Determination of the photochemical efficiency of o-quinodimethane ring closure in room-temperature solutions by using time-delayed, two-color photolysis technique

Article abstract of DOI:10.1021/jo991518s

Photochemical efficiency of o-quinodimethane (3) ring closure at room temperature was determined by using a time-delayed, two-color photolysis technique. o-Quinodimethane (3) was generated by the photolysis of 1,2- bis[(phenylseleno)methyl]benzene (1) by a KrF (248 nm) laser pulse and thus- generated 3 was photolyzed by a subsequent XeCl (308 nm)/XeF (351 nm) laser pulse with varying delay time of 0 to 3 s. The time profile of 3 was monitored by the chemical analyses of benzocyclobutene (5) (a photochemical product of 3), which was formed by a one-photon process, and the spiro dimer of 3 (4) (a thermal product of 3) in the two-color photolysis experiments. The time profile of 3 followed a second-order decay kinetics. The photochemical efficiency was obtained by the analysis of the delay-time dependence of the product yields; those of the consumption of 3 and the conversion 3 → 5 by a single pulse of the excimer laser were 81% and 5.7% for the XeCl laser, and 73% and 2.3% for the XeF laser. This difference was attributed to the different excited states involved in the photolysis. In contrast to the photolysis of 3 in argon or rigid organic matrixes, it was revealed that photochemical conversion 3 → 5 was not the main path in the solutions, and intermolecular reactions predominated.

Full text of DOI:10.1021/jo991518s

2350
J . Org. Chem. 2000, 65, 2350-2357
Deter m in a tion of th e P h otoch em ica l Efficien cy of
o-Qu in od im eth a n e Rin g Closu r e in Room -Tem p er a tu r e Solu tion s
by Usin g Tim e-Dela yed , Tw o-Color P h otolysis Tech n iqu e
Akihiko Ouchi,*,† Masako Sakuragi,† Haruo Kitahara,‡ and Maurizio Zandomeneghi§
National Institute of Materials and Chemical Research, Tsukuba, Ibaraki 305-8565, J apan, Faculty of
Education, Hirosaki University, Hirosaki 036-8560, J apan, and Department of Chemistry and Industrial
Chemistry, University of Pisa, via Risorgimento 35, I-56126 Pisa, Italy
Received September 27, 1999
Photochemical efficiency of o-quinodimethane (3) ring closure at room temperature was determined
by using a time-delayed, two-color photolysis technique. o-Quinodimethane (3) was generated by
the photolysis of 1,2-bis[(phenylseleno)methyl]benzene (1) by a KrF (248 nm) laser pulse and thus-
generated 3 was photolyzed by a subsequent XeCl (308 nm)/XeF (351 nm) laser pulse with varying
delay time of 0 to 3 s. The time profile of 3 was monitored by the chemical analyses of
benzocyclobutene (5) (a photochemical product of 3), which was formed by a one-photon process,
and the spiro dimer of 3 (4) (a thermal product of 3) in the two-color photolysis experiments. The
time profile of 3 followed a second-order decay kinetics. The photochemical efficiency was obtained
by the analysis of the delay-time dependence of the product yields; those of the consumption of 3
and the conversion 3 f 5 by a single pulse of the excimer laser were 81% and 5.7% for the XeCl
laser, and 73% and 2.3% for the XeF laser. This difference was attributed to the different excited
states involved in the photolysis. In contrast to the photolysis of 3 in argon or rigid organic matrixes,
it was revealed that photochemical conversion 3 f 5 was not the main path in the solutions, and
intermolecular reactions predominated.
In tr od u ction
matrixes at low temperatures, and its photoproduct is
reported to be benzocyclobutene [5; bicyclo[4.2.0]octa-
1,3,5-triene]. From these matrix isolation studies, it is
also believed that photochemistry of 3 is well defined.
However, it is not surprising that the photoproduct of 3
obtained by these matrix isolated experiments was a
unimolecular photoproduct, 5, because the diffusion of
the molecules is restricted in such an environment so that
intermolecular reactions are prevented. To clarify whether
conversion 3 f 5 is also the major photochemical path
in general reaction environments, i.e., the environments
in which the diffusion of 3 is not restricted, determination
of the photochemical efficiency of conversion 3 f 5 in
solutions is necessary. However, the photochemical reac-
tion of 3 in solutions has not been reported so far because
of the occurrence of a fast thermal reaction of 3 even at
-150 °C,6 and suitable experimental techniques have not
been developed for such studies. We report here the
determination of the photochemical efficiencies of conver-
sion 3 f 5 in room-temperature solutions by using a time-
delayed, two-color pulse laser photolysis technique.7
o-Quinodimethane [3; 5,6-bis(methylene)cyclohexa-1,3-
diene]1 is a thermally unstable short-lived intermediate,
and it has been extensively studied from both synthetic2,3
and physical standpoints.4 It is believed that the thermal
reactions of 3 at room temperature or below are well
established and their main products are oligomers,
polymers, and the [4 + 2] spiro dimer [4; 3′,4′-dihydro-
6-methylenespiro(2,4-cyclohexadiene-1,2′(1′H)-naphtha-
lene)].5 On the other hand, the photochemical reaction
of 3 has been only conducted in argon4f,h or rigid organic4a,b
* Corresponding author. e-mail: ouchi@nimc.go.jp.
† National Institute of Materials and Chemical Research.
‡ Hirosaki University.
§ University of Pisa.
(1) Reviews: (a) McCullough, J . J . Acc. Chem. Res. 1980, 13, 270-
276. (b) Charlton, J . L.; Alauddin, M. M. Tetrahedron 1987, 43, 2873-
2889. (c) Scaiano, J . C.; Wintgens, V.; Netto-Ferreira, J . C. Pure Appl.
Chem. 1990, 62, 1557-1564. (d) Martin, N.; Seoane, C.; Hanack, M.
