DOI: 10.1002/cphc.201500290
Articles
Photophysical Properties of Intramolecular Charge Transfer in a Tribranched Donor–p–Acceptor Chromophore Jiangpu Hu,[a] Yang Li,[a] Huaning Zhu,[a] Shuhai Qiu,[b] Guiying He,[a] Xiaozhang Zhu,*[b] and Andong Xia*[a] The photophysical properties of intramolecular charge transfer (ICT) in a novel tribranched donor–p–acceptor chromophore, triphenoxazine-2,4,6-triphenyl-1,3,5-triazine (tri-PXZ-TRZ), with thermally activated delayed fluorescence character was investigated in different aprotic solvents by steady-state spectroscopy and femtosecond and nanosecond transient absorption spectroscopy measurements. Increasing the solvent polarity led to a significant increase in the Stokes shift. The large Stokes shift in highly polar solvents was attributed to ICT properties upon excitation; this resulted in a strong interaction between the tri-PXZ-TRZ molecule and the surrounding solvent, which led to a strong solvation process. Quantum-chemical calculations and changes in the dipole moment showed that
this compound has a large degree of ICT. Furthermore, an apolar environment helped to preserve the symmetry of triPXZ-TRZ and to enhance its emission efficiency. The femtosecond and nanosecond transient absorption spectroscopy results indicated that the excited-state dynamics of this push–pull molecule were strongly influenced by solvent polarity through the formation of a solvent-stabilized ICT state. The excitedstate relaxation mechanism of tri-PXZ-TRZ was proposed by performing target model analysis on the femtosecond transient absorption spectra. In addition, the delayed fluorescence of tri-PXZ-TRZ was significantly modulated by a potential competition between solvation and intersystem crossing processes.
1. Introduction With high luminescence efficiency, thermally activated delayed fluorescence (TADF) materials have in recent years received considerable attention for various applications in terms of producing organic light-emitting diodes and fluorescence probes for bioimaging.[1–7] In contrast to traditional fluorescent and phosphorescent materials, which can only harvest the singlet (25 %) or the triplet excitons (75 %), TADF materials can theoretically harvest almost 100 % of the excitons from both the singlet and triplet excited states.[8–11] These materials can provide an additional singlet exciton generated by efficient upconversion from the lowest triplet excited state (T1) to the lowest singlet excited state (S1) through reverse intersystem crossing. Such thermally activated delayed fluorescence was first observed by Boudin in 1930[12] and was later elucidated by Parker and Hatchard.[13] Generally, to obtain highly efficient upconversion from T1 to S1, a sufficiently small energy gap between S1 and T1 (DEST) is required. Theoretically, effective sepa[a] J. Hu, Y. Li, H. Zhu, G. He, Prof. A. Xia Beijing National Laboratory for Molecular Sciences, Institute of Chemistry, Chinese Academy of Sciences, Beijing, 100190 (China) E-mail: [email protected] [b] S. Qiu, Prof. X. Zhu Beijing National Laboratory for Molecular Sciences, Key Laboratory of Organic Solids, Institute of Chemistry, Chinese Academy of Sciences, Beijing, 100190 (China) E-mail: [email protected] Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/cphc.201500290.
ChemPhysChem 2015, 16, 2357 – 2365
ration between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) will lead to a very small DEST. On the basis of this concept, a number of efficient organic TADF materials have been designed and synthesized.[2–4, 14–21] One strategy is to introduce a donor–acceptor (D-A) group into the TADF system, as it can induce intramolecular charge transfer (ICT) and enhance separation between the HOMO and LUMO upon excitation.[18, 22–26] It is well known that the local environment, especially solvents, has a profound effect on the emission behavior of ICTtype compounds.[22–29] Once these compounds are excited to highly excited states, the excess amount of energy will be transferred to the surrounding medium because of solvent reorganization (solvation). Therefore, solvent effects on the excited-state dynamics of ICT-type TADF materials could be important to understand the ICT process and its complex spectral properties. However, in solvents, the excited-state dynamics of these compounds remain largely unclear because of the fast and complicated solvation process, which is usually coupled with vibrational cooling and ICT processes after excitation, and they take place on the timescale of picoseconds and even femtoseconds. Then, to understand comprehensively the underlying emission mechanisms of TADF materials in various environments, it is necessary to perform photophysical studies on ICTtype TADF materials in media with different polarities. In this paper, a donor–p–acceptor chromophore, triphenoxazine-2,4,6-triphenyl-1,3,5-triazine (tri-PXZ-TRZ), that has been proven to have strong delayed fluorescence from ICT[18] was chosen as a model system to investigate the ICT process in sol-
2357
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles vents with varying polarities by using steady-state spectroscopy and femtosecond, as well as nanosecond, time-resolved transient absorption spectroscopy measurements. Figure 1 shows the molecular structure of tri-PXZ-TRZ, which has a phenyltriazine (TRZ) electron-acceptor group surrounded by three phenoxazine (PXZ) electron-donor groups. It was found that tri-PXZ-TRZ has representative TADF character with typical ICT
Figure 2. Normalized UV/Vis absorption and fluorescence spectra of tri-PXZTRZ in toluene and THF.
Figure 1. Molecular structure of tri-PXZ-TRZ.
properties upon excitation. Such ICT will lead to a strong interaction between the solute molecules and the surrounding solvent, which is correlated to a large Stokes shift of the steadystate spectrum. The femtosecond and nanosecond time-resolved transient absorption spectroscopy measurements provide solvent-dependent information about the relaxation dynamics, in which the formation of a solvent-stabilized ICT state (denoted as the ICT’ state) upon excitation is included.[22, 25] These measurements suggest that the solvent plays a very important role in the excited-state dynamics of the ICT state, which is helpful for us to understand changes in the upconversion efficiencies of TADF materials in different media.
2. Results and Discussion 2.1. Steady-State Absorption and Fluorescence Spectra Figure 2 shows the normalized optical absorption and fluorescence spectra of tri-PXZ-TRZ in toluene and tetrahydrofuran (THF). The steady-state spectral parameters of this molecule in
several typical solvents with different polarities are listed in Table 1. It can be seen that tri-PXZ-TRZ exhibits two absorption bands at room temperature. The absorption bands located at the high-energy side with a sharp intense absorption band around l = 285 nm in toluene and around l = 270 nm in THF are mainly attributed to the p–p* transition derived from the TRZ acceptor moiety.[21] The low-energy bands around l = 436 nm in toluene and l = 422 nm in THF originate from ICT transitions.[18, 22, 25, 30] It should be realized that there is a general slight blueshift in the absorption bands with an increase in solvent polarity, which indicates a small amount of ground-state ICT; this is attributed to strong interaction between the solvent and solute molecules.[31, 32] That is, the conjugated push–pull tri-PXZ-TRZ molecule adopts its geometry with lower conjugation in more polar solvent (e.g. THF), which leads to a slight blueshift in the absorption bands relative to that in less polar solvents (e.g. toluene) with larger conjugation (as shown in Figure 2). Unlike the small blueshift in the absorption spectrum, a very large solvent-induced redshift in the fluorescence spectrum for tri-PXZ-TRZ was observed as the solvent polarity was increased (see Figure S1, Supporting Information). Such broad, unresolved vibronic structure fluorescence bands with relatively large bathochromic shifts in different solvents are typical properties of fluorescence from the relaxed ICT (ICT’) states.[22, 25, 33] This remarkable solvatochromism indicates strong solvent effects in the ICT state of tri-PXZ-TRZ.
Table 1. Solvent effects on the absorption and emission properties of tri-PXZ-TRZ. Solvent
Parameter[a] e n
Df
toluene MS1[c] chloroform MS2[c] THF dichloromethane
2.4 3.4 4.8 6.0 7.6 8.9
0.0159 0.0894 0.155 0.177 0.210 0.217
1.494 1.477 1.443 1.432 1.405 1.421
la[b] [nm]
lf[b] [nm]
Stokes shift [cm¢1]
Quantum yield
436 432 429 426 422 420
573 604 625 638 648 651
5484 6592 7310 7800 8264 8449
0.21 0.044 0.021 0.0026 0.0011 0.0010
[a] Parameters are taken from ref. [51]; Df: Solvent parameter. [b] la : wavelength of absorption maxima of ICT absorption; lf : wavelength of fluorescence maxima. [c] MS: mixed solvents of toluene and THF; MS1 and MS2 represent toluene/THF with volume ratios of about 8:2 and 3:7, respectively. The corresponding emix and nmix were then calculated according to emix = faea + fbeb and nmix2 = fana2 + fbnb2, respectively; fa and fb are the volume percentages of each solvent.[41]
ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
2358
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles 2.2. Quantum-Chemical Calculations To obtain deeper insight into the nature of the ICT state and further to interpret the observed steady-state spectral behavior, quantum-chemical calculations for tri-PXZ-TRZ were performed by density functional theory (DFT) methods, as implemented in the Gaussian 09 package.[34] The ground-state geometry of tri-PXZ-TRZ was fully optimized at the WB97X-D/ccpVDZ level. Due to steric repulsion between the hydrogen atoms at the donor–acceptor linkage, a twist-optimized geometry of tri-PXZ-TRZ with a large dihedral angle (818) between the donor and acceptor planes was obtained (Figure S2 a). The twist structure between the PXZ and TRZ units will suppress delocalization of the electron density and will result in effective charge separation between the HOMO and LOMO. The molecular orbitals making major contributions to the six low-lying singlet excited states for tri-PXZ-TRZ are shown in Figure S3. These clearly show the charge-transfer nature of the optical HOMO–LUMO transition, which is responsible for the ICT absorption bands in Figure 2. To demonstrate the charge distribution and the local exited character, charge different density (CDD) of the tri-PXZ-TRZ molecule was also calculated by timedependent (TD) DFT at the WB97X-D/cc-pVDZ level, which allowed visualization of the difference in the electron densities upon excitation between the ground state (S0) and the excited state.[35–37] The CDD image of tri-PXT-TRZ is depicted in Figure S2 b. It can be observed that the charges are almost entirely distributed on the TRZ unit and the holes are mainly distributed on the three PXZ units; this is direct evidence for charge transfer from the HOMO to LUMO with a shift in the electron density from the outside PXZ moieties to the core TRZ moiety. 2.3. Solvation Effects: Solvent Parameters and Spectral Properties As mentioned above, the absorption spectrum of tri-PXZ-TRZ is slightly influenced by solvent polarity, whereas the extent of the redshift of the fluorescence spectrum is strongly dependent on solvent polarity. Consequently, the excited states of tri-PXZ-TRZ will relax to a strongly polar-dependent ICT state in a highly polar medium upon excitation, and this will lead to a large Stokes shift. To elaborate on the solvatochromism and the degree of ICT, as well as the polar nature of the excited states of tri-PXZ-TRZ, we employed the Lippert–Mataga equation [Eqs. (1) and (2)]:[38–40] Dn ¼ nabs ¢v em ¼
2 Dm2 Df þ constant ðhca3 Þ
ðe¢1Þ ðn2 ¢1Þ Df ¼ ¢ ð2 e þ 1Þ ð2 n2 þ 1Þ
ð1Þ ð2Þ
in which nabs and nem are the absorption and emission bands in wavenumbers, respectively; h is Planck’s constant; c is the speed of light; a is the Onsager cavity radius defined as the solvent shell around the solute molecule; and Dm = me¢mg is the difference between the excited-state (me) and ground-state ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
(mg) dipole moments as presented by Equation (1). The Df term is the orientation polarizability of the solvent deduced from the dielectric constant (e) and the refractive index (n) of the solvent as represented by Equation (2). On the basis of the solvent parameters and Stokes shifts summarized in Table 1, a plot of Stokes shifts versus the solvent polarity parameters for tri-PXZ-TRZ is shown in Figure 3.
