Double-ligand lanthanide series metal organic framework Tb-MOF for rapidly detecting enoxacin and rapid detection method for enoxacin concentration

Abstract

The invention belongs to the technical field of nano material detection, and particularly discloses a double-ligand lanthanide series metal organic framework Tb-MOF for rapidly detecting enoxacin and a rapid detection method for enoxacin concentration. According to the invention, 2, 6-pyridinedicarboxylic acid and 2-aminoterephthalic acid are used as modification ligands to construct Tb-MOF, and after enoxacin is added into a system of the Tb-MOF, the fluorescence intensity and fluorescence color of the Tb-MOF are subjected to characteristic change due to an antenna effect, so that the fluorescence colorimetric sensor is constructed. A portable detection device based on a smart phone is further designed, a cloud server data analysis system integrated with a machine learning algorithm is fused with the smart phone, and collection, correction, analysis and result display of fluorescence color data can be efficiently completed. According to the invention, the capturing and processing capability of a modern smart phone on fluorescence color signals is fully utilized, and a set of solution which is convenient to operate and user-friendly is provided for on-site detection of enoxacin.

Description

Technical Field

[0001] This invention belongs to the field of nanomaterial detection technology, and particularly relates to a rapid detection method for enoxacin dual-ligand lanthanide metal-organic framework Tb-MOF and enoxacin concentration. Background Technology

[0002] Since the advent of penicillin in 1928, antibiotics have become a core tool for preventing and treating infectious diseases. Among them, fluoroquinolones (FQs), as highly effective broad-spectrum synthetic antibiotics, are widely used in the medical and livestock farming fields due to their low cost and significant bactericidal effects. However, their widespread use has triggered a serious public health and environmental crisis: on the one hand, abuse can induce drug resistance in pathogens, significantly weakening the effectiveness of drug treatment; on the other hand, FQs have a low metabolic rate in organisms, with approximately 30%-90% excreted in the form of unchanged or active metabolites into the soil and water environment via feces and urine, and subsequently accumulating in animal-derived foods through the food chain. More seriously, long-term human consumption of food containing FQ residues may not only trigger acute adverse reactions such as allergies and intestinal flora imbalance, but also pose potential risks of teratogenicity, carcinogenicity, and mutagenicity, while promoting the formation of drug-resistant strains in the human body, posing a significant challenge to the medical industry. Therefore, to ensure the safety of animal-derived foods, developing accurate and convenient rapid detection methods for FQs to effectively monitor their residue levels is of great significance.

[0003] Among the detection methods for FQs (Fragmentation Quotients), traditional mainstream methods such as high-performance liquid chromatography (HPLC) and liquid chromatography-tandem mass spectrometry (LC-MS) possess high accuracy, but they suffer from limitations such as long detection cycles, high equipment purchase and maintenance costs, and complex operation procedures, making it difficult to meet the needs of rapid detection. Furthermore, methods such as enzyme-linked immunosorbent assay (ELISA) and electrochemical analysis, while avoiding the cumbersome operation of the aforementioned traditional methods to some extent and possessing the advantages of speed and simplicity, suffer from poor selectivity and insufficient sensitivity, limiting their applicability in low-concentration residue detection scenarios. In contrast, fluorescence spectroscopy, due to its comprehensive advantages of low cost, high sensitivity, fast response speed, and simple sample pretreatment steps, has increasingly attracted the attention of researchers in the analytical field in recent years. However, most existing fluorescence sensing methods rely on large, specialized instruments, making it difficult to achieve sensitive on-site detection. Therefore, there is an urgent need to develop a portable, sensitive, and economical on-site detection technology. Summary of the Invention

[0004] To address the aforementioned problems in the prior art, the objective of this invention is to provide a rapid detection method for enoxacin dual-ligand lanthanide metal-organic frameworks (Tb-MOFs) and enoxacin concentration. This method not only provides ultrasensitive detection of enoxacin but also enables portable, on-site fluorescence visualization detection in complex matrices.

[0005] To achieve the above objectives, the first aspect of the present invention provides a dual-ligand lanthanide metal-organic framework Tb-MOF for rapid detection of enoxacin, characterized in that the ligands of the Tb-MOF are 2,6-pyridinedicarboxylic acid and 2-aminoterephthalic acid.

[0006] The second aspect of this invention provides a method for preparing a dual-ligand lanthanide metal-organic framework (Tb-MOF) for rapid detection of enoxacin, characterized by comprising: dissolving 2,6-pyridinedicarboxylic acid, 2-aminoterephthalic acid, and triethylamine in anhydrous ethanol to obtain solution A; dissolving terbium nitrate hexahydrate in deionized water to obtain solution B; subsequently, mixing solution A and solution B, stirring the reaction in a water bath at 30-50°C, and finally washing and drying to obtain the dual-ligand lanthanide metal-organic framework (Tb-MOF).

