Читать книгу Handbook of Aggregation-Induced Emission, Volume 1 - Группа авторов - Страница 53

In addition to the observation of the EZI process that can disclose RDBR mechanism, immobilization of TPE propeller‐like conformation, especially cyclization of TPE at cis‐position, can be used to explore the RDBR process. After bridging between two phenyl rings of TPE at the cis‐position, the double bond will be unable to freely rotate due to the restriction of the bridge chain. But the phenyl rings can still freely rotate. Therefore, the effect of RDBR on the fluorescence will be clearly observed.

In 2016, Zheng’s group [43] found an ideal route for the synthesis of cis‐TPE dicycle in which the double bond rotation can be blocked at the excited state (see Figure 3.15). By intramolecular nucleophilic substitution of 2 with 1,4‐bis(bromomethyl)benzene, cis‐TPE dicycle tetraldehyde 3 could be obtained in a 43% yield. The formation of cis‐TPE dicycles between two phenyl groups at the cis‐position instead of those at the gem‐position was ascribed to the length of 1,4‐benzenedimethyl tether that was more compatible with the distance between two cis‐phenyl rings than that between two gem‐phenyl ones. With this key intermediate in hand, even TPE tetracycle 6 whose propeller‐like conformation was completely immobilized was obtained.

Noticeably, while the TPE derivatives 1 and 2 bearing no intramolecular cycle did not emit fluorescence in solution, cis‐TPE dicycles 3–5 displayed a strong emission in solution. In THF, Φ_f of 2 was 0% but that of 3 and 4 was 24 and 49%, respectively. Due to the bridge, p‐phenylenedimethyl units were short and rigid, and the propeller‐like conformation of the TPE dicycle should have been fixed. However, the enantiomers obtained by chiral high‐pressure liquid chromatography (HPLC) rapidly racemized in solution at room temperature. This result indicated that the phenyl rings of the cis‐TPE dicycles was still able to rotate although the rotation had a little restriction. In addition, the formaldehyde groups of cis‐TPE dicycle 3 reacted with chiral α‐methylbenzylamine to form imine, and positive circular dichroism (CD) signals could be induced by (R)‐α‐methylbenzylamine but negative CD signals were aroused by (S)‐α‐methylbenzylamine. The CD signals should result from a single‐handed propeller‐like conformation induced by the enantiomer of the chiral amine because neither individual 3 nor amine enantiomers could show such signals. This meant that the propeller‐like conformation of 3 could be transformed from the left‐handed helical direction to the right‐handed helical direction, demonstrating that the phenyl could freely rotate. Consequently, it could be inferred that about 25–50% Φ_f should come from the RDBR process (see Figure 3.16).

As expected, the resolved enantiomers from TPE tetracycle 6 were stable due to complete immobilization of its propeller‐like conformation. The racemate of 6 had Φ_f up to 97%, and both of the two enantiomers had a quantitative Φ_f up to 100% due to restriction of not only double bond rotation but also phenyl ring rotation. Compared with the corresponding TPE dicycle 4, TPE tetracycle 6 showed a twofold increase in fluorescence intensity, demonstrating that RDBR and RIR play equal key roles on the AIE effect.

In addition, the two enantiomers of 6, M‐6 (left‐handed helical propeller‐like configuration) and P‐6 (left‐handed helical one), emitted strong circularly polarized luminescence (CPL) signals in THF with a dissymmetric factor (g_lum) of +3.1 × 10⁻³ for M‐6 and −3.3 × 10⁻³ for P‐6, indicating that the propeller‐like conformation of TPE was maintained even at the excited state. Moreover, the |g_lum| of CPL was similar with the g_abs of absorbance (g_abs = 2(Δε/ε)) of 2.4 × 10⁻³ for M‐6 and P‐6, indicating little conformational change between the ground state and excited state. This further confirmed that no double bond rotation occurred for this TPE tetracycle (see Figure 3.17).

Figure 3.16 CD spectra of a mixture of TPE dicycle 3 and enantiomer of α‐methylbenzylamine 7 in the presence of acetic acid in 1,2‐dichloroethane.

Figure 3.17 The crystal structures of M‐6 (a) and P‐6 (b); (c) photos of 3, 4, and 6 in THF solution under a 365‐nm UV light and (d) CPL spectra of M‐6 and P‐6 in THF (1.0 × 10⁻³ M).

Source: Reproduced with permission from Ref. [43]. Copyright 2016, American Chemical Society.

