A mechanism of photodissociation of diphenylmethane to a diphenylmethyl radical in solution

Article abstract of DOI:10.1063/1.476688

The photodissociation of diphenylmethane by excitation to the S₁ state at 266 nm in n-heptane solution is studied by nanosecond fluorescence and absorption spectroscopy. The formation of the diphenylmethyl radical is identified by its fluoresce

Full text of DOI:10.1063/1.476688

A mechanism of photodissociation of diphenylmethane to a diphenylmethyl radical in
solution
M. Fujiwara, A. Yamasaki, K. Mishima, and K. Toyomi
Citation: The Journal of Chemical Physics 109, 1359 (1998); doi: 10.1063/1.476688
View online: http://dx.doi.org/10.1063/1.476688
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/109/4?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Laser-induced UV photodissociation of 2-bromo-2-nitropropane: Dynamics of OH and Br formation
J. Chem. Phys. 134, 044316 (2011); 10.1063/1.3532085
Photodissociation dynamics of hydroxybenzoic acids
J. Chem. Phys. 134, 034314 (2011); 10.1063/1.3526059
Photodissociation of the vinoxy radical through conical, and avoided, intersections
J. Chem. Phys. 117, 7198 (2002); 10.1063/1.1507587
Proton transfer reaction of 4-methyl-2,6-diacetylphenol and an analysis with AM1 potential-energy surfaces
J. Chem. Phys. 114, 1336 (2001); 10.1063/1.1329641
Photodissociation of C–H and C–O bonds of p -methoxytoluene and p -methoxybenzyl alcohol in solution
J. Chem. Phys. 107, 9354 (1997); 10.1063/1.475232
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
152.7.17.127 On: Wed, 10 Dec 2014 09:51:55

JOURNAL OF CHEMICAL PHYSICS
VOLUME 109, NUMBER 4
22 JULY 1998
A mechanism of photodissociation of diphenylmethane to a diphenylmethyl
radical in solution
M. Fujiwara, A. Yamasaki, K. Mishima, and K. Toyomi
Department of Chemistry, Faculty of Science, Hiroshima University, Kagamiyama,
Higashi-Hiroshima 739-8526, Japan
͑
Received 2 March 1998; accepted 13 April 1998͒
The photodissociation of diphenylmethane by excitation to the S state at 266 nm in n-heptane
1
solution is studied by nanosecond fluorescence and absorption spectroscopy. The formation of the
diphenylmethyl radical is identified by its fluorescence, which is induced by excitation at 308 nm,
7
Ϫ1
and by its absorption. The growth rate of (3.7Ϯ0.4)ϫ10 s for the radical is equal to the decay
7
Ϫ1
rate of (3.8Ϯ0.4)ϫ10 s for the precursor fluorescence. The quantum yield of the radical is of the
Ϫ3
order of ϳ10 . Neither dissociation to the radical nor intersystem crossing to the T state is
1
thermally activated, whereas activated internal conversion to the S state is observed. The formation
0
of the radical depends linearly on the photolysis pulse fluence. The data are consistent with a
mechanism that the molecule undergoes intersystem crossing from thermally equilibrated levels of
the S state to vibrationally excited levels of the T state at which it dissociates in competition with
1
1
vibrational relaxation. The mechanism is explained in terms of electronic coupling between the
precursor and product states. The S state does not correlate adiabatically to the ground state of the
1
C–H bond fission products, so intersystem crossing to the T state precedes dissociation. In the T1
1
state, avoided crossing between the ␲␲* ͑benzene͒ configuration and the ␴␴* ͑C–H͒ repulsive
configuration results in the adiabatic potential energy surface which evolves to the ground state of
the C–H bond fission products allowing rapid dissociation. © 1998 American Institute of Physics.
͓
S0021-9606͑98͒00428-0͔
I. INTRODUCTION
II. EXPERIMENT
Photoexcitation of methyl-substituted aromatic mol-
ecules has been known to induce C–H bond dissociation
yielding arylmethyl radicals.1 In the liquid phase ͑nonpolar
solution͒, the dissociation rate is limited by the decay rate of
For the two-pulse fluorescence measurement, a pulse of
the fourth harmonic ͑266 nm, 4–5 ns͒ of a Nd:YAG laser
͑Quanta-Ray GCR-11͒ was used as the photolysis light. The
fluorescence was induced by a second pulse ͑308 nm, 8–12
ns͒ from a XeCl excimer laser ͑Lumonics 500͒. The laser
power was attenuated with ND filters ͑Sigma FNDU-50C02-
–7
the thermally equilibrated S state, showing that dissociation
1
7
occurs via the thermally equilibrated S state. The mecha-
1
1
0
0, -20, -50͒ and measured with a thermopile detector ͑Ophir
3A-P, DGX͒. The delay time was adjusted with a digital
nism is explained in terms of electronic coupling. Since the
S1 state does not correlate adiabatically to the ground state of
delay generator ͑EG&G PAR 9650͒. The 266 and 308 nm
laser beams were focused coaxially on a 10ϫ10 mm quartz
cell to 1.2ϫ12 and 1.6ϫ4.0 mm rectangular profiles, respec-
tively. The fluorescence was collected into a monochromator
the products, intersystem crossing to the hot T state or in-
1
ternal conversion to the hot S state precedes dissociation.
