Excited-state double-proton transfer on 3-methyl-7-azaindole in a single crystal: Deuterium isotope/tunneling effect

Article abstract of DOI:10.1021/jp020411x

Unlike 7-azaindole consisting of the tetrameric configuration, 3-methyl-7-azaindole (3MAI) exists solely as intact double hydrogen-bonded dimeric forms in a single crystal. Both steady-state and time-resolved measurements down to 8.0 K reveal remarkable deuterium isotope effects on the rate of excited-state double proton transfer (ESDPT) in the N(1)-deuterated 3MAI (3MAI-d) single crystal. The rates of ESDPT for the 3MAI-d dimer resolved at < 150 K are mainly governed by the proton tunneling mechanism. At < 12 K, the nearly temperature-independent ESDPT dynamics lead us to qualitatively deduce a barrier height of ～ 1.73 kcal/mol for the 3MAI-d dimer. The results provide an ideal model to investigate the intrinsic ESDPT dynamics for 7-azaindole analogues in which the structural information is well documented.

Full text of DOI:10.1021/jp020411x

8006
J. Phys. Chem. A 2002, 106, 8006-8012
ARTICLES
Excited-State Double-Proton Transfer on 3-Methyl-7-azaindole in a Single Crystal:
Deuterium Isotope/Tunneling Effect
Wei-Shan Yu,† Chung-Chih Cheng,‡ Chen-Pin Chang,‡ Guo-Ray Wu,† Chin-Hao Hsu,† and
Pi-Tai Chou*,†,§
Department of Chemistry, National Chung-Cheng UniVersity, Chia-Yi, Taiwan, Republic of China, Department
of Chemistry, Fu Jen Catholic UniVersity, Shin Chuang, Taiwan, Republic of China, and Department of
Chemistry, National Taiwan UniVersity, Taipei, Taiwan, Republic of China
ReceiVed: February 12, 2002; In Final Form: June 8, 2002
Unlike 7-azaindole consisting of the tetrameric configuration, 3-methyl-7-azaindole (3MAI) exists solely as
intact double hydrogen-bonded dimeric forms in a single crystal. Both steady-state and time-resolved
measurements down to 8.0 K reveal remarkable deuterium isotope effects on the rate of excited-state double
proton transfer (ESDPT) in the N(1)-deuterated 3MAI (3MAI-d) single crystal. The rates of ESDPT for the
3MAI-d dimer resolved at <150 K are mainly governed by the proton tunneling mechanism. At <12 K, the
nearly temperature-independent ESDPT dynamics lead us to qualitatively deduce a barrier height of ∼1.73
kcal/mol for the 3MAI-d dimer. The results provide an ideal model to investigate the intrinsic ESDPT dynamics
for 7-azaindole analogues in which the structural information is well documented.
1. Introduction
The excited-state double proton transfer (ESDPT)1-3 in the
7-azaindole (7AI) dimer (Figure 1) has long been recognized
as a prototype to mimic the photoinduced mutation of the DNA
base pair.4,5 Much research has focused on the dynamics of
ESDPT, especially in solution phase.6-10 Early studies have
revealed that the rate of proton transfer is within 0.3-1.0 ×
1012 s-1 for the 7AI dimer in nonpolar solvents.6,7 Recently,
0.2-ps and 1.1-ps dynamics were resolved by Takeuchi and
Tahara8 and first assigned to the rates of internal conversion
and proton transfer, respectively, based on a concerted ESDPT
model. Subsequently, Chachisvilis et al.9 reported the 1-ps and
12-ps dynamics and interpreted on the basis of nonconcerted
double proton-transfer reaction, but it was then claimed that
the 12-ps component is due to the vibrational cooling.10 Then,
Fiebig et al. reassigned the 0.2-ps and 1-ps components to the
two-steps double proton transfer11 and concluded that both
trajectories of the symmetric and asymmetric vibrational motion
coupled with the solvent dynamics must be considered for the
ESDPT in the solution phase. Only when the internal energy is
small, such as at sufficiently low temperatures, can one examine
the process of tunneling and nonconcertedness. Despite different
Figure 1. Differences in photophysical properties for the 7AI-
associated structures: a. the dual hydrogen-bonded dimer (in solution)
viewpoints on the mechanism, both claimed observation of the
kinetic isotope effects on the rate of ESDPT at ambient
and b. tetrameter (in a single crystal).
temperature.
Temperature-dependent ESDPT dynamics on 7AI were once
In a 77 K methylcyclohexane glass containing concentrated 7AI,
suggested to be inaccessible due to the formation of thermo-
the dominant steady-state emission maximum at ∼460 nm was
dynamically more favorable oligomers at low temperature.12,13
reported to originate from phosphorescence of the oligomers
in which the cooperative double proton transfer is prohibited.12,13
* To whom correspondence should be addressed.
Recently, via a clever approach of applying sufficiently low
† Department of Chemistry, National Chung-Cheng University.
7AI concentrations, Catala´n and Kasha14 were able to resolve
the dimeric form at low temperature and to study its corre-
‡ Department of Chemistry, Fu Jen Catholic University.
§ Department of Chemistry, National Taiwan University.
