English

• The traditional antibiotic medications for bacterial infections have confronted increasing challenges with the emergence and widespread of multidrug-resistant bacteria, posing a serious threat to global human health and led to millions of deaths worldwide [1-5]. In particular, infectious caused by Gram-negative (G(−)) bacteria are extremely difficult to treat compared to Gram-positive (G(+)) bacterial infection resulted from their natural outer membrane, highly selective porins, and abundant efflux pumps, which confers intrinsically resistance to G(−) bacteria to many antibiotics or other antibacterial materials [6]. Therefore, this poses a serious challenge to the existing antibiotics and an urgent need to develop new antibiotics to defend against G(−) bacteria-associated infections. However, the development of new antibiotics has far behind the outbreak of antibiotic-resistant bacteria. Thus, it is of great significance to explore novel antibacterial strategies to cope with antibiotic resistance. Among them, photodynamic therapy (PDT) provides a promising bactericidal approach that is not prone to resistance, which employs the photosensitizer (PS) to generate reactive oxygen species (ROS), especially singlet oxygen (¹O₂), upon exposure to light at specific wavelength [7-9]. As the most important component in PDT, a large number of PSs have been extensively exploited and applied in recent years. Nevertheless, most of these PSs suffer from insufficient ¹O₂ generation and fluorescence quenching in the aggregated state or a hydrophilic physiological environment, which severely limits their applications in the imaging-guided PDT. In this context, PSs with aggregation-induced emission (AIE) features have been recently emerged [10-15]. As expected, these AIE PSs showed strong fluorescence and efficient ¹O₂ generation in the aggregated state owing to the minimized nonradiative decay (NR) via restriction of intramolecular motion (Fig. 1a). However, the lower molar extinction coefficients in the visible region and the shorter excitation wavelength for most AIE PSs still need to be addressed for their further applications because they have the narrow light-harvesting ability.

  Figure 1

  

  Figure 1.  (a) Simplified Jablonski diagram of singlet oxygen generation for AIE photosensitizer. (b) Simplified Jablonski diagram of singlet oxygen generation for AIE photosensitizer based on BF₂-curcuminoid. (c) Chemical structures of AIE photosensitizer based on BF₂-curcuminoid (TBBC, TBC and TBBC-C8). (d) The calculated spatial distributions of HOMO and LUMO of TBBC (left) and TBC (right). A: absorption, NR: nonradiative decay, FL: fluorescence, S₀: ground state, S₁: excited singlet state, T₁: excited triplet state, ³O₂: normal oxygen, ¹O₂: singlet oxygen, ISC: intersystem crossing.

  DownLoad: Full-Size Img PowerPoint

  Fortunately, a class of highly luminescent organoboron complexes with large molar extinction coefficients, i.e., difluoroboron β-diketonate (BF₂bdk) complexes [16, 17], may provide a possible solution to the above issues. Recently, the BF₂bdk complexes have shown extensive applications in bioimaging, organic electronics, and photovoltaics due to their rich photophysical properties. However, a common limitation of these complexes is their tendency to exhibit shorter absorption and emission wavelengths, constraining their broader application. To address this, researchers have explored the integration of BF₂ into curcumin analogues, creating BF₂-curcuminoid complexes. This modification not only shifts the absorption and emission wavelengths towards the red end of the spectrum but also increases molecular rigidity and enhances photostability. BF₂-curcuminoid complexes have been utilized in various technologies, including near-infrared (NIR) fluorescent probes [18-23], photoacoustic imaging [24-28], NIR thermally activated delayed fluorescence (TADF) photosensitizers [29], TADF switches [30], and NIR TADF organic light-emitting diodes (OLEDs) [31, 32]. This innovation in BF₂-curcuminoid design is particularly inspiring in the context of AIE PSs. By incorporating the strong electron-withdrawing BF₂bdk group into tetraphenylethene (TPE)-type AIE PSs, it is anticipated that this approach will yield novel AIE PSs that are excited by longer wavelength visible light and possess large molar extinction coefficients. Additionally, this strategy is expected to endow BF₂-curcuminoid complexes with unique AIE characteristics, enhancing their potential for broader applications.

  To validate this hypothesis, we have rationally designed and prepared two BF₂-curcuminoid-based AIE PSs (TBBC and TBC) (Figs. 1b and c), in which methoxy-substituted TPE (MTPE) groups serve as both electron donor (D) and AIE active moieties, BF₂bdk group functions as electron acceptor (A), and styrene (or ethylene) group as π-bridge in this D-π-A-π-D system, respectively. For contrast, the bulky C8 alkyl chains-substituted analogue (TBBC-C8) was also designed and synthesized to verify the speculation about the mechanism of admirable antibacterial action towards E. coli. As expected, the as-prepared TBBC, TBC and TBBC-C8 presented solvent-dependent photophysical properties with large molar extinction coefficients in solutions, and excellent AIE properties in mixtures of toluene/DMSO. Remarkably, TBBC was particularly effective in generating singlet oxygen (¹O₂) when exposed to green light at 530 nm. It also showed a notable capability for selective staining and photodynamic inactivation of E. coli, indicating its potential as an innovative antibacterial PDT agent in the fight against drug-resistant bacterial infections.

