Introduction

Since most organic molecules can only emit a single fluorescence peak, their applications in many fields such as chemistry, physics, and biology are very limited [1,2,3,4]. However, organic chromophores with excited-state intramolecular proton transfer (ESIPT) properties often emit double fluorescence peaks in experiments, with larger stokes shifts and higher detection sensitivity and fluorescence quantum yields, which makes them have great development potential in the fields of fluorescent materials, luminescent materials, and molecular logic gates [5,6,7,8]. It is well accepted that molecules with ESIPT properties are susceptible to the solvent environment [9,10,11,12]. Hence, the energy barrier of proton transfer can be effectively changed by adjusting the polarity of the solvent to achieve the ultimate goal of regularly regulating the fluorescence characteristics and providing effective help for developing new organic light-emitting materials and fluorescent probes [13,14,15,16].

In 2018, Zhang and colleagues experimentally synthesized the HBT-H, which is a derivative of 2-(2-hydroxyphenyl)benzothiazole (HBT) [17]. By measuring a large number of absorption and fluorescence spectra in many solvents with different polarities, solvatochromic effect was investigated in detail. Regrettably, experimental spectroscopic techniques often fail to give direct information on photophysical, photochemical, and geometric structures, so they do not discuss in detail why HBT-H behaves differently. Later, Yang et al. investigated for the first time the solvent-induced discoloration effect of HBT-T and the reaction mechanism of proton transfer within the excited-state molecule based on density functional theory (DFT) and time-dependent density functional theory (TD-DFT) and gave a proper quantum-level explanation [18]. Moreover, they reproduced the experimentally observed photophysical phenomena by simulating the absorption and emission spectra of molecules. The qualitative effect of solvent polarity on molecular charge transfer excitation was also found by observing the distribution of electron clouds on frontier molecular orbital (FMO), and the solvent-degenerate effect of molecules was revealed in theoretical detail.

Compared with ordinary organic molecules with single fluorescence emission, molecules with dual fluorescence emission and significant stokes shift characteristics tend to have more research value and development potential [19,20,21]. Numerous previous studies have confirmed that the photophysical properties of ESIPT systems are sensitive to the polarity of surrounding solvents. Therefore, it is a good option to adjust the luminescence properties by changing the solvent polarity without changing the molecular structure for ESIPT chromophores [22,23,24]. In this study, we not only sought to find the most suitable computational level for the molecule under study by combining theory with experiments but also established the relationship between solvent polarity, intramolecular hydrogen bond (IHB) strength, and ESIPT energy barriers.

Up to now, there is no report on the influence of ESIPT properties of the 1-acetylamino-4-hydroxyanthraquinone molecule (AcHAQ). Based on this, we take AcHAQ as the research object and systematically study the influence of solvent polarity on its ESIPT reaction process based on DFT and TD-DFT. Three aprotic solvents, strong polar solvent ACN, medium polar solvent CF, and weak polar solvent CYH, are taken as variables. Through the analysis of bond length, bond angle, absorption fluorescence property, FMO, infrared (IR) vibration analysis, natural bond orbital (NBO) distribution, and potential energy curve (PEC), the ESIPT property of AcHAQ is analyzed, hoping to provide theoretical support for the solvent photochromic effect of organic molecules.

The relevant atoms involved in the intramolecular proton transfer reaction in the excited state were numbered and shown in Fig. 1 to facilitate our subsequent discussion and analysis. The molecule contains two intramolecular hydrogen bonds O1-H2 bond and O4-H5 bond, that is, there are two positions where the ESIPT process may occur. Therefore, we consider that there are three types of ESIPT reactions: type I: H2 is transferred from O3 to O1, but H5 does not transfer; type II: H5 is transferred from N6 to O4, and H2 does not transfer; type III: H2 and H5 occur simultaneously transfer. On the other hand, the strength of the intramolecular hydrogen bonds of AcHAQ changes under different solvent polarities, which will affect the ESIPT reaction. Therefore, this paper mainly focuses on two main points to study the ESIPT characteristics of AcHAQ: First, what kind of situation does the ESIPT reaction of AcHAQ belong to? Second, what effect does the solvent polarity have on the ESIPT reaction of AcHAQ?

