Results and discussion

Design and synthesis

Constructing molecular structures with an electron donor–acceptor (D–A) backbone is widely recognized as an effective approach to enhance the internal charge transfer (ICT) effect of fluorescent molecules and thereby extending their emission wavelengths (Ersoy and Henary, 2025; Essam et al., 2021; Jaswal and Kumar, 2020; İçli et al., 2010). The rational selection and systematic integration of D-A units are critical for successful design of D-A fluorophores with high luminescence yield and red-shifted emission wavelength. Imidazole, an electron-rich heterocyclic structure containing two nitrogen atoms, has been receiving increasing attention in different fields such as bio-labeling, fluorescent sensors, organic light emitting diodes (OLEDs) and light sources due to its favorable biological activity and photoelectric properties (Tolomeu and Fraga, 2023; Verma et al., 2013). Notably, it has a unique bipolar configuration of two donor and acceptor nitrogen atoms apart via C2. Introducing electron-withdrawing or neutral substituents at the C2 position enhances the electron-donating behavior of the nitrogen-containing heterocyclic system. Conversely, functionalizing the 2-position of the imidazole scaffold with an electron-donating moiety induces electron-withdrawing properties, thereby reversing the electronic polarization of the heterocycle (Al Sharif et al., 2023). Given these unique electronic characteristics and facile modification on the five-membered ring, imidazole derivatives also serve as ideal candidates for electron donors in D–A type fluorophores. Among numerous imidazole derivatives, triphenylimidazole has caught our attention due to its characteristic property: the intramolecular rotations of multiple benzene rotors on the imidazole ring are supposed to induce nonradiative decay, thereby quenching the fluorescence of the molecule in dilute solutions. In contrast, within the aggregated state, steric hindrance from molecular packing significantly restricts these rotations, leading to restored fluorescence intensity. This restriction of intramolecular rotation (RIR) effect is widely recognized as the primary contributor to the aggregation-induced emission (AIE) phenomenon (Leung et al., 2014). Built in nonplanar 1,4,5-triphenylimidazole core structure, we conceived a kind of AIE molecular platform, TPIT-X, with a classic D-A framework and further extended its conjugation length by incorporating thiophene as a π-bridge, forming a “D-π-A” type system. By conjugating a series of acceptor units with different electron affinity and steric hindrance, the emission wavelengths and AIE properties of TPIT-X could be tuned as desired. Based on this engineering strategy, TPIT-1 to TPIT-6 were conceived in this study. The triphenylimidazole moiety, functioning as the core structural unit in this molecular series, exhibits a bipolar electronic characteristic that enables its reversible switching between electron-donating and electron-withdrawing roles. In the TPIT-1 molecule, this heterocyclic system acts as a weak electron-withdrawing unit due to π-conjugation with electron-donating diphenylethylene, which delocalizes electron density. However, the introduction of electron-withdrawing groups such as cyanide, pyridyl, and pyridinium onto the conjugated vinyl group of the imidazole transforms the tricyclic core into an electron donor. All these molecules were synthesized via facile routes. Specifically, we synthesized the core donor structure TPIT-CHO according to previous reports, from which all other target products were obtained from a single-step reaction (Scheme 1). TPIT-1 was prepared by McMurry coupling reaction between TPIT-CHO and benzophenone. TPIT-2, TPIT-3, TPIT-4, TPIT-5, TPIT-6 were obtained through condensation reactions between TPIT-CHO and 2-phenylacetonitrile, 2-(pyridin-4-yl)acetonitrile, malononitrile, 4-benzyl-1-methyl-pyridinium iodide, 1-methyl-4-(prop-2-yn-1-yl)pyridin-1-ium iodide, respectively. The iodide anions (I−) in TPIT-5 and TPIT-6 were further replaced with hexafluorophosphate ions (PF6 −) using potassium hexafluorophosphate (KPF6) in acetone. All these structures were confirmed by proton nuclear magnetic resonance (H-NMR) spectroscopy and high-resolution mass spectrometry (HRMS) (Supplementary Figurea S1–S14).

SCHEME 1: Synthetic routes to TPIT-X molecules. Reagents and conditions: (a) TiCl4, Zn, THF, reflux, 10–12 h; (b) piperidine, EtOH, reflux, 10–12 h; (c) KPF6, acetone, room temperature.