10.1021/jp020411x CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/09/2002

3-Methyl-7-azaindole in a Single Crystal
J. Phys. Chem. A, Vol. 106, No. 35, 2002 8007
hydrochloride (1.35 g) and sodium acetate trihydrate (2.3 g)
was added to a solution containing 7-azaindole-3-carboxalde-
hyde (0.80 g) in 80 mL of boiling water. After the appearance
of a transitory yellow coloration, a white solid gradually
precipitated. The mixture was then heated on the steam-bath
for ∼15 min, then cooled to yield semicarbazone (1.1 g).
Semicarbazone (1.1 g) was added to a solution containing
sodium (0.96 g) in dry diethylene glycol (40 mL). The solution
was refluxed at 205 °C under nitrogen atmosphere for 1h, and
the mixture was cooled and poured into ∼230 mL of water
containing 3.1 g of glacial acetic acid. The resulting solution,
which contained solid in suspension, was extracted with ether.
The extracts were then washed with water, dried, and evapo-
rated. The crude product was chromatographed on silica gel
(eluent: ethyl acetate) to obtain 0.41 g (55%) 3MAI. 1H NMR
(400 MHz), δ 9.66 (s, 1H), 8.28 (dd, J ) 1.6, 4.8 Hz, 1H),
7.89 (dd, J ) 1.4, 7.6 Hz, 1H), 7.08(s, 1H), 7.04 (dd, J ) 4.4,
8.0 Hz, 1H), 2.31(s, 3H). A single crystal of 3MAI-h with
dimensions of ∼0.50 × 0.45 × 0.45 mm3 was obtained by slow
evaporation in a CH3OH solution. To obtain the 3MAI-d single
crystal, a similar recrystalization procedure was repeated three
times in the CD3OD solution under the N2 atmosphere. After
monitoring the disappearance of the N(1)-H proton in 1H NMR
(in CDCl3), we concluded that ∼95% of the N-H proton had
been deuterated. The transfer of 3MAI-d to the cryostat was
handled in a N2 filled drybox to avoid the D/H exchange on
the surface of the crystal.
Figure 2. Structures of 3MAI and their corresponding proton-transfer
isomer as well as methylated derivatives.
sponding proton-transfer spectroscopy. Instead of the exclusive
proton-transfer tautomer fluorescence in the 7AI dimer, the
appearance of normal dimeric fluorescence at <200 K for the
deuterated 7AI rendered a convincing deuterium isotope effect
on the ESDPT dynamics14 (hereafter N(1)-protonated and
deuterated-7AI analogues are denoted by the suffix -h and -d,
respectively). However, from the thermodynamic point of view,
the hydrogen-bonding equilibrium should be dominated by the
enthalpic factor and hence favors the oligomer formation at
sufficiently low temperatures (e.g., 77 K and lower). It is thus
rather difficult to determine the lower-limit temperature at which
the existence of dual hydrogen-bonded dimer is free from the
interference of oligomers. This in combination with possible
perturbations resulting from solvent dielectric relaxation and/
or viscosity (e.g., solvent cage) may further smear and com-
plicate the ESDPT dynamics.
From yet another approach, the hydrogen bond in the solid
state can be well-characterized.15 It has been demonstrated that
the hydrogen-bonding structure in the single crystal provides
an ideal model for investigating proton-transfer dynamics free
from the solvent perturbation.16,17 Unfortunately, the crystal
structure of 7AI reveals tetrameric units of approximate S4
symmetry, in which molecules are associated by means of four
complementary N-H‚ ‚ ‚N hydrogen bonds18 (Figure 1b). As
a result the cooperative ESDPT is prohibited. Recently, we have
found that chemical modification at the C(3)-position of 7AI,
forming 3-iodo-7-azaindole (3IAI), exhibited a dimeric structure
in the single crystal. Steady-state approaches have shown the
dominant ESDPT process even at 10 K for the 3IAI-h dimer,
resulting in a unique tautomer emission.19 Unfortunately, studies
on ESDPT dynamics of 3IAI, particularly the deuterium isotope
effect, were rather difficult due to its photochemical liability
as well as fast S1-T1 intersystem crossing enhanced by the
iodine heavy atom effect.19,20 Thus, chemical modifications on
7AI suitable for investigating intrinsic ESDPT in the single-
crystal became one of our focuses. In this study 3-methyl-7-
azaindole (3MAI, Figure 2) was synthesized, of which the
photostability is superior to that of 3IAI. Furthermore, unlike
3IAI where the association of iodine substituents induces strong
perturbation due to the enhanced spin-orbit coupling and
resonance effects, the much more inert methyl substitution in
3MAI provides an ideal model to mimic the intrinsic ESDPT
dynamics of the 7AI dimer.
The compound 1,3-dimethyl-7-azaindole (3MM(1)AI, Figure
2) was synthesized by adding sodium hydride (57%, 60 mg) to
the THF solution containing 3MAI (0.15 g), followed by the
addition of methyl iodide (30 mg). 1H NMR (CDCl3, 400 MHz),
δ 2.32 (s, 3H), 4.10 (s, 3H), 7.0 (s, 1H), 7.17 (t, 1H), 8.05 (d,
J ) 7.2 Hz, 1H), 8.33 (d, J ) 5.2 Hz, 1H). The compound
3,7-dimethyl-1H-pyrrolo[2,3-b]pyridine (3MM(7)AI, Figure 2)
was synthesized by the reaction of 3MAI (0.12 g) and CH3I
(0.58 g) in THF under a N2 atmosphere. NaOH (2.5 N, 3 mL)
was then added and the mixture was stirred for ∼20 min to
obtain 3MM(7)AI (60 mg). 1H NMR (CDCl3, 400 MHz), δ 2.37
(s, 3H), 4.74(s, 3H); 7.37(t, 1H); 7.55(s, 1H); 7.98(d, J ) 7.64
Hz, 1H); 8.34(d, J ) 6.8 Hz, 1H).