0
Then, questions arise as to the C–H bond dissociation
channel in the liquid phase: ͑1͒ From which state does dis-
͑Ritsu MC-10N͒ at right angles with the laser beams and
sociation actually take place, the hot T state or the hot S0
1
detected with a photomultiplier ͑Hamamatsu R636͒. The sig-
nal was fed to a digital oscilloscope ͑Tektronix 2440, 500
MHz sampling͒, which was interfaced to a personal com-
puter ͑NEC PC-9801UV͒ for data storage.
state, if dissociation does not occur on the S state? ͑2͒ Does
1
no possibility exist that dissociation proceeds on the S state,
1
while the molecule stays on it? Experimental studies are nec-
essary to answer these questions.
In the present paper, the photodissociation of diphenyl-
methane in n-heptane solution is examined by nanosecond
fluorescence and absorption spectroscopy. Excitation of the
For the transient absorption measurement, the photolysis
light was the 266 nm laser pulse. The white light was pro-
vided by a Xe arc lamp ͑Ushio UXL-500D-O͒. The lamp
beam was collimated on the cell to a 0.8 mm diameter at
right angles with the laser beam. The distance from the laser
irradiation inner surface to the lamp beam center path in the
cell was 1.9 mm. The detection apparatus for the white light
was the same as for the fluorescence. A UV filter ͑Toshiba
UV-D33S, 250–400 nm͒ was placed in front of the mono-
chromator to reject visible light for the measurement in the
short wavelength region (Ͻ400 nm).
molecule to the S state at 266 nm produces the diphenyl-
1
methyl radical. Its fluorescence, which is induced by excita-
tion at 308 nm, and its absorption are observed. The rate,
quantum yield, activation energy, and number of photons for
dissociation are measured. The mechanism is identified that
the molecule undergoes intersystem crossing from the ther-
mally equilibrated S1 state to the hot T state at which it
1
dissociates in competition with vibrational relaxation.
The temperature of the cell was varied by passing water
0021-9606/98/109(4)/1359/7/$15.00
1359
© 1998 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
52.7.17.127 On: Wed, 10 Dec 2014 09:51:55
1

1360
J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Fujiwara et al.
FIG. 2. Transient absorption spectra of diphenylmethyl radical and diphe-
nylmethane triplet recorded after excitation of ͑a͒ diphenylmethane and ͑b͒
FIG. 1. Two-pulse fluorescence spectra of diphenylmethyl radical observed
by excitation with a 308 nm pulse at 1 ␮s after excitation of ͑a͒ diphenyl-
methane and ͑b͒ ␣-chlorodiphenylmethane with a 266 nm pulse.
␣-chlorodiphenylmethane with a 266 nm pulse.
in the 340–530 nm region, seen in the spectra from diphe-
from a circulator ͑Julabo F10-UC͒ through the cell holder
and measured by a thermocouple ͑Yokogawa 245903,
nylmethane, is assigned to its T state by comparison with
1
13
the band of the T state of benzene. The lifetime of the T1
1
245522͒. Unless otherwise noted, experiments were run at
state is determined to be ϳ3.4 ␮s.
room temperature ͑293 K͒.
These results show that the diphenylmethyl radical is
formed by C–H bond dissociation of diphenylmethane after
excitation with the 266 nm pulse.
Diphenylmethane ͑Tokyo Chemical, Ͼ99%͒, ␣-
chlorodiphenylmethane ͑Tokyo Chemical, Ͼ98%͒, and tolu-
ene ͑Kanto Chemical, Ͼ99%͒ were distilled. Naphthalene
͑
͑
Nacalai Tesque͒ was recrystallized from ethanol. n-Heptane
Dojindo, spectrosol͒ was used as received. The sample so-
B. Rates of radical growth, S decay, and T decay
1
1
Ϫ3
Ϫ3
lutions, containing diphenylmethane (5.5ϫ10 mol dm
)
To obtain knowledge of an excited state that leads to
dissociation, the growth rate of the diphenylmethyl radical
and ␣-chlorodiphenylmethane (5.0ϫ10 mol dmϪ3) with
Ϫ3
n-heptane, were degassed by freeze-pump-thaw cycles.
has been compared with the decay rates of the S and T1
1
states of diphenylmethane. The growth rate of the ground
state radical is evaluated from the dependence of its fluores-
cence intensity on the delay from the 266 nm ͑photolysis͒ to
the 308 nm ͑probe͒ pulse. ͑The absorption time profile of the
radical cannot be obtained in a 200 ns delay because the
precursor fluorescence interrupts observation of the absorp-
tion signal.͒ The decay rates of the thermally equilibrated S1
III. RESULTS
A. Identification of radical
In order to identify the fragment radical formed from
photodissociation of diphenylmethane, the nanosecond fluo-
rescence and absorption spectra have been measured by pho-
tolysis of diphenyl- and ␣-chlorodiphenylmethanes.
and T states of the precursor are determined from the time
1
The two-pulse fluorescence spectra, observed by excita-
tion with the 308 nm pulse after photolysis of diphenyl- and
profiles of its fluorescence and T ←T absorption, respec-
n
1
tively.