In order to further disclose the important contribution of the RDBR process to the AIE effect, gem‐TPE dicycles 7 and 8 and even the typical isomers cis‐ and gem‐TPE dicycles 9 and 10 were designed and synthesized in Zheng’s group (see Figure 3.18) [44]. Their configuration had been confirmed by the crystal structure. By comparing the fluorescence intensity of these isomers, the effect of groups, atoms, and the bridge chains on the fluorescence could be excluded. Therefore, more direct and more exact RDBR evidence could be furnished.

Figure 3.18 (a) Structures of TPE dicycle isomers 7–10 (left) and photos of their solution in THF (1.0 × 10⁻⁴ M) under a 365‐nm UV light (middle). (b) The change of ¹H NMR spectra of 7 in CDCl₃ with temperature.

Source: Reproduced with permission from Ref. [44]. Copyright 2018, American Chemical Society.

As expected, while cis‐TPE dicycles 3, 4, and 9 emitted a strong fluorescence, the gem‐TPE dicycles 7, 8, and 10 had no emission in THF under an irradiation of a 365‐nm UV light. The fluorescence quantum yields of the cis‐TPE dicycles 3, 4, and 9 were 24, 49, and 22%, while those of the gem‐TPE dicycles 7, 8, and 10 were 3.1, 5.7, and 1.9%, respectively, in THF at 25 °C. Compared with the gem‐isomer 10, the cis‐isomer 9 displayed an increase of fluorescence intensity by 11.6‐fold. From the ¹H NMR of the gem‐TPE dicycle 7 at different temperatures, the phenyl ring of the TPE unit was fixed and failed to rotate due to linkage of the bridge at the m‐position of the phenyl rings. However, the bridging unit 1,4‐benzenedioxymethyl was rapidly rotating at 25 °C because the proton peaks of both the 1,4‐benzenedioxy and methylene unit were very wide and flat. When the temperature was lowered to 0 °C, these proton resonance signals became sharp and well resolved, indicating that the free rotation of these bridge units was also restricted. However, the Φ_f of 7 was only increased to 9.0% at 0 °C and was still much less than that of cis‐isomers. Compared with the cis‐isomer, the fluorescent decrease of the gem‐isomer in solution only came from free rotation of the double bond at the excited state after all other intramolecular rotations had been restricted (see Figure 3.18 right).

Time‐resolved fluorescence decay of TPE dicycles 3, 4, and 7–10 disclosed that the fluorescence lifetime was in the range of 6.9–14 ns in THF and 13–19 ns in suspension for these TPE dicycles. In suspension, their fluorescence lifetime was always larger than that in solution. This demonstrated that there was further restriction of intramolecular rotation in solid state. By making use of the fluorescence quantum yields and lifetime values, the radiative (k_f) and nonradiative (k_nr) rate constants of 3, 4, and 7–10 (k_f, k_nr (ns⁻¹)) could be calculated. The k_f and k_nr were 0.035 and 0.112 for 3, 0.051 and 0.053 for 4, 0.004 and 0.126 for 7, 0.004 and 0.067 for 8, 0.023 and 0.080 for 9, and 0.002 and 0.103 for 10. And ratios of k_nr vs. k_f for 3, 4, and 7–10 were 3.20, 1.04, 31.5, 16.8, 3.48, and 51.5, respectively. The k_nr/k_f ratio from gem‐isomers was always much larger than that from cis‐isomers, demonstrating a more nonradiative process from gem‐isomers. This nonradiative process should be mainly ascribed to the double bond rotation.

If the double bond rotated at the excited state, one intermediate state in which sp²‐hybridized orbital planes of two carbons are vertical instead of coplanar should exist. This twisted state of the double bond should be able to be observed by femtosecond transient absorption spectra because of the decreased conjugation with one another. It was true that two excited‐state absorption (ESA) bands, which were located at <460 and >600 nm, respectively, were observed. The former should come from a twisted excited double bond and the latter came from a planar double bond at the excited state, corroborating the occurrence of the double bond rotation. Surprisingly, the two ESA bands existed both in gem‐isomers and in cis‐isomers. Two typical time components τ₁ and τ₂ from the dynamic decay process were obtained. By the global analysis, the first component τ₁ representing the rise component of ESA and associated with the geometry relaxation from the Franck–Condon (FC) configuration was the negative amplitude. The second component τ₂ represented the nonradiative internal conversion (IC) decay of the excited state. From time‐resolved spectra, an obviously growing process and the increase ending up in less than 20 ps were observed for the species at short wavelength, suggesting that the planar excited state could be transformed into the twisted excited state. Therefore, the double bond rotation at the excited state occurred for all the TPE dicycles.