2.2. Measurements. A closed-helium-cycle cryostat (Oxford,
Model CCC1104) was used for the temperature-dependent study
in the range of 298-8 K. The solid sample (i.e., single crystal)
was placed tightly between two thin quartz plates that came
into close contact with the Indium part of the cooling compart-
ment to ensure maximum thermal conduction. The solid sample
was excited by the fourth harmonic (266 nm) of the Nd:YAG
laser under a front-face configuration. The resulting fluorescence
was detected by an intensified charge coupled detector (ICCD,
Princeton Instrument, Model 576G/1) coupled with a polychro-
mator in which the grating is blazed with a maximum of 500
nm. Occasionally, to obtain excitation spectra the output of an
Nd:YAG (355 nm, 8 ns, Continuum Surlite II) pumped optical
parametric oscillator was tuned between 650 and 720 nm and
was then frequency doubled by a BBO crystal to obtain a tunable
325-360 nm excitation frequency. The intensity of the laser
pulse was measured by a joule meter, which was then normal-
ized to obtain the corrected excitation spectra. A combination
of filters was used in front of the exit polychromator to isolate
the emission wavelengths of interest. The crystal orientation
was 45° with respect to the direction of the excitation light.
2. Experimental Section
2.1. Materials. Synthesis of 3-methyl-7-azaindole (3MAI). A
precursor 7-azaindole-3-carboxaldehyde has to be prepared prior
to the synthesis of 3MAI, which was synthesized according to
the previous report.21 Subsequently, a solution of semicarbazide
Detailed fluorescence lifetime measurement has been de-
scribed in the previous report.22 Briefly, a third harmonic of
the Ti-Sapphire oscillator (100 fs, 82 MHz, Spectra Physics)

8008 J. Phys. Chem. A, Vol. 106, No. 35, 2002
Yu et al.
TABLE 1: Bond Lengths [Ångstroms] and Angle [Degrees]
for 3MAI-h in a Single Crystal
Dimer 1
N(1)-C(5)
N(2)-C(5)
C(1)-C(2)
C(3)-C(4)
C(4)-C(6)
C(6)-C(8)
N(1A)‚‚‚H(2)
1.332(2) N(1)-C(1)
1.365(2) N(2)-C(7)
1.387(2) C(2)-C(3)
1.394(2) C(4)-C(5)
1.425(2) C(6)-C(7)
1.502(2) N(2)-H(2)
2.131(2) N(1A)‚‚‚N(2)
1.335(2)
1.368(2)
1.377(2)
1.416(2)
1.367(2)
0.856(2)
2.980(2)
C(5)-N(1)-C(1) 113.95(14) C(5)-N(2)-C(7)
N(1)-C(1)-C(2) 124.99(16) C(3)-C(2)-C(1)
C(2)-C(3)-C(4) 117.77(15) C(3)-C(4)-C(5)
C(3)-C(4)-C(6) 135.71(15) C(5)-C(4)-C(6)
N(1)-C(5)-N(2) 125.88(14) N(1)-C(5)-C(4)
N(2)-C(5)-C(4) 107.54(14) C(7)-C(6)-C(4)
C(7)-C(6)-C(8) 128.49(16) C(4)-C(6)-C(8)
108.30(14)
119.98(16)
116.74(14)
107.54(14)
126.57(14)
105.41(14)
126.10(15)
C(6)-C(7)-N(2) 111.22(15) N(1A)‚‚‚H(2)-N(2) 170.90
Dimer 2
N(3)-C(13)
1.335(2) N(3)-C(9)
1.366(2) N(4)-N(15)
1.390(2) C(10)-C(11)
1.386(2) C(12)-C(13)
1.434(2) C(14)-C(15)
1.500(2) N(4)-H(4)
2.121(2) N(3A)‚‚‚N(4)
1.335(2)
1.368(2)
1.376(2)
1.415(2)
1.365(2)
0.865(2)
2.979(2)
N(4)-C(13)
C(9)-C(10)
C(11)-C(12)
C(12)-C(14)
C(14)-C(16)
N(3A)‚‚‚H(4)
C(13)-N(3)-C(9)
N(3)-C(9)-C(10)
113.75(14) C(13)-N(4)-C(15) 108.42(14)
125.03(16) C(11)-C(10)-C(9) 119.82(15)
C(10)-C(11)-C(12) 117.89(15) C(11)-C(12)-C(13) 117.01(14)
C(11)-C(12)-C(14) 135.84(14) C(13)-C(12)-C(14) 107.15(13)
N(3)-C(13)-N(4)
125.77(14) N(3)-C(13)-C(12) 126.50(14)
N(4)-C(13)-C(12) 107.71(14) C(15)-C(14)-C(12) 105.55(14)
C(15)-C(14)-C(16) 128.67(17) C(12)-C(14)-C(16) 125.77(15)
C(14)-C(15)-N(4) 111.16(15) N(3A)‚‚‚H(4)-N(4) 171.4
of hydrogen-bonding dimers (specified as dimer 1 and dimer
2) are in the same circumstances in the single crystal, we have
closely examined these two components in the X-ray data. The
results of X-ray structural analyses are shown in Figure 3 and
Table 1. The sameness in configuration of these two hydrogen-
bonded dimers under different spatial arrangements can be
supported by several critical bond lengths and angles shown in
Table 1. For example, N(1A)‚‚‚N(2) and N(1A)‚‚‚H(2)N(2) were
calculated to be 2.980 and 2.131 Å, respectively, for dimer 1.
