆
in each solvent derived from ESR spectra.
Experimental Section
Time-Resolved (TR) ESR Measurement. The TR-ESR
detection system consists of an X-band ESR spectrometer
(Varian E-112) and a boxcar integrator (Stanford SR-250) to
extract transient ESR signals synchronized with laser pulses.
The excitation light source was a XeCl excimer laser (Lambda
Physik LPX 100, 308 nm) of 20 ns duration time and of 100
mJ/pulse laser power. The laser pulse of about 20 mJ was
*
Corresponding author.
Present address: Department of Chemical and Biological Sciences,
†
Japan Women’s University, 2-8-1 Mejirodai, Bunkyo-ku, Tokyo 112-8681,
Japan.
1
0.1021/jp0121320 CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/02/2001

Solvent Effects on Benzophenone Ketyl Radical
J. Phys. Chem. A, Vol. 105, No. 42, 2001 9629
Figure 1. Transient absorption spectra of BPK and the BPK-TEA
complex in toluene measured at 0.8-1.4 µs after 355 nm laser excitation
at 295 K. The concentrations of TEA were 0, 11.8, 21.7, 57.8, and
Figure 2. Van’t Hoff plots of the complex formation reaction in
benzene, toluene, o-xylene, or anisole.
1
08.4 mM. The inset shows fitting curves of eq 1 for absorbance
changes against TEA concentration at various wavelengths.
[
Com]. The R was kept constant during a series of measure-
ments. Hence, the final form of Abs(λ) is given by the eq 1:
introduced into a photoreaction cell in the microwave cavity.
The gate of the boxcar integrator was opened from 1.5 to 2.0
µs after the laser pulse. The signals were collected at the
repetition rate of 10 Hz. The microwave power was 15 mW.
Transient Absorption Measurement. The transient absorp-
ꢀk
K [TEA] +
f
ꢀC
Abs(λ) ) RꢀC
(1)
1
+ K [TEA]
f
1
3
tion detection system has already been described elsewhere.
3
+
To determine the Kf value from eq 1, we measured Abs(λ) at
five wavelengths against TEA concentration. The Kf value for
the BPK/TEA system was determined from the least-squares
fit of the eq 1 as shown in the inset of Figure 1 using the molar
extinction coefficient, ꢀk ) 4600 M cm at the BPK
absorption peak (545 nm) and ꢀc value reported previously.
The averaged value of five Kf’s (17.1, 25.8, 30.8, 28.0, 30.3
The excitation light source was the third harmonics of a Nd -
YAG laser (Continuum Powerlight 8010, 8 ns duration time)
and the laser power was attenuated to 10 mJ or less at the
reaction cell. The laser was operated at Q-switch repetition
frequency of 1 Hz under 10 Hz flash lamp operation by using
a homemade frequency control unit for Q-switch.
Samples. All the samples, except R-phenylbenzoin, were
purchased from Tokyo Kasei and mostly used as received.
Benzophenone was recrystallized twice from n-heptane. R-Phe-
nylbenzoin was synthesized by the Grignard reaction of phe-
nylmagnesium bromide with benzil in tetrahydrofurane at 245
K and recrystallized from n-hexane. Sample solutions were
degassed by bubbling of nitrogen or Ar gas and were flowed
through the cell. Sample cells were (1) a quartz flat cell with
-
1
-1
1
4
8
-
1
-1
M
at 295 K) was adopted as the final Kf value (26.4 M ).
Similar experiments were carried out to determine Kf values in
benzene, o-xylene, or anisole. Thermodynamic parameters,
namely enthalpy and entropy changes (∆H and ∆S, respectively)
of the complex formation reaction in benzene, toluene, o-xylene,
or anisole were determined by the van’t Hoff plot of dimension-
Φ
Φ
Φ
less formation constants, Kf values () Kf × c , where c is
3
the standard concentration, 1 mol/dm ) at different temperatures
0
.5 mm optical path length attached in the microwave cavity
around 300 K as shown in Figure 2. From the slope and intercept
values of four fitted lines in Figure 2, the ∆H and ∆S values
were determined and are listed in Table 1. According to these
values, ∆G values at 298 K were calculated by the equation,
for the TR-ESR measurement and (2) a quartz cell with optical
path lengths of 3 mm for excitation laser and 10 mm for monitor
lights, respectively, for the transient absorption measurement.
∆
G ) ∆H - T∆S and were listed in Table 1. The Kf values
Results and Discussion
and thermodynamic parameters in other solvents reported
previously are also summarized in Table 1.
8
,9
Thermodynamic Parameters of Complex Formation Re-
action. Figure 1 shows a series of transient absorption spectra
of BPK obtained by the laser flash photolysis of benzophenone
in toluene with different TEA concentrations. The absorption
spectrum characteristic of BPK in the D1-D0 transition region
is clearly seen in the spectrum without TEA, while the spectra
measured in the presence of TEA consist of those due to
uncoupled BPK and the BPK-TEA complex with an isosbestic
point at 550 nm. The complex formation proceeds according
to an equilibrium reaction; BPK + TEA a complex. From these
spectra, we determined the formation constant (Kf) in toluene.
The details of the procedures to determine a Kf value have been
described in the previous papers.8 The Kf value follows the
equation, Kf ) [Com]/[BPK][TEA], where Com represents the
complex. Absorbance at a wavelength λ is given by Abs(λ) )
The transient absorption spectra of BPK with different TEA
concentrations were also measured in tetrahydrofurane and in
methanol as shown in Figure 3. In both systems, an absorption
band around 550 nm appears in the spectra, which suggests the
formation of BPK. In tetrahydrofuran, no drastic red-shift was
seen even under the very high TEA concentration (361 mM).
