Primary processes of the electronic excited states of trans-urocanic acid

Article abstract of DOI:10.1021/jp962679s

The primary photoreactivity of the excited states of trans-urocanic acid (t-UA) is investigated by ultrafast transient-absorption spectroscopy. Fundamentally different photophysics were observed when t-UA is excited at 266 nm, near the peak of the absorption spectrum, and 306 nm, in the red tail of the absorption spectrum. The data support the conclusion that the wavelength-dependent photophysics of t-UA is due to the presence of two different closely spaced electronic states. Excitation at 266 nm populates a ππ^* state that is localized on the imidazole ring. The transient data following photoexcitation of t-UA at 266 nm in both a pH 5.6 and a pH 7.2 solution are similar, even though the protonation state of the tertiary nitrogen on the imidazole ring is different at these two pH values. The data therefore support that the photophysics at pH 5.6 and pH 7.2 must involve a common excited state. Steady-state excitation spectra suggest that a proton transfer process from t-UA to the solvent occurs following the excitation at 266 nm at pH 5.6, which generates an electronically excited singlet state of the deprotonated molecule. This state is directly accessed by the 266 nm excitation of t-UA at pH 7.2. The population in this singlet state decays by intersystem crossing with a rate constant of 1.4 × 10¹¹ s^-1. Isomerization is not believed to occur from this triplet state. Excitation of t-UA at 306 nm populates an entirely different state, which leads to isomerization. From the observed ground state repopulation dynamics, the minimum rate for the excited state isomerization is 1.2 × 10¹⁰ s^-1.

Full text of DOI:10.1021/jp962679s

© Copyright 1997 by the American Chemical Society
VOLUME 101, NUMBER 6, FEBRUARY 6, 1997
LETTERS
Primary Processes of the Electronic Excited States of trans-Urocanic Acid
Bulang Li, Kerry M. Hanson, and John D. Simon*
Department of Chemistry and Biochemistry, UniVersity of California at San Diego, 9500 Gilman DriVe,
La Jolla, California 92093-0341
ReceiVed: August 30, 1996; In Final Form: NoVember 6, 1996X
The primary photoreactivity of the excited states of trans-urocanic acid (t-UA) is investigated by ultrafast
transient-absorption spectroscopy. Fundamentally different photophysics were observed when t-UA is excited
at 266 nm, near the peak of the absorption spectrum, and 306 nm, in the red tail of the absorption spectrum.
The data support the conclusion that the wavelength-dependent photophysics of t-UA is due to the presence
of two different closely spaced electronic states. Excitation at 266 nm populates a ππ* state that is localized
on the imidazole ring. The transient data following photoexcitation of t-UA at 266 nm in both a pH 5.6 and
a pH 7.2 solution are similar, even though the protonation state of the tertiary nitrogen on the imidazole ring
is different at these two pH values. The data therefore support that the photophysics at pH 5.6 and pH 7.2
must involve a common excited state. Steady-state excitation spectra suggest that a proton transfer process
from t-UA to the solvent occurs following the excitation at 266 nm at pH 5.6, which generates an electronically
excited singlet state of the deprotonated molecule. This state is directly accessed by the 266 nm excitation
of t-UA at pH 7.2. The population in this singlet state decays by intersystem crossing with a rate constant
of 1.4 × 1011 s-1. Isomerization is not believed to occur from this triplet state. Excitation of t-UA at 306
nm populates an entirely different state, which leads to isomerization. From the observed ground state
repopulation dynamics, the minimum rate for the excited state isomerization is 1.2 × 1010 s-1.
1. Introduction
Urocanic acid is a metabolizing product of histidine and
represents about 0.7% of the dry weight of the epidermis where
it can absorb both UV-A and UV-B radiation. It is synthesized
in the uppermost layer of the skin as its trans isomer,4 where,
upon exposure to UV radiation, trans-UA can isomerize to cis-
UA (Scheme 1). A photostationary state with approximately
equal amount of both isomers5 can be achieved. Because both
trans- and cis-UA absorb in the same wavelength region as
DNA, urocanic acid was initially ascribed to behave as a natural
sunscreen, protecting DNA from photodamage. As a result, it
has been used as an ingredient in suntan lotions. But in 1983,
De Fabo and Noonan6 proposed that UA could act as a mediator
for UV-induced immunosuppression. Currently, research in-
dicates that the isomerization of trans-UA to, and the subsequent
The potential health risks associated with solar UV exposure
have become major concerns worldwide because of recent
reports of reduced levels of atmospheric ozone1 and increased
incidence of skin cancers.2 UV radiation can cause sunburn,
photoaging, DNA damage, immunosuppression, and skin cancer.
