Effect of medium on the rate constant of decarbonylation of phenylacetyl radical

Article abstract of DOI:10.1023/A:1009518015555

The rate constant k_CO of decarbonylation of phenylacetyl radicals generated by photolysis of dibenzyl ketone was measured by laser flash photolysis technique in six solvents in a wide temperature range. The pre-exponential factors A and activation energies E_a of decarbonylation were found for all solvents. The k_CO value decreases with an increase in the dielectric constant ε of the solvent, whereas an increase in the ability of the solvent for hydrogen bonding increases k_CO. The results of quantum-chemical calculations confirm the mutual compensation of the contributions of specific and nonspecific solvations to the activation energy of decarbonylation in alcohols.

Full text of DOI:10.1023/A:1009518015555

Russian Chemical Bulletin, International Edition, Vol. 50, No. 2, pp. 237—240, February, 2001
237
Effect of medium on the rate constant of decarbonylation of
phenylacetyl radical
Yu. P. Tsentalovich,a« O. A. Kurnysheva,a,b and N. P. Gritsanb,c
aInternational Tomography Center, Siberian Branch of the Russian Academy of Sciences,
3
a ul. Institutskaya, 630090 Novosibirsk, Russian Federation.
Fax: +7 (383 2) 33 1399. E-mail: yura@tomo.nsc.ru
bNovosibirsk State University,
2
ul. Pirogova, 630090 Novosibirsk, Russian Federation
cInstitute of Chemical Kinetics and Combustion, Siberian Branch of the Russian Academy of Sciences,
ul. Institutskaya, 630090 Novosibirsk, Russian Federation.
3
Fax: +7 (383 2) 34 2350. E-mail: gritsan@ns.kinetics.nsc.ru
The rate constant kCO of decarbonylation of phenylacetyl radicals generated by photolysis of
dibenzyl ketone was measured by laser flash photolysis technique in six solvents in a wide
temperature range. The pre-exponential factors A and activation energies Ea of decarbonylation
were found for all solvents. The kCO value decreases with an increase in the dielectric constant ε of
the solvent, whereas an increase in the ability of the solvent for hydrogen bonding increases kCO
.
The results of quantum-chemical calculations confirm the mutual compensation of the contribu-
tions of specific and nonspecific solvations to the activation energy of decarbonylation in alcohols.
Key words: laser flash photolysis, radical fragmentation, decarbonylation, phenylacetyl
radical, quantum-chemical calculations.
It has been considered for a long time that the rate
constants of fragmentation of neutral radicals in a solu-
tion are virtually independent of the solvent nature. For
irradiated with an XeCl-excimer pulse laser (Lambda Physik
EMG 101, 308 nm, pulse energy up to 100 mJ, pulse duration
15—20 ns). The monitoring system included a DKsSh-150
xenon short-arc lamp (pulse duration 2 ms, size of the detection
beam 2½3 mm), two synchronically monitored monochroma-
tors, a Hamamatsu R955 photomultiplier, a digital two-channel
oscilloscope with an 11-bit analog-to-digital converter (LeCroy
9310A, 400 MHz, time resolution 10 ns), and a system of filters
and shutters. All solutions were purged with argon to remove
oxygen from the system prior to use and during irradiation. The
temperature of samples was varied within the 248—347 Ê
interval by a thermostabilized nitrogen flow and measured by a
thermocouple.
Quantum-chemical calculations of the geometry and heat of
formation of the phenylacetyl radical, transition state, and their
complexes with alcohols were performed by the ÐÌ3 method13
using the MNDO 92 program (the next version of the MNDO 85
program14). In addition, the geometry, electronic structure, and
solvation energy of the initial radical and transition state were
calculated by the density functional theory in the variant of the
B3LYP hybrid method15 using the GAUSSIAN 98 program.16
1
example, the decarbonylation of phenylacetyl radicals2—4
was proposed to be used as a "chemical clock" for the
quantitative determination of the absolute rates of the
competitive reactions. The authors of recent works re-
ported an increase in the reaction rates of β-scis-
sion of alkoxyl radicals5
—9
and a decrease in the rate of
carbonylation of acyl radicals10,11 in polar solvents.