  In such photosensitive system, the molecular design principles for incorporating BF₂bdk group into the AIE PSs are primarily focused on the following aspects (Fig. 1): (I) Enhancing molar extinction coefficients in favor of light-harvesting ability of AIE PSs; (II) Extending excitation wavelength to achieve long wavelength visible light excitation of AIE PSs; (III) Promoting the intersystem crossing (ISC) because active nonbonding p electrons on the BF₂bdk group are devoted to enhancing the spin-orbital coupling (SOC); (IV) Separating the highest occupied molecular orbital (HOMO)-lowest unoccupied molecular orbital (LUMO) distribution to reduce ΔE_ST (E_S1 − E_T1) for higher ISC [33], which is achieved by inserting a benzene ring between the donor and acceptor for TBBC to fine-tune their distance and torsional angle compared to the analogue TBC. As shown in Fig. 1d, the HOMO distribution of TBBC is positioned at two MTPE groups and LUMO is mainly fixed at central BF₂-curcuminoid skeleton. This indicates an efficient separation of HOMO and LUMO, thus resulting in a small ΔE_ST value (0.360 eV). For contrast, another AIE PS (TBC) without benzene ring bridging is designed and synthesized, which shows a much bigger ΔE_ST (0.429 eV) compared to TBBC resulting from its partially overlapped HOMO and LUMO (Fig. 1d and Table S1 in Supporting information). In addition, the bulky C8-substituted analogue (TBBC-C8) of TBBC is also obtained to verify the speculation about the mechanism of admirable antibacterial action towards E. coli. Unsurprisingly, no appreciable differences in the HOMO and LUMO orbital distribution and its ΔE_ST (0.349 eV) for TBBC-C8 are observed compared with TBBC (Fig. S1 and Table S1 in Supporting information), further implying that the incorporation of C8 alkyl chain may have insignificant influence on the generation of singlet oxygen. As depicted in Scheme S1 (Supporting information), TBBC and TBC were prepared by the Knoevenagel condensation reaction between MTPE-CHO (2 or 3) and BF₂ complex of pentane-2, 4-dione (4) using n-BuNH₂ as the catalysis in anhydrous toluene in the yield of 48% and 57%, respectively. A similar approach was used to synthesize TBBC-C8 in the yield of 36%. Their chemical structures were well characterized by ¹H NMR, ¹³C NMR and high-resolution mass spectrometry (HRMS) (Figs. S25–S39 in Supporting information). From ¹H NMR spectrum of TBBC, the coupling constants of double bonds were ca. 15.5 Hz, suggesting a stable trans-configuration for both ethylene bonds in TBBC.

  With these curcuminoid-BF₂ complexes in hand, their photophysical properties were investigated in solvents with different polarity. As depicted in Fig. 2a and Table S2 (Supporting information), TBBC clearly showed a maximum absorption band ranging from 498 nm to 508 nm in four various solvents, which is attributable to intramolecular charge transfer (ICT) transition from TPE groups to the central difluoroboron β-diketonate moiety [34]. The perceptible redshift was observed with increasing solvent polarity. More importantly, TBBC showed high extinction coefficients (ε = 6.31 × 10⁴–7.02 × 10⁴ L mol⁻¹ cm⁻¹) in the visible region, which indicates the outstanding visible light-harvesting capability for TBBC. Compared with TBBC, the maximum absorption wavelength of TBC revealed an obvious redshift (Δλ = ca. 20 nm) in the same solvent (Figs. S2, S4–S6 and Table S2 in Supporting information), which is attributed to increased π-conjugation for TBC without the benzene ring acting as a bridge linker. As expected, TBBC-C8 bearing four alkyl chains displayed negligible changes in ultraviolet (UV) absorption because alkyl chains have no obvious effect on the conjugated system (Figs. S3–S6 and Table S2 in Supporting information). As illustrated in Fig. 2c, Figs. S7–S9 and Table S2 (Supporting information), TBBC presented an intense orange fluorescence with high fluorescence quantum yield (Ф_F = 0.23) in weak polar toluene (poor solvent), while relatively inefficient emission with lower Ф_F (from 0.06 to 0.008) was detected in the larger polar solvents (good solvents: CH₂Cl₂, CHCl₃ and DMSO). Comparatively speaking, TBC emitted a strong red fluorescence and showed red-shifted changes (Δλ_em = 24–51 nm) in fluorescence emission in solvents with different polarity (Table S2). In addition, TBBC-C8 exhibited a similar fluorescence performance to TBBC under the same conditions (Fig. S10 and Table S2 in Supporting information).

