Picosecond time-resolved fluorescence spectroscopy of (Z)-1-(2-anthryl)-2-phenylethene and its model compounds: Understanding the photochemistry by distinguishing between the s-cis and s-trans rotamers

Article abstract of DOI:10.1021/jp993305c

The photochemical reactions (Z-E isomerization and photocyclization) of the s-cis and s-trans rotamers of (Z)-1-(2-anthryl)-2-phenylethene (Z2APE) were investigated. The absorption, steady-state and picosecond time-resolved fluorescences, and transient absorption spectra were measured in order to achieve a better understanding of the rotamer photochemistry. The spectra were compared with those of the model compounds ((Z)-1-[2-(1-methylanthryl)]-2-phenylethene: Z1Me2APE and (Z)-1-[2-(3-methylanthryl)]-2-phenylethene: Z3Me2APE). It is confirmed that the s-trans rotamer undergoes only Z→E adiabatic isomerization and that the s-cis rotamer mainly photocyclizes to give the dihydrophenanthrene-type intermediate with a much faster rate constant (k_cyc = 2.3×10¹⁰ s^-1). The reason Z2APE does not give an aromatized photocyclization product (1,2-naphtho[a]anthracene, NA) is that the return of the dihydrophenanthrene-type intermediate to the Z isomer is much faster than the oxidation to produce NA.

Full text of DOI:10.1021/jp993305c

J. Phys. Chem. A 2000, 104, 6993-7001
6993
Picosecond Time-Resolved Fluorescence Spectroscopy of (Z)-1-(2-Anthryl)-2-phenylethene
and Its Model Compounds: Understanding the Photochemistry by Distinguishing between
the s-cis and s-trans Rotamers
Takashi Karatsu,*,† Hajime Itoh,† Atsuko Nishigaki,‡ Keijiro Fukui,† Akihide Kitamura,*,†
Shigeki Matsuo,§ and Hiroaki Misawa§
Department of Materials Technology, Faculty of Engineering, and Center for Frontier Electronics and
Photonics, Chiba UniVersity, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan, and Faculty of Engineering,
The UniVersity of Tokushima, Minamijosanjima-cho, Tokushima 770-8506, Japan
ReceiVed: September 15, 1999; In Final Form: May 24, 2000
The photochemical reactions (Z-E isomerization and photocyclization) of the s-cis and s-trans rotamers of
(Z)-1-(2-anthryl)-2-phenylethene (Z2APE) were investigated. The absorption, steady-state and picosecond
time-resolved fluorescences, and transient absorption spectra were measured in order to achieve a better
understanding of the rotamer photochemistry. The spectra were compared with those of the model compounds
((Z)-1-[2-(1-methylanthryl)]-2-phenylethene: Z1Me2APE and (Z)-1-[2-(3-methylanthryl)]-2-phenylethene:
Z3Me2APE). It is confirmed that the s-trans rotamer undergoes only ZfE adiabatic isomerization and that
the s-cis rotamer mainly photocyclizes to give the dihydrophenanthrene-type intermediate with a much faster
rate constant (kcyc ) 2.3 × 1010 s-1). The reason Z2APE does not give an aromatized photocyclization product
(1,2-naphtho[a]anthracene, NA) is that the return of the dihydrophenanthrene-type intermediate to the Z isomer
is much faster than the oxidation to produce NA.
Introduction
SCHEME 1
The compound 1-(2-anthryl)-2-phenylethene (2APE) has two
rotational isomers with respect to the single bond that connects
the anthryl group to the olefinic double bond. The s-trans- and
s-cis-rotamers align their double bonds along the long and short
axes of the anthryl moiety, respectively. The rotational isomer-
ization of 2-anthrylethenes in the ground state and in the excited
states is a subject of current interest.1-24 Most of the studies
have detailed the E isomer, because of its ease of preparation
and its inactivity in the geometrical isomerization known as one-
way ZfE isomerization.23-29 This isomerization has been
studied using steady-state fluorescence,1,2 time-correlated single-
photon counting (SPC),3-9 time-gated fluorescence spectros-
copy,10-13 principal component analysis with self-modeling
(PCA-SM),3,4,14-17 and T-T absorption spectroscopy.18-21
When compared with the extensive studies on the rotamers of
the E isomers, only a few studies have been reported for the Z
isomers.22-24 In particular, the selective photochemistry of the
two rotamers of (Z)-2APE (Z2APE) was elucidated by Saltiel’s
group using the PCA-SM technique.23,24 They showed that in
toluene (1) the NEER (nonequilibration of excited rotamers)
principle1,30,31 holds for the singlet excited Z2APE, (2) that the
1(Z s-trans)* rotamer undergoes efficient adiabatic Z fE one-
1
way isomerization, but that the (Z s-cis)* rotamer does not
adiabatically isomerize to the E isomer (Scheme 1).
In our previous reports,18,19 we characterized the s-trans and
s-cis rotamers of E2APE by using their model compounds, in
which methyl groups restrict their conformations as previously
reported for the analogous naphthalene derivatives.15,32,33 We
attempted to characterize the s-trans and s-cis rotamers of
Z2APE along the same lines by comparing their properties with
those of the model compounds: (Z)-1-[2-(1-methylanthryl)]-2-
phenylethene (Z1Me2APE) and (Z)-1-[2-(3-methylanthryl)]-2-
phenylethene (Z3Me2APE). The study of the Z2APE rotamer
and its analogues is particularly difficult, because one-way ZfE
* To whom correspondence should be addressed, e-mail: karatsu@
xtal.tf.chiba-u.ac.jp, FAX: +81-43-290-3039.
