Electron transfer from cyclohexane to photoexcited aromatic ions. Generation and kinetics of high-mobility solvent holes

Article abstract of DOI:10.1021/jp9624717

In cyclohexane solutions, aromatic radical cations are produced by two-photon (248 nm) photoionization of aromatic molecules. A novel excited state chemistry is observed; i.e., excitation of these radical cations with a 532 nm laser pulse, delayed by several microseconds, produces the high mobility cation of cyclohexane. The difference of the signals with and without the 532 nm pulse shows the decay of the cyclohexane radical cation free of complications from geminate ion recombination. The biexponential nature of this decay suggests an equilibrium involving the high mobility cation, such as was proposed to explain previous experimental results.

Full text of DOI:10.1021/jp9624717

2120
J. Phys. Chem. A 1997, 101, 2120-2123
Electron Transfer from Cyclohexane to Photoexcited Aromatic Ions. Generation and
†
Kinetics of High-Mobility Solvent Holes
‡
I. A. Shkrob,* M. C. Sauer, Jr., K. H. Schmidt, A. D. Liu, J. Yan, and A. D. Trifunac*
Chemistry DiVision , Argonne National Laboratory, Argonne, Illinois 60439
X
ReceiVed: August 14, 1996; In Final Form: NoVember 25, 1996
In cyclohexane solutions, aromatic radical cations are produced by two-photon (248 nm) photoionization of
aromatic molecules. A novel excited state chemistry is observed; i.e., excitation of these radical cations with
a 532 nm laser pulse, delayed by several microseconds, produces the high mobility cation of cyclohexane.
The difference of the signals with and without the 532 nm pulse shows the decay of the cyclohexane radical
cation free of complications from geminate ion recombination. The biexponential nature of this decay suggests
an equilibrium involving the high mobility cation, such as was proposed to explain previous experimental
results.
Introduction
In a previous publication, we reported that in 4 eV (308 nm)
and 5 eV (248 nm) laser photolysis of 10 -10 mol dm
tween conformers of the cyclohexane radical cation. However,
results obtained with the delayed 532 nm pulse indicate that
the detailed nature of the observed biexponential decay is
affected by the purity of the cyclohexane. Briefly, we observed
that the decay of the signal induced by the 532 nm pulse
depended on the time elapsed after the purification of the
solvent. Most of the results were obtained with solvent for
which more than a few days had elapsed. Freshly purified
solvent was observed to result in slower decay of the conductiv-
ity. Analysis of these results indicates that an equilibrium is
still required to explain the bimodal decay, but the equilibrium
is required to involve an impurity, X, which acts as a shallow
trap as shown in reaction (1). Because the effect of this impurity
increases with time after purification, it is likely to be a result
of a slow oxidation process. Attempts to identify it have so
far been unsuccessful.
1
-
5
-6
-3
anthracene in cyclohexane the light flux dependence of the dc
conductivity signal from solvent radical cation indicated that
its production is a three-photon process. The first two photons,
absorbed by the ground state of anthracene and the first excited
singlet state, sequentially, result in ionization of anthracene. The
third photon was proposed to be absorbed by the anthracene
•
+
radical cation, resulting in production of cyclohexane by
electron transfer from the solvent to the aromatic ion. Due to
extremely rapid hopping of the charge between the neighboring
cyclohexane molecules, the solvent radical cation (also referred
to as a high mobility ion, HMI) exhibits a mobility that is ca.
5
0 times greater than the mobility of other molecular ions in
2-6
this liquid.
In high purity cyclohexane, the lifetime of the
-2
2
HMI is ∼500 ns and its mobility is (1.6-1.8) × 10 cm /(V
• +
• +
2
+ X
+ X
(1)
s). Thus, for cyclohexane, dc conductivity provides a sensitive
method of detecting solvent radical cations in a solution that
contains many other ions and excited molecules.3
The high mobility is due to rapid resonant charge transfer
between the chair form radical cation and the chair form neutral
molecule (which accounts for all but ≈10 ppm of the cyclo-
In the present work, a second laser pulse was used to excite
the aromatic ion with 2.3 eV (532 nm) photons several
microseconds after the first 248 nm pulse, thereby producing
the HMI. By determining the difference of the conductivity
kinetics observed with and without the 532 nm pulse, it is
possible to study the decay of the HMI in much detail. In
previous experiments that did not use this “delayed pulse”
technique, i.e., using just the excimer laser pulse, this decay is
more difficult to study, even in the neat solvent, because the
decay of HMI by geminate recombination and by homogeneous
neutralization must be taken into account, and in addition, part
of the conductivity is due to ions formed by the laser pulse that
9
hexane ). We cannot be certain that our previous suggestion
that reaction 2 contributes to the bimodal decay is invalid,
because if it occurs, its contribution would not be determinable
unless we knew, and were able to vary, the concentration of
impurity, X. In the purest samples, the combined concentrations
of X and the twist/boat form of cyclohexane must be about 40
µM to explain the observed signal decay.
