Rational design of anthocyanidins-directed near-infrared two-photon fluorescent probes†

Xiu-e Zhang^a, Xue Wei^b, Wei-Bo Cui^b, Jin-Pu Bai^a, Aynur Matyusup^a, Jing-Fu Guo*^a, Hui Li^b and Ai-Min Ren*^b
aSchool of Physics, Northeast Normal University, Changchun 130024, P.R. China. E-mail: guojf217@nenu.edu.cn
^bLaboratory of Theoretical and Computational Chemistry, Institute of Theoretical Chemistry, College of Chemistry, Jilin University, Liutiao Road #2, Changchun 130061, P.R. China. E-mail: renam@jlu.edu.cn

First published on 4th September 2024

---

---

1. Introduction

Nitroreductase (NTR) in hypoxic tumors is likely to be overexpressed, making it critical to precisely and in real-time monitor NTR levels.^1–3 Small molecule fluorescence imaging technology has been widely adopted in diverse fields in recent years due to its numerous merits, such as superior biocompatibility, rapid and sensitive feedback, non-invasive nature, and easy adjustment of optical properties through structural modification.^4–6 Unlike the conventional one-photon (OP) excitation probes, two-photon (TP) fluorescent probes have received considerable attention because of their excitation wavelengths in the near-infrared (NIR) region.⁷ This characteristic provides various unique advantages, including negligible background fluorescence interference, minimal photodamage to biological samples, substantially decreased photobleaching, and deeper tissue penetration.^8–12 Above all, two-photon excited fluorescence (TPEF) probes have emerged as powerful tools in both fundamental biological research and clinical applications.^13–16

Although a number of molecular fluorophores have been reported in the past few decades, their relatively short wavelength (<600 nm) still constrains their applications in biological fields.^17–20 In this case, our attention is piqued by the fact that anthocyanidin, a natural pigment, can achieve long wavelength emission, great selectivity and sensitivity for bioimaging.^21,22 A unique fluorescent molecular framework called AC-Fluor was designed in 2017 and its biological relevance was proved using two-photon deep tissue imaging, exhibiting deep penetration of up to 300 μm with negligible cytotoxicity.²³ In 2019, their group synthesized several kinds of anthocyanidins based on 1b (molecule ACF4 in the literature), as shown in Fig. 1(a).²⁴ In particular, LDO-NTR was proposed as a practical NTR-activated TPEF probe as a result of molecule 2a, also known as LDOH-4 in the literature, demonstrating favorable chemical and optical properties (λ_em = 633 nm, Φ = 0.55, pK_a = 5.13, and Stokes shift = 59 nm). Currently, there is no comprehensive and in-depth investigation about anthocyanidin derivatives and their unique electronic structures as TPEF fluorescent probes, which has caught our considerable attention and enthusiasm. Beyond that, it appears that the photoinduced electron transfer (PET) and intramolecular charge transfer (ICT) result in the fluorescence quenching of LDO-NTR. However, the underlying microscopic mechanism of fluorescence quenching still remains ambiguous. Besides, for LDO-NTR derivatives with the same substituents but different fluorophore skeletons, 1a–3a (corresponding to LDOH-1, LDOH-4, and Ctrl-2 in ref. 24) and 1b–3b (corresponding to ACF4, LDOH-3, and Ctrl-1 in ref. 24) show excellent properties such as suitable pK_a, high quantum yield, and good photostability. How do the six/five-membered rings in the backbone affect their photophysical properties differently? Regarding molecule 2a, also, is there any room to improve its optical properties, such as a small two-photon absorption (TPA) cross-section (89 GM/Φ = 161.8 GM), short emission wavelength (633 nm) and small Stokes shift (59 nm)? If so, further application opportunities would arise as it would improve the photobleaching limit and fluorescence/background ratio of anthocyanin-based TP fluorescent probes for super-resolution and deep tissue imaging, opening up broader application prospects. All of the issues aforementioned are worthy of serious attention and need to be addressed urgently.

In this work, density functional theory (DFT) and time-dependent density functional theory (TDDFT) are adopted to systematically reveal the relationship between the molecular configurations and properties of a series of anthocyanidin derivatives (molecular probes with a 4-nitrobenzyl alcohol group as the NTR reaction site and the corresponding products with a hydroxyl group), including their OPA/TPA properties, fluorescence emission properties, fluorescence probing mechanism, and solvation free energy. Then, a new series of 2a-n (n = 1–5) TPEF molecules are designed accounting for the biological applications of TP fluorescent probes. Ultimately, the PET and ICT luminescence mechanisms of the anthocyanidin derivatives are investigated by means of quantum chemistry methods. These findings will hopefully provide some meaningful insights into the rational design of novel functional fluorescent materials based on the anthocyanidin skeleton.

2. Calculation methods

The full explanation of theoretical methodologies utilized in this paper are given in the ESI,† including one-photon absorption (OPA), two-photon absorption (TPA), and solvation free energy. In addition, the computational details are summarized as follows.

The geometric structures of the ground states and excited states of the studied molecules were optimized using DFT²⁵ and TDDFT,²⁶ respectively. Vibrational frequency analysis was performed based on the optimized geometric configurations, and no imaginary frequency emerged. All of the preceding above calculations were carried out in the Gaussian 16 program.²⁷ Taking the experimental molecules 1a and 1b as references, their OPA and fluorescent emission spectra were calculated by TDDFT utilizing different functionals and basis sets. Detailed results are listed in Tables S1–S3 (ESI†). As a compromise between accuracy and computational efficiency, B3LYP/6-31G(d,p) and B3LYP/6-311+G(d) methods are chosen and employed in the subsequent geometry optimization and transition energy calculation, respectively. On the basis of quadratic response theory,²⁸ the B3LYP/6-311+G(d) and CAM-B3LYP/6-311+G(d) methods were considered to calculate the TPA properties of the a-series and b-series molecules by virtue of the DALTON²⁹ program. As can be seen in Table S4 (ESI†), for the experimental molecule 2a, the result for B3LYP (185.73 GM at 746.90 nm) is closer to the experimental value (161.8 GM at 810 nm)²⁴ than CAM-B3LYP (581.00 GM at 635.82 nm). Therefore, the TPA properties of all the subsequent molecules were calculated using B3LYP/6-311+G(d). Furthermore, a polarization continuous model (PCM)³⁰ was also used to take into account the solvent effect (aqueous and DMSO) for the experimentally synthesized molecules 1a–3a and 1b–3b. Detailed results are shown in Tables S5 and S6 (ESI†), from which it can be seen that the difference in solvents has little effect on the photophysical parameters and does not change the general trend. Therefore, the aqueous solvent was used to model solvent effects in the subsequent calculations. Previous efforts have confirmed that the M052X/6-31G(d) together with the SMD solvent model is an optimal approach for solvation free energy calculations,³¹ thus the method was applied in this work. The superimposed structures of S₀ and S₁ states were achieved using the PyMOL program.³² In addition, the radiative and nonradiative transition rates were also assessed using MOMAP³³ software which is currently available and their fluorescence quantum yields were obtained in the framework of harmonic oscillator approximation. The calculated results are collected in Tables S7 and S8 (ESI†), and a detailed discussion of them can be found in the ESI.†

3. Results and discussion

3.1. Molecular design

Firstly, a series of anthocyanidins with the same substituents but different fluorophore skeletons, 1a–3a and 1b–3b, are selected in an effort to study how the six/five-membered rings in the backbone affect their photophysical properties. The chemical structures of the studied experimental compounds are presented in Fig. 1(a). The maximum absorption peak (503.80–535.52 nm) of b-series anthocyanidins with a five-membered ring fused in the backbone has a hypsochromic shift with respect to a-series with a six-membered ring fused in the backbone (519.95–551.59 nm). Moreover, the b-series molecules exhibit a stronger fluorescence oscillator intensity (1.1269–1.2102) than that of the a-series (1.0430–1.1251). What effect do the five and six-membered rings in the backbone have on the photophysical properties? In an attempt to address the above issues, molecule 1c was also designed as a reference, and its chemical structure is shown in Fig. S1 (ESI†). Their electronic structures and photophysical properties, such as OPA and TPA spectra, are scrutinized in Sections 3.2–3.5.