[0007] Furthermore, the precipitate was washed and purified sequentially with deionized water and anhydrous ethanol, three times each, and finally dried overnight in a vacuum oven at 40°C.

[0008] Furthermore, the ratio of the sum of the amounts of 2,6-pyridinedicarboxylic acid and 2-aminoterephthalic acid to the amount of terbium nitrate hexahydrate is 2:1.

[0009] The stirring reaction takes 1-5 hours. Preferably, the stirring time is 2 hours; the heating reaction is carried out at a temperature of 40°C.

[0010] Furthermore, the amount of 2,6-pyridinedicarboxylic acid is 0.48 mmol, the amount of 2-aminoterephthalic acid is 0.02 mmol, and the amount of terbium nitrate hexahydrate is 0.25 mmol.

[0011] The following solutions were prepared: 15 mL of anhydrous ethanol; 140 μL of triethylamine; and 5 mL of deionized water.

[0012] The molar ratio of the 2,6-pyridinedicarboxylic acid and the 2-aminoterephthalic acid modified substances is (20-30):1, preferably 24:1.

[0013] Furthermore, the Tb-MOF solution modified with 2,6-pyridinedicarboxylic acid and 2-aminoterephthalic acid showed a morphology of uniformly sized small particles tightly aggregated under a transmission electron microscope, forming dense and cohesive aggregates.

[0014] A third aspect of this invention provides a rapid method for detecting enoxacin concentration, comprising the following steps:

[0015] The lanthanide metal-organic framework Tb-MOF obtained by any of the above preparation methods is prepared into a Tb-MOF solution, which is then thoroughly mixed with the enoxacin solution to be tested. Under excitation at the excitation wavelength, the concentration of the enoxacin solution to be tested is determined based on the changes in fluorescence intensity and / or fluorescence color of the mixed system.

[0016] The fluorescence intensity is determined by constructing a standard curve of enoxacin concentration and fluorescence enhancement; the fluorescence color change is determined by collecting image information of the mixed system, determining the fluorescence color change through a machine learning model, and then completing the concentration prediction of enoxacin.

[0017] Furthermore, the method for establishing the standard relationship curve includes: mixing Tb-MOF solution with enoxacin standard solutions of different concentrations respectively; after the reaction is complete, under 365nm ultraviolet light irradiation, establishing a standard relationship curve between enoxacin solution concentration and fluorescence enhancement degree based on the change in fluorescence intensity of the corresponding mixed solution and the concentration of enoxacin solution.

[0018] The standard relationship curve is as follows:

[0019] lgY = 0.229 + 0.598lgC Eno (R 2 =0.992), where Y represents the ratio of fluorescence intensity at 545 nm to fluorescence intensity at 417 nm, and C Eno This indicates the concentration of enoxacin.

[0020] Preferably, the concentration of the Tb-MOF solution is 0.4 mg / mL, and the solvent is a Tris-HCl buffer solution (0.01 mol / L) at pH 6.

[0021] Preferably, the volume ratio of Tb-MOF solution to enoxacin solution is 3:7.

[0022] Furthermore, the reaction time between the enoxacin standard solution and the Tb-MOF solution is 15 min, and the concentration range of enoxacin in the reaction system is 0-20.00 μmol / L; preferably, the concentrations of the enoxacin solution in the reaction system are 0, 0.05, 0.08, 0.10, 0.30, 0.50, 0.70, 0.90, 1.00, 2.00, 3.00, 4.00, 5.00, 7.00, 10.00, 15.00, and 20.00 μmol / L.

[0023] Preferably, in step 1, after adding enoxacin solution to Tb-MOF solution, the fluorescence absorption peak at 545 nm is enhanced, while the fluorescence absorption peak at 417 nm changes only slightly, and the fluorescence color of the mixed solution changes from blue to green.

[0024] The method for constructing the machine learning-based model includes:

[0025] The Tb-MOF solution was mixed with standard solutions of enoxacin at different concentrations. After the reaction was complete, the image information of each mixed solution was acquired under 365nm ultraviolet light. The RGB value digital matrix of the mixed solution area in the corresponding image was extracted using MATLAB software. Finally, the relationship between the RGB value of the pixel and the concentration of enoxacin was established using the random forest model (RF).

