Paper-based microfluidic analytical device for detecting beta-lactoglobulin and preparation method and application thereof

Abstract

The application discloses a paper-based microfluidic analysis device for beta-lactoglobulin detection and a preparation method and application thereof, and relates to the technical field of food detection. The paper-based microfluidic analysis device comprises a bottom plate and a water-absorbing paper loaded on the bottom plate. The water-absorbing paper comprises, in sequence along a liquid flow direction, a sample application area, an acidification area containing a buffer system with a pH range of 3-5, a deposition area with at least one layer of water-absorbing paper stacked thereon, a neutralization area containing a buffer system with a pH range of 6-8, and a detection area containing a fluorescent probe for beta-lactoglobulin detection. The fluorescent probe has the structure: R1 and R2 are selected from hydrogen or C1-C5 alkoxy, and R3 is methyl or ethyl. When the paper-based microfluidic analysis device is applied, the detection sample is subjected to casein precipitation in the acidification area, mechanical filtration in the deposition area, and recovery of the physiological pH condition before detection in the neutralization area, and finally combines with the fluorescent probe in the detection area to generate a fluorescent detection signal.

Description

Technical Field

[0001] This invention relates to the field of food testing technology, and in particular to a paper-based microfluidic analytical device for the detection of β-lactoglobulin, its preparation method, and its application. Background Technology

[0002] β-Lactoglobulin (β-LG) is the main whey protein in the milk of ruminant animals and plays a vital role in nutritional science and food technology. As a rich source of essential amino acids, its accurate quantification is crucial for assessing the nutritional value of dairy products and infant formula. Clinically, β-LG is a major cause of cow's milk protein allergy (CMPA), and for the dairy industry, accurate β-LG monitoring is equally important for process control, nutritional consistency, and compliance with allergen labeling. However, the direct, rapid, and stable quantification of β-LG in the complex, optically opaque milk matrix remains challenging.

[0003] Conventional methods for β-LG detection, such as enzyme-linked immunosorbent assay (ELISA) and liquid chromatography-mass spectrometry (LC-MS), offer high sensitivity and specificity. However, these methods rely on labor-intensive workflows, long turnaround times, and complex instruments and trained personnel, limiting field deployment and high-throughput screening. These limitations have spurred the development of simple, rapid, cost-effective, and portable β-LG analysis platforms.

[0004] Environmentally activated small molecule fluorescent probes offer an effective alternative, combining ease of operation with real-time and high signal response. The unique three-dimensional structure of β-LG provides a perfect physical basis for the rational design of cavity-targeted probes. Figure 1 The monomer of the β-cathode molecule folds into a highly stable β-cavity structure that surrounds a well-defined conical hydrophobic pouch called the calyx. This cavity is lined with nonpolar amino acid residues, providing ideal docking sites for small hydrophobic ligands. However, existing fluorescent probes have failed to fully utilize this structural feature due to their reliance on a single photophysical mechanism. Probes based solely on twisted intramolecular charge transfer (TICT) exhibit two key limitations: (i) they respond to any hydrophobic environment, leading to co-binding and false positives from other proteins or lipids, and (ii) their on-rate is limited because TICT inhibition alone cannot provide sufficient contrast between bound and unbound states.

[0005] Therefore, existing technologies still need improvement and development. Summary of the Invention

[0006] In view of the shortcomings of the prior art, the purpose of this invention is to provide a paper-based microfluidic analysis device for β-lactoglobulin detection, its preparation method and application, aiming to solve the problems of poor specificity and sensitivity of existing fluorescent probes for β-lactoglobulin detection and the lack of corresponding portable analysis devices.

[0007] The technical solution of the present invention is as follows: In a first aspect, a paper-based microfluidic analytical device for the detection of β-lactoglobulin is provided, comprising: a substrate and absorbent paper loaded on the substrate; The absorbent paper includes, in sequence along the liquid flow direction: a sample application zone, an acidification zone, a deposition zone, a neutralization zone, and a detection zone; The acidification zone contains a buffer system with a pH range of 3 to 5; The sedimentation zone is superimposed with at least one layer of absorbent paper; The neutralization zone contains a buffer system with a pH range of 6 to 8; The detection area contains a fluorescent probe for the detection of β-lactoglobulin; The fluorescent probe has the following structure: R1 and R2 are independently selected from hydrogen or C1~C5 alkoxy groups, and R3 is methyl or ethyl.

[0008] Specifically, this invention develops a paper-based microfluidic analysis device, wherein the sample application zone undergoes no additional treatment and is used to add the sample solution during detection; the acidification zone contains a buffer system with a pH range of 3-5, which can induce casein precipitation upon sample contact, reducing interference from non-target proteins; the deposition zone undergoes no additional chemical modification and mechanically filters the precipitated protein aggregates by stacking at least one layer of absorbent paper, thereby retaining the curd; the neutralization zone contains a buffer system with a pH range of 6-8, which can restore the sample to its pre-detection physiological pH conditions upon sample contact; and the detection zone contains a fluorescent probe for β-lactoglobulin detection, which can bind to the final β-lactoglobulin and generate a fluorescent detection signal.

[0009] The fluorescent probe is based on a novel design strategy that synergistically combines aggregation-induced emission (AIE) and TICT mechanisms. This dual-mechanism approach separates target recognition from signal amplification: the AIE component, restricted by intramolecular motion (RIM), generates primary fluorescence enhancement after mechanical fixation within the protein cavity, while the TICT mechanism provides secondary environmental sensitivity specific to the calyx microenvironment. This synergistic effect produces a multiplier enhancement effect—full activation of the probe requires mechanical confinement and specific dielectric properties of the β-LG cavity, thereby achieving unprecedented selectivity and signal amplification.

[0010] This fluorescent probe employs a D-π-A backbone, using triphenylamine (TPA) as a conserved electron donor and molecular rotor. The propeller-like geometry of TPA enables efficient non-radiative decay through low-barrier intramolecular rotation in an aqueous environment. When encapsulated within a hydrophobic protein cavity, RIM induces significant fluorescence enhancement. TPA's favorable electronic properties, including its high-occupied highest-occupied-motor (HOMO) orbitals and oxidative stability, allow for efficient intramolecular charge transfer (ICT) with various acceptor groups while maintaining photostability.

[0011] To systematically tune photophysical properties and binding affinity, this invention strengthens the π-skeletal framework by introducing diene bridges, thereby increasing molecular rigidity while extending conjugation. Furthermore, the acceptor binding sites are modified to fine-tune the electronic properties of each fluorescent probe. Strategic functionalization of the donor aryl substituents is used to optimize the balance between lipophilicity and cavity compatibility, minimizing nonspecific interactions while maintaining adequate steric accessibility.

[0012] The core of the recognition strategy is the introduction of a cation acceptor moiety that utilizes electrostatic interactions with acidic residues near the β-LG calyx. This design element serves multiple functions: (i) providing remote electrostatic guidance for the binding site, (ii) enhancing local binding through ionic interactions, (iii) improving water solubility, and (iv) enhancing the probe's ICT properties.

[0013] The molecular size was designed to match the geometry of the β-LG calyx, with a total length of approximately 18–19 Å and a cross-sectional diameter of <10 Å, to facilitate deep insertion into the conical pouch. The binding orientation positions the cation terminus near the acidic edge while maximizing hydrophobic contact and potential π-π stacking with aromatic residues, generating multivalent interactions, increasing affinity, restricting rotor motion, and favoring the formation of a well-defined 1:1 probe-protein complex with a high opening rate.