  Figure 2

  

  Figure 2.  (a) The UV–vis absorption spectra of TBBC in different solvents (2.0 × 10⁻⁵ mol/L). (b) The UV–vis absorption spectra of TBBC, TBC and TBBC-C8 in toluene (2.0 × 10⁻⁵ mol/L). (c) The fluorescence spectra of TBBC and TBC in toluene (2.0 × 10⁻⁵ mol/L) (λ_ex = 498 nm and 517 nm, respectively). (d, e) The fluorescence spectra of TBBC (d) and TBC (e) in different mixtures of toluene/DMSO (v/v) (2.0 × 10⁻⁵ mol/L) (λ_ex = 500 nm and 520 nm, respectively). Insert: their corresponding photographs in 0% and 90% toluene solutions under 365 nm UV light. (f) The dependence of the emission intensity on the toluene fraction (f_t). (g) The fluorescence spectra of TBBC-C8 in different mixtures of toluene/THF (v/v) (2.0 × 10⁻⁵ mol/L) (λ_ex = 500 nm). Insert: their corresponding photographs in 0% and 90% toluene solutions under 365 nm UV light. (h) The dependence of the emission intensity on the toluene fraction (f_t). (i) Time-resolved decay profiles of TBBC, TBC and TBBC-C8 in the aggregated state.

  DownLoad: Full-Size Img PowerPoint

  Whereafter, their AIE performance was evaluated in the mixed solvents (toluene/DMSO). As depicted in Figs. 2d and f, TBBC showed a weak emission in pure DMSO. As toluene fraction (f) increased from 0% to 20%, negligible fluorescence change was detected. When f was greater than 20%, the emission peak at 562 nm (f = 30%) was gradually red-shifted to 570 nm (f = 90%), and its emission intensity significantly increased and reached the maximum value at f = 90% due to the formation of the aggregates, which was accompanied by the appearance of an intense orange fluorescence (insert in Fig. 2d). Furthermore, a much higher quantum yield (Ф_f = 0.19) for TBBC was obtained in 90% water mixed solutions compared with that in pure DMSO (Ф_f = 0.008), as well as fluorescence lifetimes ranging from 0 ns to 0.55 ns (Fig. 2i). Moreover, similar AIE properties for TBC (λ_em = 624 nm, Ф_f = 0.27, τ = 0.36 ns) and TBBC-C8 (λ_em = 608 nm, Ф_f = 0.30, τ = 1.56 ns) were observed with the increase of f from 0% to 90% (Figs. 2e–i and Fig. S11 in Supporting information). Therefore, these results clearly implied that the incorporation of curcuminoid-BF₂ did not affect the AIE characteristics of OCH₃- and OC₈H₁₇-substituted TPE moieties.

  In consideration of their excellent luminescent properties in the aggregated state, the bacteria imaging capabilities for these BF₂-curcuminoids were further explored. Firstly, the hydrophobic TBBC was made into water-soluble nanoparticles (abbreviated as TBBC NPs) by encapsulation of the amphiphilic Pluronic F-127, as illustrated in Fig. 3a. The dynamic light scattering (DLS) experiment showed that the average hydrodynamic diameter of the resulting nanoparticles (TBBC NPs) was determined to be ca. 109 nm (Fig. 3c), which favors efficient accumulation at the bacterial sites for fluorescence imaging and antibacterial PDT. Subsequently, TBC NPs (average size = 83 nm) and TBBC-C8 NPs (average size = 122 nm) were also successfully fabricated through the same bottom-up approach using F-127 as the doping matrix (Figs. S12 and S13 in Supporting information).

  Figure 3

  

  Figure 3.  (a) Schematic representation of the preparation of the nanoparticles by using an amphiphilic Pluronic F-127 as the doping matrix (TBBC NPs, TBC NPs and TBBC-C8 NPs). (b) The absorption and emission spectra of TBBC NPs in water (2.0 × 10⁻⁵ mol/L, λ_ex = 488 nm). (c) DLS profiles of TBBC NPs. (d) Fluorescence and merged images of bacteria (S. aureus and E. coli) incubated with TBBC NPs for 20 min.

  DownLoad: Full-Size Img PowerPoint

  The shape and size of these prepared nanoparticles were confirmed by transmission electron microscopy (TEM), manifesting uniform spheres with the different size for TBC NPs, TBBC NPs, and TBBC-C8 NPs (Fig. S14 in Supporting information). More importantly, the as-prepared TBBC NPs displayed a broad visible light absorption ranging from 390 nm to 675 nm (Fig. 3b). The aqueous suspension of TBBC NPs emitted a bright red fluorescence (λ_em = 640 nm, Ф_f = 0.08) under irradiation with 365 nm UV light (Fig. 3b and Table S2), which is beneficial for fluorescence imaging in biomedical fields. In comparison, TBC NPs showed a red-shifted fluorescence emission (λ_em = 668 nm, Ф_f = 0.05) in aqueous solution (Fig. S15 and Table S2 in Supporting information) due to the extended conjugated system compared with TBBC. Besides, TBBC-C8 NPs exhibited a similar maximum emission wavelength at 642 nm in water, as well as a stronger red fluorescence emission than TBBC NPs (Fig. S16 and Table S2 in Supporting information), suggesting incorporation of C8 alkyl chains can prevent the intermolecular stacking to enhance the AIE effect.