Org. Prep. Proced. Int. 1991, 23, 237-272. (e) Segura, J . L.; Mart´ın,
N. Chem. Rev. 1999, 99, 3199-3246.
(2) Reviews: (a) Oppolzer, W. Angew. Chem., Int. Ed. Engl. 1977,
16, 10-23. (b) Oppolzer, W. Synthesis 1978, 793-802. (c) Funk, R. L.;
Vollhardt, K. P. C. Chem. Soc. Rev. 1980, 9, 41-61. (d) Kametani, T.;
Nemoto, H. Tetrahedron 1981, 37, 3-16.
(3) (a) Sano, H.; Ohtsuka, H.; Migita, T. J . Am. Chem. Soc. 1988,
110, 2014-2015. (b) Hoey, M. D.; Dittmer, D. C. J . Org. Chem. 1991,
56, 1947-1948. (c) Fujihara, H.; Yabe, M.; Furukawa, N. J . Org. Chem.
1993, 58, 5291-5292. (d) Sato, H.; Isono, N.; Okamura, K.; Date, T.;
Mori, M. Tetrahedron Lett. 1994, 35, 2035-2038; and references
therein.
(4) (a) Flynn, C. R.; Michl, J . J . Am. Chem. Soc. 1973, 95, 5802-
5803. (b) Flynn, C. R.; Michl, J . J . Am. Chem. Soc. 1974, 96, 3280-
3288. (c) Tseng, K. L.; Michl, J . J . Am. Chem. Soc. 1977, 99, 4840-
4842. (d) Roth, W. R.; Biermann, M.; Dekker: H.; J ochems, R.;
Mosselman, C.; Hermann, H. Chem. Ber. 1978, 111, 3892-3903. (e)
Roth, W. R.; Scholz, B. P. Chem. Ber. 1981, 114, 3741-3750. (f)
Chapman, O. L.; McMahon, R. J .; West, P. R. J . Am. Chem. Soc. 1984,
106, 7973-7974. (g) Trahanovsky, W. S.; Macias, J . R. J . Am. Chem.
Soc. 1986, 108, 6820-6821. (h) Chapman, O. L.; J ohnson, J . W.;
McMahon, R. J .; West, P. R. J . Am. Chem. Soc. 1988, 110, 501-509.
(5) For example, Errede, L. A. J . Am. Chem. Soc. 1961, 83, 949-
954.
(6) Spectroscopic studies on the photolysis of 3 have been reported
at -196 °C in organic matrixes4a,b and at 10-15 K in argon matrixes.4f,h
Irradiation (>254 nm) of 3 in the organic and argon matrixes gave
benzocyclobutene (5), which was detected by UV absorption spectrom-
etry. Compound 5 was isolated after the photolysis of 3 in the organic
matrixes but without its chemical yield.4a,b On melting of the matrix
at ca. -150 °C before the photolysis, 3 was easily converted to spiro
dimer 4 by a thermal [4 + 2] cycloaddition reaction.4a,b
(7) For example, (a) Netto-Ferreira, J . C.; Wintgens, V.; Scaiano, J .
C. Tetrahedron Lett. 1989, 30, 6851-6854. (b) Bendig, J .; Mitzner, R.
Ber. Bunsen-Ges. Phys. Chem. 1994, 98, 1004-1008. (c) Ouchi, A.;
Koga, Y. Tetrahedron Lett. 1995, 36, 8999-9002. (d) J ime´nez, M. C.;
Miranda, M. A.; Scaiano, J . C.; Tormos, R. Chem. Commun. 1997,
1487-1488.
10.1021/jo991518s CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/28/2000

Time-Delayed, Two-Color Photolysis of o-Quinodimethane
J . Org. Chem., Vol. 65, No. 8, 2000 2351
Sch em e 1
occurs when molecules are designed to give high photo-
chemical efficiency.
Resu lts
P r ecu r sor s of o-Qu in od im eth a n e 3. A necessary
requirement for the precursor of 3 in our time-delayed,
two-color photolysis experiments is high efficiency of the
formation of 3 by a single laser pulse. If the efficiency is
low, several time-delayed, two-color photolysis cycles are
necessary to obtain detectable amounts of the products,
4 and 5. However, the occurrence of secondary reactions
of 4 (vide infra) is expected to cause complication in the
analysis of the reaction. Therefore, the selection of a
suitable precursor for 3 is essential in our experiments.
The precursor used in the rigid organic matrix experi-
ments was 1,4-dihydrophthalazine (6);4a,b however, 6 is
reported to be unstable at room temperature so that 6
was excluded as a possible precursor in our experiment.
Other compounds that have been used as precursors of
31 are 2-indanone (7), 3-isochromanone (8), sultine [9; 1,4-
dihydro-2,3-benzoxathiin 3-oxide], and 1,3-dihydroisothion-
aphthene 2,2-dioxide (10).
The strategy used in our experiments for the photolysis
of 3 in room-temperature solutions was the fast genera-
tion of 3 and its photolysis before the occurrence of its
thermal reaction. We have generated the short-lived
intermediate 3 by the first laser pulse and, by the second
laser pulse, we have targeted 3 at a particular time in
the course of the reaction. The fast generation of 3 was
accomplished by KrF excimer laser photolysis of 1,2-bis-
[(phenylseleno)methyl]benzene (1) through a two-photon
process,8 and successive photolyses of 3 were conducted
by a XeCl or a XeF excimer laser (cf. Scheme 1). In our
time-delayed, two-color laser photolysis experiments at
room temperature,9 we have traced the formation of a
photochemical product, benzocyclobutene 5, and a ther-
mal product, spiro dimer 4. The photochemical efficiency
of the formation of 5 was determined by the analysis of
the delay-time dependence of the product yields.
The photochemical efficiencies of the consumption of
3 and conversion 3 f 5 by a pulse of the excimer lasers
were 81% and 5.7%, respectively, for the XeCl laser, and
73% and 2.3% for the XeF laser. This indicates that
photochemical conversion 3 f 5 was not the main path
in solution, which is in contrast to the results in the
restricted rigid matrixes.
