Amplified aggregation-induced emission via donor-acceptor-donor molecular architecture: a high-sensitivity lateral flow immunoassay platform Authors: Xuekun Bai, Hong Zhang, Rui Zhang, Zhanxu Liu, Wei Pang, Zefeng Mao, Zesheng Liu, Huanying Zhou, Zhixian Gao Published: 2026 ABSTRACT Precise control over the structure of aggregation-induced emission luminogens (AIEgens) enables rapid development of tailored luminescent molecules. In this study, 4-(4-(1,2,2-triphenylvinyl)phenyl)benzo[c][1,2,5]thiadiazole (TPB) (donor-acceptor, D–A) and 4,7-bis(4-(1,2,2-triphenylvinyl)phenyl)benzo[c][1,2,5]thiadiazole (BTB) (donor–acceptor–donor, D–A–D) were designed and synthesized using tetraphenylene as a starting unit, guided by molecular engineering principles. Theoretical calculations and experimental results demonstrate that the D–A–D structure of BTB enhances intramolecular charge transfer, induces highly twisted molecular conformation, and imparts greater structural rigidity, thereby effectively suppressing non-radiative transitions. Subsequently, the spatial confinement effect of polystyrene microspheres restricts the rotational motion of BTB, enabling the construction of fluorescent nanoparticles (NPs) with higher quantum yields and excellent versatility. When applied to a lateral flow immunoassay (LFIA) platform, BTBNPs demonstrated significantly higher sensitivity than TPBNPs and AuNPs. Moreover, the sensitive detection of melatonin in biological samples was achieved. In conclusion, this study demonstrates that D–A–D-type AIEgens exhibit superior luminescence performance and greater potential than their D–A-type counterparts for constructing high-brightness fluorescent probes. It also establishes theoretical calculation methods for AIEgens and offers a valuable reference for their design, analysis, and application.Graphical Abstract Supplementary InformationThe online version contains supplementary material available at 10.1186/s12951-026-04598-6. Introduction Point-of-care testing (POCT) relies on rapid and sensitive methods for on-site detection, playing an important role in diagnosing diseases and monitoring their progression [1, 2]. Lateral flow immunoassay (LFIA) is widely used for POCT in various fields because it is portable, easy to operate, and does not require bulky testing equipment [3, 4]. Gold nanoparticle (AuNP)-based LFIA is the most classical and widely used colorimetric diagnostic method; [5, 6] however, its low signal intensity makes it unsuitable for detecting low-concentration targets [7–9]. Consequently, the development of high-brightness signalling tags is imperative for enhancing the sensitivity of LFIA. Recently, fluorescent dyes, carbon dots, and quantum dots (QDs) have been developed to enhance brightness [10–12]. However, these fluorescent materials often experience intermolecular π-π stacking, leading to aggregation-caused quenching (ACQ) at high concentrations or in aggregated states [13, 14]. This effect weakens the fluorescence signal of LFIA and reduces its sensitivity. Conversely, aggregation-induced emission luminogens (AIEgens) exhibit extremely low fluorescence signals at low concentrations [15, 16]. When at high concentrations or upon aggregation, their intramolecular rotation or movement is restricted, which blocks nonradiative transition pathways and releases energy through radiative decay channels, significantly enhancing the fluorescence signal [17, 18]. Currently, the majority of commonly used AIEgens feature simple symmetric structures, which possess limited luminous efficiency, molecular rigidity, and charge transport capability. Research has demonstrated that molecular engineering can enhance the photophysical properties of AIEgens by adjusting their molecular structures [19, 20]. A common strategy involves introducing donor or acceptor groups into AIEgens to establish donor-acceptor structures (D–A) [21]. The intramolecular charge transfer (ICT) effect generated can be used to regulate the electron cloud distribution and optimize the emission wavelength and fluorescence intensity of AIEgens. When the intensity of ICT within AIEgens increases, it can induce conformational changes in the molecule, leading to twisted intramolecular charge transfer (TICT) [22, 23]. The twisted and rigid molecular configuration can effectively obstruct intermolecular π-π interactions and restrict intramolecular motions [24]. Currently, most studies focus primarily on the independent reporting of D–A or D–A–D type AIEgens, whereas systematic comparative studies between single-donor and multi-donor AIEgens remain relatively scarce. The structure-activity