These adverse effects of UV radiation are initiated by the
photochemistry of DNA and various other chromophores. For
example, one of the major causes of skin cancer is attributed to
photodimerization of thymine in the P53 tumor suppresser gene.3
Recently, another epidermal UV chromophore, urocanic acid
(UA) (refs 4-10 and references therein), has been implicated
as a mediator for UV-induced immunosuppression, as well.
X Abstract published in AdVance ACS Abstracts, January 15, 1997.
S1089-5639(96)02679-5 CCC: $14.00 © 1997 American Chemical Society

970 J. Phys. Chem. A, Vol. 101, No. 6, 1997
Letters
SCHEME 1: Isomerization of trans-Urocanic Acida
a The structure drawn is that present in solution at pH 5.6 (the average
pH of human sweat). The pKa of the tertiary nitrogen on the imidazole
ring is 5.8, and therefore this site is deprotonated in a pH 7.2 solution.
accumulation of, cis-UA leads to immunosuppression and
potentially cutaneous carcinogenesis.7 Thus, urocanic acid’s use
in cosmetic products has been withdrawn since the early
nineties.8
The photochemistry of trans-UA has turned out to be
complex. For example, Morrison and co-workers reported that
while the absorption of trans-UA peaks around 270 nm, the
quantum efficiency of the trans to cis isomerization peaks at
310 nm, in the low-energy tail of the absorption spectrum.
Studies show that the quantum yield for isomerization is strongly
wavelength dependent, ranging from about 0.05 at 266 nm to
0.5 at 310 nm.9 To our knowledge, there are no experimental
studies that directly probe the excited state dynamics of urocanic
acid. Our current knowledge of UA’s photophysics is provided
by steady-state absorption and emission studies,10 quantum yield
measurements for isomerization,9 and triplet-sensitization ex-
periments.11 In this paper, we examine UA’s excited state
dynamics using femtosecond and picosecond transient-absorp-
tion spectroscopy. Transient-absorption experiments were
conducted as a function of excitation and probe wavelength and
the pH of the solution. The results support the conclusion that
the wavelength-dependent photoreactivity of trans-UA near 266
and 310 nm is due to the presence of at least two distinct and
weakly coupled electronic transitions.
Figure 1. Absorption spectra of t-UA in pH 5.6 and 7.2 buffer solution.
The excitation wavelengths for transient absorption experiments are
marked. The tertiary nitrogen on the imidazole group is protonated at
pH 5.6 and deprotonated at pH 7.2.
Experimental Section
Transient-absorption experiments requiring subpicosecond
time resolution were carried out with a home-built amplified
Ti:sapphire laser system with a time resolution of approximately
200 fs. This output of the laser beam was frequency doubled
and then tripled using two BBO crystals to obtain the pump
wavelength of 266 nm. The pump pulse energy at 266 nm was
2-5 µJ. The residual of the second harmonic output was
focused into a water cell to generate a continuum, and various
10 nm band-pass filters were used to select a probe wavelength
until a signal was found (340 nm). For the 266 nm degenerate
experiments, a glass slide was used to select approximately 5%
of the tripled light for the probe beam. The degenerate 306
nm pump-probe experiments were accomplished using a home-
built optical parametric amplifier system operating at 1 kHz12
with a time resolution of 50 fs. The 306 nm light was produced
from frequency doubling the amplified 1200 nm OPA output,
which was then frequency doubled to generate approximately
1 µJ 306 nm pump pulses. Approximately 5% of this output
was used as the probe source. Experiments requiring delays
between 1 and 10 ns were carried out using a home-built Nd:
YLF regenerative amplification system that has a time resolution
of 10 ps. In addition to generating light at the various
harmonics, this laser is used to pump a home-built OPG/OPA
systems that provides tunable pulses from 300 nm to above 2
µm. Transients acquired on either laser system were found to
be independent of the polarization of the pump and probe beams.
t-UA was obtained from Aldrich and used without further
purification. The solutions were buffered using HPLC grade
water. A pH of 5.6 was maintained by 0.1 M acetate buffer. A
Figure 2. Transient absorption data at 266 nm excitation: (top) in
pH 5.6 solution; (bottom) in pH 7.2 solution. Probe wavelength is at
340 nm (a and c) and at 266 nm (b and d).
pH of 7.2 was maintained by 0.1 M potassium phosphate buffer.