[0026] Further, specifically: the collected photos are imported into MATLAB software, and the RGB values ​​of each pixel in the photos are extracted using a pixel extraction algorithm to generate the corresponding digital matrix; the digital matrix is ​​used as the feature input, and the corresponding enoxacin concentration is used as the output, and the RF algorithm is combined to train and build the model, and finally an RF model that can realize quantitative prediction of enoxacin concentration is obtained.

[0027] Preferably, the specific operation method is as follows: In the process of building the RF model, the original training set is first constructed based on the RGB values ​​and their corresponding enoxacin concentrations; then, the self-sampling method is used to extract samples from the original training set with replacement to generate multiple independent training subsets, each subset being used to train a decision tree model; after all decision trees are constructed, the RF algorithm averages the regression prediction results of each decision tree to obtain the final prediction value, thereby completing the construction of the RF model for enoxacin concentration prediction.

[0028] Furthermore, in the construction of the machine learning-based model, the reaction time between the enoxacin standard solution and the Tb-MOF solution is 15 min, and the concentration range of enoxacin in the reaction system is 0-20.00 μmol / L; preferably, the concentrations of the enoxacin solution in the reaction system are 0, 0.05, 0.08, 0.10, 0.30, 0.50, 0.70, 0.90, 1.00, 2.00, 3.00, 4.00, 5.00, 7.00, 10.00, 15.00, and 20.00 μmol / L.

[0029] Furthermore, it also includes predicting the concentration of enoxacin by collecting fluorescence color information through a smart terminal, specifically including the following steps:

[0030] Step 1: Build an RGB value conversion program and the trained RF model on a smart terminal;

[0031] Step 2: Mix the Tb-MOF solution with the enoxacin solution to be tested to obtain a mixed solution. After the reaction is complete, under 365nm ultraviolet light irradiation, the image information of the mixed solution is acquired by a smart terminal and the image information is converted into standardized RGB values. Then, based on the RF model, the smart terminal predicts the concentration of the enoxacin solution to be tested according to the RGB values ​​of the enoxacin solution to be tested.

[0032] Preferably, the smart terminal is a smartphone or a tablet computer.

[0033] Preferably, the method involves adding Tb-MOF solution and enoxacin standard solutions of different concentrations to the sample plate and mixing them thoroughly. A fluorescence photograph of the mixed system is taken using a smartphone, and the RGB values ​​of the mixed system region in the photograph are extracted via a WeChat mini-program. The extracted RGB values ​​are uploaded to the quantitative detection interface, and the cloud-based machine learning algorithm on the mini-program automatically outputs a predicted enoxacin concentration based on the input RGB values.

[0034] Compared with the prior art, the beneficial effects and advantages of the present invention are as follows:

[0035] This invention constructs a ratiometric fluorescence sensor based on a dual-ligand Tb-MOF for detecting enoxacin concentration. Compared with existing detection methods, this sensor offers advantages such as low cost, rapid response, high sensitivity, and high selectivity, while significantly reducing the detection limit. The dual-ligand strategy effectively broadens the linear range of enoxacin detection (0–20.00 μmol / L) and enhances the fluorescence color change from blue to green, making its sensitivity far exceed that of single-ligand MOF sensors. By identifying the fluorescence intensity and the color change of the solution, a simple, visual, and quantitative analysis of enoxacin concentration can be achieved. Attached Figure Description

[0036] Figure 1 This is a schematic diagram illustrating the principle and detection of the Tb-MOF fluorescence sensor of the present invention.

[0037] Figure 2 The structure of Tb-MOF is shown below. A represents the TEM image and EDX elemental mapping of Tb-MOF, where green represents Tb elements, red represents C elements, purple represents O elements, and yellow represents N elements. B represents the PXRD results of simulated and synthesized Tb-MOF. C represents the FT-IR results of DPA, ATA, and Tb-MOF. D represents the XPS spectrum of Tb-MOF.

[0038] Figure 3 In the diagram, A represents the fluorescence spectra of Tb-MOF in the presence of different concentrations of enoxacin (0-20.00 μmol / L) (excitation wavelength λex = 365 nm). B represents the fluorescence intensity ratio (IL). 545 / I 417 The linear relationship between enoxacin and Eno concentration is shown. C represents the CIE chromaticity diagrams in the presence of different concentrations of enoxacin (0–20.00 μmol / L). D is a partially enlarged CIE chromaticity diagram.

[0039] Figure 4 To investigate the mechanism of action, A shows the fluorescence emission spectra of Eno, Tb-MOF, and Tb-MOF+Eno solutions at an excitation wavelength of 365 nm. The inset shows the corresponding fluorescence images of Tb-MOF and Tb-MOF+Eno solutions under a 365 nm UV lamp. B shows the UV-Vis absorption spectra of Eno, Tb-MOF, and Tb-MOF+Eno solutions. C shows the fluorescence lifetime decay curves of Tb-MOF and Tb-MOF+Eno. D is a schematic diagram of the antenna effect and energy transfer between levels.

