Effects of pressure on the thermal back electron transfer in a dibenzylbipyridinium salt. Solvent reorganizations and dynamic solvent effects

Article abstract of DOI:10.1246/cl.2004.1542

Kinetic effects of pressure on a decay of a radical pair produced by UV irradiation of N,N′-dibenzyl-4,4′-bypyridinium dichloride were studied in acetonitrile (AN) and propylene carbonate (PC). Both in AN and PC, the process was accelerated by an increase in pressure indicating a considerable contraction of the solvation sphere during the activation. Furthermore, in PC, pressure-induced retardation was observed at P > 200 MPa and it was attributed to slow solvent thermal fluctuations at high pressures. Copyright

Full text of DOI:10.1246/cl.2004.1542

1
542
Chemistry Letters Vol.33, No.12 (2004)
Effects of Pressure on the Thermal Back Electron Transfer in a Dibenzylbipyridinium Salt.
Solvent Reorganizations and Dynamic Solvent Effects
ꢀ
ꢀ
Mehdi S. Shihab, Koji Kubota, Toru Takahashi, Yasushi Ohga, and Tsutomu Asano
Department of Applied Chemistry, Faculty of Engineering, Oita University, 700 Dannoharu, Oita 870-1192
(Received August 25, 2004; CL-040999)
Kinetic effects of pressure on a decay of a radical pair pro-
0
0
duced by UV irradiation of N,N -dibenzyl-4,4 -bypyridinium di-
chloride were studied in acetonitrile (AN) and propylene carbo-
nate (PC). Both in AN and PC, the process was accelerated by an
increase in pressure indicating a considerable contraction of the
solvation sphere during the activation. Furthermore, in PC, pres-
sure-induced retardation was observed at P > 200 MPa and it
was attributed to slow solvent thermal fluctuations at high pres-
sures.
0.0070
-
6
-7
-8
0
0
0
.0060
.0050
.0040
-9
-
10
0
1
2
3 4 5
t / s
There have been many reports on pressure dependence of
electron transfer rate processes and most of the results could
be rationalized on the basis of the transition state theory
0
1
2
3
4
5
t / s
1
(
the activation volume ꢀV . For example, the reduction of
TST). Namely, the results could be rationalized in terms of
z
Figure 1. Absorbance change and the plot of first-order rate law
ꢂ
(inset) monitored at 600 nm in PC at 10 C (90 MPa).
2
þ
4ꢁ
ꢁ1
Co(NH3)5(H2PO4) by Fe(CN)6 was retarded by an increase
z
3
ꢂ
in pressure, ꢀV ¼ 17 cm mol at 25 C in H2O, and it was at-
followed the first-order rate law at all of the conditions studied
as in the example and we could safely assume that this electron
transfer took place within a radical pair and the translational dif-
fusion of the radicals did not play an important role in the present
reaction. Figures 2 and 3 illustrate pressure dependence of the
first-order rate constant k1 in AN and PC, respectively. In both
of the solvents, the reaction was accelerated by an increase in
pressure at lower pressures, P < 100 MPa. The activation vol-
umes were estimated by fitting the results to a quadratic equation
2
tributed to desolvation caused by the charge neutralization. In
other words, solvent thermal fluctuations were fast enough to
maintain the thermodynamic equilibrium between the initial
and the transition state. Since applying pressure results in an in-
crease in viscosity and it, in turn, slows down solvent thermal
fluctuations, it is conceivable that solvent reorganization process
itself becomes rate-limiting at high pressures, namely, a reaction
may change its kinetic nature from the ‘‘activation-limited’’ to
the ‘‘fluctuation-limited’’ at high viscosities. To the best of our
knowledge, however, there has been no report of a case in elec-
tron transfer where the TST was valid at 0.1 MPa and it turned
invalid with increasing pressure. In this letter we would like to
report possibly the first observation of induced invalidation of
the TST in an electron-transfer rate process.
z
and the values at zero pressure, ꢀV0 , were estimated to be ꢁ17
3
ꢁ1
ꢂ
and ꢁ57 cm mol at 25 C in PC and AN, respectively. Since a
pair of charges was generated in this reaction, a volume contrac-
tion during activation is more or less pressure-expected. Solvent
molecules have to be reorganized to stabilize the charged spe-
4–7
cies. Kondo and co-workers measured reaction volumes ꢀV
ꢂ
in several nucleophilic substitutions in methyl iodide at 30 C
h
ν
and the values were ꢁ63 to ꢁ40 in AN and ꢁ36 to ꢁ21
R
N
N
R
R
N
+
N
R
3
ꢁ1
cm mol in PC depending on the nucleophiles. The large vol-
ume contraction results from electrostrictions around the cation
and the anion and the magnitude was always larger in AN be-
cause of its larger compressibility than PC. Similarly, in our re-
action, the charged species, pyridyl cation and chloride anion,
were generated from the corresponding neutral radicals by elec-
tron transfer. Comparing these values with our activation vol-
umes, it may be concluded that solvent reorganizations were al-
most complete at the transition state in AN but they were still un-
derway in PC.
∆
−
X
X
X
X
1
2
Scheme 1.
0
N,N -disubstituted bipyridyls 1 are known to be photochro-
mic and it is well proved that the colored species is a radical cat-
3
ion 2. In solution, this species decays rapidly by thermal back
electron transfer (Scheme 1). We studied pressure dependence
0
0
of this thermal reaction rate in N,N -dibenzyl-4,4 -bipyridyl
dichloride (R = PhCH2, X = Cl) in acetonitrile (AN, 4 ꢃ
Figures 2 and 3 reveal another significant difference be-
tween the two solvents. In AN, the reaction was accelerated
throughout the whole pressure range studied, while in PC the
pressure effect was reversed somewhere between 100 and
300 MPa. The sign of the activation volume changed from neg-
ative to positive with increasing pressure. This sort of behavior is
ꢁ
4
ꢁ4
10
M) and propylene carbonate (PC, 4{5 ꢃ 10 M) by flash
8
photolysis using Xe lamp. The reaction was followed by the ab-
sorption decrease at ꢀmax in the visible region, 570 nm (AN) and
600 nm (PC). Figure 1 shows an example of a kinetic reaction
profile and the first-order plot constructed from it. The reaction
Copyright Ó 2004 The Chemical Society of Japan