• +
• +
+
+
(2)
2
are secondary ions that do not have high mobility. Still,
multiparameter fitting was used to unravel this complex kinetics.
These earlier computer fitting results showed that for decalins
Before equilibrium is established, in several tens of nano-
seconds, fast scavenging of HMI is observed. After the
equilibrium is established, the effective scavenging constant falls
the decay kinetics of free HMI was pseudo-first-order, but for
cyclohexane bimodal decay kinetics was required.2 An ex-
2
,4
by 2-3 times.
In this article we demonstrate how this
planation was proposed which involved an equilibrium be-
complex scavenging kinetics can be investigated using two-
color laser photoconductivity.
†
Work performed under the auspices of the Office of Basic Energy
Sciences, Division of Chemical Science, US-DOE, under Contract W-31-
Experimental Section
1
09-ENG-38.
*
To whom correspondence should be addressed.
On leave from Beijing Normal University.
‡
The aromatic solutes were the highest purity available from
Aldrich and were used as received. The cyclohexane (Baxter)
X
Abstract published in AdVance ACS Abstracts, February 15, 1997.
S1089-5639(96)02471-1 CCC: $14.00 © 1997 American Chemical Society

ET from Cyclohexane to Photoexcited Aromatic Ions
J. Phys. Chem. A, Vol. 101, No. 11, 1997 2121
was passed 4-5 times through a 1 m column of activated silica
gel. The purity was checked by the UV absorption and the
lifetime of HMI. Pulses of 248 nm photons (5 eV) from a
Lambda Physik LPX 120i laser (≈20 ns fwhm, ≈30 mJ/pulse)
were used to produce the HMI in CO2-saturated solutions. CO2
serves as an electron scavenger. The solutions were irradiated
in a cylindrical 15 mL Suprasil cell with a 4 cm optical path.
The cell has two rectangular 4 × 1.2 cm platinum electrodes
separated by 0.65 cm, to which a dc voltage of 5 kV was applied.
The signal due to the displacement current was amplified 100
times and acquired on a DSA-601 digitizing signal averager.
Except for the use of a different excimer laser, the techniques
related to the 248 nm excitation are the same as we have recently
2
described.
The 248 nm pulse entered the cell from one end; the optical
system and apertures resulted in about 17 mJ per pulse over a
2
× 5 mm area throughout the length of the cell. A 3 ns fwhm,
≈
60 mJ pulse of 532 nm light from a Continuum Model 8010
Figure 1. (a) Change in the conductivity induced by the 532 nm laser
pulse (40 mJ) from a Nd:YAG laser in two-color photolysis of 8 ×
Nd:YAG laser entered the cell from the opposite end; the 532
nm beam is circular with a diameter of ∼5 mm. The zone
10-6 mol dm-3 triphenylene in CO -saturated cyclohexane. A 20 ns
2
-
2
irradiated by both lasers was exposed to ∼0.17 J cm of 248
pulse (248 nm, 30 mJ) occurred at time shown by a broad arrow; the
532 nm pulse was applied after (i) 30 ns, (ii) 270 ns (shown by an
arrow), and (iii) 2 µs. The change in the conductivity is due to formation
of HMI by photoexcitation of the triphenylene radical cation. (b) Trace
iii from (a); the conductivity induced by the 248 nm pulse is subtracted.
The difference trace (the decay kinetics of HMI) is fit by a sum of two
exponentials. The slow component with a lifetime of 260 ns is given
by the broken line; the fast component has a lifetime of 30 ns.