The influence of the medium on the decarbonylation
rate of phenylacetyl radicals has previously10 been stud-
ied only at room temperature. It seems reasonable to
find the activation energy and pre-exponential factor of
these reaction rates in various solvents and examine the
influence of the medium on these parameters in a
quantitative way. In this work, laser flash photolysis
measurements were carried out to determine the rate
constants of decarbonylation of phenylacetyl radicals
generated by photolysis of dibenzyl ketone in a wide
range of temperatures for a series of solvents. The influ-
ence of the dielectric properties of the solvents and their
ability for hydrogen bonding on the rate constant and
activation energy of this reaction was examined.
The 6-31G* basis set was used in all cases. The polarized
continuum model (PCM, Tomasi model)17 was used to calcu-
late the solvation energy.
Results and Discussion
The mechanism of photolysis of dibenzyl ketone
1
8
Experimental
R CO (where R = PhCH ) is well studied and de-
2 2
scribed by the following reactions:
Materials. Solvents were purified by repeated distillation.
Dibenzyl ketone («Aldrich») was used as received.
Pulse photolysis. A laser pulse photolysis technique has
been described previously.8,12 A solution circulating through a
rectangular quartz cell 10½3 mm in size using a flow system was
–
8
hν
<10
s
R CO
2
R CO (S )
2
R CO (T ) (1)
2 1
1
–
9
<
10
s
R CO (T )
2
RCO + R
(2)
1
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 2, pp. 229—233, February, 2001.
066-5285/00/5002-237 $25.00 © 2001 Plenum Publishing Corporation
1

2
38
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 2, February, 2001
Tsentalovich et al.
k
CO
As can be seen in Fig. 1, at room temperature
decarbonylation occurs most rapidly in hexane (nonpo-
lar solvent) and most slowly in acetonitrile.
RCO
R + CO
Products
Products
(3)
(4)
(5)
2
k1
2
2
RCO
The decarbonylation rate constants kCO at different
temperatures were determined in the following way. For
every given temperature the kinetics of changing the
absorption of benzyl radicals was detected at four differ-
ent energies of laser pulses in an interval of 1—10 mJ.
The initial concentrations of radicals were changed from
2
k2
R
k
x
RCO + R
Products
(6)
Since the recombination reactions of the radicals are
controlled by diffusion and the size of the radicals is
approximately equal, we can use the approximation19
kt = 2k1 ≈ 2k2 ≈ k . In this case, the kinetics of
formation and decay of the benzyl and phenylacetyl
radicals is described by the system of Eqs. (7) and (8).
2•10–6
to 2•10–5 mol L–1, respectively. The obtained
experimental kinetic curves were treated according to
Eq. (9). The extinction coefficient of the benzyl radical
2
x
6
500 mol–1 L cm–1 was taken from the literature.10 The
rate constants kCO and k and initial concentrations of
t
d[R]
kCO[RCO] – 2k [R]2 – 2k [R][RCO]
radicals R are variable parameters. As a rule, scatter of
=
(7)
0
t
t
dt
the calculated rate constants inside each series did not
exceed 5—7%. The calculation is exemplified in Fig. 1.
The temperature dependences of measured rate con-
stants of decarbonylation of the phenylacetyl radical kCO
for three solvents are presented in Fig. 2. The data fall
on a straight line fairly well and that makes it possible to
d R[ CO ]
dt
=
–kCO[RCO] – 2k [RCO]2 – 2k [R][RCO] (8)
t
t
The solution of this system gives the the concentra-
tion of benzyl radicals [R ] as a function of time
•
20
•
determine the activation energy of decarbonylation (Eà)
[
R ] = R (2 – exp(–k t))/(1 + 4k R t),
(9)
0
CO
t
0
and pre-exponential factor (A) in the Arrhenius equa-
tion. The activation energies and pre-exponential factors
where R is the initial concentration of benzyl radicals.
0
Transient absorption kinetics of the benzyl radical
detected at the absorption maximum (λ = 314 nm)10 is
presented in Fig. 1. The instant increase in the absorp-
tion immediately after a laser pulse corresponds to the
formation of the benzyl radical during the decomposi-
tion of the starting ketone (reaction (2)). The subsequent
increase in the signal indicates the formation of benzyl
radicals in decarbonylation of phenylacetyl radicals (re-
action (3)). The slow decrease in the signal corresponds
to the bimolecular decay of the radicals (reactions
are presented in Table 1 along with the kCO values at
∼
20 °C, which are in good agreement with the published
3,4,10
data.