[0014] This modular strategy enables orthogonal control over viscosity sensitivity, cavity complementarity, and binding thermodynamics, providing a novel fluorescent probe for highly selective β-LG detection.

[0015] Optionally, the molar ratio of the buffer system with a pH range of 3 to 5, the buffer system with a pH range of 6 to 8, and the fluorescent probe is 1:(1 to 1.5):(0.002 to 0.005). The total molar amount of the buffer system with a pH range of 3 to 5 is 1 to 50 μmol; The total molar amount of the buffer system with a pH range of 6 to 8 is 1 to 75 μmol; The total molar amount of the fluorescent probe is 2~250 nmol.

[0016] Optionally, the buffer system with a pH range of 3 to 5 is selected from one or more of the following: formic acid-formate buffer system, acetic acid-acetate buffer system, and citric acid-citate buffer system.

[0017] Optionally, the buffer system with a pH range of 6 to 8 is selected from one or more of the following: phosphate buffer system, tris(hydroxymethyl)aminomethane-hydrochloric acid buffer system, and phosphate-imidazole buffer system.

[0018] Optionally, the fluorescent probe is selected from one of the following structures: .

[0019] Optionally, the detection area may further contain a lyophilization protectant and / or a film-forming agent; The freeze-drying protectant is selected from one or more of trehalose, sucrose, lactose, chitosan, and polyols; The film-forming agent is selected from one or more of polyvinylpyrrolidone, polyethylene glycol, polyvinyl alcohol, polypropylene glycol, and polyacrylate; The total mass of the freeze-drying protectant is 1~10 mg, and the total mass of the film-forming agent is 1~20 μg.

[0020] Optionally, the absorbent paper is commercially available chromatography paper.

[0021] Optionally, the base plate is a polyethylene terephthalate (PET) sheet, but is not limited to it. The base plate only provides support, and other substances that do not react with the test sample and added reagents are also acceptable.

[0022] Secondly, a method for preparing a paper-based microfluidic analysis device as described in the first aspect is provided, comprising the following steps: Place the absorbent paper onto the base plate; A first solution is added to the acidification zone of the absorbent paper, the first solution containing a buffer system with a pH range of 3 to 5; A second solution is added to the neutralization zone of the absorbent paper, the second solution containing a buffer system with a pH range of 6 to 8; A third solution containing a fluorescent probe for β-lactoglobulin detection is added to the detection area of ​​the absorbent paper. At least one layer of absorbent paper is superimposed on the deposition area of ​​the absorbent paper to obtain the paper-based microfluidic analysis device.

[0023] Optionally, the first solution, the second solution, and / or the third solution may further contain ≤1% by volume of dimethyl sulfoxide.

[0024] Optionally, after any of the steps of adding a first solution to the acidification zone of the absorbent paper, adding a second solution to the neutralization zone of the absorbent paper, and adding a third solution to the detection zone of the absorbent paper, the process further includes: a drying process.

[0025] Optionally, the drying conditions include: a temperature of 20~45℃ and a time of 30~120 minutes.

[0026] Thirdly, the application of the paper-based microfluidic analysis device as described in any of the first aspects in detecting the β-lactoglobulin content in food is provided.

[0027] Optionally, the food is a dairy product, including one or more of milk, yogurt, and milk powder.

[0028] Fourthly, the application of the paper-based microfluidic analysis device as described in any of the first aspects in the preparation of β-lactoglobulin detection products is provided.

[0029] Beneficial Effects: This invention provides a paper-based microfluidic analytical device for β-lactoglobulin detection, its preparation method, and its application. The paper-based microfluidic analytical device includes a substrate and absorbent paper loaded on the substrate. The absorbent paper, along the liquid flow direction, sequentially includes: a sample application zone, an acidification zone containing a buffer system with a pH range of 3-5, a deposition zone with at least one layer of absorbent paper, a neutralization zone containing a buffer system with a pH range of 6-8, and a detection zone containing a fluorescent probe for β-lactoglobulin detection. When this paper-based microfluidic analytical device is used for β-lactoglobulin detection, the sample undergoes casein precipitation in the acidification zone, mechanical filtration in the deposition zone, and the neutralization zone restores the physiological pH conditions before detection. Finally, the sample binds to the fluorescent probe in the detection zone, thereby generating a fluorescent detection signal. The fluorescent probe combines aggregation-induced emission (AIE) with suppression of tortuous intramolecular charge transfer (TICT) during insertion into the β-LG calyx. Employing a D-π-A backbone and using triphenylamine as a conserved electron donor and molecular rotor, it optimizes cavity coordination and electrostatic interactions through repositioning acceptor connections, thereby systematically modulating ICT and charge distribution. This cavity-locking strategy integrates molecular recognition and motion restriction, enabling rapid analysis and specific detection of β-LG in complex dairy matrices. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the principle of the present invention; wherein, A is the crystal structure of β-LG, highlighting the hydrophobic calyx binding cavity; B is the mechanism of action of the fluorescent probe designed using the β-LG calyx structure; C is the chemical structure of four synthesized AIE-TICT probes (LGL-1 to LGL-4).

[0031] Figure 2Figure 1 shows the results of the photophysical properties of the fluorescent probes of this invention and the mechanism of AIE-TICT behavior. A and B are the normalized UV-Vis absorption spectra (A) and normalized fluorescence emission spectra (B) of LGL-1 to LGL-4 in aqueous buffer (10mM PBS containing 1% (v / v) DMSO, pH 7.4, 25℃). C is the fluorescence enhancement factor ( ) when monitoring AIE behavior in a THF / H2O mixed system (total probe concentration = 3μM). I / I 0) and THF score (ƒ THF The diagram shows the relationship between the two regions, with the inset showing a magnified view of the critical aggregation region (ƒ). THF =0~20%); D is LGL-4 (λ ex =528nm) in ƒ THF Representative fluorescence spectra at 0%, 10%, and 90%; E reveals the fluorescence spectra of LGL-1 to LGL-4 at ƒ using DLS analysis. THF =0% (molecular dissolution) and ƒ THF =99% (aggregate state) of nano-aggregate formation; F~I are the viscosity-dependent fluorescence responses of LGL-1 to LGL-4 in solvents of different viscosities, including dichloromethane (η=0.44cP), ethanol (η=1.1cP) and ethylene glycol (η=16.0cP), with Rhodamine B ethanol solution ( Using 0.65 as the reference standard, the relative fluorescence quantum yield (Φ) was determined. fl J represents radiation in different solvent environments ( ); ) and non-radioactive ( The deconvolution of the rate constant revealed the viscosity-dependent suppression of the non-radiative pathway; K is the Förster-Hoffmann analysis of fluorescence intensity and viscosity in the methanol / glycerol mixture at 25 °C, yielding the viscosity sensitivity parameter. L is an Arrhenius analysis of temperature-dependent fluorescence in acetonitrile (278~333K), providing the activation energy for the formation of TICT; all measurements were performed using 0.1% (v / v) DMSO as a co-solvent to ensure complete dissolution.

