Anomalously slow proton transport of a water molecule

Article abstract of DOI:10.1021/jp112210x

Unusually low proton-transporting ability of a water molecule has been observed in the excited-state proton transfer (ESPT) of a 7-azaindole (7AI) molecule complexed cyclically with a water molecule in diethyl ether and dipropyl ether. In contrast with ultrafast (<1 ps) proton diffusion along a systematically structured hydrogen-bond network in an aqueous solution, the proton transport of a water monomer has been observed to be extremely slow (～1 ns) because it is hard for a monomeric water molecule alone to accept or donate a proton. Thus, polar ether molecules surrounding a cyclic hydrogen-bonded 1:1 7AI-water complex (Nc) play a crucial role in the ESPT of Nc. The proton acceptance of a water molecule from the acidic amino group of 7AI via tunneling to form a hydronium ion, which is the rate-determining step, initiates ESPT, and the subsequent rapid proton donation of the hydronium ion to the basic imino group of 7AI takes place barrierlessly to complete ESPT without accumulating any intermediate. Due to the anomalously weak proton-transporting ability of a water monomer, the elaborate reorganization of the hydrogen-bond bridge in Nc to form an optimized precursor configuration is required for proton tunneling.

Full text of DOI:10.1021/jp112210x

ARTICLE
pubs.acs.org/JPCB
Anomalously Slow Proton Transport of a Water Molecule
Sun-Young Park, Hyeok Jeong, and Du-Jeon Jang*
School of Chemistry, Seoul National University, NS60, Seoul 151-742, Korea
S Supporting Information
b
ABSTRACT: Unusually low proton-transporting ability of a
water molecule has been observed in the excited-state proton
transfer (ESPT) of a 7-azaindole (7AI) molecule complexed
cyclically with a water molecule in diethyl ether and dipropyl
ether. In contrast with ultrafast (<1 ps) proton diffusion along a
systematically structured hydrogen-bond network in an aqu-
eous solution, the proton transport of a water monomer has
been observed to be extremely slow (∼1 ns) because it is hard
for a monomeric water molecule alone to accept or donate a
proton. Thus, polar ether molecules surrounding a cyclic
hydrogen-bonded 1:1 7AIꢀwater complex (Nc) play a crucial role in the ESPT of Nc. The proton acceptance of a water molecule
from the acidic amino group of 7AI via tunneling to form a hydronium ion, which is the rate-determining step, initiates ESPT, and
the subsequent rapid proton donation of the hydronium ion to the basic imino group of 7AI takes place barrierlessly to complete
ESPT without accumulating any intermediate. Due to the anomalously weak proton-transporting ability of a water monomer, the
elaborate reorganization of the hydrogen-bond bridge in Nc to form an optimized precursor configuration is required for proton
tunneling.
Thus, the anomalously high mobility of a proton in water is at-
tributed to fast interconversion between the Eigen and the Zundel
cations through well-organized H-bond chains. Then, one would
be concerned about whether a water molecule alone can trans-
port a proton facilely in the absence of H-bond networks. In this
regard, to understand the proton-transport mechanism of a water
monomer at the molecular level, we have employed a model system
of a 7-azaindole (7AI) molecule complexed cyclically with a water
molecule and investigated the ESPT of the cyclic 1:1 7AIꢀwater
complex.
Proton transfer plays a decisive role in a wide variety of chemical
and biological processes relevant to vital phenomena,1,6,17,18,29 and
amphoteric aromatic molecules such as 7AI and hydroxyquino-
lines have been the subject of growing interest to researchers who
study proton-relay dynamics.7ꢀ10,30ꢀ51 In particular, 7AI, which
is an analogue of nucleobases such as adenine and guanine, has
been explored intensively because it is a novel optical probe to
investigate the structure of proteins and the dynamics of enzy-
matic catalyses in biological systems.36ꢀ51 On one hand, because
a 7AI dimer is structurally similar to a DNA base pair, experi-
mental and theoretical mechanistic studies on the proton transfer
of the 7AI dimer have received considerable attention.36ꢀ39 On
the other hand, the normal species (N) of monomeric 7AI can
undergo ESPT with assistance of protic solvents, such as water
and alcohols, to produce the tautomeric species (T), so that
1. INTRODUCTION
Water, which is the prime agent in most biological phenomena
and chemical processes, has often been the subject of many ex-
perimental and theoretical studies.1ꢀ10 The characteristic physi-
cal and chemical properties of water, such as the high boiling point,
the strong surface tension, and the high vaporization enthalpy,
originate primarily from systematically well-structured hydrogen
(H)-bond networks.11 In an aqueous solution, H bonds break
and form continuously,12 and their breaking and forming dy-
namics is coupled closely to proton mobility in water.13ꢀ15 Because
H bonds play a pivotal role in a great number of chemical and
biological processes such as acidꢀbase reactions and proton pumps
through membraneꢀprotein channels,3,16ꢀ18 many researchers
have tried to elucidate the structural change mechanism of H-bond
networks at the molecular level.19,20 Since the famous concept of
fast proton transfer via structural diffusion in water was proposed
by Grotthuss in 1806,21 subsequent explanations have refined the
concept by invoking thermal hopping, proton tunneling, or sol-
vation effects.3,20ꢀ25 Computer simulations have shown that in-
terconversion between an Eigen cation (H3Oþ (H2O)3), whose
3
hydronium core is H-bonded to three H2O molecules, and a
Zundel cation (Hþ (H2O)2), where a proton fluctuates between
3
two water molecules, leads to fast proton migration in water.26ꢀ28
Recently, we have reported the excited-state proton transfer (ESPT)
of 7-hydroxyquinoline (7HQ) along a H-bonded water dimer in
diethyl ether (DEE) and dipropyl ether (DPE); the sequential
Grotthuss-type proton-transport mechanism of water via inter-
conversion between two hydration complexes, an an Eigen cation
and a Zundel cation, has been experimentally demonstrated.7,8
Received: December 23, 2010
Revised:
March 10, 2011
Published: April 25, 2011
r
2011 American Chemical Society
6023
dx.doi.org/10.1021/jp112210x J. Phys. Chem. B 2011, 115, 6023–6031
|

The Journal of Physical Chemistry B
ARTICLE
Scheme 1. ESPT Mechanism of a Cyclically H-Bonded 1:1
7AIꢀWater Complex in Ethers
solvated cluster species of 7AI, that is, 7AI complexed with protic
solvents via H-bonding, have also attracted a lot of interest.40ꢀ51
Thus, during the last several decades, the ESPT dynamics of 7AI
catalyzed by the H-bonded counterpart of a protic solvent, as well
as the proton-switching mechanism of 7AI dimers, has been in-
vestigated extensively. In particular, the mechanism of the sol-
vent-catalyzed ESPT of 7AI in water and alcohols has been
widely discussed as a two-step model;40ꢀ47 the first step is sol-
vent reorganization to form a cyclically H-bonded 7AIꢀsolvent
complex, and the second step is intrinsic proton transfer. In bulk
water or alcohols, because the solvent reorganization is the rate-
determining step of ESPT, the intrinsic proton transfer is hard to
be observed directly.43ꢀ45 However, in aprotic media such as n-alkanes
and ethers, a 7AI molecule associates with a water molecule to
form a cyclic complex already in the ground state, and thus, the
intrinsic ESPT of the cyclic complex can be explored directly
without being interfered with by solvent reorganization.40
Here, we report the ESPT of cyclically H-bonded 1:1
7AIꢀwater complexes (Nc) in diethyl and dipropyl ethers
(Scheme 1). The rate-determining step of the ESPT of Nc is
the proton acceptance of a water molecule from the pyrrolic
amino group via tunneling to form a hydronium ion, and the
subsequent barrierless proton donation of the transiently formed
hydronium ion to the pyridinic imino group completes the ESPT
without accumulating any intermediate. As a result, we have
found that the ESPT of 7AI along a water monomer is immensely
slowed down compared with the ESPT of 7HQ along a water
dimer.7,8 Whereas a water dimer can transport a proton rapidly
through interconversion between an Eigen cation and a Zundel
cation, a water monomer alone can barely do. Due to such
extraordinarily reduced proton-transporting ability of a water
molecule without having a H-bond network, it is hard for a water
monomer alone to accept a proton from the acidic amino group
of 7AI at the rate-determining step of the ESPT of Nc. Never-
theless, Nc can undergo ESPT in ethers via transiently forming a
charged intermediate species, that is, a deprotonated 7AI anion
H-bonded to a hydronium ion, because polar ether molecules
surrounding Nc can stabilize the charged intermediate species.
