Читать книгу Handbook of Aggregation-Induced Emission, Volume 2 - Группа авторов - Страница 19

Hydrogen sulfide (H₂S) is a natural gas with a rotten egg smell. It is poisonous, corrosive, and flammable. Exposure of H₂S with a small amount can give rise to headache, dizziness, and even death. On the other hand, H₂S is an indispensable endogenous gas in human body by metabolism. It is related to different physiological processes like cell growth, vasodilation, regulation of inflammation, and so on. The abnormal level of H₂S is associated with symptoms such as Alzheimer's diseases and diabetes [53].

Tang synthesized an AIEgen of malonitrile‐functionalized TPP (TPP‐PDCV) to act as a ratiometric fluorescent probe to detect H₂S with high sensitivity and good selectivity [54]. TPP‐PDCV shows an orange emission at 565 nm in the DMSO/PBS buffer mixture (v/v = 9 : 1) due to the TICT from the TPP part to the strong electron‐withdrawing malonitrile group. However, upon addition of NaHS, the orange emission disappears gradually, whereas a blue emission centered at 429 nm in the short‐wavelength region enhances accordingly. Such a fluorescent response (I₄₂₉/I₅₆₅) to H₂S changes less until 10 minutes, indicating that the detection is efficient and can be finished in 10 minutes (Figure 1.1). It is due to the activity of the double bond in the malonitrile group, which can undergo nucleophilic addition by H₂S. Thus, the double bond is easy to break to prohibit the TICT effect. On the other hand, TPP‐PDCV is transformed to a thiol‐substituted TPP derivative (TPP‐PSH) after addition and elimination reactions, which continues to oxidize to form the dithio‐containing derivative of TPP‐2PS. Because the resulting TPP‐2PS shows a lower polarity and lager rigidity, it displays a bad solubility in the DMSO/water mixture. The TPP‐2PS is easy to aggregate to produce a blue light signal contributed by the TPP unit.

Figure 1.1 Fluorescent detection of H₂S by TPP‐PDCV. (a) Time‐dependent PL spectra of TPP‐PDCV (25 μM) in the DMSO/PBS buffer mixture (v/v = 9 : 1) in the presence of NaHS (250 μM). (b) Plot of the relative PL intensity (I₄₂₉/I₅₆₅) versus the number of scan in 25 minutes, where I₄₂₉ and I₅₆₅ are the PL intensity at 429 and 565 nm, respectively. Inset: the photographs of the TPP‐PDCV solution before and after the addition of NaHS taken under an irradiation of a 356 nm UV light. (c) PL spectra of TPP‐PDCV in the DMSO/PBS buffer mixture (v/v = 9 : 1) with different NaHS concentrations (0–500 μM). (d) Plot of the relative PL intensity (I₄₂₉/I₅₆₅) versus NaHS concentration. For all the tests, the excitation wavelength is 372 nm.

The sensor can analyze the H₂S quantificationally. For example, the sensor possesses a response with the NaHS concentration range from 0 to 250 μM. However, in the concentration range of 0–75 and 150–225 μM, the linear responses can be observed, thus providing a platform to detect H₂S at low concentration (Figure 1.1). Besides, the sensor shows a good selectivity against other anions (e.g. AcO⁻, F⁻, ClO⁻, IO₄⁻, N₃⁻, NO₂⁻, and OH⁻, etc.), except for CO₃²⁻, which exerts subtle effect on the detection. The biothiols of cysteine, homocysteine, and glutathione also pose some effect because of the presence of thiol in the structures. However, such an influence can be neglected in comparison to the strong signal in the presence of H₂S.

Study on the influence of molecular structure on the sensing property is particularly valuable to provide the clues for designing advanced functional materials. With this regard, Tang prepared three conjugated isomers of TPP‐p‐TPE, TPP‐m‐TPE, and TPP‐o‐TPE by connecting TPP and triphenylethene at the para‐, meta‐, and ortho‐positions, respectively (Figure 1.2a) [55]. The isomers show the AIE effect due to the melding of the typical AIEgens of TPP and triphenylethene in the structures. Their conformations evolve from an extended to a folded one due to the different linkages between two units, which makes the emissions blue‐shifted and luminescence efficiency to decrease gradually. It is reasonable because the molecular conjugation gets worse as the structure changes.

