1. Introduction
Excited-state intramolecular proton transfer (ESIPT) reactions have received growing attention due to their wide range of potential applications [
1,
2,
3,
4]. Typically, the ESIPT process is represented by a keto-to-enol tautomerization that occurs via a proton transfer reaction initiated by photoexcitation with short-laser pulses [
5,
6]. A detailed understanding of ESIPT is, therefore, crucial to achieve manipulation of the ESIPT dynamics for various applications [
7,
8,
9], such as developing photo-responsive materials and bio-fluorescent markers.
The structure of 4-aminophthalimide (4-AP) contains electron acceptor moieties (>NH and –NH
2) that are conjugated with an electron donor moiety (>C=O) in an aromatic ring. Consequently, its electron density changes significantly upon electronic excitation [
10]. 4-AP can form a six-membered ring hydrogen bonding configuration with protic solvents, and this configuration can serve as a precursor for solvent-assisted ESIPT (
Scheme 1). In addition, 4-AP is highly fluorescent in nonpolar solvents and its fluorescence quantum yield remains high (0.76–0.63) regardless of the polarity in aprotic solvents; however, the fluorescence yield drops dramatically in protic solvents [
11]. The fluorescence spectra and lifetimes of 4-AP are also sensitive to the molecular environment [
10,
12,
13,
14,
15,
16]. Furthermore, 4-AP shares some important structural features with the nucleic acid bases, including the ability to undergo hydrogen bonding with protic solvent molecules and with itself. It has also been reported that 4-AP is approximately isosteric to tryptophan, a weakly fluorescent probe used in peptides and proteins, although the former exhibits a stronger fluorescence [
17]. Consequently, 4-AP has been extensively employed as a strong fluorescent probe in various environments, including biological systems, nanomaterials, and polymers, since it can be readily incorporated into larger molecules without any significant changes in its fluorescence properties [
10,
16,
18,
19,
20].
4-AP is known to exhibit strong solvatochromism. This was attributed to the significant increase in the dipole moment of its excited state relative to the ground state [
21], the formation of stronger hydrogen bonds with protic solvents in the excited state [
12,
22], and/or the formation of the enol form of 4-AP (Enol-4-AP) as a result of solvent-assisted ESIPT [
10,
23,
24,
25]. ESIPT in 4-AP was revoked by showing the same behavior as 4-AP in 4-AP derivatives wherein an n-alkyl group replaces the imide hydrogen atom [
26]. Later, Durantini et al. reported that 4-AP undergoes solvent-assisted ESIPT from the imino/amino group to the carbonyl group, leading to a keto–enol tautomerization in the excited state following excitation at 300 nm [
10]. However, ESIPT of 4-AP in the excited state by excitation at 300 nm was also revoked by a comparative experiment on 4-AP and 4-AP derivatives wherein the hydrogen atom of the imino group is replaced by an
n-butyl group [
27]. Although the controversial observation of ESIPT in 4-AP has been extensively investigated, the molecular mechanism that defines the feasibility of this process has yet to be examined.
The electronic spectrum of 4-AP consists of two distinctive bands that lead to two different excited states (S
2 and S
1). While many studies have reported the solvatochromism behavior of 4-AP upon excitation to the S
1 state, where the dipole moment and charge distribution change significantly upon excitation and, thus labeled ‘charge-transferred state’ [
10,
27], little efforts are made for the excitation to the S
2 state where the dipole moment changes to a lesser extent (i.e., the non-charge transferred state). Previously, excitation to the S
2 state has been suggested to result in keto–enol tautomerization, which was not observed upon excitation to the S
1 state [
10,
27].
As expected, the choice of solvent plays a critical role in the excited-state dynamics of 4-AP. Protic solvents, such as methanol, can form hydrogen bonds with solute molecules, significantly influencing their electronic states and relaxation pathways. In addition, hydrogen bond interactions can stabilize specific excited states and facilitate processes such as ESIPT. In contrast, aprotic solvents, including acetonitrile, lack hydrogen bonding capabilities, providing a different interaction landscape for the solute. By comparing the behaviors of 4-AP in these two solvent environments, we can gain insights into the molecular mechanism of how solvent–solute interactions modulate the dynamics of the associated excited states.