[0032] Figure 3This is a graph showing the molecular docking analysis results of the fluorescent probes of this invention with β-LG. A through D are the optimized binding conformations of LGL-1 (A), LGL-2 (B), LGL-D (C), and LGL-4 (D) within the β-LG calyx bag obtained through molecular docking simulation. The left side (first column) shows the molecular dimensions (length × width, in Å) of each probe, highlighting their linear structure compatible with the binding cavity. The middle (second column) shows the complex docked with the probe, with the probe presented as a rod-shaped model embedded in the protein surface representation. The right side (third column) indicates the calyx bag entrance and bond-interacting residues. Using the β-LG crystal structure as the acceptor, docking was performed using AutoDock Vina. The binding conformations shown represent the lowest energy conformations in cluster analysis from 10 independent docking runs.

[0033] Figure 4 This is a simulation result of the molecular docking of the fluorescent probe of the present invention with β-LG, showing the electrostatic potential surface of the four probe molecules (LGL-1 to LGL-4) at their predicted binding sites.

[0034] Figure 5 This is a graph showing the results of fluorescence titration analysis of the fluorescent probes of this invention with β-LG; where A, D, G, and J represent the evolution of the steady-state fluorescence emission spectra of LGL-1 (A), LGL-2 (D), LGL-D (G), and LGL-4 (J) as the concentration of β-LG increases; experimental conditions: fluorescent probe concentration = 3 μM, PBS buffer (10 mM, pH 7.4, 25℃), DMSO content ≤ 1% (v / v), β-LG concentration range = 0~27 μM, and the arrows indicate the direction of spectral change as the concentration of β-LG increases; B, E, H, and K represent the changes in chromaticity diagrams when different concentrations of fluorescent probes LGL-1 (B), LGL-2 (E), LGL-D (H), and LGL-4 (K) react with β-LG; C, F, I, and L are obtained by plotting the normalized fluorescence enhancement at the corresponding emission maximum (…). F / F The relationship between 0) and β-LG concentration was plotted, and corresponding combined isotherms for LGL-1 (C), LGL-2 (F), LGL-D (I), and LGL-4 (L) were constructed. The solid lines represent the nonlinear least squares fitting of the 1:1 combined model and the combined constant ( K a The error bars represent the standard deviation of three measurements and the maximum fluorescence enhancement factor.

[0035] Figure 6 This is a graph showing the concentration-dependent fluorescence response of LGL-4 to β-LG; the concentrations of β-LG from left to right are 0, 1, 5, 10 and 15 μM, and the concentration of LGL-4 is 3 μM.

[0036] Figure 7 This is a graph showing the selection and stability evaluation results of LGL-4 for β-LG detection; where A represents the fluorescence response selectivity of LGL-4 (3 μM) for β-LG (10 μM) compared to common potential interfering substances in dairy products, with tested species including metal ions (K). + Na + Mn 2+ Mg 2+ Ca 2+ Fe 2+ Cu 2+ Fe 3+ Al 3+ B is the fluorescence response of LGL-4 (3 μM) to β-LG (10 μM) in the pH range of 3–11, indicating stable performance within the physiological range; C is the reaction kinetic monitoring results, showing the fluorescence enhancement over time when LGL-4 (3 μM) and β-LG (10 μM) are mixed; D is the temperature stability assessment results of the LGL-4 / β-LG complex in the range of 4–100 °C, indicating that it maintains sensitivity under common storage and processing conditions; E is the salt tolerance assessment results at different NaCl concentrations (0–1 M); all measurements were performed in PBS buffer (10 mM, pH 7.4), λ ex =528nm, λ em =682nm.

[0037] Figure 8 This is a graph showing the fluorescence response of the LGL-4 / β-LG complex (3 μM LGL-4, 10 μM β-LG) measured in the presence of various potential interfering antibiotics (10 μM each).

[0038] Figure 9 This is a computational analysis result of the LGL-4 specificity for the off-target protein α-lactalbumin (α-LA), showing the lowest energy binding conformation of LGL-4 with α-LA predicted by molecular docking simulation, and the calculated binding free energy (ΔG). bind The concentration of LGL-4 was -5.9 kcal / mol, indicating that the interaction with α-LA was significantly weaker than that with the target β-LG, thus confirming the high selectivity of LGL-4.

[0039] Figure 10This figure shows the results of the study on the AIE-TICT behavior and β-LG binding characteristic mechanism of LGL-4; where A is the frontier molecular orbital distribution of LGL-4 obtained by DFT calculation (B3LYP / 6-31G*), showing the spatial separation between the HOMO (located on the triphenylamine donor) and LUMO (located on the quinolinon acceptor), and the calculated charge transfer parameter d. CT =12.31Å, Δ μ =11.62D; B is the TICT key coordinates identified through constrained geometry optimization scanning, revealing C 14 -C 16 The bond is the main torsional degree of freedom; C shows the relaxation of the S1 potential energy surface as a dihedral angle function of TICT and the oscillator intensity distribution, proving the formation of the dark state and the high emission probability of the planar geometry at θ=92°; D is a schematic diagram of the excited potential energy change, proving the competition between the radiative decay of the LE state and the non-radiative decay through the TICT state, activating the potential barrier. =0.21eV; E is the isothermal titration calorimeter and integral combined isotherm of LGL-4 titration of β-LG (see figure below), yielding... K d =1.27±0.08μM and n=1.13±0.05; F is a graphical analysis of continuously varying data, confirming a 1:1 binding stoichiometry between LGL-4 and β-LG in the LGL-4 / β-LG complex, with a maximum value X L =0.48±0.02; G is the time-resolved fluorescence decay curve monitored by TCSPC; H is the time-resolved fluorescence anisotropic decay of the LGL-4 / β-LG complex; I is the circular dichroism spectrum of β-LG (20 μM) in the presence and absence of LGL-4.

[0040] Figure 11 This diagram presents a schematic of a paper-based microfluidic analyzer based on the LGL-4 microarray and its validation results. A is a schematic of the microfluidic analyzer architecture; B is a representative fluorescence image of the detection region under 365nm UV excitation, showing the concentration-dependent response to β-LG (0~25μM); C is a quantitative calibration curve based on a smartphone, plotting the relationship between the ratio RGB parameter lg(R / G+R / B) and the β-lg concentration, showing a linear response (R²=0.9978); D is the temperature stability evaluation result of the microfluidic analyzer, showing consistent detection results for 20μM β-LG in spiked milk samples under storage conditions of 3~40℃, validating cold chain-free operation; E is the shelf-life evaluation result of the microfluidic analyzer, demonstrating that its analytical performance for detecting 10μM β-LG remains unchanged within 12 days of environmental storage. Detailed Implementation

[0041] This invention provides a paper-based microfluidic analytical device for the detection of β-lactoglobulin, its preparation method, and its application. To make the objectives, technical solutions, and effects of this invention clearer and more explicit, specific embodiments are provided below to further illustrate this invention.

[0042] Example 1 This embodiment provides a fluorescent probe for the detection of β-lactoglobulin and its preparation method, as detailed below: (1) Synthesis of compound 1 and compound 2: ; Compound 1 and compound 2 were synthesized according to the methods described in references 1 (DOI: 10.3390 / molecules28217309) and 2 (DOI: 10.1021 / acs.analchem.0c03199).