  Encouraged by the remarkable AIE performance and red emission in nanoaqueous solution, we selected Gram-positive S. aureus and Gram-negative E. coli as representatives to preliminarily assess the bacteria staining capabilities of these BF₂-curcuminoids. As depicted in Fig. 3d, when two kinds of bacteria were incubated with TBBC NPs, almost no fluorescence signal was detected for S. aureus, while the distinct fluorescence signal was accidentally visualized in E. coli. To the best of our knowledge, most previously reported AIE photosensitizer molecules exhibited selective imaging for Gram-positive bacteria through hydrophobic and electrostatic interactions [35-38], which is attributable to the distinctive multilayer outer membrane structures possessed by Gram-negative bacteria [39]. Additionally, the poring channels in the outer envelope of Gram-negative bacteria can expel small dye molecules. Therefore, the experimental result that TBBC NPs displayed a selective bacteria imaging for Gram-negative E. coli was very surprising to us. However, when both S. aureus and E. coli were incubated with TBC NPs and TBBC-C8 NPs under the same conditions, no fluorescence signals were detected in neither S. aureus nor E. coli, which may be attributed to the fact that the smaller TBC and larger TBBC-C8 do not readily bind to the bacterial cell membrane. Consequently, these bacterial staining results implied that TBBC as a novel AIE fluorescent dyes has great potential for application in discriminating between Gram-positive and Gram-negative bacteria.

  To speculate on the reason for the differences in their ability to target S. aureus and E. coli, zeta potentials (ζ) usually reflecting the electrical charge on the bacterial surface were subsequently determined to investigate interactions between the bacteria and these BF₂-curcuminoids. As illustrated in Fig. S17 (Supporting information), after S. aureus was incubated with TBBC, TBC and TBBC-C8, the corresponding ζ potentials of S. aureus exhibited an obvious negative shift in contrast with that of S. aureus alone, which implies no electrostatic interactions between these BF₂-curcuminoids and the negatively charged bacterial surface. Thus, no fluorescence signals for TBBC, TBC and TBBC-C8 were detected in S. aureus. Moreover, an apparent positive shift (Δζ = 4.44 eV) for the ζ potential of E. coli when incubated with TBBC compared with that of E. coli alone, which can be attributed to the electrostatic interactions between the negatively charged bacterial surface and the AIE PS TBBC, thus leading to neutralization of the negative charge on the surface of E. coli. However, the ζ potentials of E. coli displayed a negligible change after incubated with smaller TBC and larger TBBC-C8 with C8 alkyl chains compared to E. coli alone. This may be because TBC can effectively insert into the outer membrane on E. coli, thereby the positive charge was concealed inside the bacteria instead of being exposed to the surface [40, 41]. Therefore, no effect on the potentials of E. coli was detected for smaller TBC. In contrast, TBBC-C8 with C8 alkyl chains may be hindered by their interaction with bacteria due to larger steric hindrance [42, 43], thus attenuating the bacteria imaging capability of TBBC-C8. Accordingly, zeta potentials results indicated that a significant impact of the size of the AIE PS molecules on the selective bacteria imaging.

  Encouraged by the small ΔE_ST value of these BF₂-curcuminoids (TBBC, TBC and TBBC-C8), we further evaluated their ¹O₂ generation ability in water under irradiation with green light (λ = 530 nm). The commercial 9, 10-anthracenediyl-bis-(methylene)-dimalonic acid (ABDA) was utilized as an ¹O₂ indicator to monitor the process of ¹O₂ generation. As shown in Fig. 4a, upon irradiation with 530 nm green light (8.9 mW/cm²), the absorbance intensity of ABDA gradually weakened in the presence of TBBC NPs, which is resulted from the decomposition by the increasing ¹O₂ generation (Scheme S2 in Supporting information). As illustrated in Figs. S18 and S19 (Supporting information), the presence of the other two BF₂-curcuminoids NPs (TBC NPs and TBBC-C8 NPs) also caused different degrees of decrease in the absorption intensity of ABDA, while that of ABDA alone changed slightly (Fig. S20 in Supporting information). Moreover, under the treatment of these BF₂-curcuminoids NPs (TBBC NPs, TBC NPs and TBBC-C8 NPs), the higher decomposition rate of ABDA was detected compared with that of RB by monitoring the attenuation of absorption at 378 nm (Fig. 4b and Fig. S21 in Supporting information), indicating the superior ¹O₂ generation performance of these BF₂-curcuminoids under 530 nm green light irradiation. In addition, the absorbance of ABDA was decreased by 57.0% in the presence of TBBC NPs upon irradiation with 530 nm green light for 80 s, suggesting that 7.13 µmol of ABDA was consumed per 20 s when 5 µmol/L of TBBC NPs was exposed to 530 nm green light. In contrast, 1.07 µmol and 4.45 µmol of ABDA were depleted per 20 s for 5 µmol/L of TBC NPs and TBBC-C8 NPs under the same irradiation conditions, respectively. It was demonstrated that the ¹O₂ generation efficiency of TBBC NPs was superior to that of TBC NPs, which is mainly attributed to the smaller ΔE_ST value for TBBC compared with TBC. Unexpectedly, ¹O₂ generation performance for TBBC-C8 with a smaller ΔE_ST was inferior to TBBC probably due to the influence of the hydrophobic C8 alkyl chains. Subsequently, the ¹O₂ quantum yields of these BF₂-curcuminoids (TBBC, TBC and TBBC-C8) NPs in water were determined as 19%, 6%, 11% by using RB as the standard reference [44], respectively. Besides, it was noteworthy that TBBC NPs and TBC NPs displayed an excellent photostability in comparison with the commercial PS (RB) when exposed to green light probably because they were not easily oxidized by ¹O₂, as depicted in Fig. 4a and Figs. S18 and S21, whereas TBBC-C8 NPs showed a relatively poor photostability (Fig. S19), which may be attributed to the perturbation of the flexible C8 alkyl chains on the nanoparticles under green-light irradiation.