† Faculty of Engineering, Chiba University.
‡ Center for Frontier Electronics and Photonics, Chiba University.
§ The University of Tokushima.
10.1021/jp993305c CCC: $19.00 © 2000 American Chemical Society
Published on Web 07/12/2000

6994 J. Phys. Chem. A, Vol. 104, No. 30, 2000
Karatsu et al.
geometrical isomerization efficiently takes place in S1, and the
fluorescence quantum yield of the E isomer is higher than that
of the Z isomer.23,24 This difficulty was overcome in Saltiel’s
study in which the characterization of the rotamers was based
on steady-state fluorescence and photoisomerization data. We
have now succeeded in understanding the behavior of the ro-
tamers of Z2APE in the excited state using time-resolved spec-
troscopy. Our results on the reaction kinetics obtained by time-
resolved spectroscopy are consistent with the conclusions of
Saltiel et al.23,24 complementing the results from the s-trans
rotamer and providing new insight into the behavior of the s-cis
rotamer. In addition, 2APE does not yield the photocyclization
product whereas the p-methyl derivative, (E)-1-(2-anthryl)-2-
(p-methylphenyl)ethene (EpMe2APE), yields the photocycliza-
tion product.34 It has been argued that the isomerization cannot
be exclusively ZfE one-way.24,34 However, we found that
Z2APE gave a photocyclization product in the presence of
oxidizing agents and the ZfE isomerization should still seem
one-way.
laser for excitation is about 40 times greater than that for oxygen.
Therefore, air oxidation cannot compete with cycloreversion but
iodine oxidation can, and the yield of NA increases in the
presence of iodine (Scheme 3).
SCHEME 3
Conformer-specific photocyclization of arylethenes has been
reported by many groups. Therefore, we have also investigated
the reaction from this point of view.34-36 We find that the model
compound Z3Me2APE yields 7-methyl-1,2-naphtho[a]an-
thracene (7MeNA) as well as the E isomer (E3Me2APE).
However, as expected based on the free valence theory (Scheme
2),35,36 Z1Me2APE gives only the E isomer (E1Me2APE) and
does not produce 7-methylbenz[a]tetracene (MeBT).
SCHEME 2
In addition to the fluorescence, the nanosecond time scale
T-T absorption was examined for Z2APE to determine the
behavior in the triplet state. At low temperature, the T-T
absorption of the E s-trans isomer is observed just after laser
pulse excitation, then the E s-transfE s-cis adiabatic rotational
isomerization in T1 is observed as previously reported for the
E isomer.19 The spectral change indicates that the activation
energy of the one-way ZfE isomerization in T1 is smaller than
that in S1 assuming that the preexponential factors of T1 and S1
are the same. The rate of intersystem crossing becomes higher
than that of the ZfE isomerization in S1 at the applied
temperature (213 K).
Our results are complementary to the PCA-SM results23,24
and supply information on the dynamics of the photochemistry
of 2APE, including its photocyclization and one-way isomer-
ization.
Experimental Section
Olefins and Other Chemicals. 2APE, 1Me2APE, and
3Me2APE were prepared as reported previously.19,37 Spectro-
grade benzene and methylcyclohexane (Dotite) were used
without further purification. Guaranteed grade 2-methyltetrahy-
drofuran (Wako Chemicals) was purified by distillation over
lithium aluminum hydride.
Z2APE yields a negligible amount of NA in the absence of
iodine in air by the irradiation of 366 nm Hg arc light; however,
we have determined that Z2APE gave a significant amount of
NA in the presence of iodine in air, and this was due to the
strong oxidizing ability of iodine. The absolute rate of oxidation
by iodine determined in this study using a cryostat and pulsed
Photochemistry and Product Analysis. Stationary irradia-
tion was performed with a 400-W high-pressure mercury lamp

Photochemical Reactions of Z2APE
J. Phys. Chem. A, Vol. 104, No. 30, 2000 6995
using a UV-D36B glass filter (366 nm light, for direct exci-
tation). Samples were prepared in benzene or methylcyclohexane
and purged with argon or nitrogen, or degassed by three freeze-
pump-thaw cycles. The Z and E isomers and cyclized products
were analyzed by GC (Shimadzu GC-14A) with an OV-1
column (25 m) and HPLC systems (JASCO 880-PU) using a
Senshu Pak Silica-1251-R (25 cm) column with a hexane-ethyl
acetate (99:1) mixture. Absorption spectra were also measured
using an HPLC-UV detector (Shimadzu SPD-M6A).
Photochemical products were identified by their mass spectra
measured with a Shimadzu QP-2000 GC-MS spectrometer and
1
by H NMR measured with a JEOL GSX-500. Quantitative
analyses were performed by GC as mentioned above using
chrysene as the internal standard.
Following irradiation at 355 nm from a Continuum Surelite
I-10, the reaction rates of DHP for thermal reversion and reaction
with oxygen or iodine were measured in methylcyclohexane
with the use of a Hitachi U-3000 spectrophotometer using an
Oxford DN1704 cryostat.
Figure 1. Absorption spectra of Z2APE, Z1Me2APE, and Z3Me2APE
in benzene at ambient temperature.
Spectroscopy. Absorption and emission spectra were mea-
sured with the use of a Hitachi U-3000 and of an F-4010
spectrometer, respectively. Fluorescence lifetimes in the nano-
second time range were measured with a Horiba NAES-550
single-photon counting spectrometer. The incident light was set
at 360 nm and emission was collected through an L-40 or L-42
glass filter.
