Study of intramolecular hydrogen abstraction on photoexcited 1,3-dimethylanthrone in the condensed phase

Article abstract of DOI:10.1016/S0009-2614(98)00821-5

The time-resolved thermal lensing (TRTL) technique was applied to determine the reaction heat of intramolecular hydrogen abstraction for excited 1,3-dimethylanthrone (DMAT). The excited carbonyl group abstracts the hydrogen atom of the neighboring methyl group to produce an intermediate. The intermediate decayed with 1.2±0.1 μs. The heat conversion efficiency was determined to be 1.02±0.02, revealing that excited DMAT emits all the energy as heat. The time profile of a TRTL signal led to the value of the reaction heat for the intermediate, 131±6 kJ/mol. We briefly discuss candidates for the intermediate.

Full text of DOI:10.1016/S0009-2614(98)00821-5

4 September 1998
Ž
.
Chemical Physics Letters 293 1998 436–440
Study of intramolecular hydrogen abstraction on photoexcited
1,3-dimethylanthrone in the condensed phase
)
Takashi Omori, Tadashi Suzuki, Teijiro Ichimura
Department of Chemistry and Materials Science, Tokyo Institute of Technology, 2-12-1 Ohokayama, Meguro, Tokyo 152-8551, Japan
Received 28 May 1998; in final form 10 July 1998
Abstract
The time-resolved thermal lensing TRTL technique was applied to determine the reaction heat of intramolecular
Ž
.
Ž
.
hydrogen abstraction for excited 1,3-dimethylanthrone DMAT . The excited carbonyl group abstracts the hydrogen atom of
the neighboring methyl group to produce an intermediate. The intermediate decayed with 1.2"0.1 ms. The heat conversion
efficiency was determined to be 1.02"0.02, revealing that excited DMAT emits all the energy as heat. The time profile of a
TRTL signal led to the value of the reaction heat for the intermediate, 131"6 kJrmol. We briefly discuss candidates for the
intermediate. q 1998 Elsevier Science B.V. All rights reserved.
Ž
.
1. Introduction
tively. The large energy gap 86 kJrmol implies
that the structure of the trans form should be less
planar because of vinylrphenyl steric hinderance.
One thing to note here is that there may be some
uncertainty estimated from the formation quantum
yield of cis- and trans-enols. Their transient absorp-
tion spectra were practically identical. Namely, each
Reaction dynamics of intramolecular hydrogen
abstraction has been extensively studied on o-alkyl-
substituted carbonyl compounds such as 2-methyl-
Ž
. w
x
benzophenone MBP 1–5 . After excitation, the
carbonyl group abstracts the hydrogen atom of the
neighboring methyl group to produce cis- and trans-
Ž
.
quantum yield fcisrftrans s0.70r0.30 was ob-
tained by extrapolation of the transient absorption
intensities to the time ts0 on the assumption that
the extinction coefficients were identical for the two
Ž
.
enols Scheme 1 , which finally return to the parent
w
x
keto form 2–5 .
w x
In a previous Letter 6 , the time-resolved thermal
Ž
.
w x
enols 6 .
lensing TRTL technique was applied to the deter-
mination of reaction heats for the enol formation.
The enthalpy differences of the two enol forms
against the keto form were successfully obtained to
In order to solve this problem, we investigated a
system whose reaction mechanism is expected to be
similar to the cis-enolization of MBP, because the
Ž
.
Ž
.
Ž
.
be 116 kJrmol cis and 202 kJrmol trans , respec-
two benzene rings in 1,3-dimethylanthrone DMAT
Ž
.
are bridged by a methylene group Scheme 1 .
In this Letter, we report the heat of reaction for
synthesized DMAT, which is expected to produce a
)
Corresponding author. E-mail: tichimur@chem.titeh.ac.jp
0009-2614r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 98 00821-5