Compounds 7-10 had very small or no absorption
above 300 nm, but considerable absorption was observed
below 270 nm so that a KrF (248 nm) excimer laser was
used for the generation of 3. The molar absorption
coefficient of 7-10 at 248 nm are 840 M-1‚cm-1 for 7,
210 M-1‚cm-1 for 8, 175 M-1‚cm-1 for 9, and 160 M-1‚cm-1
for 10. The photolyses were conducted by a single pulse
of a KrF (1.25 × 1021 photons‚m-2‚pulse-1) excimer laser
on 10-4 M acetonitrile solutions; the consumption was
found to be 9.5% for 7, 3.8% for 8, 6.7% for 9, and 2.2%
for 10.10a An attempt to trap 3 by maleic anhydride (11)
through a cycloaddition reaction was also conducted for
7-10; 10-4 M acetonitrile solutions of 7-10 were pho-
tolyzed by a single pulse of the KrF excimer laser (1.25
× 1021 photons‚m-2‚pulse-1) in the presence of 10-3
M
11. The yield11 of the cycloadduct and the consumption
of the precursor were 4.7 and 9.4%, respectively, for 7,
1.4 and 3.3% for 8, 3.5 and 7.2% for 9, and 0.7 and 1.9%
for 10.10a This indicates that the photochemical efficiency
of 7-10 is not sufficient to use these compounds as
precursors of 3 in our time-delayed, two-color experi-
ments.
To the best of our knowledge, this is the first report
on the determination of photochemical efficiency of short-
lived intermediates by product analysis. This technique
complements conventional spectroscopic methods in the
study of short-lived intermediates especially in the cases
where many intermediates which have absorptions in the
same spectroscopic regions are generated by the photoly-
sis. Such spectral overlap of the intermediates often
We have recently developed an efficient precursor of
3, 1,2-bis[(phenylseleno)methyl]benzene (1).8 The con-
sumption of 1 and the yield11 of the cycloadduct of 3 and
11 by a single pulse of the KrF excimer laser under the
same condition as before were 91% and 43%, respectively.
Although the formation of 3 from 1 proceeds by a two-
photon process, which was confirmed by experiments on
laser fluence dependence,8 the photochemical efficiency
(8) (a) Ouchi, A.; Koga, Y. Chem. Commun. 1996, 2075-2076. (b)
Ouchi, A.; Koga, Y. J . Org. Chem. 1997, 62, 7376-7383.
(9) Ouchi, A.; Koga, Y. J . Org. Chem. 1998, 63, 6780-6781.
(10) The results were obtained from the average of (a) three and
(b) two independent runs.
(11) The yield is based on the starting material.

2352 J . Org. Chem., Vol. 65, No. 8, 2000
Ouchi et al.
dide (12) and tetrabutylammonium fluoride (13).13 Tra-
hanovsky et al.4g have shown that 3 was quantitatively
formed from 12 by using more than 25 equiv of 13.
Therefore, we have conducted a trapping experiment of
3 by 11 through generation of 3 from 10-4 M 12 and 100
equiv of 13 in acetonitrile in the presence of 10-3 M 11.
The concentrations of 12 and 11 were set equal to those
of 1 and 11 in the one-color KrF laser photolysis.
However, decomposition of the anhydride moiety occurred
under the experimental conditions so that the determi-
nation of the trapping efficiency of 3 by 11 was unsuc-
cessful.
We have therefore used fumaronitrile (14) as an
alternative trapping agent. Under the same experimental
condition as in the reaction of 3 and 11, the cycloadduct
of 3 and 14 was obtained in 82%.14 The yield of the
cycloadduct was almost the same (81%) when 200 equiv
of 13 were used. It is reported that 39%11 of the cycload-
duct of 3 and 14 is formed by a pulse of the KrF excimer
laser photolysis of 10-4 M 1 in the presence of 10-3 M 14
in acetonitrile.8 From the yield of the cycloadduct of 3
and 14 in the one-color KrF excimer laser photolysis
(39%) and the trapping efficiency of 3 by 14 (82%), we
conclude that the yield of 3 from 1 by the photolysis of a
pulse of the KrF laser was 48%.
On e-Color Excim er La ser P h otolyses. Photolysis
of 10-4 M 1 in acetonitrile was conducted by a KrF (248
nm, 1.25 × 1021 photon‚m-2‚pulse-1), XeCl (308 nm, 1.88
× 1021 photon‚m-2‚pulse-1), and XeF (351 nm, 1.88 × 1021
photon‚m-2‚pulse-1) excimer lasers. The yields11 of 4 and
5 and the consumption of 1 by a pulse of the excimer laser
were 7.0%, 6.3%, and 91%, respectively, for the KrF laser,
and 2.2%, ∼ 0% (below the detection level), and 44% for
the XeCl laser.10a Some unidentified peaks were observed
in HPLC analyses, and TLC analyses also showed the
existence of a significant amount of unidentified com-
pounds at the origin. These compounds are expected to
be oligomers and polymers.5 On the contrary, no reaction
proceeded with the XeF laser, which was due to the
absence of the absorption of 1 at 351 nm.
Tim e-Dela yed , Tw o-Color P h otolyses. Figures 2a,b
show the yields11 of spiro dimer 4 and benzocyclobutene
(5) as a function of the delay time of the two laser pulses
in the time-delayed, two-color photolyses. The horizontal
axes in Figures 2a,b are shown in a logarithmic scale in
order to indicate clearly the fast changes at short delay
times.15 The two-color photolysis was conducted by one
pair of laser pulses; the pair of laser pulses consisted of
one pulse of the KrF excimer laser and one subsequent
pulse of the XeCl or XeF excimer laser which was fired
after various delay times. The consumption of 1 in the
two-color photolyses (91-94%) was independent of the
delay times and almost the same as that in the one-color
KrF laser photolysis (91%). This indicates that the
consumption of 1 proceeded mainly by the KrF laser. On
the contrary, the yields of 4 and 5 were strongly depend-
ent on the delay time of the second XeCl or XeF laser
pulse. At short delay times, the yields of 4 and 5 were
almost the same as those of the one-color KrF laser
photolysis. However, a maximum on the yield of 5 and a
F igu r e 1. Absorption spectra of 1,2-bis[(phenylseleno)methyl]-
benzene 1 (s), spiro dimer 4 (-‚-‚-), and benzocyclobutene
5 (- - -). Concentration: 10-4 M (1, 4) and 10-3 M (5) in
acetonitrile, optical path: 10 mm. The wavelengths of the laser
emissions are indicated in the figure.
for the formation of 3 is sufficiently high. Therefore, 1
was selected as the precursor of 3 in our experiments. It
is important to note that spectroscopic methods could not
be used for tracing 3 because many other speciesssuch
as, monoradical 2a /b, phenylseleno radical, diphenyl
diselenideswhich have absorptions in the same spectro-
scopic regions were generated at the same time by the
photolyses. Although 1 was designed to give high pho-
tochemical efficiency for the generation of 3, spectral
overlap of such species made it impossible to determine
the time profile of 3 by optical spectrometry.