Optical properties

The optical properties of TPIT-1 to TPIT-6 were firstly investigated by UV-vis absorption and photoluminescence (PL) spectroscopy. As illustrated in Figures 1a,b, the absorption maxima of TPIT-1 to TPIT-6 progressively red-shifted from 388 nm to 507 nm, while their emission maxima spanned from 490 nm (green) to 695 nm (deep red). These results demonstrate successful tuning of emission wavelengths across almost entire visible spectrum by solely modifying the acceptor moieties within the D-π-A backbone. The redshift in both absorption and emission can be attributed to the gradually increasing electron-withdrawing capability of the acceptor units (phenyl < cyano < pyridyl), which effectively strengthens the ICT effect. The AIE characteristics of these fluorescent probes were also investigated in tetrahydrofuran (THF)/water mixed solvents at different water contents (f w). As a typical example showed in Figures 1c,d, TPIT-4 consistently exhibited negligible fluorescence intensity when the water fraction (f w) increased from 0% to 70%, whereas a marked 30.3-fold enhancement emerged when f w exceeded 90%. Similar trends were observed for other derivatives (Supplementary Figures S15-S19), together with the photographs of their powders under 365 nm excitation (Supplementary Figure S20), confirming the typical AIE characteristics of these molecules. This phenomenon can be attributed to the RIR mechanism, where the restriction of intramolecular rotation in aggregated state suppresses non-radiative decay pathways. Notably, The fluorescence quenching of TPIT-5 and TPIT-6 at low water content (f w < 60%) is primarily attributed to efficient electron transfer from the imidazole donor to the N-methylpyridinium acceptor, which induces substantial charge separation and promotes the formation of a twisted intramolecular charge-transfer (TICT) state. Under these conditions, increased solvent polarity facilitates a nearly orthogonal donor–acceptor conformation through excited-state bond rotation, ultimately leading to emission attenuation (McDonald et al., 2025; Wang et al., 2021). In contrast, when the water fraction exceeds 60%, molecular aggregation effectively restricts internal bond rotation, thereby suppressing the TICT pathway and, consistent with the RIR mechanism, triggering a pronounced AIE response (Bhuin et al., 2025; Zhang et al., 2024; Zhu et al., 2025). In conclusion, these results prove that the emission window and AIE performance of this molecular “platform” can be precisely programmed through rational selection of acceptor units with distinct electron-withdrawing abilities and steric effects.

FIGURE 1: (a) Normalized UV-Vis absorption spectra of TPIT-1 to TPIT-6 in tetrahydrofuran (THF). (b) Normalized fluorescence emission spectra of TPIT-1 to TPIT-6 in THF. (c) Fluorescence emission spectra of TPIT-4 in water/THF mixtures with varying water fractions (f w). (d) Plot of PL peak intensities of TPIT-4 as a function of f w in water/THF mixtures. I 0 and I denote the PL intensities of the compounds in pure THF and in water/THF mixtures, respectively. Inset: Fluorescent photographs of TPIT-1 to TPIT-6 in THF/water mixtures (f w = 90%) under 365 nm excitation.

Cell imaging

It is well known that long-wavelength emitting fluorophores can effectively avoid biological autofluorescence, reduce photon scattering, and improve deep-tissue penetration, thereby enhancing imaging contrast and resolution. Therefore, we selected TPIT-6, which emits in deep red and NIR-I region, as the fluorescent probe for in vivo imaging. To overcome the poor solubility of TPIT-6 in aqueous media, we prepared a suspension of TPIT-6 nanoparticles (TPIT-6 NPs) at a concentration of 2 mg/mL using DSPE-PEG2000 lipids. As illustrated in Figure 2a, dynamic light scattering (DLS) and transmission electron microscopy (TEM) characterization confirm that TPIT-6 NPs exhibit a uniform, nearly spherical morphology with an average diameter of approximately 200 nm. Further fluorescence spectroscopy analysis (Figure 2b) reveals that TPIT-6 NPs have a maximum absorption wavelength at 506 nm, with an emission peak located at 697 nm. These results demonstrate the successful establishment of the TPIT-6 NPs deep-red-emitting probe system.

FIGURE 2: (a) Dynamic light scattering (DLS) diameter of TPIT-6 NPs; Inset: Transmission electron microscopy (TEM) image of the NPs. (b) UV-Vis absorption and photoluminescence spectra of TPIT-6 NPs. (c) Dynamic light scattering (DLS) diameter of TPIT-6 NPs-Ab; Inset: Transmission electron microscopy (TEM) image of the NPs. (d) Comparison of the zeta potential between TPIT-6 NPs and TPIT-6 NPs-Ab. Error bars: standard deviation (SD), experimental times: n = 3.