  Figure 4

  

  Figure 4.  The absorption spectra of ABDA in the presence of TBBC NPs under 530 nm green light irradiation (a). The decomposition rates of ABDA in the absence and presence of RB, TBBC NPs and TBC NPs under 530 nm green light irradiation (b), where A₀ and A are the absorbance of ABDA at 378 nm, [RB] = [TBBC NPs] = [TBC NPs] = [TBBC-C8 NPs] = 5 × 10⁻⁶ mol/L, [ABDA] = 5 × 10⁻⁵ mol/L. Photographs of the agar plates (c) and CFU reduction (d) of S. aureus and E. coli with/without TBBC NPs (2 × 10⁻⁶ mol/L) and green light (8.9 mW/cm²) treatment (e). Error bars: mean ± SD (n = 3). The SEM images of E. coli treated with or without TBBC NPs (2 × 10⁻⁶ mol/L) and green light irradiation (8.9 mW/cm²).

  DownLoad: Full-Size Img PowerPoint

  Inspired by the outstanding performance of ¹O₂ generation and bacterial discrimination of the presented BF₂-curcuminoids, we further evaluated their photodynamic antibacterial activity by the classical plate colony-counting method [45, 46]. Gram-positive S. aureus and Gram-negative E. coli were utilized as representative bacteria to probe their in vitro photodynamic antimicrobial properties under green light (8.9 mW/cm²) irradiation, respectively. From Fig. 4c, both S. aureus and E. coli grew and reproduced smoothly on the agar plates in dark or green-light irradiation conditions when not treated with TBBC NPs. When both bacterial strains were treated with TBBC NPs and then stored in the dark, the colony-forming unit (CFU) of S. aureus and E. coli decreased by approximately 8% and 6% (Fig. 4d), respectively, implying its inconspicuous dark toxicity toward S. aureus and E. coli. In the presence of both TBBC NPs (2.0 µmol/L) and green light irradiation, E. coli was killed effectively with a nearly 100% CFU reduction rate of (Figs. 4c and d), while the photodynamic antimicrobial effect on S. aureus was negligible, and only ca. 9% CFU reduction rate was detected in comparison with the control groups. The photodynamic antimicrobial results thus indicated that TBBC NPs can eliminate selectively E. coli over S. aureus via antibacterial PDT. Subsequently, photodynamic antibacterial effects of TBBC NPs have been tested at two lower concentrations (i.e., 1.0, 0.5 µmol/L). As shown in Fig. S22 (Supporting information), under the treatment of green light irradiation, about 60% of E. coli were killed at the concentration of 1.0 µmol/L, and almost 100% of E. coli survived when the concentration of TBBC NPs was decreased 0.5 µmol/L. Whereafter, scanning electron microscopy (SEM) was further employed to explore the detailed morphological changes of E. coli upon treatment with TBBC NPs and green light irradiation. As illustrated in Fig. 4e, the morphology of E. coli remained intact with smooth bodies when treated with only green light, while the treatment with TBBC NPs made the originally smooth bodies of E. coli become partially wrinkled probably owing to strong electrostatic interactions between TBBC NPs and E. coli. After E. coli was treated with both TBBC NPs and green light, the bacterial shape changed dramatically along with the fusion and shrinkage of cell walls, which definitely provided evidence that ¹O₂ generated by TBBC NPs exerted toxicity to E. coli. These SEM results also revealed that the destruction of bacterial cell might be the main cause of the death of E. coli. By comparison, TBC NPs showed only almost 90% bactericidal rate against E. coli and an unconspicuous antimicrobial effect on S. aureus (Fig. S23 in Supporting information), suggesting a similar selective photodynamic antimicrobial activity to TBBC NPs. However, TBBC-C8 NPs with four C8 alkyl chains hardly presented any photodynamic antimicrobial activity against E. coli and S. aureus (Fig. S24 in Supporting information). According to the previous reports [47-49], for TBBC, TBC and TBBC-C8, reasons for the discrepancy in photodynamic antimicrobial activity against E. coli may be attributed to the difference in molecular size of these AIE PSs. As was mentioned above, the poring channels in the outer envelope of Gram-negative bacteria (E. coli) can expel small dye molecules. Therefore, for the smaller-size TBC which showed partial photodynamic antimicrobial activity against E. coli, some of the PS molecules (TBC) may be pumped by the poring channels in the outer envelope of E. coli after exerting the photodynamic antibacterial effect to some extent, thus resulting in ca. 90% bactericidal rate against E. coli. As for C8 alkyl chains-functionalized TBBC-C8, the outer envelope of E. coli may not be crossed due to its larger size, leading to almost no photodynamic antimicrobial activity against E. coli.