Absor p tion Sp ectr a . Figure 1 shows the absorption
spectra of 1,2-bis[(phenylseleno)methyl]benzene (1), ben-
zocyclobutene (5), and spiro dimer 4. The molar absorp-
tion coefficient (ꢀ) of 1 at the three laser wavelengths was
16800 (at 248 nm), 1800 (at 308 nm), and 20 (at 351 nm)
M-1‚cm-1. Similarly, ꢀ of 5 was 640 (at 248 nm) and 0
(at 308 and 351 nm) M-1‚cm-1, and that of 4 was 1100
(at 248 nm), 6100 (at 308 nm), and 6 (at 351 nm)
M-1‚cm-1. o-Quinodimethane (3) has been reported4a-c,12
to have strong absorption at <240 nm, an absorption
minimum at ca. 300 nm, a maximum at 373 nm (ꢀ >
3000), and the absorption edge at ca. 430 nm. The ratio
of the ꢀ of 3 at the three laser emissions was reported to
be 3.8:1:3.8 (248 nm: 308 nm: 351 nm)4b and 1:2.5 (308
nm: 351 nm).12
Th e Yield of o-Qu in od im eth a n e 3 by a P u lse of
th e Kr F Excim er La ser . Direct measurement of the
yield of 3 generated by a pulse of the KrF excimer laser
was not possible by means of conventional chemical
analysis because of its short lifetime. Spectroscopic
techniques were also not applicable because other species
which had absorption in the same wavelength region as
3 were generated at the same time. We have therefore
used a trapping technique to obtain the yield of 3.
We have reported trapping of 3 by maleic anhydride
(11), a cycloaddition reaction, in the one-color KrF laser
(1.25 × 1021 photons‚m-2‚pulse-1) photolysis of 1. The
yield11 of the cycloadduct with one pulse of the KrF laser
was 41%.8 However, to obtain the yield of 3, it is
necessary to know the trapping efficiency of 3 by 11. To
determine the trapping efficiency, we used a standard
procedure for the generation of 3, i.e., the reaction of
[o-[(trimethylsilyl)methyl]benzyl]trimethylammonium io-
(13) (a) Ito, Y.; Nakatsuka, M.; Saegusa, T. J . Am. Chem. Soc. 1980,
102, 863-865. (b) Ibid. 1982, 104, 7609-7622.
(14) The yields of the cycloadduct of 3 and 14 were 81.6 ( 0.8% (100
equivof 13) and 81.4 ( 1.4% (200 equiv of 13).
(12) Fujiwara, M.; Mishima, K.; Tamai, K.; Tanimoto, Y.; Mizuno,
K.; Ishii, Y. J . Phys. Chem. A 1997, 101, 4912-4915.
(15) The figures with horizontal axes in linear scale are shown in
Supporting Information.

Time-Delayed, Two-Color Photolysis of o-Quinodimethane
J . Org. Chem., Vol. 65, No. 8, 2000 2353
F igu r e 4. Increased yield11 of benzocyclobutene 5 as a
function of the XeCl excimer laser fluence.10b Concentration:
10-4 M 1 in acetonitrile. Optical path: 1 mm. Delay time: 2
ms. KrF laser fluence: 1.25 × 1021 photons‚m-2‚pulse-1 (100
mJ ‚cm-2‚pulse-1).
F igu r e 2. Yields11 of spiro dimer 4 and benzocyclobutene 5
as a function of the delay time between the (a) KrF and XeCl
lasers, and (b) KrF and XeF lasers.10b 4: open symbols; 5:
closed symbols. Combinations of lasers: KrF-XeCl (O, b);
KrF-XeF (0, 9). Concentration: 10-4 M 1 in acetonitrile.
Optical path: 1 mm. Laser fluence: 1.25 × 1021 (KrF) and 1.88
× 1021 (XeCl and XeF) photons‚m-2‚pulse-1
.
F igu r e 5. Absorption of the XeF laser as a function of the
delay time. Concentration: 10-4 M 1 in acetonitrile. Optical
path: 1 mm. Laser fluence: 1.25 × 1021 (KrF) and 1.88 × 1021
(XeF) photons‚m-2‚pulse-1
.
F lu en ce Dep en d en ce on th e Yield of Ben zocy-
clobu ten e 5 in th e Tim e-Dela yed , Tw o-Color P h o-
tolysis. To know the number of photons necessary for
the additional formation of 5, an experiment on the XeCl
laser fluence dependence was conducted in the time-
delayed, two-color photolysis. Figure 4 shows the in-
creased yield of 5 as a function of the XeCl laser fluence
at the fixed delay time of 2 ms. As seen in the figure, the
increased yield of 5 was proportional to the first power
of the XeCl laser fluence. This result indicates that the
formation of 5 by the XeCl laser irradiation was a one-
photon process.