In addition, the angle of N(1A)‚‚‚H(2)N(2) was calculated to
be 170.9°. In comparison, similar results of 2.979 Å (N(3A)‚‚
‚N(4)) and 2.121 Å (N(3A)‚‚‚H(4)N(4)) with an N(3A)‚‚‚H(4)N-
(4) angle of 171.4°were obtained for dimer 2. These, in
combination with similar bond angles and bond distances among
the rest of the atoms between dimer 1 and 2 (see Table 1)
indicate that within the standard deviation these two hydrogen-
bonding dimers are in the same configuration and hence should
behave in practically the same way.
To ensure the same crystal structure in both fluorescence and
X-ray measurements, we first determined the structure by X-ray
spectroscopy, followed by performing temperature-dependent
fluorescence spectroscopy and dynamics of the same single
crystal. During the fluorescence measurement an ultrahigh
vacuum was maintained in the cryostat so that the D/H isotope
exchange was negligible. In the case of 3MAI-d, great care was
taken to perform the X-ray measurement in dry air conditions,
and the results are attached in the Supporting Information.
Comparing X-ray data shown in Table 1 and Supporting
Information, it is obvious that single crystals of 3MAI-h and
3MAI-d used in the fluorescence studies are isomorphous. The
X-ray crystallography was also performed at -80 °C. The results
showed negligible differences in structure from that obtained
at 295 K, indicating no temperature-dependent structural
transformation for 3MAI in a single crystal.
Figure 3. a. The projection of 3MAI-h dimer in a unit cell of the
single crystal. b. Two types of spatially arranged dimeric forms,
specified as dimer 1 and 2. Bond distance and angle are in angstroms
and degree (degrees), respectively.
coupled with a pulse picker (NEOS, model N17389) was used
as an excitation source, giving a tunable wavelength in the range
of 255-275 nm with a repetition rate of 8 MHz. An Edinburgh
OB 900-L single-photon counter was used as a detecting system.
The resolution of the time-correlated photon-counting system
is limited by the detector response of ∼30 ps. The fluorescence
decays were analyzed by the sum of exponential functions with
an iterative convolution method reported previously,23 which
allows partial removal of the instrument time broadening and
consequently renders a temporal resolution of ∼15 ps.
3. Results
3.1. Structure of 3MAI in a Single Crystal. As shown in
Figure 3, the crystal structure of 3MAI-h is composed of a cyclic
dimer via dual N-H‚‚‚N hydrogen bonds. Parallel cyclic dimers
are stacked via the interaction between their ring π-systems to
form alternating slabs. The orientation of the parallel molecules
in one slab is nearly perpendicular to those in adjacent slabs.
The space group is p21/C, and Z ) 8. Thus, there are two kinds
of crystallographically independent molecules. Each of them
forms a hydrogen-bonding dimer by the operation of the
inversion center, while there is no symmetry relation between
the two forms of dimers. To examine whether these two kinds

3-Methyl-7-azaindole in a Single Crystal
J. Phys. Chem. A, Vol. 106, No. 35, 2002 8009
Figure 4. (s) The fluorescence of the 3MAI-h crystal as a function
of temperature at a. 298, b. 240, c. 195, d. 90, e. 50, and f. 8 K. (- -)
The absorption spectra of 3MAI-h in a solid form at room temperature.
Inset: The plot of ln[kobs - (kr + knr)] versus the reciprocal of the
temperature.
Figure 6. (s) The fluorescence of 3MAI-d crystal as a function of
temperature at a. 298, b. 200, c. 165, d. 135, e. 110, f. 85, g. 35, h. 12,
and i. 8 K. The excitation spectrum of the 3MAI-d crystal monitored
at the F2 (• •) and F1 bands (o o). The excitation wavelength was tuned
from 360 to 325 nm. (- -) The absorption spectra of 3MAI-d in a solid
form at room temperature. Inset: The plot of tautomer emission
intensity at 500 nm versus the temperature (in Kelvin).
fluorescence radiative decay rate, the result simply indicates
the decrease in nonradiative deactivation rates (vide infra), which
is consistent with the steady-state measurement, showing that
the intensity of the tautomer emission increased as the temper-
ature decreased (see Figure 4a-f).
Remarkable temperature-dependent fluorescence spectra were
observed for the 3MAI-d single crystal. Dual emission became
obvious at <200 K, in which a normal emission band (specified
as the F1 band) maximized at 395 nm gradually appeared (Figure
6). Conversely, the 500-nm tautomer emission (specified as the
F2 band) intensity versus temperature was slightly irregular (see
insert of Figure 6); it increased upon decreasing the temperature
from 298 to 190 K, followed by a decrease in the intensity at
150-120 K. At < 12 K the intensity as well as the ratio of
dual emission was nearly temperature independent. By tuning
the excitation wavelength, the resulting excitation spectrum,
within experimental error, was independent of the emission
wavelength monitored at either the F1 or F2 band, which is also
effectively identical with the absorption profile (see Figure 6).