[0043] (2) Synthesis of compound LGL-1: ; Compound 1 (299.2 mg, 1 mmol) and compound 3 (1-ethyl-4-methylquinoline iodide, CAS: 605-59-4, 300.0 mg, 1 mmol) were dissolved in anhydrous DMF (2 mL), and two drops of methanesulfonic acid (CH3SO3H) were added. The mixture was reacted at 150 °C for 3 hours under nitrogen. After cooling to room temperature, the reaction mixture was poured into saturated brine, and the precipitate was collected by filtration. The crude product was purified by silica gel chromatography (dichloromethane / methanol) to give LGL-1 in 46% yield. The structure of LGL-1 was determined by... 1 H NMR, 13 Verification was performed using C NMR and HRMS, and the structural characterization data are as follows: 1 H NMR (DMSO- d 6, 400 MHz, ppm) δ 8.91 (d, J =6.8 Hz, 1 H), 8.38 (d, J = 8.9 Hz, 1 H), 8.29 (d, J = 8.2 Hz, 1 H), 8.10 (d, J = 3.1 Hz, 1 H), 7.95 (d, J = 3.8 Hz, 1 H), 8.63 (d, J = 3.3 Hz, 1 H), 7.46 (d, J= 3.7 Hz, 2 H), 7.30-7.27 (m, 4 H), 7.14-7.11 (m, 6 H), 7.06-6.97 (m, 3 H), 6.95(d, J = 15.8 Hz, 1 H), 6.93 (d, J = 16.1 Hz, 1 H), 6.82 (d, J = 9.4 Hz, 1 H), 4.72 (q, J = 4.2 Hz, 2 H), 1.56 (t, J = 4.5 Hz, 3 H). 13 C NMR (101 MHz, DMSO- d 6) δ (ppm) = 154.60, 147.58 (2C), 147.18, 144.71, 143.46, 137.68, 135.09, 132.89, 131.01, 130.85, 130.13, 129.43, 129.36, 129.00 (4C), 128.89 (2C), 126.85, 125.09 (4C), 124.93 (2C), 123.38, 121.79 (2C), 121.58, 54.00, 12.50. HRMS calculated molecular weight [M] + The value was 453.2325, and the found value was 453.2337.

[0044] (3) Synthesis of compound LGL-2: ; Compound 1 (299.2 mg, 1 mmol) and compound 4 (1,2-dimethylquinoline iodide, CAS: 876-87-9, 285.3 mg, 1 mmol) were dissolved in anhydrous ethanol (10 mL), and piperidine (0.10 mmol) was added. The mixture was reacted at 85 °C for 5 hours under nitrogen. After cooling to room temperature, the reaction mixture was poured into saturated brine, and the precipitate was collected by filtration and washed with water and ethanol. The crude product was purified by silica gel chromatography (dichloromethane / methanol) to give LGL-2 in 41% yield. The structure of LGL-2 was determined by... 1 HNMR, 13 Verification was performed using C NMR and HRMS, and the structural characterization data are as follows: 1 H NMR (DMSO- d 6, 400 MHz, ppm) δ8.69 (d, J= 6.8 Hz, 1 H), 8.32 (d, J = 9.1 Hz, 1 H), 8.23 ​​(d, J = 10.1 Hz, 1 H), 7.97 (d, J = 6.4 Hz, 1 H), 7.92 (d, J = 5.1 Hz, 1 H), 7.69 (d, J = 16.7 Hz, 1 H), 7.47 (d, J = 15.4 Hz, 2 H), 7.30-7.26 (m, 4 H), 7.14-7.11 (m, 6 H), 7.08-7.00(m, 5 H), 6.93(d, J = 14.9 Hz, 1 H), 4.44 (s, 3 H). 13 C NMR (101 MHz, DMSO- d 6) δ(ppm) = 150.70, 147.58 (2C), 147.43, 147.18, 140.58, 139.57, 137.93, 132.24, 131.87, 130.77, 129.46, 129.14, 129.00 (4C), 128.89 (2C), 127.87, 127.43, 125.09 (4C), 124.93 (2C), 121.79 (2C), 121.26, 119.33, 40.72. HRMS calculated molecular weight [M] + The value was 439.2169, and the found value was 439.2173.

[0045] (4) Synthesis of compound LGL-3: ; Compound 2 (359.4 mg, 1 mmol) and compound 3 (1-ethyl-4-methylquinoline iodide, CAS: 605-59-4, 300.0 mg, 1 mmol) were dissolved in anhydrous DMF (2 mL), and two drops of methanesulfonic acid (CH3SO3H) were added. The mixture was reacted at 150 °C for 3 hours under nitrogen. After cooling to room temperature, the reaction mixture was poured into saturated brine, and the precipitate was collected by filtration. The crude product was purified by silica gel chromatography (dichloromethane / methanol) to give LGL-3 in 49% yield. The structure of LGL-3 was determined by... 1 H NMR, 13 Verification was performed using C NMR and HRMS, and the structural characterization data are as follows: 1H NMR (DMSO- d 6, 400 MHz, ppm) δ 8.89 (d, J =5.7 Hz, 1 H), 8.39 (d, J = 7.7 Hz, 1 H), 8.29 (d, J = 10.8 Hz, 1 H), 8.10 (d, J =3.7 Hz, 1 H), 7.95 (d, J = 5.9 Hz, 1 H), 7.63 (d, J = 5.5 Hz, 1 H), 7.47 (d, J =6.1 Hz, 2 H), 7.16-7.13 (m, 6 H), 7.02 (d, J = 6.8 Hz, 1 H), 6.96-6.90 (m, 6 H), 6.85(d, J = 12.8 Hz, 1 H), 4.72 (q, J = 4.2 Hz, 2 H), 3.78 (s, 6 H) 1.56 (t, J = 4.5 Hz, 3 H). 13 C NMR (101 MHz, DMSO- d 6) δ (ppm) = 159.10 (2C), 156.60, 147.10, 144.71, 143.46, 140.98 (2C), 137.68, 135.09, 132.89, 131.01, 130.85, 130.13, 129.43, 129.36, 128.89 (2C), 126.85, 124.30 (4C), 123.38, 121.79 (2C), 121.58, 114.64 (4C), 55.39 (2C), 54.00, 12.90. HRMS calculated molecular weight [M] + The value was 513.2537, and the found value was 513.2543.

[0046] (5) Synthesis of compound LGL-4: ; Compound 2 (359.4 mg, 1 mmol) and compound 4 (1,2-dimethylquinoline iodide, CAS: 876-87-9, 285.3 mg, 1 mmol) were dissolved in anhydrous ethanol (10 mL), and piperidine (0.10 mmol) was added. The mixture was reacted at 85 °C for 5 hours under nitrogen. After cooling to room temperature, the reaction mixture was poured into saturated brine, and the precipitate was collected by filtration and washed with water and ethanol. The crude product was purified by silica gel chromatography (dichloromethane / methanol) to give LGL-4 in 34% yield. The structure of LGL-2 was determined by... 1 HNMR, 13 Verification was performed using C NMR and HRMS, and the structural characterization data are as follows: 1 H NMR (DMSO- d 6, 400 MHz, ppm) δ8.69 (d, J = 4.1 Hz, 1 H), 8.32 (d, J = 6.3 Hz, 1 H), 8.23 ​​(d, J = 10.1 Hz, 1 H), 7.96 (d, J = 4.2 Hz, 1 H), 7.90 (d, J = 3.3 Hz, 1 H), 7.68 (d, J = 15.2 Hz, 1 H), 7.48 (d, J = 16.3 Hz, 2 H), 7.16-7.13 (m, 6 H), 7.08-7.03 (m, 3 H), 6.95-6.90(m, 5 H), 6.93(d, J = 14.9 Hz, 1 H), 4.44 (s, 3 H), 3.78 (s, 6 H). 13 C NMR (101MHz, DMSO- d 6) δ (ppm) = 159.10 (2C), 150.70, 147.43, 147.10, 140.98 (2C), 140.58, 139.57, 137.93, 132.24, 131.87, 130.77, 129.46, 129.14, 128.89 (2C), 127.87, 127.43, 124.30 (4C), 121.79 (2C), 121.26, 119.33, 114.64 (4C), 55.39 (2C), 40.72. HRMS calculated molecular weight [M] +The value was 499.2380, and the found value was 499.2397.