Fluorescence spectra and lifetimes in the picosecond time
range were measured as described below. A Ti:sapphire laser
(Spectra-Physics Tsunami, 1 mW, 150 fs fwhm, 4 MHz
repetition) and a pulse selector (S-P model 3980) pumped by
an Ar+ laser (S-P BeamLok model 2580) were used to obtain
380 nm incident light. Emission was focused by a lens (f30)
on a monochromator (Oriel multispec 257, grating 150 g/mm,
2.6 nm resolution at a 100 µm slit width) through an optical
fiber (Oriel 77529). The detectors were a streak scope (Hamamat-
su Photonics C4334-01) or a single photon counting detector
(Hamamatsu Photonics MCP-PMT C2773) with a PC-module
for time-correlated single photon counting (Becker & Hickl
GmgH). Trigger signals were taken by a pin-photodiode through
a delay unit (Hamamatsu Photonics C1097). Low-temperature
experiments were performed with the cryostat.
T-T absorption spectra were measured with excitation at 355
nm (Continuum SL I-10, 6 ns fwhm, 20 mJ per pulse at 5 Hz
repetition) using a detection system (Tokyo Instruments)
composed of a multichannel diode array (Princeton IRY-
512G: 18 ns gate width) with a SPEX 270M monochromator
(resolution: 0.3 nm/channel). Decay profiles measured with a
photomultiplier (Hamamatsu Photonics SR928) were analyzed
by the Marquardt nonlinear least-squares fitting method. Low-
temperature experiments were performed with the cryostat.
Figure 2. Time resolved fluorescence spectra of (a) E2APE and (b)
E1Me2APE (solid line) and E3Me2APE (dashed line) in nitrogen
purged methylcyclohexane at ambient temperature. The spectra were
obtained by excitation at 380 nm and averaged between 0 and 100 ps
after the laser pulses.
just after the laser pulse (average of CCD camera image through
0-100 ps). The spectra are essentially time-independent, at least
up to 400 ps. The spectra of E2APE are known to be time-
dependent, even obeying the NEER principle due to the different
lifetimes of each rotamer;5,6 however, no spectral change was
observed in this short time scale. For E1Me2APE and
E3Me2APE, the spectra are time-independent up to 400 ps,
consistent with a single rotamer for each and are identical to
the steady-state fluorescence spectra,2 because the E isomer does
not isomerize to the Z isomer and methyl substitution prevents
rotational isomerization.
Results and Discussion
Absorption Spectra. Figure 1 shows the absorption spectra
of Z2APE, and its model compounds Z1Me2APE and Z3Me2-
APE in benzene for comparison. All compounds have broad
absorption bands at 300 nm with bands with vibrational structure
between 340 and 400 nm. The vibrational fine structure of
Z2APE is very similar to that of Z1Me2APE but not at all like
that of Z3Me2APE.
Picosecond Time-Resolved Fluorescence Spectra. The time-
resolved fluorescence spectra were measured for the Z and E
isomers of 2APE, 1Me2APE, and 3Me2APE in nitrogen-purged
methylcyclohexane at various temperatures between 200 and
295 K. Figure 2 shows the fluorescence spectra of the E isomers
Figure 3a-c shows the fluorescence spectra with their typical
time profile of the Z isomers of 2APE, 1Me2APE, and
3Me2APE, respectively, at 200 K. The spectra are time-

6996 J. Phys. Chem. A, Vol. 104, No. 30, 2000
Karatsu et al.
E s-trans excited state;23,24 however, such adiabatic ZfE
isomerization through S1 is prohibited at 200 K. For the spectra
measured above 230 K, even at 0-100 ps, the contribution of
the E isomer is significant, and a gradual increase in intensity
of the E fluorescence was observed in the time scale of 100 ps.
The observation of the E fluorescence just after laser excitation
is unreasonable considering the lifetime of the Z isomer.
Therefore, the emission comes from the re-excitation of the E
isomer generated by adiabatic ZfE isomerization through S1
(at higher temperature) or T1 after the intersystem crossing (the
rate of isomerization in T1 is still high at 200 K, and it is very
difficult to prevent this factor since we could not use a flow
cell in a cryostat).
The time-resolved spectra and decay profile of Z3Me2APE
at 200 K are shown in Figure 3b. The spectrum at 0-100 ps is
assigned to the Z isomer, but the spectrum is mixed with the E
isomer at 300-400 ps. This was also seen in the decay profile
(Figure 3b inset). The decay consists of two components. The
short-lived (τs ) 200 ps) major component is the fluorescence
of the Z isomer, and the minor other is the contribution of the
E isomer as an impurity (purity of Z isomer is above 98%;
however, the emission intensity of the Z isomer is very weak,
therefore, such an influence becomes important). The reason
fluorescence from the Z isomer of Z3Me2APE has such a shorter
lifetime than that of Z1Me2APE is that the photocyclization
reaction takes place at a much faster rate than the adiabatic ZfE
isomerization in S1. Therefore, the decay rate mainly corre-
sponds to the rate of the cyclization. This was also confirmed
by the efficient formation of 7MeNA.
Figure 3c shows the behavior of Z2APE. Here, Z2APE has
two rotamers, s-cis and s-trans, and the behavior of Z2APE is
very similar to the mixed behavior of Z1Me2APE (model of
the Z s-trans isomer) and Z3Me2APE (model of the Z s-cis
isomer). At 200 K, the decay profile (Figure 3c inset) has two
components with an equal contribution of the rotamers which
have 150 ( 22 ps and 2 ( 0.8 ns lifetimes and almost a
negligible contribution of the E isomer as seen in the time-
resolved spectra (Figure 3c). This means that the decay profile
and the spectra consist of the fluorescence of the short-lived Z
s-cis rotamer and the long-lived Z s-trans rotamer.