␣
-chlorodiphenylmethanes with the 266 nm pulse, are shown
The fluorescence time profile of diphenylmethane, ob-
served by excitation with the 266 nm pulse, is shown in Fig.
3͑a͒. The decay rate of the precursor fluorescence is deter-
in Fig. 1. Excitation with the 308 nm pulse alone does not
induce the fluorescence. The two spectra exhibit a band at
7
Ϫ1
570 nm, which is assigned to the diphenylmethyl radical and
mined to be (3.8Ϯ0.4)ϫ10 s . The dependence of the
fluorescence intensity of the diphenylmethyl radical on the
delay of the 308 nm pulse, recorded by photolysis of diphe-
nylmethane with the 266 nm pulse, is shown in Fig. 3͑b͒.
The growth rate of the radical fluorescence is calculated to be
8
–10
is similar to the bands reported previously.
The fluores-
cence lifetimes, determined from photolysis of diphenyl- and
-chlorodiphenylmethanes, are of a value of 230Ϯ20 ns and
␣
8,9
agree with the previous lifetimes. The fluorescence band
disappears at a 100 ␮s delay from the 266 to the 308 nm
pulse.
The transient absorption spectra, recorded by excitation
of diphenyl- and ␣-chlorodiphenylmethanes with the 266 nm
pulse, are shown in Fig. 2. A band at 328 nm, appearing in
7
Ϫ1
(3.7Ϯ0.4)ϫ10 s
.
It is evident that the growth rate of the radical fluores-
cence is equal to the decay rate of the precursor fluorescence.
This shows that dissociation occurs via the thermally equili-
brated S state of diphenylmethane.
1
the two spectra, is assigned to the diphenylmethyl radical
The absorption time profile of the T state of diphenyl-
methane is observed by excitation with the 266 nm pulse.
1
and is similar to the bands obtained previously.6
,8,11,12
This
absorption band disappears at a 100 ␮s delay. A broad band
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
The decay rate of the T ←T absorption is ϳ2.9
n
1
152.7.17.127 On: Wed, 10 Dec 2014 09:51:55

J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Fujiwara et al.
1361
2
66 nm pulse, though the extinction coefficient of the diphe-
nylmethyl cation is comparable with that of the diphenyl-
methyl radical. No absorption due to a stable product is seen
in the steady-state spectrum for the sample solution of diphe-
nylmethane after repetitive irradiation with the 266 nm pulse,
showing that no chemical intermediate yielding is practically
formed. The observation suggests that the S state of diphe-
1
nylmethane relaxes mainly by photophysical processes. The
energy dissipation pathway of the S state consists of four
1
channels: ͑1͒ intersystem crossing; ͑2͒ internal conversion;
͑
3͒ fluorescence; and ͑4͒ dissociation forming the diphenyl-
Ϫ3
methyl radical with the low (ϳ4.5ϫ10 ) yield. Then, the
internal conversion yield for diphenylmethane is calculated
to be ϳ0.38, which is similar to the value of 0.41 for
toluene.17
FIG. 3. Time evolution of fluorescence intensities of ͑a͒ diphenylmethane
and ͑b͒ diphenylmethyl radical produced after excitation of diphenyl-
methane with a 266 nm pulse. Fluorescence of diphenylmethyl radical is
induced by excitation with a 308 nm pulse, of which a delay time is an
abscissa for ͑b͒. ͑c͒ Time evolution of a 266 nm pulse. Monitoring wave-
lengths: ͑a͒ 300 nm and ͑b͒ 570 nm.
D. Thermal activation
If dissociation occurs after intersystem crossing to the T1
state or internal conversion to the S state, the yield of dis-
sociation will be affected by the efficiency of intersystem
crossing or internal conversion. This efficiency may depend
upon temperature. Alternatively, if dissociation proceeds on
0
ϫ10 sϪ1. It is clear that the growth rate of the radical fluo-
5
the S surface after thermal equilibration, there will exist a
1
barrier along the reaction coordinate. It is because the S1
state is an associative state with ␲␲* ͑benzene͒ character.
rescence is much faster than the decay rate of the T ←T
n
1
absorption. This excludes a possibility that the diphenylm-
The energy required for crossing over this barrier may be
provided by thermal excitation. To clarify these points, the
temperature dependence of the yields of dissociation, inter-
system crossing, and fluorescence has been investigated
along with that of the decay rate of the thermally equilibrated
S1 state for diphenylmethane.
ethyl radical is formed via the thermally equilibrated T state
of diphenylmethane.
