Chinese Journal of Chemical Physics  2018, Vol. 31 Issue (6): 749-755

#### The article information

Ya-ping Wang, Chun-hua Li, Bing Zhang, Chen Qin, Song Zhang

Ultrafast Investigation of Excited-State Dynamics in Trans-4-methoxyazobenzene Studied by Femtosecond Transient Absorption Spectroscopy

Chinese Journal of Chemical Physics, 2018, 31(6): 749-755

http://dx.doi.org/10.1063/1674-0068/31/cjcp1806155

### Article history

Accepted on: August 28, 2018
Ultrafast Investigation of Excited-State Dynamics in Trans-4-methoxyazobenzene Studied by Femtosecond Transient Absorption Spectroscopy
Ya-ping Wanga, Chun-hua Lia, Bing Zhangb, Chen Qinc, Song Zhangb
Dated: Received on June 26, 2018; Accepted on August 28, 2018
a. School of Electronic Science & Applied Physics, Hefei University of Technology, Hefei 230009, China;
b. State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences, Wuhan 430071, China;
c. Key Laboratory of Mineral Luminescent Material and Microstructure of Xinjiang, School of Physics and Electronic Engineering, Xinjiang Normal University, Urumqi 830054, China
*Author to whom correspondence should be addressed. Chen Qin, E-mail:zhangsong@wipm.ac.cn; Song Zhang, E-mail:qinchen5555@sina.com
Abstract: The ultrafast photoisomerization and excited-state dynamics of trans-4-methoxyazobenzene (trans-4-MAB) in solutions were investigated by femtosecond transient absorption spectroscopy and quantum chemistry calculations. After being excited to the S$_2$ state, the two-dimensional transient absorptions spectra show that cis-4-MAB is produced and witnessed by the permanent positive absorption in 400$-$480 nm. Three decay components are determined to be 0.11, 1.4 and 2.9 ps in ethanol, and 0.16, 1.5 and 7.5 ps in ethylene glycol, respectively. The fast component is assigned to the internal conversion from the S$_2$ to S$_1$ state. The other relaxation pathways are correlated with the decay of the S$_1$ state via internal conversion and isomerization, and the vibrational cooling of the hot S$_0$ state of the cis-isomer. Comparing of the dynamics in different solvents, it is demonstrated that the photoisomerization pathway undergoes the inversion mechanism rather than the rotation mechanism.
Key words: Photoisomerization    Trans-4-methoxyazobenzene    Femtosecond transient absorption spectroscopy
Ⅰ. INTRODUCTION

The $trans$-$cis$ photoisomerization reaction of azobenzene and its derivatives has attracted significant attention in optical information storage, nonlinear optics, molecular switches, artificial photocontrolled proteins and DNA hybridization [1-10]. The high potential of azobenzene compounds rests on the large changes in molecular structure, dipole moment, refractive index, absorption spectrum and dielectric constant by photoisomerization [11]. Therefore, to reach many ambitious rational design goals for azobenzene devices, it is necessary to have detailed knowledge about reaction dynamics of the trans-cis photoisomerization behavior.