Figure 1. Maximum-normalized absorption spectra of 7AI in DEE
(top) and DPE (bottom) having H2O concentrations indicated inside.
measurements of kinetic isotope effects. In every sample, the
concentration of 7AI was kept at 50 μM, which is low enough to
prevent the formation of dimers or oligomers of 7AI; that is,
practically all of the 7AI molecules exist as monomers.
2.2. Measurements. Absorption spectra were measured with
a UV/vis spectrophotometer (Scinco, S3100). Emission spectra
were obtained using a home-built fluorometer consisting of a Xe
lamp of 75 W (Acton Research, XS432) with a monochromator
of 0.15 m (Acton Research, Spectrapro150) and a photomulti-
plier tube (Acton Research, PD438) attached to a monochromator
of 0.30 m (Acton Research, Spectrapro300). To obtain time-
resolved fluorescence kinetic profiles, a mode-locked Nd:YAG
laser (Quantel, Pizzicato) with the pulse duration of 25 ps was
employed for excitation, and a streak camera of 10 ps (Hamamatsu,
C2830) attached to a CCD detector (Princeton Instruments,
RTE128H) was used for detection. Samples were excited with
288 nm pulses generated from a Raman shifter, filled with CH4
gas of 15 atm and pumped by the fourth-harmonic pulses of
266 nm from the laser. Emission wavelengths were selected by
combining band-pass and cutoff filters. Fluorescence kinetic con-
stants were extracted by fitting kinetic profiles to computer-
simulated exponential curves convoluted with instrument response
functions. The sample temperature was controlled using a re-
frigerated bath circulator (Jeio Tech, RC-10V). Unless specified
otherwise, all of the static and kinetic measurements were carried
out at room temperature.
2. EXPERIMENTAL SECTION
2.1. Materials. While 7AI (g99%) was used as purchased
from Sigma-Aldrich, DEE (anhydrous, g99.7%) and DPE
(anhydrous, g99%), purchased from Sigma-Aldrich, were stored
over molecular sieves of 4 Å for 24 h prior to use to remove even a
trace amount of humidity. Water (g18 MΩ cm) distilled
triply with a Milli-Q system (Millipore) was used. The protic
1H (H) atoms of both 7AI and water were exchanged with
2H (D) atoms by dissolving 7AI in DEE or DPE containing D2O
(isotropic purity g 99.9%, received from Sigma-Aldrich) for the
3. RESULTS AND DISCUSSION
3.1.Steady-StateAbsorptionandEmissionSpectra.Figure1
shows the absorption spectra of 7AI in DEE and DPE with a
variation of water content; the lowest absorption band of N in
both DEE and DPE, havingthe bandmaximum at288 nm, shifts to
the red and grows at 310 nm as the concentration of water
increases. The spectral changes of the N-absorption band imply
6024
dx.doi.org/10.1021/jp112210x |J. Phys. Chem. B 2011, 115, 6023–6031

The Journal of Physical Chemistry B
ARTICLE
shifts from 345 nm in neat DEE to 351 nm in DEE containing
200 mM water, while it shifts from 343 nm in neat DPE to
347 nm in DPE having 200 mM of water (Figure S1 in the Sup-
porting Information). This suggests that 7AI monomers are
converted into Nc, which can undergo ESPT to generate T*, as
the concentration of water increases. An isoemissive point at around
470 nm is observed in both DEE and DPE (the insets of Figure 2a),
implying that only two emitting species, that is, N* and T*, exist.
Thus, we infer that no reaction intermediate accumulates during
the ESPT process of N* along a water molecule to yield T*, although
an intermediate state may form transiently (see below). On the
other hand, we have figured out that the intensity of T* fluorescence
is much weaker in 7AIꢀwater complexes than that in 7AIꢀalcohol
complexes.40 This suggests that the proton-transporting solvent,
such as water or an alcohol-mediating ESPT, plays an important
role in determining the ESPT dynamics of 7AI and that the pro-
ton-transporting ability of water is quite different from that of an
alcohol (see below). Chapman and Maroncelli have reported
that T* fluorescence is not observable in neat water because the
nonradiative relaxation of T* is even much faster than the
formation of T* via ESPT of Nc*.43 However, for the ESPT of
monohydrated 7AI, T* fluorescence has been observed, although
its intensity is very weak.46 These imply that the proton-accepting
and proton-donating behaviors of a monomeric water molecule are
quite different from those of bulk water having methodically well-
organized H-bond networks.
We have deduced the formation constant of Nc (KNc) by
monitoring T* fluorescence; KNc can be estimated by eq 1.52
1
c
¼
ꢀ KNc
ð1Þ
C
F ꢀ F0
where F and F0 denote fluorescence intensity at 530 nm mea-
sured at the molar water concentrations of C and 0, respectively,
and c is a proportionality constant. A linear relationship between
1/C and 1/(F ꢀ F0) confirms that a 7AI molecule associates with
a water molecule to form a cyclic 1:1 complex (Figure 2b) and
that the fluorescence of proton-translocated T* originates en-
tirely from Nc*. While KNc values are found to be 11.3 and 5.5 Mꢀ2
in H2O- and D2O-added DEE, respectively, they are 14.3 and
13.7 Mꢀ1 in H2O- and D2O-added DPE, respectively. We have
found out that the observed KNc values in ethers are much
smaller than the previously reported formation constant of a 1:1
7AIꢀmethanol complex in n-heptane (50 Mꢀ1).40 This implies
that H bonds between a 7AI molecule and a water molecule are
weaker than those between a 7AI molecule and a methanol mol-
ecule, and we attribute this to the polar solvent effect of ethers;
DEE (ε = 4.27 cP at 20 °C) and DPE (ε = 3.38 cP at 20 °C) are
much more polar than n-heptane (ε = 1.92 cP at 20 °C).
Moreover, while an alcohol molecule can form H bonds with a
7AI molecule only, a water molecule can form an additional H bond
with an ether molecule besides H bonds with a 7AI molecule.
3.2. Time-Resolved Fluorescence Kinetic Profiles. The
fluorescence kinetic profiles of 7AI in DEE and DPE containing
50 mM of water have been obtained with variation of protic hy-
drogen isotopes as well as monitoring of wavelengths (Figure 3).