Secondly, it is worth highlighting that 2a exhibits superior properties for biological imaging, with a better fluorescence enhancement factor (27.8-fold) for phenol/phenolate states than 2b (23.7-fold), and its excellent properties under simulated physiological conditions (2a: λ_em = 633 nm, Φ = 0.55, pK_a = 5.13; 2b: λ_em = 620 nm, Φ = 0.70, pK_a = 5.60).²⁴ In this context, some design strategies are put forward by us based on the experimentally synthesized molecule 2a in the hope of improving properties such as the emission wavelength and TPA cross-sections. In addition, molecules 4a and 5a (as shown in Fig. S7, ESI†) are intended to function as a comparison for 2a to identify the effects from the various substitution sites (a detailed discussion is provided in the ESI†). Eventually, the five compounds (2a-n (n = 1–5)) are designed by modifying the R₁ substituent group on the basis of molecule 2a with a six-membered ring that is not fully conjugated, and Fig. 1(b) offers a reference to their unique molecular structures. Most notably, dialkylamino groups, which are extensively utilized as electron-donating groups in traditional fluorophores, are regarded as the rotating groups resulting in twisted intramolecular charge transfer (TICT).³⁴ The non-radiative decays of some fluorophores coincide with the formation of TICT states, hence they may be a quencher of much highly effective fluororescence.^35–37 It has been reported that performing alkyl cyclization and azetidine substitution can effectively suppress TICT formation.^36–40 Thereby, 2a-1 and 2a-2 are designed to make every effort to improve the emission properties. To achieve a larger TPA response, 2a-3, 2a-4 and 2a-5 are also designed by extending the fused backbone. Along with the first part of the work, the research on the photophysical properties of the designed molecules is also included in Sections 3.2–3.5.

Thirdly, the fluorescence quenching of LDO-NTR upon light excitation is regarded as being due to the combined effects from both photoinduced electron transfer (PET) and intramolecular charge transfer (ICT). However, the underlying mechanism of the fluorescence quenching phenomenon of the probe still remains ambiguous. We have carried out a detailed theoretical study on the ICT and PET processes of LDO-NTR to rationalize this phenomenon. Furthermore, as illustrated in Fig. 1(c), for the designed 2a-n (n = 1–5) series of molecules, 2a-3 has the most excellent properties among them, its corresponding probe molecule 2a-3-NTR is designed for detecting the NTR reaction, and its fluorescence quenching mechanism is also predicted by using the same theoretical method. This part is described in Section 3.6.

3.2. Geometrical optimization

Geometric configurations are intimately tied to electronic structures and further photophysical properties. Therefore, some selected geometric parameters of the studied molecules at the optimized S₀ and S₁ states are summarized in Table S9 (ESI†), and their superimposed structures are depicted in Fig. S2 (ESI†). Upon conducting a comparative analysis, it can be concluded that the a-series of molecules relative to the b-series with the same substituents but different fluorophore skeletons, (1) possess a longer interatomic distance (C₂–C₃, C₁₀–C₁₃) (the numbering of atoms is shown in Fig. 1), indicating that the a-series molecules might enhance the molecular structure stability by weakening the interaction between the B and D rings.⁴¹ (2) The dihedral angle (DHA1: 6-13-12-11 is 147–150 degrees for the a-series of molecules or 6-12-11-10 are 177–178 degrees for b-series molecules) of a-series molecules is not as close to 180 degrees as that of b-series molecules, which demonstrates that the former exhibit less planarity. (3) They have the lowest root-mean-square deviation (RMSD) values, suggesting that a-series molecules would experience smaller structural distortion upon excitation. Compared to 2a, the differences in geometric parameters at S₀ states are extremely minor for the designed molecules (2a-1-2a-5). However, the RMSD values of 2a-3 and 2a-5 are close to zero, which implies that there is almost no structural distortion during the transition from the S₀ to S₁ states. (4) From the DHA2 in Table S9 (ESI†), the carboxyl–phenyl fragment of a-series molecules twists relatively less during the S₀ → S₁ process than that of b-series molecules. Overall, the six-membered ring in a-series molecules minimizes the geometry rotation upon light excitation with respect to b-series molecules. Nevertheless, as shown in Table S10 (ESI†), the reorganization energy of a-series molecules (1a: 1262.04 cm⁻¹, 2a: 1370.09 cm⁻¹, 3a: 1101.87 cm⁻¹) is greater than that of b-series molecules (1b: 1187.46 cm⁻¹, 2b: 1338.33 cm⁻¹, 3b: 1089.32 cm⁻¹). It is surprising that the six-membered ring in a-series molecules makes the reorganization energy increase during the S₁ → S₀ process. This will be discussed in more detail in the following section, leading to an increase in the Stokes shift.

3.3. Frontier molecular orbitals

To further investigate the properties of the electronic structures, molecular orbitals were calculated using the DFT//B3LYP/6-31G(d,p) level of theory. The 10 frontier molecular orbitals (FMOs) of the studied molecules are depicted in Fig. 2, in which Fig. 2(a) shows that compared to the b-series with the same substituents but different fluorophore skeletons, a-series molecules have a lower LUMO energy level and obviously smaller energy gaps (ΔE_HOMO–LUMO) between the HOMO and the LUMO. The above phenomenon indicates that the introduction of a six-membered ring in the fluorophore skeleton makes a-series molecules have lower LUMO levels than b-series molecules. As shown in Fig. S3 (ESI†), it is observed that the HOMO and LUMO of the a-series and b-series molecules have extremely similar electronic cloud distributions. In order to scrutinize the different roles of the five and six-membered rings added in the backbone, 1a, 1b, and 1c (unsubstituted on the backbone) are taken as examples, and the difference in their LUMOs is clarified as follows:

On the one hand, as can be seen from Fig. S4 (ESI†), the LUMO distribution of 1a and 1b is basically the same, primarily composed of antibondings and nonbondings, with the most intense antibonding located between the O₁–C₂ bonds. Table S9 (ESI†) indicates that 1b has a shorter O₁–C₂ bond length (1.3297 Å) than both 1a (1.3422 Å) and 1c (1.3465 Å). This difference means that 1b has stronger antibonding in the LUMO, which ultimately results in a higher LUMO energy level. On the other hand, the same reason mentioned above also explains the wider HOMO–LUMO gap of the b-series molecules, and the larger HOMO–LUMO gap also further implies that they might be responsible for a hyperchromic OPA spectrum.