The photoisomerization dynamics of azobenzene [12-22] and its derivatives [23-31] has been investigated by quantum chemistry calculation and various spectroscopies in recent decades. For azobenzene, there exist an intense $\pi\pi^*$ band (the S$_2$ state) peaked at 316 nm and a weak n$\pi^*$ band (the S$_1$ state) around 440 nm [12]. In the 1980s, Rau proposed that upon n$\pi^*$ excitation, $trans$-azobenzene isomerized via inversion with in-plane bending of a CNN bond; while upon $\pi\pi^*$ excitation, a torsional reaction channel around the N=N double bond became active [13, 14]. However, Tahara and coworkers studied the excited-state dynamics on the S$_2$ state of $trans$-azobenenzene under $\pi\pi^*$ excitation by femtosecond time-resolved fluorescence absorption, they concluded that no isomerization occurred in the S$_2$ state or during S$_2$$\rightarrowS_1 relaxation and proposed inversion for the isomerization mechanism in the S_1 state regardless of difference in initial photoexcitation [15]. As demonstrated here, there is a controversy about the isomerization dynamics with the \pi\pi^* excitation. For azobenzene derivatives, the different substituent groups have been proven to lead the red shift of the \pi\pi^* state [23-31]. Ikeda and co-workers researched a liquid crystalline trans-azobenzene (4-butyl-4-methoxyazobenzene, abbreviated as BMAB) in solution by femtosecond transient absorption and picosecond single-photon timing fluorescence spectroscopies. Their results indicated that a fast internal conversion from the S_2 (\pi\pi^*) state to the S_1 (n\pi^*) state occurred within 200 fs and the photoisomerization occurring on the S_1 state was completed within 10 ps [23]. However, Reid et al. studied the trans-4-dimethylaminoazobenzene (trans-4-DMAAB) with the \pi\pi^* state excitation by subpicosecond pump-probe spectroscopy. They observed a bi-exponential decay with time constants of \sim0.8 and \sim10 ps which are assigned to the lifetime of the lowest-lying \pi\pi^* state and the ground-state vibrational relaxation, respectively [24]. Moreover, Hirose et al. determined that the n\pi^* state of trans-4-aminoazobenzene (trans-4-AAB) was populated from the initially excited \pi\pi^* state within 0.2 ps and then decayed with \sim0.6 and \sim2 ps which cannot be distinguished from the corresponding decay channels [25]. These studies show that different substituent groups can also affect isomerization dynamics with the \pi\pi^* excitation. Recently, a methoxyl group substituted azobenzene, trans-4-methoxyazobenzene (trans-4-MAB) is of particular interest because of its many potential applications in portable information systems and electronics [32, 33]. The trans-4-MAB can be doped into a nematic liquid crystal to make the material which contains the 45^{\circ} twisted nematic (TN) and the photo-induced isotropic (PHI) states. Both states can be rapidly switched to each other by the isomerization [32]. Norikane and co-workers have reported that crystals of trans-4-methoxyazobenzene also can be used to make artificial self-propelled systems with the aid of photoisomerization [33]. However, the isomerization dynamics of the trans-4-MAB in solution is still not clear. In this work, we investigate the ultrafast excited state dynamics and photoisomerization of trans-4-MAB in ethanol and ethylene glycol with the \pi\pi^* excitation by femtosecond transient absorption spectroscopy combined with quantum chemical calculations. The spectra are measured until the delay time up to 300 ps to obtain more complete dynamics information. The quantum chemical calculations are used to present the molecular orbitals, oscillator strength and static absorption spectra. Following the excitation at 345 nm, ultrafast excited-state dynamics associated with the photoisomerization of trans-4-MAB are observed and analyzed in detail. Ⅱ. EXPERIMENTAL AND COMPUTATIONAL DETAILS Commercially available trans-4-MAB (98% purity) was used for experiments without further purification. Ethanol and ethylene glycol (99% purity) with viscosities of 1.2 and 13.5 mPa\cdots at 20 ℃ were used as solvents [34]. The fresh sample was prepared for transient absorption with concentration of 1 mmol/L at room temperature. The steady absorption spectra were recorded with concentration of 0.25 mmol/L on the UV-Vis spectrometer (INESA, L6) in a 1-mm quartz cell. Femtosecond transient absorption apparatus which is based on the Ti:sapphire laser system has been described elsewhere [26, 27]. Briefly, the seed beam is generated by a commercial Ti:sapphire oscillator pumped by a CW second harmonic of an Nd:YVO_4 laser, and then is amplified by an Nd:YLF pumped regenerative amplifier to generate a 1-kHz pulse train centered at 800 nm with 35-fs pulse width and the energy up to 1 mJ/pulse as fundamental pulse. A fraction of the fundamental laser was focused into a 1 mm thick BBO crystal, yielding pulses at 400 nm with energy of 130 μJ, which are used to pump the NOPA. The pulse with \sim610 nm from the NOPA needs to be temporally compressed in order to obtain the minimum pulse width compatible with their bandwidth. The excitation wavelength of 345 nm with energy of \sim1.5 μJ was generated by sum frequency of the 800 nm fundamental pulse and the \sim610 nm from the compressed NOPA. The other fraction of the fundamental light at 800 nm was focused into a 1 mm sapphire to generate a white continuum (390-720 nm) which would be split into the probe and the reference beam by a metallic-coated beamsplitter. The pump and probe pulses were overlapped into the sample cell at an angle of <4^{\circ} and the reference beam was focused on the sample at a different spot. For eliminating polarization and photoselection effects, the relative polarization between the pump and probe beam was set at the magic angle (54.7^{\circ}) for all the measurements. A computer-controlled linear translation stage was used to set the pump-probe time delay. Each step from the linear translation stage represented the time delay of \sim2.08 fs. The transmitted probe and reference spectra were taken by a CCD camera (PI-MAX) equipped with a spectrometer (Princeton, SpectraPro 2500i). The pump-induced stimulated Raman scattering (SRS) and cross-phase modulation (XPM) signals were measured independently for the pure solvent. All the transient absorption spectra were obtained by subtracting the SRS and XPM contributions. The instrumental response function (IRF) of the system, determined by the SRS signal, was typically better than 200 fs. The temporal resolution in our experiments was estimated to be 40 fs which was about 20% of IRF. All quantum chemical calculations were performed by the Gaussian 09W software package [35]. The ground geometries of the trans- and cis-4-MAB were optimized by ab initio density-functional theory with the B3LYP function and 6-311G++(d, p) basis set. The stationary points were verified by the vibrational frequencies analysis. The effect of the bulk solvent was considered by the polarizable continuum model (PCM) [36]. The molecular orbitals, vertical excitation energy, oscillator strengths, and absorption spectroscopy were also performed by the TD-DFT/B3LYP for trans-4-MAB and cis isomer. Ⅲ. RESULTS AND DISCUSSION A. Quantum chemical calculations The two lowest excited singlet states of trans-4-MAB in ethanol and ethylene glycol were performed by DFT calculations, respectively, as listed in Table Ⅰ. The vertical excitation energies of the S_1 and S_2 states are 2.51 and 3.54 eV in ethanol, respectively, while they are 2.63 and 3.24 eV in ethylene glycol. In both solvents, the first transition is from HOMO-1 to LUMO and the oscillator strength is about the order of <10^{-4}. And the second transition is from HOMO to LUMO with the oscillator strength of 0.8528 and 0.9611, respectively. From the CI coefficients, we can clearly obtain that the first transition (the S_1 state) is n\pi^* character while the second transition (the S_2 state) is \pi\pi^* character. FIG. 1 shows the molecular orbitals of the HOMO-1, HOMO, and LUMO of trans-4-MAB in ethanol which present the n, \pi and \pi^* character, respectively. Compared with the values in solvent-less environment, the vertical excitation energies of two states are obviously lower both in ethanol and ethylene glycol. Table Ⅰ Orbital transition, vertical excitation energy, configuration-interaction (CI) coefficients and oscillator strength (f) of the two lowest excited singlet states of trans-4-MAB in ethanol and ethylene glycol calculated by the TD-DFT/B3LYP with 6-311G++(d, p) basis set.  