[0047] Example 2 This embodiment performs performance testing and analysis on the fluorescent probe prepared in Example 1, as detailed below: (1) The specific methods include: 1. Methods for plotting Job's plot curves While keeping the total concentration of protein and fluorescent probe constant, the ratio of protein to fluorescent probe in the solution was changed, and a Job's plot curve was drawn with fluorescence intensity as the vertical axis and the protein and probe ratio as the horizontal axis. The binding stoichiometry was obtained based on the change in fluorescence intensity.

[0048] 2. Determination of fluorescence quantum yield Fluorescence quantum yield of fluorescent probes ( The determination was performed in dilute solution using the standard relative method, with Rhodamine B in ethanol as a reference. =0.65), and calculated using the following equation (reference DOI: 10.1016 / j.snb.2023.134008): ; in, and The fluorescence quantum yields of Rhodamine B and the fluorescent probe prepared in Example 1 are represented, respectively. and The absorbance intensities at 570 nm for Rhodamine B and the fluorescent probe prepared in Example 1 are respectively (dilution solution). and (between 0.01 and 0.1). and The values ​​represent the integrated emission intensities of Rhodamine B and the sample, respectively. Value by / and / The slope of the curve (at least 5 points) is given. and This is the refractive index of the solvent. (Subscript) S and R The fluorescent probe and Rhodamine B prepared in Example 1 are represented respectively.

[0049] 3. Rotational energy barrier ( Determination of ) Temperature-dependent fluorescence of LGL-4 (3 μM) was recorded in acetonitrile at 278–333 K. At each temperature... TThe relative quantum yield was measured on samples with oxygen control, background subtraction, and internal filter correction. Amplitude-weighted lifetime (Reference DOI: 10.1038 / s41467-023-40716-w). The radiative and non-radiative rate constants are obtained by the following equation: ; ; Throughout the research scope It is nearly constant, therefore confirmed. The changes mainly come from .

[0050] By With 1 / T (K) -1 Fitting to a straight line, performing Arrhenius analysis on the non-radiative channels, and calculating using the following formula: ; Among them, the slope is from It is concluded that .

[0051] 4. Determination of the detection limit Fluorescence intensity of the probe was recorded at a series of protein concentrations, with three independent replicates for each level (n=3). For each concentration, a calibration curve (intensity versus protein concentration) was constructed using the average intensity. Slope k It is obtained through linear least squares regression within the linear response range. The limit of detection (LOD) is calculated as follows: ; in, This is the standard deviation of probe fluorescence intensity in the absence of protein (blank). This standard corresponds to a signal three times the baseline noise level.

[0052] (2) The test and analysis results specifically include: 1. Photophysical characterization of fluorescent probes The four fluorescent probes (LGL-1~4) prepared in Example 1 all exhibited extensive charge transfer absorption in a water-miscible medium, and emitted light in both the visible and red light regions. Figure 2 (A and B). Normalized spectra show absorption peaks between 496-528 nm and emission peaks between 661-682 nm, accompanied by large Stokes shifts (154-165 nm), which is consistent with the characteristics of D-π-A ICT.

[0053] To investigate the AIE properties of these probes, this embodiment uses probes with different THF volume fractions (ƒ). THF The fluorescence behavior of the probe was systematically monitored in a water / THF mixture. Figure 2 As shown in Figure C, in pure aqueous solution, all probes show almost no fluorescence, but the fluorescence intensity increases significantly with the gradual addition of THF. This fluorescence enhancement effect is most pronounced in LGL-4, at ƒ THF When =90%, its fluorescence on-state ratio ( I / I 0) is 643 ( Figure 2 (D). Simultaneously, dynamic light scattering (DLS) measurements were performed in this embodiment to characterize the particle size distribution in different solvent environments. In a high water content environment (ƒ... THF In the mixture with 0% THF, no obvious aggregation was observed, indicating that the probe exists in a well-solvated molecular form. Conversely, in the mixed system with higher THF content (ƒ), no aggregation was observed. THF In the case of 99%, due to poor solvent conditions, the probe self-assembled into obvious aggregates. Figure 2 (E). This aggregated emission characteristic is a typical feature of the AIE phenomenon.

[0054] To evaluate the TICT effect of these probes, this example investigated the probes' response to ambient viscosity. In a series of solvents with progressively increasing viscosity (dichloromethane (η=0.44 cP) < ethanol (η=1.1 cP) < ethylene glycol (η=16.0 cP)), a significant viscosity-dependent increase in fluorescence intensity was observed for all probes. Figure 2 (F~I). For compound LGL-4, the relative fluorescence quantum yield (Φ) fl The emission value jumps from 0.01 in dichloromethane to 0.32 in ethylene glycol. Notably, the emission maximum remains largely unchanged, indicating that viscosity primarily suppresses the nonradiative de-excitation pathway (i.e., the formation of TICT states) rather than altering the energy of the emission locally excited (LE) states.

[0055] Rate constant analysis further supports this mechanism. Radiation was measured in different solvents. ) and non-radioactive ( The decay rate constant is used to elucidate the mechanistic basis of probe behavior. For example... Figure 2 As shown in Figure J, increasing solvent viscosity has an effect on... The impact is negligible, but it will lead to A significant decrease. For LGL-4, 6.08 ns from dichloromethane -1 Reduced to 0.22 ns in ethylene glycol -1This represents a 27-fold reduction. This selective quenching of nonradiative channels is a hallmark of the RIM-dominated mechanism, in which a viscous environment physically hinders the intramolecular rotation required to form the “dark” TICT state.

[0056] Quantitative viscosity sensitivity was determined using Förster-Hoffmann analysis. Within the moderate viscosity range (1–100 cP), lgI and lgη showed a linear relationship, with the slope from LGL⁻¹ to LGL⁻⁴ (…). The values ​​were 0.76, 0.88, 0.89, and 0.95, respectively, for all R values. 2 All ≥0.99 ( Figure 2 (Middle K). The monotonically increasing value indicates that the rotor characteristics are gradually enhanced, and the coupling between intramolecular torsion and nonradiative decay is becoming increasingly tight.

[0057] Temperature dependence studies conducted in acetonitrile (278~333K) showed that It exhibits a linear Arrhenius relationship with 1 / T ( Figure 2 (L). As the temperature increases, all probes... Both increased; LGL-4 showed the smallest slope and the lowest rotational energy barrier. Consistent (0.23 eV for LGL-4, 0.26–0.28 eV for LGL-1–3). This low barrier makes LGL-4 exceptionally sensitive to the environment; it ensures that the probe is in a “closed” state in solution, but also means that even with partially restricted movement within the protein groove, this simple nonradiative pathway can be effectively blocked, thereby generating a maximized “on” signal.

[0058] In summary, these spectral data demonstrate that the fluorescent probes of Example 1 construct an AIE-TICT framework that combines tunable viscosity sensitivity with strong environmentally dependent "on" capability. Among them, LGL-4 possesses the highest dynamic range and is the preferred probe for cavity activation sensing.