(
)
T. Omori et al.rChemical Physics Letters 293 1998 436–440
437
Scheme 1.
Ž
.
sole enol and will discuss the enthalpy for the enols
between MBP and DMAT.
lamp Ushio UXL-300DO; 300 W for a monitoring
light source.
Ž
DMAT was synthesized as follows. 5- m-
.
xylyl magnesium bromide, prepared by the proce-
2. Experimental
dure with 5-bromo-m-xylene and magnesium turn-
Ž
ings in dry ether, was added dropwise under anhy-
.
The experimental apparatus for the TRTL mea-
drous condition into a well-stirred suspension of
w
x
surement has been described elsewhere 6,7 . A XeCl
phthalic anhydride in dry benzene. The reaction mix-
ture was refluxed for 3 h. The solvents were removed
by evaporation and the residue was collected and
treated with an aqueous solution of Na2CO3 and
dichloromethane. The aqueous fraction was acidified
with hydrochloric acid and the precipitation was
filtered off and washed with water. The precipitation
Ž
excimer laser 110 mJrpulse at 308 nm, 10 ns pulse
.
duration was used as an excitation light source. The
sample solution flowed into a cuvette NSG T-59FL-
10; 10 mm optical path length . A He–Ne laser
Uniphase 1103P; 2 mW used as a probe light for a
thermal lens which was focused in front of the
sample cuvette with a 30 mm focal length lens
collinear to the excitation beam focused with a 200
mm focal length lens. The probe light passing through
a pinhole Corion 2401; 300 mm diameter was
detected by a photomultiplier tube Hamamatsu
Ž
.
Ž
.
Ž
.
Ž .
of 2- 3,5-dimethylbenzoyl benzoic acid a was puri-
fied by triturating with ethanol and obtained as a
Ž .
yellow powder. The acid a and KOH were added to
Ž
.
activate zinc and the reaction mixture was refluxed
for ;12 h. The reaction mixture was filtered off, the
residue was washed well with aqueous ammonia
solution and the combined filtrate was acidified with
Ž
.
R928 . The output signals were converted into volt-
age with a 50 or 500 V load register, measured by a
Ž
Ž
.
digital oscilloscope Sony Tektronix TDS380P; 2
hydrochloric acid. 2- 3,5-dimethylbenzyl benzoic
.
Ž .
GHzrS and transferred to a personal computer. The
TRTL signals were averaged over 50 shots.
acid b was collected, washed with water and re-
crystallized from methanol. The precipitating acid
Ž .
Transient absorption spectra were measured with
the conventional laser flash photolysis system con-
sisting of the excimer laser and a steady-state Xe
b was added under anhydrous conditions to conc.
H2 SO4 at 08C and the mixture was kept at this
temperature for 12 h with magnetic stirring. The

(
)
438
T. Omori et al.rChemical Physics Letters 293 1998 436–440
product was isolated by column chromatography with
benzene. Vaporizing the solvent, DMAT was ob-
tained as pale-yellowish crystals. DMAT was puri-
fied by recrystallization with methanol five times
Ž
.
purity 99.5% .
Ž
.
Ethanol Kanto Chemical, GR grade was used
Ž
from a freshly opened bottle. Pyridazine Tokyo
Kasei, GR grade was purified by trap-to-trap distil-
.
lation under a vacuum. All samples were deaerated
Ž
.
by bubbling Ar gas purity 99.95% purged by
ethanol vapor for 30 min before use. All measure-
ments were carried out at room temperature. Absorp-
tion spectra were measured with a double beam
Ž
.
spectrometer Jasco Ubest V-550 .
Fig. 1. Transient absorption spectra of DMAT in ethanol at 0 ns
Ž
.
Ž
.
Ž
.
Ž
.
Ž .
v , 350 ns I , 800 ns ' , 1.7 ms \ and 4.9 ms l , after
the 308 nm excitation. Three peaks are observed at 315, 360 and
425 nm. The time profiles of the absorption monitored at 360 and
425 nm are shown in the inset. The best-fit curves are shown as
the solid lines. The lifetime was obtained to be 1.2"0.1 ms.
3. Results and discussion
The 1H-NMR measurement was carried out to
determ ine the equilibrium constant for
anthroneranthranol. In the case of anthrone, the
kinetics and thermodynamics of the tautomerization
between anthrone and anthranol are well understood
nm are shown in the inset of Fig. 1. The lifetime of
the reaction intermediate, which has the absorption
maximum at 360 nm, was determined with a single-
exponential equation to be 1.2"0.1 ms. The solid
lines show the best-fitting curves. The time constant
of the decay at 425 nm was identical to that at 360
nm. The absorption residue was not observed in the
spectral range monitored. Therefore, it suggests that
the absorption should be attributed to a chemical
species. The candidate for the intermediate may be a
triplet DMAT, a biradical, or an enol, which will be
discussed later.
w
x
8,9 . However, the NMR signal resulting from an-
thranol was not observed on DMAT in deuterated
Ž
.
methanol less than the instrumental resolution . It
Ž
was found that the equilibrium lies to the left )
.
99% .
A TRTL signal of DMAT in ethanol was mea-
sured with the 308 nm irradiation to obtain the heat
of reaction. The time evolution of the TRTL signal
Ž
.
Ž
.
was composed of fast Ufast and of slow
U
slow
Fig. 1 shows transient absorption spectra of
components shown in Fig. 2. The lens signal stayed
flat for ;1 ms and decayed due to the thermal
diffusion. The slow rising component was success-
fully analyzed with the single-exponential equation.
The risetime of slow component was determined to
be 1.2"0.1 ms. This value is identical with the
lifetime for the intermediate obtained with the flash
photolysis experiment. Hence, the fast component
corresponds to the heat for the intermediate forma-
tion and the slow one to the heat released with the
intermediate reverting to the parent keto form.
DMAT in ethanol with the laser excitation of 308
nm. The spectra have three peaks at 315, 360 and
425 nm, which appears in the similar spectral region
w
x
for cis- and trans-enols of MBP 3–6 . The spectra
shown in Fig. 1 are different from the reported one
for 1,4-dimethylanthrone observed in ethanol at 77 K
w
x
10 . Our results well agree with the spectra of not
only MBP enols but also 2-methylacetophenone and
w
x
2-methylbenzaldehyde enols 11 . The time profiles
of the transient absorption monitored at 360 and 425