FIG. 1 Molecular orbitals of HOMO-1, HOMO, and LUMO of trans-4-DEAAB in ethanol calculated by B3LYP/6-311G++(d, p). B. Steady absorption spectrum The steady absorption spectra of trans-4-MAB in ethanol and ethylene glycol were measured and shown in FIG. 2. Both absorption spectra are very similar to each other. The first absorption band is a board band with a center wavelength of \sim430 nm and assigned to the S_1 state. In fact, the first absorption band can also be observed in the range of 400-480 nm in azobenzene [16, 17] and other derivatives, such as trans-4-aminoazobenzene in ethanol [25] and 4-nitro-4'-(ethyl (2-hydroxyethyl) amino) azobenzene (DR1) in 2-fluorotoluene [28]. For trans-azobenzene, the oscillator strength of the n\pi^* transition is measured to be the order of 10^{-2} and larger than the calculated value [16]. Compared with azobenzene, trans-4-MAB has a lower symmetry because of the substitution. And the experimental value of the oscillator strength by n\pi^* transition is estimated to be much larger than the calculated value. Although the calculated result is <10^{-4}, the absorption in 400-480 nm is still attributed to the n\pi^* state (the S_1 state). The second absorption band with the center wavelength of \sim345 nm shows much more intensity and is attributed to the \pi\pi^* transition. In FIG. 2, we can clearly observe that the energy gap between the S_1 and S_2 states of the trans-4-MAB is closer than that of azobenzene, in which the S_1 and S_2 states locate at \sim440 nm and 316 nm, respectively [12]. It is probably deduced that the substitution destroys the symmetry and leads to a red-shift of the \pi\pi^* state. The steady absorption spectrum of cis-4-MAB in ethanol is also performed by calculations using TD-DFT/B3LYP, as shown in FIG. 2 (black line). The first and second absorption bands of cis-4-MAB locate in 400-500 nm and 260-350 nm, respectively, which is consistent with the experimental results measured by Scaiano et al. [37]. In their experiments, they monitored the cis-trans (thermal) isomerization of methoxyazobenzene by UV-visible spectroscopy and obtained that the first absorption band of cis-4-MAB coupling with nanoparticles is in 400-500 nm and the second absorption band is about in 275-375 nm.  FIG. 2 Steady absorption and emission spectra of the trans-4-MAB in ethanol and ethylene glycol, and the calculated absorption spectrum of cis-4-MAB in ethanol. C. The ultrafast dynamics of {trans}-4-MAB The two-dimensional transient absorptions spectrum of trans-4-MAB in ethanol was obtained at \lambda_{\textrm{pump}}=345 nm with delay time up to 10 ps and 300 ps, respectively, shown in FIG. 3. The pump wavelength coincides with the peak of the \pi\pi^* absorption band. The detection wavelengths range from 390 nm to 720 nm. The two-dimensional spectrum shows an intensive excited state absorption (ESA) in the range of 390-440 nm and a weak absorption in the longer wavelength. It is noticed that there is a much weaker positive signal in the range of 400-480 nm until the longest experimental delay time (\Delta$$t_{\textrm{max}}$=300 ps). This permanent signal is assigned to the product of isomerization, because the first absorption band of the $cis$-4-MAB has been proven to locate exactly in this range by our calculations and experiments [37]. FIG. 4 displays the absorption spectra of the time evolution at selected delay time. After being excited to the S$_2$ state by 345 nm, the absorption in the 550-720 nm arises immediately and reaches its maximum at delay time of $\sim$75 fs (FIG. 4(a)). It indicates that the signal in this range includes the dynamic information of the initially excited S$_2$ state. Following that, the signal in 550-720 nm decreases along with the increase of signal in 390-450 nm which attains its maximum at $\sim$265 fs. It can be deduced that a direct dynamical conversion from the S$_2$ state into the other state occurs with the ultrafast timescale. FIG. 4(b) also shows that the absorption in the shorter wavelength decays within $\sim$3 ps and maintains a constant ($\sim$0.003) in the 400-480 nm until the decay time up to 300 ps.

 FIG. 3 Two-dimensional transient absorption spectrum of $trans$-4-MAB in ethanol with delay time within 10 ps and 300 ps (insert).
 FIG. 4 Transient absorption spectra of $trans$-4-MAB in ethanol at different delay times. The arrows indicate the tendency of the temporal evolution.