Absor p tion of th e Secon d La ser P u lse in th e
Tim e-Dela yed , Tw o-Color P h otolysis. The delay time
dependence on the absorption of the second laser (XeF
laser) was measured at 351 nm. The XeF laser was used
as the second laser to eliminate the absorption of the
starting material 1 and the products 4 and 5. Figure 5
shows the absorbance at 351 nm as a function of the delay
time in the two-color photolysis. The horizontal axis in
the figure is shown in logarithmic scale in order to show
clearly the fast change at short delay times. As seen in
the figure, the time profile of the absorption at delay
times >10-2 s was consistent with that of the yield of 5
(cf., Figure 2a,b); both the absorption and the yield of 5
decreased in this time domain. The leveling off of the
absorbance at 3.2 × 10-3 is due to the absorption of
F igu r e 3. Increased yield11 of benzocyclobutene 5 as a
function of the delay time between the KrF and XeF lasers
with second-order kinetic fitting.10b Concentration: 10-4 M 1
in acetonitrile. Optical path: 1 mm. Laser fluence: 1.25 × 1021
(KrF) and 1.88 × 1021 (XeF) photons‚m-2‚pulse-1
.
minimum on the yield of 4 were observed at the delay
time of ca. 20 ms. Also in the two-color experiments, TLC
analyses showed the existence of a significant amount
of compounds at the origin,5 most likely oligomers and
polymers. Some unidentified byproducts were also de-
tected by HPLC analyses in the two-color photolyses.
Figure 3 shows the increased yield of 5 as a function
of the delay time of the KrF and XeF lasers. The
increased yield is defined as the difference in the yield11
of 5 between the time-delayed, two-color photolysis and
the one-color KrF laser photolysis. The horizontal axis
in Figure 3 is shown in linear scale so that this time
profile corresponds to the decay curve of 3 in conventional
flash photolysis experiments16 because the increased yield
of 5 reflects the concentration of 3 at each delay time.
As seen in Figure 3, 3 decreased with second-order
kinetics. This is in accord with the previous flash pho-
tolysis studies and is explained by the thermal dimer-
ization of 3.1c,4d,g
(16) Transient absorption was observed at 360-370 nm by the
photolysis of 1 by using a pulse of a 266-nm laser, and the absorption
is in good accord with that of the previous reports:4a-c,12 unpublished
result. However, the decay of 3 was difficult to observe because of the
overlapping of several species such as phenylseleno radicals, diphenyl
diselenide, intermediate 2a /b, and other unidentified species.

2354 J . Org. Chem., Vol. 65, No. 8, 2000
Ouchi et al.
Sch em e 2
diphenyl diselenide. However, the time profiles of the
absorption (Figure 5) and the yield of 5 (Figure 2a,b) were
quite different at shorter delay times. In the case of the
absorption, a rapid increase occurred between 0 and 10-7
s and showed a gradual decrease between 10-7 and 10-2
s, whereas a gradual increase in the yield of 5 was
observed between 10-7 and 10-2 s. These results indicate
the existence of a transient species that has a shorter
lifetime than 3; the transient species absorbed 351-nm
light but was not photochemically converted to 3.
b
a
c
c
a
b
P h otoch em ica l Sta bility of Ben zocyclobu ten e 5
a n d th e Sp ir o Dim er 4. Photolyses of 4 and 5 were
conducted in 10-4 M acetonitrile solutions with one pulse
of the KrF, XeCl, or XeF excimer laser.17 The reaction
conditions were the same as those in the one-color laser
photolyses of 1. In the case of 4, no decomposition was
observed with a pulse of the KrF and XeF lasers and 10%
decomposition with the XeCl laser. When 4 was irradi-
ated in the presence of 1 equiv of diphenyl diselenide,
the decomposition slightly increased to 14% with the
XeCl laser and 5% with the XeF laser. However, in all
cases, 5 was not observed after the laser irradiations.
This indicates that the reverse reaction 4 f 3 was not
involved in the photolysis. In the case of 5, no decomposi-
tion was observed in all the three laser irradiations.17
d
laser irradiation proceeded through a one-photon process
so that the photochemical process was 3 f 5 and not 2a /b
f 3 f 5.
In the case of the two-color photolyses, the existence
of the maximum for the yield of 5 in Figure 2a,b means
that highest concentration of 3 was reached at ca. 20 ms
delay time. This result indicates the existence of a slow
process for the generation of 3. On the other hand, the
formation of a considerable amount of 5 by the one-color
KrF laser photolysis implies the existence of another very
fast process for the generation of 3 which enables 3 to
absorb third photon within the same laser pulse (<30
ns). The origin of the two processes is not clear at present
due to the lack of information in this time domain.
Rea ction P a th s of o-Qu in od im eth a n e 3. Scheme 2
shows the reaction paths of 3 in the time-delayed, two-
color photolysis. First, 3 is generated by a pulse of the
KrF laser. When the photolysis of 3 is conducted by a
pulse of the XeCl/XeF laser, a part of 3 (100c %, 0 e c e
1) undergoes photochemical processes and the rest [100-
(1 - c) %] reacts by thermal processes. In the photo-
chemical processes, a part of 3 forms 5 (100b %, 0 e b e
1), and the rest [100(1 - b) %] leads to the formation of
other photochemical byproducts. On the other hand, in
the thermal processes, the formation of 4 (100a %, 0 e a
e 1) and other byproducts [100(1 - a) %] are expected.
From the control experiments, partial decomposition
of 4 was observed by the second XeCl/XeF laser irradia-
tion especially in the presence of diphenyl diselenide.
Therefore, we assume that 100d % (0 e d e 1) of 4 formed
by the thermal process are decomposed by the second
laser pulse. It was also proved by the control experiments
that 5 was not generated by the XeCl/XeF laser photoly-
ses of 4.
Tim e P r ofile of o-Qu in od im eth a n e 3. The time
profiles of 3-5 in the one-color KrF laser and the time-
delayed, two-color photolyses are shown schematically in
Schemes 3a and 3b, respectively. The rise time of 3 is
neglected in these figures because the rise time is much
shorter than the decay time (cf., Figures 2a,b and 3) so
that the time profile of each species can be approximated
as shown in Schemes 3a,b. As shown in Scheme 3a, when
1 is irradiated by a pulse of the KrF laser (at t ) 0), 3 is
generated by a two-photon process with the concentration
of [QM]0, and at the same time 5 is formed with the
concentration of [CB]s by a three-photon process via 3.