relationship between these two categories, in terms of molecular structure, charge transfer capability, intermolecular forces, and fluorescence properties, remains poorly defined. In-depth investigation into such structural differences and the advantages of configurations is therefore of great theoretical guiding significance for the design and development of high-performance AIE fluorescent probes. Herein, tetraphenylethylene (TPE), a typical AIEgen was used as the starting material and an electron-donating unit. The typical strong electron-accepting unit, benzothiadiazole (BTD), was introduced to TPE to construct a D–A type AIEgen 4-(4-(1,2,2-triphenylvinyl)phenyl)benzo[c][1,2,5]thiadiazole (TPB) and a classic symmetric D–A–D type AIEgen 4,7-bis(4-(1,2,2-triphenylvinyl)phenyl)benzo[c][1,2,5]thiadiazole (BTB). Using TPE, TPB, and BTB as model molecules, the effects of different molecular structures on ICT, intramolecular rotational motion, and intermolecular interactions were investigated. Experimental results and theoretical calculations demonstrate that BTB exhibits stronger ICT, more twisted molecular conformation, and greater backbone stiffness than TPB. This also results in stronger intermolecular interactions, which in turn generate stronger restriction of intramolecular motions with higher quantum yields (Scheme 1A). Thus, incorporating multiple donors can effectively improve the TICT and luminescence performance. Subsequently, BTB-based fluorescent nanoparticles (BTBNPs) were constructed using a region-restriction strategy (Scheme 1B). As a high brightness fluorescent probe, BTBNPs were combined with LFIA for the sensitive detection of melatonin (MLT) in biomedical samples (Scheme 1C), demonstrating superior performance compared to TPBNPs and traditional AuNPs. This study presents a molecular design strategy for AIEgen engineering, proposes a method for analyzing the properties of AIEgens from multiple perspectives, and confirms that D-A-D type AIEgens have stronger luminescent properties. Finally, a high-brightness fluorescent probe and a more sensitive POCT sensing platform were constructed, offering valuable references for the design, analysis, and application of AIEgens. Scheme 1(A) Chemical structures of TPB and BTB. (B) The preparation of BTBNPs and BTBNPs@Ab. (C) Schematic of detection mechanism for circadian rhythm-associated biomarkers by BTBNPs-LFIA sensing platform (A) Chemical structures of TPB and BTB. (B) The preparation of BTBNPs and BTBNPs@Ab. (C) Schematic of detection mechanism for circadian rhythm-associated biomarkers by BTBNPs-LFIA sensing platform Results and discussion Exploring the properties of AIEgens The synthesis steps for the AIEgens are shown in Fig. S1 and their nuclear magnetic resonance spectroscopy (NMR) results are shown in Figs. S2-S7. The molecular structures of the obtained AIEgens are shown in Fig. 1A. D–A-type structure was formed by combining a BTD acceptor unit and a TPE donor unit. Subsequently, by continuously introducing the donor unit, a multi-donor D–A–D structure was obtained. The molecular structure changes affect the luminescence properties of AIEgens. As shown in Fig. 1B, the maximum UV-vis absorption peak of TPE in tetrahydrofuran (THF) is detected at 311 nm, which is caused by its intrinsic π-π* transition [25]. After introducing BTD, a broad and weak absorption peak emerges at 376 nm, which is attributed to the ICT from TPE to BTD in the TPB molecule. When another TPE molecule was introduced to form BTB, similar characteristic peaks were observed. Notably, the ICT absorption band of BTB was red-shifted to 420 nm with increased intensity, indicating that the double-donor structure significantly enhances the electron transition ability and further strengthens ICT [26]. Similar phenomena were observed in Fig. 1C. The maximum emission wavelengths of TPE, TPB, and BTB in THF exhibited a successive red shift (446, 512, and 528 nm, respectively), with corresponding gradual color changes, suggesting that the addition of acceptor and double-donor units significantly enhances the ICT effect. The fluorescence spectra of TPB and BTB were measured in various polar solvents (Figs. 1D and S8). As the polarity of the solvent increased, the fluorescence spectra gradually redshifted, indicating more pronounced ICT characteristics [27]. The CIE chromaticity diagram shown in Fig. 1E confirms this conclusion. Meanwhile, the spectrum gradually broadened with increasing solvent polarity, implying changes in the initial excited state, which may induce the transformation from ICT to TICT [28]. In contrast, the spectrum of