  To summarize, we successfully prepared three D-π-A-π-D type AIE photosensitizers (TBBC, TBC and TBBC-C8) based on BF₂-curcuminoid with yields ranging from 36% to 57%. These molecules incorporated substituted TPE groups as electron donors and AIE active moieties, BF₂bdk group as an electron acceptor, and a styrene (or ethylene) group as a π-bridge. Initial assessment of their photophysical properties revealed solvent-dependent behavior with large molar extinction coefficients in solvents of varying polarity and excellent AIE properties in mixed solvents, indicating that the inclusion of curcuminoid-BF₂ moieties did not compromise their AIE performance. Following this, the hydrophobic BF₂-curcuminoids were assembled into water-soluble nanoparticles (TBBC NPs, TBC NPs, and TBBC-C8 NPs) with average hydrodynamic diameters ranging from 83 nm to 122 nm, emitting bright red fluorescence (640–668 nm) under 365 nm UV light. Surprisingly, bacterial staining experiments revealed that TBBC NPs selectively imaged Gram-negative E. coli, while no fluorescence signals were detected for TBC NPs and TBBC-C8 NPs in either S. aureus or E. coli. Zeta potential (ζ) results indicated a significant impact of the AIE PS molecule size on selective bacteria imaging. Moreover, TBBC demonstrated efficient singlet oxygen generation when exposed to green light at 530 nm (8.9 mW/cm²), attributed to its smaller singlet-triplet energy gap (ΔE_ST). Notably, it was noteworthy that TBBC NPs and TBC NPs displayed an excellent photostability compared with the RB under green-light irradiation, whereas TBBC-C8 NPs showed a relatively poor photostability. More importantly, TBBC showed effective in vitro selective photodynamic killing of E. coli, likely due to its appropriate molecular size compared to TBC and TBBC-C8. In brief, TBBC holds great potential as a novel green light-excited AIE photosensitizer for discriminating and selectively sterilizing Gram-positive and Gram-negative bacteria.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Ziyong Li: Writing – review & editing, Writing – original draft, Funding acquisition, Formal analysis, Data curation. Jinzhao Song: Data curation. Xinyu Gao: Data curation. Xiaoxie Ma: Formal analysis, Data curation. Keyu Liu: Data curation. Ziwei Ma: Data curation. Qilian Wang: Data curation. Xinliang Zeng: Data curation. Haining Zhang: Data curation. Pei Zhang: Data curation. Hui Guo: Writing – review & editing, Writing – original draft, Supervision. Jun Yin: Writing – review & editing, Writing – original draft, Supervision.

  Acknowledgments

  The authors acknowledge financial support from National Natural Science Foundation of China (No. 32101150), Key Scientific Research Project of Higher Education of Henan Province (No. 22A430007), Natural Science Foundation of Henan Province (No. 222300420501), the Science and Technology Project of Henan Province (No. 242102230119), and Innovation and Entrepreneurship Training Program for College students in China (No. 202310482001).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110073.

  
  
• 1. [1]

     E. Svoboda, Nature 586 (2020) S58. doi: 10.1038/d41586-020-02887-0

  2. [2]

     X. Li, H. Bai, Y. Yang, et al., Adv. Mater. 31 (2019) 1805092. doi: 10.1002/adma.201805092

  3. [3]

     J.A. Bartell, L.M. Sommer, J.A. Haagensen, et al., Nat. Commun. 10 (2019) 629. doi: 10.1038/s41467-019-08504-7

  4. [4]

     E.D. Brown, G.D. Wright, Nature 529 (2016) 336-343. doi: 10.1038/nature17042

  5. [5]

     S.B. Levy, B. Marshall, Nat. Med. 10 (2004) S122-S129. doi: 10.1038/nm1145

  6. [6]

     A. Luther, M. Urfer, M. Zahn, et al., Nature 576 (2019) 452-458. doi: 10.1038/s41586-019-1665-6

  7. [7]

     J. Zhao, W. Wu, J. Sun, S. Guo, Chem. Soc. Rev. 42 (2013) 5323-5351. doi: 10.1039/c3cs35531d

  8. [8]

     S.S. Lucky, K.C. Soo, Y. Zhang, Chem. Rev. 115 (2015) 1990-2042. doi: 10.1021/cr5004198

  9. [9]

     X. Zhao, J. Liu, J. Fan, et al., Chem. Soc. Rev. 50 (2021) 4185-4219. doi: 10.1039/d0cs00173b

  10. [10]

      X. He, L.H. Xiong, Z. Zhao, et al., Theranostics. 9 (2019) 3223. doi: 10.7150/thno.31844

  11. [11]

      D. Wang, M.M.S. Lee, W. Xu, et al., Theranostics 8 (2018) 4925. doi: 10.7150/thno.27787

  12. [12]

      Z. Li, S. Chen, Y. Huang, et al., Chem. Eng. J. 450 (2022) 138087. doi: 10.1016/j.cej.2022.138087

  13. [13]

