Article pubs.acs.org/JPCA
Intersystem Crossing Pathway in Quinoline−Pyrazole Isomerism: A Time-Dependent Density Functional Theory Study on Excited-State Intramolecular Proton Transfer Yu-Hui Liu,*,†,‡ Sheng-Cheng Lan,† Chaoyuan Zhu,*,‡ and Sheng-Hsien Lin‡ †
Department of Physics, College of Mathematics and Physics, Bohai University, Jinzhou 121013, China Department of Applied Chemistry, Institute of Molecular Science and Center for Interdisciplinary Molecular Science, National Chiao-Tung University, Hsinchu 30050, Taiwan
‡
ABSTRACT: The dynamics of the excited-state intramolecular proton-transfer (ESIPT) reaction of quinoline−pyrazole (QP) isomers, designated as QP-I and QP-II, has been investigated by means of time-dependent density functional theory (TDDFT). A lower barrier has been found in the potential energy curve for the lowest singlet excited state (S1) along the proton-transfer coordinate of QP-II compared with that of QP-I; however, this is at variance with a recent experimental report [J. Phys. Chem. A 2010, 114, 7886−7891], in which the authors proposed that the ESIPT reaction would only proceed in QP-I due to the absence of a PT emission for QP-II. Therefore, several deactivating pathways have been investigated to determine whether fluorescence quenching occurs in the PT form of QP-II (PT-II). The S1 state of PT-II has nπ* character, which is a well-known dark state. Moreover, the energy gap between the S1 and T2 states is only 0.29 eV, implying that an intersystem crossing (ISC) process would occur rapidly following the ESIPT reaction. Therefore, it is demonstrated that the ESIPT could successfully proceed in QP-II and that the PT emission would be quenched by the ISC process. methods by Dunietz and coworkers.29 It is notable that their calculations demonstrate that the effect of conjugation can inhibit the hydrogen-atom transfer reaction at both the ground and excited states. Recently, a series of quinoline−pyrazole (QP) isomers with different π-conjugation structures (see scheme 1 in ref 30; the structures of QP-I and QP-II are shown in Figure 1) have been synthesized and subjected to spectroscopic measurements by Chung and coworkers.30 It is interesting that only QP-I displays a proton-transfer emission at ∼560 nm in its emission spectrum. Despite their similar molecular structures, no other emissions are seen for QP-II and QP-III besides the normal fluorescence. Chung and coworkers30 concluded that only QP-I undergoes the ESIPT process. The absence of ESIPT in QP-III may be readily understood because it adopts a nonplanar structure due to steric hindrance. For QP-II, however, it may be premature to rule out an ESIPT reaction solely on the basis of its lack of proton-transfer emission and quenching of its excited state. A number of organic molecules with a heterocyclic group formed by an intramolecular hydrogen bond have been investigated to gain insight into the dynamics of ESIPT,31−35 and different internal conversion (IC) rates have been found in many molecules showing ESIPT, such as 2-(2′-hydroxyphenyl)benzothiazole (HBT),31,32 TINUVIN-P,33 and 2-(2′-hydroxyphenyl)-4-methylthiazole.34 In particular, for TINUVIN-P, the
1. INTRODUCTION Excited-state intramolecular proton transfer (ESIPT) is a phototautomerization process,1−7 which has been extensively studied to gain insight into fundamental photophysical and photochemical processes because of their potential applications in optical devices as well as the fundamental importance of proton transfer in chemistry and biology.8−11 The intramolecular hydrogen bond between proton donor and acceptor plays an important role in the ESIPT reaction;12−16 however, little is known about electronic excited-state hydrogen bonds because their structures and dynamics are difficult to analyze by both theoretical and experimental means. Recently, Zhao and Han determined theoretically that intermolecular hydrogen bonds between solute and alcohol molecules can be significantly strengthened in the electronic excited state upon photoexcitation.17−22 Within this excited-state hydrogen-bond strengthening theory, we have previously studied the dynamics of excited-state proton-transfer (ESPT) reactions of several photoacid molecules such as 6-hydroxyquinoline.23−26 The important role of excited-state hydrogen bonding has been revealed, and novel mechanisms have been proposed. The π-conjugation structure would play an important in the photochemistry processes. It has been demonstrated that the barriers of the excited-state intramolecular hydrogen-atom transfer can be effectively tuned by the π-conjugation structure.27,28 Moreover, the conjugation effects on the thermodynamics of ground-state and lowest-singlet intermolecular double hydrogen-atom transfer reactions in 7-azaindole have been investigated with ab initio electronic structure © XXXX American Chemical Society
Received: April 13, 2015 Revised: May 30, 2015
A
DOI: 10.1021/acs.jpca.5b03557 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
excited-state energies, we checked the S1 → S0 emission energies for excited-state QP-I, QP-II, PT-I, and PT-II (as shown in Figure 1) by using the second-order approximate coupled-cluster (CC2) method40 at the TZVP level with the resolution-of-the-identity (RI) approximation.41,42 This was performed with the TURBOMOLE program suite.43,44
3. RESULTS AND DISCUSSION To clarify how π conjugation affects the ESIPT of QPs, we used DFT and TDDFT methods to calculate the structures of QP-I and QP-II in the ground state and the lowest singlet excited state at the TZVP/B3LYP level. It is known that B3LYP can be quite accurate for aromatic molecules, but it also can fail sometimes in an unpredictable manner when a charge-transfer (CT) state is involved.45,46 Recently, a series of so-called longrange-corrected (LC) functionals have been widely used to investigate the excited states of large molecules.47−50 The Coulomb operator in LC functional has been separated into long- and short-range parts to correct the underestimate of the energy of the excited CT state by conventional TDDFT. Hence, an LC-corrected CAM-B3LYP functional was also employed to improve the credibility of our work. Simulated absorption and emission spectra of QP-I obtained with both the B3LYP and CAM-B3LYP functionals are shown in Figure 2 and are compared with the experimental results.30
Figure 1. Optimized geometric structures of QP-I, QP-II, and their PT forms PT-I and PT-II in the ground state as well as the denotations of some atoms.