To properly describe the dynamics of $trans$-4-MAB in ethanol, we performed singular value decomposition (SVD) analysis on the 2D data matrix. The resulting kinetic amplitude vectors were globally fitted. Decay-associated difference spectra (DADS) with three time components of 0.11, 1.4, and 2.9 ps were obtained from the global analysis, as shown in FIG. 5. The DADS reflects the relative spectral contributions of each time component. The global analysis result shows a good match with the experimental traces over the whole spectro-temporal range. The DADS with timescale of 0.11 ps presents the relatively large positive amplitudes in 550-720 nm. It is in accordance with the region of the transient absorption of the S$_2$ state. As a result, the 0.11-ps component is attributed to the decay of the S$_2$ state. Similar time scales of the S$_2$ state are also obtained in azobenzene and other derivatives [18, 19, 25, 28]. A large number of studies about azobenzene and its derivatives show that there may be two decay channels of the excited S$_2$ state ($\pi\pi^*$) [15, 20, 21, 24-30]. The first channel is the internal conversion to the S$_1$ state of the trans-isomer and the second one is the isomerization to the cis-isomer. For trans-azobenzene, the isomerization time has been proven to be 0.5-12.5 ps depending on the solvents [15, 19, 22]. The para-substitution group causing the heavier phenyl may make the isomerization even more difficult. In consequence, it is impossible that the isomerization of the $trans$-4-MAB completes in such a fast 0.11 ps timescale. Furthermore, Tahara et al. demonstrated that the quantum yield of the internal conversion from the S$_2$ state to the S$_1$ state of trans-azobenzene was almost unity by femtosecond time-resolved fluorescence technology [15]. As mentioned above, a methoxyl-substitution makes the energy gap between $\pi\pi^*$ (S$_2$) and n$\pi^*$ (S$_1$) states much smaller than that of trans-azobenzene. Hence, the electronic relaxation of the S$_2$$\rightarrowS_1 becomes easier in trans-4-MAB. It is concluded that the appropriate decay channel of the S_2 state in trans-4-MAB is the internal conversion to the S_1 state. FIG. 4(a) shows that the signal in 550-720 nm originating from the S_2 state decreases with the increase of signal in 390-450 nm. Now, we can assume that the absorption in 390-450 nm originates from the S_1 state. The amplitude of the DADS with lifetime of 1.4 ps is positive and relatively stronger in the range of the 390-450 nm. It is clearly indicated that the 1.4-ps component is the contribution of the S_1 state which may decay by internal conversion to the trans-isomer S_0 state and isomerization to the cis-isomer S_0 state. The DADS with lifetime of the 2.9 ps shows the main absorption in the 400-450 nm which is the same region of the permanent absorption of the product. Moreover, it also corresponds to the first static absorption band of the cis-4-MAB approximately. In consequence, the 2.9-ps component is attributed to the vibration cooling in the electronic ground state of the cis-isomer. Similar results can be observed in DR1 [28]. The vibration cooling of the hot S_0 state in the region of the permanent absorption of cis-DR1 occurs in \sim6 ps. The species-associated different spectra (SADS) are also presented in FIG. 5. The species of the S_2 and S_1 states are contributed to be the most in the longer and shorter wavelength, respectively, while the species of the hot S_0 state of the cis isomer exists mainly in 400-500 nm. The results of the time evolutions of trans-4-MAB in ethanol at representative wavelengths are listed in Table Ⅱ. Several representative decay traces at selected probe wavelengths of trans-4-MAB in ethanol are shown in FIG. 6. The fitting data are consistent well with the original data.  FIG. 5 Decay-associated difference spectra (DADS) and species-associated different spectra (SADS) of trans-4-MAB in ethanol. Table Ⅱ Results of the global fit analysis of the absorption-time profiles for trans-4-MAB in ethanol.  