1
C. Branching of energy flow from S state
1
To know the branching of the energy dissipation pro-
cesses from the S state, the quantum yields of dissociation,
The intensities of the diphenylmethyl absorption, Tn
1
intersystem crossing, and fluorescence have been estimated
for diphenylmethane.
←T absorption, and fluorescence are measured as well as
1
the fluorescence decay rate by excitation of diphenylmethane
with the 266 nm pulse. The fluorescence intensity is inte-
grated by time. In Fig. 4 are shown the semilogarithmic plots
of the absorption and fluorescence intensities and the fluo-
rescence decay rate as a function of the reciprocal of tem-
perature. When temperature is raised, the intensities of the
radical absorption, T ←T absorption, and fluorescence de-
The yields of the diphenylmethyl radical and the T state
1
for diphenylmethane are measured by comparing their absor-
bances with the absorbance of the T state for naphthalene as
1
the standard by excitation with the 266 nm pulse. The fluo-
rescence yield for diphenylmethane is measured by compari-
son of its fluorescence intensity, which is integrated by time,
with that of toluene. The extinction coefficients for the
n
1
crease, whereas the fluorescence decay rate increases. The
increase of the decay rate of the thermally equilibrated S1
state on raising temperature is attributed to thermally acti-
vated internal conversion. Since intersystem crossing, inter-
nal conversion, and fluorescence compete from the thermally
diphenylmethyl radical and the T state of naphthalene are
1
4
3
Ϫ1
Ϫ1
14
4
3
2
.7ϫ10 dm mol cm
͑335 nm͒ and 1.4ϫ10 dm
mol cm ͑415 nm͒, respectively. The extinction coeffi-
Ϫ1
Ϫ1
15
cient for the T state of diphenylmethane is assumed to be
1
3
3
Ϫ1
Ϫ1
1
.0ϫ10 dm mol cm ͑360 nm͒ from that for the T1
equilibrated S state, this results in the decrease of the yields
1
1
3
state of benzene. The intersystem crossing yield for naph-
of intersystem crossing and fluorescence at high temperature.
To describe the temperature dependence in Fig. 4, the acti-
1
5
thalene is 0.68, and the fluorescence yield for toluene is
1
6
Ϫ1
0
.14. Using these values, an estimate for the diphenyl-
vation energy of 1150Ϯ190 cm for internal conversion is
Ϫ3
methyl yield of ϳ4.5ϫ10 is made from dissociation of
diphenylmethane. The C–H bond dissociation channel is
shown to be a minor one. The intersystem crossing yield for
diphenylmethane is estimated to be ϳ0.42 and is similar to
the value of 0.45 for toluene.17 The fluorescence yield for
diphenylmethane is measured to be ϳ0.20.
satisfactorily used. The yields of dissociation, intersystem
crossing, and fluorescence depend similarly on temperature,
and no thermal activation for dissociation or intersystem
crossing is found to be necessary. The S and S surfaces
1
0
may undergo crossing at a geometry accessible by thermal
excitation of quanta of promoter vibrations in the S state.
1
No absorption assignable to another chemical intermedi-
ate than the diphenylmethyl radical is observed in the tran-
sient spectra by excitation of diphenylmethane with the
The observation of activated internal conversion and
temperature-independent intersystem crossing for diphenyl-
methane is in accord with the previous one for benzene.18
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
152.7.17.127 On: Wed, 10 Dec 2014 09:51:55

1362
J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Fujiwara et al.
FIG. 5. Dependence of a fluorescence intensity of diphenylmethyl radical on
͑
a͒ 266 nm and ͑b͒ 308 nm pulse fluences. Fluorescence is induced by
excitation with a 308 nm pulse at 1 ␮s after excitation of diphenylmethane
FIG. 4. Temperature dependence of ͑a͒ an absorption intensity of diphenyl-
methyl radical, ͑b͒ an absorption intensity of diphenylmethane triplet, ͑c͒ a
fluorescence intensity of diphenylmethane, and ͑d͒ a fluorescence decay rate
of diphenylmethane observed by excitation with a 266 nm pulse. Monitoring
wavelengths: ͑a͒ 330 nm, ͑b͒ 360 nm, ͑c͒ and ͑d͒ 300 nm.
with a 266 nm pulse. Monitoring wavelength: 570 nm.