At temperatures higher than 230 K, the spectra and the decay
profiles are very similar to those of the Z1Me2APE as described
in the section on Z1Me2APE.
Figure 3. Time resolved fluorescence spectra (a, b, and c) and their
time profiles (inset) of Z1Me2APE, Z3Me2APE, and Z2APE, respec-
tively, at 200 K in nitrogen-purged methylcyclohexane. The spectra of
solid and dashed lines were averaged on 0-100 ps, and 300-400 ps,
respectively, after the laser pulses. The time profiles were averaged
between 400 and 430 nm.
The spectra expected by the PCA-SM method for the Z
isomer23,24 are assigned to the Z s-trans isomer in toluene. In
this report, the emission maximum of the spectrum is shifted
by about 10 nm to a longer wavelength than that reported for
the Z s-trans isomer by the PCA-SM method.24 This is because
our spectrum is assigned to the Z s-cis rotamer, and this 10 nm
shift is for the longer wavelength of the s-cis rotamer compared
to the s-trans rotamer. This is also seen in the rotamers of the
E isomer.
Fluorescence from the upper excited state (S2fS0) is reported
for E2APE by PCA-SM method,16 and this emission was also
observed in methylcyclohexane. However, no further investiga-
tion has been done concerning the contribution of this excited
state to the ZfE isomerization in this study.
The reaction mechanism is presented by the following
equations. After excitation, each Z rotamer (equation numbers
a and b and the superscripts zsc and zst of the rate constant
correspond to the s-cis and s-trans rotamers, respectively) does
the radiative (f; fluorescence) and nonradiative decays (isc,
intersystem crossing; nd, nonradiative decay indicated by
subscripts).
dependent up to 400 ps. Figure 3a shows the fluorescence
spectra of Z1Me2APE; the spectra are assigned to the fluores-
cence of the Z isomer. At this temperature, the solution (in
methylcyclohexane) still has fluidity;38 however, we could not
estimate the effective viscosity. The intensity between 0 and
100 ps is smaller than that between 300 and 400 ps, because
the former includes the rise in the fluorescence in the excitation
laser-pulse width. The spectrum at 300-400 ps contains a small
amount of the fluorescence of the E isomer seen as a shoulder
around 420 nm. The spectrum at 0-100 ps is broad and
structureless as previously reported for Z2APE using the PCA-
SM method.23,24 Therefore, the decay profile (Figure 3a inset)
is mainly due to the decay of the Z isomer and the lifetime is
about 2 ns at 200 K. This value is slightly smaller than that in
toluene,23 but similar to that reported in the methylcyclohexane-
3-methylpentane mixed solvent.29 In Saltiel’s reports, the Z
s-trans excited state showed a quadratic concentration depen-
dence of the fluorescence quenching of the adiabatically formed

Photochemical Reactions of Z2APE
J. Phys. Chem. A, Vol. 104, No. 30, 2000 6997
108 at 292 K is obtained, and also 3.1 × 106 s-1 at 200 K is
expected from their data. These values agree well with the results
for Z1Me2APE (model of s-trans). However, in the cases of
Z3Me2APE and Z2APE, the observed fluorescence has a much
shorter decay component in the low-temperature range (around
200 K), and we believe that this is due to the contribution of
fast photocyclization (kcyczsc) to the slow ZfE isomerization.
To determine the Arrhenius parameters of photocyclization, the
temperature dependence of the observed faster decay component
was examined. However, this is extremely difficult because the
increasing temperature causes increasing fluorescence of the
trans isomer which disturbs the observation of the short lifetime
component. Therefore, the parameters were roughly estimated
for Z3Me2APE and Z2APE. For Z3Me2APE, the values of A
) 6.8 × 1010 s-1 and Ea ) 8 kJ mol-1 were roughly estimated
using values of 1310 ( 71 and 886 ( 63 ps at 210 and 230 K,
hV
Z s-cis 98 1(Z s-cis)*
(1a)
(1b)
hV
Z s-trans 98 1(Z s-trans)*
kfzsc
1(Z s-cis)*
8 Z s-cis
(2a)
(2b)
(3a)
(3b)
(4a)
(4b)
kfzst
1(Z s-trans)* 8 Z s-trans
zsc
kisc
1(Z s-cis)*
1(Z s-trans)*
1(Z s-cis)*
8 3(Z s-cis)*
zst
kisc
8 3(Z s-trans)*
zsc
knd
8 Z s-cis
zst
knd
zsc
respectively. These correspond to kcyc ) 2.5 × 109 and 5.7
1(Z s-trans)*
8 Z s-trans
× 108 s-1 at 290 and 200 K, respectively. For Z2APE, the
observed short lifetime component is assigned to the Z s-cis,
and the values of A ) 4.9 × 1012 s-1 and Ea ) 13 kJ mol-1
were also estimated roughly using the values of 830 ( 57 and
376 ( 53 ps at 190 and 210 K, respectively. The A factor for
Z3Me2APE is slightly smaller than the typical values of the
isomerization. This may have large experimental uncertainity
because of experimental difficulty described above. These
Saltiel et al. have shown that only the singlet excited state of
the Z s-trans rotamer undergoes efficient ZfE adiabatic
photoisomerization.23,24 The absence of this process in the Z
s-cis rotamer is probably due to its efficient cyclization to give
DHNA (eq 6). Photocyclization of the s-trans isomer (eq 7) may
be neglected as expected on the basis of MO calculations.35,36
Szst
k ZfE
zsc
activation parameters correspond to kcyc ) 2.3 × 1010 and
1(Z s-trans)*
8 1(E s-trans)*
(5)
(6)
(7)
2.0 × 109 s-1 at 290 and 200 K, respectively.
zsc
kcyc
1(Z s-cis)*
8 DHNA
From the above kinetic treatment, it was found that the Z
s-cis isomer mainly photocyclizes at 200 K and that the ZfE
adiabatic isomerization competes with the cyclization at ambient
temperature. However, for the Z s-trans isomer, the rate constant
reflects the ZfE adiabatic isomerization because negligible
cyclization occurred.
zst
kcyc
1(Z s-trans)*
8 DHBT
For Z3Me2APE and Z1Me2APE, each olefin takes the s-cis
and s-trans conformation, respectively, due to steric hindrance
of the methyl groups; therefore, eq 6 is important only for
Z3Me2APE.