In accordance with the steady-state approach, a remarkable
deuterium isotope effect on the ESDPT dynamics was also
observed in the 3MAI-d single crystal. While the rise component
of the F1 band was beyond the instrument response of ∼ 15 ps
at 298-8.0 K, at <150 K the decay was resolvable and revealed
drastic temperature dependence. As shown in Figure 7A and
Table 2, the results clearly showed a decrease of the single-
exponential decay rate constant, kobsN, upon decreasing the
Figure 5. The fluorescence and fluorescence excitation spectra of
3MM(1)AI (s) and 3MM(7)AI (- -) in the crystal form at ambient
temperature.
3.2. Spectroscopy and Dynamics in the Single Crystal. In
contrast to a normal fluorescence (λmax ∼ 370 nm) observed in
7AI,19 3MAI-h in a single crystal exhibited a unique, large
Stokes-shifted fluorescence (λmax ∼ 500 nm) throughout 298-
8.0 K (Figure 4). In comparison, 3MM(1)AI (see Figure 2),
which serves as a nonproton-transfer model for 3MAI, revealed
a fluorescence with a peak maximum at 410 nm (τf ∼ 0.9 ns)
in the solid state. Conversely, 3MM(7)AI, a model compound
of the proton-transfer tautomer, exhibited room-temperature
fluorescence maximized at 510 nm (Figure 5, τf ∼ 2.1 ns), of
which the spectral and dynamical features resemble those
observed in the 3MAI-h single crystal. These steady-state
approaches unambiguously lead to two concluding remarks: 1.
ESDPT is apparently operative in the 3MAI-h single crystal,
resulting in a 500-nm proton-transfer tautomer emission. 2. Due
to the lack of any observable normal emission, the rate of
ESDPT must be fast even at a temperature of as low as 8.0 K.
The latter viewpoint was supported by the subsequent time-
resolved measurement. When monitored at the 500-nm emission
band, the rise component of the tautomer fluorescence was too
fast to be resolved (, 15 ps) throughout 298-8.0 K. Con-
versely, the decay of the tautomer emission can be well fitted
by single exponential kinetics. The lifetime-revealed slight
temperature dependence, varying from 2.7 (298 K) to 7.4 ns
(8.0 K). From the standpoint of a temperature-independent
temperature. For instance, kobs measured to be 3.8 × 109 s-1
N
at 8 K is smaller than that of 2.6 × 1010 s-1 measured at 100
K by ∼7-fold. Conversely, the fluorescence dynamics of the
F2 band resolved at <150 K could only be fitted by dual
exponential kinetics, which consist of a rise and a decay
component as indicated by the fitted negative and positive
preexponential factors, respectively. When monitored at >520
nm region free from the interference of the normal emission,
the absolute values for both rise and decay preexponential
factors, within experimental error, are similar (see Table 2).
Slight differences are possibly attributed to the existence of ∼5%
of nondeuterated 3MAI-h in which fast ESDPT takes place.
Therefore, the fitted preexponential value for the decay com-
ponent is greater than that for the rise component. Table 2 also

8010 J. Phys. Chem. A, Vol. 106, No. 35, 2002
Yu et al.
Figure 8. The plot of ln[kobs - (kr + knr + knr(T))] versus the reciprocal
of the temperature for a. the decay of the F1 band (2) and b. rise kinetics
of the F2 band (O) in a 3MAI-d single crystal at e150 K. See text for
the definition of kr + knr + knr(T).
4. Discussion
For the F2 band, the observed temperature-dependent decay
rate can be expressed as
Figure 7. A. The time-dependent fluorescence decay dynamics of the
3MAI-d crystal monitored at the F1 band (380 nm) at various
temperatures of a. 150, b. 100, c. 70, d. 40, e. 20, and f. 10K. B. The
rise dynamics of the F2 band (520 nm) monitored at a. 190, b. 150, c.
70, and d. 10 K.
kobs ) kr + knr + knr(T)
(1)
where knr denotes the temperature independent radiationless
decay rate constant, possibly involving internal conversion,
intersystem crossing, etc. The temperature-dependent radiation-
less decay rate constant knr(T) can be further expressed as an
Arrhenius type of thermally deactivated pathway, namely, knr-
TABLE 2: Temperature-Dependent Fluorescence Relaxation
Dynamics for 3MAI-d in a Single Crystal
(T) ) Ae-E /RT where Ea is the barrier for the thermally
a
fluorescence lifetime
deactivated process. As indicated by the time-resolved measure-
ment, the lifetime of the tautomer emission (the F2 band) was
invariable at <12 K. Thus, it is reasonable to assume that knr-
(T)is negligible, and the decay rate of 1.36 × 108 s-1 measured
at, e.g., 8 K is equivalent to kr + knr. The plot of ln[kobs - (kr
+ knr)] versus the reciprocal of the temperature reveals straight-
line behavior (see insert of Figure 4), supporting the validity of
the thermal deactivation pathway for the tautomer emission. A
best linear least-squares fit for the insert of Figure 4 gives Ea
and the frequency factor to be 0.35 kcal/mol and 5.2 × 108
s-1, respectively.