FIG. 6 Representative time traces of the transient absorption spectra of trans-4-MAB recorded upon excitation at 345 nm in ethanol (left) and ethylene glycol (right). The components of the 0.11, 1.4 and 2.9 ps in ethanol are shown in violet, blue and green line, respectively. The components of the 0.16, 1.5, and 7.5 ps in ethylene glycol are also shown in violet, blue and green line, respectively. The excited-state dynamics of trans-4-MAB in ethylene glycol has also been obtained to further elucidate a photoisomerization mechanism. The characters of the transient absorption spectra are much similar to those in ethanol and not shown here anymore. The global analysis also shows three components with lifetimes of 0.16, 1.5 and 7.5 ps, listed in Table Ⅲ. Several representative decay traces at selected probe wavelengths of trans-4-MAB in ethylene glycol are shown in FIG. 6. The fitting results are consistent well with the experimental data. The components of 0.16 and 1.5 ps are assigned to the lifetime of the S_2 and S_1 state, respectively, which is consistent with those in ethanol. Although the viscosity of the ethylene glycol (13.5) is much different from the ethanol (1.2), the S_1-state lifetime still shows no obvious difference. It means that the n\pi^* (S_1-state) lifetime is independent of the solvent viscosity. Sawada et al. pointed out that the effect of the solvent viscosity was weaker on the inversion process than the rotation process because the change of the volume in the inversion process was smaller than that in the rotation process [25]. Similar statement was proposed on the mechanism isomerization of trans-stilbene [38]. Therefore, the solvent independence of the transient absorption spectra of the n\pi^* excited trans-4-MAB demonstrates that the photoisomerization pathway undergoes the inversion mechanism rather than the rotation mechanism. Moreover, the 7.5-ps component is attributed to the vibration cooling of the hot ground state of cis-4-MAB which is obviously different from that in ethanol. It may be caused by different state density of the hot S_0 state in different solvents. Table Ⅲ Time components obtained from global analysis for trans-4-MAB in ethanol and ethylene glycol. Ⅳ. CONCLUSION In this work, we reported the ultrafast photoisomerization and excited-state dynamics of trans-4-methoxyazobenzene (trans-4-MAB) in ethanol and ethylene glycol by femtosecond transient absorption spectroscopy combined with quantum chemistry calculations. After being excited to the S_2 state, the two-dimensional transient absorptions spectra were obtained. The excited state absorption (ESA) of the S_2 and S_1 states were observed in different range in the whole detected wavelength. The time evolutions of \pi$$\pi$$^*$excited trans-4-MAB in both solvents were extracted and fitted by the global analysis. It shows three components with lifetimes of 0.11, 1.4 and 2.9 ps in ethanol and 0.16, 1.5 and 7.5 ps in ethylene glycol, respectively. The fast component in both solvents was assigned to the internal conversion from the S$_2$ state to the S$_1$ state. The 1.4 and 1.5 ps were correlated with the decay of the S$_1$ state by internal conversion and isomerization. The cis-4-MAB is produced and witnessed by the permanent positive absorption in the 400-480 nm. The solvent independence of the transient absorption spectra of the n$\pi^*$ excited $trans$-4-MAB demonstrated that the photoisomerization pathway undergoes the inversion mechanism rather than the rotation mechanism. The slow component of 2.9 and 7.5 ps was attributed to the vibrational cooling of the hot S$_0$ state in the cis-isomer. The vibrational cooling of the hot S$_0$ state of $cis$-4-MAB in ethanol is obviously faster than that in ethylene glycol. It may be caused by different state densities of the hot S$_0$ state in different solvents. A general photoinduced mechanism for the S$_2$-excited $trans$-4-MAB is drawn and shown in FIG. 7.

 FIG. 7 The general photoinduced isomerization mechanisms of the S$_2$-excited trans-4-MAB in solutions. The numbers in brackets are lifetimes in ethylene glycol.