The E isomers also undergo radiative and nonradiative decays
(eqs 8-10). In these equations, the superscripts esc and est
indicate the E s-cis and s-trans rotamers, respectively.
The photochemical paths of the s-trans isomer have been
studied in detail by Saltiel et al.23,24 Here, we roughly apply
the kinetic data obtained above to Saltiel’s precise kinetic
analysis of the photochemical paths of Z2APE in toluene24 that
the cyclization is the main path for the s-cis isomer. The rate
constant obtained for the cyclization is almost 2 orders of
magnitude greater than the rate constants of the other paths such
as fluorescence, intersystem crossing, and even adiabatic ZfE
isomerization. Quantitative comparison of the rate constants
from this study with those reported earlier23,24 is hampered by
the change in solvent; however, let us consider the rough
estimation of the reaction paths. It makes clear the outline of
the reaction mechanisms. In the case of 2APE, the quantum
yields of fluorescence (Φf) of the Z and E isomers obtained in
the steady state are reported to be 0.59 and 0.87, respectively,39
in benzene at ambient temperature by us and 0.52 and 0.89,
respectively,14,16,23 in toluene by Saltiel et al. The difference in
Φf reported for the Z isomer was pointed out by the generation
of the E isomer during the measurement since no flow cell was
used in our earlier experiment.24 It is reported by the PCA
method24 that the Φf is composed of 25% (Φf ) 0.13)
fluorescence of the Z isomer and 75% (Φf ) 0.39) fluorescence
of the E isomer generated by the adiabatic ZfE isomerization.
Therefore, at least 44% [(0.39/0.88) × 100] of the Z isomer
adiabatically isomerizes to the E isomer in an adiabatic way.
The intersystem crossing quantum yields (Φisc) of the Z and E
isomers are reported to be 0.17 and 0.11, respectively.39 Here,
again, if one assumes that the observed triplet state was due to
the isomerized E isomer, about 5% (0.11 × 0.44 × 100) of the
triplets was generated through the E s-trans singlet excited state
after the singlet adiabatic ZfE isomerization. Therefore, almost
kfesc
1(E s-cis)*
8 E s-cis
(8a)
(8b)
kfest
1(E s-trans)* 8 E s-trans
esc
knd
1(E s-cis)*
8 E s-cis
(9a)
est
knd
1(E s-trans)*
8 E s-trans
(9b)
esc
kisc
1(E s-cis)*
8 3(E s-cis)*
(10a)
(10b)
est
kisc
1(E s-trans)*
8 3(E s-trans)*
The lifetimes of the singlet excited state Z isomers are given by
τSzsc ) 1/(kfzsc + kisczsc + kndzsc + k ZfESzsc + kcyczsc) (11)
τSzst ) 1/(kfzst + kisczst + kndzst + k ZfE
)
(12)
Szst
based on the results of obeying the NEER principle of the Z
isomers as reported and no cyclization of the Z s-trans isomers.
Szsc
In Saltiel’s report, kZfE
is negligible and the temperature
effect on kZfESzst was extensively investigated in toluene.23 From
Szst
the results of the s-trans rotamer of Z2APE, kZfE ) 1.7 ×

6998 J. Phys. Chem. A, Vol. 104, No. 30, 2000
Karatsu et al.
12% of the (0.17-0.05) triplets is generated through the
intersystem crossing of the Z isomer, then triplet adiabatic ZfE
isomerization occurred. In addition, the ratio production of the
E s-trans rotamer through the singlet adiabatic ZfE isomer-
ization and triplet adiabatic ZfE isomerization after intersystem
S
crossing changes by the temperature, since kZfE is dependent
on temperature. In this assumption, we did not consider the
change of reaction paths by changing energy levels of S2 and
S1 state of the E s-trans isomer by temperature.
S
S
S
Values of 67% [kZfE /(kZfE + kisc + kf)], 9% [kisc/(kZfE
+
S
kisc + kf)] and 24% [kf/(kZfE + kisc + kf)] were then expected
S
at ambient temperature using the rate constants kZfE ) 1.7 ×
108 s-1, kisc ) 2.4 × 107 s-1, and kf ) 6.0 × 107 s-1 reported
for the s-trans Z2APE in toluene.23 These fit reasonably well
to the values, 64% [(44/69) × 100], 17% [(12/69) × 100], and
19% [(13/69) × 100] obtained above for the ZfE isomerization,
intersystem crossing, and fluorescence, respectively, for the Z
s-trans isomer.