F2 (λmax ) 500 nm)a
temperature F1 (λmax )395 nm)
(K)
τ (ns)
τ1 (ns)
a1
τ2 (ns)
a2
150
100
70
40
20
10
8
NA
6.66 1.000
0.038
0.061
0.127
0.210
0.260
0.263
0.040 -0.208 6.80 0.239
0.060 -0.205 7.15 0.235
0.130 -0.010 7.28 0.014
0.226 -0.011 7.33 0.017
0.260 -0.011 7.36 0.013
0.265 -0.011 7.38 0.015
a The F2 band was fitted by dual exponential kinetics expressed as
-1
-1
t
t
F(t) ) a1e-τ + a2e-τ
.
1
2
For the F1 band, the observed temperature-dependent decay
rate can be expressed as
shows an interesting correlation between decay and rise
components for the F1 and F2 bands, respectively, at various
temperatures, in which the rise time of the F2 band, within
experimental error, was the same as the decay time of the F1
band. In summary, the time-resolved measurement in combina-
tion with steady-state approaches led us to conclude that both
F1 (normal emission) and F2 (tautomer emission) bands originate
from the same ground-state species, i.e., the normal dimeric
form, and the F1 band is apparently the precursor of the F2 band.
Thus, the existence of a nonnegligible ESDPT barrier for the
3MAI-d dimer is apparent, and the deuterium isotope effect
becomes significant on the rate of ESDPT at sufficiently low
temperatures. The irregular temperature-dependent tautomer
emission intensity for 3-MAI-d can thus be rationalized by two
competing relaxation processes: 1. As the temperature de-
creases, the decrease of the ESDPT rate results in a smaller
yield of population to the tautomer state, 2. The tautomer
fluorescence quantum yield tends to increase at lower temper-
ature due to the decrease of the rate of radiationless deactiva-
tions. Details on the relaxation dynamics will be discussed in
the following section.
kobs ) kr + knr + knr(T) + kpt(T)
(2)
where knr(T)denotes the thermally deactivated pathways except
for the rate of proton transfer kpt(T). It is not feasible to obtain
kr + knr + knr(T) for the normal emission (i.e., the F1 band) of
3MAI in a single crystal because the dynamics of relaxation of
the excited normal species are still dominated by the rate of
double proton-transfer reaction even at low temperatures.
Alternatively, we have attempted to obtain kr + knr + knr(T)
values by performing the temperature-dependent relaxation
dynamics of the 3MM(1)AI crystal in which the excited-state
proton-transfer reaction is prohibited due to the lack of pyrrolic
hydrogen. The results showed a slightly temperature-dependent
decay rate from 4 × 108 s-1 (150 K) to 1.1 × 108 s-1 (8 K).
Taking the kobs value of the 3MM(1)AI crystal to be kr + knr
knr(T) for the 3MAI-d dimer, a plot of ln[kobs - (kr + knr
+
+
knr(T))] versus 1/T is shown in Figure 8. Apparently, the
contribution of kr + knr + knr(T) is rather small in this plot simply
due to its much smaller value in comparison to the rate of double
proton-transfer reaction. If the ESDPT dynamics, kpt(T), mainly

3-Methyl-7-azaindole in a Single Crystal
J. Phys. Chem. A, Vol. 106, No. 35, 2002 8011
A rough estimate of a0 can be obtained from the average
N-H‚‚‚N hydrogen bond distances,24 the estimated displacement
of the proton transfer (denoted d in Figure 9) and classical
turning points for the zero-point motion of the N-H (or N-D)
harmonic vibration from the equilibrium position. For the case
of 7AI, a0, has been estimated to be 0.27 and 0.36 Å for N(1)-H
and N(1)-D substituents, respectively.3,11,25 The same a0 values
were used for 3MAI in the following calculation, assuming that
methyl substitution at the C(3) position has negligible perturba-
tion on the dimeric structure. Since the tunneling rate was
measured to be independent of the temperature in the range of
8-12 K, the thermally available energy E can be neglected.
Accordingly, the barrier height E0 could be deduced by applying
known reaction-coordinate frequency ν, effective mass m and
the estimated a0 value to eq 4. For the case of 3MAI-d, taking
ν to be ∼6.0 × 1013 s-1 (∼2000 cm-1), an E0 value of 605
cm-1 (1.73 kcal/mol) was thus deduced. At a relatively high
temperature such as >20 K (see Figure 8) the available thermal
energy is nonnegligible. The actual tunneling rate has to be
calculated by the sum of all possible thermally populated states
and is thus complicated. Given the difference in ZPE of ∼400
cm-1 between N-H and N-D stretching, a barrier height of
Figure 9. A simplified one-dimensional PES regarding the ESDPT
based on a bound-free potential curve. See text for the detailed
description of notations.
incorporate a barrier crossing, i.e., a classical thermally activated
process, straight-line behavior is expected upon plotting ln[kobs
- (kr + knr + knr(T))] as a function of the reciprocal of the
temperature. In contrast, the plot depicted in Figure 8a revealed
an asymptote-like curve, which approached a constant value at
<12 K. A similar pattern was also obtained when the rise
dynamics of the F2 band (i.e., the tautomer emission) are plotted
against the reciprocal of the temperature (see Figure 8b).