Fig. 1
figure 1

The molecular configuration of AcHAQ. The significant atoms were numbered

Methods

The ground-state and excited-state structures of AcHAQ in acetonitrile (ACN), chloroform (CF), and cyclohexane (CYH) were separately optimized under the DFT and TD-DFT methods. Becke’s three-parameter hybrid exchange functional with Lee–Yang–Parr gradient-corrected correlation (B3LYP functional) and the 6–311 + + G(d) were utilized in both the DFT and TD-DFT methods [25,26,27]. It should be mentioned that there was no constrain to all the geometric parameters during optimization processes, and the vibrational analysis was performed to confirm the absence of an imaginary mode in all the local minima. To better simulate surrounding environments, all of the calculations in the present work were based on the polarizable continuum model (PCM) using the integral equation formalism variant (IEFPCM). In order to choose the most suitable functional, the maximum absorption peaks of the derivative of AcHAQ, namely 1-amino-4-hydroxyanthraquinone, were calculated using seven different functionals and corresponding results are listed in Table 1. It can be found that the functional B3LYP is most consistent with experiment data [28,29,30,31,32]. Hence, it was chosen to calculate the spectral properties of AcHAQ. Based on the optimized structures in ground and excited states, principal bond lengths, bond angles, IR vibration spectra, NBO population, absorption, and fluorescence properties of AcHAQ in three solvents were obtained [33,34,35]. Moreover, to more intuitively observe the changes of the two hydrogen bond lengths and the energy change trend of the AcHAQ during the ESIPT process in three solvents, PECs in the ground and excited states of AcHAQ were scanned by keeping a series of set values for the O3-H2 bond and the N6-H5 bond [36, 37]. Besides, all the calculations in this work were performed based on the Gaussian 16 software package [38].

Table 1 Maximum absorption peaks (nm) computed by seven different functional and experiment data of 1-amino-4-hydroxyanthraquinone

Results and discussion

Geometric configurations

We optimized the geometry of AcHAQ in three solvents with considerable polar differences, ACN, CF, and CYH, using the TD-DFT/B3LYP method, and compared the main bond parameters related to the ESIPT reaction. The bond lengths of H2-O1 and H5-O4 and bond angles of O1-H2-O3 and O4-H5-N6 are listed in Table 2. In the three solvents, the variation trends of the bond parameters of AcHAQ are the same. From the ground to the excited state, the bond lengths of O1-H2 and O4-H5 are shortened, but the bond length O1-H2 changed more greatly; the bond angles of O1-H2-O3 and O4-H5-N6 enlarged, but the bond angle O1-H2-O3 altered more. Taking the AcHAQ in ACN as an example, a decrease of 0.069 Å occurred for the bond length O4-H5 upon the photoexcitation, while the bond angle O1-H2-O3 enlarged from 147.5 to 152.7°. The bond length O1-H2 diminished from 1.634 Å in the ground state to 1.527 Å in the excited state, a decrease of 0.097 Å, and the bond angle O4-H5-N6 increased from 139.1 to 143.8°. Generally speaking, the shorter the hydrogen bond length, the stronger the hydrogen bond intensity, which is more favorable for the occurrence of ESIPT reaction. Both IHB strengths of O1H2 and O4H5 increased, but that of O1H2 strengthened more, and its bond length was shorter than that of O4H5, implying that the ESIPT process prefers to occur along the O1H2 IHB.