-
2
nm light and ∼0.3 J cm of 532 nm light. Both lasers were
run at 2 Hz; the jitter between the positions in time of the two
pulses was ca. (3 ns. This jitter determines the time resolution
of the kinetics presented here. For data collection, the DSA-
601 was triggered either by a portion of the light from the 248
nm pulse (to observe the signal induced by it and the delayed
signal induced by the 532 nm pulse) or by the 532 nm light (to
observe the signal induced by the 532 nm pulse). In the former
case, the signal could be observed with just the 248 nm pulse
or with both pulses. In the latter case, a subtraction was done
with the digitizer: the signal using just the 248 nm pulse
However, a two-photon process can also fit the observations if
•
+
the lifetime of an intermediate state (Ar2 *) of the aromatic,
produced by the first 532 nm photon , is about 4 ns, and if the
quantum yield for producing HMI when this state absorbs the
second 532 nm photon is about 0.01. This ambiguity arises
mainly because of the high flux of 2.3 eV light relative to the
concentration of the aromatic radical cation.
(blocking the 532 nm pulse from the cell with a shutter) was
subtracted from the signal obtained with both pulses. Each
difference signal presented here was obtained by averaging the
results from three samples; each sample received 16-248 nm
pulses and 8-532 nm pulses, the 532 nm pulses occurring at a
predetermined delay after every second 248 nm pulse.
The HMI (solvent hole) produced by the 532 nm light by
charge transfer from the excited aromatic radical cation migrates
2-3 nm away from the transfer site before thermalization. Upon
thermalization (which occurs in 1-2 ps), the charge on the hole
will sometimes return to the aromatic molecule formed in the
transfer; this process results in the regeneration of the ions of
the aromatic molecule. This geminate reaction is over in <1
Results and Discussion
Photophysics. CO2-saturated cyclohexane containing aro-
8
matic solutes whose radical cations have significant absorption
ns. Those HMI that escape this geminate recombination can
7
at 532 nm, such as triphenylene (Tp), biphenyl, fluoranthene,
eventually react homogeneously with aromatic solute and
-
5
-3
and anthracene, were pulsed with 248 nm light. The 532 nm
pulse was introduced several microseconds later. At these delay
times, most HMI generated by the first pulse from the excimer
impurity (∼10 mol dm ). The concentration of free ions in
our laser experiment (at the time of 532 nm laser pulse) is on
-
8
-3
the order of 10 mol dm , and the second-order neutralization
2
2
laser have been scavenged by the aromatic solute and impurity.
of HMI can be neglected. In addition to the occurrence of
The geminate decay due to charge recombination with anions
“hole injection” in cyclohexane, we have observed, using the
techniques described above, the same process in two other
liquids, decalin and isooctane.
2,5
is also essentially completed.
The photoexcitation with the
2
.3 eV light promotes the transfer of the valence band electron
from the solvent to the aromatic ion; the photophysics and
energetics of this process will be a subject of a future
publication. We believe that, following the transfer, a short-
lived highly mobile “intrinsic” hole with excess energy of at
least a few tenths of an electronvolt is formed. From the
magnitude of the signals observed, we estimate that less than
Decay Kinetics. Figure 1a shows typical kinetics obtained
in the two-color experiment. The 532 nm light is absorbed by
an aromatic ion created by the 248 nm pulse yielding a
conductivity signal from HMI, which then “converges” to the
conductivity from ions of the original mobility. The conductiv-
ity signal induced by the 532 nm laser pulse is a large fraction
of the signal induced by the 248 nm pulse. The amplitude of
the 532 nm signal correlates qualitatively with the absorbance
expected for the aromatic radical cations that were produced
from the aromatics listed earlier; this signal is linear with the
photon flux. The 532 nm pulse alone was determined to
produce no conductivity signal in any of the systems studied.
The possibility exists that the 532 nm pulse could ionize a
neutral product from the 248 nm pulse, for example, an aromatic
singlet or triplet state. The occurrence of this can be ruled out
1
0% of the aromatic radical cations are converted to HMI at
the highest 532 nm intensity used. For most solutes used, a
single 2.3 eV photon can produce a hole with 0.5-1 eV excess
energy. However, for perylene one photon should not be
sufficient to ionize cyclohexane. The signal from all solutes is
linear with and proportional to the light flux between 0.02 and
-
2
0
.2 J cm , within experimental error. Simulations have been
made which show that if the production of HMI is a single-
photon process, a quantum yield of just a few percent is required.

2
122 J. Phys. Chem. A, Vol. 101, No. 11, 1997
Shkrob et al.