According to the classical transition state theory, the
influence of the medium on the reaction rate constant
can be considered as the difference of the solvation
energies for the initial reactants and transition state.
The results obtained can simply be explained qualita-
tively: since the nonpolar benzyl radical is formed from
the polar phenylacetyl radical upon decarbonylation, the
transition state is less polar than the initial radical.
The dielectric constant (ε) is used as a mea-
sure of solvent polarity in simplest models. The
(
4)—(6)), which is well observed in a longer time scale.
∆D
1
ln(kCO
)
2
0
0
.08
1
1
1
1
1
6
3
5
4
3
2
.04
2
1
3
0
–
1
0
1
2
3
t/µs
T –1•10 /K
3
–1
3
.0
3.2
3.4
3.6
Fig. 1. Kinetic curves of formation and decay of benzyl radicals
314 nm) in hexane (1), methanol (2), and acetonitrile (3) at
room temperature. Calculated values (solid lines) are obtained
from Eq. (9).
(
Fig. 2. Arrhenius plots for the rate constant of carbonylation of
phenylacetyl radicals in hexane (1), isopropyl alcohol (2), and
acetonitrile (3).

Decarbonylation of phenylacetyl radical
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 2, February, 2001
239
Table 1. Rate constants for decarbonylation of phenylacetyl radicals at ∼ 20 °C and parameters
of the Arrhenius equation (log (A), E ) for this reaction in various solvents
a
Solvent
ε –1
ε + 1
.
ÅT N
(Ref. 1)
kÑΕ106 (295 Ê)
log(À)
Åa
/kJ mol
–
1
2
s–1
Hexane
Acetic
Isopropyl alcohol 0.460
Ethanol
Methanol
Acetonitrile
0.188
acid0.388
0.009
0.648
0.546
0.654
0.762
0.460
(4.5±0.3)
(2±0.2)
(2.6±0.2)
(2.4±0.2)
(2.2±0.2)
(1.4±0.1)
12.2±0.3
11.7±0.3
12.5±0.3
11.5±0.2
12.1±0.3
13.1±0.4
31.5±0.5
30.5±0.6
34.2±0.7
28.9±0.4
32.2±0.6
39.6±0.7
0.470
0.477
0.480
Kirkwood formula1 is used to relate the solvation en-
Ea/kJ mol–1
ergy to ε
4
0
6
ln(k) – ln(k ) =
0
#
2
a
2
F
I
∆
G
1
NA ε − 1
µ
µ
solv
#
=
−
=
−
G
−
J
,
(10)
3
3
#
RT
4πε0 RT 2ε + 1
H
ra
r
K
3
6
where k and k0 are the rate constants in the gas phase
and in a hypothetical solvent with ε = 1, ∆G solv is the
#
3
free solvation energy, NA is Avogardo´s number, and µa,
µ , r , and r are the dipole moments and radii of the
5
32
28
#
a
#
1
initial radical and transition state, respectively. Model
10) is very simplified and considers a spherical molecule
2
4
(
with the point dipole moment in the center, ignoring the
real shape, charge distribution, and specific interactions
of the dissolved molecule with solvent molecules, for
example, the ability of the solvent to form hydrogen
bonds. As follows from data in Table 1, the Kirkwood
formula is poorly applicable to the description of the
solvent effect on the decarbonylation rate constant: de-
spite the Kirkwood parameters (ε – 1)/(2ε + 1) for the
reactions in acetonitrile and alcohols are very close, the
activation energies for the reactions in these solvents
differ noticeably.
A more universal approach is based on the use of the
empirical parameters that characterize the solvation ca-
pability of the solvent. Among these parameters, the
Dimroth ET scale1 is most popular. The empirical pa-
rameters take into account both the electrostatic and
specific interactions between the solvent and solvate. In
ETN
0
0.2
0.4
0.6
Fig. 3. The plots of the activation energy (Ea) of decarbonylation
of phenylacetyl radicals vs. ET parameter. Solvents: hexane
1), acetic acid (2), isopropanol alcohol (3), ethanol (4), metha-
N
(
nol (5), and acetonitrile (6).
bonds have the same effect on the rate constant. Assume
that the initial radical is more polar than the transition
state, whereas the formation of hydrogen bonds stabi-
lizes the transition state more strongly than the initial
radical does. In this case, the specific and nonspecific
solvations can mutually be compensated. This is pre-
cisely the effect which is observed for the decarbonylation
of phenylacetyl radicals.