Table 2 Obtained significant bond lengths (Å) and bond angles (°) of AcHAQ in three solvents

Although in the three solvents the change trends of bond parameters of AcHAQ are the same, there are differences in value, so there are different effects on the ESIPT process of AcHAQ. As the polarity of the solvent increases, that is, the solvent changes from CYH to CF and then to ACN, the bond length O1H2 in the excited state gradually shortens, from 1.553 Å in CYH to 1.539 Å in CF, and then reduced to 1.527 Å in CAN, which implies that the intensity of O1-H2O3 IHB has gradually improved as the solvent polarity heightened.

Absorption and fluorescence spectra

Using the TD-DFT/B3LYP method, with 6–311 + + G(d) as the basis set, we simulated the absorption and fluorescence spectra of AcHAQ in the three solvents, ACN, CF, and CYH, as shown in Fig. 2. The transition properties of excited states of AcHAQ are listed in Table 3. Among the three solvents, the maximum absorption peaks of AcHAQ were located at 483, 487, and 489 nm, respectively. Apparently, as the solvent polarity decreases, the maximum absorption peak is red-shifted.

Fig. 2
figure 2

Absoprtion and fluorescence spectra of AcHAQ in three solvents. Herein, K means the proton transfer tautomer

Table 3 Calculated transition properties of AcHAQ in three different solvents via TD-DFT/B3LYP/6–311 + + G(d) method

In addition, the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) were also calculated and exhibited in Fig. 3. It can be found that the S1 states of AcHAQ show π → π* features, in which the HOMO and LUMO exhibit the π and π* properties, respectively. From Fig. 3, it can be seen that the HOMO–LUMO gap of AcHAQ in ACN is 2.951 eV, which is higher than 2.936 eV in CF and 2.919 eV in CYH. Table 4 lists the fluorescence properties of AcHAQ in three different solvents. The fluorescence peaks of AcHAQ in three different solvents of ACN, CF, and CYH are 565, 568, and 571 nm, respectively, and the corresponding stokes shifts are 81, 81, and 82 nm, respectively. The above calculation results show that the change of solvent polarity has a significant effect on the absorption and fluorescence properties of AcHAQ.

Fig. 3
figure 3

Frontier molecular orbitals (FMOs) and corresponding energy gap values (eV) of AcHAQ in three solvents

Table 4 Fluorescence properties of AcHAQ in three solvents. Herein, K means the proton transfer tautomer

IR vibrational spectra

It is well accepted that the IHB strengthening or weakening could be also revealed via evaluating the IR vibrational spectra shifts of characteristic vibrational modes involved in the IHBs [39, 40]. We have also calculated the IR vibrational spectra of AcHAQ and exhibited the stretching vibrational modes of O3-H2 and N6-H5 as shown in Table 5. It can be clearly seen that, after the photoexcitation, two vibrational peaks of AcHAQ have apparent red shifts, which indicates that the two hydrogen bonds have been strengthened. We have also noticed that the red-shift values for O3-H2 hydrogen are above 400 cm−1, while that of the N6-H5 hydrogen bond do not exceed 200 cm−1. The enhancement of the O1H2-O3 hydrogen bond is more robust than the O4H5-N6 hydrogen bond. The magnitude of the bond strengthening is much greater, and the intramolecular proton transfer is more likely to occur.

Table 5 Calculated stretching vibrational models of O3–H2 and N6–H5 in three different solvents at S0 and S1 states (cm−1)

Among the three solvents, the red-shift value of the O3-H2 hydrogen bond is 495 cm−1 in ACN, 448 cm−1 in CF, and 400 cm−1 in CYH. It can be found that when the AcHAQ transitions from the S0 to the S1 state, the red-shift value of the IR vibrational peak for the O1H2-O3 hydrogen bond increases with the increase of the solvent polarity, indicating that IHBs are more stronger in ACN, which is beneficial to the occurrence of the ESIPT process and is consistent with the results obtained from geometric parameters.