[0059] 2. Computer simulation analysis of probe-protein interactions To gain a deeper understanding of the observed sensing behavior, molecular docking simulations were performed to model the interactions between each probe (LGL-1~4) in Example 1 and the typical hydrophobic calyx structure of β-LG. The simulations revealed significant differences in binding modes and affinities across the entire probe family. Figure 3 ).

[0060] All four probes were dimensionally compatible with the β-LG calyx, but only LGL-2, LGL-3, and LGL-4 formed stable deep-binding conformations. These probes tended towards a typical “head-first” binding orientation, where the cation acceptor group penetrated deep into the hydrophobic cavity, while the large TPA donor portion remained near the solvent-exposed inlet. Crucially, this geometry aligned the central D-π-A torsion axis directly with the narrowest part of the cavity (the “contraction site”), where steric hindrance from adjacent residues was maximized, thus restricting its rotational degrees of freedom. In stark contrast, LGL-1 failed to achieve stable insertion. Its predicted conformation showed the cation group hovering on the protein surface, making only surface and brief van der Waals contacts with the calyx edge. These unique binding modes were quantified in the calculated binding affinity (docking fraction), showing a trend toward increasing stability: -6.7 (LGL-1), -8.1 (LGL-2), -8.8 (LGL-3), and -10.0 kcal·mol⁻¹. -1 (LGL-4). Detailed analysis of intermolecular contacts provided a structural basis for this trend. The shallow binding of LGL-1 was primarily dominated by nonspecific hydrophobic interactions. Conversely, the deep binding of LGL-2–4 enabled its extended π-conjugated system to establish favorable π-π and π-σ stacking interactions with aromatic residues (such as Phe105) in the β-LG cavity. Furthermore, the methoxy substituents on LGL-3 and LGL-4 acted as hydrogen bond acceptors, forming crucial interactions with the side chain of Lys69 in the protein neck (O···HN distances of 3.23 Å ​​and 3.02 Å, respectively). This specific hydrogen bond played a crucial anchoring role, further locking the probe's torsional axis at the position of maximum spatial constraint.

[0061] Analysis of the electrostatic potential surface confirmed the above findings. Figure 4 The significant negative potential field at the β-LG calyx effectively guided the position of the probe's cation head. Notably, the electrostatic complementarity gradually increased from LGL-1 to LGL-4, consistent with the optimal localization of LGL-4, which maximizes direct ion interactions. Overall, these docking results provide a reliable molecular model explaining the superior performance of LGL-4, attributing its high on-state ratio to deep hydrophobic encapsulation, specific hydrogen bonding, and strong electrostatic complementarity—all factors working together to firmly constrain the probe to its flexible rotor.

[0062] 3. Quantification of probe protein binding affinity and sensing performance Fluorescent titration was used to quantitatively assess the binding interaction between the probes (LGL-1~4) prepared in Example 1 and β-LG. Stepwise addition of β-LG (0~2 μM) to each probe solution resulted in a single, saturated fluorescence enhancement. Figure 5 (A, D, G, J). Throughout the titration, the emission maximum remained constant, and no spectral changes or linear alterations were detected, indicating a clear transition from a non-emission-free state to a single high-emission probe-protein complex. Notably, among all probes, LGL-4 exhibited the most pronounced response, achieving an 809-fold fluorescence enhancement upon β-LG saturation. Visual examination of the LGL-4 solution revealed a sharp color shift from dark red to intense bright red fluorescence with increasing protein concentration. Figure 6 Colorimetric analysis further confirmed the spectral purity of the response, and the fluorescence coordinates showed a linear trajectory on the CIE plot. Figure 5 (B, E, H, K).

[0063] The obtained titration curve was subjected to nonlinear regression analysis and processed according to a 1:1 binding model to extract the binding constant ( ). K a Analysis results show that LGL-1's K a The value is 1.27 × 10 4 M -1 LGL-2 K a The value is 1.56 × 10 4 M -1 LGL-3 K a The value is 4.18 × 10 4 M -1 ; while LGL-4's K a The value was significantly higher, reaching 7.38 × 10⁻⁶. 5 M -1 ( Figure 5 (C, F, I, L). Notably, the titration data for LGL-4 were very well described by the unit-point combination model, exhibiting a stable goodness of fit (R²) across the entire concentration range. 2 >0.99). In contrast, the goodness of fit of LGL-1, LGL-2, and LGL-3 showed greater dispersion, a characteristic consistent with their weaker binding affinity.

[0064] Crucially, the affinity tendency (LGL-1) determined through experiments... <LGL-2<LGL-3 The binding energy of LGL-4 closely matches the predictions of molecular docking simulations, providing strong experimental validation for the binding model proposed in this invention. The superior affinity of LGL-4 confirms its deeper insertion and specific anchoring interaction in the β-LG calyx, directly leading to the formation of a more stable complex.

[0065] Based on the high affinity and significant fluorescence enhancement of LGL-4, this example evaluated its analytical performance for detecting β-LG. The calculated limit of detection (LOD) was 8.2 nM. This high sensitivity highlights the potential of LGL-4 as a practical tool for the quantitative detection of β-LG.

[0066] 4. Verification of interfering factors To verify that LGL-4 is a stable and selective fluorescent probe, the selectivity of LGL-4 for β-LG was evaluated using two complementary methods. First, a direct interference experiment examined the probe's response to potential interfering substances commonly found in dairy products, including metal ions, amino acids, and common milk proteins. The results showed that common interfering substances, including bovine serum albumin (BSA), casein, and α-lactalbumin, did not elicit significant fluorescent responses, and the signal intensities produced by other proteins were less than 2% of the signal produced by an equimolar amount of β-LG. Figure 7 (A). Secondly, competitive binding experiments assessed whether commonly used antibiotics in dairy farming would disrupt the established LGL-4 / β-LG complex. The results showed that commonly used antibiotics failed to interfere with the binding of the probe to the protein, indicating that this LGL-4 / β-LG interaction is stable. Figure 8 Furthermore, computational models confirmed this specific selectivity; docking simulations of LGL-4 with α-lactalbumin predicted a weak and superficial binding conformation (-5.9 kcal·mol⁻¹). - ¹), which is energy disadvantageous compared to the LGL-4 conformation deeply inserted into the groove of β-LG. Figure 9 ).

[0067] Subsequently, the effects of key environmental parameters on the probe-protein complex were investigated. pH-dependent analysis showed that the complex exhibited optimal performance over a wide physiological range (pH 5.5 to 8.0), while the signal intensity decreased at lower and higher pH values. Figure 7 (B). This phenomenon is consistent with the known pH-dependent characteristics of β-LG calyx, in which protonation regulates electrostatic potential and overall protein conformation, which is crucial for probe recognition and binding.

[0068] Kinetic analysis showed that the binding process was rapid. After mixing LGL-4 (3 μM) and β-LG (10 μM), the fluorescence signal reached over 95% of its maximum intensity within 100 seconds and remained stable thereafter, confirming the rapid binding equilibrium and excellent photostability under continuous illumination. Figure 7 (C)

[0069] To confirm the dependence on correctly folded protein structure, the thermal stability of the complex was evaluated. The fluorescence signal remained high and stable in the range of 4–40 °C, but exhibited a sharp S-shaped decrease at higher temperatures, characteristic of protein thermal denaturation. Figure 7 (Middle D). This result directly links fluorescence activation to the structural integrity of the β-LG calyx, because unfolding disrupts the binding bag and releases the probe into a non-emission state.