Similarly, FMO analysis for the designed molecules 2a-n (n = 1–5) is also performed. As illustrated in Fig. 2(b), it is clear that compared with the experimental molecule 2a (2.65 eV), the ΔE_HOMO–LUMO of 2a-1 (2.54 eV) decreases dramatically, while 2a-2 (2.76 eV) exhibits the opposite behavior. This is due to the fact that 2a-1 shows an increased contribution of the alkyl cyclization substituent in the HOMO relative to the 2a (as shown in Fig. S11, ESI†), and the substituent is not directly involved in the components of the LUMO of 2a-1, thereby raising the HOMO significantly and LUMO slightly. Thereby, 2a-1 has a smaller ΔE_HOMO−LUMO than 2a, which favors the shift of the electronic spectrum towards longer wavelengths. However, 2a-2 exhibits the exact opposite situation. This is due to diethylamino groups with alkyl cyclization, which limits the degree of free rotation of alkane in the substituent and also increases the number of C atoms in the p-electron delocalization of the HOMO, thus favoring the shift in the electronic spectrum towards longer wavelengths. In addition, the ΔE_HOMO−LUMO for three molecules (2a-3 (2.50 eV), 2a-4 (2.65 eV) and 2a-5 (2.59 eV)) can also be minimized by further π-conjugation to some extent with respect to 2a-1, 2a-2 and 2a. Specifically, 2a-3 has the smallest ΔE_HOMO−LUMO, which is advantageous for achieving emission wavelengths within the biological transparent window range.

3.4. OPA and emission spectral properties

Following the optimized S₀ geometries, the OPA and fluorescence emission properties are calculated, and the specific results along with the corresponding experimental values are summarized in Table 1. The simulated absorption and emission spectra of these molecules are drawn in Fig. 3. It is apparent that the maximum absorption and fluorescence emission peaks of all the studied molecules are mainly composed of a HOMO → LUMO configuration. On the one hand, as shown in Fig. 3(a), a-series molecules exhibit a significant red-shift in the maximum wavelength of OPA and fluorescence emission when compared to b-series molecules with the same substituted groups, and the corresponding oscillator strengths are slightly smaller than those of b-series molecules. It is worth noting that all the oscillator strengths of the two series of molecules are larger than 1, regardless of the light absorbing or emitting process. Obviously, the difference in the HOMO and LUMO between a-series and b-series molecules is responsible for this phenomenon. The addition of a five-membered ring, which shortens the O₁–C₂ antibonding length in the LUMO in b-series molecules, leads to its higher LUMO with respect to that of the a-series molecules. This consequently results in a comparatively wider HOMO–LUMO gap in b-series molecules, which accounts for the red-shift in the electronic spectra of a-series molecules in comparison to b-series molecules. As can be seen in Table 1, the introduction of a five-membered ring dramatically raises the LUMO level in b-series molecules, increasing the HOMO → LUMO transition energy and transition dipole moment. As a result, b-series molecules have a stronger oscillator strength than a-series molecules in terms of formula (1) in the ESI.† In addition, as compared to b-series molecules, the Stokes shifts of a-series molecules show an increasing trend. Fluorescent probes with larger Stokes shifts are more suitable for biological applications due to minimized self-quenching and fluorescence detection error arising from excitation backscattering effects.^42,43 As demonstrated in Section 3.2, this phenomenon presumably occurs because the six-membered rings make the a-series molecules geometrically more flexible and pliable than the b-series molecules when they are excited by light.

---

On the other hand, as shown in Fig. 3(b), except for 2a-2, the maximum OPA and fluorescence emission spectra wavelengths of all the designed molecules are red-shifted compared to 2a. For 2a-1 the strong electron-donor capability of the R₁ substituent decreases the HOMO–LUMO gap, leading to a spectroscopic red-shift. For 2a-3-2a-5 obtained by further π-conjugation, the OPA and emission spectra are also red-shifted relative to 2a-1-2a-2 and 2a, respectively. It is worth highlighting that 2a-3 has the longest emission wavelength (691.42 nm) and quite larger Stokes shifts (110.88 nm), which are more advantageous for bioimaging detection.

3.5. TPA spectra properties

As a two-photon biofluorescent probe and the corresponding product after recognizing the NTR reaction, it is necessary to have a large two-photon absorption cross-section, which implies high photobleaching limits during the fluorescence detection process for NTR and thus ensures the lifetime of the probe. To evaluate the two-photon response of the studied molecules, the detailed physical parameters concerning the TPA spectra properties are listed in Table 2. The calculated maximum TPA cross section of 2a is 185.73 GM, which is substantially consistent with the experimentally measured result (TPA cross section = 89 GM/Φ = 161.8 GM).²⁴ All of the studied compounds exhibit significant TPA peaks (110–785 GM) from 650 nm to 1100 nm, making them appropriate for bioimaging applications. The TPA wavelengths of a-series molecules are noticeably red-shifted (ca. 29–50 nm) in comparison to those of b-series molecules with the same substituted groups, yet the maximum TPA cross-sections slightly decline (ca. 50–140 GM). As concerns the designed molecules 2a-1-2a-5, compared with 2a, the TPA cross section values of the designed molecules 2a-3, 2a-4 and 2a-5 are greatly improved by ca. 4–5 times. In particular, 2a-3, whose TPA cross-section reaches 957.36 GM, is a promising candidate for high-performance two-photon fluorescent probes. However, how does extending the π conjugation or introducing substituents enhance the TPA cross section values for these studied molecules? Herein, a generalized three states model (3SM) is adopted to analyze the molecular two-photon response, which can be described as follows:^44–47

δ3SMTP = δiiTP + δffTP + 2δifTP	(3.1)

	(3.2)

	(3.3)

	(3.4)

where,

Xii = (1 + 2cos2θif0i)	(3.5)

Xff = (1 + 2cos2θff0f)	(3.6)

Xif = cosθif0icosθff0f + cosθ0f0icosθffif + cosθff0icosθif0f	(3.7)

in which, the ground state, the intermediate excited state, and the final state are noted by the indexes 0, i, and f, respectively. The transition dipole moments between the two states are represented by the μ terms. The difference in the excitation energies can be calculated using the following formula; where ω_i means the excitation energy from the ground state to the i excited states. The angle terms (θ^if_0i) are obtained between the μ⁰ⁱ and μ^if vectors. The δⁱⁱ_TP terms always have positive values, however, due to the dipole moments being relatively different in direction, the δ^if_TP terms exhibit two entirely different answers for positive and negative values, corresponding to the constructive and destructive interference effects, respectively. In other words, an increased or decreased TPA cross section could result from the relative distinct orientation between the various optical channels.

---

As shown in Fig. S12 (ESI†), the calculated δ^3.SM values are largely in agreement with the results predicted by the response theory, indicating that the 3SM results can be applied for quantitative analysis of the relationship between molecular structures and TPA spectra properties. To begin with, the possible TPA transition channels are analyzed. Table 2 illustrates that 1a, 2a, 2b, and 2a-2 have only one kind of transition channel (S₀ → S₁ → S₂), since their maximum two-photon response occurs in the S₂ states. However, two possible transition channels (S₀ → S₁ → S₃ and S₀ → S₂ → S₃) may be involved in the TPA excitation processes for 1b, 3a, 3b, 2a-4, and 2a-5, whose final TPA states are S₃. As far as 2a-1 and 2a-3 are concerned, their higher TPA final states (S₄) lead to more complicated transition pathways. Subsequently, the relevant transition dipole moments of the compounds with multiple possible TPA transition channels are evaluated, and the results are given in Table S12 (ESI†). For 1b, 3a, and 3b, we find that μ⁰² and μ²³ are far smaller than μ⁰¹ and μ¹³, and thus the TPA cross sections, derived from a positive correlation with the product of μ⁰² and μ²³ in the process of S₀ → S₂ → S₃, are also almost negligible. Consequently, it can be deduced that their most probable TPA transition channels are S₀ → S₁ → S₃. However, for all the designed molecules, the other possible transition channels also make a substantial contribution to the total TPA cross sections. Based on the above consideration, several important physical parameters of the designed molecules corresponding to all possible TPA transition channels are calculated. Subsequently, the excitation energies and transition dipole moments are provided in Table 3.