      B. Xue, X. Geng, H. Cui, et al., Chin. Chem. Lett. 34 (2023) 108140. doi: 10.1016/j.cclet.2023.108140

  14. [14]

      Y. Jiang, Z. Zeng, J. Yao, et al., Chin. Chem. Lett. 34 (2023) 107966. doi: 10.1016/j.cclet.2022.107966

  15. [15]

      L. Huang, D. Qing, S. Zhao, et al., Chem. Eng. J. 430 (2022) 132638. doi: 10.1016/j.cej.2021.132638

  16. [16]

      P.Z. Chen, L.Y. Niu, Y.Z. Chen, Q.Z. Yang, Coord. Chem. Rev. 350 (2017) 196-216. doi: 10.1016/j.ccr.2017.06.026

  17. [17]

      X. Zeng, X. Ma, J. Dong, et al., Angew. Chem. 135 (2023) e202312618. doi: 10.1002/ange.202312618

  18. [18]

      P. Zhang, Z.Q. Guo, C.X. Yan, W. -H. Zhu, Chin. Chem. Lett. 28 (2017) 1952-1956. doi: 10.1016/j.cclet.2017.08.038

  19. [19]

      S. Shen, W. Xu, J. Lu, et al., Chin. Chem. Lett. 35 (2024) 108360. doi: 10.1016/j.cclet.2023.108360

  20. [20]

      X. Liu, Adv. Agrochem 2 (2023) 1-2. doi: 10.1016/j.aac.2023.02.004

  21. [21]

      Z. Li, S. Hou, H. Zhang, Q. et al., Adv. Agrochem 2 (2023) 79-87. doi: 10.1016/j.aac.2023.02.001

  22. [22]

      D. Li, T. Shen, X. Xue, et al., Sci. Chin. Chem. 66 (2023) 2329-2338. doi: 10.1007/s11426-023-1661-6

  23. [23]

      X. Ma, Y. Huang, W. Chen, et al., Angew. Chem. Int. Ed. 62 (2023) e202216109. doi: 10.1002/anie.202216109

  24. [24]

      S. Zeng, X. Liu, Y.S. Kafuti, et al., Chem. Soc. Rev. 52 (2023) 5607-5651. doi: 10.1039/d2cs00799a

  25. [25]

      X. Ma, Y. Huang, A. Li, et al., ChemMedChem 18 (2023) e202300204. doi: 10.1002/cmdc.202300204

  26. [26]

      J. Zhang, C. Yue, Y. Ke, H. Qu, L. Zeng, Adv. Agrochem 2 (2023) 127-141. doi: 10.3390/wevj14050127

  27. [27]

      Y. Liao, Y. Liang, Y. Huang, et al., Chin. Chem. Lett. 35 (2024) 109092. doi: 10.1016/j.cclet.2023.109092

  28. [28]

      X. Zeng, Y. Huang, J. Dong, et al., Yang, Adv. Agrochem 1 (2022) 73-84. doi: 10.1016/j.aac.2022.08.001

  29. [29]

      Z. Li, X. -G. Yang, H. Zhang, et al., Inorg. Chem. Front. 9 (2022) 4281-4287. doi: 10.1039/d2qi01112c

  30. [30]

      Z. Li, J. -R. Zhang, X. -K. Tian, et al., Chem. Sci. 13 (2022) 9381-9386. doi: 10.1039/d2sc02662g

  31. [31]

      D. -H. Kim, A. D'aléo, X. -K. Chen, et al., Nat. Photonics 12 (2018) 98-104. doi: 10.1038/s41566-017-0087-y

  32. [32]

      H. Ye, D.H. Kim, X. Chen, et al., Chem. Mater. 30 (2018) 6702-6710. doi: 10.1021/acs.chemmater.8b02247

  33. [33]

      S. Xu, W. Wu, X. Cai, et al., Chem. Commun. 53 (2017) 8727-8730. doi: 10.1039/C7CC04864E

  34. [34]

      X. Ding, Q. Wang, D. Chen, et al., Adv. Agrochem 2 (2023) 364-370. doi: 10.1016/j.aac.2023.08.002

  35. [35]

      Z. Zhang, M. Kang, H. Tan, et al., Chem. Soc. Rev. 51 (2022) 1983-2030. doi: 10.1039/d1cs01138c

  36. [36]

      Z. Li, X. Zeng, C. Gao, et al., Coord. Chem. Rev. 497 (2023) 215451. doi: 10.1016/j.ccr.2023.215451

  37. [37]

      H. Zhang, C. He, L. Shen, et al., Chin. Chem. Lett. 34 (2023) 108160. doi: 10.1016/j.cclet.2023.108160

  38. [38]

      Z. Li, X. Gao, H. Zhang, et al., Chin. Chem. Lett. 34 (2023) 107645. doi: 10.1016/j.cclet.2022.06.068

  39. [39]

      Y. Huang, W. Chen, J. Chung, et al., Chem. Soc. Rev. 50 (2021) 7725-7744. doi: 10.1039/d0cs01340d

  40. [40]