TPE exhibited a blue shift (Fig. S9), showing no ICT characteristics. Moreover, with increasing solvent polarity, the Stokes shift of BTB gradually increased (Fig. 1F). Based on the relationship between fluorescence emission and solvent orientation polarizability (Table S1 and Fig. S10), the slope of BTB is greater than that of TPB, indicating stronger ICT in BTB due to its double-donor structure [22]. Next, the solvent effect was studied using THF as a good solvent and H₂O as an inert solvent. When the fraction of water (fw) in Figs. 1G-H was 0–30%, the fluorescence intensity of BTB gradually decreased, and the spectrum showed a red shift. This is attributed to the fact that AIEgens remain in an unaggregated state, where their TICT effect facilitates free torsion of the molecular skeleton. Such torsional motion enhances non-radiative energy dissipation within the excited state, ultimately resulting in fluorescence quenching [28]. Conversely, owing to the AIE effect, the spectrum underwent a blue shift again after 40%, and the fluorescence increased sharply. This moment, AIEgens are in an aggregated state, intermolecular stacking interactions significantly suppress molecular torsion while maintaining their highly distorted molecular conformations. This effectively blocks the non-radiative relaxation channels induced by TICT, ultimately achieving a remarkable enhancement in AIE fluorescence. As fw increased, the fluorescence intensity of TPE increased monotonically (Fig. S11), suggesting the absence of TICT within the TPE molecule. TICT in the TPB and BTB is caused by the D–A structure. The quantum yields (QY) of TPE, TPB, and BTB in the solid state were 0.32%, 10.27%, and 16.52% (Fig. 1I), respectively, while their corresponding fluorescence lifetimes decreased (Fig. 1J). These results demonstrate that molecular engineering effectively enhances the QY and photoluminescence performance. Compared to TPB, BTB exhibits more pronounced TICT characteristics and a higher QY, implying that the D–A–D structure significantly enhances the luminescence properties of AIEgens. Fig. 1Photophysical properties of AIEgens. (A) Molecular formula, (B) UV-vis absorption spectra, and (C) fluorescence spectra of TPE, TPB, and BTB (λex: 365 nm). (D) Fluorescence spectra (λex: 365 nm), (E) CIE chromaticity diagrams, and (F) Stokes shifts of BTB in organic solvents with different polarities. (G) Fluorescence spectra of BTB in the presence of different fractions of water (fw) (λex: 365 nm). (H) AIE effects of TPB and BTB, (F: fluorescence intensity, F′: maximum fluorescence intensity). (I) QYs of TPE, TPB, and BTB in the solid state. (J) Fluorescence decay curves of TPE, TPB, and BTB. Photophysical properties of AIEgens. (A) Molecular formula, (B) UV-vis absorption spectra, and (C) fluorescence spectra of TPE, TPB, and BTB (λex: 365 nm). (D) Fluorescence spectra (λex: 365 nm), (E) CIE chromaticity diagrams, and (F) Stokes shifts of BTB in organic solvents with different polarities. (G) Fluorescence spectra of BTB in the presence of different fractions of water (fw) (λex: 365 nm). (H) AIE effects of TPB and BTB, (F: fluorescence intensity, F′: maximum fluorescence intensity). (I) QYs of TPE, TPB, and BTB in the solid state. (J) Fluorescence decay curves of TPE, TPB, and BTB. Structural optimization and single-molecule simulation of AIEgens To further explore the luminescence properties of the AIEgens, their geometric structures were optimized using density functional theory (DFT) at the B3LYP/6-31G(d) level. With the introduction of BTD, the dihedral angle in TPE changed from 47.138° to 45.726°. Upon the formation of BTB, the dihedral angle in the tetraphenylethylene group recovered to 47.813°, and the dihedral angle between the donor and acceptor units increased from 35.303° to 36.223° (Figs. S12 and 2A). These changes in angle indicate that the molecular conformation constantly evolves during molecular engineering [29]. In particular, the D–A–D structure induces a more twisted molecular backbone in BTB, which may affect the intermolecular interactions and thus influence its luminescent properties. The frontier orbitals of the three AIEgens were calculated using DFT. For TPE, the electron clouds in the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) are distributed uniformly over the molecular skeleton, without obvious charge separation (Fig. S12). In contrast, for TPB and BTB, the electron clouds in the HOMO are mainly localized on the donor moieties, while those in the LUMO are distributed on the acceptor moieties (Fig. 2B). The distinct charge separation between HOMO and LUMO indicates