Therefore, we consider that the Kf vlaue in tetrahydrofurane
must be very small compared to those in benzene, toluene,
o-xylene, or anisole. It is noteworthy that BPK yield (BPK
absorbance) seems to decrease when the TEA concentration
becomes higher. It has been known that BP forms a complex
with amines and that the rapid S1 quenching occurs in the BP-
TEA complex. This may be why the relative BPK yield
slightly depends on the TEA concentration. In methanol, it has
been reported that BP anion as well as BPK is formed in the
photoreduction of BP with TEA. A largely red-shifted broad
band appearing around 625 nm in the spectra is attributed to
,9
1
5
ꢀ
k[BPK] + ꢀc[Com] where ꢀk and ꢀc are the molar extinction
1
6
coefficients of BPK and the complex, respectively. We introduce
a constant R, which equals to the concentration of [BPK] +

9630 J. Phys. Chem. A, Vol. 105, No. 42, 2001
Kawai et al.
TABLE 1: Thermodynamic Parameters for Complex Formation Reaction at 298 K and Hyperfine Coupling Constants of BPK
in Various Solvents
c
ηa
cp
hfcc G
K
M
f
-1
∆G
kcal mol
∆H
kcal mol
∆S
b
b
-1
-1
cal K-1 mol-1
group
Ad
solvent
n-hexane
n-heptane
cyclohexane
o-chlorotoluene
p-chlorotoluene
ꢀ
r
n
D
a
OH
a
m
a
p
a
o
0.30 1.89 1.37486 103
0.39 1.92 1.38764 101
0.89 2.02 1.42623 99
-2.70
-2.69
-2.68
-2.57
-2.47
-2.38
-6.6
-7.5
-4.1
-3.5
-2.0
-13.2
-16.5
-5.2
-3.4
+1.3
2.08
(0.04
1.23
(0.03
3.69
(0.05
3.27
(0.04
0.96 4.45 1.5238
0.84 6.08 1.5199
83
70
60
1
,2-dichloroethane 0.78 10.7 1.4448
B
benzene
0.60 2.28 1.50112 32e
-2.06
-10.5 ( 1.72
-28.3 ( 5.8
1.21
3.58
(0.03
3.28
(0.02
(0.04
2
.45
e
toluene
o-xylene
anisole
0.56 2.38 1.49693 23
-1.86
-1.88
-1.65
-14.6 ( 1.99
-10.9 ( 0.78
-8.06 ( 1.71 -21.5 ( 5.8
-42.8 ( 6.8
-30.3 ( 2.7
(0.10
1.22
3.68
(0.05
3.26
(0.05
e
0.63 2.57 1.50545 24
(0.05
e
1.15 4.33 1.51700 16
C f
tetrahydrofurane
methanol
ethanol
0.47 7.58 1.40716
0.54 32.7 1.32840
1.07 24.6 1.36143
2.04 19.9 1.3772
2.78
2.91
3.03
3.42
1.23
1.23
1.23
1.24
3.64
3.64
3.66
3.69
3.20
3.21
3.23
3.23
2
-propanol
a
Values at 298 K.28 Reference 28.
b
c
denote hfcc’s of ortho, meta, and para protons, respectively. d
a
o
, a
m
, and a
p
K
f
and thermodynamic
e
f
parameters are from ref 9. Calculated from thermodynamic parameters at 298 K. hfcc’s are from ref 7.
shifts of ca. 10 nm due to the BPK-TEA complex formation.
Similar small red-shifts were also observed in ethanol and in
2
-propanol. Therefore, we conclude that the small red-shifts in
these solvents are not due to complex formation but due to
solvent effect on the ππ* transition of BPK caused by an
addition of TEA into these solvents. As no complex formation
was recognized in these solvents even at very high TEA
concentrations (>300 mM), we conclude that the Kf values in
-
1
these solvents are much smaller than 16∼32 M in benzene,
toluene, o-xylene, or anisole.
In our previous report, we showed that the solvent dependence
of Kf and the thermodynamic parameters ∆G, ∆H, and ∆S in
the complex formation reaction (BPK + TEA a complex) is
well explained by considering the solvation effect due to the
reaction field model. The thermodynamic parameter is expressed
by the following equations:
0
re
s
re
G ) G + ∆G
re
0
com
s
Gcom ) G + ∆G
com
where Gre and Gcom represent Gibbs free energies of reactants
0
re
0
com
and complex, respectively. G and G
mean Gibbs free
energies of nonsolvated reactants and complex, respectively.
s
re
s
∆
G
and ∆G
represent Gibbs free energy changes for the
com
solvation of reactants and complex, respectively. Then, Gibbs
free energy change of the complex formation reaction, ∆G is
9
given by
0
0
S
S
Figure 3. Transient absorption spectra of BPK in (a) tetrahydrofurane
and (b) methanol measured at 0.8-1.4 µs after 355 nm laser excitation
at 298 K. The concentrations of TEA were described in the figure.