ESDPT was further deduced to be ∼205 cm-1 (∼0.58 kcal/
H
mol) for the 3MAI-h dimer. Plugging E0 of 205 cm-1, a0
)
0.27 Å, m ) 1, and ν ) 2800 cm-1 (8.4 × 1013 s-1) to eq 4,
the proton tunneling rate, ktunnel, was calculated to be 4.3 ×
The concave asymptote for the kpt(T)-versus-1/T plot in
combination with a nearly constant ESDPT rate at 8-12 K led
us to conclude that the dynamics of ESDPT for the 3MAI-d
crystal are not governed by a thermally activated process, but
more plausibly proceed through a proton-tunneling mechanism.
To apply such theories, simplified approximation about the
ESDPT potential energy surface (PES) is necessary. Both
steady-state and dynamic approaches indicate that the ESDPT
reaction in the 3MAI dimer is exergonic and irreversible.
Accordingly, a simplified one-dimensional PES regarding the
ESDPT can be represented by a bound-free potential curve
shown in Figure 9, in which the proton-tunneling rate is given
by
1012 s- 1 for 3MAI-h at <12 K. This time scale of ktunnel
)
- 1
233 fs is too fast to be resolved by our picosecond time-
correlated single photon counting system. On the basis of the
energy barrier of 0.58 and 1.73 kcal/mol, respectively, rates of
ESDPT for 3MAI-h and 3MAI-d were calculated to be
negligibly small via a thermally activated process at <12 K.
Certainly, it should be noted that the proton tunneling could
rarely be satisfactorily described as a one-dimensional process.
The motion of atoms between which the proton is being
exchanged may have a major effect on the tunneling rate.26-32
Although the above results conclude a finite barrier to ESDPT,
it is rather small, and subtle changes in the proton-transfer
distance or energetics could have large effects on the proton-
transfer rates. This recognition has led to the introduction of
two-dimensional approaches in which the second dimension
represents the effective motion of the other atoms during the
proton-transfer reaction. Theoretical approaches33-35 have re-
vealed that the KIE (kinetic isotope effect) temperature depen-
dence may be caused by thermally excited reactants and/or
products that modulate the barrier width and shape. Although
two-dimensional models may produce a more qualitative insight
into the kinetic isotope effect and the temperature dependence
of the transfer process, they still require the introduction of
empirical parameters and thus cannot render quantitative predic-
tions. For more precise quantitative work it is necessary to
consider all degrees of freedom of the global system for an
overall proton-transfer reaction.
a0
x
2m
p
ktunnel ) ν exp -
V(x)1/2 dx
(3)
∫
[
]
- a0
where V(x) is the energy barrier along the reaction coordinate,
ν is the bound vibrational frequency, i.e., the reaction-coordinate
frequency of the N-H (or N-D) mode in the case of 3MAI, m
is the effective mass of the proton or deuterium, and a0 is the
distance from the center of the barrier to the classical vibrational
turning points. Equation 3 can easily be evaluated if the potential
barrier is further represented by a semiempirical, one-dimen-
sional approach based on a parabolic shape-like activation
2
energy, i.e., V(x) ) E0(1 - x2/a0 ) where E0 denotes the
maximum barrier height above the zero point energy (ZPE).
The resolvable ESDPT time scales for 3MAI-d are in the range
of few tens to hundreds picoseconds. Thus, it is reasonable to
assume a fast thermal equilibrium prior to the ESDPT reaction.
Further support of this viewpoint can be given by varying the
excitation wavelength from 255 to 275 nm, giving the same
results on the ESDPT dynamics (not shown here). Accordingly,
the tunneling rate for a molecule possessing thermal energy E
above ZPE can thus be expressed as
At this stage, we are unable to differentiate the one-step versus
two-step, nonconcerted ESDPT mechanism. Although the
experimental results indicate a single decay and rise component
for normal and tautomer emissions, respectively, it is also
possible that ESDPT takes place via a two-step mechanism, and
the rate in one of the steps is too fast to be resolved by the
picosecond resolution. Alternatively, fluorescence upconversion
with femtosecond resolution may provide valuable information.
Unfortunately, the temperature-dependent study in the 3MAI
single crystal is intrinsically difficult due to the photolysis and/
πa0
ktunnel ) ν exp -
2m(E - E)
(4)
x
[
0
]
p

8012 J. Phys. Chem. A, Vol. 106, No. 35, 2002
Yu et al.
(11) Fiebig, T.; Chachisvilis, M.; Manger, M.; Zewail, A. H.; Douhal,
A.; Garcia-Ochoa, I.; de La Hoz Ayuso, A. J. Phys. Chem. A. 1999, 103,
7419.
or thermal reaction upon high repetition rate and high power
laser excitation. Focus on this subject is currently in progress.
(12) Bulska, H.; Chodkowska, A. J. Am. Chem. Soc. 1980, 102, 3259.
(13) Bulska, H.; Grabowska, A.; Pakula, B.; Sepiol, J.; Waluk, J.; Wild,
U. P. J. Lumin. 1984, 29, 65.
(14) Catala´n, J.; Kasha, M. J. Phys. Chem. A 2000, 104, 10812.
(15) Steiner, T. Angew. Chem., Int. Ed. 2002, 41, 48.
(16) Sekikawa, T.; Kobayashi, T.; Inabe, T. Phys. Chem. A 1997, 101,
644.
(17) Sekikawa, T.; Kobayashi, T.; Inabe, T. J. Phys. Chem. B 1997,
101, 10645.