Ⅴ. ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China (No.21603049, No.11674355, No.11705043, No.21327804, No.11364043), the Fundamental Research Funds for the Central Universities (No.JZ2015HGBZ0532), the Industry-University-Research Fund of Hefei University of Technology Xuancheng Campus (No.XC2016JZBZ11), and the Natural Science Foundation of Xinjiang Uygur Autonomous Region (No.2016D01A058).

Reference
 [1] T. Ikeda, and O. Tsutsumi, Science 268 , 1873 (1995). DOI:10.1126/science.268.5219.1873 [2] D. Bléger, T. Liebig, R. Thiermann, M. Maskos, J. P. Rabe, and S. Hecht, Angew. Chem. Int. Ed. 50 , 12559 (2011). DOI:10.1002/anie.201106879 [3] K. G. Yager, and C. J. Barrett, J. Photochem. Photobiol. A 182 , 250 (2006). DOI:10.1016/j.jphotochem.2006.04.021 [4] H. Shoji, K. TsubakiI, M. Kondo, and N. Kawatsuki, Mol. Cryst. Liq. Cryst. 563 , 131 (2012). DOI:10.1080/15421406.2012.689146 [5] J. Liou, T. Chang, T. Lin, and C. Yu, Opt. Express 19 , 6756 (2011). DOI:10.1364/OE.19.006756 [6] A. Khan, C. Kaiser, and S. Hecht, Angew. Chem. Int. Ed. 45 , 1878 (2006). DOI:10.1002/(ISSN)1521-3773 [7] D. G. Flint, J. R. Kumita, O. S. Smart, and G. A. Woolley, Chem. Biol. 9 , 391 (2002). DOI:10.1016/S1074-5521(02)00109-6 [8] G. A. Woolley, Acc. Chem. Res. 38 , 486 (2005). DOI:10.1021/ar040091v [9] X. G. Liang, K. Fujioka, Y. Tsuda, R. Wakuda, and H. Asanuma, Nucleic Acids Symp. Ser. 52 , 19 (2008). DOI:10.1093/nass/nrn010 [10] H. Nishioka, X. Liang, and H. Asanuma, Chem. A Eur. J. 16 , 2054 (2010). DOI:10.1002/chem.v16:7 [11] N. Tamai, and H. Miyasaka, Chem. Rev. 100 , 1875 (2000). DOI:10.1021/cr9800816 [12] M. Quick, A. L. Dobryakov, M. Gerecke, C. Richter, F. Berndt, I. N. Ioffe, A. A. Granovsky, R. Mahrwald, N. P. Ernsting, and S. A. Kovalenko, J. Phys. Chem. B 118 , 8756 (2014). DOI:10.1021/jp504999f [13] H. Rau, and E. Lueddecke, J. Am. Chem. Soc. 104 , 1616 (1982). DOI:10.1021/ja00370a028 [14] H. Rau, and Y. Q. Shen, J. Photochem. Photobiol. A: Chem. 42 , 321 (1988). DOI:10.1016/1010-6030(88)80075-3 [15] T. Fujino, S. Yu, Arzhantsev, and T. Tahara, J. Phys. Chem. A 105 , 8123 (2001). DOI:10.1021/jp0110713 [16] C. W. Chang, Y. C. Lu, T. T. Wang, and E. W. G. Diau, J. Am. Chem. Soc. 126 , 10109 (2004). DOI:10.1021/ja049215p [17] H. M. D. Bandara, and S. C. Burdette, Chem. Soc. Rev. 41 , 1809 (2012). DOI:10.1039/C1CS15179G [18] J. Azuma, N. Tamai, A. Shishido, and T. Ikeda, Chem. Phys. Lett. 288 , 77 (1998). DOI:10.1016/S0009-2614(98)00251-6 [19] T. Fujino, S. Y. Arzhansev, and T. Tahara, Bull. Chem. Soc. Jpn. 75 , 1031 (2002). DOI:10.1246/bcsj.75.1031 [20] T. Schultz, J. Quenneville, B. Levine, A. Toniolo, and