∆G ) (∆G - ∆G ) + (∆G - ∆G )
com re com re
0
0
The first term (∆G - ∆G ) is constant for all solvents. The
com
re
S
S
the BP anion. If an appreciable amount of BPK forms a complex
with TEA in methanol, we expect that the red-shifted band of
the complex appears around 560 nm with an isosbestic point
near 550 nm as observed in toluene (Figure 1). However, no
such red-shifted band of BPK-TEA complex is recognized in
the spectrum even at the very high TEA concentration (400
mM), which indicates that the Kf value in methanol is also very
small. It is noteworthy that red-shifts in the BPK absorption
band at high TEA concentration are only about 3.5 and 2 nm
in tetrahydrofurane and in methanol, respectively, as seen in
Figure 3. These red-shifts are very small compared to the red-
second term (∆G
- ∆G ) arises upon solvation and the
com
re
solvent dependence of this energy could be qualitatively
explained by considering solvation energy based on the reaction
field model. According to the reaction field model, solvation
1
7
energy, ∆E, is expressed by the following equation:
2
M
∆
E ) g(n,ꢀr)
3
R
2
2
ꢀr - 1
n - 1
2n + 1
n - 1
where g(n,ꢀ ) )
+ 2
-
(2)
r
(
)
2
2
2
2n + 1 n + 2

Solvent Effects on Benzophenone Ketyl Radical
J. Phys. Chem. A, Vol. 105, No. 42, 2001 9631
Figure 4. A correlation between the g(n,ꢀ
r
) value of solvent and the
∆
G values in the BPK-TEA complex formation reaction at 298 K.
Closed marks correspond to the ∆G values in n-hexane, n-heptane,
cyclohexane, o-chlorotoluene, p-chlorotoluene, or 1,2-dichloroethane
from ref 9. Open marks correspond to the ∆G values in benzene,
toluene, o-xylene, or anisole.
Here, M and R denote dipole moment and cavity radius of solute,
ꢀr and n mean dielectric constant and refractive index of solvent,
respectively. It has already been demonstrated in our previous
study that the plots of the ∆G values (see Table 1) against g(n,
ꢀ
r) show good linear correlation. This feature is demonstrated
Figure 5. Time-resolved ESR spectra of BPK in (a) n-heptane, (b)
in Figure 4 with the plots shown by rectangular closed marks
fitted by a broken line. In the present study, we added the plots
of ∆G values in benzene, toluene, o-xylene, or anisole with
rectangular open marks in Figure 4. It is noteworthy that the
plots of the ∆G values in benzene, toluene, o-xylene, or anisole
are not on the broken straight lines but seem to show another
linear dependence on g(n, ꢀr) as indicated by a solid line. The
slopes and intercepts of the broken and solid straight lines are
fairly different. These plots strongly suggest that there must be
something different in complex formation reaction between the
solvents in our previous work and in the present work.
ESR Spectroscopy of BPK and the Complex in Various
Solvents. Since thermodynamic parameters for the complex
formation reaction in benzene, toluene, o-xylene, or anisole seem
different from those in other general solvents, a certain property
in the former solvents must be different from that in the latter
solvents. BPK has two phenyl rings and a hydroxy group which
could rotate depending on the solvent. Therefore, we assume
that the conformation of BPK depends on the solvent and plays
an important role in the complex formation reaction. To obtain
information of the BPK conformation, we performed TR-ESR
measurements of BPK in various solvents. Figure 5 shows TR-
ESR spectra of BPK in n-heptane, o-chlorotoluene, or 1,2-
dichloroethane together with the simulated spectrum. In a series
of TR-ESR measurements, BPK was generated by the laser
photolysis of R-phenylbenzoin instead of the photoreduction of
benzophenone because (1) no hydrogen abstraction of ben-
zophenone occurs in some of the solvents used in the present
study, (2) the photolysis of R-phenylbenzoin efficiently produces
BPK (ΦBPK ) 0.24 in benzene),18 and (3) CIDEP (chemically
induced dynamic electron polarization)19 of BPK produced by
the photolysis of R-phenylbenzoin shows net emission without
remarkable hyperfine dependent CIDEP. Though the S/N ratio
and the line width depend on the case, no significant difference
of hyperfine structure was found among these spectra. Figure
1
,2-dichloroethane, and (c) o-chlorotoluene at 298 K. BPK was
produced by the 308 nm laser photolysis of R-phenylbenzoin (12 mM)
and the gate was opened from 1.5 to 2.0 µs after the laser pulse. In the
simulation, the parameters of aOH ) 2.08 G, a
G, a ) 3.60 G are employed and the line shape is Lorentzian with a
line width of 0.20 G.
o
) 3.27 G, a
m
) 1.23
p
different from that in Figure 5. The aOH value determined from
Figure 5 is about 2.08 G while that from Figure 6 is about 2.45
G. As for the remaining hfcc’s of BPK, no remarkable difference
was found, which is in good agreement with the previous ESR
study by Wilson. Similar experiments were carried out for
BPK in various solvents and the resultant hfcc values are listed
in Table 1. The hfcc’s of BPK in tetrahydrofurane and in
alcohols reported by Wilson are also listed in Table 1.