[0070] Finally, the crucial role of electrostatics was confirmed by titrating the complex with NaCl. The fluorescence signal gradually decreased with increasing ionic intensity up to 1.0 M. Figure 7 (E). This attenuation is mainly attributed to the shielding of electrostatic attraction between the cationic portion of the probe and the negatively charged residues at the calyx edge, which is crucial for guiding and stabilizing the binding state.

[0071] In summary, the above experiments provide corresponding and stable validation of the proposed sensing mechanism. High selectivity, rapid kinetics, and predictable dependence on pH, temperature, and ionic strength confirm that the fluorescence response of LGL-4 is based on its specific, conformation-dependent deep binding with β-LG. It is precisely in this sterically confined and electrostatically favorable environment that the rotation within the LGL-4 molecule is effectively restricted, thereby initiating a strong fluorescence response.

[0072] 5. Mechanism Explanation To elucidate the molecular basis of the AIE-TICT phenomenon in LGL-4, comprehensive quantum chemical calculations were performed in this embodiment. Frontier molecular orbital analysis revealed a significant spatial separation between the HOMO (primarily located in the TPA donor region) and the LUMO (concentrated in the cation acceptor unit). This significant charge separation manifests as a considerably large charge transfer distance (d). CT =12.31Å) and significant changes in excited-state dipole moment (Δ μ =11.62D), clearly confirming the ICT characteristics of LGL-4 ( Figure 10 (A)

[0073] System analysis of the torsional potential surface revealed that C 14 -C 16 The bond is the key coordinate that triggers TICT. A relaxation potential scan along the S1 surface reveals a non-emitting TICT state in the near-vertical conformation (θ = 92°), in stark contrast to the highly emitting planar and quasi-planar conformations, which exhibit considerably greater oscillator strength (at θ = -6°). f =0.83)( Figure 10 (B~C). The minimum value of the LE state and the dark TICT ( A moderate energy barrier (0.21 eV) facilitates efficient nonradiative decay in low-viscosity polar media, while maintaining repressibility through conformational confinement within spatially confined hydrophobic binding sites. This is the core mechanism behind the observed AIE-TICT behavior. Figure 10 (D). It is worth noting that the calculated barrier height is in very good agreement with the experimental results. =0.23eV) Figure 2 (L). Maximum experimental absorption (λ) max =528nm; E abs =2.35eV) corresponds to the process of vertical excitation into the LE state. After that, the population of the excited state can be distributed between radiative transition and crossing the energy barrier to enter the dark TICT state. Its branching ratio is determined by both the ambient viscosity and polarity.

[0074] Thermodynamic characterization was performed using isothermal titration calorimetry (ITC), yielding a stoichiometric coefficient n = 1.13 ± 0.05 and a dissociation constant. K d =1.27±0.08μM ( Figure 10 (E). The determined dissociation constant and the value obtained by fluorescence titration ( K d =1 / K a The binding affinity was very consistent with that of Job's plot (1.35 μM), thus independently verifying the binding affinity. Furthermore, the continuous variation analysis of the Job's plot curve in X... L The maximum value is displayed when the value is 0.48 ± 0.04. Figure 10 (F). These complementary methods together confirmed a 1:1 binding stoichiometry between LGL-4 and β-LG.

[0075] Time-resolved fluorescence spectroscopy revealed significant changes in photophysical properties after the formation of the LGL-4 / β-LG complex. The fluorescence lifetime increased from the τ of the free dye. f =0.09±0.01ns (close to the value of the instrument response function) increases to τ in the protein-bound state. f =3.3±0.1ns ( Figure 10 (G), which corresponds to the non-radiative decay pathway being suppressed by more than 30-fold. The rotational relaxation time τ is obtained by fitting the anisotropic decay trajectory using a single exponential function. r =14.2±0.1ns ( Figure 10 The concentration of H in β-LG significantly exceeded expectations for free dyes, indicating substantial motility restriction within the protein-binding pouch. Furthermore, the circular dichroism spectrum of β-LG remained almost unchanged after the addition of LGL-4. Figure 10 (I) indicates that the formation of the complex does not disrupt the protein's native secondary structure.

[0076] 6. Comparison with existing β-LG fluorescent probes Further comparisons were made between LGL-4 and existing β-LG fluorescent probes, and the results are shown in Table 1.

[0077] Table 1. Comparison of fluorescent probe performance for β-LG detection

[0078] Example 3 To translate the high performance of the LGL-4 into a practical, field-operable form, this embodiment designs a multi-region paper-based microfluidic analysis device (µPAD). Details are as follows: The μPAD was constructed using Whatman CF4 chromatographic paper due to its uniform pore structure and reproducible capillary flow characteristics. A 125 μm thick polyethylene terephthalate (PET) support layer was used to ensure structural integrity and prevent leakage into the device. Functional regions were constructed sequentially by controlling reagent solution deposition using a calibrated micropipette. During the reagent solution deposition step of adjacent regions, the regions were dried at 37°C for 60 minutes under light protection.

[0079] This paper-based microfluidic analyzer integrates the complete sample-to-response workflow into a single capillary-driven test strip. Figure 11(A). It comprises five consecutive regions: (i) the sample application zone, (ii) the acidification zone, (iii) the deposition zone, (iv) the neutralization zone, and (v) the detection zone. The sample application zone undergoes no additional treatment; the acidification zone is treated with 20.0 μL of citric acid / citrate buffer (0.50 M, pH 4.6) to induce casein precipitation upon sample contact; the deposition zone undergoes no additional chemical modification, and CF4 chromatographic paper is laminated onto primary CF4 chromatographic paper to mechanically filter the precipitated protein aggregates and retain the curd; the neutralization zone is treated with 25 μL of Tris buffer (0.50 M, pH 7.4) to provide a buffer capacity of 12.5 μmol to restore the physiological pH conditions before detection; the detection zone is functionalized with 40.0 μL of an optimized probe formulation, which is a PBS (10 mM, pH 7.4) solution containing LGL-4 (0.50 mM), supplemented with trehalose (10% w / v) as a lyophilization protectant, and polyvinylpyrrolidone-40 (0.20% w / v) as a film-forming agent. In this deposition scheme, 20.0 nmol of probe loading was achieved in each paper-based microfluidic analyzer, while simultaneously forming a uniform glassy matrix. This eliminated the coffee ring effect and ensured the stability of the probe during storage at room temperature. DMSO (≤1% v / v) was added to all reagent solutions to enhance paper wettability without disrupting the hydrophobic wax barrier used to define flow path boundaries.

[0080] When a milk sample is applied, it flows autonomously through the test strip, undergoes pretreatment on the device to produce a clear, pH-adjusted whey fraction, and reaches the detection area for the detection of β-LG. The specific measurement method is as follows: 1.0 mL of milk sample diluted 10 times with deionized water is loaded into the sample application area of ​​the μPAD using a micropipette. The sample migrates in the μPAD for 5 minutes at room temperature (25°C) via capillary action. After sample migration is complete, the detection area is illuminated in a dark environment using a portable UV lamp (365 nm). A fluorescence image is captured using a smartphone camera (exposure time: 1 / 4 second, ISO: 400) positioned 10 cm above the device. The RGB values ​​of the detection area are extracted using Color Grab software, and the ratio parameter lg(R / G + R / B) is calculated.