T. J. Martinez, J. Am. Chem. Soc. 125 , 8098 (2003). DOI:10.1021/ja021363x [21] I. Conti, M. Garavelli, and G. Orlandi, J. Am. Chem. Soc. 130 , 5216 (2008). DOI:10.1021/ja710275e [22] H. Satzger, S. Spörlein, C. Root, J. Wachtveitl, W. Zinth, and P. Gilch, Chem. Phys. Lett. 372 , 216 (2003). DOI:10.1016/S0009-2614(03)00364-6 [23] J. Azuma, N. Tamai, A. Shishido, and T. Ikeda, Chem. Phys. Lett. 288 , 77 (1998). DOI:10.1016/S0009-2614(98)00251-6 [24] S. G. Mayer, C. L. Thomsen, M. P. Philpott, and P. J. Reid, Chem. Phys. Lett. 314 , 246 (1999). DOI:10.1016/S0009-2614(99)01151-3 [25] Y. Hirose, H. Yui, and T. Sawada, J. Phys. Chem. A 106 , 3067 (2002). DOI:10.1021/jp0138375 [26] Y. Wang, S. Zhang, S. Sun, K. Liu, and B. Zhang, Chin. J. Chem. Phys. 26 , 651 (2013). DOI:10.1063/1674-0068/26/06/651-655 [27] Y. Wang, S. Zhang, S. Sun, K. Liu, and B. Zhang, J. Photochem. Photobiol. A 309 , 1 (2015). DOI:10.1016/j.jphotochem.2015.04.015 [28] J. Bahrenburg, K. Röttger, R. Siewertsen, F. Renth, and F. Temps, Photochem. Photobiol. Sci. 11 , 1210 (2012). DOI:10.1039/c2pp05400k [29] M. Hagiri, N. Ichinose, C. Zhao, H. Horiuchi, H. Hiratsuka, and T. Nakayama, Chem. Phys. Lett. 391 , 297 (2004). DOI:10.1016/j.cplett.2004.04.093 [30] Y. J. Lee, S. I. Yang, D. S. Kang, and S. Joo, Chem. Phys. 361 , 176 (2009). DOI:10.1016/j.chemphys.2009.05.021 [31] J. Wachtveitl, T. Näggele, B. Puell, W. Ninth, M. Krüger, S. Rudolph-Böhner, D. Oesterhelt, and L. Moroder, J. Photochem. Photobio. A 105 , 283 (1997). DOI:10.1016/S1010-6030(96)04572-8 [32] C. Wang, H. Jau, and T. Lin, Opt. Lett. 37 , 2730 (2012). DOI:10.1364/OL.37.002730 [33] Y. Norikane, S. Tanaka, and E. Uchida, CrystEngComm 18 , 7225 (2016). DOI:10.1039/C6CE00738D [34] M. A. Haidekker, T. P. Brady, D. Lichlyter, and E. A. Theodorakis, Bioorg. Chem. 33 , 415 (2005). DOI:10.1016/j.bioorg.2005.07.005 [35] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Jr. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, N. J. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochter-ski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian 09, Revision A02, Wallingford, CT: Gaussian Inc., (2009). [36] J. Tomasi, and M. Persico, Chem. Rev. 94 , 2027 (1994). DOI:10.1021/cr00031a013 [37] G. L. Hallett-Tapley, C. D' Alfonso, N. L. Pacioni, C. D. McTiernan, M. González-Béjar, O. Lanzalunga, E. I. Alarcona, and J. C. Scaiano, Chem. Commun. 49 , 10073 (2013). DOI:10.1039/c3cc41669k [38] D. H. Waldeck, Chem. Rev. 91 , 415 (1991). DOI:10.1021/cr00003a007

a. 合肥工业大学电子科学与应用物理学院，合肥 230009;
b. 中国科学院武汉物理与数学研究所，波谱与原子分子物理国家重点实验室，武汉 430071;
c. 新疆师范大学物理与电子工程学院，新疆矿物发光材料及其微结构实验室，乌鲁木齐 830054