---

The detailed results illustrate that there is a relatively minor difference in the excitation energies between a-series and b-series molecules. To aid in visualizing the extent to which various dipole moment vectors and angle terms contribute to the TPA cross-section, Fig. S13 (ESI†) is plotted. The findings reveal that for both a-series and b-series of molecules, the transition dipole moments from the S₀ to intermediate state (μ⁰ⁱ) and the angle term of the initial state (Xⁱⁱ) are much larger than the other dipole moments and angle terms. The fact that δⁱⁱ is positively correlated with the square of both μ⁰ⁱ and Xⁱⁱ alongside explosive growth, as demonstrated in formula (3.2), also helps to explain why δⁱⁱ is dominant in δ^3SM. As shown in Fig. S13(a) (ESI†), the b-series molecules display significant increases in their transition dipole moments (μ⁰ⁱ and μ^if), and state dipole moment (μ^ff) when compared to the a-series molecules. This leads to a larger two-photon absorption, and thus rationalizing that the b-series molecules have larger TPA cross sections. Indeed, μ^if plays a more crucial role in this process for the b-series molecules. In other words, the transition dipole moment from the intermediate state to the final state makes the compounds bridged by a five-membered ring, the b-series molecules, more advantageous to increase the TPA cross section. This cause is in line with the results of the FMOs analysis. Immediately following this, specific analysis of the three pairs of experimental molecules is performed. For both 1a and 1b, X^if are greater than 0, while the values of 1b exceed 1a. Although Xⁱⁱ and X^ff are marginally smaller for 1b, the prominent dipole moments still leave 1b with a larger TPA cross section. The cases of 3a and 3b are similar to the above discussion. As for 2a and 2b, they both have negative X^if values, with the latter being smaller. However, the TPA cross section of 2b prevails over that of 2a due to the larger dipole moments and angle terms (Xⁱⁱ and X^ff).

Based on 2a, 2a-1 and 2a-2 are obtained by internal cycloalkylation and external cyclization of the substituted amine groups, respectively. 2a-3-2a-5 are designed by extending the π-conjugation, gaining better rigidity and considerably larger dipole moment vectors (μ^0f, μ^if and μ^ff) and decreased excitation energy (Table 3), which accounts for their higher TPA cross sections. Among a series of the designed molecules, 2a-3 has the most significant TPA cross section because of its largest transition dipole moment (μ⁰ⁱ) and smallest excitation energies (ΔE_iΔE_f).

3.6. Theoretical validation of the fluorescence mechanism

The NTR-activated TP fluorescent probe LDO-NTR was experimentally observed to exhibit fluorescence enhancement with a 310-fold emission turn-on response upon the addition of NTR.²⁴ Furthermore, it was hypothesized that the negligible fluorescence of LDO-NTR originated from the combined effects of photoinduced electron transfer (PET) and internal charge transfer (ICT). However, to the best of our knowledge, details about the fluorescence quenching mechanism still remain unclear. Obviously, it is of paramount importance to making clear the sensing mechanism used to recognize targets for more purposeful synthesis of probes in the future. Thus, in the subsequent discussion, on the one hand, the fluorescence quenching mechanism of LDO-NTR is theoretically analyzed and confirmed from the following perspectives, including the frontier molecular orbitals (FMOs) theory, the natural population analysis (NPA), and the calculations of ICT and PET rates. On the other hand, for the 2a-3 molecule with the superior properties among all the designed molecules, the corresponding probe 2a-3-NTR is designed and its fluorescence quenching mechanism is also predicted using the same method as described above.

With the purpose of more accurately characterizing the properties of the probes and products, the experimental molecules (the probe LDO-NTR and product 2a) are used as examples, and their one-photon absorption (OPA) and fluorescence emission properties are calculated at various functionals (TPSSH, B3LYP, PBE0, M06, BMK, M062X, CAM-B3LYP, wB97XD) and basis sets, and the detailed results are compiled in Tables S13 and S14 (ESI†). B3LYP/6-31G(d,p) is ultimately chosen as the optimal method for the subsequent calculations (although the emission peak is not quite accurate for the LDO-NTR probe due to the self-interaction problem of the B3LYP functional, the peak position of the probe relative to its reaction product is not quite important for an off–on probe). The OPA and emission properties of the probes and products are illustrated in Table S15 (ESI†), indicating that the transition from the S₀ to S₄ state for the LDO-NTR probe corresponds to the maximum absorption peak in the OPA process, which is composed of HOMO → LUMO+2. Meanwhile, the electron transitions from the S₀ to the first three excited states (S₁–S₃) of LDO-NTR with minimal oscillator strengths. From the frontier molecular orbitals depicted in Fig. S14 (ESI†), it can be noticed that the recognition group (4-nitrobenzyl alcohol) is not involved in the excitation process from the S₀ to the excited states, thus LDO-NTR is partially excited. In contrast, for the product 2a, the transition from S₀ to S₁ with the largest oscillator strength mainly involves the HOMO and LUMO, which are both delocalized throughout the molecule. The emission process almost experiences a similar situation, where the oscillator strength of the S₁ state is zero, i.e. the transition probability is zero, and the radiative transition in the S₁ → S₀ process is completely inhibited. The excited state S₁ then returns to the ground state in a nonradiative way. Besides, the designed 2a-3-NTR molecule probe has a similar case to LDO-NTR.

Now the question is whether the nonradiative decays occur through a PET process or ICT process or both? And why does it happen in this particular system? Initially, the Rehm–Weller formula, which reads ΔG_ET = E⁰(D⁺/D) − E⁰(A/A⁻) − ΔE_0,0 + w, must be satisfied for PET to occur. In the following discussion, the detailed calculations for physical parameters such as E⁰(D⁺/D), E⁰(A/A⁻), and ΔE_0,0, etc. are given. The results showed that the probes have a rather strong driving force (−ΔG_ET) in the PET process (LDO-NTR: −0.3806 eV, 2a-3-NTR: −0.3139 eV), while the corresponding reaction products fail to meet this point. Sequentially, in order to address the question of why it happens in this system, the activated fluorophore components (anthocyanin) and recognition groups (4-nitrobenzyl alcohol) of the probes are called the Flu* and receptor, respectively. Their frontier molecular orbital energy levels are calculated on the basis of the whole molecule,^48–50 with the results displayed in Fig. 4. For the LDO-NTR probe, the HOMO energy of the receptor is −2.62 eV, which is higher than the Flu* (−5.05 eV). Therefore, upon excitation of LDO-NTR, i.e., after the Flu* is excited, an electron can be transferred from the HOMO of the receptor to the HOMO of the Flu* through the Rehm–Weller formula. Furthermore, an electron lying in the LUMO of Flu* cannot instantaneously go back to the HOMO of Flu* itself, resulting in the quenching of the fluorescence, that is, photoinduced electron transfer (PET) takes place. In contrast, for the product 2a, the hydroxyl group in the product 2a has a small twist angle relative to the entire molecule, so the electron cloud distributions of the HOMO and LUMO cannot be completely separated and therefore are delocalized throughout the whole molecule, which inhibits the PET process, and a dramatic fluorescence phenomenon occurs in 2a. As can be seen in Fig. 4, the designed probe 2a-3-NTR and product 2a-3 exhibit similar behavior to the experimental molecules, suggesting that 2a-3-NTR is also involved in the PET process.