Since molecular structural dynamics after photoexcitation are quite crucial for understanding the excited state properties, direct experimental observation of the structural dynamics is necessary to unveil the mechanistic details of environment-induced excited state properties. In this context, time-resolved infrared (TRIR) spectroscopy can detect structural dynamics in their excited states, in addition to proton coordination during the ESIPT processes. This is possible because TRIR spectroscopy is sensitive not only to the molecular structure but also to the solvation environment [
28,
29]. Femtosecond TRIR spectroscopy has been utilized in obtaining a molecular picture of the photophysical and photochemical pathways involved in the excitation of interested compounds in solution [
30,
31,
32,
33,
34].
In this study, the dynamics of photoexcited 4-AP in methanol and acetonitrile are investigated using femtosecond TRIR spectroscopy to explore the mechanistic feasibility of ESIPT in 4-AP. Excitation wavelengths of 300 and 350 nm are used to excite the 4-AP molecule to the S2 and S1 states, respectively, and the resulting structures are probed with mid-IR pulses in the spectral region of 1800–1450 cm−1, where the spectral signatures of the related compounds are distinctive. Deuterated solvents (i.e., CD3OD and CD3CN) are subsequently used to shift the solvent absorption away from the spectral window of interest. Moreover, a solution 4-AP in CH3OH is also studied in a limited spectral region to test the isotope effect in the presence of ESIPT; this limited spectral region is required due to overlap with the strong solvent absorption. Detailed excitation dynamics are obtained to explain the negligible ESIPT of 4-AP in protic solvents.
2. Results and Discussion
The UV-Vis absorption spectra of 4-AP in both undeuterated and deuterated methanol and acetonitrile are shown in
Figure 1a. It can be seen that the use of a deuterated solvent did not alter the UV-Vis spectra, implying that the electronic transitions of 4-AP are not affected by deuteration. In both solvents, the absorption spectra consist of two prominent bands that correspond to two different transitions, namely the band at ~310 nm, which is assigned to the S
0 → S
2 (ππ*) transition, and the band at ~360 nm, which is attributed to the S
0 → S
1 intramolecular charge transfer (ICT) transition [
10]. Although the position of the ~310 nm band is similar in both solvents (~2 nm difference), the position of the band at ~360 nm changes significantly, exhibiting a 12 nm difference between acetonitrile and methanol. This observation is consistent with previous reports, wherein band positions of 306.2 and 357.1 nm were reported for acetonitrile, and positions of 308.2 and 369.2 nm were reported in methanol [
10]. For 4-AP, the transition occurring at ~360 nm is known to result in a significant change in the dipole moment, which is manifested as a large solvatochromism [
10] Using time-dependent density functional theory (TD-DFT) calculations with the ωB97XD functional and the pcseg-1 basis set, the dipole moment and electrostatic potential (ESP) of 4-AP were determined in the S
0, S
1, and S
2 states. According to these calculations, the dipole moment of 4-AP changes by 6.0 D for the S
1 state and by 3.7 D for the S
2 state upon excitation. The ESP of 4-AP shown in
Figure 1b displays that the charge distribution in the S
1 state is considerably more polarized compared to the S
0 and S
2 states. Although methanol and acetonitrile have similar polarity with dielectric constants of 32.63 and 36.69, respectively, at 298K [
35,
36], only the protic solvent methanol can form a hydrogen bond with 4-AP. Therefore, the larger redshift of the S
1 peak of 4-AP in methanol was attributed to a greater stabilization of the more polarized S
1 state by the hydrogen bonding of 4-AP with methanol. The S
1 peak is significantly more sensitive to the hydrogen bonding ability of the solvent than the S
2 peak [
10].