Although the kinetic profiles of N* fluorescence monitored at
380 nm show single-exponential decay, those of T* fluorescence
monitored at 550 nm have single-exponential rise and decay. If
N* fluorescence showed double-exponential decay and its fast
decay time matched with the rise time of T* fluorescence, we
could undoubtedly regard the fast decay time as the ESPT time of
Figure 2. (a) Emission spectra, with excitation at 288 nm, of 7AI in DEE
(top) and DPE (bottom) having H2O concentrations indicated inside.
(Insets) Magnified emission spectra to show tautomeric fluorescence.
(b) Plots of [water]ꢀ1 versus (F ꢀ F0)ꢀ1 in DEE (left) and DPE (right),
where F is the fluorescence intensity at 530 nm. The best-fitted lines
yield 10.6 Mꢀ1 with H2O and 5.5 Mꢀ1 with D2O for the formation
constants (KNc) of a 1:1 7AIꢀwater complex in DEE, and they yield
14.3 Mꢀ1 with H2O and 13.7 Mꢀ1 with D2O for KNc in DPE.
that 7AI molecules associate with water molecules via H bonding
to produce 7AIꢀwater complexes in both ethers. In the ground
state, the stable structure of 7AIꢀalcohol complexes in n-alkanes
has been reported to have a cyclically H-bonded 1:1 geometry.40
Similarly, a 7AI molecule is supposed to form a H-bonded cyclic
complex of Nc with a water molecule in aprotic media of ethers
(see below), as shown in Scheme 1. On the basis of the fact that
the absorption band of T has not been observed even at a high
concentration of water (g200 mM), we infer that Nc does not
undergo proton transfer in the ground state. However, the emission
spectra of Figure 2a show that the proton transfer of N* through a
water monomer takes place to produce T*; only one emission band
from N* appears at 345 nm in pure ethers, while a new additional
band from T* grows at around 515 nm with the addition of water.
This indicates that the excited-state tautomerization of 7AI in DEE
and DPE is operative in the presence of water.
With the concentration increment of water dissolved in ethers,
the N* emission band shifts to the red (Figure S1 in the Sup-
porting Information), and the intensity of T* fluorescence increases
gradually (Figure 2a); the band maximum of N* fluorescence
6025
dx.doi.org/10.1021/jp112210x |J. Phys. Chem. B 2011, 115, 6023–6031

The Journal of Physical Chemistry B
ARTICLE
Table 1. Fluorescence Time Constants of 7AI in Water-
Added Ethers
mediuma [water]/mM isotopeb λemc/nm rise time/ps decay time/ps
DEE
0
H
H
380
380
550
380
550
380
550
380
550
380
380
550
380
550
380
550
380
550
d
2600
1880
1860
2320
2020
1250
1380
2140
1950
2970
2070
3350
2850
3300
1480
1880
2670
3500
20
1050
1900
1050
1900
D
H
D
50
DPE
0
H
H
20
1170
2380
1170
2380
D
H
D
50
a The dielectric constants of DEE and DPE are 4.27 and 3.38, respec-
tively, at 20 °C.61 b H and D are protic 1H and 2H atoms, respectively.
c Wavelength of monitored fluorescence. d Instant.
Figure 3. Fluorescence kinetic profiles of 7AI in DEE (top) and DPE
(bottom) having H2O (open) and D2O (closed). The concentration of
water was kept at 50 mM in every sample, which was excited at 288 nm
and monitored at 380 nm (blue) for N* fluorescence and at 550 nm
(red) for T* fluorescence. Solid lines are the best-fitted curves to extract
time constants given in Table 1.
two decay times of N* fluorescence might be observed individu-
ally if they were largely different. For this reason, N* fluorescence
monitored at 380 nm gives only one observable decay time where
the ESPT time of Nc and the relaxation time of excited monomers
or noncyclic complexes are convoluted. Hence, we take the rise
time of T* as the ESPT time of Nc; the ESPT times in DEE are
1050 ps with H and 1900 ps with D, while those in DPE are 1170
ps with H and 2380 ps with D (Table 1).
Nc and the slow decay time as the relaxation time of excited 7AI
monomers or noncyclic 7AIꢀwater complexes that cannot
undergo proton transfer. However, on the basis of the fact that
Figure 2a shows only two emission bands of N* and T* with an
isoemissive point at around 470 nm and that the rise of T*
fluorescence at 550 nm has only one component, we consider
that the ESPT of Nc to produce T* takes place in a concerted
fashion, rather than in a stepwise fashion, without accumulating
any intermediate species and that the rise time of T* fluorescence
corresponds to the ESPT time of Nc (see below).
In aprotic media of DEE and DPE, a 7AI molecule and a water
molecule forms a cyclically H-bonded complex, that is, Nc,
already in the ground state because of the low solubility of water
in ethers. Thus, the intrinsic ESPT of 7AI along a water monomer
in ethers can be observed directly without being interfered by
slow solvation such as solvent reorganization to form Nc. While
the absorption spectra of 7AI in water-added DEE do not show
tautomeric absorption (Figure 1), its emission spectra show T*
fluorescence at around 515 nm (Figure 2). This suggests that
only the ESPT of Nc* produces T*, although Nc has already formed
in the ground state. Moreover, no intermediate species has been
observed in the emission spectra (Figure 2), and all of the ob-
served rise times of T* fluorescence in both ethers are invariant
regardless of changes in water content (Table 1). In these regards,
we attribute the observed rise time of T* fluorescence at 550 nm
to the ESPT of Nc*. The ESPT time in H2O-added DEE has been
observed to be as long as 1050 ps, which is close to the relaxation
time of excited 7AI monomers in DEE without having water
(2600 ps) (Table 1). Accordingly, the ESPT time of Nc and the
relaxation time of excited 7AI monomers or noncyclic 7AIꢀwater
complexes are too similar to be experimentally distinguishable;
The ESPT of 7HQ along a water dimer in DEE has been
reported to take place through interconversion between Eigen
and Zundel cations on a time scale of 38 ps,7,8 which is much
faster than the ESPT of 7HQ along a methanol dimer in n-
heptane (62 ps).34 Because of the unique properties of water where
each molecule can have as many as four H bonds, protons can be
hydrated stably and transferred rapidly through densely well-
ordered H-bond chains in water. The ESPT pathway of 7HQ
along a water dimer is known to be opposite to that of 7HQ along
an alcohol dimer due to different properties of water and an
alcohol, which are primary solvents in proton transport;7,8,34
a
water molecule initiates ESPT by accepting a proton from 7HQ,
whereas an alcohol molecule does so by donating a proton to
7HQ. Similarly, although the proton donation of an alcohol
molecule to the pyridinic imino group of 7AI occurs as the first
step in the ESPT of a 1:1 7AIꢀalcohol complex,40 we consider
that in the ESPT of Nc, that is, a 1:1 7AIꢀwater complex, the
proton acceptance of a water molecule from the pyrrolic amino
group of 7AI takes place first (see below). However, the ESPT
time of 7AI along a water monomer in ethers is 30 times longer
than that of 7HQ along a water dimer in ethers7,8 and 10 times
longer than that of 7AI along an alcohol monomer in n-heptane.40 In
an aqueous solution, proton diffusion occurs very fast via inter-
conversion between Eigen and Zundel cations through system-
atically well-structured H-bond networks; this is well-known as
the proton-hopping mechanism.20 However, a water monomer
6026
dx.doi.org/10.1021/jp112210x |J. Phys. Chem. B 2011, 115, 6023–6031

The Journal of Physical Chemistry B
ARTICLE
in the absence of a H-bond network hardly accepts or donates a
proton because of its extremely weak proton-transporting ability.