(18) Dufour, P.; Dartiguenave, Y.; Dartiguenave, M.; Dufour, N. Lebuis,
A. M.; Belanger-Gariepy, F.; Beauchamp, A. L. Can. J. Chem. 1990, 68,
193.
(19) Chou, P. T.; Liao, J. H.; Wei, C. Y.; Yang, C. Y.; Yu, W. S.; Chou,
Y. H. J. Am. Chem. Soc. 2000, 122, 986.
(20) After prolong photolysis of 3IAI the crystal turned yellow,
accompanied by a decrease of the proton-transfer tautomer emission. Under
the same experimental conditions similar photolysis reaction did not occur
in 3MAI. It is thus believed that iodine substituents play a role for the
observed photochemistry in 3IAI.
5. Conclusion
In conclusion, both steady-state and time-resolved measure-
ments down to 8.0 K have revealed a remarkable deuterium
isotope effect on the rate of ESDPT in the 3MAI-d single crystal.
A barrier height of ESDPT of 1.73 and 0.58 kcal/mol was
estimated for the 3MAI-d and 3MAI-h dimers, respectively, in
a single crystal, and the ESDPT rates are mainly governed by
a proton/deuterium tunneling mechanism. The results provide
a prototype to mimic the intrinsic ESDPT dynamics of the 7AI-
like dual hydrogen-bonded dimer with complete structural
information, which are believed to bring up a broad spectrum
of interest in the field of proton-transfer studies.
Acknowledgment. This work was supported by the National
(21) Robison, M. M.; Robison, B. L. J. Am. Chem. Soc. 1956, 78, 1247.
(22) Chou, P. T.; Chen, Y. C.; Yu, W. S.; Chou, Y. H.; Wei, C. Y.;
Cheng, Y. M. J. Phys. Chem. A 2001, 105, 1731.
(23) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991-1024.
(24) Vinogradov, S. N.; Linnell, R. H. Hydrogen Bonding; Van
Nostrand-Reinhold: New York, 1971; Chapter 7.
(25) Kim, S. K.; Breen, J. J.; Willberg, D. M.; Peng, L. W.; Heikal, A.;
Syage, J. A.; Zewail, A. H. J. Phys. Chem. 1995, 99, 7421.
(26) Fernandez-Ramos, A.; Smedarchina, Z.; Siebrand, W.; Zgierski,
M. Z.; Rios, M. A. J. Am. Chem. Soc. 1999, 121, 6280.
(27) Bicerano, J.; Schaefer, H. F., III; Miller, W. H. J. Am. Chem. Soc.
1983, 105, 2550.
Science Council NSC (90-2113-M-002-055).
Supporting Information Available: X-ray data for 3MAI-h
and 3MAI-d. This material is available free of charge via the
Internet at http://pubs.acs.org.
References and Notes
(1) Taylor, C. A.; El-Bayoumi, M. A.; Kasha, M. Proc. Nat. Acad.
Sci. U.S.A. 1969, 63, 253.
(2) Ingham, K. C.; Abu-Elgheit, M.; El-Bayoumi, M. A. J. Am. Chem.
Soc. 1971, 93, 5023
(3) Ingham. K. C.; El-Bayoumi, M. A. J. Am. Chem. Soc. 1974, 96,
1674.
(4) Watson, J. D.; Crick, F. H. C. Nature (London) 1953, 171, 964.
(5) (a) Lowdin, P.-O. AdV. Quantum Chem. 1965, 2, 213. (b) Lowdin,
P.-O. ReV. Mod. Phys. 1963, 35, 724.
(6) Hetherington, W. M., III; Micheels, R. H.; Eisenthal, K. B. Chem.
Phys. Lett. 1979, 66, 230.
(28) Siebrand, W.; Wildman, T. A.; Zgierski, M. Z. J. Am. Chem. Soc.
1984, 106, 4083, 4089.
(29) Smedarchina, Z.; Siebrand, W. Chem. Phys. 1993, 170, 347.
(30) Benderskii, V. A.; Goldanskii, V. I.; Makarov, D. E. Phys. Rep.
1993, 233, 195.
(31) Benderskii, V. A.; Makarov, D. E.; Wight, C. A. Chemical
Dynamics at Low Temperatures; J. Wiley & Sons: New York, 1994; p
385.
(32) Liu, Y.-P.; Lu, D.-h.; Gonzalez-Lafont, D. G.; Truhlar, D. G.;
Garrett, B. C. J. Am. Chem. Soc. 1993, 115, 7806
(33) Bruno, W. J.; Bialek, W. Biophys. J. 1992, 63, 689.
(34) Antoniou, D.; Schwartz, S. D. Proc. Nat. Acad. Sci. U.S.A. 1997,
94, 12360.
(7) Share, P.; Sarisky, M.; Pereira, M.; Repinec, S.; Hochstrasser, R.
M. J. Lumin. 1991, 48/49, 204.
(8) Takeuchi, S.; Tahara, T. Chem. Phys. Lett. 1997, 277, 340.
(9) Chachisvilis, M.; Fiebig, T.; Douhal, A.; Zewail, A. H.J. Phys.
Chem. A. 1998, 102, 669.
(10) Takeuchi, S.; Tahara, T. J. Phys. Chem. A. 1998, 102, 7740.
(35) Borgis, D.; Hynes, J. T. J. Chem. Phys. 1991, 94, 3619.