The concentration of 5 did not change in the course of
the reaction, because 5 was stable in room-temperature
Discu ssion
Rea ction Mech a n ism on th e F or m a tion of Sp ir o
Dim er 4 a n d Ben zocyclobu ten e 5. In the case of one-
color photolyses, the consumption of 1 in the XeCl laser
photolysis was lower than that with the KrF laser despite
the larger number of photons in the former pulse. This
result can be explained by the difference in the molar
extinction coefficient of 1 (16800 M-1‚cm-1 at 248 nm and
1800 M-1‚cm-1 at 308 nm) and also by the participation
of different excited states at each laser wavelength.8b The
generation of 3 by the laser photolysis is evidenced by
the trapping of 3 by various dienophiles, and it is also
proved that 3 is formed by a two-photon process.8 The
formation of 518 is rationalized by further photolysis of 3
within the same laser pulse (Scheme 1: Laser 1 ) Laser
2 ) Laser 3). This is consistent with the formation of 5
by conventional light photolysis of 3 in low-temperature
matrixes.4a,b,f,h
The increase of the yield of 5 in the time-delayed, two-
color photolyses (cf., Figure 2) is explained by the
photolysis of 3 by the second XeCl/XeF laser pulse, i.e.,
the photolysis of two-photon intermediate 3 to form 5
(Scheme 1: Laser 1 ) Laser 2 ) KrF laser, Laser 3 )
XeCl/XeF laser). However, another possibility is the
photolysis of one-photon intermediate 2a /b through a
two-photon process (Scheme 1: Laser 1 ) KrF laser,
Laser 2 ) Laser 3 ) XeCl/XeF laser) because the
presence of intermediate 2a /b was also observed by flash
photolysis experiments.16 The experiment on the XeCl
laser fluence dependence in the two-color photolysis
showed that the increased yield of 5 was proportional to
the first power of the XeCl laser fluence (cf., Figure 4).
This result indicates that the formation of 5 by the XeCl
(17) The conversions10a of 4 and 5 were 0.6 (+0.6, -0.8) % and 1.4
(+1.1, -1.0) % for the KrF laser, 9.9 (+1.4, -1.0) % and 0 (+0.6, -1.0)
% for the XeCl laser, 0 (+1.6, -0.9) % and 0 (+0.7, -1.1) % for the
XeF laser.
(18) It is important to note that 5 is not formed by the thermal
reaction of 3 below room temperature: cf. Cava; M. P.; Deana, A. A.
J . Am. Chem. Soc. 1959, 81, 4266-4268.

Time-Delayed, Two-Color Photolysis of o-Quinodimethane
Sch em e 3
J . Org. Chem., Vol. 65, No. 8, 2000 2355
solutions. However, thus-generated 3 undergoes dimer-
ization by a thermal [4 + 2] cycloaddition to give spiro-
dimer 4. When the concentrations of 3 and 4 at time t
are [QM]t and [DI]t, respectively, we obtain
[DI]i given above, we obtain
a ) 2[DI]i/[QM]0 ) 0.15
This indicates that the efficiency of the thermal process
3 + 3 f 4 was 15%.
[DI]t ) (a/2)([QM]0 - [QM]t)
(a: const.; 0 e a e 1) (1)
o-Quinodimethane (3) showed second-order decay, both
by our experiment and by flash photolysis studies, and
this was explained by the thermal dimerization which
led to 4.1c,4d,g However, as shown above, the efficiency of
the thermal reaction 3 + 3 f 4 was only 15% so that
other bimolecular processes, which probably lead to the
higher molecular weight compounds, must be involved
in the thermal processes.
At t ) ∞, when the concentration of 4 is [DI]i, eq 1
becomes
[DI]i ) (a/2)[QM]0
because, [QM]∞ ) 0.
(1′)
P h otoch em ica l Decom p osition of Sp ir o Dim er 4
in th e Rea ction Mixtu r e. The control experiments
revealed that the photochemical decomposition of 4 was
accelerated by the presence of diphenyl diselenide. This
was rationalized by the reaction of 4 and photochemically
generated phenylseleno radicals. In the reaction mixture,
the presence of radical species other than phenylseleno
radicals was anticipated so that the extent of the
decomposition of 4 in the reaction mixture was expected
to be larger.
Scheme 3b is the time profile of 3-5 in the time-
delayed, two-color photolysis. When the second laser is
irradiated at the delay time of t, the concentrations of 3
and 5 immediately change from [QM]t f [QM]Lt and [CB]s
f [CB]L , respectively, due to the photochemical reaction
t
3 f 5. Thus, we obtain
[CB]L - [CB]s ) b([QM]t - [QM]L )
t
t
(b: const.; 0 e b e 1) (2)
When the second laser is fired in the two-color pho-
[QM]L ) (1 - c)[QM]t (c: const.; 0 e c e 1) (3)
tolysis, the concentration of 4 changes from [DI]t f [DI]L
t
t
due to decomposition, so that
Equations 2 and 3 are reduced to
[DI]L ) (1 - d)[DI]t (d: const.; 0 e d e 1) (5)
[QM]t ) (1/bc)([CB]L - [CB]s)
(4)
t
t
The final concentration of 4, [DI]iLt, is
Equation 4 indicates that the value ([CB]L - [CB]s) is
t
proportional to the concentration of 3 at the delay time
t, [QM]t.