Chemistry Letters Vol.33, No.12 (2004)
1543
6
2
1
0
1
2
5
4
3
2
1
-
-
4
1
0 °C
0 °C
25 °C
3
5 °C
15 °C
0
0
50
100
150
200
250
0
100
200
P/MPa
300
P/MPa
Figure 2. Pressure dependence of the first-order rate constant
ꢁ
1
Figure 3. Pressure dependence of the first-order rate constant
k1/s ) in PC.
(
k1/s ) in AN.
ꢁ1
(
usually an indication of a shift of the rate-determining step, how-
ever, the present reaction is a one-step rate process and the pres-
sure-induced retardation was difficult to understand if we assume
fast solvent thermal fluctuations at all pressures. The most rea-
sonable explanation would be that solvent thermal fluctuations
became too slow to maintain the thermodynamic equilibrium be-
tween the initial and the transition state and so-called dynamic
References and Notes
1
A. Draljaca, C. D. Hubbard, R. van Eldik, T. Asano, M. V.
Basilevsky, and W. J. le Noble, Chem. Rev., 98, 2167
(1998) and earlier reviews cited therein.
2
3
M. Marinez and M.-A. Pitarque, J. Chem. Soc., Dalton
Trans., 1995, 4107.
M. Nanasawa, in ‘‘Organic Photochromic and Thermochro-
mic Compounds. Vol. 1. Main Photochromic Families,’’
ed. by J. C. Crano and R. J. Guglielmetti, Plenum, New York
(1999), Chap. 9, p 341.
9
solvent effects were observed. The rate-controlling factor might
z
have changed from the activation free energy, ꢀG , at low pres-
sures to the rate of solvent thermal fluctuations at high pressures.
As a measure of relative solvent fluctuation rate, the bulk
solvent viscosity is often used. Although high-pressure viscosity
of the present solvents are not available, the viscosity of PC and
4
5
6
7
8
9
Y. Kondo, A. Zanka, and S. Kusabayashi, J. Chem. Soc.,
Perkin Trans. 2, 1985, 827.
Y. Kondo, S. Hirano, and N. Tokura, J. Chem. Soc., Perkin
Trans. 2, 1979, 1738.
Y. Kondo, T. Yamada, and S. Kusabayashi, J. Chem. Soc.,
Perkin Trans. 2, 1981, 414.
T. Yamada, Y. Kondo, and S. Kusabayashi, Bull. Chem. Soc.
Jpn., 55, 2911 (1982).
At pressures about 210 MPa, it was not possible to obtain a
sufficiently large absorption increase by the irradiation.
N. Agmon and J. J. Hopfield, J. Chem. Phys., 78, 6947
(1983); N. Agmon and J. J. Hopfield, J. Chem. Phys., 79,
2043 (1983).
ꢁ
3
ꢁ4
AN are 2:53 ꢃ 10 Pa s and 3:41 ꢃ 10 Pa s, respectively at
ꢂ
2
5 C, and their pressure coefficients would be larger in PC
which has a five-membered ring with a methyl substituent and
a carbonyl oxygen than in linearly shaped AN. Therefore, it
would be reasonable to observe dynamic solvent effects in PC
but not in AN. The fact that the pressure where the maximum
ꢂ
rate constant was observed shifted from 210 MPa at 40 C to
ꢂ
150 MPa at 10 C supports this view.
Further kinetic measurements at other temperatures and in
other solvents are under way and the results will be reported
when they are completed.
Published on the web (Advance View) October 30, 2004; DOI 10.1246/cl.2004.1542