The concentration of t-UA was on the order of 70 µM, and the
solutions were flowed through a 1 mm quartz flow cell in order
to prevent artifactual signals that could arise from the photo-
degradation of the sample. The sample was changed after every
data accumulation. A typical run required 20 min, and in this
time period, only a negligible amount of t-UA is converted to
other photochemical products.
Results
Figure 1 shows the absorption spectra of t-UA in pH 5.6 and
pH 7.2 solutions. The excitation wavelengths used in this study
are indicated in the figure. Figure 2a-d presents time-resolved
bleaching and absorption data for t-UA in both pH 5.6 (2a,b)
and pH 7.2 (2c,d) buffer solutions. The excitation wavelength
for all four data plots is 266 nm. This wavelength is near the
absorption maximum of t-UA in the pH 5.6 solution and is on
the blue side of the absorption maximum (λmax ) 280 nm) of
t-UA in the pH 7.2 solution (Figure 1). The probe wavelengths
are 266 nm (2b,d) and 340 nm (2a,c). The data observed for
the two different pHs are similar. A longer time study (data
not shown) of the pump-probe experiment at 266 nm indicates
no change in the transient signal up to a delay of 8 ns, the longest

Letters
J. Phys. Chem. A, Vol. 101, No. 6, 1997 971
whereas the absorption maximum is highly dependent upon
pHsat pH 5.6 the absorption maximum peaks at 266 nm, and
at pH 7.2 the absorption peaks at 280 nm (Figure 1). The
excitation spectrum for either solution, however, peaks at 280
nm. Therefore, we can conclude that the excited state reached
upon excitation at 266 nm is deprotonated on the tertiary
nitrogen regardless of pH. Furthermore, the isomerization study
reported by Morrison and co-workers shows little pH depen-
dence,9 further supporting the generation of a common excited
state at the two pH values. The common excited state reached
upon excitation at 266 nm for pH 5.6 and pH 7.2 must result
from a change in the excited state pKa of the tertiary nitrogen
on the imidazole. The absorption transition must then decrease
the electron density on the imidazole ring, making the nitrogen
more acidic. Similar phenomena have been observed in a
variety of compounds such as aryl alcohols, which show
significant decrease of pKa in their excited electronic state
compared to the ground state. For example, 2-naphthol18 is a
weak acid in the ground state (pKa ) 9.46) but is a much
stronger acid upon excitation to its lowest excited singlet state
(pK*a ) 2.80). These observations along with UA’s strong
oscillator strength near 266 nm (104) supports the conclusion
that this transition arises most likely from a π f π* transition
within the imidazole ring at either pH.
Figure 3. Transient absorption data of t-UA in pH 5.6 solution
observed for the degenerate 306 nm pump/probe experiment.
delay time studied. Figure 3 shows the transient dynamics
observed for a degenerate 306 nm pump-probe experiment of
t-UA at pH 7.2. In this case, the initial bleach of the absorption
shows a complete recovery on the picosecond time scale.
Discussion
The data shown in Figure 2 are therefore consistent with the
following argument. The dominant contribution to the dynamics
observed at the probe wavelength of 340 is the Sx f SN
absorption of t-UA (with the nitrogen of the imidazole ring
deprotonated). (We use Sx instead of S1 to describe the lower
energy state of this transient because excitation at 306 nm at
both pH 5.6 and pH 7.2 accesses a lower energy state, which is
likely to be a singlet. The dynamics then show that intersystem
crossing occurs more rapidly than internal conversion meaning
that these two low-energy excited states are weakly coupled.)