      H. Yuan, Z. Liu, L. Liu, et al., Adv. Mater. 26 (2014) 4333-4338. doi: 10.1002/adma.201400636

  41. [41]

      H. Bai, H. Yuan, C. Nie, et al., Angew. Chem. Int. Ed. 54 (2015) 13208-13213. doi: 10.1002/anie.201504566

  42. [42]

      Y. Wang, T.S. Corbitt, S.D. Jett, et al., Langmuir 28 (2012) 65-70. doi: 10.1021/la2044569

  43. [43]

      H. Cai, X. Wu, L. Jiang, et al., Chin. Chem. Lett. 35 (2024) 108946. doi: 10.1016/j.cclet.2023.108946

  44. [44]

      J. Ge, M. Lan, B. Zhou, et al., Nat. Commun. 5 (2014) 4596. doi: 10.1038/ncomms5596

  45. [45]

      A. Kaprelyants, D. Kell, J. Appl. Bacteriol. 72 (1992) 410-422. doi: 10.1111/j.1365-2672.1992.tb01854.x

  46. [46]

      M. Dong, R. Tang, J. Li, et al., Chin. Chem. Lett. 35 (2024) 108539. doi: 10.1016/j.cclet.2023.108539

  47. [47]

      K.R. Raghupathi, R.T. Koodali, A.C. Manna, Langmuir 27 (2011) 4020-4028. doi: 10.1021/la104825u

  48. [48]

      F. Perreault, A.F. De Faria, S. Nejati, M. Elimelech, ACS Nano 9 (2015) 7226-7236. doi: 10.1021/acsnano.5b02067

  49. [49]

      L. Zheng, J. Li, M. Yu, et al., J. Am. Chem. Soc. 142 (2020) 20257-20269. doi: 10.1021/jacs.0c10771

• 

  Figure 1  (a) Simplified Jablonski diagram of singlet oxygen generation for AIE photosensitizer. (b) Simplified Jablonski diagram of singlet oxygen generation for AIE photosensitizer based on BF₂-curcuminoid. (c) Chemical structures of AIE photosensitizer based on BF₂-curcuminoid (TBBC, TBC and TBBC-C8). (d) The calculated spatial distributions of HOMO and LUMO of TBBC (left) and TBC (right). A: absorption, NR: nonradiative decay, FL: fluorescence, S₀: ground state, S₁: excited singlet state, T₁: excited triplet state, ³O₂: normal oxygen, ¹O₂: singlet oxygen, ISC: intersystem crossing.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) The UV–vis absorption spectra of TBBC in different solvents (2.0 × 10⁻⁵ mol/L). (b) The UV–vis absorption spectra of TBBC, TBC and TBBC-C8 in toluene (2.0 × 10⁻⁵ mol/L). (c) The fluorescence spectra of TBBC and TBC in toluene (2.0 × 10⁻⁵ mol/L) (λ_ex = 498 nm and 517 nm, respectively). (d, e) The fluorescence spectra of TBBC (d) and TBC (e) in different mixtures of toluene/DMSO (v/v) (2.0 × 10⁻⁵ mol/L) (λ_ex = 500 nm and 520 nm, respectively). Insert: their corresponding photographs in 0% and 90% toluene solutions under 365 nm UV light. (f) The dependence of the emission intensity on the toluene fraction (f_t). (g) The fluorescence spectra of TBBC-C8 in different mixtures of toluene/THF (v/v) (2.0 × 10⁻⁵ mol/L) (λ_ex = 500 nm). Insert: their corresponding photographs in 0% and 90% toluene solutions under 365 nm UV light. (h) The dependence of the emission intensity on the toluene fraction (f_t). (i) Time-resolved decay profiles of TBBC, TBC and TBBC-C8 in the aggregated state.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Schematic representation of the preparation of the nanoparticles by using an amphiphilic Pluronic F-127 as the doping matrix (TBBC NPs, TBC NPs and TBBC-C8 NPs). (b) The absorption and emission spectra of TBBC NPs in water (2.0 × 10⁻⁵ mol/L, λ_ex = 488 nm). (c) DLS profiles of TBBC NPs. (d) Fluorescence and merged images of bacteria (S. aureus and E. coli) incubated with TBBC NPs for 20 min.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  The absorption spectra of ABDA in the presence of TBBC NPs under 530 nm green light irradiation (a). The decomposition rates of ABDA in the absence and presence of RB, TBBC NPs and TBC NPs under 530 nm green light irradiation (b), where A₀ and A are the absorbance of ABDA at 378 nm, [RB] = [TBBC NPs] = [TBC NPs] = [TBBC-C8 NPs] = 5 × 10⁻⁶ mol/L, [ABDA] = 5 × 10⁻⁵ mol/L. Photographs of the agar plates (c) and CFU reduction (d) of S. aureus and E. coli with/without TBBC NPs (2 × 10⁻⁶ mol/L) and green light (8.9 mW/cm²) treatment (e). Error bars: mean ± SD (n = 3). The SEM images of E. coli treated with or without TBBC NPs (2 × 10⁻⁶ mol/L) and green light irradiation (8.9 mW/cm²).

  下载: 全尺寸图片 幻灯片
•