[DI]iL ) [DI]L + (a/2)[QM]L
t
(6)
t
t
Figure 3 shows the increased yield of 5 as a function
of the delay time. In practice, t ) 0 corresponds to the
delay time where the concentration of 3 becomes its
maximum, i.e., at ca. 20 ms in Figure 3. The increased
From eqs 1, 1′, 3, 5, and 6, we have
[QM]t ) {2/a(c - d)}{(1 - d)[DI]i - [DI]iL
}
(7)
t
yield of 5 is proportional to the value ([CB]L - [CB]s) so
t
that the decay curve in Figure 3 corresponds to the decay
curve of 3 obtained by spectroscopic methods. The curve
in Figure 3 shows second-order decay, which is consistent
with the results of spectroscopic studies of 3 conducted
with different precursors.1c,4d,g
At t ) ∞, [QM]t becomes zero so that the value d
derived from eq 7 is
d ) 1 - ([DI]iL∞/[DI]i)
(8)
Efficien cy of th e Th er m a l F or m a tion of Sp ir o
Dim er 4. We have revealed that 48%11 of 3 was formed
from 10-4 M 1 by a pulse of the KrF laser. Therefore,
[QM]0 ) 4.8 × 10-5 M. The value [DI]i is related to the
yield of 4 obtained by the one-color KrF excimer laser
photolysis, which was [DI]i ) 3.5 × 10-6 M (7.0% yield of
4 from 10-4 M 1; two molecules of 3 are necessary for
the formation of 4). From eq 1′ and the values [QM]0 and
because the case c ≈ d only leads to a trivial solution.
The concentration [DI]iL is related to the yield of 4 at
∞
longer delay time where the yield becomes constant (cf.
Figure 2),19 which was 1.8 × 10-6 M (3.5% yield of 4 from
10-4 M 1; two molecules of 3 are necessary for the
(19) The leveling off of the yield of 4 is clearly observed when the
delay time is shown in a linear scale. Cf. Supporting Information.

2356 J . Org. Chem., Vol. 65, No. 8, 2000
Ouchi et al.
formation of 4) for the XeCl laser and 2.5 × 10-6 M (5.0%
a, c, and d into eq 10, we obtain
bXeCl ) 0.071 and bXeF ) 0.031
yield of 4 from 10-4 M 1) for the XeF laser.
By substituting [DI]iL and [DI]i values to eq 8, we
∞
obtain
The values bXeCl and bXeF indicate that the efficiency
of the photochemical conversion 3 f 5 by a pulse of the
XeCl laser was 7.1% and that by the XeF laser was 3.1%.
The net photochemical efficiency for the photochemical
conversion 3 f 5 is presented by the product bc so that
dXeCl ) 0.50 and dXeF ) 0.29
These values dXeCl and dXeF indicate that a larger
portion of 4 was decomposed by the XeCl laser than by
the XeF laser. This tendency is in accord with the control
experiments, but the extent of the decomposition was 4-
to 6-fold larger than those of the control experiments
conducted in the presence of diphenyl diselenide.
Efficien cy of th e P h otoch em ica l F or m a tion of
Ben zocyclobu ten e 5 fr om o-Qu in od im eth a n e 3. The
ratio of photochemical and thermal processes by the
irradiation of the second laser pulse is indicated by the
values c and 1 - c, respectively. From eqs 1′ and 7 at t )
0, the value c is
bcXeCl ) 0.057 and bcXeF ) 0.023
Therefore the net photochemical efficiency is 2.5-fold
higher for the XeCl laser than the XeF laser. These
results cannot be explained by the difference of the molar
absorptivity of 3 at each wavelength because the reported
molar absorptivity is 2.5-12 to 3.8-fold4b larger at 351 nm
than at 308 nm and this tendency is opposite to that of
the values bcXeCl and bcXeF. The reported λmax of the first
absorption band of 3 is 370 nm, and the laser wavelength
of 308 nm is located just at the minimum of its shorter
wavelength side.1c,4b,12 Therefore, the involvement of
higher excited state is expected in the photochemical
conversion 3 f 5. Indeed, participation of the S2 state
for the conversion 3 f 5 has been postulated;4b,c however,
S0 f S2 transition is still not explicitly observed by
spectroscopic means, and it is believed that the transition
should be buried under more intense S0 f S1 band. We
think that our result on the wavelength dependence of
the photochemical efficiency of the formation of 5 is
consistent with this prediction on the involvement of the
S0 f S2 transition for the conversion 3 f 5.
c ) 1 - ([DI]iL0/[DI]i)
(9)
Concentration [DI]iL is related to the yield of 4 at the
0
delay time of 0 (in practice, as indicated before, at the
delay time of ca. 20 ms). When the second laser was the
XeCl laser, the yield of 4 at t ) 0 was 1.3% so that [DI]iL
0
) 6.5 × 10-7 M. Similarly, when the second laser was
the XeF laser, [DI]iL ) 9.5 × 10-7 M. By substituting
0
the values [DI]iL and [DI]i () 3.5 × 10-6 M) into eq 9,
0
we obtain
The values bXeCl and bXeF clearly indicate that photo-
chemical conversion 3 f 5 is not the main path in the
solution. Intermolecular reactions are likely to be the
main photochemical pathways in such environments, and
we are now investigating the intermolecular photochemi-
cal products in the reaction.
cXeCl ) 0.81 and cXeF ) 0.73
These values of cXeCl and cXeF indicate that the propor-
tion of the photochemical process of 3 by a pulse of the
XeCl laser was 81% and that of the XeF laser was 73%.
The rest of 3 reacted by thermal processes. Although the
molar absorptivity of 3 was 2.5-12 to 3.8-fold4b larger at
351 nm than at 308 nm, the efficiency of the consumption
of 3 was ca. 1.1-fold larger when the XeCl laser was used
instead of the XeF laser. The tendency on the molar
absorptivity is opposite to that of the c values, and the
result is best explained by the participation of different
excited states.20
Con clu sion
Photochemical efficiencies of o-quinodimethane (3) at
room temperature were determined by using the time-
delayed, two-color excimer laser photolyses technique.
The time profile of 3 was monitored by chemical analysis
of benzocyclobutene (5) (a photochemical product of 3)
and the spiro dimer 4 (a thermal product of 3) in the two-
color photolysis experiments. The time profile of 3 fol-
lowed second-order decay kinetics. Photochemical effi-
ciencies were obtained by analysis of the delay-time
dependence of the product yields; those of the consump-
tion of 3 and the conversion 3 f 5 by a single pulse of an
excimer laser were 81% and 5.7% for the XeCl laser, and
73% and 2.3% for the XeF laser. The difference between
the two lasers was attributed to different excited states
involved. In contrast to the photolysis of 3 in argon or
rigid organic matrixes, it was revealed that photochemi-
cal conversion 3 f 5 was not the main path in the
solutions and intermolecular reactions seemed to be the
main paths.