Figure 3. Stokes shift versus solvent polarity (Df) for tri-PXZ-TRZ. The solid line is the least-squares fit result.
The Stokes shifts depend approximately linearly on the solvent polarities. To quantitatively determine the polarity of the excited ICT state, we examined the dipole moment change between the excited-state and ground-state dipole moments (Dm = me¢mg) from the solvatochromic shift in polar solvents on the basis of Equations (1) and (2). Estimated from quantumchemical calculations by using the DFT method at the WB97XD/cc-pVDZ level, the Onsager radius was found to be 7.27 æ for tri-PXZ-TRZ. According to the slope obtained from the fitting results shown in Figure 3, the effective dipole moment change (Dm) was calculated to be about 23.3 D. Such a large dipole moment change for tri-PXZ-TRZ indicates remarkable intramolecular charge-transfer character,[2, 18, 42–44] which is consistent with the DFT calculation results. In addition, as shown in Figure 2, there is only a slight blueshift in the absorption spectrum, but a very large redshift in the fluorescence spectrum with an increase in solvent polarity, which indicates that the dipole moment in the excited state is much larger than that in the ground state. Therefore, the difference (Dm) mainly reflects the change in the dipole moment in the excited state. The solvatochromic shift in polar solvents is due to the formation of a polar ICT excited state. 2.4. Anisotropy Measurements To investigate further the environmental polarity effect on the nature of ICT in tri-PXZ-TRZ, steady-state fluorescence excitation anisotropy experiments were performed in matrices with different polarities. The anisotropy values (r) were obtained by measuring the fluorescence excitation spectrum in the ICT absorption range from l = 380 to 500 nm on the basis of the steady-state absorption spectrum (Figure 2). To avoid rotation of chromophores during excited-state relaxation, the molecules were embedded in proper polymer matrices. By choosing ap-
2359
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles propriate matrices with specific polarities, we can provide a more comprehensive explanation on the polarity-dependent excited character of tri-PXZ-TRZ. For our study, Zeonex E48R (Zeonex) and poly-THF were chosen as matrices to perform the anisotropy experiments, as the polarities of Zeonex and poly-THF are similar to the polarities of methylcyclohexane and THF, respectively.[45] Figure 4 plots the fluorescence excitation
Figure 4. Fluorescence excitation anisotropy (c) and normalized fluorescence excitation spectra (a) of tri-PXZ-TRZ in Zeonex and poly-THF. The fluorescence excitation spectra were measured by monitoring the fluorescence wavelength at 530 nm in Zeonex and at 550 nm in poly-THF.
spectra and the corresponding anisotropy spectra of tri-PXZTRZ in these two matrices. In apolar Zeonex, the anisotropy constant values of tri-PXZ-TRZ are around 0.1–0.2 throughout the ICT excitation band, which indicates that the C3 symmetry of the ICT states in tri-PXZ-TRZ is preserved in the low-polar matrix, in which the ICT redistributes the excitation energy among the three spatially degenerate transition dipole moments.[46] However, in poly-THF the anisotropy values are between 0.3 and 0.4 in the region of the ICT band, which suggests that the excited state is localized in one branch of the tri-PXZ-TRZ molecule in a polar medium. This may be due to the fact that a matrix with high polarity will lead to torsional disorder and, hence, reduce the global symmetry of the three branches in tri-PXZ-TRZ. The slight blueshift in the absorption spectrum displayed in Figure 2 recorded in THF relative to the spectrum recorded in toluene also suggests the formation of a polar excited state that is localized in one branch of the molecule. 2.5. Solvent Polarity Effect on Fluorescence Quantum Yields and Lifetimes Table 1 lists the fluorescence quantum yields (Ff) of tri-PXZTRZ in various solvents. As plotted in Figure S4, the Ff values decrease dramatically with an increase in solvent polarity. This is consistent with the energy gap law, for which an increased Stokes shift in a more polar solvent is often accompanied by a decrease in fluorescence quantum yield due to an increase in the radiationless rate.[47, 48] This indicates that an apolar solvent could facilitate fluorescence transition and that the suppression of the radiationless process would be greater than that in a polar solvent. As a result, an apolar environment will benefit ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
the fluorescence process and thereby increase the emission efficiency of tri-PXZ-TRZ. To study the effect of solvent on fluorescence lifetime, we measured the fluorescence decay of tri-PXZ-TRZ at its fluorescence maximum in the respective solvents. The fluorescence decay curves of tri-PXZ-TRZ in toluene and THF are depicted in Figure S5. In toluene, there are two clear components of the decay curve. The first component was assigned to prompt fluorescence, which has a lifetime of about 17 ns. The second component was assigned to delayed fluorescence, which exhibited a lifetime of around 1 ms. These values agree with the former report.[18] Notably, the lifetime of delayed fluorescence is much longer than that of prompt fluorescence, because delayed population of the excited singlet state originates from the long-lived T1 state through reverse intersystem crossing.[2, 18] However, in contrast to the case in toluene, only a dramatic decrease in the decay curve was observed, and the slow decay component disappeared in THF. Consequently, delayed population of the excited singlet state was largely decreased because of the strong interaction between the excited solute molecules and the solvent molecules, which efficiently suppresses the intersystem crossing (ISC) process.[49, 50] The lifetime in THF is too short to be measured by the same instrument as that used to measure the lifetime in toluene, so time-correlated single photon counting (TCSPC) was performed to measure the fluorescence lifetime in THF. The fluorescence lifetime was measured to be only 1.4 ns (see Figure S6), and no delayed fluorescence was found in THF. Furthermore, the dramatic decrease in the decay curve in higher polarity solvents was mainly attributed to a solvation-induced strong radiationless transition process, as this makes up a larger portion of the whole deactivation process of the excited state than the lower polarity solvent. We calculated the radiationless transition rate in both solvents on the basis of the quantum yields and lifetimes of the prompted and delayed fluorescence [Eq. (S1)]. In toluene, the radiationless rate was 2.25 Õ 108 s¢1. In THF, the delayed fluorescence rate was neglected, in view of the absence of a delayed component (Figure S5), and thus, the radiationless transition rate was 6.48 Õ 1011 s¢1, which is markedly faster than that in toluene. To further investigate the solvation process, femtosecond time-resolved transient absorption spectroscopy measurements were employed to provide additional solventdependent dynamic information about the excited state of triPXZ-TRZ. 2.6. Femtosecond Time-Resolved Transient Absorption Spectroscopy Femtosecond time-resolved transient absorption spectroscopy measurements of tri-PXZ-TRZ in different solvents were performed to gain more insight into the polar-dependent ICT state after excitation at l = 400 nm. Figure 5 illustrates the broad transient absorption spectra of tri-PXZ-TRZ obtained at different time delays from 0 fs to 1 ns in toluene and in THF. The transient spectra in both solvents are composed of a very broad excited-state absorption (ESA, positive transient absorption signal) that over the whole spectral region overlaps with
2360
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles that there is a dramatic decay in the transient spectra in the following time domain of 3 to 1000 ps (Figure 5 d). This is quite different from the case in toluene. Such decay mainly originates from fast radiationless relaxation in the highly polar solvent, which leads to a low fluorescence quantum yield in THF. As mentioned above, it is clear that the excited tri-PXZ-TRZ molecules will relax through solvation and ISC processes simultaneously at the initial time delay. Hence, there is a potential competition between solvationcoupled ICT decay and ISC during the relaxation processes. Figure 5. Evolution of the transient absorption spectra of tri-PXZ-TRZ at different delay times: a, b) in toluene and c, d) in THF. Furthermore, the energy surface of the ICT state will decrease much faster in polar solvents ground-state bleach (GSB, negative transient absorption signal) than in apolar solvents due to the strong solvation process. Acand stimulated emission (SE, negative transient absorption cordingly, kinetic models of tri-PXZ-TRZ in different solvents signal). were reasonably proposed, as shown in Figure 6. For a detailed Taking a closer look at the initial transient absorption spectra understanding of the femtosecond transient absorption specin toluene, the dip around l = 450 nm can be readily assigned tra, target analysis based on the proposed kinetic model was to the ground-state bleach according to the ICT absorption in then employed to extract the time independent correlations the steady-state absorption spectrum. Broad ESA bands from these transient absorption spectroscopy data. The species around l = 525 and 750 nm can be found over the whole meaassociated difference spectra (SADS) and concentration kinetics sured time window. Fast dynamics occur in the time range of for tri-PXZ-TRZ were extracted and are plotted in Figure 7. To approximately 0–10 ps (Figure 5 a), and this is characterized by show the quality of fitting, kinetics at selected wavelengths of an increase in the ESA bands, which may be attributed to soll = 450, 525, 575, and 750 nm in toluene and at l = 450, 525, vation and/or ISC. Furthermore, the SE band exhibits a redshift 600, and 750 nm in THF are shown in Figure S7 together with from an initial position of l = 550 nm to a final position that a global fit of all the collected time traces. Three components nearly corresponds to the steady-state fluorescence spectrum were required for an adequate fit of the transient spectroscopy of l = 575 nm with an increase in the delay time; this is indicadata by using the applied target dynamic schemes (Figure 6). tive of the existence of a solvation process that could gradually The estimated rate constants are summarized in Table 2. lower the excited-state potential energy surface with time In toluene, the first component was assigned to the ICT delay. However, no clear decays or shifts in the spectra were state upon excitation at l = 400 nm. The ICT state decays into two independent excited-state populations, the solvent-stabifound (Figure 5 b) in the following 10–1000 ps, which suggests the formation of a long-lived state. lized ICT state (ICT’ state) and the T1 state, with rate constants In contrast to the spectra in toluene, interesting transient of (2.98 ps)¢1 and (11.5 ps)¢1 by solvation and ISC processes, siabsorption features were observed in THF. From Figure 5 c, the multaneously. As shown in Figure 7 a, it is not surprising to see initial process occurs in the timeframe of about 0–3 ps, which is much faster than that in toluene. This was attributed to a stronger interaction between the solute and solvent molecules, which facilitates the solvation process.[51] Furthermore, the SE band exhibits a larger redshift from l = 550 to about 600 nm with an increase in the delay times, which indicates that the potential energy surface of the ICT state in the polar solvent is much lower than that in the apolar solvent, and this caused a bathochromic shift in the steady-state fluorescence spectra as the polarity increased. Furthermore, a local minimum is observed at about l = 720 nm at 3 ps (Figure 5 c, d), and it is also attributed to decay of stimulated emission from Figure 6. Proposed relaxation models of tri-PXZ-TRZ used in target analysis: the solvation-coupled ICT (ICT’) state. It should also be realized a) in toluene and b) in THF. ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
2361