Table S1 in the Supporting Information designates that for water,
the proton-donating and the proton-accepting abilities of a mono-
mer (Rm and βm) are enormously smaller than the respective
ones of bulk (R and β). As is well known, the unusual physical
features of water, for example, the high boiling point and the
strong surface tension, originate chiefly from methodically well-
organized huge H-bond networks. Thus, we infer that the dynamics
of water-involved reactions depends largely on the presence and
the size of a H-bond network. On the other hand, the observed
anomalously slow ESPT time also proves that a 7AI molecule
associates with a water molecule to form a 1:1 complex; if a 7AI
molecule associated with two water molecules to form a 1:2
complex, the ESPT dynamics would be observed to be similar to
the ESPT dynamics of a 1:2 7HQꢀwater complex showing a very
fast ESPT time of 38 ps, as reported earlier.7,8 Gas-phase ultrafast
pumpꢀprobe studies have revealed that the ESPT of 7AI
clustered with more than two water molecules takes place within
several picoseconds.50
The Rm value of water is very similar to the Rm value of an
alcohol, whereas the βm value of water is much smaller than the
βm value of any alcohol (Table S1 in the Supporting Infor-
mation). For the ESPT of a 1:1 7AIꢀalcohol complex,40 the rate
constant of ESPT (kpt) tends to increase with Rm of the alcohol
because the rate-determining step is proton transfer from the
alcohol molecule to 7AI; however, the value of kpt for the ESPT
of Nc does not follow this tendency (Figure S2 in the Supporting
Information). For this reason, we have proposed that in contrast
with the ESPT of the 7AIꢀalcohol complex, the ESPT of a 1:1
7AIꢀwater complex is initiated by proton transfer from the
pyrrolic amino group of 7AI to the water molecule and that the
proton acceptance of the water molecule, having very weak βm, to
form a hydronium ion is the rate-determining step. Besides, the
aprotic polar media of DEE and DPE can solvate protons (or hy-
dronium ions) effectively by their electronegative oxygen atoms.
Considering the preferential solvation of cations by ethers and
the well-known simplest continuum theory of ionic solvation,
that is, the Born model55 where the solvation free energy depends
linearly on the reciprocal of the radius of a spherical ion, we infer
that the excess solvation free energy will be larger with a delocalized
anion over a deprotonated 7AI moiety and a localized cation on
the hydronium ion than that in the opposite case, that is, the for-
mation of a protonatedꢀ7AI cation and a hydroxyl ion. Sakota
and Sekiya have suggested that because 7AI becomes acidic in the
first excited singlet state, the πꢀπ* excitation of 7AI makes the
NꢀH bond of the pyrrole ring break easily.56 Furthermore, a
theoretical study on the ESPT of a 1:1 7AIꢀwater complex has
recently reported that in the transition state, the proton of the
pyrrolic amino group of 7AI is located close to the water molecule in
Nc while the proton H-bonded to the pyridinic imino group of
7AI still belongs to the water molecule;48 thus, it is very plausible
that a hydronium ion, rather than a hydroxyl ion, forms tran-
siently during ESPT (see below). Therefore, we suggest that the
ESPT of a 1:1 7AIꢀwater complex is initiated by the proton
acceptance of a water molecule from the pyrrolic amino group of
7AI to form a hydronium ion and a deprotonatedꢀ7AI anion
transiently, as shown in Scheme 1, and that the observed rise time
of T* fluorescence becomes the ESPT time of Nc. On the other
hand, while the rise time of T* fluorescence is as long as 1050 ps
in DEE, its decay time is just 1380 ps (Table 1); although the
observed T* decay time is a typical one,40 it is very close to the T*
Figure 4. Logarithmic values of the rate constants of the T* fluores-
cence rise, that is, the ESPT rate constants of Nc (kpt), versus water
concentrations in DEE (open) and DPE (closed) with variation of protic
hydrogen isotopes (circles for H and squares for D).
rise time that is unusually long. Correspondingly, the intensity of
T* fluorescence at 515 nm is very weak (Figure 2a). Thus, we
consider that the anomalously weak proton-transporting ability
of a water monomer in the absence of a H-bond network brings
about the extraordinarily slow ESPT of Nc to yield very weak T*
emission despite the ordinary decay time of T* fluorescence.
At low concentrations of water (e80 mM), the ESPT time of
Nc is invariant regardless of water content dissolved in ethers
(Table 1 and Figure 4), implying that the collision-induced forma-
tion of Nc does not occur within the lifetime of Nc*. In other
words, Nc only, already existing in the ground state, undergoes
proton transfer upon excitation. However, the ESPT time of Nc
varies with the media of ethers, which are secondary solvents
surrounding Nc; in particular, an ether molecule forms a H bond
with the water monomer of Nc (Scheme 1). We attribute the
different ESPT times of Nc in DEE and DPE to the dissimilar
dielectric constants of two ethers.8 In a polar reaction such as
proton or electron transfer, if the reactant and the product have
different polarities, the reaction rate is generally sensitive to
medium polarity.8,36,53 In this regard, during the ESPT process of
Nc, the polar media of ethers are expected to lower the energy of
an ionic intermediate state that is formed transiently but not
observed (see below). Of note is that the solvation of an in-
termediate species by a solvent medium is completed typically in
1 ps. The ionic intermediate state can be stabilized better in the more
polar medium of DEE (ε = 4.27 at 20 °C) than that in the less
polar medium of DPE (ε = 3.38 at 20 °C), and thus, the more
polar DEE helps Nc* undergo proton transfer relatively more
readily via forming charged intermediate species transiently; the
ESPT of Nc has been observed to occur faster in DPE than in
DEE on that account. Both the acidity, correlated with R, and the
basicity, correlated with β, of the two secondary solvents, that is,
DEE and DPE, are supposed to affect the ESPT rate of Nc as well
as the stabilization of the charged intermediate species. However,
because two ethers have zero acidity and almost the same basicity,54
we consider that the dissimilar ESPT times in two ethers come
mainly from the polarity difference between two secondary solvents.
Thus, in the ESPT of Nc, the polar character of the secondary
solvent (ether) molecule, which is H-bonded to the primary solvent
(water) molecule, enhances the proton-accepting ability of a
monomeric water molecule at the rate-determining step that is
proton transfer from 7AI to water. On the other hand, the exchange
of all of the protic H atoms with D atoms has revealed that the
6027
dx.doi.org/10.1021/jp112210x |J. Phys. Chem. B 2011, 115, 6023–6031

The Journal of Physical Chemistry B
ARTICLE
kinetic isotope effect (KIE), calculated as kpt(H)/kpt(D), of the
ESPT of Nc is 1.8 in DEE and 2.0 in DPE (Table 1 and Table S1
in the Supporting Information). Generally, in the condensed
phase, vibrationally assisted quantum tunneling by heavy-atom
motions helps ESPT to occur facilely, and KIE tends to decrease
with the increment of medium viscosity.34,40 However, the KIE in
the ESPT of Nc is observed to be slightly larger in DPE than that
in DEE, although the viscosity of DPE (η = 0.396 cP) is larger
than that of DEE (η = 0.224 cP). Accordingly, it can be inferred
that medium polarity plays a more important role than medium
viscosity does in the ESPT of 7AI along a monomeric water mol-
ecule. Moreover, the value of KIE in the ESPT of Nc is even much
smaller than that in the ESPT of a 1:1 7AIꢀalcohol complex (Table
S1 in the Supporting Information).40 Thus, we consider that
although the ESPT of Nc occurs via tunneling, there is something
more important than tunneling in the ESPT process because of
the immensely weak proton-accepting ability of a water mole-
cule; that is configurational optimization for efficient proton
tunneling by the reorganization of the H-bond bridge in Nc prior
to intrinsic ESPT (see below).