IC time is as short as 150 fs, which would lead to effective quenching of the excited state; however, a recent experimental study by femtosecond transient infrared spectroscopy by Batista and coworkers35 suggests that cis/trans isomerization is not involved in the ESIPT process of HBT. Their results strongly negate the notion of an ultrafast IC pathway in the 90° twisted structure of excited HBT proposed by Barbatti and coworkers.32 Hence, all possible nonradiative transition processes should be considered to reveal the mechanism of the ESIPT process. In this work, we focus on the influence of the different directions of π-conjugation in ESIPT reaction; however, both the grounded and excited QP-III present the nonplanar structures,30 which suggests that the QP-III is not comparable to the ESIPT reaction of QP-I and QP-II due to the broken πconjugation. Therefore, the QP-III is not considered in the work. The excited-state dynamics of the QP-I and QP-II molecules has been studied by DFT/TDDFT at the CAMB3LYP/TZVP level. To ascertain whether the ESIPT reaction could occur in QP-II, we have calculated potential energy curves for different electronic states of both QP-I and QP-II along the ESIPT coordinate. The results reveal that the PT reaction of QP-II could occur favorably in the S1 state with a lower barrier than that for QP-I. To assess the pathway of electronic deactivation leading to fluorescence quenching of QP-II following the PT reaction in the S1 state, we also calculated the potential energy curves associated with twisting motion of the products of the ESIPT reaction as well as the energies of the corresponding triplet states.
Figure 2. Simulated emission spectra of QP-I and its PT form PT-I calculated by using the B3LYP (dashed line) and CAM-B3LYP (line) functionals. The inset shows the simulated absorption of QP-I, which was obtained by expanding the calculated emission energies in a Lorentzian shape and applying the corresponding coefficients. The vertical lines denote the corresponding peaks observed experimentally.30
2. THEORETICAL METHODS Ground-state geometry optimizations of all species in this work were performed by using density functional theory (DFT) with Becke’s three-parameter hybrid exchange function and the Lee−Yang−Parr gradient-corrected correlation functional (B3LYP functional)36 as well as its long-range corrected functional CAM-B3LYP.37 The triple-ζ valence quality with one set of polarization functions (TZVP) was chosen as the basis set throughout the calculations.38 The excited-state electronic structures were calculated by time-dependent density functional theory (TDDFT) at both the B3LYP/TZVP and CAM-B3LYP/TZVP levels. All of these electronic structure calculations were carried out using the Gaussian 09 program suite.39 Additionally, to obtain more accurate values of the
For the absorption spectra of QP-I, the simulated results with the B3LYP functional are in good agreement with the experimental absorption band localized the region of 310− 350 nm.30 The emission of QP-I in the S1 state is calculated to be 366 and 324 nm by using B3LYP and CAM-B3LYP, respectively. The peaks calculated with the CAM-B3LYP functional are all slightly blue-shifted. It should be caused by the long-range correction, which can avoid the underestimate of the energy of the excited CT state by conventional TDDFT but also rise up the energies of non-CT state. When the ESIPT reaction is associated with charge transfer, the B3LYP functional predictably fails. The excitation energy of PT-I was calculated as 919 nm, at great variance with the 570 nm in the B
DOI: 10.1021/acs.jpca.5b03557 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A experimental emission spectrum;30 however, the peak at 576 nm simulated by CAM-B3LYP correctly characterizes PT-I in the S1 state. Hence, although the peaks in the absorption and emission spectra of QP-I are slightly different from the experimental spectra, CAM-B3LYP would clearly be the better choice in this work. Moreover, to confirm the credibility of CAM-B3LYP, the excitation energies of QP-1, PT-1, QP-II, and PT-II were also calculated by RICC2, which is a higher quality method but also more expensive in terms of computing resources. As listed in Table 1, all of the results obtained with
similar to the variation in the C1−C2 bond length in the ESIPT reaction of HBT reported by Batista and coworkers.35 To assess whether ESPT would occur in the QP-II molecule, potential energy (PE) curves along the H−N2 bond for different electronic states of both QP-I and QP-II were calculated and are shown in Figure 3. Each point in the Figures
Table 1. Calculated S1 → S0 Emission Energies (in nm) of QP-I, QP-II, PT-I, and PT-II Obtained using TDDFT (with B3LYP and CAM-B3LYP Functionals) and RICC2 Methods As Well As the Corresponding Emission Peaks in the Experimental Spectra30 species
B3LYP
CAM-B3LYP
RICC2
exp.
QP-I PT-I QP-II PT-II
366 919 353 768
324 576 322 477
325 651 314 462
362 570 338
CAM-B3LYP are much closer to those obtained with RICC2, thus proving the credibility of the CAM-B3LYP functional. Therefore, CAM-B3LYP was used for all further calculations in this work. The structures of QP-I and QP-II in the ground state and the lowest singlet excited state as well as the products of their ESIPT reactions, PT-I and PT-II, in the lowest singlet excited state were optimized and are shown in Figure 1. The populations of non-hydrogen-bonded species were expected to be