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles ative to the rate constant in toluene, such a large rate constant indicates that the radiationless transition makes up a much larger portion of this relaxation process, which is consistent with the fluorescence quantum yields and lifetime results. 2.7. Nanosecond Time-Resolved Transient Absorption Spectroscopy Nanosecond time-resolved transient absorption spectroscopy measurements were performed to further interpret the solvation-dependent emission characteristics of tri-PXZ-TRZ. Figure 8 shows the nanosecond transient Figure 7. SADS from target analysis of tri-PXZ-TRZ: a) in toluene and b) in THF. Concentrations of the SADS comabsorption spectra obtained in ponents as a function of time: c) in toluene and d) in THF. toluene and THF. There are two negative bands at around l = 445 and 575 nm and three positive bands at around l = 400, 510, and 750 nm in toluene. As depicted in Figure S8, the tranTable 2. Rate constants estimated from target analysis of tri-PXZ-TRZ in sient absorption spectra of tri-PXZ-TRZ are similar to each different solvents. other at delay times of 500 ns (from the nanosecond measurek2 [ps¢1] k3 [ps¢1] Solvent k1 [ps¢1] ments) and 1 ns (from the femtosecond measurements). In detail, the negative bands at about l = 445 and 575 nm were toluene 3.0 0.2 11.5 1.0 > 1000 THF 1.0 0.1 30.3 3.0 319 30 readily assigned to GSB and SE, respectively, according to the steady-state and femtosecond transient absorption spectra mentioned above. The three positive bands were attributed to the ESA band of the long-lived T1 state considering the microthat the SADS corresponding to the ICT’ state and the T1 state second scale of the decay process. The lifetime of the T1 state are similar to each other because of the small energy gap is about 0.95 ms, obtained by fitting the time trace of the ESA (DEST) between them. The third component (> 1 ns) was attributed to decay from the ICT’ state to the ground state. Three components were also used for an adequate fit of the transient spectroscopy data in THF. On the basis of the results in toluene, the three components were assigned to the ICT, ICT’, and T1 states. However, the solvation rate constant was (1.1 ps)¢1 in THF, which is much larger than that in toluene, whereas the ISC rate constant was (30.3 ps)¢1 in THF, which is much smaller than that in toluene. As a consequence, in highly polar THF the formation of the ICT’ state is very fast, because strong solvation occurs, which leads to a much lower population of the T1 state relative to that in an apolar solvent. In addition, the strong solvation process may lead to a decrease in the energy of the ICT’ state to a level that is lower than that of the T1 state, and this can also cause a lower population of the T1 state. As displayed in Figure 7 b, the shape of the SADS corresponding to the T1 state becomes abnormal, and this is due to the low population of the T1 state (Figure 7 d). Thus, the emission behavior of tri-PXZ-TRZ can be significantly modulated by solvent polarity through competition between solvation and ISC. The third rate constant of (319 ps)¢1 was assigned to Figure 8. Nanosecond time-resolved transient difference absorption spectra the relaxation rate from the ICT’ state to the ground state. Relof tri-PXZ-TRZ at different delay times: a) in toluene and b) in THF. ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
2362
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles band at l = 510 nm (Figure S9), which is close to the lifetime of delayed fluorescence. It is known that triplet excitons can be inactivated by triplet oxygen molecules. In consequence, to determine the importance of the triplet excited state in triPXZ-TRZ, nanosecond transient absorption spectroscopy experiments were performed after oxygenation of a toluene solution by bubbling oxygen through it. All signals were quenched after oxygen was bubbled into the toluene solution (Figure S10), which indicates that the triplet excited state plays a very significant role in this relaxation process. However, all of the signals also disappeared if the nanosecond transient absorption spectroscopy experiment was performed in THF (Figure 8 b). According to the results of the femtosecond transient absorption spectroscopy experiments, quenching of the signals originates from the strong solvation process, which leads to suppression of the ISC process. Overall, in highly polar solvents the strong solvation process will effectively quench the delayed fluorescence of tri-PXZ-TRZ.
3. Conclusions
Steady-State Absorption and Fluorescence Spectroscopy Measurements Absorption and fluorescence spectra were measured by using a UV/Vis spectrophotometer (U-3010, Hitachi Japan) and a fluorescence spectrophotometer (F-4600, Hitachi Japan) at ambient temperature, respectively. Fluorescence quantum yield (Ff) measurements were performed by the comparative method by using the Ff of fluorescein in 0.1 m NaOH (Ff = 0.9) as the reference.[52] In these measurements, all samples were bubbled with nitrogen for 15 min. Fluorescence excitation anisotropy spectra were measured by a fluorescence spectrophotometer (F-4600, Hitachi Japan) with two polarizers in excitation and detection light routes. The anisotropy value (r) was calculated with Equation (3): r¼
In summary, we presented the results of the spectral properties and dynamics of the ICT states of a tribranched donor–p–acceptor chromophore (i.e. tri-PXZ-TRZ) by means of steady-state spectroscopy and femtosecond, as well as nanosecond, transient absorption spectroscopy measurements. The steady-state spectral results showed that increasing the solvent polarity led to a slight blueshift in the absorption bands and to a remarkably large redshift in the fluorescence bands for tri-PXZ-TRZ. Chemical calculation results showed that tri-PXZ-TRZ has typical ICT character upon excitation. The large change in the dipole moment ( 23.3 D) of tri-PXZ-TRZ between the ground state and the excited state also proved the formation of a polar-dependent ICT state. This ICT state resulted in a strong interaction between the solute molecules and the surrounding solvent, which induced a solvation process, and this was correlated to a large Stokes shift in the steady-state spectrum. In addition, the strong solvation effects and dynamics of tri-PXZTRZ were also investigated by femtosecond and nanosecond transient absorption spectroscopy measurements. It was found that the solvent polarity had a significant impact on the dynamics of the ICT state of tri-PXZ-TRZ. The solvation process was much stronger in a polar solvent than in an apolar solvent upon excitation, but the ISC process was weaker in a polar solvent than in an apolar solvent . Consequently, the emission behavior of tri-PXZ-TRZ could be efficiently modulated by choosing proper polar solvents through this potential competition between solvation and ISC.
Experimental Section Materials The synthesis and purification of tri-PXZ-TRZ were done completely as previously reported.[18] The chemical structure and purity were identified by NMR spectroscopy, MALDI-TOF-MS, and elemental analysis. All the aprotic solvents including toluene, chloroform, THF, and dichloromethane used in this work were of analytical reChemPhysChem 2015, 16, 2357 – 2365
agent grade or higher and were purchased from the Beijing Chemical Plant without further purification. Fluorescein was purchased from Sigma–Aldrich. Zeonex E48R (Zeon Japan) and poly-THF (Aladdin) were used as polymer matrices in anisotropy detection.
www.chemphyschem.org
Ik ¢GI? Ik þ 2GI?
ð3Þ
in which I k and I ? are the polarized fluorescence intensities parallel and perpendicular to excitation polarization, respectively, and G (G = I ? /I k ) is the geometrical factor of fluorescence spectrophotometer if the excitation is vertically polarized. To avoid fast rotation of the molecule during the fluorescence lifetime measurements, triPXZ-TRZ was immobilized in Zonex E48R and poly-THF matrices.
Fluorescence Lifetime Measurements Fluorescence lifetime of tri-PXZ-TRZ in toluene was measured with an LP920 laser flash photolysis spectrometer (Edinburgh Instruments UK) in a fluorescence mode by using the excitation wavelength of 355 nm. Samples were prepared by bubbling with nitrogen for 15 min for each measurement and adjusted to an absorbance of approximately 0.3 OD in 10 mm path length quartz cuvettes at the laser wavelength used. The fluorescence lifetimes in different solvents were measured at their maximum emission wavelength shown in steady-stated fluorescence spectra. The fluorescence lifetime in THF was also measured by a time-correlated single-photon counting (TCSPC) spectrometer (F900, Edinburgh Instrument). The sample was excited at l = 380 nm. The instrument response function (IRF) of the detection system was about 600 ps.
Femtosecond Transient Absorption Spectroscopy Measurements Femtosecond transient absorption spectroscopy measurements with approximately 90 fs time resolution were performed with a home-built femtosecond broadband pump-probe setup. Details of the instrument are described elsewhere.[53] In brief, 1 mJ, 40 fs pulse at 800 nm with a repetition rate of 500 Hz was obtained from a regeneratively amplified Ti:sapphire laser (Coherent Legend Elite USA) with a bandwidth (full width at half maximum) of about 30 nm. The output of the laser beam was split to generate pump and probe beam pulses with a beam splitter (90 and 10 %). The 400 nm pump pulse was produced by doubling a portion of the
2363
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles 800 nm pulse with a 0.5 mm thick BBO (type I) crystal. The pump power was about 100 nJ pulse¢1 (spot size of 120 mm in each case) during transient absorption spectroscopy measurements. The probe beam was sent to a computer-controlled optical delay system and then focused onto a 2 mm thick water cell to generate a white light continuum that was split into two beams by using a broadband 50/50 beamsplitter as the signal and reference beams. The focused signal and pump beams were overlapped into a flow sample cuvette with 1 mm beam path length, and the reference beam was passed through the unexcited part of the sample. Signal and reference beams were detected with a fiber-coupled dual-channel spectrometer (Avantes AvaSpec-2048-2-USB2). For each pump pulse, spectral intensities of the probe and the referoff on ence in the absence of excitation in the sample (Iprobe and Iref , reon on spectively) and in the presence of the pump (Iprobe and Iref , respectively) were measured. Then, the change in the optical density (DOD) of the transient absorption spectra (for a given time delay) was calculated from Equation (4): DODðt; lÞ ¼ log
off on Iprobe ðt; lÞ Iref ðt; lÞ off on Iprobe ðt; lÞ Iref ðt; lÞ
! ð4Þ
Every spectrum was recorded 200 times, and the averaged spectrum was used in further data analysis. The concentration of the sample was adjusted to an absorbance of 0.3 OD at l = 400 nm in a 1 mm path length quartz cuvette. No photodegradation was observed after femtosecond transient absorption spectroscopy measurements. In femtosecond transient absorption spectroscopy measurements, the population dynamics modeling toolbox software, developed by van Wilderen et al.,[54] was used to analyze the differential absorbance DA(t,l) as a function of wavelength and time delay. Spectral chirp in the transient absorption spectra were corrected for group velocity dispersion of the probe beam. Target model analysis was utilized to globally analyze the transient absorption spectroscopy data, which is a superposition of different spectral components, el(l), weighted by their concentration, cl(t) [Eq. (5)]:[55] DAðt; lÞ ¼
n X
cl ðt Þel ðlÞ
ð5Þ
l¼1
Nanosecond Transient Absorption Spectroscopy Measurements Nanosecond transient absorption spectra were obtained by using an LP920 laser flash photolysis spectrometer (Edinburgh Instruments UK). In this setup, samples were excited by using the 355 nm output from a pulse Nd:YAG laser (10 Hz, 8 ns) (continuum Surelite) as the excitation source. Samples were specifically prepared (bubbled with nitrogen or oxygen for 15 min unless otherwise stated) for each measurement and adjusted to an absorbance of around 0.3 OD in 10 mm path length quartz cuvettes at the laser wavelength used.