To check the assumption on the mutual compensa-
tion of two contributions to the difference of free activa-
tion energies of decarbonylation, we performed quan-
tum-chemical calculations of nonspecific solvation of
the radical and TS and heat of formation of complexes
of the radical with alcohol molecules.
fact, a good linear correlation between the E (30) pa-
T
rameter and logarithm of the reaction rate constant was
observed for several reactions, in particular, for the
β-scission of alkoxyl radicals.7,8
The plot of the activation energy of decarbonylation
of phenylacetyl radicals vs. ETN parameter is shown in
Fig. 3. It is seen that an attempt to describe the influ-
ence of the medium using this parameter gives worse
results than the application of the Kirkwood formula.
Below we present the C—C and C=O bond lengths
and spin densities on the C and O atoms in the
phenylacetyl radical and in TS of decarbonylation
along with spin density distribution calculated by the
B3LYP/6-31G* method.
The E N parameter is proportional to the energy of
T
electron transition of a standard compound in various
solvents. An increase in the polarity of the solvent and
formation of hydrogen bonds increase this parameter.
Therefore, a linear correlation between the ETN and Ea
parameters of the specific reaction is observed only when
the electrostatic interactions and formation of hydrogen
1
0.10
0.56
CH2
0.25
CH
2
•
0.12
0
.60
O
1.19
1
.54
2.05 0.21
O
1.16
C
•
C
Phenylacetyl radical
Transition state

2
40
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 2, February, 2001
Tsentalovich et al.
It is seen that an essential (0.5 Å) elongation of the
This work was financially supported by the Russian
Foundation for Basic Research (Project Nos. 99-03-
32753 and 99-03-33488) and INTAS (Grant 96-1269).
R—ÑO bond cleaved in the reaction occurs in TS. The
other bond lengths change insignificantly. In addition,
the spin density is substantially redistributed from the
C atom of the carbonyl group to the C atom of the
methylene group. The difference in the electron energies
References
of the TS and initial radical amounts to 37.9 kJ mol–1
.
1
. C. Reichardt, Solvents and Solvent Effects in Organic Chem-
istry, VSH, Weinheim, 1990.
Taking into account the difference in energies of zero
vibrations, it is equal to 29.9 kJ mol–1, and the differ-
ence in enthalpies is 30.4 kJ mol–1, which agrees well
with the activation energy measured by us in the nonpo-
lar solvent hexane (31.5±0.5 kJ mol–1).
2. D. Griller and K. U. Ingold, Acc. Chem. Res., 1980, 13, 317.
3. L. Lunazzi, K. U. Ingold, and J. C. Scaiano, J. Phys.
Chem., 1983, 87, 529.
4. N. J. Turro, I. R. Gould, and B. H. Baretz, J. Phys. Chem.,
1
983, 87, 531.
The electron density redistribution changes the di-
5
. G. D. Mendenhall, L. C. Stewart, and J. C. Scaiano, J. Am.
Chem. Soc., 1982, 104, 5109.
–
30
pole moment from µ = 7.7•10
µ# = 5.3•10
C m in the radical to
a
–1
–
30
C m in the TS. The PM3 calculations
6. P. Neta, M. Dizdaroglu, and M. G. Simic, Isr. J. Chem.,
1984, 24, 25.
7. D. V. Avila, C. E. Brown, K. U. Ingold, and J. Lusztyk,
J. Am. Chem. Soc., 1993, 115, 466.
demonstrate a noticeable decrease in the dipole moment
of the TS of decarbonylation. Therefore, taking into
account only electrostatic interactions with the solvent,
an increase in the activation energy with increasing
solvent polarity can be expected. The effect of solvent on
the difference between free solvation energies of the
8
. Yu. P. Tsentalovich, L. V. Kulik, N. P. Gritsan, and A. V.