The efficiency of the photochemical conversion 3 f 5
in the photochemical processes is related to the value b.
The value b obtained from eqs 4 and 7 is
b ) {a(c - d)/2c}([CB]L - [CB]s)/
t
{(1 - d)[DI]i - [DI]iLt} (10)
In principle, the value of b can be obtained from any
delay time t; however, the most reliable values for [CB]L
t
and [DI]iL are obtained at t ) 0 where [CB]L gives its
t
t
maximum and [DI]iLt gives its minimum because relative
experimental errors should be smallest. When the second
laser was the XeCl laser, the yield of 5 was 9.1% (from
10-4 M 1) so that [CB]L0 ) 9.1 × 10-6 M. Similarly, when
the second laser was the XeF laser, [CB]L ) 7.4 × 10-6
Exp er im en ta l Section
0
M. The yield of 5 in the one-color KrF laser photolysis
was 6.3% (from 10-4 M 1) so that [CB]s ) 6.3 × 10-6 M.
By substitution of these concentrations and the values
1,2-Bis[(phenylseleno)methyl]benzene (1) and the cycload-
duct of 3 and 11, and that of 3 and 14, were prepared according
to the reported procedures.21 Detailed descriptions on the
(20) Cf. Excitation spectrum of 3 in ref 4a,b.
(21) Reference 8b, Supporting Information.

Time-Delayed, Two-Color Photolysis of o-Quinodimethane
J . Org. Chem., Vol. 65, No. 8, 2000 2357
preparation of compounds 4, 5, 9, 10, and 12 are given as
Supporting Information. Tetrabutylammonium fluoride hy-
drate (13) and 2-indanone were purchased from Aldrich
Chemical Co. Inc., fumaronitrile (14) from Wako Chemicals,
3-isochromanone (8) from Tokyo Kasei Kogyo Co., Ltd., and
spectroscopic grade acetonitrile from Cica-Merck/Kanto Chemi-
cals.
ducted by using the KrF (1.25 × 1021 photon‚m-2‚pulse-1) and
XeF (1.88 × 1021 photon‚m-2‚pulse-1) lasers. The experimental
setup and the solutions were the same as those in the time-
delayed, two-color laser photolyses. The energies of the trans-
mitted XeF laser was measured with the biplanar phototube
and the storage scope, and the absorbance was calculated
based on the transmitted energy of the solvent acetonitrile.
On e-Color Excim er La ser P h otolyses of 1 a n d Decom -
p osition of 4, 5, a n d 7-10. The photolyses of 1, 4, 5, and
7-10 were conducted using the same experimental setup and
procedure as those of the time-delayed, two-color laser pho-
tolyses but with a single excimer laser. The fluence of the
lasers were the same as that in the two-color experiments.
Tr a p p in g of o-Qu in od im eth a n e 3 by F u m a r on itr ile 14.
The reactions were conducted by following the reported
procedure4g with a 10-4 M [o-[(trimethylsilyl)methyl]benzyl]-
trimethylammonium iodide (12) acetonitrile solution and 100
or 200 equiv of tetrabutylammonium fluoride (13) (1.0 M in
acetonitrile) in the presence of 10-3 M 14.
Tim e-Dela yed , Tw o-Color La ser P h otolyses. The reac-
tions were conducted on 0.05 mL of 10-4 M 1 acetonitrile
solutions under a nitrogen atmosphere at room temperature
by using a synthetic quartz cuvette of 10-mm width and 1-mm
optical path. The first laser pulse was generated by a Lambda
Physik EMG 201 MSC [KrF, 1.25 × 1021 photons‚m-2‚pulse-1
(100 mJ ‚cm-2‚pulse-1), pulse width (fwhm): 30 ns] and the
second pulse by a Lambda Physik EMG 102 MSC [XeCl, 1.88
× 1021 photons‚m-2‚pulse-1 (121 mJ ‚cm-2‚pulse-1), pulse width
(fwhm): 14 ns] or by a Lambda Physik COMPex 102 [XeF,
1.88 × 1021 photons‚m-2‚ pulse-1 (106 mJ ‚cm-2‚pulse-1), pulse
width (fwhm): 19 ns] excimer lasers. The pulse energy was
measured by means of a Gentec ED-500 joulemeter and a
Tektronic T912 10 MHz storage oscilloscope. The delay time
was defined as the time between two peaks of the KrF and
XeCl or XeF laser pulses and varied from 0 ns to 3 s. The delay
times between the two laser pulses were controlled by a
Stanford Research Systems DG535 four-channel digital delay/
pulse generator. The delay times of the two laser pulses were
measured with a Hamamatsu Photonics R1193U-55 biplanar
phototube and a Iwatsu Electric TS-8123 storage scope.
F lu en ce Dep en d en ce of th e Tim e-Dela yed , Tw o-Color
La ser P h otolysis. The photolyses were conducted by using
the KrF [1.25 × 1021 photons‚m-2‚pulse-1 (100 mJ ‚cm-2
‚pulse-1)] and XeCl [varied between 0.78 and 2.33 × 1021
photons‚m-2‚pulse-1 (50-150 mJ ‚cm-2 ‚pulse-1)] lasers at the
delay time of 2 ms.
An a lyses of th e P r od u cts. The yields of 4, 5, cycloadduct
of 3 and 14, and consumption of 1 were determined by HPLC
analysis (detection wavelengths of UV detectors: 254 and 300
nm) on a Merck Supersphere 60 RP-8e column (244 mm × 4
mm i.d.), and their retention times were compared with
authentic samples.
Ack n ow led gm en t. We thank the Agency of Indus-
trial Science and Technology (MITI J apan) for financial
support.
Su p p or tin g In for m a tion Ava ila ble: Preparation of com-
pounds 4, 5, 9, 10, and 12, and Figure 2 with the delay time
in linear scale (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
Tr a n sm itta n ce of th e Secon d La ser P u lse in th e Tim e-
Dela yed , Tw o-Color P h otolysis. The photolyses were con-
J O991518S