In addition, the possibility of the ICT mechanism was proposed in the recognition reaction of NTR by the fluorescent probe LDO-NTR, accompanied by the PET process, which will also be theoretically unravelled below. As shown in Fig. S15 (ESI†), each probe (LDO-NTR and 2a-3-NTR) and product (2a and 2a-3) is divided into three fragments, and their natural population analysis (NPA) is presented in Table S16 (ESI†). Evidently, the recognition group (4-nitrobenzyl alcohol) in the probes functions as an electron acceptor. After the reaction with NTR, the 4-nitrobenzyl alcohol group is hydrolyzed to generate a phenolic compound, with the hydroxyl group serving as an electron donor. Analyzing the differences in natural charge population (ΔQ) between the S₁ and S₀ states of each probe's fragment (ΔQ: fragment I < 0, II > 0, III > 0), indicates that the negative charges increase in fragment I and decrease in II and III. While the products have the opposite effect (ΔQ: fragment I > 0, II < 0, III > 0), meaning that the negative charges decrease in fragment I and III and increase in II. Thus, it may be inferred that intramolecular charge transfer (ICT) is involved in the excited state relaxation process of all the probes and products.

The above discussion simply demonstrates the existence of the PET and ICT mechanisms by the FMO theory and NPA analysis. An oxidation–reduction reaction dyad is formed between the receptor and fluorophore in the probe because it has an extra receptor component compared with the corresponding product. The Rehm–Weller formula reveals that an electron in the higher HOMO level of the receptor would be transferred to the HOMO of the excited fluorophore under the perturbation interaction of its hole–electron pair. Because of the sufficiently large negative free energy, this redox reaction can occur spontaneously. As a result, the PET process happens in the probe. The next questions are, what are the reaction rates if the probes undergo PET and ICT processes? Do the fluorophores still have enough fluorescence in their excited states to compete with the PET? From the following Marcus equation,^51–53 the rates are determined to further confirm the occurrence of the two mechanisms.

	(3.8)

where Planck's constant and Boltzmann's constant are represented by h and k_B, respectively. The absolute temperature (T) is set to 298 K. λ and ΔG⁰ stand for the reorganization energy and free energy change, respectively. The symbol V denotes the electronic coupling matrix element.

The Rehm–Weller equation^54,55 can be utilized to calculate the reaction free energy (ΔG_ET) in the PET process.

ΔGET = E0(D+/D) − E0(A/A−) − ΔE0,0 + w	(3.9)

where the oxidation potential of the electron-donor, E⁰(D⁺/D), is defined as the energy difference between its optimized ionic and neutral states. Likewise, E⁰(A/A⁻) represents the energy difference between the optimized neutral and ionic states of the electron-acceptor fragment. As shown in Fig. S15 and S16 (ESI†), the probes are separated into donor and acceptor segments based on the findings listed in Table S16 (ESI†). The ICT and PET process share the same precise divisions for the probes. The zero–zero transition energy is denoted by ΔE_0,0. One way to estimate the w term in the formula is by using eqn (3.10):

	(3.10)

where the dielectric constants in the vacuum and water solvent are represented by ε₀ and ε, respectively. The electron charge is referred to as e, and the distance between the donor and acceptor is symbolized by R.

The following formula can be used to determine the reorganization energy in the PET process:⁵⁶

	(3.11)

in which r₁ and r₂ denote the radii of the electron donor and acceptor, respectively. The symbol n corresponds to the refractive index of the water solvent.

The expression of ΔG⁰ during the ICT process is as follows:⁵⁷

ΔGCT = EEA(A) − EIP(D) − ΔE0-0 − Eb	(3.12)

in this case, the ionization potential of the donor and electron affinity of the acceptor, respectively, are represented by E_IP(D) and E_EA(A). The energy of the lowest excited state of the free donor is marked by ΔE_0-0, while the exciton binding energy of the whole molecule is expressed as E_b.

The following equation determines the reorganization energy in the ICT process:

λ = E(A−) − E(A) + E(D) − E(D+)	(3.13)

here, E(A⁻) and E(A) are the energies of the neutral acceptor at anionic and optimal ground state geometries, respectively. E(D) and E(D⁺) note the energies of the radical cation donor in neutral and optimal cation geometries, respectively.

Ultimately, the Generalized Mulliken–Hush method⁵⁸ can be used to estimate the electronic coupling term V of these two rate equations.

	(3.14)

where, μ₁₂ is the electron transition dipole moment between the initial state and final state. The dipole moment difference and the energy difference between the two states are symbolized by Δμ₁₂ and ΔE₁₂, respectively.

The aforementioned equations are used to calculate the PET and ICT rates based on the S₀ → S₁ process in this paper, and the theoretically simulated related physical parameters are provided in Table 4. It can be seen that the ICT rate for LDO-NTR is 1.02 × 10¹⁴ s⁻¹, which is noticeably faster than the PET rate (7.36 × 10⁹ s⁻¹). Stated otherwise, the probe LDO-NTR may effortlessly transition from a local excited state to an ICT state, that is, undergo charge redistribution, after which it can either radiate a photon for fluorescence or undergo PET through electron transfer. As a result of the fluorescence process being forbidden (the oscillator strength of S₁ → S₀ is zero, see Table S15, ESI†), ICT happens upon light excitation and PET occurs immediately thereafter. Simultaneously, the situation remains almost exactly similar for the designed probe 2a-3-NTR. Consequently, the negligible fluorescence of the probes results from the combined effects of PET and ICT. These findings unmistakably show the dynamic trend of the PET and ICT processes. Hence, it is possible to draw the following conclusions: (1) since the fluorescence emission process is prohibited for LDO-NTR, PET occurs between the fluorophore and receptor, whether by calculating the kinetic rates or by estimating their states using the Rehm–Weller formula. After LDO-NTR reacts with NTR, the HOMO and LUMO electron clouds are delocalized throughout the product 2a. Subsequently, the PET process is interrupted, resulting in a strong fluorescence. Furthermore, both the probe LDO-NTR and product 2a involve the ICT process. (2) The designed probe 2a-3-NTR might share the same fluorescence quenching mechanism as LDO-NTR.

---

---

3.7. Solvation free energy

Bearing in mind that two-photon fluorescent probes are usually used for biological tissues, good solubility is essential. As listed in Table 5, the solvation free energy (ΔG_solv) values of all the studied molecules are calculated in water solution. The detailed results indicate that the ΔG_solv of the designed probe 2a-3-NTR (−19.50 kcal mol⁻¹) and product 2a-3 (−55.04 kcal mol⁻¹) remain almost comparable to those of the experimental probe LDO-NTR (−18.46 kcal mol⁻¹) and product 2a (−56.16 kcal mol⁻¹). This implies that the design strategy of replacing the dialkylamino group by alkyl cyclization and expanding the π-conjugation fails to worsen the water solubility for LDO-NTR-based probes.