NBO analysis

Investigating the charge distribution during the optical excitation process has always been an effective way to analyze the molecular dynamic [20, 41]. The charge distribution on O1, O3, O4, and N6 atoms of AcHAQ in the S0 and S1 states was calculated based on NBO populations and listed in Table 6. For all the studied solvents, the charge localized on the O3 and N6 atoms was decreased, while that on the O1 and O4 was increased. For instance, the charge distributed on the O3 and N6 of AcHAQ in ACN has lessened from − 0.677 a.u. and − 0.612 a.u. in the S0 state to − 0.626 a.u. and − 0.547 a.u. in the S1 state, indicating that the capacity of the O3 and N6 atoms to bind the proton has been weakened upon the photoexcitation. Besides, the charge distribution on the O1 and O4 atoms in ACN has enlarged from − 0.634 a.u. and − 0.619 a.u. in the S0 state to − 0.667 a.u. and − 0.659 a.u. in the S1 state, which implies the enhanced ability of the O1 and O4 atom to attract protons, thereby triggering the ESIPT behavior.

Table 6 NBO charge distribution (a.u.) of the O1, O3, O4, and N6 atoms in the three solvents at the S0 and S1 states

PESs

To further explore the ESIPT mechanism in the three solvents, we scanned the PESs of AcHAQ via varying the O3-H2 and N6-H5 distance and allowing all other degrees of freedom to relax freely towards the minimum energy as shown in Fig. 4. As presented in Fig. 4, the Z-direction indicates the energy of the molecule, and the positive direction of the left coordinate is the direction of O3-H2 bond lengthening, which represents the reaction process of proton transfer from O3 to O1, i.e., type I reaction. Moreover, the positive direction of the right coordinate is the direction of N6-H5 bond lengthening, which represents the reaction process of proton transfer from N6 to O4, i.e., type II reaction.

Fig. 4
figure 4

Potential energy surfaces (PESs) for the S0 state of AcHAQ in three solvents along with the increasing bond lengths of O3–H2 and N6–H5

It can be found from Fig. 4 that the energy increases monotonically with the elongation of O3-H2 and N6-H5 bonds, which indicates that proton transfer processes of AcHAQ in the three solvents cannot occur in the ground state. This is well understood. The AcHAQ is in the low-energy S0 state, and the hydrogen bond is relatively stable, making it difficult for proton transfer reactions to occur. But when we looked at the AcHAQ in the S1 state, the situation became different, as shown in Fig. 5. First, in different solvents, the energy barrier exists only in the direction of the type I reaction, indicating that the intramolecular proton transfer reaction of the excited state of the AcHAQ only occurs between the O1 and O3 atoms. Second, we found that the energy barrier for the type I reaction is the lowest in ACN, followed by CF, and highest in CYH, that is, the ESIPT reaction is more likely to occur in strongly polar solvents.

Fig. 5
figure 5

Potential energy surfaces (PESs) for the S1 state of AcHAQ in three solvents along with the increasing bond lengths of O3–H2 and N6–H5

Conclusion

In this work, three different solvents, ACN, CF, and CYH, were selected to change the polarity of the solvent environment where the AcHAQ was located so as to study the influence of the solvent effect on the ESIPT process of the AcHAQ. Using DFT and TD-DFT, the ESIPT mechanism of AcHAQ in solvents of different polarities was deeply studied, and the main parameters closely related to the ESIPT process, such as bond lengths and bond angles, were obtained. Absorption and fluorescence spectra, FMOs, IR vibrational spectra, NBO populations, and PESs were analyzed. Finally, it is concluded that the ESIPT process of AcHAQ is mainly a type I reaction, and the type II and type III reactions hardly occur. Besides, AcHAQ is more likely to undergo ESIPT reaction in ACN, while it is difficult to proceed in CYH, indicating that strong polar solvents have a reinforcing effect on the IHB of the AcHAQ, thereby promoting the ESIPT process. This conclusion provides a theoretical basis for regulating the luminescence properties of organic molecules by changing the polarity of the solvent.