As stated above, the isomerization quantum yield of t-UA is
wavelength dependent. In addition, the emission spectra are
wavelength dependent,10 and the photoacoustic calorimetry
signals also depend on the excitation wavelength.13 We consider
two possible explanations for this wavelength-dependent pho-
toreactivity. First, the broad and structureless UV spectrum
could be due to the presence of multiple ground state rotamers
that have different absorption spectra and different excited state
reactivities. Such a model has been used to describe the
wavelength-dependent behavior of bilirubin,14 diphenylbutadi-
ene,15 and dihexatriene.16 Second, the absorption spectrum of
t-UA could reflect the superposition of two separate absorption
transitions. The two excited states accessed would then have
different reactivities, and the photophysics from the higher
energy state would occur on a time scale short compared to
internal conversion. Such a model has been used to describe
the behavior of cinnamic acid.17 This molecule is similar in
structure to urocanic acid, with the imidazole replaced by a
benzyl group, and has two electronic transitions: a ππ*
dominating the absorption spectrum near the maximum at 280
nm and a weak nπ* transition near the red tail of the absorption
profile (>300 nm). The transient absorption data suggest that
urocanic acid’s photochemistry can be described by a model
that is similar to that used to describe the wavelength-dependent
behavior of cinnamic acid, namely, that the UV absorption band
is comprised of overlapping and weakly coupled electronic
transitions.
For the photoexcitation of t-UA at 266 nm, similar transient
absorption dynamics are observed for both pH 5.6 and 7.2
solutions (Figure 2), suggesting that a common excited state of
t-UA is generated in both solutions. Note though that the ground
state configurations do differ at these two pHs values. At pH
5.6 the tertiary nitrogen on the imidazole ring is protonated,
and at pH 7.2 this site is deprotonated. We propose that at pH
5.6 t-UA undergoes rapid (<200 fs) deprotonation in the excited
state to generate an electronically excited molecule, which is
similar to that created by the direct excitation of t-UA at pH
7.2. This proposed reaction is supported by the fluorescence
excitation study reported by Shukla and Mishra10 and confirmed
by our laboratory. In these studies, the excitation maximum is
found to be independent of the pH (pH 3.2-11) of the solution,
The decay in the transient signal reflects the subsequent
intersystem crossing process, which can be fit to a first-order
kinetic process with a rate constant of 1.4 × 1011 s-1. For a
probe wavelength of 266 nm, an initial bleaching of the ground
state population is observed, followed by a recovery of the signal
to about half of its negative-time value. The recovery dynamics
observed in the degenerate 266 nm experiment could, in
principle, arise from either the repopulation of the ground state
or the population of a long-lived transient absorbing at the same
wavelength as the pump wavelength. A triplet state has been
shown to exist by Morrison’s triplet sensitization11 studies, and
the lowest-energy triplet state is estimated to lie approximately
250 kJ/mol above the ground state. We have confirmed this
value by photoacoustic experiments.13 The time scale of the
rise at 266 nm is the same as that of the decay of the data
obtained at 340 nm. Such evidence when combined supports
the argument that the short-lived transient at 340 nm and the
long-lived transient absorbing at 266 nm result from rapid
intersystem crossing to a long-lived electronically excited triplet
state with Tx f Tn absorption at 266 nm.
The 266 nm pump/340 nm probe transient absorption data
are slightly different at the two pH values studied. Both contain
a 7 ps component, and yet we observe an initial fast decay
component (∼1 ps) at pH 7.2 which is absent at pH 5.6. We
tentatively assign the fast component observed at pH 7.2 to a
vibrational relaxation process. At pH 7.2, the electronic excited
state of t-UA lies lower in energy than it does at pH 5.6. This
is evidenced by the absorption spectrum of t-UA, which red
shifts when the pH of the solution is raised (see Figure 1).
Therefore, excitation at 266 nm generates a molecule with more
excess energy at pH 7.2 than at pH 5.6. As a result, pronounced

972 J. Phys. Chem. A, Vol. 101, No. 6, 1997
Letters
vibrational relaxation dynamics appear. At pH 5.6, 266 nm
excitation still produces a vibrationally hot (but to a lesser
degree) molecule. However, the ultrafast deprotonation reaction
that occurs at this pH provides the molecule with an addition
channel by which it can dissipate the excess vibrational energy.