---

4. Conclusions

In this paper, a series of anthocyanidins with different backbones and substituents are chosen and designed as NTR probes, and their photophysical properties are thoroughly investigated. Subtle differences in photophysical nature resulting from non-conjugated modifications in their substituents and backbones are fully revealed. The research on a series of anthocyanidins with the same substituents but different fluorophore skeletons 1a–3a and 1b–3b clarifies the effect of six/five-membered rings in the backbone on their photophysical properties. The results indicate that six/five-membered rings fused in the backbone have considerable differences in deep electronic occupied orbitals and geometrical spatial hindrance upon excitation. In comparison with the five-membered ring fused in the backbone, the six-membered ring in the backbone makes molecular geometry relatively easy to relax. However, a-series molecules have a higher reorganization energy, generating more energy loss upon light excitation, which allows the reaction products to detect NTR by a larger Stokes shift. More importantly, the loss of fluorescence intensity is extremely low when the Stokes shift is increased. These features are highly useful for high-resolution NTR detection. On this basis, we have designed 2a-n (n = 1–5) series probe molecules, by using different substituent and backbone π conjugation modifications to improve their two-photon absorption and fluorescence performance. As a result, 2a-3 has a better figure of merit: its emission wavelength, Stokes shift and TPA cross section are as high as 691.42 nm, 110.88 nm and 957.36 GM, respectively, making its probe molecule 2a-3-NTR a promising candidate for high-performance two-photon fluorescence probes. In addition, we also demonstrate how the experimental and designed anthocyanin probes quench fluorescence, and the involvement of ICT and PET mechanisms in the NTR detection are elucidated and validated by using the static state and dynamic rate. The findings show that after being excited by light, both the probes and the products reach an excited state, and the ICT process occurs instantaneously. The probe molecule then undergoes a PET process, causing fluorescence quenching, while the product emits photons from the excited state and then returns to the S₀ state due to the lack of PET structural feature. Given the larger Stokes shift (110.88 nm), two-photon absorption cross section (957.36 GM), and longer fluorescence wavelength (691.42 nm), the designed two-photon fluorescence probe molecule 2a-3-NTR is considered the most promising candidate molecule for detecting NTR. Further experimental confirmation is expected.

Author contributions

Xiu-e Zhang: investigation, data curation, and writing – original draft. Xue Wei: methodology, writing – review & editing. Wei-Bo Cui: writing – review & editing. Jin-Pu Bai: writing – review & editing. Aynur Matyusup: writing – review & editing. Jing-Fu Guo: supervision and software. Hui Li: project administration and funding acquisition. Ai-Min Ren: supervision and writing – review & editing.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Natural Science Foundation of Jilin Province of China (no. 20240101167JC), the Key Research and Development Project of Jilin Provincial Department of Science and Technology (no. 20240302015GX), the 2020-JCJQ project (GFJQ2126-007) and the Natural Science Foundation of China (no. 21473071, 21173099, and 20973078).