(
)
T. Omori et al.rChemical Physics Letters 293 1998 436–440
439
factor, the laser power dependence of the TRTL
signal intensity at 8 ms was measured at various
concentrations of DMAT, shown in Fig. 3. Good
linear relations between the lens signal intensity and
the laser power appeared in the laser power range
less than 10 mJ, revealing that the TRTL signals
were generated with one-photon absorption. The
slopes in Fig. 3 were plotted against the absorptivity,
1y10yOD, in Fig. 4. The a value was successfully
determined to be 1.02"0.02 by comparing the slope
with pyridazine, which is a calorimetric standard for
photothermal measurements. The unit value indicates
that all the energy absorbed by DMAT is released as
heat. Namely, the exited DMAT is relaxed to the
triplet state through intersystem crossing, followed
by an intramolecular hydrogen abstraction reaction
to yield the enol via the biradical state. The enol
formed, eventually returns the parent DMAT in the
ground state.
Fig. 2. The TRTL signal of DMAT in ethanol with the 308 nm
Ž
.
Ž
.
irradiation. The signal shows the fast Ufast and the slow U
slow
components. The slow component rises single-exponentially,
which was analyzed with a least-squares fitting. The risetime was
determined to be 1.2"0.1 ms, which was identical to the lifetime
of the reaction intermediate.
The ratio of the slowly rising signal intensity
The observed TRTL signal intensity is described
as
Ž
.
Ž
.
U is ex-
total
U
to the total signal intensity
slow
pressed as
UsKaIL 1y10yOD
,
1
Ž .
Ž
.
U
slowrUtotal sfisc fr ErrEex
,
2
Ž .
where K is the instrumental factor including the
thermal properties of the solvent, IL the laser energy,
OD the absorbance of the sample at 308 nm and a
denotes the heat conversion efficiency, namely, the
fraction of the energy released as heat against the
total energy absorbed. To eliminate the instrumental
where Eex is the energy of the excitation light, Er
the energy difference between the intermediate and
the parent DMAT in the ground state, and fisc and
fr the quantum yields for intersystem crossing and
Fig. 4. The plots of UrIL against of the absorptivity, 1–10yOD
;
Ž
.
Ž .
Ž .
Fig. 3. The plots of U
against the laser power IL at several
DMAT U
v , DMAT U
` and pyridazine ' . Good
total
total
slow
Ž
.
Ž
.
Ž .
concentrations of DMAT; OD 308 nm s0.610 I , 0.463 ' ,
linear relations are also observed. The a value was obtained to be
1.02"0.02 by comparing the slopes of pyridazine and DMAT
Ž
.
Ž
.
Ž
.
Ž .
0.345 \ , 0.218 l , 0.142
relations are observed. The solid lines show the best fitting.
` and 0.066 v . Good linear
U
.
total