Yurkovskaya, J. Phys. Chem. A, 1998, 102, 7975.
9
. M. Weber and H. Fischer, J. Am. Chem. Soc., 1999, 121, 7381.
1
0. Yu. P. Tsentalovich and H. Fischer, J. Chem. Soc., Perkin
Trans. 2, 1994, 729.
reactions in hexane ∆G#
and acetonitrile ∆G#
was
ACN
hex
estimated by formula (10). The radius of the initial
radical estimated from the molar volume of phenyl-
11. M. Salzmann, Yu. P. Tsentalovich, and H. Fischer, J. Chem.
Soc., Perkin Trans. 2, 1994, 2119.
acetaldehyde similar in size was accepted as r = 3.20 Å,
12. I. F. Molokov, Yu. P. Tsentalovich, A. V. Yurkovskaya, and
R. Z. Sagdeev, J. Photochem. Photobiol. A: Chemistry, 1997,
a
and that of the transition complex, taking into account
the elongation of the R—CO bond, was taken as
r# = 3.40 Å. The obtained value of the difference
1
10, 159.
1
3. J. J. P. Stewart, J. Comp. Chem., 1989, 10, 209; J. J. P.
Stewart, J. Comp. Chem., 1989, 10, 221.
∆
G#
– ∆G#
= 1.6 kJ mol–1 is much lower than the
ACN
hex
14. A. A. Bliznyuk and A. A. Voityuk, Zh. Strukt. Khim., 1986,
7, 190 [J. Struct. Chem. (USSR), 1986, 27 (Engl. Transl.)].
experimentally determined difference in activation ener-
gies of the decarbonylation rate constants in hexane and
acetonitrile (see Table 1). This is due to the imperfection
of the model of point dipole in the spherical cavity.
Unlike estimations by formula (10), the ÐÑÌ model
2
15. A. D. Becke, J. Chem. Phys., 1993, 98, 1372.
16. 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, 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.6, Gaussian, Inc., Pittsburgh PA, 1998.
(
Polarized Continuum Model) takes into account a more
realistic shape of the cavity in the solvent and the real
charge distribution in the radical. The calculations by this
model give the difference in free solvation energies of the
transition state and initial phenylacetyl radical in aceto-
nitrile and hexane as ∆G#ACN – ∆G#hex = 4.2 kJ mol–1
,
which agrees well with the experimental data.
The difference in enthalpies of formation of com-
plexes of the initial radical and TS with a solvent
molecule can be accepted as a measure that character-
izes the influence of formation of hydrogen bonds with
the solvent on the reaction rate constant. The PM3
calculations for the reaction in methanol show that the
heat of formation of a complex with the radical
1
7. S. Miertus, E. Scrocco, and J. Tomasi, Chem. Phys., 1981,
55, 117; R. Cammi and J. Tomasi, J. Chem. Phys., 1994,
101, 3888; M. Cossi, V. Barone, R. Cammi, and J. Tomasi,
Chem. Phys. Lett., 1996, 255, 327.
7.2 kJ mol–1) is much lower than that of a complex
18. J. C. Dalton and N. J. Turro, Ann. Rev. Phys. Chem., 1970,
21, 499; P. S. Engel, J. Am. Chem. Soc., 1970, 92, 6074;
W. K. Robbins and R. H. Eastman, J. Am. Chem. Soc.,
(
with the TS (12.9 kJ mol–1). Therefore, unlike specific
solvation, the formation of hydrogen bonds increases the
decarbonylation rate constant. The calculated effects of
electrostatic interactions with the solvent and formation
of hydrogen bonds are close in value and unlike in sign.
Thus, the quantum-chemical calculations confirm our
hypothesis about the mutual compensation of the contri-
butions of specific and nonspecific solvation to the free
activation energy of decarbonylation in alcohols.
1
970, 92, 6076; N. C. Yang, E. D. Feit, N. N. Nui, N. J.
Turro, and J. C. Dalton, J. Am. Chem. Soc., 1970, 92, 6974.
19. H. Fischer and H. Paul, Acc. Chem. Res., 1987, 20, 200.
20. J.-K. Vollenweider, H. Fischer, J. Hennig, and R. Leuschner,
Chem. Phys., 1985, 97, 217.
Received June 6, 2000;
in revised form September 12, 2000