Acknowledgements This work was supported by the 973 Program (2013CB834604), National Natural Science Foundation of China (NSFC) (21173235, ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
91233107, 21127003, 21333012 , and 21373232), and the Chinese Academy of Sciences (XDB12020200). Keywords: donor–p–acceptor · charge transfer · reaction mechanisms · solvation · solvatochromism [1] S. M. Bachilo, A. F. Benedetto, R. B. Weisman, J. R. Nossal, W. E. Billups, J. Phys. Chem. A 2000, 104, 11265 – 11269. [2] H. Uoyama, K. Goushi, K. Shizu, H. Nomura, C. Adachi, Nature 2012, 492, 234 – 238. [3] X. Xiong, F. Song, J. Wang, Y. Zhang, Y. Xue, L. Sun, N. Jiang, P. Gao, L. Tian, X. Peng, J. Am. Chem. Soc. 2014, 136, 9590 – 9597. [4] T. Nakagawa, S.-Y. Ku, K.-T. Wong, C. Adachi, Chem. Commun. 2012, 48, 9580 – 9582. [5] J. Chan, S. C. Dodani, C. J. Chang, Nat. Chem. 2012, 4, 973 – 984. [6] C. Baleiz¼o, M. N. Berberan-Santos, ChemPhysChem 2010, 11, 3133 – 3140. [7] H. Nakanotani, T. Higuchi, T. Furukawa, K. Masui, K. Morimoto, M. Numata, H. Tanaka, Y. Sagara, T. Yasuda, C. Adachi, Nat. Commun. 2014, 5, 1 – 7. [8] M. A. Baldo, D. F. O’Brien, M. E. Thompson, S. R. Forrest, Phys. Rev. B 1999, 60, 14422 – 14428. [9] M. Segal, M. A. Baldo, R. J. Holmes, S. R. Forrest, Z. G. Soos, Phys. Rev. B 2003, 68, 075211. [10] B. H. Wallikewitz, D. Kabra, S. G¦linas, R. H. Friend, Phys. Rev. B 2012, 85, 045209. [11] C. Baleiz¼o, M. N. Berberan-Santos, ChemPhysChem 2011, 12, 1247 – 1250. [12] S. Boudin, J. Chim. Phys. Phys. 1930, 27, 285 – 290. [13] C. A. Parker, C. G. Hatchard, Trans. Faraday Soc. 1961, 57, 1894 – 1904. [14] H. Tanaka, K. Shizu, H. Miyazaki, C. Adachi, Chem. Commun. 2012, 48, 11392 – 11394. [15] Q. Zhang, J. Li, K. Shizu, S. Huang, S. Hirata, H. Miyazaki, C. Adachi, J. Am. Chem. Soc. 2012, 134, 14706 – 14709. [16] J. Lee, K. Shizu, H. Tanaka, H. Nomura, T. Yasuda, C. Adachi, J. Mater. Chem. C 2013, 1, 4599 – 4604. [17] K. Nasu, T. Nakagawa, H. Nomura, C. J. Lin, C. H. Cheng, M. R. Tseng, T. Yasuda, C. Adachi, Chem. Commun. 2013, 49, 10385 – 10387. [18] H. Tanaka, K. Shizu, H. Nakanotani, C. Adachi, Chem. Mater. 2013, 25, 3766 – 3771. [19] R. Ishimatsu, S. Matsunami, T. Kasahara, J. Mizuno, T. Edura, C. Adachi, K. Nakano, T. Imato, Angew. Chem. Int. Ed. 2014, 53, 6993 – 6996; Angew. Chem. 2014, 126, 7113 – 7116. [20] T. Komino, H. Tanaka, C. Adachi, Chem. Mater. 2014, 26, 3665 – 3671. [21] H. Tanaka, K. Shizu, H. Nakanotani, C. Adachi, J. Phys. Chem. C 2014, 118, 15985 – 15994. [22] Y. Gong, X. Guo, S. Wang, H. Su, A. Xia, Q. He, F. Bai, J. Phys. Chem. A 2007, 111, 5806 – 5812. [23] G. Ramakrishna, T. Goodson III, J. Phys. Chem. A 2007, 111, 993 – 1000. [24] G. Ramakrishna, A. Bhaskar, T. Goodson III, J. Phys. Chem. B 2006, 110, 20872 – 20878. [25] M. Jia, X. Ma, L. Yan, H. Wang, Q. Guo, X. Wang, Y. Wang, X. Zhan, A. Xia, J. Phys. Chem. A 2010, 114, 7345 – 7352. [26] L. Yan, X. Chen, Q. He, Y. Wang, X. Wang, Q. Guo, F. Bai, A. Xia, D. Aumiler, S. Vdovic´, S. Lin, J. Phys. Chem. A 2012, 116, 8693 – 8705. [27] M. Glasbeek, H. Zhang, Chem. Rev. 2004, 104, 1929 – 1954. [28] C. Reichardt, Chem. Rev. 1994, 94, 2319 – 2358. [29] K. Santhosh, A. Samanta, ChemPhysChem 2012, 13, 1956 – 1961. [30] S.-J. Chung, K.-S. Kim, T.-C. Lin, G. S. He, J. Swiatkiewicz, P. N. Prasad, J. Phys. Chem. B 1999, 103, 10741 – 10745. [31] D. J. Stewart, M. J. Dalton, R. N. Swiger, J. L. Fore, M. A. Walker, T. M. Cooper, J. E. Haley, L.-S. Tan, J. Phys. Chem. A 2014, 118, 5228 – 5237. [32] F. Wìrthner, G. Archetti, R. Schmidt, H.-G. Kuball, Angew. Chem. Int. Ed. 2008, 47, 4529 – 4532; Angew. Chem. 2008, 120, 4605 – 4608. [33] X. Guo, S. Wang, A. Xia, H. Su, J. Phys. Chem. A 2007, 111, 5800 – 5805. [34] Gaussian 09 (Revision A.2), 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,
2364
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles
[35] [36]
[37] [38] [39] [40] [41] [42] [43] [44]
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. 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. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford, CT, 2009. M. Sun, P. Kjellberg, F. Ma, T. Pullerits, Chem. Phys. Lett. 2005, 401, 558 – 564. Y. Wan, L. Yan, Z. Zhao, X. Ma, Q. Guo, M. Jia, P. Lu, G. Ramos-Ortiz, J. L. Maldonado, M. Rodrguez, A. Xia, J. Phys. Chem. B 2010, 114, 11737 – 11745. N.-K. Persson, M. Sun, P. Kjellberg, T. Pullerits, O. Ingans, J. Chem. Phys. 2005, 123, 204718. E. Lippert, Z. Naturforsch. A 1955, 10, 541 – 545. N. Mataga, Y. Kaifu, M. Koizumi, Bull. Chem. Soc. Jpn. 1956, 29, 465 – 470. N. Mataga, Bull. Chem. Soc. Jpn. 1963, 36, 654 – 662. B. Li, R. Tong, R. Zhu, F. Meng, H. Tian, S. Qian, J. Phys. Chem. B 2005, 109, 10705 – 10710. R. Ishimatsu, S. Matsunami, K. Shizu, C. Adachi, K. Nakano, T. Imato, J. Phys. Chem. A 2013, 117, 5607 – 5612. A. Kapturkiewicz, J. Herbich, J. Karpiuk, J. Nowacki, J. Phys. Chem. A 1997, 101, 2332 – 2344. V. A. Galievsky, S. I. Druzhinin, A. Demeter, P. Mayer, S. A. Kovalenko, T. A. Senyushkina, K. A. Zachariasse, J. Phys. Chem. A 2010, 114, 12622 – 12638.
ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
[45] H. J. Verhey, B. Gebben, J. W. Hofstraat, J. W. Verhoeven, J. Polym. Sci. Part A 1995, 33, 399 – 405. [46] W. Verbouwe, M. Van der Auweraer, F. C. De Schryver, J. J. Piet, J. M. Warman, J. Am. Chem. Soc. 1998, 120, 1319 – 1324. [47] D. Panda, A. Datta, J. Chem. Phys. 2006, 125, 054513. [48] J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Springer Science, New York, 2006. [49] E. V. Sitzmann, J. G. Langan, D. Griller, K. B. Eisenthal, Chem. Phys. Lett. 1989, 161, 353 – 360. [50] J. Wang, J. Kubicki, H. Peng, M. S. Platz, J. Am. Chem. Soc. 2008, 130, 6604 – 6609. [51] L. Reynolds, J. A. Gardecki, S. J. V. Frankland, M. L. Horng, M. Maroncelli, J. Phys. Chem. 1996, 100, 10337 – 10354. [52] J. E. Rogers, T. M. Cooper, P. A. Fleitz, D. J. Glass, D. G. McLean, J. Phys. Chem. A 2002, 106, 10108 – 10115. [53] M. Zhou, S. Vdovic, S. R. Long, M. Z. Zhu, L. Y. Yan, Y. Y. Wang, Y. L. Niu, X. F. Wang, Q. J. Guo, R. C. Jin, A. D. Xia, J. Phys. Chem. A 2013, 117, 10294 – 10303. [54] L. J. G. W. van Wilderen, C. N. Lincoln, J. J. van Thor, PLoS One 2011, 6, e17373. [55] I. H. M. van Stokkum, D. S. Larsen, R. van Grondelle, Biochim. Biophys. Acta Bioenerg. 2004, 1657, 82 – 104.