[0081] The μPAD exhibited a significant concentration-dependent fluorescence response to β-LG (0~25 μM). Figure 11 (B). This ratio output produced an excellent linear calibration curve (R) across the entire operating range. 2 =0.9978), proving its suitability for quantitative analysis ( Figure 11(C). Crucially, the µPAD's detection limit was calculated to be 0.076 μM. This sensitivity not only exceeds the 0.1 μM threshold typically required for food safety analysis methods but is also far below the 5.4 μM safety threshold set to protect individuals with milk allergies, highlighting the device's practicality in allergen screening. The µPAD's analytical performance has been rigorously validated for practical applications. Furthermore, the device exhibits excellent thermal stability; the analytical signal of a 20 μM β-LG sample shows negligible variance over a wide temperature range (3–40 °C), making it suitable for refrigerated and environmentally sensitive samples. Figure 11 (D). Furthermore, long-term stability was confirmed by analyzing the same dairy samples over 12 days; the device exhibited excellent daily reproducibility with minimal signal drift and a relative standard deviation (RSD) of ≤5.0% for intra-day measurements, validating the stability and reliability of this paper-based microfluidic analyzer for quantitative analysis. Figure 11 (E).

[0082] Example 4 This embodiment uses the paper-based microfluidic analysis device prepared in Example 3 to detect real samples, as detailed below: First, method validation was performed using β-LG-free infant formula as a typical complex matrix, and a standard addition recovery experiment was conducted. After sample dissolution, a predetermined concentration of β-LG standard was added to the matrix, and the matrix was thoroughly homogenized before μPAD analysis. The assay method was the same as in Example 3, with fluorescence images of the detection area acquired under 365 nm UV excitation, and quantification performed using a smartphone-based ratio parameter lg(R / G+R / B) calibrated according to external standards. Recovery studies (Table 2) showed excellent accuracy (98–108%) across the entire analytical range, with RSD ≤ 4.7% (n=3), while the background signal from the blank sample was negligible, confirming the specificity and matrix compatibility of the method.

[0083] Table 2 Peak recovery of β-LG in β-LG-free infant formula

[0084] Subsequently, the method performance was evaluated using seven commercial dairy products, including different formulations: two conventional milk samples (skimmed and whole milk), one ultra-high temperature (UHT) processed milk, one yogurt beverage, and three infant formula formulations (including powder variants).

[0085] Commercial liquid dairy samples (milk, yogurt beverages, and UHT-processed products) were prepared by direct dilution. Briefly, 1.0 mL of the original sample was transferred to a microcentrifuge tube and diluted 10-fold with 9.0 mL of deionized water. The diluted sample was vortexed for 30 seconds to ensure homogeneity. Powdered samples (infant formula and milk powder) required reconstitution before analysis. 10.0 ± 0.1 g of powder was accurately weighed using an analytical balance and transferred to a 250 mL flask. Deionized water (100 mL) was gradually added while stirring to prevent clumping. The mixture was vigorously shaken for 5 minutes, then sonicated for 10 minutes to ensure complete dissolution. For all samples, centrifugation at 10000 × g for 3 minutes at 4°C was performed to remove the lipid layer before analysis.

[0086] After appropriate dilution, samples were analyzed in parallel using the μPAD and HPLC methods described in Example 3. The HPLC method employed C18 reversed-phase separation, UV detection at 214 nm, and calibration against an external β-LG standard. Comparative analysis (Table 3) showed excellent consistency between the methods within the original matrix concentration range of approximately 30–170 μM, with absolute deviations typically below 9%. This strong correlation confirms the previously elucidated selective binding mechanism and validates the applicability of the μPAD from Example 3 for routine β-LG screening and quality control in dairy processing.

[0087] Table 3. Applications of β-LG in commercial dairy products: μPAD and HPLC

[0088] In Table 3, the deviation (%) = (μPAD - HPLC) / HPLC × 100%.

[0089] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A paper-based microfluidic analytical device for the detection of β-lactoglobulin, characterized in that, include: The base plate and the absorbent paper loaded on the base plate; The absorbent paper includes, in sequence along the liquid flow direction: a sample application zone, an acidification zone, a deposition zone, a neutralization zone, and a detection zone; The acidification zone contains a buffer system with a pH range of 3 to 5; The sedimentation zone is superimposed with at least one layer of absorbent paper; The neutralization zone contains a buffer system with a pH range of 6 to 8; The detection area contains a fluorescent probe for the detection of β-lactoglobulin; The fluorescent probe has the following structure: R1 and R2 are independently selected from hydrogen or C1~C5 alkoxy groups, and R3 is methyl or ethyl.

2. The paper-based microfluidic analysis device according to claim 1, characterized in that, The molar ratio of the buffer system with a pH range of 3 to 5, the buffer system with a pH range of 6 to 8, and the fluorescent probe is 1:(1 to 1.5):(0.002 to 0.005). The total molar amount of the buffer system with a pH range of 3 to 5 is 1 to 50 μmol; The total molar amount of the buffer system with a pH range of 6 to 8 is 1 to 75 μmol; The total molar amount of the fluorescent probe is 2~250 nmol.

3. The paper-based microfluidic analysis device according to claim 1, characterized in that, The buffer system with a pH range of 3 to 5 is selected from one or more of the following: formic acid-formate buffer system, acetic acid-acetate buffer system, and citric acid-citate buffer system; The buffer system with a pH range of 6 to 8 is selected from one or more of the following: phosphate buffer system, tris(hydroxymethyl)aminomethane-hydrochloric acid buffer system, and phosphate-imidazolium buffer system. The fluorescent probe is selected from one of the following structures: 。 4. The paper-based microfluidic analysis device according to claim 1, characterized in that, The detection area also contains a freeze-drying protectant and / or a film-forming agent; The freeze-drying protectant is selected from one or more of trehalose, sucrose, lactose, chitosan, and polyols; The film-forming agent is selected from one or more of polyvinylpyrrolidone, polyethylene glycol, polyvinyl alcohol, polypropylene glycol, and polyacrylate; The total mass of the freeze-drying protectant is 1~10 mg, and the total mass of the film-forming agent is 1~20 μg.

5. The method for preparing the paper-based microfluidic analysis device according to any one of claims 1 to 4, characterized in that, Including the following steps: Place the absorbent paper onto the base plate; A first solution is added to the acidification zone of the absorbent paper, the first solution containing a buffer system with a pH range of 3 to 5; A second solution is added to the neutralization zone of the absorbent paper, the second solution containing a buffer system with a pH range of 6 to 8; A third solution containing a fluorescent probe for β-lactoglobulin detection is added to the detection area of ​​the absorbent paper. At least one layer of absorbent paper is superimposed on the deposition area of ​​the absorbent paper to obtain the paper-based microfluidic analysis device.

6. The method for preparing the paper-based microfluidic analysis device according to claim 5, characterized in that, The first solution, the second solution, and / or the third solution also contain ≤1% by volume of dimethyl sulfoxide.

7. The method for preparing the paper-based microfluidic analysis device according to claim 5, characterized in that, After any of the steps of adding a first solution to the acidification zone of the absorbent paper, adding a second solution to the neutralization zone of the absorbent paper, and adding a third solution to the detection zone of the absorbent paper, the process further includes: a drying process.

8. The method for preparing the paper-based microfluidic analysis device according to claim 7, characterized in that, The drying conditions include a temperature of 20-45°C and a time of 30-120 minutes.

9. The application of a paper-based microfluidic analysis device as described in any one of claims 1 to 4 in detecting the β-lactoglobulin content in food.

10. The application of a paper-based microfluidic analysis device as described in any one of claims 1 to 4 in the preparation of β-lactoglobulin detection products.

Images