Figure 2 shows the equilibrium Fourier transform infrared (FT-IR) spectra of 4-AP in undeuterated and deuterated acetonitrile and methanol at room temperature. It was found that CH
3CN and CH
3OH exhibit strong absorption bands at ~1400 cm
−1 due to the CH
3 bending mode, which also produces an absorbance of >0.5, even at 1550 cm
−1 under the current conditions. As a result, the measured sample absorbance is unreliable below 1550 cm
−1 and so the data for this spectral region are not shown for 4-AP in CH
3CN and CH
3OH. Consequently, deuterated solvents were used to extend the spectral region up to 1450 cm
−1. Although the deuterated solvent altered the vibrational spectrum of 4-AP in methanol, the vibrational spectrum of 4-AP in CD
3CN was almost identical to that in CH
3CN. Based on these results, it is apparent that the electronic and vibrational transitions of 4-AP are not affected by the deuteration of acetonitrile; thus, only the solution of 4-AP in CD
3CN was considered hereafter for the sake of simplicity. In contrast, solutions of 4-AP in both CD
3OD and CH
3OH are investigated due to the differences in their vibrational spectra. The two major bands at 1763 and 1703 cm
−1 in CD
3OD, 1763 and 1718 cm
−1 in CH
3OH, and 1763 and 1726 cm
−1 in CD
3CN correspond to the symmetric and asymmetric C=O stretching modes of 4-AP, respectively. The band positions for the symmetric C=O stretching mode are identical in acetonitrile and methanol, whereas the bands corresponding to the asymmetric C=O stretching mode were found to be solvent dependent. DFT calculations performed for 4-AP in the presence of one methanol molecule showed that methanol is hydrogen bonded to 4-AP (optimized structure in
Figure 2). In addition, these calculations show that the asymmetric C=O stretching mode undergoes a red-shift, indicating that this band originates from hydrogen bonding between methanol and the C=O and imine moieties of 4-AP. The asymmetric C=O band of 4-AP in CD
3OD is even more red-shifted than that in CH
3OH due to the deuteration of the 4-AP amine and imine groups. As can be seen in
Figure 2, the signals corresponding to NH
2 bending of 4-AP in CD
3CN and CH
3OH (i.e., at 1635 and 1649 cm
−1, respectively) are absent in CD
3OD due to the deuteration of the amine group. The remaining bands at 1618, 1598, and 1504 cm
−1 in CD
3OD, 1617 and1593 cm
−1 in CH
3OH, and 1617, 1597, and 1501 cm
−1 in CD
3CN correspond to the C=C stretching modes of the 4-AP benzene ring. The small band at ~1750 cm
−1 (red dashed lines in
Figure 2) was assigned to a combination of NH
2 scissoring (or C–N–C stretching in CD
3OD) and OC–N–CO scissoring. These distinctive bands of 4-AP in the 1800–1450 cm
−1 region were used as reference peaks in the vibrational analysis of the excited state due to the fact that they are observable in the same region across all IR spectra, even in the presence of the tautomerized compound.
One key objective of this work was to establish a detailed picture of the roles of different electronic states in the solvent-assisted ESIPT process of 4-AP in protic solvents. The electronic absorption spectra (
Figure 1) demonstrated the possibility of selective photoexcitation from the S
0 state of 4-AP to either the S
2 or S
1 state. The TRIR spectra of 10 mM 4-AP solutions in CD
3OD and CH
3OH were recorded at 293 K over the spectral ranges of 1800–1450 and 1800–1550 cm
−1, respectively. More specifically, the spectra were recorded from 0.3 ps to 100 ns after excitation with a 350 or 300 nm pulse, and the resulting two-dimensional (2D)-contour maps are displayed in
Figure 3 and
Figure 4. The TRIR spectra clearly display evolving negative-going (blue) and positive-going (red) features. The negative-going features (bleach), appearing immediately after photoexcitation and having the same band positions as the equilibrium absorption spectrum of 4-AP, arise from population depletion of the ground state upon photoexcitation. The amplitudes of the bleach bands decreased over time, essentially disappearing within 100 ns, suggesting that the depleted ground-state population recovered within this timeframe. Some new absorption features also developed immediately after photoexcitation, while others appeared later. However, all absorption features were observed to decay during the bleach recovery time, suggesting that all excited states relax or new species return to the reactant within the 100 ns period.