Received: April 3, 2015 Revised: April 25, 2015 Published online on May 28, 2015
2365
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles
Photophysical Properties of Intramolecular Charge Transfer in a Tribranched Donor–p–Acceptor Chromophore Jiangpu Hu,[a] Yang Li,[a] Huaning Zhu,[a] Shuhai Qiu,[b] Guiying He,[a] Xiaozhang Zhu,*[b] and Andong Xia*[a] The photophysical properties of intramolecular charge transfer (ICT) in a novel tribranched donor–p–acceptor chromophore, triphenoxazine-2,4,6-triphenyl-1,3,5-triazine (tri-PXZ-TRZ), with thermally activated delayed fluorescence character was investigated in different aprotic solvents by steady-state spectroscopy and femtosecond and nanosecond transient absorption spectroscopy measurements. Increasing the solvent polarity led to a significant increase in the Stokes shift. The large Stokes shift in highly polar solvents was attributed to ICT properties upon excitation; this resulted in a strong interaction between the tri-PXZ-TRZ molecule and the surrounding solvent, which led to a strong solvation process. Quantum-chemical calculations and changes in the dipole moment showed that
this compound has a large degree of ICT. Furthermore, an apolar environment helped to preserve the symmetry of triPXZ-TRZ and to enhance its emission efficiency. The femtosecond and nanosecond transient absorption spectroscopy results indicated that the excited-state dynamics of this push–pull molecule were strongly influenced by solvent polarity through the formation of a solvent-stabilized ICT state. The excitedstate relaxation mechanism of tri-PXZ-TRZ was proposed by performing target model analysis on the femtosecond transient absorption spectra. In addition, the delayed fluorescence of tri-PXZ-TRZ was significantly modulated by a potential competition between solvation and intersystem crossing processes.
1. Introduction With high luminescence efficiency, thermally activated delayed fluorescence (TADF) materials have in recent years received considerable attention for various applications in terms of producing organic light-emitting diodes and fluorescence probes for bioimaging.[1–7] In contrast to traditional fluorescent and phosphorescent materials, which can only harvest the singlet (25 %) or the triplet excitons (75 %), TADF materials can theoretically harvest almost 100 % of the excitons from both the singlet and triplet excited states.[8–11] These materials can provide an additional singlet exciton generated by efficient upconversion from the lowest triplet excited state (T1) to the lowest singlet excited state (S1) through reverse intersystem crossing. Such thermally activated delayed fluorescence was first observed by Boudin in 1930[12] and was later elucidated by Parker and Hatchard.[13] Generally, to obtain highly efficient upconversion from T1 to S1, a sufficiently small energy gap between S1 and T1 (DEST) is required. Theoretically, effective sepa[a] J. Hu, Y. Li, H. Zhu, G. He, Prof. A. Xia Beijing National Laboratory for Molecular Sciences, Institute of Chemistry, Chinese Academy of Sciences, Beijing, 100190 (China) E-mail: [email protected] [b] S. Qiu, Prof. X. Zhu Beijing National Laboratory for Molecular Sciences, Key Laboratory of Organic Solids, Institute of Chemistry, Chinese Academy of Sciences, Beijing, 100190 (China) E-mail: [email protected] Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/cphc.201500290.
ChemPhysChem 2015, 16, 2357 – 2365
ration between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) will lead to a very small DEST. On the basis of this concept, a number of efficient organic TADF materials have been designed and synthesized.[2–4, 14–21] One strategy is to introduce a donor–acceptor (D-A) group into the TADF system, as it can induce intramolecular charge transfer (ICT) and enhance separation between the HOMO and LUMO upon excitation.[18, 22–26] It is well known that the local environment, especially solvents, has a profound effect on the emission behavior of ICTtype compounds.[22–29] Once these compounds are excited to highly excited states, the excess amount of energy will be transferred to the surrounding medium because of solvent reorganization (solvation). Therefore, solvent effects on the excited-state dynamics of ICT-type TADF materials could be important to understand the ICT process and its complex spectral properties. However, in solvents, the excited-state dynamics of these compounds remain largely unclear because of the fast and complicated solvation process, which is usually coupled with vibrational cooling and ICT processes after excitation, and they take place on the timescale of picoseconds and even femtoseconds. Then, to understand comprehensively the underlying emission mechanisms of TADF materials in various environments, it is necessary to perform photophysical studies on ICTtype TADF materials in media with different polarities. In this paper, a donor–p–acceptor chromophore, triphenoxazine-2,4,6-triphenyl-1,3,5-triazine (tri-PXZ-TRZ), that has been proven to have strong delayed fluorescence from ICT[18] was chosen as a model system to investigate the ICT process in sol-
2357
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles vents with varying polarities by using steady-state spectroscopy and femtosecond, as well as nanosecond, time-resolved transient absorption spectroscopy measurements. Figure 1 shows the molecular structure of tri-PXZ-TRZ, which has a phenyltriazine (TRZ) electron-acceptor group surrounded by three phenoxazine (PXZ) electron-donor groups. It was found that tri-PXZ-TRZ has representative TADF character with typical ICT
Figure 2. Normalized UV/Vis absorption and fluorescence spectra of tri-PXZTRZ in toluene and THF.
Figure 1. Molecular structure of tri-PXZ-TRZ.
properties upon excitation. Such ICT will lead to a strong interaction between the solute molecules and the surrounding solvent, which is correlated to a large Stokes shift of the steadystate spectrum. The femtosecond and nanosecond time-resolved transient absorption spectroscopy measurements provide solvent-dependent information about the relaxation dynamics, in which the formation of a solvent-stabilized ICT state (denoted as the ICT’ state) upon excitation is included.[22, 25] These measurements suggest that the solvent plays a very important role in the excited-state dynamics of the ICT state, which is helpful for us to understand changes in the upconversion efficiencies of TADF materials in different media.
2. Results and Discussion 2.1. Steady-State Absorption and Fluorescence Spectra Figure 2 shows the normalized optical absorption and fluorescence spectra of tri-PXZ-TRZ in toluene and tetrahydrofuran (THF). The steady-state spectral parameters of this molecule in
several typical solvents with different polarities are listed in Table 1. It can be seen that tri-PXZ-TRZ exhibits two absorption bands at room temperature. The absorption bands located at the high-energy side with a sharp intense absorption band around l = 285 nm in toluene and around l = 270 nm in THF are mainly attributed to the p–p* transition derived from the TRZ acceptor moiety.[21] The low-energy bands around l = 436 nm in toluene and l = 422 nm in THF originate from ICT transitions.[18, 22, 25, 30] It should be realized that there is a general slight blueshift in the absorption bands with an increase in solvent polarity, which indicates a small amount of ground-state ICT; this is attributed to strong interaction between the solvent and solute molecules.[31, 32] That is, the conjugated push–pull tri-PXZ-TRZ molecule adopts its geometry with lower conjugation in more polar solvent (e.g. THF), which leads to a slight blueshift in the absorption bands relative to that in less polar solvents (e.g. toluene) with larger conjugation (as shown in Figure 2). Unlike the small blueshift in the absorption spectrum, a very large solvent-induced redshift in the fluorescence spectrum for tri-PXZ-TRZ was observed as the solvent polarity was increased (see Figure S1, Supporting Information). Such broad, unresolved vibronic structure fluorescence bands with relatively large bathochromic shifts in different solvents are typical properties of fluorescence from the relaxed ICT (ICT’) states.[22, 25, 33] This remarkable solvatochromism indicates strong solvent effects in the ICT state of tri-PXZ-TRZ.
Table 1. Solvent effects on the absorption and emission properties of tri-PXZ-TRZ. Solvent
Parameter[a] e n
Df
toluene MS1[c] chloroform MS2[c] THF dichloromethane
2.4 3.4 4.8 6.0 7.6 8.9
0.0159 0.0894 0.155 0.177 0.210 0.217
1.494 1.477 1.443 1.432 1.405 1.421
la[b] [nm]
lf[b] [nm]
Stokes shift [cm¢1]
Quantum yield
436 432 429 426 422 420
573 604 625 638 648 651
5484 6592 7310 7800 8264 8449
0.21 0.044 0.021 0.0026 0.0011 0.0010
[a] Parameters are taken from ref. [51]; Df: Solvent parameter. [b] la : wavelength of absorption maxima of ICT absorption; lf : wavelength of fluorescence maxima. [c] MS: mixed solvents of toluene and THF; MS1 and MS2 represent toluene/THF with volume ratios of about 8:2 and 3:7, respectively. The corresponding emix and nmix were then calculated according to emix = faea + fbeb and nmix2 = fana2 + fbnb2, respectively; fa and fb are the volume percentages of each solvent.[41]
ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
2358
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles 2.2. Quantum-Chemical Calculations To obtain deeper insight into the nature of the ICT state and further to interpret the observed steady-state spectral behavior, quantum-chemical calculations for tri-PXZ-TRZ were performed by density functional theory (DFT) methods, as implemented in the Gaussian 09 package.[34] The ground-state geometry of tri-PXZ-TRZ was fully optimized at the WB97X-D/ccpVDZ level. Due to steric repulsion between the hydrogen atoms at the donor–acceptor linkage, a twist-optimized geometry of tri-PXZ-TRZ with a large dihedral angle (818) between the donor and acceptor planes was obtained (Figure S2 a). The twist structure between the PXZ and TRZ units will suppress delocalization of the electron density and will result in effective charge separation between the HOMO and LOMO. The molecular orbitals making major contributions to the six low-lying singlet excited states for tri-PXZ-TRZ are shown in Figure S3. These clearly show the charge-transfer nature of the optical HOMO–LUMO transition, which is responsible for the ICT absorption bands in Figure 2. To demonstrate the charge distribution and the local exited character, charge different density (CDD) of the tri-PXZ-TRZ molecule was also calculated by timedependent (TD) DFT at the WB97X-D/cc-pVDZ level, which allowed visualization of the difference in the electron densities upon excitation between the ground state (S0) and the excited state.[35–37] The CDD image of tri-PXT-TRZ is depicted in Figure S2 b. It can be observed that the charges are almost entirely distributed on the TRZ unit and the holes are mainly distributed on the three PXZ units; this is direct evidence for charge transfer from the HOMO to LUMO with a shift in the electron density from the outside PXZ moieties to the core TRZ moiety. 2.3. Solvation Effects: Solvent Parameters and Spectral Properties As mentioned above, the absorption spectrum of tri-PXZ-TRZ is slightly influenced by solvent polarity, whereas the extent of the redshift of the fluorescence spectrum is strongly dependent on solvent polarity. Consequently, the excited states of tri-PXZ-TRZ will relax to a strongly polar-dependent ICT state in a highly polar medium upon excitation, and this will lead to a large Stokes shift. To elaborate on the solvatochromism and the degree of ICT, as well as the polar nature of the excited states of tri-PXZ-TRZ, we employed the Lippert–Mataga equation [Eqs. (1) and (2)]:[38–40] Dn ¼ nabs ¢v em ¼
2 Dm2 Df þ constant ðhca3 Þ
ðe¢1Þ ðn2 ¢1Þ Df ¼ ¢ ð2 e þ 1Þ ð2 n2 þ 1Þ
ð1Þ ð2Þ
in which nabs and nem are the absorption and emission bands in wavenumbers, respectively; h is Planck’s constant; c is the speed of light; a is the Onsager cavity radius defined as the solvent shell around the solute molecule; and Dm = me¢mg is the difference between the excited-state (me) and ground-state ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
(mg) dipole moments as presented by Equation (1). The Df term is the orientation polarizability of the solvent deduced from the dielectric constant (e) and the refractive index (n) of the solvent as represented by Equation (2). On the basis of the solvent parameters and Stokes shifts summarized in Table 1, a plot of Stokes shifts versus the solvent polarity parameters for tri-PXZ-TRZ is shown in Figure 3.