The logarithmic plot of the fluorescence intensity of the
diphenylmethyl radical versus the 266 nm pulse fluence
Ϫ2
(
0.26–21 mJ cm ) is shown in Fig. 5͑a͒. The efficiency of
From the result that neither dissociation nor intersystem
crossing requires thermal activation, dissociation is shown to
follow intersystem crossing and to take place from the vibra-
S ←S excitation with the 266 nm pulse increases from
n
1
0
2
2
.0028 to 0.21, when the fluence is increased from 0.26 to
Ϫ2
1 mJ cm . The efficiency of D ←D excitation with the
1
0
tionally excited T state. A channel is precluded that disso-
1
66 nm pulse is negligible in this fluence range. Observed
ciation occurs from the vibrationally excited S state, for the
0
points are linearly fitted with a slope of 0.96Ϯ0.11. It is
evident that dissociation occurs by 266 nm one-photon exci-
tation.
reason that dissociation is not thermally activated. If this
were to occur, thermally activated internal conversion must
precede dissociation, and the latter must be thermally acti-
vated. A channel is impossible that dissociation proceeds by
The fluorescence intensity of the diphenylmethyl radical
is plotted as a function of the 308 nm pulse fluence
crossing over a barrier on the S surface after thermal equili-
1
Ϫ2
(
0.20–18 mJ cm ) in Fig. 5͑b͒. The efficiency of D ←D
1
0
bration. For this to occur, thermal excitation must be re-
quired for dissociation. The experiment observes that disso-
ciation is not thermally activated.
excitation with the 308 nm pulse increases from 0.0071 to
0
.64, and that of D ←D excitation from 0.0088 to 0.80, on
n
1
Ϫ2
increasing the fluence from 0.20 to 18 mJ cm . Linear fit-
ting results in a slope of 0.88Ϯ0.12. It is shown that one 308
nm photon is required to populate the fluorescent state of the
E. One- or two-photon photochemistry
radical. Saturation behavior appearing at a higher fluence
The number of photons required for excitation of diphe-
nylmethane to its dissociative state has been determined
from the dependence of the fluorescence intensity of the
diphenylmethyl radical on the fluences of the 266 nm ͑pho-
tolysis͒ and 308 nm ͑probe͒ pulses. Beer’s law complication
Ϫ2
(
Ͼ18 mJ cm ) is understood to result from depletion of the
D and D states.
0
1
Ϫ2
is removed at the dilute radical concentration (Ͻ10 absor-
IV. DISCUSSION
bance͒. A rough estimate for the efficiencies of excitation of
A. Bond dissociation energy
the states into higher ones is made using the following ex-
3
3
Ϫ1
Ϫ1
tinction coefficients: 4.2ϫ10 dm mol cm ͑266 nm͒ for
Diphenylmethane dissociates by one-photon excitation
the S state of diphenylmethane assumed from the value for
to the S state at 266 nm ͑Fig. 5͒. The vibrational ground
1
1
13
2
3
Ϫ1
Ϫ1
Ϫ1
the S1 state of benzene; 6.8ϫ10 dm mol cm ͑266
level of the S state possesses energy of 37 100 cm . The
1
3
3
Ϫ1
Ϫ1
nm͒ and 6.0ϫ10 dm mol cm ͑308 nm͒ for the D state
bond dissociation energy for the benzylic C–H bond is esti-
0
14
4
3
Ϫ1
Ϫ1 19–23
of the diphenylmethyl radical; and 1.5ϫ10 dm mol
mated to be 28 900Ϯ700 cm
.
Clearly, the energy of
Ϫ1
cm
͑8308 nm͒ for the D1 state of the diphenylmethyl
the vibrational ground level of the S state is sufficient to
1
radical.
fragment the molecule into the radical.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
152.7.17.127 On: Wed, 10 Dec 2014 09:51:55

J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Fujiwara et al.
1363
B. Dissociation rate and yield
Dissociation occurs by intersystem crossing to vibra-
tionally excited levels of the T state for diphenylmethane.
1
11
Ϫ1
Since vibrational relaxation is extremely fast, ϳ10
s , in
the liquid phase, this leads to the result that the dissociation
rate is limited by the intersystem crossing rate from ther-
mally equilibrated levels of the S state ͑Fig. 3͒. The actual
1
dissociation rate from vibrationally excited levels of the T1
Ϫ3
state is not resolved. The low dissociation yield of ϳ10
results from competition with vibrational relaxation in the T1
state.
C. Barrier to dissociation
For diphenylmethane, since the T state is an associative
1
state with ␲␲*(benzene) character, dissociation will occur
from another dissociative state such as the ␴␴*(C–H) state.
The ␴␴*(C–H) state is the lowest excited state in alkane,
and the transition leading to it is of energy of
Ϫ1
ϳ65 000 cm . Thus, when dissociation proceeds on the T1
potential energy surface, avoided crossing must occur be-
tween the ␲␲*(benzene) and ␴␴*(C–H) configurations.
This forms a barrier along the C–H bond fission coordinate.