Finally, the missing 31% [100 - (44 + 12 + 13)] ) [1 -
Zst
Zst
Zst
(ΦZfE + Φisc + Φf )], which we initially thought is a de-
activation process at the twisted geometry, must be the efficiency
of the photocyclization reaction as pointed out by Saltiel.23
Considering the quantum efficiencies and kinetic values at 290
K mentioned above, kcyc is exclusively faster than the other
S
processes such as kZfE , kisc, and kf. Therefore, it is reasonably
explained that the s-cis isomer solely undergoes photocyclization
and that the s-trans isomer undergoes ZfE isomerization,
intersystem crossing, and fluorescence. In addition, the large
kcyc value hampers estimating the other rate constants such as
kisc and kf. The values of 31% and 69% then correspond to the
population of the s-cis and s-trans isomers. These values show
a slightly more equal population of each isomer than the reported
value of around 80% of the s-trans isomer at ambient temper-
ature (∆H ) 3.9 kJ mol-1);23 however, our rough estimation
based on the experimental value is acceptable for understanding
entire photochemical paths at this moment, but a more precise
study of solvent effects should be done.
Cyclization Reaction and Behavior of the Dehydro-
phenanthrene-Type Intermediates. Figure 4 shows the change
in the absorption spectra of Z3Me2APE, Z2APE, ZpMe2APE,
and ZmMe2APE, before and after irradiation in argon-purged
methylcyclohexane. For Z3Me2APE, the intermediate is stable
enough to determine the reaction rate using a UV-visible
spectrophotometer at temperatures between 200 K and room
temperature with a cryostat. The sample solution was irradiated
by 355 nm laser pulses (totally 50-100 pulses by 10 Hz), then
the change followed by a spectrophotometer. The windows of
the cryostat were covered by aluminum foil between each
measurement. The plot of absorbance versus time was pseudo-
first order. For Z2APE and ZpMe2APE, the spectral changes
were very fast at room temperature; therefore, we have to
observe these changes at temperatures below 250 K. Especially
for ZmMe2APE, the spectral change due to the formation of
an intermediate was very small even at low temperature. The
decrease in the absorbance at the maximum with time is due to
the thermal ring opening reaction of the dihydrophenanthrene-
type intermediates to the Z isomers (eq 13)
Figure 4. Absorption spectral changes of Z3Me2APE (a), Z2APE (b),
ZpMe2APE (c), and ZmMe2APE (d) by the 355 nm laser irradiation
at 290, 200, 250, and 250 K, respectively, in methylcyclohexane.
are found to be A ) 3.2 × 102, 1.6 × 102, and 5.6 × 106 s-1
,
and Ea ) 4.3, 8.1 and 14.9 kJ mol-1, for Z2APE, Z3Me2APE,
and Z1APE, respectively. These A factors are very small for
the unimolecular process; however, similar values are derived
from the Ea values and the rate constant at the temperature
reported.36 The reason is not clear at this moment, and more
work is apparently needed.
kop
DHNA
98 Z s-cis
(13)
Figure 5 shows an Arrhenius plot of kopn for the intermediates
generated from Z3Me2APE, Z2APE, and (Z)-1-(1-anthryl)-2-
phenylethene (Z1APE), which we reported,40 for comparison.
The preexponential factors (A) and the activation energies (Ea)
DHNA (14b,14c-dihydro-1,2-naphth[a]anthracene) yields an
aromatized product, NA (1,2-naphth[a]anthracene) by oxidation

Photochemical Reactions of Z2APE
J. Phys. Chem. A, Vol. 104, No. 30, 2000 6999
Figure 5. Arrhenius plots of the ring opening rate constant (kop) for
DHP type intermediate of Z3Me2APE and Z2APE in methylcyclo-
hexane determined in this study and Z1APE for comparison.
Figure 7. The absorption spectra of BT and NA in ethanol, reported
in ref 41 and 42, respectively, and the absorption spectrum of product
in benzene measured in this study.
of iodine and light intensity. The rate was 5.0 × 10-4 s-1 under
room light when [I2] ) 1.9 × 10-4 mol dm-3 under these
experimental conditions. This rate is about 6 times faster than
the cycloreversion (7.8 × 10-5 s-1). The rate is expected to be
greater under irradiation by a mercury lamp (366 nm) using a
merry-go-round type photoreactor from which we obtained the
photochemical product since the iodine atom is continuously
generated by irradiation. We are doing more experiments to
elucidate the oxidation mechanism by iodine.
Table 1 shows the products and yields by the irradiation of
ZAPEs in air or in the presence of iodine in air. NA and 7MeNA
were isolated and identified by the usual spectroscopic tech-
niques, especially UV-vis absorption spectroscopy (Figure 7).
Their spectra were compared with the spectra reported for NA42
and benz[a]tetracene (BT).43 In case of ZpMe2APE and
ZmMe2APE, the products were identified only by GC-MS
because of their low yields, and the yields were estimated based
on the products (1MeNA, 2MeNA, and 3MeNA) having the
same sensitivity to GC that 7MeNA has. The reason the yields
for MeNA are different due to the methyl substitution at various
positions is now under investigation using MO calculations. It
is interesting that, in case of Z1Me2APE, NA having no methyl
group was observed by GC-MS. The loss of the methyl or
other groups by cyclization is sometimes observed as described
in the literature.40 In all cases, some oxygen-containing products
were observed by GC-MS (the structures are expected to be
like anthraquinone).
Figure 6. Oxidation rate constant for dihydrophenanthrene type
intermediate of Z3Me2APE by oxygen (koxo) in methylcyclohexane at
ambient temperature.
in the presence of an oxidant such as molecular oxygen (eq 14)
or iodine (eq 15). The spectral change was accelerated in the
presence of the oxidants.