Figure 3. Stokes shift versus solvent polarity (Df) for tri-PXZ-TRZ. The solid line is the least-squares fit result.
The Stokes shifts depend approximately linearly on the solvent polarities. To quantitatively determine the polarity of the excited ICT state, we examined the dipole moment change between the excited-state and ground-state dipole moments (Dm = me¢mg) from the solvatochromic shift in polar solvents on the basis of Equations (1) and (2). Estimated from quantumchemical calculations by using the DFT method at the WB97XD/cc-pVDZ level, the Onsager radius was found to be 7.27 æ for tri-PXZ-TRZ. According to the slope obtained from the fitting results shown in Figure 3, the effective dipole moment change (Dm) was calculated to be about 23.3 D. Such a large dipole moment change for tri-PXZ-TRZ indicates remarkable intramolecular charge-transfer character,[2, 18, 42–44] which is consistent with the DFT calculation results. In addition, as shown in Figure 2, there is only a slight blueshift in the absorption spectrum, but a very large redshift in the fluorescence spectrum with an increase in solvent polarity, which indicates that the dipole moment in the excited state is much larger than that in the ground state. Therefore, the difference (Dm) mainly reflects the change in the dipole moment in the excited state. The solvatochromic shift in polar solvents is due to the formation of a polar ICT excited state. 2.4. Anisotropy Measurements To investigate further the environmental polarity effect on the nature of ICT in tri-PXZ-TRZ, steady-state fluorescence excitation anisotropy experiments were performed in matrices with different polarities. The anisotropy values (r) were obtained by measuring the fluorescence excitation spectrum in the ICT absorption range from l = 380 to 500 nm on the basis of the steady-state absorption spectrum (Figure 2). To avoid rotation of chromophores during excited-state relaxation, the molecules were embedded in proper polymer matrices. By choosing ap-
2359
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles propriate matrices with specific polarities, we can provide a more comprehensive explanation on the polarity-dependent excited character of tri-PXZ-TRZ. For our study, Zeonex E48R (Zeonex) and poly-THF were chosen as matrices to perform the anisotropy experiments, as the polarities of Zeonex and poly-THF are similar to the polarities of methylcyclohexane and THF, respectively.[45] Figure 4 plots the fluorescence excitation
Figure 4. Fluorescence excitation anisotropy (c) and normalized fluorescence excitation spectra (a) of tri-PXZ-TRZ in Zeonex and poly-THF. The fluorescence excitation spectra were measured by monitoring the fluorescence wavelength at 530 nm in Zeonex and at 550 nm in poly-THF.
spectra and the corresponding anisotropy spectra of tri-PXZTRZ in these two matrices. In apolar Zeonex, the anisotropy constant values of tri-PXZ-TRZ are around 0.1–0.2 throughout the ICT excitation band, which indicates that the C3 symmetry of the ICT states in tri-PXZ-TRZ is preserved in the low-polar matrix, in which the ICT redistributes the excitation energy among the three spatially degenerate transition dipole moments.[46] However, in poly-THF the anisotropy values are between 0.3 and 0.4 in the region of the ICT band, which suggests that the excited state is localized in one branch of the tri-PXZ-TRZ molecule in a polar medium. This may be due to the fact that a matrix with high polarity will lead to torsional disorder and, hence, reduce the global symmetry of the three branches in tri-PXZ-TRZ. The slight blueshift in the absorption spectrum displayed in Figure 2 recorded in THF relative to the spectrum recorded in toluene also suggests the formation of a polar excited state that is localized in one branch of the molecule. 2.5. Solvent Polarity Effect on Fluorescence Quantum Yields and Lifetimes Table 1 lists the fluorescence quantum yields (Ff) of tri-PXZTRZ in various solvents. As plotted in Figure S4, the Ff values decrease dramatically with an increase in solvent polarity. This is consistent with the energy gap law, for which an increased Stokes shift in a more polar solvent is often accompanied by a decrease in fluorescence quantum yield due to an increase in the radiationless rate.[47, 48] This indicates that an apolar solvent could facilitate fluorescence transition and that the suppression of the radiationless process would be greater than that in a polar solvent. As a result, an apolar environment will benefit ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
the fluorescence process and thereby increase the emission efficiency of tri-PXZ-TRZ. To study the effect of solvent on fluorescence lifetime, we measured the fluorescence decay of tri-PXZ-TRZ at its fluorescence maximum in the respective solvents. The fluorescence decay curves of tri-PXZ-TRZ in toluene and THF are depicted in Figure S5. In toluene, there are two clear components of the decay curve. The first component was assigned to prompt fluorescence, which has a lifetime of about 17 ns. The second component was assigned to delayed fluorescence, which exhibited a lifetime of around 1 ms. These values agree with the former report.[18] Notably, the lifetime of delayed fluorescence is much longer than that of prompt fluorescence, because delayed population of the excited singlet state originates from the long-lived T1 state through reverse intersystem crossing.[2, 18] However, in contrast to the case in toluene, only a dramatic decrease in the decay curve was observed, and the slow decay component disappeared in THF. Consequently, delayed population of the excited singlet state was largely decreased because of the strong interaction between the excited solute molecules and the solvent molecules, which efficiently suppresses the intersystem crossing (ISC) process.[49, 50] The lifetime in THF is too short to be measured by the same instrument as that used to measure the lifetime in toluene, so time-correlated single photon counting (TCSPC) was performed to measure the fluorescence lifetime in THF. The fluorescence lifetime was measured to be only 1.4 ns (see Figure S6), and no delayed fluorescence was found in THF. Furthermore, the dramatic decrease in the decay curve in higher polarity solvents was mainly attributed to a solvation-induced strong radiationless transition process, as this makes up a larger portion of the whole deactivation process of the excited state than the lower polarity solvent. We calculated the radiationless transition rate in both solvents on the basis of the quantum yields and lifetimes of the prompted and delayed fluorescence [Eq. (S1)]. In toluene, the radiationless rate was 2.25 Õ 108 s¢1. In THF, the delayed fluorescence rate was neglected, in view of the absence of a delayed component (Figure S5), and thus, the radiationless transition rate was 6.48 Õ 1011 s¢1, which is markedly faster than that in toluene. To further investigate the solvation process, femtosecond time-resolved transient absorption spectroscopy measurements were employed to provide additional solventdependent dynamic information about the excited state of triPXZ-TRZ. 2.6. Femtosecond Time-Resolved Transient Absorption Spectroscopy Femtosecond time-resolved transient absorption spectroscopy measurements of tri-PXZ-TRZ in different solvents were performed to gain more insight into the polar-dependent ICT state after excitation at l = 400 nm. Figure 5 illustrates the broad transient absorption spectra of tri-PXZ-TRZ obtained at different time delays from 0 fs to 1 ns in toluene and in THF. The transient spectra in both solvents are composed of a very broad excited-state absorption (ESA, positive transient absorption signal) that over the whole spectral region overlaps with
2360
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles that there is a dramatic decay in the transient spectra in the following time domain of 3 to 1000 ps (Figure 5 d). This is quite different from the case in toluene. Such decay mainly originates from fast radiationless relaxation in the highly polar solvent, which leads to a low fluorescence quantum yield in THF. As mentioned above, it is clear that the excited tri-PXZ-TRZ molecules will relax through solvation and ISC processes simultaneously at the initial time delay. Hence, there is a potential competition between solvationcoupled ICT decay and ISC during the relaxation processes. Figure 5. Evolution of the transient absorption spectra of tri-PXZ-TRZ at different delay times: a, b) in toluene and c, d) in THF. Furthermore, the energy surface of the ICT state will decrease much faster in polar solvents ground-state bleach (GSB, negative transient absorption signal) than in apolar solvents due to the strong solvation process. Acand stimulated emission (SE, negative transient absorption cordingly, kinetic models of tri-PXZ-TRZ in different solvents signal). were reasonably proposed, as shown in Figure 6. For a detailed Taking a closer look at the initial transient absorption spectra understanding of the femtosecond transient absorption specin toluene, the dip around l = 450 nm can be readily assigned tra, target analysis based on the proposed kinetic model was to the ground-state bleach according to the ICT absorption in then employed to extract the time independent correlations the steady-state absorption spectrum. Broad ESA bands from these transient absorption spectroscopy data. The species around l = 525 and 750 nm can be found over the whole meaassociated difference spectra (SADS) and concentration kinetics sured time window. Fast dynamics occur in the time range of for tri-PXZ-TRZ were extracted and are plotted in Figure 7. To approximately 0–10 ps (Figure 5 a), and this is characterized by show the quality of fitting, kinetics at selected wavelengths of an increase in the ESA bands, which may be attributed to soll = 450, 525, 575, and 750 nm in toluene and at l = 450, 525, vation and/or ISC. Furthermore, the SE band exhibits a redshift 600, and 750 nm in THF are shown in Figure S7 together with from an initial position of l = 550 nm to a final position that a global fit of all the collected time traces. Three components nearly corresponds to the steady-state fluorescence spectrum were required for an adequate fit of the transient spectroscopy of l = 575 nm with an increase in the delay time; this is indicadata by using the applied target dynamic schemes (Figure 6). tive of the existence of a solvation process that could gradually The estimated rate constants are summarized in Table 2. lower the excited-state potential energy surface with time In toluene, the first component was assigned to the ICT delay. However, no clear decays or shifts in the spectra were state upon excitation at l = 400 nm. The ICT state decays into two independent excited-state populations, the solvent-stabifound (Figure 5 b) in the following 10–1000 ps, which suggests the formation of a long-lived state. lized ICT state (ICT’ state) and the T1 state, with rate constants In contrast to the spectra in toluene, interesting transient of (2.98 ps)¢1 and (11.5 ps)¢1 by solvation and ISC processes, siabsorption features were observed in THF. From Figure 5 c, the multaneously. As shown in Figure 7 a, it is not surprising to see initial process occurs in the timeframe of about 0–3 ps, which is much faster than that in toluene. This was attributed to a stronger interaction between the solute and solvent molecules, which facilitates the solvation process.[51] Furthermore, the SE band exhibits a larger redshift from l = 550 to about 600 nm with an increase in the delay times, which indicates that the potential energy surface of the ICT state in the polar solvent is much lower than that in the apolar solvent, and this caused a bathochromic shift in the steady-state fluorescence spectra as the polarity increased. Furthermore, a local minimum is observed at about l = 720 nm at 3 ps (Figure 5 c, d), and it is also attributed to decay of stimulated emission from Figure 6. Proposed relaxation models of tri-PXZ-TRZ used in target analysis: the solvation-coupled ICT (ICT’) state. It should also be realized a) in toluene and b) in THF. ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
2361