The experiment shows that neither dissociation nor in-
tersystem crossing is thermally activated ͑Fig. 4͒. After un-
FIG. 6. A state correlation diagram for C–H bond fission of diphenyl-
methane. States of a benzene ring of diphenylmethane and a benzyl part of
diphenylmethyl radical are represented with corresponding ones of benzene
(D6h) and benzyl radical (C2v), respectively. Symmetry representations
relative to a symmetry plane are noted for singlet ͑solid lines͒ and triplet
dergoing intersystem crossing to the T state without thermal
1
excitation, the molecule gains a large amount of vibrational
energy corresponding to the energy gap between the vibra-
tional ground levels of the S1 and T1 states. Then, it can
cross over the barrier to the repulsive state. A fraction of
those molecules which undergo intersystem crossing are
forced to dissociate.
͑
broken lines͒ states in parentheses. A benzene ring is perpendicular to a
symmetry plane, on which a C–H bond lies. RϭC H .
6
5
benzene ring. The symmetry of the potential energy surfaces
along the C–H bond fission coordinate is noticed.7
Photoexcitation promotes diphenylmethane to the
,24–29
D. Adiabatic potential energy surfaces
1
S ͓ B ,␲␲*(benzene)͔ state which is of AЉ symmetry
1
2u
The photodissociation pathway of p-methoxytoluene
forming the p-methoxybenzyl radical in the liquid phase has
been investigated recently.7 The C–H bond dissociation
channel occurs via thermally equilibrated levels of the S1
with respect to the symmetry (C ) plane perpendicular to the
s
2
excited benzene ring. The ground ( B ) and excited
2
2
͓ A ,␲␲*(benzyl)͔ states of the diphenylmethyl radical
2
possess AЈ and AЉ symmetries, respectively. The symmetry
2
state. The mechanism is explained that the S state does not
of the ground-state ( S) H atom is AЈ. The S (AЉ) state of
1
1
correlate adiabatically to the ground state of the products,
the precursor does not correlate adiabatically to the ground
2
2
allowing intersystem crossing to the T state or internal con-
singlet ( B ϩ S,AЈ) state of the asymptotic products, but
1
2
2
2
version to the S state to precede dissociation. The mecha-
does correlate to the excited singlet ( A ϩ S,AЉ) state of
0
2
nism is identified more precisely by the present experiments.
The electronic energy is considered to migrate between
the products. If the S state were to participate in C–H bond
fission, a symmetry-imposed energy barrier must be over-
come. Therefore, C–H bond fission cannot occur adiabati-
1
the two benzene rings of diphenylmethane in the S state,
1
because its absorption and fluorescence spectra are similar to
those of toluene. One of the two benzyl parts of the diphe-
nylmethyl radical is assumed to have such a planar structure
as the benzyl radical does, just after it is produced by C–H
bond fission from excitation of the corresponding benzene
ring. The most favorable geometry of stretch of the C–H
bond that leads to the planar structure of the benzyl part is
one in which it occurs in a plane perpendicular to the exited
benzene ring ͑see Fig. 6͒. The energy of this geometry is
comparable to that of other geometries of the stretched C–H
bond, even if they are distorted by vibrations. Thus, the ap-
proximation is made that the C–H bond fission coordinate
cally from the S potential energy surface. Instead, intersys-
1
3
tem crossing proceeds to the T ͓ B ,␲␲*(benzene)͔ state
1
1u
which is of AЈ symmetry relative to the symmetry (Cs)
plane perpendicular to the excited benzene ring. The
␴␴*(C–H) repulsive triplet state possesses AЈ symmetry.
For C–H bond fission to occur, the electronic configuration
is required to change from being localized on the benzene
(␲␲*) moiety to being associated with the C–H (␴␴*)
moiety. The configuration interaction matrix elements that
facilitate such a change in the electronic wave function may
be small but cannot be zero. Then, crossing between the
3
␲␲*(benzene)( B1u ,AЈ) and ␴␴*(C–H) ͑triplet, AЈ͒ elec-
lies in the symmetry (C ) plane perpendicular to the exited
tronic configurations is avoided from the requirement of the
s
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
152.7.17.127 On: Wed, 10 Dec 2014 09:51:55

1364
J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Fujiwara et al.
symmetry. This results in the AЈ triplet potential energy sur-
face which evolves adiabatically from ␲␲*(benzene) char-
acter in the Franck–Condon region to ␴␴*(C–H) character
beyond the barrier along the C–H bond fission coordinate.
Rapid C–H bond fission can occur directly on the resulting
sion channel for diphenylmethane is distinguished by inter-
system crossing to the T1 state with subsequent adiabatic
crossing to a ␴␴*(C–H) triplet state.
The dissociation mechanism of the C–O bond of
p-methoxybenzyl alcohol has been investigated by excitation
3
7
AЈ adiabatic potential energy surface from the T ( B ) state
to the S state at 266 nm in solution. The C–O bond fission
1
1u
1
2
2
of the precursor to the ground triplet ( B ϩ S) state of the
channel has occurred from vibrationally excited levels of the
S1 state. This makes a contrast to the C–H bond fission
channel, which proceeds via thermally equilibrated levels of
2
asymptotic products.