the directional electron transfer from HOMO to LUMO, forming a typical ICT process, which also perfectly confirms the D–A structures of TPB and BTB [21]. Moreover, the HOMO energy levels gradually increase, whereas the LUMO energy levels gradually decrease. The energy gaps of TPE, TPB, and BTB are 4.12, 2.97, and 2.81 eV, respectively (Fig. 2C), which correspond to the red shift in the absorption and emission spectra upon the introduction of the acceptor and double donors. Therefore, receptor/donor-based molecular engineering can effectively regulate the absorption and emission wavelengths of AIEgens, ultimately yielding luminescent molecules with tunable colors. To further illustrate the potential differences between the D-A-D and D-A molecules, time-dependent density functional theory (TD-DFT) calculations were performed on TPB and BTB. By simulating the molecular geometric configurations of the ground and excited states, the root-mean-square deviation (RMSD) was calculated to illustrate the atomic displacement between the two conformations. As shown in Fig. 2D, the RMSD of TPB is 0.6521 Å, indicating obvious molecular motion and flexible molecular conformation. The RMSD of BTB (0.2477) is significantly lower than that of TPB (Fig. 2E), suggesting a more rigid molecular conformation [30]. Thus, the D–A–D molecular structure significantly restricts intramolecular rotational motion. The bond dissociation energy (BDE) further confirms the enhanced structural rigidity of BTB. As shown in Fig. 2F, BDE calculations for five bonds in both molecules reveal that BTB consistently exhibits higher BDE values than TPB in both ground and excited states. In other words, more energy is required to break the chemical bonds in BTB, further indicating its more rigid molecular structure. In addition to covalent bond interactions, non-covalent interactions in molecules have a significant impact on the luminescence properties of AIEgens. An analysis of reduced density gradient (RDG) provides insight into these interactions. As shown in Figs. 2G and S13, the overall RDG distributions of TPB and BTB were similar. However, when Sign(λ₂)ρ is greater than 0.02, the RDG distribution in TPB is slightly more dispersed, while that of BTB is more concentrated, implying the more concentrated non-covalent interaction regions and stronger intramolecular interactions in BTB [31]. The interaction region indicator (IRI) analysis (Fig. S14) shows obvious steric hindrance effects in both TPB and BTB [32]. However, the double-donor structure of BTB induces significant steric hindrance on both sides of the molecule, which considerably restricts the intramolecular rotation of BTB and enhances its molecular rigidity [34]. This further supports that BTB has a more stable molecular structure. Furthermore, the molecular motion conformations and geometric changes of AIEgens in the excited states can be quantitatively characterized by the reorganization energy (λ). As shown in Figs. 2H-I, the total λ values of TPB and BTB are 9961 and 5517 cm⁻¹, respectively, indicating that TPB has stronger intramolecular motion in the monodisperse state. In addition, BTB has a smaller reorganization energy contribution in the low-frequency region, implying no obvious rotational motion in its molecular skeleton [33]. Correspondingly, the rotational motion of the dihedral angle contributes the least to the reorganization energy (22.3%), confirming the relatively rigid molecular structure of BTB. Meanwhile, Fig. 2J shows that both the hole reorganization energy (λh) and electron reorganization energy (λe) of BTB are smaller than those of TPB, indicating that BTB undergoes less geometric structural change during the charge injection process and demonstrates better conformational stability [34]. To further analyze the luminescence and spatial orientation trends of TPB and BTB, their ground-state and transition dipole moments were evaluated. As shown in Fig. 2K, BTB exhibits stronger dipole moments than TPB, indicating a greater tendency for ordered aggregation and a higher transition intensity in the excited state, which collectively provide a structural basis for its high fluorescence efficiency. In summary, theoretical calculations of RMSD, BDE, λ, etc. demonstrate that the D–A–D type BTB has a more rigid molecular skeleton and less pronounced intramolecular rotational motion, which likely contribute to its enhanced AIE performance. Fig. 2Single-molecule simulation of AIEgens. (A) Optimized molecular structure, and (B) orbital distribution of TPB and BTB. (C) Energy levels of TPE, TPB, and BTB. Overlaps of the optimized ground