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles ative to the rate constant in toluene, such a large rate constant indicates that the radiationless transition makes up a much larger portion of this relaxation process, which is consistent with the fluorescence quantum yields and lifetime results. 2.7. Nanosecond Time-Resolved Transient Absorption Spectroscopy Nanosecond time-resolved transient absorption spectroscopy measurements were performed to further interpret the solvation-dependent emission characteristics of tri-PXZ-TRZ. Figure 8 shows the nanosecond transient Figure 7. SADS from target analysis of tri-PXZ-TRZ: a) in toluene and b) in THF. Concentrations of the SADS comabsorption spectra obtained in ponents as a function of time: c) in toluene and d) in THF. toluene and THF. There are two negative bands at around l = 445 and 575 nm and three positive bands at around l = 400, 510, and 750 nm in toluene. As depicted in Figure S8, the tranTable 2. Rate constants estimated from target analysis of tri-PXZ-TRZ in sient absorption spectra of tri-PXZ-TRZ are similar to each different solvents. other at delay times of 500 ns (from the nanosecond measurek2 [ps¢1] k3 [ps¢1] Solvent k1 [ps¢1] ments) and 1 ns (from the femtosecond measurements). In detail, the negative bands at about l = 445 and 575 nm were toluene 3.0 0.2 11.5 1.0 > 1000 THF 1.0 0.1 30.3 3.0 319 30 readily assigned to GSB and SE, respectively, according to the steady-state and femtosecond transient absorption spectra mentioned above. The three positive bands were attributed to the ESA band of the long-lived T1 state considering the microthat the SADS corresponding to the ICT’ state and the T1 state second scale of the decay process. The lifetime of the T1 state are similar to each other because of the small energy gap is about 0.95 ms, obtained by fitting the time trace of the ESA (DEST) between them. The third component (> 1 ns) was attributed to decay from the ICT’ state to the ground state. Three components were also used for an adequate fit of the transient spectroscopy data in THF. On the basis of the results in toluene, the three components were assigned to the ICT, ICT’, and T1 states. However, the solvation rate constant was (1.1 ps)¢1 in THF, which is much larger than that in toluene, whereas the ISC rate constant was (30.3 ps)¢1 in THF, which is much smaller than that in toluene. As a consequence, in highly polar THF the formation of the ICT’ state is very fast, because strong solvation occurs, which leads to a much lower population of the T1 state relative to that in an apolar solvent. In addition, the strong solvation process may lead to a decrease in the energy of the ICT’ state to a level that is lower than that of the T1 state, and this can also cause a lower population of the T1 state. As displayed in Figure 7 b, the shape of the SADS corresponding to the T1 state becomes abnormal, and this is due to the low population of the T1 state (Figure 7 d). Thus, the emission behavior of tri-PXZ-TRZ can be significantly modulated by solvent polarity through competition between solvation and ISC. The third rate constant of (319 ps)¢1 was assigned to Figure 8. Nanosecond time-resolved transient difference absorption spectra the relaxation rate from the ICT’ state to the ground state. Relof tri-PXZ-TRZ at different delay times: a) in toluene and b) in THF. ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
2362
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles band at l = 510 nm (Figure S9), which is close to the lifetime of delayed fluorescence. It is known that triplet excitons can be inactivated by triplet oxygen molecules. In consequence, to determine the importance of the triplet excited state in triPXZ-TRZ, nanosecond transient absorption spectroscopy experiments were performed after oxygenation of a toluene solution by bubbling oxygen through it. All signals were quenched after oxygen was bubbled into the toluene solution (Figure S10), which indicates that the triplet excited state plays a very significant role in this relaxation process. However, all of the signals also disappeared if the nanosecond transient absorption spectroscopy experiment was performed in THF (Figure 8 b). According to the results of the femtosecond transient absorption spectroscopy experiments, quenching of the signals originates from the strong solvation process, which leads to suppression of the ISC process. Overall, in highly polar solvents the strong solvation process will effectively quench the delayed fluorescence of tri-PXZ-TRZ.
3. Conclusions
Steady-State Absorption and Fluorescence Spectroscopy Measurements Absorption and fluorescence spectra were measured by using a UV/Vis spectrophotometer (U-3010, Hitachi Japan) and a fluorescence spectrophotometer (F-4600, Hitachi Japan) at ambient temperature, respectively. Fluorescence quantum yield (Ff) measurements were performed by the comparative method by using the Ff of fluorescein in 0.1 m NaOH (Ff = 0.9) as the reference.[52] In these measurements, all samples were bubbled with nitrogen for 15 min. Fluorescence excitation anisotropy spectra were measured by a fluorescence spectrophotometer (F-4600, Hitachi Japan) with two polarizers in excitation and detection light routes. The anisotropy value (r) was calculated with Equation (3): r¼
In summary, we presented the results of the spectral properties and dynamics of the ICT states of a tribranched donor–p–acceptor chromophore (i.e. tri-PXZ-TRZ) by means of steady-state spectroscopy and femtosecond, as well as nanosecond, transient absorption spectroscopy measurements. The steady-state spectral results showed that increasing the solvent polarity led to a slight blueshift in the absorption bands and to a remarkably large redshift in the fluorescence bands for tri-PXZ-TRZ. Chemical calculation results showed that tri-PXZ-TRZ has typical ICT character upon excitation. The large change in the dipole moment ( 23.3 D) of tri-PXZ-TRZ between the ground state and the excited state also proved the formation of a polar-dependent ICT state. This ICT state resulted in a strong interaction between the solute molecules and the surrounding solvent, which induced a solvation process, and this was correlated to a large Stokes shift in the steady-state spectrum. In addition, the strong solvation effects and dynamics of tri-PXZTRZ were also investigated by femtosecond and nanosecond transient absorption spectroscopy measurements. It was found that the solvent polarity had a significant impact on the dynamics of the ICT state of tri-PXZ-TRZ. The solvation process was much stronger in a polar solvent than in an apolar solvent upon excitation, but the ISC process was weaker in a polar solvent than in an apolar solvent . Consequently, the emission behavior of tri-PXZ-TRZ could be efficiently modulated by choosing proper polar solvents through this potential competition between solvation and ISC.
Experimental Section Materials The synthesis and purification of tri-PXZ-TRZ were done completely as previously reported.[18] The chemical structure and purity were identified by NMR spectroscopy, MALDI-TOF-MS, and elemental analysis. All the aprotic solvents including toluene, chloroform, THF, and dichloromethane used in this work were of analytical reChemPhysChem 2015, 16, 2357 – 2365
agent grade or higher and were purchased from the Beijing Chemical Plant without further purification. Fluorescein was purchased from Sigma–Aldrich. Zeonex E48R (Zeon Japan) and poly-THF (Aladdin) were used as polymer matrices in anisotropy detection.
www.chemphyschem.org
Ik ¢GI? Ik þ 2GI?
ð3Þ
in which I k and I ? are the polarized fluorescence intensities parallel and perpendicular to excitation polarization, respectively, and G (G = I ? /I k ) is the geometrical factor of fluorescence spectrophotometer if the excitation is vertically polarized. To avoid fast rotation of the molecule during the fluorescence lifetime measurements, triPXZ-TRZ was immobilized in Zonex E48R and poly-THF matrices.
Fluorescence Lifetime Measurements Fluorescence lifetime of tri-PXZ-TRZ in toluene was measured with an LP920 laser flash photolysis spectrometer (Edinburgh Instruments UK) in a fluorescence mode by using the excitation wavelength of 355 nm. Samples were prepared by bubbling with nitrogen for 15 min for each measurement and adjusted to an absorbance of approximately 0.3 OD in 10 mm path length quartz cuvettes at the laser wavelength used. The fluorescence lifetimes in different solvents were measured at their maximum emission wavelength shown in steady-stated fluorescence spectra. The fluorescence lifetime in THF was also measured by a time-correlated single-photon counting (TCSPC) spectrometer (F900, Edinburgh Instrument). The sample was excited at l = 380 nm. The instrument response function (IRF) of the detection system was about 600 ps.
Femtosecond Transient Absorption Spectroscopy Measurements Femtosecond transient absorption spectroscopy measurements with approximately 90 fs time resolution were performed with a home-built femtosecond broadband pump-probe setup. Details of the instrument are described elsewhere.[53] In brief, 1 mJ, 40 fs pulse at 800 nm with a repetition rate of 500 Hz was obtained from a regeneratively amplified Ti:sapphire laser (Coherent Legend Elite USA) with a bandwidth (full width at half maximum) of about 30 nm. The output of the laser beam was split to generate pump and probe beam pulses with a beam splitter (90 and 10 %). The 400 nm pump pulse was produced by doubling a portion of the
2363
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles 800 nm pulse with a 0.5 mm thick BBO (type I) crystal. The pump power was about 100 nJ pulse¢1 (spot size of 120 mm in each case) during transient absorption spectroscopy measurements. The probe beam was sent to a computer-controlled optical delay system and then focused onto a 2 mm thick water cell to generate a white light continuum that was split into two beams by using a broadband 50/50 beamsplitter as the signal and reference beams. The focused signal and pump beams were overlapped into a flow sample cuvette with 1 mm beam path length, and the reference beam was passed through the unexcited part of the sample. Signal and reference beams were detected with a fiber-coupled dual-channel spectrometer (Avantes AvaSpec-2048-2-USB2). For each pump pulse, spectral intensities of the probe and the referoff on ence in the absence of excitation in the sample (Iprobe and Iref , reon on spectively) and in the presence of the pump (Iprobe and Iref , respectively) were measured. Then, the change in the optical density (DOD) of the transient absorption spectra (for a given time delay) was calculated from Equation (4): DODðt; lÞ ¼ log
off on Iprobe ðt; lÞ Iref ðt; lÞ off on Iprobe ðt; lÞ Iref ðt; lÞ
! ð4Þ
Every spectrum was recorded 200 times, and the averaged spectrum was used in further data analysis. The concentration of the sample was adjusted to an absorbance of 0.3 OD at l = 400 nm in a 1 mm path length quartz cuvette. No photodegradation was observed after femtosecond transient absorption spectroscopy measurements. In femtosecond transient absorption spectroscopy measurements, the population dynamics modeling toolbox software, developed by van Wilderen et al.,[54] was used to analyze the differential absorbance DA(t,l) as a function of wavelength and time delay. Spectral chirp in the transient absorption spectra were corrected for group velocity dispersion of the probe beam. Target model analysis was utilized to globally analyze the transient absorption spectroscopy data, which is a superposition of different spectral components, el(l), weighted by their concentration, cl(t) [Eq. (5)]:[55] DAðt; lÞ ¼
n X
cl ðt Þel ðlÞ
ð5Þ
l¼1
Nanosecond Transient Absorption Spectroscopy Measurements Nanosecond transient absorption spectra were obtained by using an LP920 laser flash photolysis spectrometer (Edinburgh Instruments UK). In this setup, samples were excited by using the 355 nm output from a pulse Nd:YAG laser (10 Hz, 8 ns) (continuum Surelite) as the excitation source. Samples were specifically prepared (bubbled with nitrogen or oxygen for 15 min unless otherwise stated) for each measurement and adjusted to an absorbance of around 0.3 OD in 10 mm path length quartz cuvettes at the laser wavelength used.