7
a
Figure 7 shows TR-ESR spectra of BPK in the presence of
TEA (100 mM) in toluene and in n-hexane together with
simulated spectra. Reflecting the complex formation of BPK
in the presence of TEA, TR-ESR spectra with TEA and without
TEA (see Figures 5 and 6) show significant difference. In Figure
7, parts a and b, the concentration ratio of the BPK-TEA
complex to uncoupled BPK is about 3 and 5, respectively, which
are calculated from Kf values, and the carrier of the spectra
should be dominated by the BPK-TEA complex. The simulated
spectrum of BPK-TEA complex well reproduced the observed
ones and the hfcc of the complex was determined as shown in
Table 2. The spectrum of the complex is characterized by a
large aOH value of about 4.95 G. Similar experiments were
carried out for the complexes in various solvents and the hfcc
of the complex are almost the same. In accordance with the
previous work by Davidson and Wilson, the aOH value of BPK
in neat TEA was determined to be 5.48 G which is even larger
than the aOH value of the BPK-TEA complex determined in
the present study (see Table 2). When we increased the
concentration of TEA, the aOH value of BPK slightly increased
up to about 5.5 G in various solvents. This may be due to a
sort of solvent effect of TEA because the TEA concentration
more than 100 mM should be considered more likely as a
mixture of two solvents than a solvent with dissolved TEA.
6
shows TR-ESR and simulated spectra of BPK in benzene,
toluene, or o-xylene obtained by the same procedure as those
in Figure 5. The hyperfine structure of BPK in Figure 6 is very

9632 J. Phys. Chem. A, Vol. 105, No. 42, 2001
Kawai et al.
Figure 6. Time-resolved ESR spectra of BPK in (a) toluene, (b)
benzene, and (c) o-xylene at 298 K. BPK was produced by the 308
nm laser photolysis of R-phenylbenzoin (12 mM) and the gate was
opened from 1.5 to 2.0 µs after the laser pulse. In the simulation, the
Figure 7. Time-resolved ESR spectra of BPK in the presence of TEA
(100 mM) in (a) toluene and (b) n-hexane at 298 K. BPK was produced
by the 308 nm laser photolysis of R-phenylbenzoin (12 mM) and the
gate was opened from 1.5 to 2.0 µs after the laser pulse. The simulated
spectra; the top is for the BPK-TEA complex in toluene or in n-hexane,
and the bottom is for the spectrum of a 5:1 mixture of the complex
and uncoupled BPK.
parameters of aOH ) 2.45 G, a
are employed and the line shape is Lorentzian with a line width of
.20 G.
o
) 3.26 G, a
m
) 1.22 G, a
p
) 3.68 G
0
The broadening of the spectra is seen in Figure 7, which may
be partly due to the blend of different spectra of uncoupled BPK
and BPK-TEA complex coexisting. This feature is demonstrated
in Figure 7 by the simulation of the spectrum for 5:1 mixture
of uncoupled BPK and the complex.
TABLE 2: Hyperfine Coupling Constants of BPK-TEA
Complex in Toluene and BPK in Neat TEA
hfcc G
aOH
am
ap
ao
It is noteworthy that the aOH value of the complex is about
.95 G which is remarkably larger than that of BPK of 2.08 or
.45 G while the hfcc’s of the complex except the aOH value
BPK-TEA complex in toluene
or n-hexane (298 K)
4.95
(0.10
5.48
1.22
(0.03
1.22
3.67
(0.05
3.66
3.22
(0.03
3.21
4
2
BPK in neat TEA
(295 K)
are almost the same with the corresponding hfcc’s of BPK.
Figure 8 shows plots of hfcc’s of uncoupled BPK and of the
complex against ∆G for the complex formation. This plots
clearly suggests that there is a correlation between ∆G for the
complex formation reaction and the aOH value of BPK.
Therefore, we focus on the aOH value in the following.
According to the aOH value of BPK, we divide the solvents into
three groups as indicated in Table 1. Groups A, B, and C include
the solvents in which the aOH values of BPK are about 2.08,
discussed on the basis of the hfcc’s and line width of BPK and
7
,18
their temperature dependence.
According to the calculation
by Wilson, the aOH value due to the coupling between oxygen
of CO and the 1s orbital of hydroxy-proton (H1s) is only -0.64
7
a
G assuming a planer conformation of BPK. This is much
smaller than observed values in the solvents of groups A, B,
and C. In the ESR study by Wilson, it was concluded that the
aOH value of BPK is dominantly determined by the overlap
between the H1s and the 2p orbital of C13 in the COH group
together with the overlaps between H1s and 2p orbital of C1
and C7 in C-CO. The numbers of carbon atoms is referenced
to Scheme 1.
2
.45, and >2.7 G, respectively. The solvents belonging to each
group are marked in Table 1; Group A includes n-hexane,
n-heptane, cyclohexane, 1,2-dichloroethane, o- and p-chloro-
toluene; group B includes benzene, toluene, o-xylene, and
anisole; and group C includes tetrahydrofurane and aliphatic
alcohols such as ethanol and 2-propanol. It is interesting to
mention that the linear dependence of the ∆G value on g(n, ꢀr)
are found within group A and within group B as shown in Figure
These overlaps are characterized by two different dihedral
angles, θ and φ, which are described in Figure 9. The former is
the angle between OH bond axis and the plane through the C1,7-
C13-O group and the latter is the angle between phenyl ring
and the plane through the C1,7-C13-O group. Larger overlaps
of H1s and 2p orbitals of C13, C1, and C7 results in a larger aOH
value and thus, these angles are the important factors to control
the aOH value. In the following, we discuss these angles with
respect to possible BPK conformation.
4
. No correlation is found between the solvent properties and
the aOH value of the complex which is constant at about 4.95
G.
Factors Determining aOH Values of BPK and the Complex.
There are a number of ESR studies on BPK in various solvents
where the conformation and internal motion of BPK were

Solvent Effects on Benzophenone Ketyl Radical
J. Phys. Chem. A, Vol. 105, No. 42, 2001 9633
Figure 8. Plots of hfcc’s of BPK and the complex against the ∆G for
the complex formation reaction.