To equip the NPs with active targeting capability towards SCC tumor, we functionalized TPIT-6 NPs by coupling anti-EGFR antibodies, resulting in TPIT-6 NPs-Ab. Characterization by DLS and TEM confirmed that the resulting TPIT-6 NPs-Ab also exhibited a uniform near-spherical morphology with an average diameter of approximately 200 nm (Figure 2c). As shown in Figure 2d, the zeta potentials of TPIT-6 NPs-Ab and TPIT-6 NPs were measured to be −39 mV and −29 mV, respectively. According to previous reports, a higher absolute zeta potential value typically indicates better suspension stability, as enhanced electrostatic repulsion between particles reduces their tendency to aggregate (Jarzynska et al., 2025). This finding suggests that antibody conjugation moderately improves the stability of the nanoparticles. In addition, neither TPIT-6 NPs nor TPIT-6 NPs-Ab exhibited any significant variation in fluorescence intensity following a 10-day storage period in PBS buffer at 37 °C, indicating excellent stability under physiological conditions (Supplementary Figure S21). Before initiating cell experiments, the cytotoxicity of TPIT-6 NPs and TPIT-6 NPs-Ab toward normal cells was evaluated in mouse smooth muscle cells (SMCs) using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. As shown in Supplementary Figures S22 and S23, after 24 h of co-incubation with TPIT-6 NPs and TPIT-6 NPs-Ab under dark, respectively, the viability of SMCs remained above 90% in both cases even when the probe concentration was increased to 100 μM. These findings indicate the low cytotoxicity of TPIT-6 NPs and TPIT-6 NPs-Ab toward normal cells and tissues. To further assess the effectiveness of anti-EGFR antibodies in SCC targeting, SCC7 cells were incubated with either TPIT-6 NPs or TPIT-6 NPs-Ab. As shown in Figure 3, confocal laser scanning microscopy demonstrated that TPIT-6 NPs-Ab specifically bound to SCC7 cells, while TPIT-6 NPs showed minimal binding under identical conditions. It is noteworthy that both NPs appear to predominantly localize on the cell surface or within intercellular spaces, demonstrating that nanospheres of this size exhibit limited cellular internalization efficiency.

FIGURE 3: (a) Confocal fluorescence images of SCC7 cells treated with (A) TPIT-6-NPs (10 μM) and (b) TPIT-6-NPs-Ab (10 μM). Red color indicates the fluorescence signal in the red channel (680–720 nm, λex = 506 nm). Merge: Merged images with red channel and bright-field images. B) Mean fluorescence intensity of TPIT-6-NPs and TPIT-6-NPs-Ab. Error bars: standard deviation (SD), experimental times: n = 3, **P < 0.01. Scale bar: 40 μm.

Ex vivo tissue imaging

We next explored whether the TPIT-6 NPs-Ab probe could serve as a rapid diagnostic tool for intraoperative tissue biopsy. As shown in Scheme 2, we designed a convenient testing workflow to obtain diagnostic information including tumor margin discrimination and tumor cell metastasis status during the surgical procedure, thereby assisting surgeons in rapid pathological diagnosis. In detail, suspected tumor specimens resected from the surgical margin were immediately placed in PBS containing 5% fetal bovine serum, then transferred to PBS containing 50 μM TPIT-6 NPs-Ab and incubated at room temperature for 15 min. Following incubation, the specimen was rinsed twice with PBS buffer to remove unbound probes and then imaged using an IVIS LumiNova II imaging system. Ex vivo experiments were then conducted in C3H mice models to evaluate the performance of our method in the pathological diagnosis of SCC subcutaneous tumor resection. As shown in Figure 4, after treatment with TPIT-6 NPs-Ab, certain areas of the suspected tumor tissue specimens became highly luminescent and the margins of these areas are quite clear under imaging system. To validate the accuracy of the tumor margin indicated by fluorescence, tissue samples were also collected from the edges of the fluorescent regions and processed into paraffin sections. Subsequent hematoxylin and eosin (H&E) staining of these sections revealed the coexistence of both tumor cells and normal tissues under microscopic examination. Accordingly, the fluorescent regions reasonably corresponded to the actual tumor locations. These findings indicate that TPIT-6 NPs-Ab-based fluorometric assay can discriminate tumor margins rapidly through ex vivo specimens lighted by fluorescence during resection surgeries.

SCHEME 2: Experimental procedure for suspected tissue examination. Tissue specimens from the surgical margin were incubated with TPIT-6-NPs-Ab (50 μM) at room temperature for 15 min, washed twice with PBS, and then imaged using an IVIS LumiNova II system.

FIGURE 4: Fluorescence and white-light images of an SCC7 tumor tissue slice following ex vivo incubation with TPIT-6-NPs-Ab, along with hematoxylin and eosin (H&E) staining of the paraffin section for pathological evaluation. Imaging channel: 695–770 nm. λex = 500 nm. Scale bar: 2 mm.