The TRIR spectra were globally fitted using the basis spectra shown in
Figure 5 by adjusting the amplitude of the basis spectra whilst maintaining the center frequency and width of each band in the spectra constant. The basis spectra were assigned to specific chemical species based on quantum calculations and possible reaction dynamics. In addition to the equilibrium spectrum of 4-AP (denoted as S
0), the use of two additional basis spectra that were assigned to 4-AP in the S
1 state was sufficient to reproduce the TRIR spectra at 350 nm. More specifically, these spectra included one basis spectrum (denoted as S
1-0) for 4-AP in the S
1 state immediately after photoexcitation and before the solvent reorganizes to a new electronic configuration of the S
1 state, and a second basis spectrum (denoted as S
1-1) for 4-AP in the S
1 state with the fully optimized solvent configuration for the S
1 state. The TRIR spectra at 300 nm, which were excited to the S
2 state, also required three basis spectra, namely S
0, S
1-1, and S
2 shown in
Figure 5. The two basis spectra S
0 and S
1-1 for the TRIR spectra at 300 nm were the same as those used to fit the TRIR spectra at 350 nm. Although the S
1-0 and S
1-1 basis spectra were required for the excitation of 4-AP to the S
1 state, one basis spectrum was sufficient for the excitation of 4-AP to the S
2 state. This can be attributed to the fact that the spectrum for the S
2 state immediately after photoexcitation is approximately the same as that for 4-AP in the S
2 state with the fully adjusted solvent configuration. As can be seen in
Figure 1b, the ESP of 4-AP in the S
2 state is not significantly different from that in the S
0 state, and thus, the solvent configuration does not change significantly after the transition of S
0 → S
2, resulting in little difference between the spectra in the S
2 state before and after optimization of the solvent configuration. In the case of excitation to the S
1 state, the ESP changes dramatically; thus, the spectra before and after optimization (i.e., S
1-0 and S
1-1) are required. In the global fitting of both TRIR spectra at 350 and 300 nm in a given solvent, two basis spectra, S
0 and S
1-1 were common. As shown in
Figure 3 and
Figure 4, the TRIR spectra at 350 and 300 nm in CD
3OD and CH
3OH were well reproduced by the sum of the three basis spectra (i.e., S
0, S
1-0, and S
1-1 for 350 nm; S
0, S
1-1, and S
2 for 300 nm) of the corresponding solvent in
Figure 5.
The vibrational frequencies of 4-AP in various electronic states were calculated using the DFT method with the ωB97XD functional and the pcseg-1 basis set. The basis spectrum of S0 represents the equilibrium FT-IR spectrum of 4-AP in methanol. The S2, S1-0, and S1-1 basis spectra were obtained by optimizing the calculated vibrational frequencies of 4-AP in the corresponding states. The calculated vibrational spectra consisted of major bands corresponding to the C=O and C=C stretching modes. The position and width of each band were optimized to globally fit the TRIR spectra. The basis spectra of S1-0 and S2 were optimized during global fitting of the TRIR spectra at 350 and 300 nm, respectively, and the basis spectrum of S1-1 was optimized during simultaneous global fitting of the two TRIR spectra at 350 and 300 nm. Initially, the integrated area of each band in the basis spectra was kept proportional to the calculated oscillator strength in the global fitting and was then slightly optimized in the final fitting.
The spectrum for S1-0 was calculated using the solvent configuration in the S0 state, whilst the 4-AP molecule was in the S1 state. The spectrum for S1-1 was calculated after the solvent configuration was fully optimized for 4-AP in the S1 state using TD-DFT. The same notation as that for the S1 state can be used for 4-AP in the S2 state, i.e., S2-0 and S2-2 were calculated immediately after excitation from the S0 state to the S2 state, and with the fully optimized solvent configuration in the S2 state, respectively. The spectrum of S2-2 was indistinguishable from that of S2-0; thus, the spectra were denoted as S2 in both cases. The Gibbs free energy of 4-AP for the S2-2 state was calculated to be 0.11 kcal/mol lower than the S2-0 state, resulting in the S2-2 spectrum being insignificantly different from that of the S2-0 state. However, the Gibbs free energy of 4-AP for the S1-1 state was calculated to be 1.8 kcal/mol lower than that of the S1-0 state, resulting in the S1-1 spectrum being significantly different from that of the S1-0 state. The insignificant difference between the spectra of S2-0 and S2-2 allowed a single S2 spectrum to be used in both cases.