SCHEME 1: Scavenging Kinetics of Solvent Radical
TABLE 1: Rate Constants for Reaction of the Cyclohexane
Radical Cation with Various Reactants at 296 K
a
Cations in Cyclohexane
1
1
11
k
HM,S × 10 ,
k
HM,S × 10 ,
-
1
3
-1
-1
3
-1
reactant, S
mol dm s
reactant, S
decalinc
cyclohexene
2-propanol
cyclohexene oxide
triethylamine
mol dm s
b
triphenylene
anthracene
biphenyl
fluoranthene
perylene
benzene
4.2
3.6
4.2
3.1
3.3
3.0
3.8
2.3
1.4
3.1
1.8
b
b
b
b
•
+
of the solvent radical cation and X , a normally diffusing ion
a
Obtained by multitrace fitting of the scavenging kinetics using eq
(Scheme 1). The evolution of their concentrations is given by
1
0
-1
3
-1
7
with kND ) 1.5 × 10 mol dm s . The estimated accuracy of
these values is ≈(10%. The 532 nm pulse was delayed by 2-5 µs
•+
b
d[HMI]/dt ) -(k + k )[HMI] + k [X ]
(3)
relative to the 248 nm pulse. Was used both as a scavenger and a
1
3
-1
c
photosensitizer. 67% cis, 33% trans.
•
+
•+
d[X ]/dt ) k [HMI] - (k + k )[X ]
(4)
(5)
(6)
(7)
1
-1
4
because the decay of the difference signal does not exhibit the
•
-
expected geminate decay (of CO2 with the cation that would
have to be produced) followed by second-order ion recombina-
tion; also, the 532 nm pulse produced no “permanent” change
in conductivity, i.e., no net change in the free ion concentration
resulted. Furthermore, the fact that the difference signal does
not decrease appreciably when the delay between the pulses is
changed from 10 ns to several microseconds rules out ionization
of an aromatic singlet state. Involvement of aromatic triplet
states was ruled out by experiments where the triplet quencher,
O2, was used instead of CO2, and the difference signal for a
k3 ) kHM,S[S] + kHM,Tp[Tp] + k5
k ) k + kND([S] + [Im] + [Tp])
4
2
k5 ) kHM,Im[Im]
where the three kHM’s and kND are scavenging rate constants
for high mobility and normally diffusing forms, respectively,
with added scavenger, [S], any impurity, [Im] (which is a
permanent trap for HMI, in contrast to X), and aromatic
photosensitizer, [Tp]; k2 is the rate of transformation of X• to
a normally diffusing ion that can no longer transfer charge to
2
-15 µs delay was not eliminated.
Figure 1b shows the difference trace obtained as described
+
in the Experimental Section. The 532 nm pulse was applied 2
µs after the 248 nm pulse. Because the additional conductivity
induced by the 532 nm photons is entirely due to the HMI (and
is proportional to its concentration), the trace in Figure 1b
represents the decay kinetics of HMI. This decay is not
monoexponential, that is, the decay kinetics of HMI in cyclo-
hexane is not, as was believed in the past,5,6 pseudo-first-order.
We found that the kinetics can be fit by a sum of two
exponential functions, at any concentration of the aromatic
sensitizer and added scavenger. For the data shown in Figure
1
0
-1
3
-1
cyclohexane. For kND we assumed 1.5 × 10 mol dm s ,
typical for a diffusion-controlled scavenging reaction in
2
-6
-1
cyclohexane.
Let us introduce τ ) (k1 + k-1) , the time
constant of the equilibrium, and K ) k1/k-1, the “equilibrium
constant”. Note that k1 is actually kHM,X[X], and therefore K
and τ depend on [X] as we have observed in experiments where
better purification, which reduces [X], causes the decay of the
conductivity to be slower and thereby alters the values of K
and τ. The solution of eqs 3 and 4 gives
-
6
-3
1
b for cyclohexane solution with 8 × 10 mol dm of
triphenylene, the fast component has a lifetime of ca. 30 ns
corresponding to reactions depleting HMI during the approach
to equilibrium), and the slow component has the lifetime of
60 ns (corresponding to the disappearance of HMI after
[
HMI]/[HMI]t)0 ) const exp(-k t) + (1 - const) exp(-k t)
s
f
(8)
(
where kf,s ) (τ- + k + k ( r)/2 are the rate constants of the
1
3
4
2
fast and slow components, respectively, const ) (k-1 + k -
4
equilibrium is established). The latter lifetime is somewhat
shorter than the lifetime of HMI in neat cyclohexane. This is
expected, since the aromatic solutes are known to react with
HMI in cyclohexane (our measurements give the scavenging
constants ca. 4 × 10 mol dm s for low-IP aromatic
solutes, Table 1). The fast component has a lifetime that is
close to the lifetime, determined by transient absorption spec-
trophotometry, for charge transfer from HMI to aromatic solute
ks)/r, and
-
1
2
1/2
r ) {(τ + k + k ) - 4(k k + k k + k k )}
(9)
3
4
-1
3
1
4
3 4
11
-1
3
-1
Analysis of eq 7 shows that a family of decay curves for
different scavenger concentrations [S] (see Figure 2) cannot be
fit in a single way. While the scavenging constants kHM,S and
kHM,Tp can be found unambiguously, the values of K and τ
depend on [X], [Im], and the value of k2. The scavenging of
HMI by impurity Im shifts the equilibrium to the left side, while
4
molecules. As is emphasized above, the fast component cannot
result from the geminate reaction of HMI with the parent
aromatic molecule, as this reaction is complete in 1 ns. In
8
•+
the transformation of X by k2 shifts it to the right side. To
refs 2 and 4, we argued that the lifetime of the fast component
is close to the relaxation time of the equilibrium.