Figure 10. Schematic drawing of the BPK structure at the minimum
energy. The bond lengths and angles are determined by DFT calculation
with B3LYP/6-31G* level of theory. See text for details.
TABLE 3: The Calculated Results for BPK with B3LYP/
6-31G* Level of Theory
atom and
position
atomic
spin density (F)
atomic
charge
a
carbon
1
2
3
4
5
6
7
8
9
0
1
2
3
-0.123
0.155
-0.0817
0.165
-0.0782
0.146
-0.143
0.174
-0.0923
0.187
-0.0918
0.169
0.0803
-0.153
-0.0855
-0.0878
-0.0926
-0.121
0.0904
-0.106
-0.0894
-0.0866
-0.0945
-0.130
0.190
1
1
1
1
0.563
hydrogen
2
3
4
5
6
8
9
-0.00737
0.00329
-0.00773
0.00306
-0.00658
-0.00769
0.00350
-0.00875
0.00347
0.0887
0.0878
0.0864
0.0887
0.100
0.0989
0.0838
0.0823
0.0836
0.0933
0.316
Figure 9. Schematic representation of BPK with two characteristic
angles, θ and φ. “Plane” ) the plane through the C1,7-C13-O group.
10
11
12
SCHEME 1
-0.00738
0.00261
0.0785
hydroxy hydrogen
oxygen
-0.525
a
Atomic charges were obtained by Mulliken population analysis and
sum of Mulliken charges equals zero.
Figure 10 and the charge and spin densities of this conformation
are summarized in Table 3. According to the DFT calculation,
energy minimum conformation requires that a hydroxy-proton
locates over the C1,7-C13-O plane with θ ) 23° (θ ) 0 means
OH axis is on the plane through C1,7-C13-O group) and that
two phenyl rings twist by 48° with φ ) 22 and 26° (φ ) 0
means that phenyl ring is on the plane through the C1,7-C13-O
group) and hydroxy-proton is facing to the phenyl with the larger
φ. This result is qualitatively understood as follows. If hydroxy-
The most stable conformation of isolated BPK in the gas
phase was calculated by B3LYP/6-31G* level of theory
provided by the Gaussian 94 program.21 The schematic drawing
of the conformation of BPK by this calculation is shown in

9634 J. Phys. Chem. A, Vol. 105, No. 42, 2001
Kawai et al.
a smaller φ value, θ value must become larger. Hence, when
one of these angles increases the aOH value, the other must
decrease the aOH value. The difference of the aOH value between
the solvents of groups A and B is determined as a result of
these two contributions. Either of the solvated BPK conforma-
tions (I) or (II) in Figure 11 corresponds to either of BPK in
solvent group A or B. Since the difference of the aOH value is
very small (0.37 G), we could not tell which of φ and θ is
dominate in the deviation of the aOH value. Unfortunately, we
could not have any experimental evidence to determine which
conformations of (I) or (II) corresponds to BPK conformations
in the solvent of group A or B.
The BPK conformation in the solvent group C (alcohol and
tetrahydrofurane) may be different from those in solvent groups
A and B. This is because BPK can form a H-bond with alcohol
and tetrahydrofurane. In the solvent group C, the aOH value is
not constant and rather shows dependence on the solvent
viscosity as already suggested by Wilson. For larger aOH values
observed in solvent group C (2.7∼3.1G) compared to those in
solvent groups A and B (2.08 and 2.45G), the viscosity of the
solvent may be an important key controlling the rotation of
phenyl rings, but we cannot propose a clear mechanism at this
moment.
Figure 11. Schematic drawing of the possible conformations of BPK
in various solvents and in the BPK-TEA complex.
7a
proton of BPK was on the plane through C1,7-C13-O group
and the BPK was almost planer, the steric hinderence between
hydroxy- and ortho protons must be large. This requires θ and
φ with nonzero values. The DFT calculation suggests that the
BPK is not planer and supports that overlaps of H1s and 2p
orbital of C13, C1, and C7 is not zero to give nonzero aOH value.
It deserves short comment that there is the internal rotation
of two phenyl rings of BPK. This was predicted by the fact
that hfcc’s of proton at ortho and meta positions are almost
equivalent at room temperature. Though detail analysis of BPK
internal rotation has not been reported as far as we know, some
studies have mentioned the internal rotation of benzophenone
anion. The rate of geometry change due to internal rotation of
two phenyls was calculated from the line width changes of a
series of ESR spectra at different temperatures and was estimated
Structure of the BPK-TEA Complex. The small difference
in the aOH value of BPK (2∼3G) among solvent groups A, B,
and C has been discussed above. Now, we consider the large
change in aOH during complex formation (4.95 G). For the aOH
value of the complex, we again consider two factors, θ and φ.
The φ factor is, however, unlikely to be important in this case.
Usually, the hfcc of a γ proton is less than 1 G for carbon-
centered radicals. In BPK, spin density on C1,7 is ca. -0.12 to
-0.14 according to the DFT calculation (see Table 3). For such
a small spin density, hfcc may be much smaller than 1 G. This
value is fairly smaller than the aOH value of 4.95 G of the
complex. On the other hand, hfcc due to a â proton usually
could become such a large value. The equation for hfcc due to
6
-1
22
to proceed on the order of 10 s around 200 K. This report
indicates that in BPK the internal rotation occurs along the
C-CO bond. In the recent study of jet-cooled BPK by means
2
2
a â proton is approximately given by
2
3
of laser-induced fluorescence spectroscopy, it has been
-
1
concluded that BPK has a low frequency (74 cm ) torsional
motion of these phenyls, which suggests that the potential of
two phenyl rings of BPK along the internal rotation angle is
shallow. According to these results, one can find that two
phenyls of BPK rather freely rotate and it is very likely that the
angle φ corresponding to the potential minimum conformation
of BPK is easily changed depending on the solvent, which
results in the slight difference in aOH value.