To explore the possible formation of Enol 4-AP after photoexcitation via solvent-assisted ESIPT, the spectra of Enol 4-AP were calculated in the S
1 state before and after full optimization of the solvent configuration to Enol 4-AP in the S
1 state, denoted as Enol S
1-0 and Enol S
1-1, respectively. The basis spectra denoted as S
0, S
1-0, S
1-1, and S
2 represent the spectra of the keto form of 4-AP in various electronic states. As shown in
Figure 5, the calculated spectral features of Enol 4-AP in the S
1 state are quite different from those of the keto form, being significantly red-shifted and showing congestion in the lower frequency region. These observations indicate that, when present, Enol 4-AP can be readily distinguishable in the TRIR spectra. No distinct absorption features were observed in the TRIR spectral region < 1540 cm
−1 (see
Figure 3) wherein the presence of Enol 4-AP should be evident. This indicates that ESIPT does not proceed in a methanolic solution of 4-AP excited at either 350 or 300 nm.
Global fitting of the TRIR spectra using the basis spectra resulted in time-dependent amplitude changes in the basis spectra, which reflected population changes in the corresponding species after the photoexcitation of 4-AP. The amplitude of the basis spectrum represents the sum of the integrated areas of the bands in the basis spectra used to fit the TRIR spectra and is proportional to the sum of the integrated extinction coefficients of the bands in the spectrum multiplied by the population of the corresponding species. Thus, the time-dependent population changes in the species can be obtained from the recovered time-dependent amplitude changes in the basis spectra used to fit the TRIR spectra once the integrated extinction coefficients of the corresponding basis spectra are determined [
37].
The kinetic model (
Figure 6b) was used to describe the reaction dynamics of excited 4-AP in methanol. This model was selected due to its ability to reproduce the time-dependent amplitude changes in the basis spectra obtained from the global fitting of the TRIR spectra at both 350 and 300 nm. The pump wavelength-dependent and time-dependent fractional population changes were obtained by simultaneously fitting the time-dependent amplitude changes for both the 350 and 300 nm excitations to the kinetic model by adjusting the integrated extinction coefficients of all species and rate constants for all processes between species. As shown in the figure, the time-dependent fractional population changes in all the species involved in the photoexcitation of 4-AP at 350 and 300 nm, as obtained from the corresponding time-dependent amplitude changes in the basis spectra and the fitted integrated extinction coefficients of the corresponding species, were well reproduced by the kinetic model with the time constants shown in the
Figure 6b. It can be seen from
Figure 6b that the S
1-1 → S
0 relaxation is common in a given solvent for excitation at both 350 and 300 nm. In other words, the time constants of the S
1-1 → S
0 relaxation are the same for 4-AP in a given solvent excited at both 350 and 300 nm. Time constants for the S
2 → S
1-1 and the S
1-0 → S
1-1 relaxations are identical, which happens to be the same value of 13 ± 2 ps, for 4-AP in both CD
3OD and CH
3OH, while the S
1-1 → S
0 relaxation was found to be ~2.3-times faster in CH
3OH than CD
3OD. The fitted relative integrated extinction coefficients obtained from
Figure 6b were 0.45 ± 0.02, 0.95 ± 0.02, and 0.94 ± 0.02 for the S
2, S
1-1, and S
1-0 spectra relative to that of S
0, respectively. Notably, these correlate with the calculated values of 0.45, 0.95, and 0.94 for the S
2, S
1-1, and S
1-0 states, respectively.