After intersystem crossing to the T state, the precursor
1
gains a large amount of vibrational energy equal to the en-
the S state. The difference between the mechanisms is due
1
ergy gap between the S and T states. If the barrier to dis-
to that of electronic coupling. For C–O bond dissociation,
avoided crossing on the S1 surface between the ␲␲*
(benzene)(AЉ) and np͑O͒␴*͑C-O͒(AЉ) configurations leads
to the adiabatic pathway to the ground state of the products.
1
1
sociation is so low that vibrational quanta other than the
C–H bond stretch quantum are highly excited, the nuclear
geometry is distorted and the symmetry restriction is dis-
turbed. Even in such a case, since both the T state of the
For C–H bond dissociation, the S (AЉ) state does not corre-
1
1
precursor and the ground triplet state of the products are the
lowest triplet states at the respective C–H bond lengths and
are correlated, the nuclear dynamics can sample the trajec-
tory toward C–H bond fission.
late adiabatically to the ground (AЈ) state of the products,
allowing intersystem crossing to the T state to occur, and
1
adiabatic crossing on the T1 surface from the ␲␲*
(benzene)(AЈ) to the ␴␴*(C–H)(AЈ) state results in the
pathway to the products.
On the S surface, intersystem crossing cannot compete
1
with vibrational relaxation, so dissociation follows vibra-
tional relaxation. This accounts for the experimental observ-
able that the dissociation rate is determined by the decay rate
The formation mechanism of the 3,5-dimethylbenzyl
radical from mesitylene has been observed by excitation of
the T state at 248, 337, or 405 nm in glassy matrices at 77
1
5
of thermally equilibrated levels of the S state. On the T1
K. The mechanism by excitation of the T state in glassy
1
1
surface, an adiabatic reaction pathway is allowed to lead
rapid dissociation. Since there exists a barrier to dissociation
on the T1 surface, statistical partitioning of excess vibra-
tional energy is required for crossing over it. This predicts
that dissociation occurs from vibrationally excited levels of
the T1 state after intramolecular vibrational redistribution.
matrices is different from that by one-photon excitation in
solution. The T1 state may live long enough to absorb a
photon regardless of the low dissociation efficiency in low-
temperature matrices.
The dissociation mechanism of toluene to the benzyl
radical has been studied by excitation to the S state at 193
3
Ϫ3
1–4
The low dissociation yield of ϳ10 is considered to result
nm in the gas phase. The bond fission channel, which is a
from competition with vibrational relaxation of which the
dominant one (ϳ0.75 yield͒, has occurred by internal con-
1
1
Ϫ1
rate is ϳ10
s
in the T state.
version to highly vibrationally excited levels of the S state.
1
0
6
Ϫ1
The measured rate (ϳ2.0ϫ10 s ) has been predicted by
the unimolecular reaction theory on the assumption that the
electronic energy is distributed over vibrational modes. The
channel by internal conversion does not occur in the liquid
E. Other dissociation mechanisms
The dissociation mechanism of diphenylmethane to the
diphenylmethyl radical has been observed by two-photon ex-
citation at 248 nm followed by ionization and deprotonation
6
Ϫ1
phase. Such slow dissociation (ϳ10 s ) may not compete
with vibrational relaxation (ϳ1011
s ) on the S0 state in the
Ϫ1
6
in acetonitrile, ethanol, and water solutions. The mechanism
liquid phase.
by two-photon excitation at 266 nm is not detected in
n-heptane solution by the present experiments. Ionization is
efficient in polar solutions such as acetonitrile, ethanol, and
water. If dissociation via ionization has a much higher quan-
tum yield than that via intersystem crossing, only the former
will be observed in polar solutions. Ionization is less effec-
tive in a nonpolar solution such as n-heptane. This may ac-
count for the results that dissociation via intersystem cross-
ing is observed and that dissociation via ionization is not
detected in a nonpolar solution.
V. CONCLUSIONS
Excitation of diphenylmethane to the S state at 266 nm
1
induces dissociation, forming the diphenylmethyl radical in
n-heptane solution. The growth rate of the radical is equal to
the decay rate of the thermally equilibrated S state. Thermal
1
activation is not required for dissociation to the radical or
intersystem crossing to the T1 state. The formation of the
radical follows one-photon excitation. The observation
shows a mechanism that the molecule undergoes intersystem
The dissociation mechanism of the C–X bond (XϭCl,
Br͒ of 1- and 2-(X-methyl͒naphthalenes has been seen by
crossing from the thermally equilibrated S state to the hot
3
0,31
1
excitation to the S state at 266 or 299 nm in solution.
2
T1 state at which it dissociates in competition with vibra-
tional relaxation.
The amount of excess vibrational energy ͑4900 or 700 cmϪ1͒
has not affected dissociation. The mechanism of 1- and
2
-͑X-methyl͒naphthalenes may be similar to that of diphe-
nylmethane. The C–X bond fission channel for 1- and
-͑X-methyl͒naphthalenes is characterized by intersystem
1
N. Ikeda, N. Nakashima, and K. Yoshihara, J. Chem. Phys. 82, 5285
͑
1985͒.