Hâ
2
π
2
aOH
)
F13 cos
- θ
(3)
(
)
2
where F13 is the atomic spin density on C13 which is estimated
to be ca. +0.56 according to the DFT calculation, and Hâ is
the coefficient for hfcc of a â proton whose typical value is 50
2
2
G. The corresponding hfcc due to a â proton can be 14 G at
the maximum under the assumption that spin density is constant
for any θ value. This value is much larger than the aOH value
of the complex (4.95 G). Therefore, the aOH value of the
complex is dominantly controlled by the factor θ. Since
uncoupled BPK is rather a planer structure with relatively small
θ value as shown in Figure 10, aOH value can easily be twice
by taking a larger θ value. For example, aOH value is ca. 2.14
and 5.07 G for θ ) 23 and 37°, respectively. With the aOH
values of uncoupled BPK (2.08 and 2.45 G) and BPK-TEA
complex (4.95G), θ value was estimated to be 22.7, 24.7, and
36.5° according to eq 3. In the estimation of the θ value, we
omitted the contribution to the aOH value from the coupling
between hydroxy proton and C1,7 or O atoms. Contributions from
these couplings are about -0.6 G for an O atom and on the
order of +0.1 G for C1,7 atoms (γ proton case) as mentioned
above. These contributions may cancel out partially or totally
and thus, we consider that the aOH value can be estimated only
by eq 3.
Dependence of BPK Conformation on the Solvent. The
aOH values measured in the solvents of groups A and B are
different by about 0.37 G, which implies that the BPK
conformation in solvent group A is slightly different from that
in the solvent group B. No dependence of aOH value upon the
static physical properties of solvent such as dielectric constant
and viscosity are recognized (see Table 1). The two distin-
guished conformations of BPK in solvent groups A and B may
be determined by the local solvent effect. When BPK dissolves
into the solvent, BPK with the most stable conformation in the
gas phase deforms to become stable solvated BPK in the solvent
cage. The deformation due to solvation may be different between
solvent groups A and B, which is illustrated in Figure 11. Two
structural factors are responsible for slightly different aOH values;
One is the twisted angle, φ, of two phenyl rings and the other
is the θ value. Theoretically, smaller aOH is responsible for
smaller θ and φ values. In the case of BPK, there is a steric
hinderence between hydroxy- and ortho protons and thus, if
BPK has a larger φ value, θ value must become smaller and if
To make sure that the angle θ becomes larger when BPK
forms a complex with TEA, we calculated the optimized

Solvent Effects on Benzophenone Ketyl Radical
J. Phys. Chem. A, Vol. 105, No. 42, 2001 9635
TABLE 4: The aOH and θ Values of BPK and BPK-Amine
Complexes in Benzene
θ estimated by
a
OH value
θ determined by
DFT calculation
radicals
a
OH
assuming eq 3
BPK in solvent A)
BPK (in solvent B)
BPK/SBA
2.08
2.45
3.63
4.95
22.7°
24.7°
30.6°
36.5°
23°
23°
30°
36°
BPK/TEA
complexes given by DFT calculation. A good agreement
between the θ values determined by the aOH values and DFT
calculation were seen for BPK and BPK complexes. This result
strongly suggests that the θ value increases when BPK forms
complex with amine while other parts of BPK conformations
do not show significant change.
Solvation Effect on the Complex Formation. It is note-
worthy that a good correlation between Kf (or ∆G) and aOH
values is seen as indicated in Table 1 and in Figure 8. When
the aOH value of BPK is 2.08 G (group A), the Kf value is in
-
1
the range of 60∼103 M . If aOH of BPK is 2.45 G (group B),
-
1
the Kf value is in the range of 16∼32 M . For BPK with aOH
values of more than 2.7 G (group C), the Kf value is almost
zero and complex formation was hardly recognized in the
experiment. A smaller Kf value in group C solvents is easily
understood because BPK is already stabilized by hydrogen bond
formation between BPK and the solvent.
The BPK-TEA complex, TEA, and uncoupled BPK are
involved in the complex formation reaction. We examine which
of them mainly contributes to the solvent effects on the ∆G
value. Solvation energy by the reaction field is given by the
Figure 12. Schematic drawing of the BPK-TEA structure calculated
by DFT method with B3LYP/6-31G* level of theory. Hydrogen atoms
in TEA are excluded in order to simplify the figure.
2
3
term, M /R g(n, ꢀr). Therefore, dipole moment (M) and mean
radius (R) of the molecule are essentially important in the
solvation energy. We estimate the M and R values of BPK, TEA,
and the complex. The radii were calculated according to the
structure of the BPK-TEA complex as shown in Figure 12. The
bulky alkyl group of amine interacting with twisted phenyl rings
may control the equilibrium conformation of the complex,
especially the position of hydroxy-proton. In TEA, there is a
bulky ethyl group whose interaction with phenyl rings of BPK
is not negligible and TEA is thus, encouraged to locate in an
upper position over C1,7-C13-O plane, which is accomplished
by taking a larger θ value in the complex than in uncoupled
BPK. Reflecting this feature for the complex, DFT calculation
gave a θ value of 36° which is 1.6 times larger than that of
uncoupled BPK (θ ) 23°). In the complex, the conformation
of C1,7-C13-O group becomes slightly pyramidal. Hence, we
assume the planner C1,7-C13-O conformation by using the
C1,7-O plane for the calculation of the θ value.