References

1. Y. Li, Y. Sun, J. Li, Q. Su, W. Yuan, Y. Dai, C. Han, Q. Wang, W. Feng and F. Li, Ultrasensitive Near-Infrared Fluorescence-Enhanced Probe for in Vivo Nitroreductase Imaging, J. Am. Chem. Soc., 2015, 137, 6407–6416 CrossRef CAS PubMed .
2. F. Liu, H. Zhang, K. Li, Y. Xie and Z. Li, A Novel NIR Fluorescent Probe for Highly Selective Detection of Nitroreductase and Hypoxic-Tumor-Cell Imaging, Molecules, 2021, 26, 4425 CrossRef CAS PubMed .
3. Y. Shi, S. Zhang and X. Zhang, A novel near-infrared fluorescent probe for selectively sensing nitroreductase (NTR) in an aqueous medium, Analyst, 2013, 138, 1952 RSC .
4. S. Li, D. Cheng, L. He and L. Yuan, Recent Progresses in NIR-I/II Fluorescence Imaging for Surgical Navigation, Front. Bioeng. Biotechnol., 2021, 9, 768698 CrossRef PubMed .
5. Y. Liu, Y. Li, S. Koo, Y. Sun, Y. Liu, X. Liu, Y. Pan, Z. Zhang, M. Du, S. Lu, X. Qiao, J. Gao, X. Wang, Z. Deng, X. Meng, Y. Xiao, J. S. Kim and X. Hong, Versatile Types of Inorganic/Organic NIR-IIa/IIb Fluorophores: From Strategic Design toward Molecular Imaging and Theranostics, Chem. Rev., 2022, 122, 209–268 CrossRef CAS PubMed .
6. F. Ma, C.-C. Li and C.-Y. Zhang, Nucleic acid amplification-integrated single-molecule fluorescence imaging for in vitro and in vivo biosensing, Chem. Commun., 2021, 57, 13415–13428 RSC .
7. L. Xu, J. Zhang, L. Yin, X. Long, W. Zhang and Q. Zhang, Recent progress in efficient organic two-photon dyes for fluorescence imaging and photodynamic therapy, J. Mater. Chem. C, 2020, 8, 6342–6349 RSC .
8. L. Wu, J. Liu, P. Li, B. Tang and T. D. James, Two-photon small-molecule fluorescence-based agents for sensing, imaging, and therapy within biological systems, Chem. Soc. Rev., 2021, 50, 702–734 RSC .
9. X. Wang, P. Li, Q. Ding, C. Wu, W. Zhang and B. Tang, Observation of Acetylcholinesterase in Stress-Induced Depression Phenotypes by Two-Photon Fluorescence Imaging in the Mouse Brain, J. Am. Chem. Soc., 2019, 141, 2061–2068 CrossRef CAS PubMed .
10. L. Yuan, L. Wang, B. K. Agrawalla, S.-J. Park, H. Zhu, B. Sivaraman, J. Peng, Q.-H. Xu and Y.-T. Chang, Development of Targetable Two-Photon Fluorescent Probes to Image Hypochlorous Acid in Mitochondria and Lysosome in Live Cell and Inflamed Mouse Model, J. Am. Chem. Soc., 2015, 137, 5930–5938 CrossRef CAS PubMed .
11. H. M. Kim and B. R. Cho, Two-Photon Probes for Intracellular Free Metal Ions, Acidic Vesicles, And Lipid Rafts in Live Tissues, Acc. Chem. Res., 2009, 42, 863–872 CrossRef CAS PubMed .
12. X. Wu, R. Wang, S. Qi, N. Kwon, J. Han, H. Kim, H. Li, F. Yu and J. Yoon, Rational Design of a Highly Selective Near-Infrared Two-Photon Fluorogenic Probe for Imaging Orthotopic Hepatocellular Carcinoma Chemotherapy, Angew. Chem., Int. Ed., 2021, 60, 15418–15425 CrossRef CAS PubMed .
13. A.-M. Caminade, A. Zibarov, E. Cueto Diaz, A. Hameau, M. Klausen, K. Moineau-Chane Ching, J.-P. Majoral, J.-B. Verlhac, O. Mongin and M. Blanchard-Desce, Fluorescent phosphorus dendrimers excited by two photons: synthesis, two-photon absorption properties and biological uses, Beilstein J. Org. Chem., 2019, 15, 2287–2303 CrossRef CAS PubMed .
14. Y. Tang, X. Kong, A. Xu, B. Dong and W. Lin, Development of a Two-Photon Fluorescent Probe for Imaging of Endogenous Formaldehyde in Living Tissues, Angew. Chem., Int. Ed., 2016, 55, 3356–3359 CrossRef CAS PubMed .
15. L. Yuan, F. Jin, Z. Zeng, C. Liu, S. Luo and J. Wu, Engineering a FRET strategy to achieve a ratiometric two-photon fluorescence response with a large emission shift and its application to fluorescence imaging, Chem. Sci., 2015, 6, 2360–2365 RSC .
16. H. M. Kim and B. R. Cho, Small-Molecule Two-Photon Probes for Bioimaging Applications, Chem. Rev., 2015, 115, 5014–5055 CrossRef CAS PubMed .
17. W. Niu, L. Guo, Y. Li, S. Shuang, C. Dong and M. S. Wong, Highly Selective Two-Photon Fluorescent Probe for Ratiometric Sensing and Imaging Cysteine in Mitochondria, Anal. Chem., 2016, 88, 1908–1914 CrossRef CAS PubMed .
18. C. Chen, L. Zhou, W. Liu and W. Liu, Coumarinocoumarin-Based Two-Photon Fluorescent Cysteine Biosensor for Targeting Lysosome, Anal. Chem., 2018, 90, 6138–6143 CrossRef CAS PubMed .
19. T. Peng and D. Yang, Construction of a Library of Rhodol Fluorophores for Developing New Fluorescent Probes, Org. Lett., 2010, 12, 496–499 CrossRef CAS PubMed .
20. S. Xu, H.-W. Liu, X.-X. Hu, S.-Y. Huan, J. Zhang, Y.-C. Liu, L. Yuan, F.-L. Qu, X.-B. Zhang and W. Tan, Visualization of Endoplasmic Reticulum Aminopeptidase 1 under Different Redox Conditions with a Two-Photon Fluorescent Probe, Anal. Chem., 2017, 89, 7641–7648 CrossRef CAS PubMed .
21. D. Liu, Y. Lv, M. Chen, D. Cheng, Z. Song, L. Yuan and X. Zhang, A long wavelength emission two-photon fluorescent probe for highly selective detection of cysteine in living cells and an inflamed mouse model, J. Mater. Chem. B, 2019, 7, 3970–3975 RSC .
22. L. Qiao, Y. Yang, J. Cai, X. Lv, J. Hao and Y. Li, Long wavelength emission fluorescent probe for highly selective detection of cysteine in living cells, Spectrochim. Acta, Part A, 2022, 264, 120247 CrossRef CAS PubMed .
23. T. Ren, W. Xu, F. Jin, D. Cheng, L. Zhang, L. Yuan and X. Zhang, Rational Engineering of Bioinspired Anthocyanidin Fluorophores with Excellent Two-Photon Properties for Sensing and Imaging, Anal. Chem., 2017, 89, 11427–11434 CrossRef CAS PubMed .
24. R. Peng, J. Yuan, D. Cheng, T. Ren, F. Jin, R. Yang, L. Yuan and X. Zhang, Evolving a Unique Red-Emitting Fluorophore with an Optically Tunable Hydroxy Group for Imaging Nitroreductase in Cells, in Tissues, and in Vivo, Anal. Chem., 2019, 91, 15974–15981 CrossRef CAS PubMed .
25. E. Runge and E. K. U. Gross, Density-Functional Theory for Time-Dependent Systems, Phys. Rev. Lett., 1984, 52, 997–1000 CrossRef CAS .
26. J. Autschbach, T. Ziegler, S. J. A. van Gisbergen and E. J. Baerends, Chiroptical properties from time-dependent density functional theory. I. Circular dichroism spectra of organic molecules, J. Chem. Phys., 2002, 116, 6930–6940 CrossRef CAS .
27. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato, A. V. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. J. Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov, T. A. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas, J. B. Foresman and D. J. Fox, Gaussian 16, Revision B.01, Gaussian, Inc., Wallingford CT, 2016 Search PubMed .
28. P. Sałek, O. Vahtras, T. Helgaker and H. Ågren, Density-functional theory of linear and nonlinear time-dependent molecular properties, J. Chem. Phys., 2002, 117, 9630–9645 CrossRef .
29. K. Aidas, C. Angeli, K. L. Bak, V. Bakken, R. Bast, L. Boman, O. Christiansen, R. Cimiraglia, S. Coriani, P. Dahle, E. K. Dalskov, U. Ekström, T. Enevoldsen, J. J. Eriksen, P. Ettenhuber, B. Fernández, L. Ferrighi, H. Fliegl, L. Frediani, K. Hald, A. Halkier, C. Hättig, H. Heiberg, T. Helgaker, A. C. Hennum, H. Hettema, E. Hjertenaes, S. Høst, I.-M. Høyvik, M. F. Iozzi, B. Jansík, H. J. A. A. Jensen, D. Jonsson, P. Jørgensen, J. Kauczor, S. Kirpekar, T. Kjaergaard, W. Klopper, S. Knecht, R. Kobayashi, H. Koch, J. Kongsted, A. Krapp, K. Kristensen, A. Ligabue, O. B. Lutnaes, J. I. Melo, K. V. Mikkelsen, R. H. Myhre, C. Neiss, C. B. Nielsen, P. Norman, J. Olsen, J. M. H. Olsen, A. Osted, M. J. Packer, F. Pawlowski, T. B. Pedersen, P. F. Provasi, S. Reine, Z. Rinkevicius, T. A. Ruden, K. Ruud, V. V. Rybkin, P. Sałek, C. C. M. Samson, A. S. de Merás, T. Saue, S. P. A. Sauer, B. Schimmelpfennig, K. Sneskov, A. H. Steindal, K. O. Sylvester-Hvid, P. R. Taylor, A. M. Teale, E. I. Tellgren, D. P. Tew, A. J. Thorvaldsen, L. Thøgersen, O. Vahtras, M. A. Watson, D. J. D. Wilson, M. Ziolkowski and H. Ågren, The Dalton quantum chemistry program system: the Dalton program, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2014, 4, 269–284 CAS .