Acknowledgements This work was supported by the 973 Program (2013CB834604), National Natural Science Foundation of China (NSFC) (21173235, ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
91233107, 21127003, 21333012 , and 21373232), and the Chinese Academy of Sciences (XDB12020200). Keywords: donor–p–acceptor · charge transfer · reaction mechanisms · solvation · solvatochromism [1] S. M. Bachilo, A. F. Benedetto, R. B. Weisman, J. R. Nossal, W. E. Billups, J. Phys. Chem. A 2000, 104, 11265 – 11269. [2] H. Uoyama, K. Goushi, K. Shizu, H. Nomura, C. Adachi, Nature 2012, 492, 234 – 238. [3] X. Xiong, F. Song, J. Wang, Y. Zhang, Y. Xue, L. Sun, N. Jiang, P. Gao, L. Tian, X. Peng, J. Am. Chem. Soc. 2014, 136, 9590 – 9597. [4] T. Nakagawa, S.-Y. Ku, K.-T. Wong, C. Adachi, Chem. Commun. 2012, 48, 9580 – 9582. [5] J. Chan, S. C. Dodani, C. J. Chang, Nat. Chem. 2012, 4, 973 – 984. [6] C. Baleiz¼o, M. N. Berberan-Santos, ChemPhysChem 2010, 11, 3133 – 3140. [7] H. Nakanotani, T. Higuchi, T. Furukawa, K. Masui, K. Morimoto, M. Numata, H. Tanaka, Y. Sagara, T. Yasuda, C. Adachi, Nat. Commun. 2014, 5, 1 – 7. [8] M. A. Baldo, D. F. O’Brien, M. E. Thompson, S. R. Forrest, Phys. Rev. B 1999, 60, 14422 – 14428. [9] M. Segal, M. A. Baldo, R. J. Holmes, S. R. Forrest, Z. G. Soos, Phys. Rev. B 2003, 68, 075211. [10] B. H. Wallikewitz, D. Kabra, S. G¦linas, R. H. Friend, Phys. Rev. B 2012, 85, 045209. [11] C. Baleiz¼o, M. N. Berberan-Santos, ChemPhysChem 2011, 12, 1247 – 1250. [12] S. Boudin, J. Chim. Phys. Phys. 1930, 27, 285 – 290. [13] C. A. Parker, C. G. Hatchard, Trans. Faraday Soc. 1961, 57, 1894 – 1904. [14] H. Tanaka, K. Shizu, H. Miyazaki, C. Adachi, Chem. Commun. 2012, 48, 11392 – 11394. [15] Q. Zhang, J. Li, K. Shizu, S. Huang, S. Hirata, H. Miyazaki, C. Adachi, J. Am. Chem. Soc. 2012, 134, 14706 – 14709. [16] J. Lee, K. Shizu, H. Tanaka, H. Nomura, T. Yasuda, C. Adachi, J. Mater. Chem. C 2013, 1, 4599 – 4604. [17] K. Nasu, T. Nakagawa, H. Nomura, C. J. Lin, C. H. Cheng, M. R. Tseng, T. Yasuda, C. Adachi, Chem. Commun. 2013, 49, 10385 – 10387. [18] H. Tanaka, K. Shizu, H. Nakanotani, C. Adachi, Chem. Mater. 2013, 25, 3766 – 3771. [19] R. Ishimatsu, S. Matsunami, T. Kasahara, J. Mizuno, T. Edura, C. Adachi, K. Nakano, T. Imato, Angew. Chem. Int. Ed. 2014, 53, 6993 – 6996; Angew. Chem. 2014, 126, 7113 – 7116. [20] T. Komino, H. Tanaka, C. Adachi, Chem. Mater. 2014, 26, 3665 – 3671. [21] H. Tanaka, K. Shizu, H. Nakanotani, C. Adachi, J. Phys. Chem. C 2014, 118, 15985 – 15994. [22] Y. Gong, X. Guo, S. Wang, H. Su, A. Xia, Q. He, F. Bai, J. Phys. Chem. A 2007, 111, 5806 – 5812. [23] G. Ramakrishna, T. Goodson III, J. Phys. Chem. A 2007, 111, 993 – 1000. [24] G. Ramakrishna, A. Bhaskar, T. Goodson III, J. Phys. Chem. B 2006, 110, 20872 – 20878. [25] M. Jia, X. Ma, L. Yan, H. Wang, Q. Guo, X. Wang, Y. Wang, X. Zhan, A. Xia, J. Phys. Chem. A 2010, 114, 7345 – 7352. [26] L. Yan, X. Chen, Q. He, Y. Wang, X. Wang, Q. Guo, F. Bai, A. Xia, D. Aumiler, S. Vdovic´, S. Lin, J. Phys. Chem. A 2012, 116, 8693 – 8705. [27] M. Glasbeek, H. Zhang, Chem. Rev. 2004, 104, 1929 – 1954. [28] C. Reichardt, Chem. Rev. 1994, 94, 2319 – 2358. [29] K. Santhosh, A. Samanta, ChemPhysChem 2012, 13, 1956 – 1961. [30] S.-J. Chung, K.-S. Kim, T.-C. Lin, G. S. He, J. Swiatkiewicz, P. N. Prasad, J. Phys. Chem. B 1999, 103, 10741 – 10745. [31] D. J. Stewart, M. J. Dalton, R. N. Swiger, J. L. Fore, M. A. Walker, T. M. Cooper, J. E. Haley, L.-S. Tan, J. Phys. Chem. A 2014, 118, 5228 – 5237. [32] F. Wìrthner, G. Archetti, R. Schmidt, H.-G. Kuball, Angew. Chem. Int. Ed. 2008, 47, 4529 – 4532; Angew. Chem. 2008, 120, 4605 – 4608. [33] X. Guo, S. Wang, A. Xia, H. Su, J. Phys. Chem. A 2007, 111, 5800 – 5805. [34] Gaussian 09 (Revision A.2), 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,
2364
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles
[35] [36]
[37] [38] [39] [40] [41] [42] [43] [44]
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. 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. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford, CT, 2009. M. Sun, P. Kjellberg, F. Ma, T. Pullerits, Chem. Phys. Lett. 2005, 401, 558 – 564. Y. Wan, L. Yan, Z. Zhao, X. Ma, Q. Guo, M. Jia, P. Lu, G. Ramos-Ortiz, J. L. Maldonado, M. Rodrguez, A. Xia, J. Phys. Chem. B 2010, 114, 11737 – 11745. N.-K. Persson, M. Sun, P. Kjellberg, T. Pullerits, O. Ingans, J. Chem. Phys. 2005, 123, 204718. E. Lippert, Z. Naturforsch. A 1955, 10, 541 – 545. N. Mataga, Y. Kaifu, M. Koizumi, Bull. Chem. Soc. Jpn. 1956, 29, 465 – 470. N. Mataga, Bull. Chem. Soc. Jpn. 1963, 36, 654 – 662. B. Li, R. Tong, R. Zhu, F. Meng, H. Tian, S. Qian, J. Phys. Chem. B 2005, 109, 10705 – 10710. R. Ishimatsu, S. Matsunami, K. Shizu, C. Adachi, K. Nakano, T. Imato, J. Phys. Chem. A 2013, 117, 5607 – 5612. A. Kapturkiewicz, J. Herbich, J. Karpiuk, J. Nowacki, J. Phys. Chem. A 1997, 101, 2332 – 2344. V. A. Galievsky, S. I. Druzhinin, A. Demeter, P. Mayer, S. A. Kovalenko, T. A. Senyushkina, K. A. Zachariasse, J. Phys. Chem. A 2010, 114, 12622 – 12638.
ChemPhysChem 2015, 16, 2357 – 2365
www.chemphyschem.org
[45] H. J. Verhey, B. Gebben, J. W. Hofstraat, J. W. Verhoeven, J. Polym. Sci. Part A 1995, 33, 399 – 405. [46] W. Verbouwe, M. Van der Auweraer, F. C. De Schryver, J. J. Piet, J. M. Warman, J. Am. Chem. Soc. 1998, 120, 1319 – 1324. [47] D. Panda, A. Datta, J. Chem. Phys. 2006, 125, 054513. [48] J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Springer Science, New York, 2006. [49] E. V. Sitzmann, J. G. Langan, D. Griller, K. B. Eisenthal, Chem. Phys. Lett. 1989, 161, 353 – 360. [50] J. Wang, J. Kubicki, H. Peng, M. S. Platz, J. Am. Chem. Soc. 2008, 130, 6604 – 6609. [51] L. Reynolds, J. A. Gardecki, S. J. V. Frankland, M. L. Horng, M. Maroncelli, J. Phys. Chem. 1996, 100, 10337 – 10354. [52] J. E. Rogers, T. M. Cooper, P. A. Fleitz, D. J. Glass, D. G. McLean, J. Phys. Chem. A 2002, 106, 10108 – 10115. [53] M. Zhou, S. Vdovic, S. R. Long, M. Z. Zhu, L. Y. Yan, Y. Y. Wang, Y. L. Niu, X. F. Wang, Q. J. Guo, R. C. Jin, A. D. Xia, J. Phys. Chem. A 2013, 117, 10294 – 10303. [54] L. J. G. W. van Wilderen, C. N. Lincoln, J. J. van Thor, PLoS One 2011, 6, e17373. [55] I. H. M. van Stokkum, D. S. Larsen, R. van Grondelle, Biochim. Biophys. Acta Bioenerg. 2004, 1657, 82 – 104.
Received: April 3, 2015 Revised: April 25, 2015 Published online on May 28, 2015
2365
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