2
5
equation of Spernol and Wirtz,
1
/3
3
× w
ø
R )
(4
)
πFN
A
where w, F, NA, and ø are molecular weight, density, Avogadro’s
constant, and free volume, respectively. We adopted a free
volume of 0.74 as used in their calculation. Since the density
of BPK and the complex are unknown, we assume that the
densities of them are unity. The M values were obtained from
DFT calculation with the minimum energy conformations of
BPK, TEA, and the complex. Since the θ value of the BPK-
TEA complex obtained by DFT calculation agrees well with
the experimentally determined θ value as shown in Table 4,
we believe that the conformation of the BPK-TEA complex
given by DFT calculation is close to the real conformation of
the complex. According to these methods, the M and R values
are obtained as 1.48 D and 3.77 Å for BPK, 0.55 D and 3.44 Å
for TEA, and 3.53 D and 4.37 Å for the complex, respectively.
The dipole moment of BPK is mainly generated by positively
charged C13 and hydroxy-proton, and negatively charged oxygen
(see Table 3). Dipole moment of TEA is due to negatively
charged nitrogen and positively charged end-carbons of each
ethyl part. The complex is formed by the H-bond between the
hydroxy-proton of BPK and the nitrogen atom in TEA and two
dipoles tends to locate rather parallel giving large total dipole
moment of the complex. Calculated M values of BPK and TEA
In our previous study,10 we reported that the complex of BPK
with sec-butylamine (SBA) shows an aOH value of about 3.9
G, which is smaller than 4.95 G of BPK-TEA complex and
that any difference were found in the remaining hfcc’s. This
may suggest that the θ value of BPK-SBA complex is smaller
than that of the BPK-TEA complex. With the aOH value of the
BPK-SBA complex under the assumption that F13 is +0.56 and
Hâ is 50 G, we estimated the θ value as 30.6° which is
apparently smaller than that of the BPK-TEA complex (36.5°).
The alkyl group of SBA is less bulky than that of TEA and the
steric hindrance between alkyl and phenyl groups should be
smaller in the BPK-SBA complex than in the BPK-TEA
complex, which results in a smaller θ value of the BPK-SBA
complex. DFT calculation of the BPK-SBA complex also shows
that the θ value of the complex is 30° which is larger than that
of BPK (23°) but smaller than that of the BPK-TEA complex
2
6
(36°). These experimentally determined θ values from the aOH
is close to the observed M values for BPK (1.5 D) in benzene
2
7
values were summarized in Table 4, together with θ values in
the energy minimum conformations of BPK or BPK-amine
and for TEA (0.66 D). Slight differences may be due to the
deformation of these molecules in the solvent. According to

9636 J. Phys. Chem. A, Vol. 105, No. 42, 2001
Kawai et al.
the M and R values mentioned above, we obtained the relative
solvation energy of three species as 149:41:7(complex:BPK:
TEA). This means that contribution of TEA is almost negligible
compared to BPK and the complex. Therefore, we consider that
the solvation effect is mainly controlled by the complex and
BPK. Since conformation of the complex is almost identical
between the solvent groups A and B, it is unlikely that the
solvation energy of the complex differs between the two. Thus,
we conclude that the different solvent dependence of ∆G value
between solvent group A and B as shown in Figure 4 is mainly
due to the different conformation of uncoupled BPK.
(b) Baumann, H.; Schumacher, K. P.; Timpe, H.-J.; Reh a´ k, V.; Chem. Phys.
Lett. 1982, 89, 315. (c) Adam, W.; Oestrich, R. S.; Chem. Ber. 1992, 125,
2
463.
6) (a) Terazima, M.; Okamoto, K.; Hirota, N. J. Phys. Chem. 1993,
97, 13387. (b) Ukai, A.; Hirota, N.; Terazima, M. Chem. Phys. Lett. 2000,
19, 427.
7) (a) Wilson, R. J. Chem. Soc. (B) 1968, 84. (b) Davidson, R. S.;
Wison, R. J. Chem. Soc. (B) 1970, 71.
8) Kajii, Y.; Itabashi, H.; Shibuya, K.; Obi, K. J. Phys. Chem. 1992,
96, 7244.
(9) Abe, T.; Kawai, A.; Kajii, Y.; Shibuya, K.; Obi, K. J. Phys. Chem.
A 1999, 103, 1457.
10) Kawai, A.; Aoki, A.; Kobori, Y.; Obi, K. J. Phys. Chem. 1996,
00, 10021.
(
3
(
(
(
1
(
(
11) Kawai, A.; Kobori, Y.; Obi, K. Chem. Phys. Lett. 1993, 215, 203.
12) The molecular plane of BPK means the plane defined by three
Conclusion
carbon and one oxygen atoms within the C-C(OH)-C moiety of BPK.
13) Kajii, Y.; Fujita, M.; Hiratuska, H.; Obi, K.; Mori, Y.; Tanaka, I.