Biexponential scavenging kinetics of HMI in cyclohexane
can be understood quantitatively in terms of an equilibrium
model similar to that suggested in refs 2-4. Bearing in mind
compensate for these shifts, the equilibrium constant must be
changed accordingly. Lower values of K result if k2 ) 0; higher
values result if [Im] ) 0. Because of the involvement of the
impurity X, the significance of the experimental values of K
and τ is diminished, and we give only values determined from
results on scavenging of the HMI as a function of [triphenylene]
to show the trends. Fixing k2 ) 0, the results of the fitting
gave K ∼ 0.2 and τ ∼ 100 ns for the higher purity solvent and
K ∼ 0.4 and τ ∼ 50 ns for the less pure solvent (smaller [X]).
Fixing [Im] ) 0, the results of the fitting gave K ∼ 1 and τ ∼
120 ns for the higher purity solvent and K ∼ 0.8 and τ ∼ 50 ns
for the lower purity.
-
8
that the concentrations of free ions in our experiments are ∼10
-
3
mol dm , it may be seen that reaction 1 cannot change the
concentration of X (or of the twist/boat cyclohexane molecule,
if reaction 2 is involved) to an appreciable extent, because the
concentration of X must be more than 2 orders of magnitude
higher to give the observed decay kinetics. Therefore, equi-
librium (1) can be reduced to that between the chair form (HMI)

ET from Cyclohexane to Photoexcited Aromatic Ions
J. Phys. Chem. A, Vol. 101, No. 11, 1997 2123
The biexponential decay of the conductivity signal from the
high mobility cyclohexane cation is shown to be consistent with
the existence of an equilibrium between this cation and an ion
of a shallow hole trap (impurity ion or a conformer ion).
Establishment of this equilibrium is responsible for a fast
component in the decay of the high mobility ion, as we proposed
on the basis of observations on the pulse radiolysis of cyclo-
4
hexane solutions.
Acknowledgment. We acknowledge stimulating discussions
with Drs. J. M. Warman, M. P. de Haas, and W. F. Schmidt.
References and Notes
(
1) Liu, A.; Sauer, Jr., M. C.; Trifunac, A. D. J. Phys. Chem. 1993,
7, 11265.
2) Sauer, Jr., M. C.; Shkrob, I. A.; Yan, J.; Schmidt, K.; Trifunac, A.
9
(
D. J. Phys. Chem. 1996, 100, 11325.
Figure 2. Multitrace fitting of the scavenging kinetics using eq 7, for
a family of difference traces obtained by 532 nm photolysis of 8 ×
(3) Trifunac, A. D.; Sauer, Jr., M. C.; Shkrob, I. A.; Werst, D. W.
Acta Chem. Scand., in press.
-
6
-3
1
0
mol dm triphenylene with variable amounts of decalin, a good
(4) Shkrob, I. A.; Sauer, Jr., M. C.; Trifunac, A. D. J. Phys. Chem.
1996, 100, 7237.
(5) Warman, J. M. The Dynamics of Electrons and Ions in Non-Polar
Liquids, IRI 134-81-23; Delft, The Netherlands, 1981; The Study of Fast
Processes and Transient Species by Electron-Pulse Radiolysis; Baxendale,
J. H., Busi, F., Eds.; Reidel: Dordrecht, The Netherlands, 1982.
scavenger of HMI, in cyclohexane. The 532 nm pulse was applied 2
µs after the 248 nm pulse. The decay curve obtained for a solution
without decalin is given by the filled circles (the top trace). The
bimodality of the scavenging kinetics is well distinguished: in the first
5
0 ns the scavenging rates are notably higher than at the later times.