As can be seen in
Figure 6b, the solvent configuration is fully optimized to 4-AP in the S
1 state with a time constant of 13 ± 2 ps. This state remains until it relaxes into the S
0 state with time constants of 14 ± 2 and 6 ± 2 ns in CD
3OD and CH
3OH, respectively. Since all excited 4-AP species are in the S
1 state upon 350 nm excitation, the excited 4-AP at this wavelength undergoes a simple relaxation to the S
0 state following picosecond optimization of the solvent configuration to 4-AP in the excited S
1 state (
Figure 6b). The excitation of 4-AP at 350 nm represents an ICT transition, which results in a significant change in the dipole moment (∆μ ≈ 6 D) and the ESP (
Figure 1). Immediately after photoexcitation, 4-AP reaches the S
1 state, residing in non-relaxed solvent environments since the solvent molecule configuration remains in the S
0 state. As the solvent molecules become reorganized based on the new charge distribution of 4-AP in the S
1 state, 4-AP becomes stabilized. The time constant of 13 ± 2 ps for the S
1-0 → S
1-1 transition, therefore, represents the solvent reorganization time of methanol and is comparable to the dielectric relaxation time of methanol measured by the time-resolved fluorescence Stokes shift method (i.e., 17 ps) [
38], thereby consolidating the current experimental data. The solvent reorganization time of CD
3OD was identical to that of CH
3OH, as was its internal conversion time from the S
2 to the S
1 state of 4-AP. As shown in
Figure 5, solvent reorganization manifested as spectral evolution and dynamic peak shifts of the bands in the S
1-0 to S
1-1 basis spectra. Fluorescence decay measurements identified the excited state lifetime of 4-AP in CH
3OH to be ~7 ns [
11], which was proposed to be ~3 times lower than that in CD
3OD [
27]. Therefore, the S
1-1 lifetimes in CD
3OD and CH
3OH (i.e., 14 ± 2 and 6 ± 2 ns, respectively) were consistent with the previously reported values.
The fluorescence quantum yield of 4-AP in CH
3OH was reported to be 0.1 [
39], while that in CD
3OD has been reported as ~0.3 [
27]. The longer lifetime of the S
1 state of 4-AP in CD
3OD supports a higher fluorescence quantum yield, indicating that the shorter lifetime of the S
1 state in CH
3OH arises from faster nonradiative relaxation. The solvent deuterium effect on the fluorescence lifetime was attributed to the participation of the solvent proton in the nonradiative decay channels of the excited molecule [
27].
According to
Figure 6b, the 4-AP S
2 state generated upon 300 nm excitation relaxes into the S
1 state via internal conversion with a time constant of 13 ± 2 ps. Optimization of the solvent configuration to the S
2 state was undetectable because the spectrum of the S
2 state with the optimized solvent configuration was indistinguishable from that with the pre-optimized solvent configuration. Because the internal conversion time from the S
2 to the S
1 state is comparable to, happens to be the same as, the solvent reorganization time of the S
1 state, the spectrum for the S
1 state with a pre-optimized solvent configuration (S
1-0) is not observable during internal conversion from the S
2 to the S
1 state. Therefore, the simple relaxation scheme of S
2 → S
1-1 → S
0 is sufficient for describing the decay of 4-AP excited to the S
2 state.
It has been reported that substituting hydrogen for deuterium preserves the electrostatic potential but significantly slows the ESIPT process [
40]. In the current study, the relaxation times of the initially excited states at 300 or 350 nm were identical in both CD
3OD and CH
3OH. If proton transfer was involved in these relaxation processes, the time constant would depend on solvent deuteration. However, the independence of the initial relaxation time on solvent deuteration confirmed that proton transfer was not involved in the initial evolution of the TRIR spectra at 300 and 350 nm.