J. Phys. Chem. 1987, 91, 2791.
We studied the complex formation reaction of BPK with TEA
in various solvents by means of transient absorption and TR-
ESR method. According to TR-ESR spectra of uncoupled BPK
and the complex together with DFT calculation, we concluded
that BPK has at least three different conformations depending
on the solvent. The conformation of COH group and twisted
angle of phenyls in BPK depends on the solvent. On the other
hand, the conformation of the BPK-TEA complex does not
depend on the solvent. Conformation of BPK is strongly
controlled by the solvation while that of the BPK-TEA complex
may be controlled only by the interaction between BPK and
TEA.
The dihedral angle between OH bond axis and the plane
through the C1,7-C13-O group changes from 22.7° and 24.7°
to 36.5° by the complex formation. A good correlation between
BPK conformation and the ∆G value in the complex formation
reaction was found, which suggests that the ∆G value is roughly
controlled by the conformation of uncoupled BPK.
(
(14) Miyasaka, H.; Mataga, N. Bull. Chem. Soc. Jpn. 1990, 63, 131.
(15) (a) Miyasaka, H.; Nagata, T.; Kiri, M.; Mataga, N. J. Phys. Chem.
1
992, 96, 8060. (b) Miyasak, H.; Morita, K.; Kamada, K.; Mataga, N. Bull.
Chem. Soc. Jpn. 1990, 63, 3385.
16) Devadoss, C.; Fessenden, R. W. J. Phys. Chem. 1991, 95, 7253.
(
(17) (a) Onsager, L. J. Am. Chem. Soc. 1936, 58, 1486. (b) Wilson, J.
N. Chem. ReV. 1939, 25, 377. (c) McRae, E. G. J. Phys. Chem. 1957, 61,
562.
(
18) Shrestha, N. K.; Yagi, E. J.; Takatori, Y.; Kawai, A.; Kajii, Y.;
Shibuya, K.; Obi, K. J. Photochem. Photobiol., A Chem. 1998, 116, 179.
19) For example; Muus, L. T.; Atkins, P. W.; McLauchlan, K. A.;
(
Pedersen, J. B. Chemically Induced Magnetic Polarization; Reidel: Dor-
drecht, 1977.
(
20) (a) Miyagawa, K.; Murai, H.; I’Haya, Y. J. Chem. Phys. Lett. 1985,
1
18, 140. (b) Miyagawa, K.; I’Haya, Y. J.; Murai, H. Nippon Kagaku Kaishi
1989, 8, 1358.
(21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G.
A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski,
V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M. H.; Gonzalez, C.; Pople, J. A. Gaussian 94; Gaussian, Inc.:
Pittsburgh, PA, 1995.
Acknowledgment. The authors express their thanks to Dr.
S. Yamago (Kyoto University) and Prof. E. Nakamura (Uni-
versity of Tokyo) for their kind aid in synthesis. The present
work is partly defrayed by the Grant-in-Aid for Scientific
Research (No.10740269) from the Ministry of Education,
Science, Sports and Culture of Japan.
(22) (a) Takeshita, T.; Hirota, N. J. Chem. Phys. 1969, 51, 2146. (b)
Bernardi, F.; Guerra, M.; Pedulli, G. F. J. Phys. Chem. 1974, 78, 2144.
(
23) Wada, S.; Matsushita, Y.; Obi, K. J. Phys. Chem. A 1997, 101,
423.
24) Atherton, N. M. Principles of Electron Spin Resonance; Ellis
Horwood Limited: Chichester, U.K., 1993.
2
(
References and Notes
(
25) Spernol, von A.; Wirtz, K. Z. Natureforsch. 1953, 8a, 522.
(
1) (a) Obi, K.; Yamaguchi, H. Chem. Phys. Lett. 1978, 54, 448. (b)
(26) Fukui, T.; Shimamori, H. Unpublished data measured by a time-
Takeda, K.; Kajii, Y.; Shibuya, K.; Obi, K. J. Photochem. Photobiol., A
Chem. 1998, 115, 109.
resolved microwave dielectric absorption method. For the experimental
method, see Shimamori, H.; Uegaito, H.; Houdo, K. J. Phys. Chem. 1991,
5, 7664.
(
2) Nagarajan, V.; Fessenden, R. W. Chem. Phys. Lett. 1984, 112, 207.
9
(3) Hiratsuka, H.; Yamazaki, Y.; Maekawa, Y.; Hikida, T.; Mori, Y.
J. Phys. Chem. 1986, 90, 774.
4) (a) Johnston, L. J.; Lougnot, D. J.; Wintgens, V.; Scaiano, J. C. J.
(27) CRC Handbook of Chemistry and Physics, 1988-1989, 69th ed.;
(
Weast, R. C., Ed.; CRC Press: Boca Raton, FL, 1988.
(28) (a) Murove, S. L.; Charmichael, I.; Hug, G. L. Handbook of
Photochemistry; Marcel Dekker Inc.: New York, 1993. (b) CRC Handbook
of Chemistry and Physics, 1995-1996, 76th ed.; Lide, D. R., Ed.; CRC
Press: Boca Raton, FL, 1995.
Am. Chem. Soc. 1988, 110, 518. (b) Redmond, R. W.; Scaiano, J. C.;
Johnson, L. J. J. Am. Chem. Soc. 1990, 112, 398.
(5) (a) Baumann, H.; Merckel, C.; Timpe, H.-J.; Graness, A.; Klein-
schmidt, J.; Gould, I. R.; Turro, N. J.; Chem. Phys. Lett. 1984, 103, 497.