(6) Hummel, A.; Luthjens, L. H. J. Chem. Phys. 1973, 59, 654. Zador,
Table 1 summarizes the results of least-squares fits using eq
on multitrace data sets such as those shown in Figure 2, for
E.; Warman, J. M.; Hummel, A. Chem. Phys. Lett. 1973, 23, 363; 75, 914;
J. Chem. Phys. 1975, 62, 3897; J. Chem. Soc., Faraday Trans. 1 1979. de
Haas, M. P.; Warman, J. M.; Infelta, P. P.; Hummel, A. Chem. Phys. Lett.
7
several scavengers. The rate constants of scavenging of HMI
by the photosensitizers and scavengers were determined by
varying their concentrations. Importantly, the values of K and
τ obtained from the least-squares fitting were independent of
the scavenger, despite the 3-fold spread in scavenging rate
constants (see Table 1). We believe that the values of kHM,S
given in Table 1 are, so far, the most accurate HMI reaction
rate constants obtained.
1975, 31, 382; Chem. Phys. Lett. 1976, 43, 321; Can. J. Chem. 1977, 55,
2249. Luthjens, L. H.; de Leng, H. C.; van den Ende, C. A. M.; Hummel
A. Proc. 5th Symp. Radiat. Chem. 1982, 471. Hummel, A.; Luthjens, L. H.
J. Radioanal. Nucl. Chem., Art. 1986, 101, 293. Anisimov, O. A.; Warman,
J. M.; de Haas, M. P.; de Leng, H. Chem. Phys. Lett. 1987, 137, 365.
Warman, J. M.; de Leng H. C.; de Haas, M. P.; Anisimov, O. A. Radiat.
Phys. Chem. 1990, 36, 185.
(7) Shida, T. Electronic Absorption Spectra of Radical Ions, Elsevier:
New York, 1988.
(
8) In the asymptotic limit (t > τD) the decay kinetics of HMI in the
geminate pairs with aromatic molecules is given by P(t) ) P∞{1 + (τD/
Conclusion
πt)1/2}, where P is the escape probability, τ ) R2/D ∼ 30 ps, R is the
∞
D
2
scavenging radius (≈0.9 nm), D is the diffusion coefficient of HMI (∼3
This work provides definitive evidence for the occurrence
of “hole injection”, the transfer of the valence band electron
from the solvent to a photoexcited solute ion, and is an
interesting example of new chemistry of highly excited radical
cations of aromatic molecules. The quantum yield of the
process is only a few percent or less; if this low yield is due to
a high probability of recombination, rather than to an in-
efficiency in the electron transfer, one can be estimate that the
mean separation between the solvent hole and the aromatic
molecule need be only somewhat less than a nanometer.
Presumably, this separation could result from movement of the
hole while it still has some excess energy. On the other hand,
if the inefficiency is entirely due to the electron transfer, the
separation would have to be several nanometers.
-4
2
11
×
10 cm /s). If the space distribution of HMI around the aromatic solute
2
immediately after thermalization of the hole is Gaussian, p(r) ∝ r exp(-
2
[
2
r/b] ), the escape probability can be estimated as P∞ ) erfc (R/b). For b ∼
-3 nm, P∞ ∼ 0.5-0.65; a value of P∞ only ≈0.1 or greater is needed to
explain the observed yield of HMI.
9) Kellie, G. M.; Riddell, F. G. Top. Stereochem. 1974, 8, 225.
(
Squillacote, M.; Sheridan, R. S.; Chapman, O. L.; Anet, F. A. L. J. Am.
Chem. Soc. 1967, 89, 760; 1975, 97, 3244. Pickett, H. M.; Strauss, H. L.
J. Am. Chem. Soc. 1970, 92, 7281.
(10) Iwasaki, M.; Toriyama, K.; Nunome, K. Faraday Discuss. Chem.
Soc. 1984, 78, 19; J. Chem. Soc., Chem. Commun. 1984, 143. Lund, A.;
Lindgren, M.; Lunell, S.; Maruani, J. in Molecules in Physcis, Chemistry,
and Biology; Kluwer: Amsterdam, Vol. 3, 1989; p 259. Lunell, S.; Huang,
M. B.; Claesson, O.; Lund, A. J. Chem. Phys. 1985, 82, 5121.
(
11) Rice, S. A. Diffusion-Limited Reactions. In Chemical Kinetics;
Bamford, C. H., Tipper, C. F. H., Compton, R. G., Eds.; Elsevier: New
York, 1985; Vol. 25, Chapter 1.