DHNA koxo[O2]8 NA
DHNA koxi[I2]8 NA
(14)
(15)
For Z3MeAPE, the oxidation rate constants of the DHP-type
intermediates were determined by a Stern-Volmer-type plot
of the rate versus the concentration of oxygen in solution (Figure
6). Only the intermediate generated from Z3MeAPE was
relatively long-lived enough to determine the rate constants at
ambient temperature. The sample solution containing oxygen
or iodine was irradiated by 355 nm laser pulses (totally 50-
100 pulses by 10 Hz), then the spectrum followed by a
spectrophotometer at temperatures between 200 K and room
temperature using a cryostat. The sample solution was irradiated
by 355 nm laser pulses, and the change was followed by a
spectrophotometer. The windows of the cryostat were covered
by aluminum foil between each measurement. The plot of
absorbance versus time was pseudo-first order. The absolute
Conformer-specific photocyclizations of arylethenes have
been reported by several groups.35,36,40 Especially, concerning
ZpMe2APE, Laarhoven et al. reported the formation of 2MeNA
in the presence of iodine.34 This is in contrast to our initial report
that no formation of NA occurred upon the irradiation of
Z2APE.25-28 This difference is discussed in relation to the one-
way ZfE isomerization, and the formation of 2MeNA is due
to the accumulation of the product during the reversible Z-E
isomerization.24 However, this study clarified why Z2APE did
not give NA under the experimental conditions examined.
T-T Absorption Spectra. The adiabatic ZfE geometrical
isomerization in the triplet excited state was examined by T-T
absorption measurements. For the E isomers, our previous report
showed that adiabatic s-trans to s-cis rotational isomerization
occurs in the triplet excited state. For the Z isomer, the adiabatic
ZfE isomerization in the singlet excited state is the main path;
geometrical isomerization after intersystem crossing (eqs 16a
and 16b) is not the dominant path, and only T-T absorption
due to the E s-cis rotamer was observed at 290 K. However,
rate constants, koxo, was determined to be 0.050 mol dm-3 s-1
.
However, oxidation of dihydrophenanthrene by iodine is
proposed to occur through a radical abstraction mechanism
initiated by dissociation of iodine by light.41 The rate of
oxidation of DHNA is expected to depend on the concentration

7000 J. Phys. Chem. A, Vol. 104, No. 30, 2000
Karatsu et al.
TABLE 1: Photocyclization Yields of (Z) 2-Anthrylethenes in the Presence and Absence of Iodinea
a Irradiated at 366 nm in methylcyclohexane (5.0 × 10-3 mol dm-3) under air ([O2] ) 2.6 × 10-3 mol dm-3).
Conclusions
In conclusion, we have determined the photochemistry of
Z2APE by picosecond time-resolved fluorescence spectra by
comparison of the spectra with those of two model compounds.
The s-trans isomer is more stable in the ground state. Based on
the fluorescence spectra, kinetic values, and analysis of the
photoproducts, the s-cis rotamer solely photocyclizes in the
singlet excited state, and the s-trans rotamer undergoes ZfE
adiabatic isomerization. The cyclization process has a much
smaller activation energy than that of the ZfE adiabatic
isomerization. This conclusion essentially coincides with the
earlier work done by Saltiel’s group using principal component
analysis. We have concluded that the efficient cyclization
reaction prevents ZfE isomerization from the s-cis rotamer.
References and Notes
(1) Mazzucato, U.; Momiccioli, F. Chem. ReV. 1991, 91, 1679-1719,
and references therein.
Figure 8. T-T absorption spectra of Z2APE in deaerated methylcy-
clohexane at 300 ns, 5 µs, and 20 µs after laser excitation at 213 K.
(2) Fischer, G.; Fischer, E. J. Phys. Chem. 1981, 85, 2611-2613.
(3) Spalletti, A.; Bartocci, G.; Masetti, F.; Mazzucato, U.; Cruciani,
G. Chem. Phys. 1992, 160, 131-144.
S
kZfE was suppressed and kisc became the main path at low
temperature. To estimate the population of each rotamer, the
T-T absorption spectra were measured at low temperature (213
K) as shown in Figure 8. At 300 ns after the laser pulse, the
spectrum is mainly due to the E s-trans rotamer, and a spectral
change due to the E s-transfs-cis adiabatic rotational isomer-
ization was observed as seen in the rotational isomerization of
the reported E isomer.18-20 However, a spectral change due to
the ZfE adiabatic isomerization as expressed in eqs 16a and
16b could not be observed because the isomerization is faster
than the instrument resolution even at 213 K.
(4) Bartocci, G.; Mazzucato, U.; Spalletti, A. Chem. Phys. 1996, 202,
367-376.
(5) Ghiggino, K. P.; Skilton, P. F.; Fischer, E. J. Am. Chem. Soc. 1986,
108, 1146-1149.
(6) Wismontski-Knittel, T.; Das, P. K. J. Phys. Chem. 1984, 88, 1163-
1168.
(7) Wismontski-Knittel, T.; Das, P. K. J. Phys. Chem. 1984, 88, 1168-
1173.
(8) Arai, T.; Karatsu, T.; Sakuragi, H.; Tokumaru, K.; Tamai, N.;
Yamazaki, I. Chem. Phys. Lett. 1989, 158, 429-434.
(9) Arai, T.; Karatsu, T.; Sakuragi, H.; Tokumaru, K.; Tamai, N.;
Yamazaki, I. J. Photochem. Photobiol. A: Chem. 1992, 65, 41-51.
(10) Kang, T. J.; Etheridge, T.; Jarzeba, W.; Barbara, P. F. J. Phys.
Chem. 1989, 93, 1876-1881.
(11) Brearley, A. M.; Strandjord, A. J. G.; Flom, S. R.; Barbara, P. F.