The value of kpt for the ESPT of Nc in water-added ethers is
independent of water content up to 80 mM, but then, it increases
with the concentration of water (Figure 4 and Table S2 in the
Supporting Information). At a low concentration of water
(e80 mM), the water molecule of Nc has a negligible chance
to be H-bonded to another water molecule, so that it is H-bonded to
an ether molecule to remain as a water monomer. However, at a
high water concentration (>80 mM), the water molecule of Nc
can be H-bonded to another water molecule to form a dimeric
water chain. In other words, whereas the secondary solvent mol-
ecule, which is H-bonded to the water molecule of Nc, is always
an ether molecule at a low concentration of water, it could be a
water molecule at a high concentration of water. We need to note
that the 7AIꢀwater complex having a dimeric water chain at a
high concentration of water still has the cyclic 1:1 geometry, that
is, Nc, rather than the 1:2 geometry (see below) and that the
complex is still surrounded by ether molecules. Thus, an extra
water molecule H-bonded to the water molecule of Nc does not
transport a proton directly in the ESPT of Nc but helps the water
molecule of Nc accept a proton facilely. Of note is that an ether
molecule can form a H bond with a proton or a hydronium ion
but cannot relay a proton on its own. After a proton is transferred
from the acidic amino group of 7AI to the water molecule of Nc,
it cannot be diffused away via the dimeric water chain because Nc
is isolated in an aprotic medium of DEE or DPE. Accordingly, the
proton hydrated in the dimeric water chain has no choice but to
move to the basic group of 7AI.
noteworthy that the reported ESPT time of a 1:1 7AIꢀalcohol
complex is close to that of a 1:2 7HQꢀalcohol complex.34,40 The
ESPT of 7AI clustered with more than two water molecules has
been reported to occur within several picoseconds.50 We con-
sider that the kpt value of Nc increases with the increasing con-
centration of water over 80 mM because the number of 7AIꢀ
water complexes having a dimeric water chain increases. On the
other hand, when deuterated water (D2O) was added to an ether,
the kpt value of Nc remained the same regardless of D2O content
up to the highest concentration in this work. This suggests that
the D2O molecule of Nc has a negligible chance to be D-bonded
to another D2O molecule even at the high D2O concentration of
200 mM because D-bonding ability is generally much weaker
than H-bonding ability. Recall that the KNc value of a 1:1 7AIꢀ
H2O complex is almost two times larger than the value of a 1:1
7AIꢀD2O complex in DEE (Figure 2b). Similarly, the associa-
tion of a 7AIꢀD2O complex with another D2O molecule is
considered to be substantially weaker than the association of a
7AIꢀH2O complex with another H2O molecule. Thus, the con-
centration of 7AIꢀD2O complexes having a dimeric D2O chain
is too low to increase the kpt value of Nc noticeably at the D2O
concentration of e200 mM.
3.3. Proton Inventory Experiments. On the basis of the
result that the excitation of Nc at 288 nm gives rise to only two
emission bands at 345 nm for N* fluorescence and at 515 nm for
T* fluorescence with an isoemissive point at around 470 nm
(Figure 2a), we have suggested that the ESPT of Nc takes place
concertedly without accumulating any intermediate. In the con-
certed process of ESPT, then it would be an issue whether two
hydrogen atoms participating in the ESPT of Nc move synchro-
nously. Proton inventory experiments for the kpt of Nc with a
variation of deuteration degrees of protic hydrogen atoms system-
atically can give a clue to this issue.36,40,57ꢀ59 Nc can have four
different types of isotopic exchange, Hꢀ1N N H (HH),
7
3 3 3
Hꢀ1N N D (HD), Dꢀ1N N H (DH), and Dꢀ1N
7
7
3 3 3
3 3 3
7N D (DD), where the two successive hydrogen atoms in
3 3 3
parentheses denote protic hydrogen atoms bound covalently to
the nitrogen atom of the pyrrole ring and H-bonded to the nitrogen
atom of the pyridine ring in turn (Scheme 1). The ESPT rate
constants of those four types of Nc are denoted as kHH, kHD, kDH
,
and kDD, respectively. Because the rate-determining step is
proton transfer from the pyrrolic amino group (1NꢀH) to the
water molecule of Nc, we can deduce eqs 2 and 3 for kpt according
to the protic hydrogen isotope of the pyrrolic amino group.
HH
d½H ꢂ=dt ¼ ꢀðXHk þ XDkHDÞ½H ꢂ
ð2Þ
ꢁ
ꢁ
With the increasing number of water molecules H-bonded to
the water molecule of Nc, the value of kpt would become larger
because the water molecule of Nc can accept a proton from 7AI
easily with assistance of the H-bonded water molecules; the water
molecule of Nc does not exist as a monomer anymore, so that the
proton-accepting ability of the water molecule of Nc increases.
Nevertheless, even with 200 mM water, the ESPT time of Nc is as
long as 900 ps (Table S2 in the Supporting Information), implying
that the cyclic 1:1 geometry of 7AIꢀwater complexes is still
preserved even in the presence of excess water molecules. If a 7AI
molecule formed a cyclic 1:2 complex with two water molecules,
ultrafast proton transport through interconversion between Eigen
and Zundel cations would take place, and thus, the ESPT time
of a 1:2 7AIꢀwater complex would be similar to the reported
ESPT time of a 1:2 7HQꢀwater complex (38 ps in DEE).7 It is
DH
d½D ꢂ=dt ¼ ꢀðXHk þ XDkDDÞ½D ꢂ
ð3Þ
ꢁ
ꢁ
where * denotes either H or D, XH is [H]/([H] þ [D]), and XD is
1 ꢀ XH. Figure S3 in the Supporting Information shows that kpt
decreases with the increase of XD. In an isolated system such as a
7AIꢀwater complex in an ether, the collision-induced formation
of Nc is too slow to take place; that is, the exchange of any protic
hydrogen isotope in Nc does not occur within the lifetime of Nc*.
The rise component of T* fluorescence giving kpt can be further
decomposed into two rise components of kH and kD as eq 4.