state (red), excited state (green) geometries, and RMSD values of (D) TPB and (E) BTB in the gas phase. (F) BDE of TPB and BTB in ground and excited states. (G) Sign(λ2)ρ-colored independent gradient model scatter plot of BTB. Plots of the calculated reorganization energy versus the normal mode wavenumber of (H) TPB and (I) BTB. (Inset: the contribution of bond length, bond angle, and dihedral angle to the total λ). (J) Hole reorganization energy and electron reorganization energy of TPB and BTB. (K) Dipole moment of TPB and BTB in the ground state and transition Single-molecule simulation of AIEgens. (A) Optimized molecular structure, and (B) orbital distribution of TPB and BTB. (C) Energy levels of TPE, TPB, and BTB. Overlaps of the optimized ground state (red), excited state (green) geometries, and RMSD values of (D) TPB and (E) BTB in the gas phase. (F) BDE of TPB and BTB in ground and excited states. (G) Sign(λ2)ρ-colored independent gradient model scatter plot of BTB. Plots of the calculated reorganization energy versus the normal mode wavenumber of (H) TPB and (I) BTB. (Inset: the contribution of bond length, bond angle, and dihedral angle to the total λ). (J) Hole reorganization energy and electron reorganization energy of TPB and BTB. (K) Dipole moment of TPB and BTB in the ground state and transition Simulation of AIEgens crystals and analysis of intermolecular interactions To achieve a comprehensive understanding of the intermolecular interactions between AIEgens, the Polymorph module within Materials Studio was used to forecast and simulate the crystallographic properties of TPB and BTB, drawing on methodologies from prior research [35, 36]. Additionally, representative dimer structures and molecular clusters were systematically constructed using CrystalExplorer software [37]. Figs. 3A-B show the simulated dimeric structures of TPB and BTB. The distance d3 between two TPB molecules (3.322 Å) indicates strong C-H···N interactions. In BTB, d4 and d6 correspond to relatively strong C-H···π and C-H···N interactions, respectively. Moreover, most of the distances in BTB were shorter than those in TPB. Figures 3C-D clearly show the relationship between the dimers and the Hirshfeld surface, wherein BTB is more closely connected (the red dots represent the intermolecular interactions). The presence of additional red dots in Fig. 3D indicates that BTB has more intermolecular interaction sites. Therefore, the strong intermolecular interactions in BTB effectively restrict the rotational motion of the molecules and enhance radiative energy transition, resulting in a stronger luminescence effect. In addition, intermolecular interactions can be quantified using Hirshfeld surface analysis. The corresponding quantification results and fingerprint plots are shown in Fig. 3E, where H-H (65.8%) and C-H/H-C (23%) contribute to the main interaction forces. These results indicate that the dual-donor system can enhance intermolecular interactions. To accurately describe the intermolecular forces among multiple molecules, a multimolecular cluster model centered on a single molecule was constructed using CrystalExplorer (Figs. 3F-G). BTB exhibits stronger intermolecular forces (-95.7 kJ/mol) than TPB (-45.9 kJ/mol, Fig. 3H), imparting greater rigidity to the BTB molecular cluster and effectively restricting intramolecular rotational motion. The above calculation results demonstrate that high-performance D–A–D-type AIEgen (BTB) has been successfully achieved by rational molecular engineering. The strong intermolecular interactions in BTB restrict the intramolecular rotational motion and promote radiative transition, making it a promising candidate for the development of AIE fluorescent probes with enhanced luminescence performance. Fig. 3Crystalline analyses and intermolecular interactions. Typical dimers of (A) TPB and (B) BTB from their corresponding crystal-packing structures. Hirshfeld surfaces of (C) TPB and (D) BTB. (E) Hirshfeld surface analysis and fingerprint plot for BTB crystals. Crystal stacking of (F) TPB and (G) BTB. (H) Summary of the intermolecular interactions in TPB and BTB clusters. (Eele, electrostatic energy; Epol, polarization energy; Edis, dispersion energy; Erep, exchange-repulsion energy; Etol, the total energy of interaction between surrounding molecules. Unit: kJ/mol.) Crystalline analyses and intermolecular interactions. Typical dimers of (A) TPB and (B) BTB from their corresponding crystal-packing structures. Hirshfeld surfaces of (C) TPB and (D) BTB. (E) Hirshfeld surface analysis and fingerprint plot for BTB crystals. Crystal stacking of (F) TPB and (G) BTB. (H) Summary of the intermolecular interactions in TPB and BTB clusters. (Eele, electrostatic energy; Epol, polarization energy; Edis, dispersion energy; Erep, exchange-repulsion energy; Etol, the total energy of interaction between surrounding molecules. Unit: kJ/mol.) Preparation and characterization of fluorescent probes In this study, the BTBNPs-based fluorescent probe was prepared using a swelling method (Scheme 1B) [38]. As shown in Fig. 4A, polystyrene microspheres (PS) exhibit a spherical structure with a diameter of 302 nm. Moreover, the transmission electron microscopy (TEM) images show a transparent region around the spheres, indicating that the PS has a hollow structure. The TEM image of BTBNPs shows that they appear black overall without a transparent morphology, indicating that the BTB swelled into the interior of the PS. The BTBNPs maintained a uniform spherical morphology as their particle size increased to 331 nm (Fig. 4B). Subsequently, N2 adsorption-desorption isotherm was used to study the swelling process. As shown in Fig. 4C, the typical IV hysteresis loop indicates the mesoporous structure of PS, which facilitates the loading and confinement of BTB. The Brunauer–Emmett–Teller (BET) surface area and pore volume of PS are 375.6 m2/g and 0.9372 cm3/g, respectively (Fig. 4D). After loading BTB, the BET surface area and pore volume of BTBNPs decreased to 19.1 m2/g and 0.0643 cm3/g, respectively, indicating that BTB swelled into the pores of PS. This was further confirmed by the decrease in the Barrett–Joyner–Halenda (BJH) pore diameter from 13.5 to 6.2 nm. Compared to PS, BTBNPs exhibited an increased zeta potential (Fig. 4E) and higher resistance (Fig. 4F), indicating the successful preparation of BTBNPs. Next, the optical properties of BTBNPs were characterized. As shown in Fig. 4G, a characteristic absorption peak at 430 nm was observed for BTBNPs, which red-shifted compared to the absorption peak of BTB. Meanwhile, the fluorescence was significantly enhanced for BTBNPs (Fig. 4H), indicating that a large number of BTB were confined in the pores of PS in an aggregated state. By optimizing the concentration of BTB during the preparation of BTBNPs, the fluorescence intensity reached its maximum value at 1 mg/mL (Fig. 4I), indicating that the PS pores have reached a saturation state. The fluorescence spectrum (Fig. 4J) shows that the optimal excitation and emission wavelengths are 420 and 530 nm, respectively. Moreover, the QY was determined to be 57.99% (Fig. 4K), which was significantly higher than that of BTB. The spatial confinement of PS restricts the molecular motion of BTB, improving the fluorescence performance of BTBNPs. To demonstrate the high brightness of BTBNPs in LFIA, serial dilutions of BTBNPs were applied onto a nitrocellulose membrane and compared with the performance of TPBNPs and traditional AuNPs. As shown in Fig. 5A, BTBNPs maintained visible fluorescence even after being diluted 256-fold, whereas the maximum dilution factors for TPBNPs and AuNPs were 64-fold and 8-fold, respectively. This indicates the superior brightness of BTBNPs, which is expected to significantly improve the detection sensitivity of LFIA. Furthermore, BTBNPs maintained high fluorescence intensity in NaCl solutions of different concentrations (Fig. 4L), at different pH values, and after different storage times (Fig. S15), demonstrating excellent stability and promising application potential. Fig. 4Characterization and analysis of BTBNPs. Particle size distribution of (A) PS and (B) BTBNPs. (Inset: SEM images and TEM images of (A) PS and (B) BTBNPs.) (C) N2 adsorption/desorption isotherms (Inset: pore size distributions.), (D) BET surface area and total pore volume, (E) zeta potentials, (F) EIS, (G) UV-vis absorption spectra (inset: photographs of PS and BTBNPs under light.), and (H) fluorescence spectra (λex: 420 nm) (inset: photographs of PS and BTBNPs under 365 nm UV light.) of PS and BTBNPs. (I) Fluorescence spectra of BTBNPs with different dosages of BTB (0.2, 0.4, 0.6, 0.8, 1.0, 1.2 mg in 10 mg of PS) (λex: 420 nm). (J) Three-dimensional fluorescence spectra. (K) QY of BTBNPs in the solid state. (L) Fluorescence intensity of BTBNPs at different concentrations of NaCl Characterization and analysis of BTBNPs. Particle size distribution of (A) PS and (B) BTBNPs. (Inset: SEM images and TEM images of (A) PS and (B) BTBNPs.) (C) N2 adsorption/desorption isotherms (Inset: pore size distributions.), (D) BET surface area and total pore volume, (E) zeta potentials, (F) EIS, (G) UV-vis absorption spectra (inset: photographs of PS and BTBNPs under light.), and (H) fluorescence spectra (λex: 420 nm) (inset: photographs of PS and BTBNPs under 365 nm UV light.) of PS and BTBNPs. (I) Fluorescence spectra of BTBNPs with different dosages of BTB (0.2, 0.4, 0.6, 0.8, 1.0, 1.2 mg in 10 mg of PS) (λex: 420 nm). (J) Three-dimensional fluorescence spectra. (K) QY of BTBNPs in the solid state. (L) Fluorescence intensity of BTBNPs at different concentrations of NaCl Construction of fluorescent LFIA sensing platform To apply BTBNPs in an LFIA sensing platform, the carboxyl groups on the surface of BTBNPs were activated using the EDC/NHS method [39] and subsequently conjugated with monoclonal antibodies to prepare the immunofluorescent probe BTBNPs@Ab. The successful preparation of BTBNPs@Ab was verified by changes in particle size (Fig. 5B), zeta potential (Fig. 5C), and electrochemical impedance spectroscopy (EIS) response (Fig. 5D) compared to BTBNPs. Next, the LFIA test strips were marked with MLT@BSA and goat anti-mouse IgG on the test (T) and control (C) lines, respectively, to construct a competitive LFIA-sensing platform (Scheme 1C). In the presence of negative samples without the MLT, BTBNPs@Ab directly bound to MLT@BSA, causing the T line to emit bright fluorescence. In contrast, in the presence of positive samples containing the target, BTBNPs@Ab preferentially bound to the target, occupying the binding sites, which prevents interactions with MLT@BSA on the T line and results in a decrease in the fluorescence intensity. Therefore, the POCT of the target is achieved by detecting the fluorescence signal of the T line. Fig. 5Characterization and optimization of BTBNPs@Ab. (A) Images of BTBNPs, TPBNPs, and AuNPs at various dilution multiples titrated on nitrocellulose membranes. (B) Particle size distribution, (C) zeta potentials, and EIS of BTBNPs and BTBNPs@Ab. Optimization of BTBNPs-LFIA parameters: (E) pH, (F) Ab amounts, and (G) BTBNPs amounts. (Data are mean ± standard deviation, N = 3 for all groups) Characterization and optimization of BTBNPs@Ab. (A) Images of BTBNPs, TPBNPs, and AuNPs at various dilution multiples titrated on nitrocellulose membranes. (B) Particle size distribution, (C) zeta potentials, and EIS of BTBNPs and BTBNPs@Ab. Optimization of BTBNPs-LFIA parameters: (E) pH, (F) Ab amounts, and (G) BTBNPs amounts. (Data are mean ± standard deviation, N = 3 for all groups) Fluorescent LFIA sensing platform for POCT of MLT Circadian rhythm disorder, a common physiological condition, is caused by long-term irregular work and rest that disrupts the body’s biological clock [40, 41]. MLT is a hormone highly associated with circadian rhythms, showing an obvious correlation between its content and the rhythmic state [42, 43]. When exogenous MLT is dosed, the MLT concentration in the human body increases significantly. Precise detection of MLT levels not only facilitates a clear understanding of its metabolic and excretion profiles in vivo but also enables the evaluation of bioavailability, and further provide reliable guidance for clinical medication. Therefore, the LFIA platform constructed using BTBNPs as signal probe (BTBNPs-LFIA) enables POCT of MLT for rapid detection of the rhythm state of the body. To obtain better detection performance, the experimental conditions for BTBNPs-LFIA were optimized. As shown in Fig. 5E, the T-line exhibits the highest brightness and the highest fluorescence inhibition rate at pH 9, likely due to the optimal antigen–antibody binding efficiency under this condition. In Fig. 5F, the inhibition rate of the T line reached its maximum when the labeling amount of Ab was 20 µg. Similarly, when BTBNPs@Ab was 7 µg (Fig. 5G) and the incubation time was 6 min (Fig. S16), the inhibition rate reached its maximum, indicating complete binding of BTBNPs@Ab to MLT. Under optimal conditions, the fluorescence intensity of the T-line in BTBNPs-LFIA gradually decreased with MLT concentration (Fig. 6A), showing a good linear relationship in the range of 3 ng/mL to 10 µg/mL (R2 = 0.9836). The linear equation was y = -573.1 × lg x + 2373, and the limit of detection (LOD) could reach 1.82 ng/mL. To highlight the excellent performance of BTBNPs-LFIA, a fluorescent LFIA platform (TPBNPs-LFIA) and a colorimetric LFIA platform (AuNPs-LFIA) were subsequently constructed using TPBNPs and AuNPs as signal probes, respectively, and applied to the detection of MLT (The optimized experimental conditions for them are shown in Figs. S17-S18). Similarly, the T-line signals decreased with increasing MLT concentrations. Among them, TPBNPs-LFIA showed a linear relationship with R2 = 0.9889 in the range of 10-3000 ng/mL. The