T. Yuzawa, H. Hamaguchi / Journal of Molecular Structure 976 (2010) 414–418
415
surface to the trans. The possibility of this triplet quantum chain
process has been pointed out from the concentration dependence
of the quantum yield of photoisomerization [4,13,14]. In the pres-
ent paper, we use time-resolved infrared spectroscopy to directly
monitor the triplet quantum chain process in the photoisomeriza-
tion of 9-cis retinal.
2. Experimental
The experimental arrangements for AC-coupled nanosecond
time-resolved infrared spectroscopy with a dispersive spectrome-
ter (JASCO TRIR-1000) have been described in detail in previous pa-
pers [15–17]. A photovoltaic MCT detector (Kolmar Technologies,
Inc. KV103-1-A-1-SMA) was used for fast infrared detection. The
time resolution was about 50 ns. Signals from the MCT detector
are first amplified with a preamplifier and then averaged on a dig-
ital oscilloscope (Tektronix DSA602). A sampling rate of data acqui-
sition was every 40 ns for the experiments described here. The
third harmonic of a cw Q-switched Nd:YLF Laser (Spectra Physics
TFR, 5 ns pulse width, 190 Hz repetition rate, 30 mJ at 349 nm)
was used for photoexcitation. Infrared absorption spectra were ob-
tained at intervals of 4 cmꢀ1 with a spectral resolution of 16 cmꢀ1
.
A cyclohexane solution of retinal was circulated with peristaltic
pump to eliminate the influence of the photoisomerization prod-
ucts. A flowing sample cell with two BaF2 windows was used. Solu-
tions were bubbled with argon gas during the measurements in
order to eliminate the effect of oxygen. Samples of all-trans-retinal
and 9-cis-retinal were purchased from Sigma and used as received.
Cyclohexane was purchased from Wako Pure Chemical Industry,
Ltd. and used without further purification. All measurements were
carried out at room temperature.
Fig. 1. Time-resolved infrared difference spectra of photoexcited 9-cis retinal in
cyclohexane. (a) 0 ns–0.4
(e) 1.6 s–2 s, (f) 2 s–4
12 s, (k) 12 s–14 s, (l) 14
l
l
s, (b) 0.4
s, (g) 4
s–16
l
s–0.8
s–6 s, (h) 6
s and (m) 16
l
s, (c) 0.8
s–8
s–18
l
l
s–1.2
s, (i) 8
s.
l
l
s, (d) 1.2
l
s–1.6
l
s–
s,
l
l
l
l
l
l
s–10 s, (j) 10 l
l
l
l
l
l
l
l
l
3. Results and discussion
resolve the fast rise (880 ps) of the all-trans triplet state from the
9-cis.
3.1. Time-resolved infrared difference spectra
The negative band at 1144 cmꢀ1 (Fig. 2a) represents the deple-
tion of 9-cis retinal in the ground state. Neither all-trans T1 nor all-
trans S0 show a band at 1144 cmꢀ1. The temporal profiles of this
band observed for three different concentrations are shown in
Fig. 3. The signal can be separated into two temporal components.
One is the instantaneous decrease of the signal (denoted as I1 in the
trace Fig. 3a), which is ascribed to the instantaneous depletion of 9-
cis S0 by the photoexcitation. The other is denoted as I2 in the trace
Fig. 3b, which is ascribed to the delayed depletion of 9-cis S0 in the
microsecond time regime. The second component suggests that
there is a pathway that consumes the ground state of 9-cis retinal
in the microsecond regime. The contribution of the I2 component
to the whole amount of the 9-cis S0 depletion depends on the con-
centration of retinal. At a high concentration (Fig. 3a, 3.2 mM), the
amplitude of I2 is comparable to that of I1. At a low concentration
(Fig. 3c, 0.28 mM), the amplitude of I2 becomes negligible. The
Time-resolved infrared difference spectra of photoexcited 9-cis
retinal in cyclohexane (3.5 mM) are shown in Fig. 1. These differ-
ence spectra, which correspond to the light-on minus light-off sig-
nal, were obtained directly with the AC-coupled method [15–17].
Spectra were averaged over the delay time ranges shown in the
caption. The wavenumber regions around 1450 cmꢀ1 and
1030 cmꢀ1 were blocked by the strong absorption bands of cyclo-
hexane. Negative peaks at 1670, 1588, 964 cmꢀ1 show the deple-
tion of 9-cis retinal in the ground state by the photoexcitation.
Positive peaks at 1602, 1552, 1184 and 940 cmꢀ1 represent the
production of transient species. After the delay of 18
the observed difference spectrum shows no further temporal
change. The difference spectrum after 18 s is identical with the
ls (Fig. 1m)
l
difference spectrum between the all-trans isomer and the 9-cis iso-
mer in the ground state. The negative peaks at 1592, 1144, and
960 cmꢀ1 are ascribed to 9-cis S0. The positive peaks at 1570,
1164, 1130, and 972 cmꢀ1 are ascribed to all-trans S0. It is thus
indicated that the photoexcitation of 9-cis retinal in cyclohexane
solution causes the photoisomerization to the all-trans isomer.
Temporal profiles of the signal observed at four different wave-
numbers are shown in Fig. 2. The best fit of those curves with the
single exponential function convoluted with the instrumental re-
sponse are shown as solid lines in Fig. 2. A global fitting analysis
intensity of the 1144 cmꢀ1 band decreases at the rate of 5
ls,
which is the same as that of the decay of all-trans T1. This fact indi-
cates that an interaction between the 9-cis S0 and all-trans T1 is in-
volved in the photoprocess.
3.2. SVD analysis
of these curves gives a single time constant of 5
l
s. The positive
The time-resolved infrared difference spectra shown in Fig. 1
have been analyzed with the singular value decomposition (SVD)
method. The decomposed spectral and temporal components ob-
tained are shown in Fig. 4 for the largest four singular values,
which are 0.00952, 0.00463, 0.00170, 0.00166. It is obvious from
peaks at 1184 cmꢀ1 (Fig. 2c) and 1600 cmꢀ1 (Fig. 2d) are assigned
to the all-trans triplet state [10]. They show instantaneous rise and
along with decay with the time constant of 5
ls. The response of
the present time-resolved infrared apparatus is not sufficient to
Yuzawa, Tetsuro