ꢀexpðꢀkpttÞ ¼ ꢀfXH expðꢀkHtÞ þ XD expðꢀkDtÞg ð4Þ
Then, experimentally observed parameters of kH and kD consist
of the above-described four different kpt values according to eqs
6028
dx.doi.org/10.1021/jp112210x |J. Phys. Chem. B 2011, 115, 6023–6031

The Journal of Physical Chemistry B
ARTICLE
Table 2. Comparison of Isotope-Dependent Rate Constants
Extracted From Figure 5
kHD/kDH
kHH/kHD
kHH/kDH
kHD/kDD
kDH/kDD
medium
(R1)
(R2)
(R3)
(R4)
(R5)
DEE
DPE
1.35
1.28
1.16
1.31
1.57
1.68
1.57
1.59
1.16
1.24
transition state, and vibrationally adiabatic energy surfaces in the
first excited singlet state depend largely on isotopic substitutions,
resulting in a large variational effect even at 0 K; the adiabatic
surfaces are flat, and thus, the tunneling effect is not significantly
large. Similarly, our results show that although the ESPT of Nc in
ethers takes place in an asymmetrically concerted manner via
tunneling, tunneling contribution to ESPT is not substantially
large. In other words, there is something more important than
tunneling in the ESPT of Nc because of the extremely small βm
value of a water molecule (see below).
Figure 5. Plots of kH (circles) and kD (squares) with variation of XD
for Nc in DEE (opens) and DPE (closed). The total concentration of
water was kept at 50 mM in every sample. Solid lines are the best linear
fittings to obtain kHH, kHD, kDH, and kDD as (1060 ps)ꢀ1, (1230 ps)ꢀ1
(1660 ps)ꢀ1, and (1930 ps)ꢀ1, respectively, in DEE and as (1180 ps)ꢀ1
(1550 ps)ꢀ1, (1980 ps)ꢀ1, and (2460 ps)ꢀ1, respectively, in DPE.
,
,
To understand the ESPT mechanism of Nc at the molecular
level and the medium polarity effect on the ESPT dynamics, we
need to compare the ratios given in Table 2 thoroughly. First, we
have figured out that kHD/kDH (R1) in DEE (1.35) is larger than
R1 in DPE (1.28). As described above, if the ESPT of Nc took
place concertedly and symmetrically, R1 should be unity. How-
ever, R1 is found to be significantly larger than unity in both DEE
and DPE, so that the ESPT of Nc in ethers has been suggested to
occur concertedly and yet asymmetrically. Thus, we can estimate
the asymmetric degree of the ESPT reaction by considering how
much the value of R1 deviates from unity. On the basis of the
result that the value of R1 is larger in DEE than that in DPE, we
suggest that the asymmetric character of ESPT is stronger in
DEE than that in DPE. We attribute this to the larger dielectric
constant of DEE (4.27 at 20 °C) than that of DPE (3.38 at
20 °C); the more polar DEE molecules lower the energy of the
ionic intermediate state, which forms transiently during ESPT, as
shown in Scheme 1, better than the less polar DPE molecules do.
Second, we have found out that either one of kHH/kHD (R2) or
5 and 6, respectively.
kH ¼ kHH þ ðkHD ꢀ kHHÞXD
ð5Þ
ð6Þ
kD ¼ kDH þ ðkDD ꢀ kDHÞXD
As shown in Figure 2b, the value of KNc varies considerably
with protic hydrogen isotopes, so that we have calibrated XH and
XD considering different KNc values with H and D. When XD is
0.5, its calibrated value becomes 0.342 in DEE and 0.489 in DPE.
The linear correlations of kH and kD with calibrated XD, shown in
Figure 5, yield kHH, kHD, kDH, and kDD as (1060 ps)ꢀ1, (1230
ps)ꢀ1, (1660 ps)ꢀ1, and (1930 ps)ꢀ1, respectively, in DEE and
(1180 ps)ꢀ1, (1550 ps)ꢀ1, (1980 ps)ꢀ1, and (2460 ps)ꢀ1, respec-
tively, in DPE. If the ESPT of Nc occurs concertedly and syn-
chronously, the rule of the geometric mean that kHD = kDH
=
(kHHkDD 1/2 should hold.40,57ꢀ59 However, the rule is not valid
)
in both DEE and DPE, indicating that two protons participating
in the ESPT of Nc do not move at the same time but move in
succession. Considering results described so far, thus, we con-
clude that the ESPT of 7AI via a water molecule occurs in an
asymmetrically concerted fashion; two protons taking part in the
ESPT of Nc move asynchronously, although ESPT occurs con-
certedly. Such asymmetric character usually appears in a con-
certed reaction when a single transition state is involved in the
rate-determining step with forming an intermediate transiently
but without accumulating the intermediate.7,8,34,40
A number of theoretical researchers have carried out calcula-
tions for the ESPT of 7AIꢀwater or 7AIꢀalcohol complexes at
various numbers of proton-transporting solvent (water or an
alcohol) molecules.48,49 To explore the ESPT mechanism of 7AI
in detail, they have studied how many solvent molecules are
required to form the most stable cyclic complex, which species
between a cationic intermediate and an anionic intermediate is
preferred to form during ESPT, and which geometry the inter-
mediate species has in the transition state. Recently, Duong and
Kim have performed a detailed ab initio reaction dynamics study
on the ESPT of a 1:1 7AIꢀwater complex using variational
transition-state theory calculations including multidimensional
tunneling.48 They have found that two protons participating in
the H-bond bridge of the cyclic 1:1 7AIꢀwater complex are
transferred concertedly and yet asynchronously. The position of
the variational transition state is very different from the conventional
k
k
DH/kDD (R5) is significantly smaller than either one of kHH
/
DH (R3) or kHD/kDD (R4) in both DEE and DPE, although R2
and R5 are very similar to each other and R3 and R4 are also close
to each other. Those four ratios of R2ꢀR5 represent kinetic
single-isotope effects (KSIEs); R3 and R4 indicate the KSIEs of
the hydrogen atom bound covalently to the nitrogen atom of the
pyrrole ring, while R2 and R5 signify the KSIEs of the hydrogen
atom H-bonded to the nitrogen atom of the pyridine ring. The
result that R3 and R4 are larger than R2 and R5 designates that
the ESPT rate is dependent mostly on the protic hydrogen isotope
of the pyrrolic amino group regardless of the other hydrogen isotope
H-bonded to the pyridinic nitrogen atom; that is, observed KIE
values are correlated primarily with the first single-proton transfer
from 7AI to the water monomer of Nc. In this regard, the ob-
served KIE values of 1.8 and 2.0 in DEE and DPE, respectively,
are not small for single-proton transfer, implying that the first
single-proton transfer occurs via tunneling (see below). Even-
tually, these results support our suggestion that the double-proton
transfer of Nc* takes place in an asymmetrically concerted manner
involving the rate-determining tunneling process; the first proton
transfer occurs via tunneling through a barrier at the rate-determin-
ing step, and the second proton transfer takes place barrierlessly.
3.4. Arrhenius Plots. Figure 6 shows the Arrhenius plots of
kpt for the ESPT of Nc in water-added DEE, giving activation
6029
dx.doi.org/10.1021/jp112210x |J. Phys. Chem. B 2011, 115, 6023–6031

The Journal of Physical Chemistry B
ARTICLE
ESPT of a 1:1 7AIꢀwater complex is reduced, resulting in a
small KIE.