Considering the structure presented in
Figure 6a, wherein 4-AP is hydrogen bonded to a methanol molecule in the ground state, with CH
3OH bridging between the C=O and imide N–H moieties of 4-AP, the excited 4-AP possesses an optimal precursor configuration for solvent-assisted ESIPT. Furthermore, the charge of the accepting oxygen increases (see the ESPs in
Figure 1), thereby strengthening the hydrogen bonds in the 4-AP–methanol complex and enhancing the probability of proton transfer upon excitation to the S
1 state [
24,
41]. Therefore, the S
1 state of 4-AP is favored in the context of solvent-assisted ESIPT. However, as shown in
Figure 6, the calculated Gibbs free energy of the Enol S
1 state is ~10 kcal/mol higher than that of the keto-form S
1 state. Consequently, the lack of solvent-assisted ESIPT in the S
1 state of 4-AP arises for energetic reasons. The majority of ESIPT processes have been observed for enol-to-keto tautomerizations due to the fact that the excited enol is higher in energy [
4,
42]. This indicates that the ESIPT of 4-AP requires an uncommon keto-to-enol tautomerization to proceed [
40]. The higher energy of Enol 4-AP in the S
1 state is consistent with the majority of molecules possessing higher energy in this form and state. Upon the excitation of 4-AP to the S
2 state, the energy of which is higher than that of the Enol S
1 state, the energetics of ESIPT are satisfied. However, the proton transfer process from the S
2 state should compete with the S
2 → S
1 internal conversion with a time constant of 13 ± 2 ps. In other words, proton transfer should proceed with a comparable time of 13 ps to be successful, requiring the acidity and basicity of the related atoms to increase and the hydrogen bonding configuration with the solvent molecule to be optimized. However, the ESP of the S
2 state is similar to that of the S
1 state, indicating that hydrogen bonding is not dramatically strengthened in the S
2 state compared to the S
0 state. Therefore, proton transfer is slower than the internal conversion time of 13 ps, resulting in a negligible ESIPT for 4-AP in the S
2 state.
To differentiate between the possible solvent-assisted ESIPT of 4-AP in the excited S
2 state, the dynamics of 4-AP were investigated in a polar aprotic solvent excited at 300 nm. More specifically, the TRIR spectra of a 10 mM 4-AP solution in CD
3CN were obtained between 1800 and1450 cm
−1 and at 293 K over a wide time range of 0.3 ps to 100 ns after excitation with a 300 nm pulse; the corresponding 2D contour map is shown in
Figure 8. As described above, the early dynamics of 4-AP are expected to be the same in CD
3CN and CH
3CN because the electronic and vibrational spectra of 4-AP are not affected by deuteration. Since the spectral window is broader in CD
3CN, experiments were performed for a solution of 4-AP in this solvent. From
Figure 8, it can be seen that the negative-going features, which appeared immediately after photoexcitation, matched well with the bands in the equilibrium FT-IR spectrum of 4-AP in CD
3CN; positive-going features, appearing immediately or later, evolved over time. The TRIR spectra were globally fitted using the three basis spectra presented in
Figure 9a, namely the equilibrium FT-IR spectrum of 4-AP in CD
3CN (S
0), the spectrum of 4-AP in the S
1 state (S
1), and the spectrum of 4-AP in the S
2 state (S
2). Since the dielectric relaxation time of acetonitrile is known to be ~0.2 ps [
43], the solvent reorganizes immediately to the new electronic configuration upon electronic excitation. Consequently, all of the observed spectra for 4-AP in CD
3CN exhibited a fully optimized solvent configuration in the electronic state; thus, all spectra recorded in acetonitrile were denoted as S
n.
Time-dependent amplitude changes in the basis spectra, obtained from global fitting of the TRIR spectra, were analyzed in the same manner as for 4-AP in methanol using
Figure 6b. Since CD
3CN does not form hydrogen bonds with 4-AP, and such bonding is required to enable the ESIPT of 4-AP, the enol form is not produced. As shown in
Figure 9b, the time-dependent fractional population changes in all species involved in the photoexcitation of 4-AP in CD
3CN at 300 nm are well reproduced by
Figure 6b shown in
Figure 6b.
According to
Figure 6b, all 4-APs in the 300 nm excited S
2 state decay to the S
1 state via internal conversion with a time constant of 9 ± 2 ps, while the 4-AP in the S
1 state decays to the S
0 state with a time constant of 10 ± 2 ns, indicating that the excited state life time of the S
1 state of 4-AP in CD
3CN is 10 ± 2 ns. This S
1 lifetime is consistent with the reported value of 14 ns measured using the fluorescence decay time of 4-AP in CH
3CN excited at 375 nm [
22]. When excited to the S
2 state, 4-AP undergoes the same dynamics in both protic and aprotic solvents, namely simple relaxation to the ground state via the S
1 state. These results, therefore, confirm that no ESIPT proceeds even for excitation to the S
2 state.