linear equation was y = -383.4 × lg x + 1579, and the LOD was 5.01 ng/mL (Fig. 6B). AuNPs-LFIA exhibited a linear relationship with R2 = 0.9781 in the range of 30 ng/mL-10 µg/mL. The linear equation was y = -2304 × lg x + 9023, and the LOD was 18.7 ng/mL (Fig. 6C). Although all three types of LFIA could successfully detect MLT, the LOD of BTBNPs-LFIA was 2.75-fold and 10.27-fold lower than that of TPBNPs-LFIA and AuNPs-LFIA, respectively, showing better detection performance (Fig. 6D). Compared with the reported detection methods for MLT, the BTBNPs-LFIA strategy established in this work exhibits higher sensitivity (Table S2). This is attributed to the excellent AIE properties of the D–A–D type BTB, which enhance the fluorescence brightness, thereby enabling BTBNPs-LFIA to have a wider detection range and lower LOD. The detection signals of BTBNPs-LFIA for non-target substances showed no significant difference compared to the control group (Fig. 6E), indicating that BTBNPs-LFIA had satisfactory selectivity for MLT detection. When stored at room temperature for 7 days, BTBNPs-LFIA showed high recoveries ranging from 84.97% to 119.08% (Fig. 6F), suggesting the good stability of the constructed POCT platform. To evaluate the practical application potential of BTBNPs-LFIA, it was used for the detection of MLT in artificially spiked saliva and urine. As shown in Table S3, the recovery ranges of MLT in saliva and urine were 94.60% − 109.53% with relative standard deviations (RSDs) of less than 4.86%. A commercial kit was used to verify the detection results. As shown in Table S4, the consistency between the two methods confirms that BTBNPs-LFIA can reliably detect MLT in biological samples and holds promise for POCT of rhythm disorder states. Fig. 6Sensing performance of BTBNPs-LFIA for MLT. (A) Photographs, detection results and standard curves of BTBNPs-LFIA for the detection of different concentrations of MLT. (B) Photographs, detection results and standard curves of TPBNPs-LFIA for the detection of different concentrations of MLT. (C) Photographs, detection results and standard curves of AuNPs-LFIA for the detection of different concentrations of MLT. (D) Comparison of the LODs of BTBNPs-LFIA, TPBNPs-LFIA, and AuNPs-LFIA for the detection of MLT. (E) Responses of BTBNPs-LFIA for the detection of MLT and non-targets (1 µg/mL). (F) Stability of BTBNPs-LFIA. (Data are mean ± standard deviation, N = 3 for all groups) Sensing performance of BTBNPs-LFIA for MLT. (A) Photographs, detection results and standard curves of BTBNPs-LFIA for the detection of different concentrations of MLT. (B) Photographs, detection results and standard curves of TPBNPs-LFIA for the detection of different concentrations of MLT. (C) Photographs, detection results and standard curves of AuNPs-LFIA for the detection of different concentrations of MLT. (D) Comparison of the LODs of BTBNPs-LFIA, TPBNPs-LFIA, and AuNPs-LFIA for the detection of MLT. (E) Responses of BTBNPs-LFIA for the detection of MLT and non-targets (1 µg/mL). (F) Stability of BTBNPs-LFIA. (Data are mean ± standard deviation, N = 3 for all groups) Conclusion In this study, molecular regulation methods were employed to prepare two AIEgens, i.e., TPB (D–A) and BTB (D–A–D), as model molecules. The experimental results proved that the D–A–D type BTB exhibited a stronger TICT effect and a higher QY. Subsequent simulations and calculations proved that BTB possessed a more rigid molecular structure, a more distorted molecular conformation, and stronger intermolecular interactions, providing a theoretical basis for its improved luminescence performance. Highly luminescent BTB was incorporated with PS, where its rotational motion was effectively restricted, contributing to enhanced fluorescence. The resulting BTBNPs were used as a high brightness fluorescent probe to construct LFIA sensing platforms to detect MLT, showing good practical performance and high sensitivity. The excellent luminescence performance of BTB also improved the sensitivity of BTBNPs-LFIA, compared to that of TPBNPs-LFIA and AuNPs-LFIA, confirming the feasibility of using molecular engineering to regulate AIEgens and improve the sensitivity of LFIA. In summary, this study demonstrates from multiple perspectives that the luminescence performance of D–A–D-type AIEgens is superior to that of D–A-type AIEgens. Our findings provide constructive guidance for the rational design of novel AIEgens with strong luminescence, which is expected to accelerate the development of fluorescence-based POCT. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1. Supplementary Material 1.