4. CONCLUSIONS
The extraordinarily slow excited-state proton-transfer (ESPT)
dynamics of 7-azaindole (7AI) along a water molecule has been
investigated in aprotic polar media of diethyl ether (DEE) and
dipropyl ether (DPE). The ESPT of a cyclically hydrogen
(H)-bonded 1:1 7AIꢀwater complex (Nc) in ethers occurs in an
asymmetrically concerted fashion, rather than in a stepwise manner,
via forming a charged intermediate species transiently with no
accumulation of the intermediate. The ESPT of Nc is initiated by
the proton acceptance of the water molecule from the pyrrolic
amino group via tunneling to form a hydronium ion, which is the
rate-determining step, and completed by the subsequent barrier-
less proton donation of the hydronium ion to the pyridinic imino
group. Whereas proton diffusion in an aqueous solution having
methodically well-organized H-bond networks has been known
to take place swiftly through interconversion between Eigen and
Zundel cations, proton transport by the water molecule of Nc has
been observed to be extremely slow (∼1 ns) because a water
monomer has very weak proton-accepting ability in the absence
of a H-bond network. The ESPT rate constant of Nc is larger in
DEE than that in DPE because the more polar DEE molecules
help the water monomer of Nc accept a proton relatively more
facilely from 7AI at the rate-determining step by lowering the
energy of ionic intermediates, that is, a deprotonated 7AI anion
and a hydronium ion, which form transiently during ESPT. The
asymmetric character of ESPT is stronger in DEE than that in
DPE due to the larger dielectric constant of DEE. The ESPT rate
is invariant regardless of water content dissolved in an ether up to
80 mM, but then, it increases with the water content. On one
hand, at a low concentration of water (e80 mM), because the water
molecule of Nc has a negligible chance to form a H bond with
another water molecule, it is H-bonded to an ether molecule.
However, at a high concentration of water (>80 mM), the water
molecule of some Nc can be H-bonded to another water molecule
to form a dimeric water chain, and thus, it can accept a proton readily
from 7AI because of its enhanced proton-accepting ability. On
the other hand, the ESPT rate is independent of deuterated water
(D2O) content up to the highest concentration of D2O in this
work (200 mM). This suggests that the association of a 7AIꢀD2O
complex with another D2O molecule is much weaker than the
association of a 7AIꢀH2O complex with another H2O molecule
due to the lower hydrogen-bond ability of the D2O molecule.
Small but significant kinetic isotope effects of ESPT have been
observed, while the observed activation energy of the ESPT of Nc
with H as a protic hydrogen isotope is similar to that with D. In
conclusion, because of the anomalously weak proton-transport-
ing ability of a water monomer without having a H-bond network,
the elaborate reorganization of the H-bond bridge in Nc to form
an optimized precursor configuration with proper H-bond dis-
tances and suitable bond angles is necessary for proton tunneling,
resulting in anomalously slow proton transport of a water molecule.
Figure 6. Arrhenius plots for kpt of 7AI in DEE having 50 mM water.
The best-fit lines reveal that while the activation energy (Ea) and the pre-
exponential factor (Af) with H as protic hydrogen atoms (open) are 0.96
kcal molꢀ1 and 6.0 ꢃ 109 sꢀ1, respectively, Ea and Af with D (closed) are
1.06 kcal molꢀ1 and 2.7 ꢃ 109 sꢀ1, respectively.
energies (Ea) and preexponential factors (Af). Although the value
of kpt with H is significantly larger than the value with D, deduced
Ea values with H and D are very similar to each other and quite
small. This implies that KIE, that is, kpt(H)/kpt(D), is affected
chiefly by the ratio of preexponential factors, Af(H)/Af(D). Actually,
the observed KIE value for the ESPT of Nc in DEE (1.8) is close
to the Af(H)/Af(D) ratio (2.2). On the other hand, KIE is in-
dependent of temperature within our experimental errors, although
kpt is dependent on temperature to some extent (Figure 6). These
types of Arrhenius plots have been reported in solvent-mediated
and enzymatic proton-transfer reactions.7,8,34,40,60 In general,
temperature-independent and large KIEs in enzymes have been
explained with a model employing vibrationally enhanced proton
tunneling. When the tunneling contribution to ESPT is large
enough to be equally effective for both H and D, the ratio of
Af(H)/Af(D) would become even much greater than unity.
However, our observed KIE values for the ESPT of Nc (1.8 in
DEE and 2.0 in DPE) are even much smaller than the values for
the ESPT of 1:1 7AIꢀalcohol complexes in n-heptane.40 None-
theless, we consider that the KIE values are substantial for single-
proton transfer, and thus, tunneling is operative through an ap-
preciable barrier. Within the limit of the BornꢀOppenheimer
approximation, the potential energy hardly changes with isotope
substitution, so that KIE originates mainly from changes of nuclear
motions on the same potential energy surface.
Usually, configurational optimization in which solvent fluctua-
tions play the main role is a prerequisite to proton tunneling, and
solvent fluctuations decrease with the decrement of temperature
due to the reduced solvent motions at low temperature. Accord-
ingly, in the ESPT of Nc, the reorganization of the H-bond bridge
with a little activation energy is required to form an optimized
precursor configuration having proper H-bond distances and
angles for pretunneling. As the temperature decreases, the kpt value
decreases somewhat because configurational optimization, which
allows a proton to tunnel through the potential energy barrier, is
retarded. Whereas intrinsic proton transfer via tunneling is isotopi-
cally sensitive, configurational optimization for pretunneling is
isotopically insensitive; that is, configurational optimization is
orthogonal to intrinsic proton transfer in the reaction coordi-
nates of the potential hypersurface. Consequently, due to the
extremely weak proton-transporting ability of a monomeric water
molecule, more elaborate configurational optimization is required
for proton tunneling, and thus, the tunneling contribution in the
’ ASSOCIATED CONTENT
S
Supporting Information. Proton-accepting and proton-
b
donating abilities, fluorescence time constants, steady-state emis-
sion spectra, kpt versus Rm of proton-transporting solvents, and
isotope-composition-dependent fluorescence kinetic profiles.
6030
dx.doi.org/10.1021/jp112210x |J. Phys. Chem. B 2011, 115, 6023–6031

The Journal of Physical Chemistry B
ARTICLE
This material is available free of charge via the Internet at http://
pubs.acs.org.
(30) Park, S.-Y.; Yu, H.; Park, J.; Jang, D.-J. Chem.—Eur. J. 2010,
16, 12609.
(31) Park, S.-Y.; Jang, D.-J. J. Am. Chem. Soc. 2010, 132, 297.
(32) Park, S.-Y.; Lee, Y.-S.; Jang, D.-J. Phys. Chem. Chem. Phys. 2008,
’ AUTHOR INFORMATION
10, 6703.
(33) Yu, H.; Kwon, O.-H.; Jang, D.-J. J. Phys. Chem. A 2004,
108, 5932.
(34) Kwon, O.-H.; Lee, Y.-S.; Yoo, B. K.; Jang, D.-J. Angew. Chem.,
Int. Ed. 2006, 45, 415.
Corresponding Author
*E-mail: djjang@snu.ac.kr. Tel: þ82-(2)-880-4368. Fax: þ82-(2)-
875-6624.
(35) Tolbert, L. M.; Solntsev, K. M. Acc. Chem. Res. 2002, 35, 19.
(36) Kwon, O.-H.; Zewail, A. H. Proc. Natl. Acad. Sci. U.S.A. 2007,
104, 8703.
’ ACKNOWLEDGMENT
(37) Takeuchi, S.; Tahara, T. Proc. Natl. Acad. Sci. U.S.A. 2007,
104, 5285.
This work was financially supported by research grants through
the National Research Foundation of Korea by the Ministry
of Education, Science, and Technology (2010-0015806 and
2011-0001216). S.-Y.P. acknowledges the Seoul fellowship and
the BK21 scholarship as well.