Figure 3 displays the transient absorption data for a degenerate
pump-probe experiment at 306 nm. Following a bleach at t
) 0, a full recovery of the signal is observed. The recovery
dynamics can be described by a first-order kinetic process with
a rate constant of 1.2 × 1010 s-1, which is approximately 10
times greater than the intersystem-crossing rate constant found
when pumping at 266 nm and probing at 340 or 266 nm. The
dramatic difference in the two rate constants resulting from the
two different pump wavelengths is indicative of the wavelength-
dependent photoreactivity exhibited by urocanic acid (recall that
the isomerization of trans- to cis-UA is dependent upon
excitation wavelength). Morrison determined an isomerization
quantum yield of nearly 0.5 at 310 nm but only 0.05 near 266
nm.9 It is reasonable then to ascribe the dynamics at 306 nm
to be unique from those found at 266 nm, and therefore we
attribute the dynamics at 306 nm to result from ground state
repopulation, with no intersystem crossing occurring upon
excitation of t-UA at this wavelength. Because cis- and trans-
UA absorb similarly at 306 nm, the dynamics cannot be used
to determine whether or not isomerization occurs. The isomer-
ization quantum yield of 0.5 at approximately 306 nm places a
lower limit on the excited state isomerization rate constant of
1.2 × 1010 s-1. Further study is required in order to directly
monitor the excited state isomerization process and assess
whether or not there is an energy barrier to isomerization along
the excited state isomerization reaction coordinate. We are
currently constructing a tunable UV source that is to be pumped
by our femtosecond Ti-sapphire system, and such studies will
be reported at a later date.
Acknowledgment. We thank Chris Bardeen and Vladislav
Yakovlev for their help with the 306 nm experiments in the
laboratory of Professor Kent Wilson. We thank Peijun Cong
for technical assistance. This work is supported by the National
Institutes of Health.
References and Notes
(1) Kerr, J. B.; McElroy, C. T. Science 1993, 262, 1032.
(2) Ko, C. B.; Walton, S.; Keczkes, K.; Bury, H. P. R.; Nicholson, C.
Br. J. Dermatol. 1994, 130, 269.
(3) Ziegler, A.; Jonason, S.; Leffell, D. M.; Simon, J. A.; Sharma, H.
W.; Kimmelman, J.; Remington, L.; Jacks, T.; Brash, D. E. Nature 1994,
372, 773.
(4) Scott, I. R. Biochem. J. 194, 829.
(5) Gibbs, N. K.; Norval, M.; Traynor, N. J.; Wolf, M.; Johnson, B.
E.; Crosby, J. Photochem. Photobiol. 1993, 57, 584.
(6) De Fabo, E. C.; Noonan, F. P. J. Exp. Med. 1983, 157, 84-98.
(7) Matheson, M. J.; Reeve, V. E. Photochem. Photobiol. 1991, 53,
639.
(8) Review by the Cosmetic Ingredient Expert Panel. J. Am. Coll.
Toxicol. 1995, 14, 386.
(9) Morrison, H.; Bernasconi, C.; Pandey, G. Photchem. Photobiol.
1984, 40, 549.
(10) Shukla, M. K.; Mishra, P. C. Spectrochim. Acta 1995, 51A, 831.
(11) Morrison, H.; Bernasconi, A. C.; Fagan, G. Photochem. Photobiol.
1980, 32, 711.
(12) Yakovlev, V. V., Kohler, B., Wilson, K. R. Opt. Lett. 1994, 19,
2000.
(13) Hanson, K. M.; Li, B.; Simon, J. D. J. Am. Chem. Soc., in press.
(14) Braslavsky, S. E.; Holzwarth, A. R.; Schaffner, K. Angew. Chem.,
Int. Ed. Engl. 1983, 22, 656.
(15) Wallace-Williams, S. E.; Moller, S.; Goldbeck, R. A.; Hanson, K.
M.; Lewis, J.W.; Yee, W. A.; Kliger, D. S. J. Phys. Chem. 1993, 97, 9587.
(16) Saltiel, J.; Sears, D. F.; Sun, Y.-P.; Choi, J.-O. J. Am. Chem. Soc.
1992, 114, 3607.
(17) Ullman, E. F.; Babad, E.; Singh, P. J. Am. Chem. Soc. 1969, 91,
5792.
(18) Lawrence, M.; Marzzacco, C. J.; Morton, C.; Schwab, C.; Halpern,
A. M. J. Phys. Chem. 1991, 95, 10294.