Photoisomerization of trans-Urocanic Acid in a Supersonic Jet
J. Am. Chem. Soc., Vol. 123, No. 5, 2001 965
Table 1. Vibronic Features in Excitation (S1) and Emission (S0) Spectra of trans-UA, Compared to a Published Imidazole Infrared Spectrum
published imidazole
a
S0 (cm-1
226
)
IR spectrum11 (cm-1
)
S1 (cm-1
)
assignments
ethylene bend
237
506
606
736
538/551
636/631
735/728
476, 514 (Y)
620
719
1066
1281
1376
1576
1056/1074/1120/1130
924/954/987
imidazole normal
modes
1252
1325
1518
1214/1180 (Z)
1338/1366 (X)
1519
1786
1876
1694 (Z + Y)
1814/1842 (X + Y)
combination bands?
other propionic acid vibrations?
a The letters X, Y, and Z in parentheses represent possible fundamentals contributing to combination bands, e.g. X + Y.
between 1600 and 1800 cm-1, but the imidazole infrared
spectrum does not contain any peaks between 1600 and 3000
cm-1. These features may be combination bands of some of
the imidazole fundamental vibrations or vibrations arising from
the rest of the molecule, like the carbonyl stretch. A few peaks
in the excitation spectrum, labeled X and Z in the table, appear
to be combination bands with the +476 cm-1 mode, labeled Y.
The emission spectra resulting from excitation of the vibra-
tional peaks in this region are very similar and resemble the
emission spectrum obtained by exciting the origin. At higher
energies, the spectra become broadened, but retain similar
features. This is likely a consequence of intramolecular vibra-
tional relaxation (IVR). The molecule is initially excited to a
Franck-Condon allowed bright state. As the excitation energy
is increased, the density of dark states originating from other
vibrational modes of the molecule increases, which may mix
with the bright state via anharmonic couplings. Each of these
dark states has large Franck-Condon factors with corresponding
ground-state levels, i.e., large ∆V ) 0 Franck-Condon factors.
The many overlapping transitions with slightly different fre-
quencies result in a broadened emission spectrum. The IVR
broadening of the emission spectra occurs around +1000 cm-1
above the origin, when the first two features in the emission
spectrum appear to merge together. However, the unrelaxed
vibrational peaks at +620 and +1180 cm-1 suggest that IVR
in this region is mode selective.
Region II. The second region in the fluorescence excitation
spectrum involves an abrupt appearance of features with
increasing width and decreasing intensity as the excitation
energy increases. We will discuss two arguments as to why we
believe that we are observing trans to cis isomerization in this
region.
Initially, the broad excitation features alternate with sharp
features, then at about 1600 cm-1 all features become broad.
Excitation of these broad features results in dual emission which
contains the standard IVR broadened trans-UA emission with
some blue-shifted components. Dual emission signifies coupling
to another electronic state, which we believe is the cis-UA S1
state. The blue-shifted emission would then be fluorescence of
the cis-UA S1 state back to the cis-UA ground state. The
presence of some sharp peaks initially means that at the onset
of level mixing of the trans-UA S1 with the cis-UA S1, the
density of states of acceptor levels has not reached the statistical
limit, so accidental degeneracies are observed.
energy increased. Quantitative studies were performed to
measure the lifetimes of each vibrational peak, and the conclu-
sion was that the lines broadened due to competition of the
radiationless isomerization channel with radiative decay. Fur-
thermore, the onset of IVR occurred a few hundred cm-1 before
isomerization, which led to the conclusion that the isomerization
was not mode selective.
The assumption that isomerization is occurring in Region II
allows us to make a few comments on the isomerization process
in urocanic acid. The onset of the broad features exhibiting dual
emission occurs around 1400 cm-1, or 4 kcal/mol. We ap-
proximate this as the isomerization barrier. Emission spectra
exhibiting IVR broadened features occurs around 1100 cm-1
,
which means that like stilbene, isomerization of trans-UA is
not mode selective. The bluest emission peak of the cis-UA
emission occurs at 33 800 cm-1. This is likely the cis-UA origin
transition which is about 1200 cm-1 higher in energy than the
trans-UA origin. Several calculations have shown that the cis-
UA ground state is more stable than the trans-UA ground state
because of the intramolecular hydrogen bond that may form
between the ring imidazole and the carbonyl oxygen. The
calculation of Page et al.7 assigns a difference of 3 kcal/mol, or
about 1000 cm-1, between the cis-UA and trans-UA ground
states. If this is the case, the S1 state will be about 200 cm-1
higher in energy than the trans S1 state, which means that the
cis-to-trans isomerization barrier is a few hundred cm-1 lower
than the trans-to-cis barrier.
Region III. The third region is comprised of the increasingly
broadened peaks from Region II, with a growing in of a broad,
structureless feature that covers about 3000 cm-1. The broad
structure is likely due to excitation of the S2 electronic state of
trans-UA. The emission spectrum resulting from excitation of
this state is S1 emission minus the blue-shifted components
characteristic of cis-UA emission, which seems to indicate that
isomerization is occurring to a significantly lesser extent in
Region III than in Region II. According to Kasha’s rule, S2
emission rarely occurs, because the density of states available
at higher electronic states is great enough that some other
radiationless process will occur much faster than fluorescence.
The most likely radiationless processes are internal conversion
(IC) to the S1 state and intersystem crossing (ISC) to a triplet
state.
We do not believe we are observing ISC because the emission
is shifted at most a hundred cm-1 from S1 emission. The triplet
state would have to lie coincidentally close to the S1 state for
ISC to be occurring. Furthermore, we attempted to measure a
lifetime for this state but it was shorter than our instrument
response of 10-20 ns, which is too short to be a triplet state.
Ruling out ISC would lead us to believe that we are observing
IC to the S1 state. The problem with this idea is that if S2 is
The second argument we have for the occurrence of isomer-
ization in this region is the similarity of this portion of the
spectrum to the isomerization region of trans-stilbene.8,9 In the
frequency domain, the excitation spectrum of stilbene showed
an abrupt broadening of peaks at the isomerization barrier that
increased in width and decreased in intensity as the excitation