(38) Sekiya, H.; Sakota, K. J. Photochem. Photobiol. C 2008, 9, 81.
(39) Folmer, D. E.; Wisniewski, E. S.; Hurley, S. M.; Castleman,
A. W. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12980.
(40) Kwon, O.-H.; Lee, Y.-S.; Park, H. J.; Kim, Y.; Jang, D.-J. Angew.
Chem., Int. Ed. 2004, 43, 5792.
(41) Kwon, O.-H.; Jang, D.-J. J. Phys. Chem. B 2005, 109, 8049.
(42) Kwon, O.-H.; Jang, D.-J. J. Phys. Chem. B 2005, 109, 20479.
(43) Chapman, C. F.; Maroncelli, M. J. Phys. Chem. 1992, 96, 8430.
(44) Mente, S.; Maroncelli, M. J. Phys. Chem. A 1998, 102, 3860.
(45) Chen, Y.; Gai, F.; Petrich, J. W. J. Am. Chem. Soc. 1993,
115, 10158.
’ REFERENCES
(1) Geissler, P. L.; Dellago, C.; Chandler, D.; Hutter, J.; Parrinello,
M. Science 2001, 291, 2121.
(2) Chen, H.; Voth, G. A.; Agmon, N. J. Phys. Chem. B 2010,
114, 333.
(46) Chou, P.-T.; Martinez, M. L.; Cooper, W. C.; McMorrow, D.;
Collins, S. T.; Kasha, M. J. Phys. Chem. 1992, 96, 5203.
(47) Chou, P.-T.; Yu, W.-S.; Wei, C.-Y.; Cheng, Y.-M.; Yang, C.-Y.
J. Am. Chem. Soc. 2001, 123, 3599.
(3) Mohammed, O. F.; Pines, D.; Nibbering, E. T. J.; Pines, E. Angew.
Chem., Int. Ed. 2007, 46, 1458.
(4) Siwick, B. J.; Cox, M. J.; Bakker, H. J. J. Phys. Chem. B 2008,
112, 378.
(48) Duong, M. P. T.; Kim, Y. J. Phys. Chem. A 2010, 114, 3403.
(49) Chaban, G. M.; Gordon, M. S. J. Phys. Chem. A 1999, 103, 185.
(50) Folmer, D. E.; Wisniewski, E. S.; Stairs, J. R.; Castleman, A. W.
J. Phys. Chem. A 2000, 104, 10545.
(51) Smirnov, A. V.; English, D. S.; Rich, R. L.; Lane, J.; Teyton, L.;
Schwabacher, A. W.; Luo, S.; Thornburg, R. W.; Petrich, J. W. J. Phys.
Chem. B 1997, 101, 2758.
(5) Tanner, C.; Thut, M.; Steinlin, A.; Manca, C.; Leutwyler, S.
J. Phys. Chem. A 2006, 110, 1758.
(6) Bhattacharyya, K. Chem. Commun. 2008, 2848.
(7) Park, S.-Y.; Lee, Y.-S.; Kwon, O.-H.; Jang, D.-J. Chem. Commun.
2009, 926.
(8) Park, S.-Y.; Kim, B.; Lee, Y.-S.; Kwon, O.-H.; Jang, D.-J. Photo-
chem. Photobiol. Sci. 2009, 8, 1611.
(52) Tanimoto, Y.; Itoh, M. Chem. Phys. Lett. 1978, 57, 179.
(53) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265.
(54) Kamlet, M. J.; Abboud, J.-L. M.; Abraham, M. H.; Taft, R. W.
J. Org. Chem. 1983, 48, 2877.
(9) Park, S.-Y.; Kwon, O.-H.; Lee, Y.-S.; Jang, D.-J. J. Phys. Chem. A
2009, 113, 10589.
(10) (a) Park, S.-Y.; Kwon, O.-H.; Kim, T. G.; Jang, D.-J. J. Phys.
Chem. C 2009, 113, 16110. (b) Park, S.-Y.; Lee, Y.-S.; Jang, D.-J. Phys.
Chem. Chem. Phys. 2011, 13, 3730–3736.
(55) Born, M. Z. Phys. 1920, 1, 45.
(56) Sakota, K.; Sekiya, H. J. Phys. Chem. A 2009, 113, 2663.
(57) Schowen, R. L. Angew. Chem., Int. Ed. Engl. 1997, 36, 1434.
(58) Klein, O.; Aguilar-Parrilla, F.; Lopez, J. M.; Jagerovic, N.;
Elguero, J.; Limbach, H.-H. J. Am. Chem. Soc. 2004, 126, 11718.
(59) Gerritzen, D.; Limbach, H.-H. J. Am. Chem. Soc. 1984, 106, 869.
(60) Kohen, A.; Cannio, R.; Bartolucci, S.; Klinman, J. P. Nature
1999, 399, 496.
(11) Jeffrey, G. A. An Introduction to Hydrogen Bonding; Oxford
University Press: Oxford, U.K., 1997; pp 135ꢀ153.
(12) Rahman, A.; Stillinger, F. H. J. Am. Chem. Soc. 1973, 95, 7943.
(13) Agmon, N. Chem. Phys. Lett. 1995, 244, 456.
(14) Lapid, H.; Agmon, N.; Petersen, M. K.; Voth, G. A. J. Chem.
Phys. 2005, 122, 014506.
(15) Markovitch, O.; Chen, H.; Izvekov, S.; Paesani, F.; Voth, G. A.;
Agmon, N. J. Phys. Chem. B 2008, 112, 9456.
(16) Mohammed, O. F.; Pines, D.; Dreyer, J.; Pines, E.; Nibbering,
E. T. J. Science 2005, 310, 83.
(61) CRC Handbook of Chemistry and Physics, 78th ed.; Lide, D. R.,
Ed.; CRC: Boca Raton, FL, 1999.
(17) Mathias, G.; Marx, D. Proc. Natl. Acad. Sci. U.S.A. 2007, 104,
6980.
(18) Wraight, C. A. Biochim. Biophys. Acta 2006, 1757, 886.
(19) Swanson, J. M. J.; Maupin, C. M.; Chen, H.; Petersen, M. K.;
Xu, J.; Wu, Y.; Voth, G. A. J. Phys. Chem. B 2007, 111, 4300.
(20) Marx, D. ChemPhysChem 2006, 7, 1848.
(21) de Grotthuss, C. J. T. Ann. Chim. 1806, 58, 54.
(22) Huckel, E. Z. Elektrochem. Angew. Phys. Chem. 1928, 34, 546.
(23) Stearn, A. E.; Eyring, H. J. Chem. Phys. 1937, 5, 113.
(24) Bernal, J. D.; Fowler, R. H. J. Chem. Phys. 1933, 1, 515.
(25) Huggins, M. L. J. Phys. Chem. 1936, 40, 723.
(26) Marx, D.; Tuckerman, M. E.; Hutter, J.; Parrinello, M. Nature
1999, 397, 601.
(27) Eigen, M. Angew. Chem., Int. Ed. Engl. 1964, 3, 1.
(28) Zundel, G.; Metzger, H. Z. Phys. Chem., Neue Folge 1968,
58, 225.
(29) Lill, M. A.; Helms, V. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 2778.
6031
dx.doi.org/10.1021/jp112210x |J. Phys. Chem. B 2011, 115, 6023–6031