(
)
440
T. Omori et al.rChemical Physics Letters 293 1998 436–440
the reaction, respectively. From the slopes of U
1.02 " 0.02, revealing that an excited DMAT
slow
and U
in Fig. 4, the ratio of UslowrU
was
molecule emits all the energy absorbed as heat. The
total
total
Ž
.
Ž
.
determined to be 0.31"0.01. The fisc fr Er product
was obtained as 131"6 kJrmol, using the values of
TRTL signal had the fast Ufast and the slow U
slow
components. The risetime of slow component was
determined to be 1.2"0.1 ms, which was identical
with the lifetime for the intermediate obtained with
Ž
.
0.31 and Eex 384 kJrmol . The fisc value was
estimated to be unity because the rate for the triplet
formation of anthrone was reported to be quite large
the flash photolysis experiment. The UslowrU
ra-
total
10
;10 sy1 12–14 . The fr value was also re-
Ž
. w
x
tio has given the reaction heat for the intramolecular
hydrogen abstraction, Er s131"6 kJrmol. The
DMAT enol is probably the most appropriate species
for the observed intermediate. The energy of the
DMAT enol seemed to be similar to the cis-enol of
MBP.
garded as unity since the hydrogen abstraction rate
constant for anthrone is reported to be as large as
w x
benzophenone 9 and the abstraction from the neigh-
boring methyl group in DMAT is thought to be the
w x
same as in the case of MBP 6 . Therefore, the
reaction enthalpy change Er was determined to be
131"6 kJrmol.
Acknowledgements
Here, we may consider a candidate for the inter-
mediate of DMAT. The intermediate should not be
We are indebted to Professor Y. Fujimoto, Mr. N.
Ž
the triplet DMAT, because the lifetime 1.2"0.1
Ž
.
Hara Tokyo Institute of Technology for their help
with the sample synthesis.
.
ms is much longer than the reported triplet lifetime
170 ns for anthrone. In addition, the triplet energy
Ž
.
w
x
301 kJrmol 15 for anthrone, is much larger than
the value of 131"6 kJrmol described above. The
candidate for the intermediate may be the biradical
References
Ž
.
or the enol form Scheme 1 . The relative energy of
the DMAT biradical to the ground keto form will be
larger than 131"6 kJrmol, since these values for
w x
1
N.J. Turro, Modern Molecular Photochemistry, Benjamin
Cummings, Menlo Park, CA, 1978.
w x
2
Ž
.
P.G. Sammes, Tetrahedron 32 1976 405, and references
therein.
similar biradicals of 2-methylacetophenone and 2-
w x
Ž
.
3
P.K. Das, J.C. Scaiano, J. Photochem. 12 1980 85.
1
w
x
methylbenzaldehyde are over 200 kJrmol
11 .
w x
4
P.K. Das, M.V. Encinas, R.D. Small, J.C. Scaiano, J. Am.
Accordingly, the DMAT enol is the most appropriate
species for the intermediate. The energy of the inter-
mediate is similar to the cis-enol of MBP, Ecis s116
Ž
.
Chem. Soc. 101 1979 6965.
w x
5
T. Nakayama, K. Hamanoue, T. Hidaka, M. Okamoto, H.
Ž
.
Teranishi, J. Photochem. 24 1984 71.
w x
6
T. Suzuki, U. Okuyama, T. Ichimura, Chem. Phys. Lett. 266
w x
kJrmol 6 . The fact also suggests that the interme-
Ž
.
1997 107.
diate should be the enol form. In order to further
elucidate the reaction mechanism of the excited
DMAT, we are carrying out ultrafast flash photolysis
and ESR measurements.
w x
7
T. Suzuki, U. Okuyama, T. Ichimura, J. Phys. Chem. A 101
Ž
.
1997 7047.
w x
8
J.C. Netto-Ferreira, D. Weir, J.C. Scaiano, J. Photochem.
Ž
.
Photobiol. A: Chem. 48 1989 345.
w x
9
R.W. Redmond, J.C. Scaiano, J. Photochem. Photobiol. A:
Ž
.
Chem. 49 1989 203.
w
w
w
w
x
10 M.A. Garcia-Garibay, A. Gamarnik, R. Bise, L. Pang, W.S.
4. Summary
Ž
.
Jenks, J. Am. Chem. Soc. 117 1995 10264.
x
11 T. Suzuki, U. Okuyama, T. Omori, T. Ichimura, in prepara-
tion.
We measured transient absorption spectra and
thermal lens of DMAT in ethanol. The heat conver-
sion efficiency a was successfully obtained to be
x
12 D.E. Damschen, C.D. Merrit, D.L. Perry, G.W. Scott, L.D.
Ž
.
Talley, J. Phys. Chem. 82 1978 2268.
x
Ž
.
13 T. Kobayashi, S. Nagakura, Chem. Phys. Lett. 42 1976
429.
w
w
x
Ž
.
14 G.W. Scott, L.D. Talley, Chem. Phys. Lett. 53 1977 431.
x
1 The energy for the biradicals is estimated to be in the range of
200–220 kJrmol.
15 S.L. Murov, I. Carmichael, G.L. Hug, Handbook of Photo-
chemistry, 2nd ed., Marcel Dekker, New York, 1993.