Interestingly, TPP‐o‐TPE is an easy‐to‐form organic porous crystal thanks to its fold conformation locked by the strong intramolecular C–H···N hydrogen bond. It is an easy‐to‐form DCM‐captured porous crystal (Figure 1.2b). In the crystal, every four molecules coordinately generate a cylinder‐like pore with a volume of around 0.27 nm³, which is enough for accommodating two DCM molecules. The pores are distributed uniformly in the crystal and connected in a straight line as long‐ranged nanochannels. The porous crystal produces a fixed framework structure regardless of the guest molecules. For example, each pore can also capture two THF molecules, with the structure of crystal less changed (Figure 1.2c). However, the size of the pore is tunable to fit different volumes of guest molecules.

Considering the difference in the influence of molecular conjugation and porosity caused by the isomerization effect, it is desirable to investigate their sensing behavior. 2,4,6‐Trinitrophenol (picric acid, PA) is chosen as the first analyte because it is a well‐known model explosive and the accurate detection of explosive meets current needs in antiterrorist and protection of country safety. The detection is basically carried out with the nanoaggregates in solution because of the strong emissions of AIE isomers in the aggregate state. Upon addition of PA, the isomers show similar quenching behavior as the concentration of PA increases. The lifetime of the three sensors has a negligible change before and after analyte addition, demonstrating that a static quenching model dominates the sensing mechanism. Since the process normally takes place from the ground state of isomers to the excited state of PA, no excited‐state behaviors of sensors such as fluorescence resonance energy transfer (FRET) and photoinduced electron transfer (PET) will occur, while the interaction between the sensor and the analyte plays a crucial role. Overall, the quenching effect of TPP‐p‐TPE is somewhat higher than those of the others at high PA concentration. It is due to the stronger Lewis acid–base interactions preferred to take place between PA and AIE isomers with better conjugation. On the other hand, although TPP‐o‐TPE possesses the worst molecular conjugation, its quenching effect is a bit larger than TPP‐m‐TPE at a PA concentration of 400–500 μM, which is because of its best porosity that increases the binding capacity between TPP‐o‐TPE and PA. Therefore, both factors collectively determine the detection of PA by nanoprobes (Figure 1.2d).

The study is then focused on the sensing of another analyte of Ru³⁺ by the nanoaggregates because it is important in catalysis but harmful to the environment and toxic to the human beings (Figure 1.2e) [55]. Different from the above research, the isomers show a partial quenching behavior upon addition of Ru³⁺. The quenching keeps linear at a low‐concentration range of Ru³⁺ and reaches plateaus at high concentrations. Besides, the quenching constant of TPP‐p‐TPE, TPP‐m‐TPE, and TPP‐o‐TPE increases accordingly. The lifetime study indicates that an obvious decrease of lifetime of sensors is observed after addition of Ru³⁺. Thus, the quenching mechanism is mainly ascribed to a dynamic (collisional) quenching model, which requires a close contact of the excited molecules and the analyte. The Dexter energy transfer is dominated because a short distance of the donor and the acceptor may help the electron exchange between them. On the other hand, the possible influence of FRET on the sensing is ruled out because all the emissions should be quenched if it exists. It seems that the change of the quenching constant of the sensors is related to the variation in the PL quantum efficiency of isomers. It is more likely that the interaction of the isomers and Ru³⁺ plays a major role and a much strong interaction will be produced between Ru³⁺ and isomers with better conjugation.

Figure 1.2 Fluorescent detection of PA and Ru³⁺ by AIE isomers. (a) Molecular structures of AIE isomers. (b) DCM‐ and (c) THF‐containing organic porous crystals of TPP‐o‐TPE. (d) Plots of I₀/I‐1 versus PA concentration. Inset: molecular structure of PA. (e) Plots of I₀/I‐1 versus Ru³⁺ concentration. Inset: molecular structure of PA. In (d) and (e), I₀ and I are the peak intensity of probes before and after the addition of analytes, respectively, and the excitation wavelengths of TPP‐p‐TPE, TPP‐m‐TPE, and TPP‐o‐TPE are 350, 336, and 332 nm, respectively.

Although the quenching efficiency of TPP‐m‐TPE is larger than that of TPP‐o‐TPE at low concentrations, it displays the same quenching extent at high concentrations. As the TPP‐o‐TPE is an easy‐to‐form pore in the aggregates, the sensor will encapsulate more Ru³⁺ into the nanoaggregates, which will enhance their interactions. On the other hand, the presence of 3D voids in the probes may help Ru³⁺ to diffuse and contact more fluorogens to annihilate the emission. Both factors enable the TPP‐o‐TPE to possess similar quenching extent with TPP‐m‐TPE, while the plateaus of the former will be reached at a higher Ru³⁺ concentration. However, the isomerization effect has less influence on the selectivity of the sensors.

For examples, by addition of other metal ions like Ag⁺, Fe³⁺, and Cu²⁺, etc., no fluorescent response can be observed.