However, there was an obvious difference of the double bond rotation between the cis‐ and gem‐isomers. As an index of the rotation, the component τ₁ was 6 ps for gem‐10 and obviously shorter than 15 ps of cis‐9. This should be ascribed to the more freely rotating double bond in gem‐isomers than in cis‐isomers. At 21 and 14 ps, the rotation was accomplished because the maximum intensity of the absorption spectra of cis‐isomer 9 and gem‐isomer 10 at the excited state was reached, respectively. It was found that the absorption maximum wavelength of the gem‐isomer 10 was shortened by 15 nm compared with that of the cis‐isomer 9. Moreover, the area ratio of short‐wavelength band vs. long‐wavelength band was much larger for the gem‐isomer 10 than that for the cis‐isomer 9, further corroborating that the double bond of the gem‐isomer rotated more freely than that of the cis‐isomer. Therefore, the gem‐isomer showed lower fluorescence quantum yield than the cis‐one because of the freer double bond rotation at the excited state and more nonradiative decay (see Figure 3.19).

Recently, Zheng et al. [45] have exploited the RDBR mechanism to improve the sensitivity of DNA detection and enhance the chiroptical properties from TPE AIEgens. In this regard, cis‐TPE macrocycle diquaternary ammoniums 11 were designed and synthesized. As a comparison, gem‐isomers 12 were also prepared. As soon as the two ammonium arms at the cis‐position simultaneously hold on one DNA chain by electrostatic attraction, the formed cycle together with the original cycle will completely immobilize the double bond rotation at the excited state and arouse the enhancement of the AIE effect (see Figure 3.20).

Figure 3.19 (a) Femtosecond transient absorption spectra of 9 and 10 at the respective maximum intensity. (b) Diagrammatic sketch of the normal excited state (cis^* and gem^*) and twisted excited state (cis^** and gem^**) of the double bond.

Source: Reproduced with permission from Ref. [44]. Copyright 2018, American Chemical Society.

Figure 3.20 Structures of cis‐ and gem‐TPE macrocycle diquaternary ammoniums 11 and 12.

Due to the limitation of crown ether cycle, both cis‐11 and gem‐12 display weak emission in solution. However, while the quantum yield of gem‐isomer 12 was 1.5%, the cis‐TPE ammonium 11 had a Φ_f of 3.0%, which was a 2.0‐fold stronger than that of the gem‐one. This should be ascribed to the partial limitation of the double bond rotation in cis‐isomers but no restriction of the double bond rotation in the gem‐one. When cis‐TPE ammonium cis‐11a or cis‐11b was added to the solution of calf thymus DNA (CT‐DNA), strong new CD signals were induced. In addition to CD signals of DNA itself at short wavelengths, one bisignate band from about 370 nm (+) to 310 nm (−), which should be ascribed to the single‐handed propeller‐like conformation of the TPE unit, appeared. Conversely, gem‐isomers gem‐12a and gem‐12b showed weak CD signals from the TPE unit and did not form a new bisignate band. Probably because of more flexibility of the TPE unit in the gem‐isomer than in the cis‐one, DNA was unable to induce the stable single‐handed propeller‐like conformation.

Outstandingly, strong CPL emission was observed in the drop‐cast film from a mixture of cis‐11a and cis‐11b with CT‐DNA in water, while the gem‐isomer–CT‐DNA film emitted no CPL signals. The CPL dissymmetric factor (g_lum) of 0.0028 and 0.016 for cis‐11a and cis‐11b, respectively, was much larger than that from the mixture of DNA with other AIEgens. Even in solution, strong CPL light was emitted from a mixture of cis‐isomer with CT‐DNA in water but no CPL signals were found for the mixture of gem‐isomers with CT‐DNA. With fish sperm DNA (FS‐DNA), a similar result was obtained. Considering the structure of the cis‐isomers, the CPL enhancement should result from the more RDBR process of cis‐isomers in which the formed cycle in situ together with the original cycle in the cis‐position firmly restricts the double bond rotation not only at the ground state but also at the excited state (see Figure 3.21).

Given that the obvious interaction of the TPE diammoniums with DNA, they should be excellent sensor for the detection of DNA. It was truly that the fluorescence of TPE macrocycle diammoniums in water was increased when FS‐DNA was added into the solution at 1.0 × 10⁻⁶ M. But the solution from a mixture of cis‐isomer with FS‐DNA showed stronger fluorescence than that from the corresponding gem‐isomer and FS‐DNA. The fluorescence intensity was linearly increased in the range of DNA concentration less than 1.0 × 10⁻⁸ M. As a result, the detection limit for DNA analysis was obtained. It was found that the detection limits were 123, 74, 496, and 235 pM for 11a, 11b, 12a, and 12b, respectively. The cis‐isomers had always much lower detection limit than the gem‐ones. And the detection limitation of 74 pM from cis‐isomer is among the best results from AIE DNA sensors. The higher sensitivity of cis‐isomers than that of gem‐isomers should also come from the RDBR mechanism. As shown in Figure 3.22, the restriction of a double bond of cis‐isomers upon binding to DNA chain enhanced the AIE effect. Therefore, the sensitivity was significantly increased.