2
2
3
K. Tsukiyama and R. Bersohn, J. Chem. Phys. 86, 745 ͑1987͒.
Y. Kajii, K. Obi, I. Tanaka, N. Ikeda, N. Nakashima, and K. Yoshihara, J.
Chem. Phys. 86, 6115 ͑1987͒.
crossing to an upper triplet state which is itself or crosses to
a ␴␴*(C–H) dissociative triplet state. The C–H bond fis-
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
52.7.17.127 On: Wed, 10 Dec 2014 09:51:55
1

J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Fujiwara et al.
1365
4
19
20
U. Brand, H. Hippler, L. Lindemann, and J. Troe, J. Phys. Chem. 94, 6305
M. J. Rossi, D. F. McMillen, and D. M. Golden, J. Phys. Chem. 88, 5031
͑1984͒.
F. G. Bordwell, J.-P. Cheng, and J. A. Harrelson, Jr., J. Am. Chem. Soc.
110, 1229 ͑1988͒.
S. E. Stein and R. L. Brown, J. Am. Chem. Soc. 113, 787 ͑1991͒.
F. G. Bordwell, X. Zhang, and J.-P. Cheng, J. Org. Chem. 56, 3216
͑1990͒.
5
K. W. Haider, E. Migirdicyan, M. S. Platz, N. Soundararajan, and A.
Despres, J. Am. Chem. Soc. 112, 733 ͑1990͒.
J. L. Faria and S. Steenken, J. Phys. Chem. 97, 1924 ͑1993͒.
M. Fujiwara and K. Toyomi, J. Chem. Phys. 107, 9354 ͑1997͒.
A. Bromberg, K. H. Schmidt, and D. Meisel, J. Am. Chem. Soc. 106,
6
7
8
21
22
͑
1991͒.
3
056 ͑1984͒.
23
24
25
V. D. Parker, J. Am. Chem. Soc. 114, 7458 ͑1992͒.
L. Salem, J. Am. Chem. Soc. 96, 3486 ͑1974͒.
N. J. Turro, W. E. Farneth, and A. Devaquet, J. Am. Chem. Soc. 98, 7425
9
0
1
D. Weir and J. C. Scaiano, Chem. Phys. Lett. 128, 156 ͑1986͒.
R. W. Redmond and J. C. Scaiano, Chem. Phys. Lett. 166, 20 ͑1990͒.
J. Bartl, S. Steenken, H. Mayr, and R. A. McClelland, J. Am. Chem. Soc.
1
1
͑
1976͒.
1
12, 6918 ͑1990͒.
2
2
2
2
6
7
8
9
1
1
2
3
S. S. Hunnicutt, L. D. Waits, and J. A. Guest, J. Phys. Chem. 95, 562
͑1991͒.
M. D. Person, P. W. Kash, S. A. Schofield, and L. J. Butler, J. Chem.
Phys. 95, 3843 ͑1991͒.
M. D. Person, P. W. Kash, and L. J. Butler, J. Chem. Phys. 97, 355
͑1992͒.
T. L. Myers, N. R. Forde, B. Hu, D. C. Kitchen, and L. J. Butler, J. Chem.
Phys. 107, 5361 ͑1997͒.
D. F. Kelley, S. V. Milton, D. Huppert, and P. M. Rentzepis, J. Phys.
Chem. 87, 1842 ͑1983͒.
E. F. Hilinski, D. Huppert, D. F. Kelley, S. V. Milton, and P. M. Rentz-
epis, J. Am. Chem. Soc. 106, 1951 ͑1984͒.
J. T. Banks and J. C. Scaiano, J. Am. Chem. Soc. 115, 6409 ͑1993͒.
N. Nakashima, M. Sumitani, I. Ohmine, and K. Yoshihara, J. Chem. Phys.
7
2, 2226 ͑1980͒.
1
1
4
5
I. W. C. E. Arends, P. Mulder, K. B. Clark, and D. D. M. Wayner, J. Phys.
Chem. 99, 8182 ͑1995͒.
I. Carmichael and G. L. Hug, in CRC Handbook of Organic Photochem-
istry, edited by J. C. Scaiano ͑CRC, Boca Raton, FL, 1989͒, Vol. 1, Chap.
1
6.
1
6
30
31
R. B. Cundall and L. C. Pereira, J. Chem. Soc., Faraday Trans. 2 68, 1152
1972͒.
͑
1
1
7
8
R. R. Hentz and R. M. Thibault, J. Phys. Chem. 77, 1105 ͑1973͒.
T. A. Gregory and W. P. Helman, J. Chem. Phys. 56, 377 ͑1972͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
52.7.17.127 On: Wed, 10 Dec 2014 09:51:55
1