30. J. Tomasi, B. Mennucci and R. Cammi, Quantum Mechanical Continuum Solvation Models, Chem. Rev., 2005, 105, 2999–3094 CrossRef CAS PubMed .
31. J. Ho, A. Klamt and M. L. Coote, Comment on the Correct Use of Continuum Solvent Models, J. Phys. Chem. A, 2010, 114, 13442–13444 CrossRef CAS PubMed .
32. Schrodinger, The PyMOL Molecular Graphics System, Version 1.8, 2015 Search PubMed .
33. H. Ma, Q. Peng, Z. An, W. Huang and Z. Shuai, Efficient and Long-Lived Room-Temperature Organic Phosphorescence: Theoretical Descriptors for Molecular Designs, J. Am. Chem. Soc., 2019, 141, 1010–1015 CrossRef CAS PubMed .
34. C. Wang, W. Chi, Q. Qiao, D. Tan, Z. Xu and X. Liu, Twisted intramolecular charge transfer (TICT) and twists beyond TICT: from mechanisms to rational designs of bright and sensitive fluorophores, Chem. Soc. Rev., 2021, 50, 12656–12678 RSC .
35. Z. R. Grabowski, K. Rotkiewicz and W. Rettig, Structural Changes Accompanying Intramolecular Electron Transfer: Focus on Twisted Intramolecular Charge-Transfer States and Structures, Chem. Rev., 2003, 103, 3899–4032 CrossRef PubMed .
36. Z. Ye, W. Yang, C. Wang, Y. Zheng, W. Chi, X. Liu, Z. Huang, X. Li and Y. Xiao, Quaternary Piperazine-Substituted Rhodamines with Enhanced Brightness for Super-Resolution Imaging, J. Am. Chem. Soc., 2019, 141, 14491–14495 CrossRef CAS PubMed .
37. X. Liu, Q. Qiao, W. Tian, W. Liu, J. Chen, M. J. Lang and Z. Xu, Aziridinyl Fluorophores Demonstrate Bright Fluorescence and Superior Photostability by Effectively Inhibiting Twisted Intramolecular Charge Transfer, J. Am. Chem. Soc., 2016, 138, 6960–6963 CrossRef CAS PubMed .
38. J. B. Grimm, A. K. Muthusamy, Y. Liang, T. A. Brown, W. C. Lemon, R. Patel, R. Lu, J. J. Macklin, P. J. Keller, N. Ji and L. D. Lavis, A general method to fine-tune fluorophores for live-cell and in vivo imaging, Nat. Methods, 2017, 14, 987–994 CrossRef CAS PubMed .
39. T. Karstens and K. Kobs, Rhodamine B and rhodamine 101 as reference substances for fluorescence quantum yield measurements, J. Phys. Chem., 1980, 84, 1871–1872 CrossRef CAS .
40. G. Jones, W. R. Jackson, C. Y. Choi and W. R. Bergmark, Solvent effects on emission yield and lifetime for coumarin laser dyes. Requirements for a rotatory decay mechanism, J. Phys. Chem., 1985, 89, 294–300 CrossRef CAS .
41. L. Shen, X. Ding, T. He, X. Hao, J. Guo, L. Zou and A. Ren, New insight into the effects of N^N ligand isomerization and methyl modification on the phosphorescence properties of Cu(I) complexes with (1-(2-pyridyl)pyrazole/imidazole) ligands, New J. Chem., 2018, 42, 3660–3670 RSC .
42. L. Yuan, W. Lin and H. Chen, Analogs of Changsha near-infrared dyes with large Stokes Shifts for bioimaging, Biomaterials, 2013, 34, 9566–9571 CrossRef CAS PubMed .
43. X. Peng, F. Song, E. Lu, Y. Wang, W. Zhou, J. Fan and Y. Gao, Heptamethine Cyanine Dyes with a Large Stokes Shift and Strong Fluorescence: A Paradigm for Excited-State Intramolecular Charge Transfer, J. Am. Chem. Soc., 2005, 127, 4170–4171 CrossRef CAS PubMed .
44. M. D. M. Alam, M. Chattopadhyaya, S. Chakrabarti and K. Ruud, Chemical Control of Channel Interference in Two-Photon Absorption Processes, Acc. Chem. Res., 2014, 47, 1604–1612 CrossRef CAS PubMed .
45. A. Avramopoulos, R. Zaleśny, H. Reis and M. G. Papadopoulos, A Computational Strategy for the Design of Photochromic Derivatives Based on Diarylethene and Nickel Dithiolene with Large Contrast in Nonlinear Optical Properties, J. Phys. Chem. C, 2020, 124, 4221–4241 CrossRef CAS .
46. M. D. M. Alam, M. Chattopadhyaya and S. Chakrabarti, Solvent induced channel interference in the two-photon absorption process—a theoretical study with a generalized few-state-model in three dimensions, Phys. Chem. Chem. Phys., 2012, 14, 1156–1165 RSC .
47. P. Cronstrand, Y. Luo and H. Ågren, Generalized few-state models for two-photon absorption of conjugated molecules, Chem. Phys. Lett., 2002, 352, 262–269 CrossRef CAS .
48. Z. Xu, A.-M. Ren, J.-F. Guo, X.-T. Liu, S. Huang and J.-K. Feng, A Theoretical Investigation of Two Typical Two-Photon pH Fluorescent Probes, Photochem. Photobiol., 2013, 89, 300–309 CrossRef CAS PubMed .
49. H. Lu, S. Zhang, H. Liu, Y. Wang, Z. Shen, C. Liu and X. You, Experimentation and Theoretic Calculation of a BODIPY Sensor Based on Photoinduced Electron Transfer for Ions Detection, J. Phys. Chem. A, 2009, 113, 14081–14086 CrossRef CAS PubMed .
50. H. Salman, S. Tal, Y. Chuvilov, O. Solovey, Y. Abraham, M. Kapon, K. Suwinska and Y. Eichen, Sensitive and Selective PET-Based Diimidazole Luminophore for Zn^II Ions: A Structure−Activity Correlation, Inorg. Chem., 2006, 45, 5315–5320 CrossRef CAS PubMed .
51. H. Guo, Y. Jing, X. Yuan, S. Ji, J. Zhao, X. Li and Y. Kan, Highly selective fluorescent OFF–ON thiol probes based on dyads of BODIPY and potent intramolecular electron sink 2,4-dinitrobenzenesulfonyl subunits, Org. Biomol. Chem., 2011, 9, 3844 RSC .
52. H. Zong, X. Wang, J. Quan, C. Tian and M. Sun, Photoinduced charge transfer by one and two-photon absorptions: physical mechanisms and applications, Phys. Chem. Chem. Phys., 2018, 20, 19720–19743 RSC .
53. X. Zhang, L. Chi, S. Ji, Y. Wu, P. Song, K. Han, H. Guo, T. D. James and J. Zhao, Rational Design of d-PeT Phenylethynylated-Carbazole Monoboronic Acid Fluorescent Sensors for the Selective Detection of α-Hydroxyl Carboxylic Acids and Monosaccharides, J. Am. Chem. Soc., 2009, 131, 17452–17463 CrossRef CAS PubMed .
54. J. Cody, S. Mandal, L. Yang and C. J. Fahrni, Differential Tuning of the Electron Transfer Parameters in 1,3,5-Triarylpyrazolines: A Rational Design Approach for Optimizing the Contrast Ratio of Fluorescent Probes, J. Am. Chem. Soc., 2008, 130, 13023–13032 CrossRef CAS PubMed .
55. H. Sunahara, Y. Urano, H. Kojima and T. Nagano, Design and Synthesis of a Library of BODIPY-Based Environmental Polarity Sensors Utilizing Photoinduced Electron-Transfer-Controlled Fluorescence ON/OFF Switching, J. Am. Chem. Soc., 2007, 129, 5597–5604 CrossRef CAS PubMed .
56. T. Fournier, S. M. Tavender, A. W. Parker, G. D. Scholes and D. Phillips, Competitive Energy and Electron-Transfer Reactions of the Triplet State of 1-Nitronaphthalene: A Laser Flash Photolysis and Time-Resolved Resonance Raman Study, J. Phys. Chem. A, 1997, 101, 5320–5326 CrossRef CAS .
57. G. J. Kavarnos and N. J. Turro, Photosensitization by reversible electron transfer: theories, experimental evidence, and examples, Chem. Rev., 1986, 86, 401–449 CrossRef CAS .
58. M. Rust, J. Lappe and R. J. Cave, Multistate Effects in Calculations of the Electronic Coupling Element for Electron Transfer Using the Generalized Mulliken–Hush Method, J. Phys. Chem. A, 2002, 106, 3930–3940 CrossRef CAS .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cp02067g

This journal is © the Owner Societies 2024