Figure 3.21 CPL spectra of a drop‐cast film (a) and solution (b) from a mixture of cis‐TPE isomers and gem‐TPE isomers with CT‐DNA in water.

When the cis‐TPE macrocycle diammonium was synthesized using octaethylene glycol as a bridge, the resultant crown ether cycle is large enough to allow the EZI. As shown in Figure 3.23, the as‐prepared cis‐13 could be converted into trans‐13 under an irradiation of a 365‐nm portable UV lamp both in organic solvent and in water. Under a 365‐nm light from one fluorophotometer, the absorption spectrum of cis‐13 in CH₂Cl₂ had a gradual absorbance increase at 355 and 278 nm but a constant decrease at 315 nm with irradiation time, and showed an obvious isosbestic point at 336 and 302 nm, indicating the conversion from cis‐isomer to trans‐one. In CDCl₃, after irradiation by a 365‐nm portable UV lamp for one hour, another set of signals appeared beside signals of cis‐13 in the ¹H NMR spectrum. Especially, besides the signals of the aromatic proton near to the oxygen atom (6.71 ppm, double), benzyl methylene (4.82 ppm, single), and the ethylene proton close to the aromatic ring (4.09 ppm, triple), new peaks that were well separated and had the same shape and split with that of cis‐isomer appeared at a lower field. The integral area of these new peaks was all almost equal to that of cis‐13, suggesting a 50% conversion of cis‐isomer to trans‐one. In DMSO, the converted trans‐isomer (about 20%) by light could completely come back to cis‐isomer under heating at 180 °C. This result suggested that the double bond of the TPE derivatives was easy to rotate in the excited state.

Figure 3.22 Schematic diagram of the binding of TPE cycle diammoniums 11 (a) and 12 (b) to the DNA strand.

The TPE ammonium 13 was tested for the detection of fish sperm DNA. It was found that cis‐13 had a high sensitivity (139 pM) and high‐intensity ratio I_max/I₀ (6.6), whereas a mixture of cis/trans‐13 about 1 : 1 displayed a low‐sensitivity (326 pM) and low‐intensity ratio I_max/I₀ (4.5), indicating that the cis‐isomer was a much better DNA sensor than the trans‐one due to more restriction of intramolecular motion including the double bond rotation (see Figure 3.23).

There are more examples in which TPE cycle at the cis‐position could emit fluorescence in solution due to the restriction of double bond rotation. Rathore et al. synthesized a class of stilbenoprismands 14 via an intramolecular McMurry coupling reaction, which contained a TPE core‐bearing cycle at the cis‐side (see Figure 3.24 left) [46]. While the parent TPE showed no detectable emission, the solution of 14 showed significant emission under similar conditions although there were two substituted phenyl rings that could freely rotate. By the direct coordination of TPE tetraimidazolium salts with silver or gold, Hahn and Strassert [47] prepared TPE dicycles 15 on the cis‐side. The TPE dicycles 15 showed strong emission in solution but TPE tetraimidazolium salts emit no fluorescence, demonstrating the important role of the RDBR process on the fluorescence (see Figure 3.24 right).

Hu et al. [48] reported one TPE derivative dicycle 17, in which two pairs of the phenyl rings at the cis‐position were connected by two diacetylenes. While the no cyclized TPE reactant 16 emitted very weak fluorescence in solution, the emission of 17 is very strong even in dilute solution. This fluorescence enhancement after cyclization at the cis‐position should mainly come from the RDBR process (see Figure 3.25).

Figure 3.23 (a) The cis‐/trans‐isomerization of 13 under a 365‐nm light irradiation and heating. (b) Change in UV–vis spectrum of compounds cis‐13 in dichloromethane under an irradiation of 365‐nm light from fluorophotometer for different periods. [cis‐13] = 1.0 × 10⁻⁵ M. (c) The ¹H NMR spectra of cis‐13 in CDCl₃ before and after irradiation by a 365‐nm portable UV lamp for one hour.

Figure 3.24 The structure of AIEgens 14 and 15.

Figure 3.25 The synthesis of the emissive molecule 17.