Chem. Phys. Lett. 1985, 113, 43-48.
(12) Flom, S. R.; Nagarajan, V.; Barbara, P. F. J. Phys. Chem. 1986,
90, 2085-2092.
Tzsc
k ZfE
3(Z s-cis)*
8 3(E s-cis)*
(16a)
(16b)
Tzst
k ZfE
3(Z s-trans)*
8 3(E s-trans)*

Photochemical Reactions of Z2APE
J. Phys. Chem. A, Vol. 104, No. 30, 2000 7001
(13) Brearley, A. M.; Flom, S. R.; Nagarajan, V.; Barbara, P. F. J. Phys.
Chem. 1986, 90, 2092-2099.
(14) Saltiel, J.; Zhang, Y.; Sears, D. F., Jr.; Choi, J.-O. Res. Chem.
Intermed. 1995, 21, 899-921, and references therein.
(15) Saltiel, J.; Eaker, D. W. J. Am. Chem. Soc. 1984, 106, 7624-7626.
(16) Saltiel, J.; Zhang Y.; Sears, D. F., Jr. J. Phys. Chem. 1997, 101,
7053-7060.
(17) Saltiel, J.; Choi, J.-O.; Sears, D. F., Jr.; Eaker, D. W.; O’Shea, K.
E.; Garcia, I. J. Am. Chem. Soc. 1996, 118, 7478-7485.
(18) Karatsu, T.; Yoshikawa, N.; Kitamura, A.; Tokumaru, K. Chem.
Lett. 1994, 381-384.
(28) Arai, T.; Tokumaru, K. Bull. Chem. Soc. Jpn. 1995, 68, 1065-
1087.
(29) Mazzucato, U.; Spalletti, A.; Bartocci, G. Coord. Chem. ReV. 1993,
125, 251-260, and references therein.
(30) Jacobs, H. J.; Havinga, E. AdV. Photochem. 1979, 11, 305-373,
and references therein.
(31) Birks, J. B.; Bartocci, G.; Aloisi, G. G.; Dellonte, S.; Barigelletti,
F. Chem. Phys. 1980, 51, 113-120.
(32) Sun, Y.-P.; Sears, D. F., Jr.; Saltiel, J.; Mallory, F. B.; Mallory, C.
W.; Buser, C. J. Am. Chem. Soc. 1988, 110, 6974-6984.
(33) Saltiel, J.; Choi, J.-O.; Sears, D. F., Jr.; Eaker, D. W.; Mallory, F.
B.; Mallory, C. W. J. Phys. Chem. 1994, 98, 13162-13170.
(34) Laarhoven, W. H.; Cuppen, Th. J. H. M.; Nivard, R. J. F.
Tetrahedron 1970, 4865-4881.
(19) Karatsu, T.; Itoh, H.; Yoshikawa, N.; Kitamura, A.; Tokumaru, K.
Bull. Chem. Soc. Jpn. 1999, 72, 1837-1849.
(20) Krongauz, V.; Castel, N.; Fischer, E. J. Photochem. 1987, 39, 285-
300.
(35) Laarhoven, W. H. in Organic Photochemistry, Padwa, A., Ed.;
Marcel Dekker: New York, 1989; Vol. 10, p. 163, and references cited
therein.
(36) Muszkat, K. A. Top. Curr. Chem. 1980, 88, 89-144, and references
therein.
(37) Karatsu, T.; Tsuchiya, M.; Arai, T.; Sakuragi, H.; Tokumaru, K.
Bull. Chem. Soc. Jpn. 1994, 67, 3030-3039.
(38) von Salis, G. A.; Labhart, H. J. Phys. Chem. 1968, 72, 752-754.
(39) Arai, T.; Karatsu, T.; Tsuchiya, M.; Sakuragi, H.; Tokumaru, K.
Chem. Phys. Lett. 1988, 149, 161-166.
(40) Karatsu, T.; Kitamura, A.; Zeng, H.; Arai, T.; Sakuragi, H.;
Tokumaru, K. Bull. Chem. Soc. Jpn. 1995, 68, 920-928.
(41) Mallory, F. B.; Mallory, C. W. Org. React. 1984, 30, 1-456.
(42) Clar, E.; Stewart, D. G. J. J. Am. Chem. Soc. 1952, 74, 6235-
6238.
(21) Laarhoven, W. H.; Cuppen, Th. J. H. M.; Castel, N.; Fischer, E. J.
Photochem. Photobiol. A: Chem. 1989, 49, 137-141.
(22) Saltiel, J.; Tarkalanov, N.; Sears, D. F., Jr. J. Am. Chem. Soc. 1995,
117, 5586-5587.
(23) Saltiel, J.; Zhang, Y.; Sears, D. F., Jr. J. Am. Chem. Soc. 1997,
119, 11202-11210.
(24) Saltiel, J.; Zhang, Y.; Sears, D. F., Jr. J. Am. Chem. Soc. 1996,
118, 2811-2817.
(25) Arai, T.; Tokumaru, K. Chem. ReV. 1993, 93, 23-39, and references
therein.
(26) Karatsu, T.; Arai, T.; Sakuragi, H.; Tokumaru, K. Chem. Phys. Lett.
1985, 115, 9-15.
(27) Arai, T.; Karatsu, T.; Misawa, H.; Kuriyama, Y.; Okamoto, H.;
Hiresaki, T.; Furuuchi, H.; Zeng, H.; Sakuragi, H.; Tokumaru, K. Pure Appl.
Chem. 1988, 60, 989-998.
(43) Clar, E.; Lombardi, L. Ber. 1932, 65, 1411-1420.