[0040] Figure 5 A schematic diagram of a machine learning algorithm-assisted fluorescence visualization method for the quantitative detection of enoxacin.

[0041] Figure 6 The linear relationship between actual and predicted enoxacin concentrations in real samples based on an RF regression model is shown: A represents honey, B represents milk, and C represents beef.

[0042] Figure 7 Images showing different concentrations of enoxacin in honey samples. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0044] This invention discloses a dual-ligand lanthanide metal-organic framework (Tb-MOF) for rapid detection of enoxacin and a rapid detection method for enoxacin concentration. The core of this method is a Tb-MOF fluorescence colorimetric sensor constructed based on the antenna effect, which simultaneously induces fluorescence enhancement and color change in the mixed solution. During this process, the concentration of enoxacin is directly related to the degree of fluorescence enhancement and the change in fluorescence color. Specifically, as the concentration of enoxacin increases, the fluorescence intensity of the solution gradually increases, and the fluorescence color gradually changes from blue to green. Based on the above principle, this invention can achieve the following three main objectives: (1) By measuring the degree of fluorescence enhancement with a fluorescence spectrophotometer and establishing a linear equation, accurate prediction of enoxacin concentration in complex matrices can be achieved; (2) The RGB values ​​of the fluorescence color of the solution can be extracted and combined with a random forest (RF) model to complete a visual prediction of the concentration of enoxacin in complex matrices; (3) Fluorescence photos can be taken using a self-developed portable fluorescence sensing platform, and the RGB values ​​of the fluorescence color of the solution in the photos can be extracted by a mini-program and uploaded to the cloud. After calculation by the cloud algorithm, rapid prediction of enoxacin concentration can be achieved.

[0045] The following are specific examples and Figure 1-7 The method for ultrasensitive detection of enoxacin using a machine learning-assisted smartphone fluorescence visual sensing platform according to the present invention will be described in detail.

[0046] Step (1): Synthesis of Tb-MOF

[0047] Solution A was prepared by dissolving 0.48 mmol of 2,6-pyridinedicarboxylic acid (DPA), 0.02 mmol of 2-aminoterephthalic acid (ATA), and 140 μL of triethylamine in 15 mL of anhydrous ethanol. Solution B was prepared by dissolving 0.25 mmol of Tb(NO3)3·6H2O in 5 mL of deionized water. Solution A and solution B were then mixed and transferred to a round-bottom flask. The mixture was stirred in a water bath at 40 °C for 2 hours. The resulting white precipitate was collected and washed three times each with deionized water and anhydrous ethanol. Finally, the precipitate was dried overnight (12 h) in a vacuum oven at 40 °C to obtain the Tb-MOF material.

[0048] The prepared Tb-MOF was scanned using a transmission electron microscope (TEM), and the obtained TEM images are shown below. Figure 2 As shown in Figure A, Tb-MOFs exhibit small, densely distributed aggregated particles. Furthermore, elemental mapping results further confirm that Tb, C, O, and N are uniformly dispersed in Tb-MOFs. Figure 2 Powder X-ray diffraction (PXRD) analysis of Tb showed that the synthesized Tb-MOF was consistent with the simulated crystal form. Figure 2FT-IR analysis of the C-base revealed the formation of coordination bonds, with 2,6-pyridinedicarboxylic acid (DPA) and 2-aminoterephthalic acid (ATA) at 1200–1300 cm⁻¹. - The α-carboxylic acid exhibits a strong CO stretching vibration absorption peak in the 1-cm range, while the absorption peak of Tb-MOF in this region is relatively weak. Meanwhile, 2,6-pyridinedicarboxylic acid shows an absorption peak at 1695.98 cm⁻¹. - At position 1 and 2-aminoterephthalic acid at 1685.48 cm - The strong C=O tensile vibration peak at point 1 shifts to 1604.48 cm⁻¹ in Tb-MOF. - 1. These shifts and intensity changes in the characteristic peaks indicate that coordination bonds were successfully formed between the ligands and metal ions during the preparation of Tb-MOF. Figure 2 XPS analysis of the medium-D model further characterized the elemental composition of the Tb-MOF, and its full spectrum clearly showed the characteristic peaks of the four constituent elements: Tb, C, N, and O.

[0049] Step (2): Quantitative analysis of enoxacin using Tb-MOF fluorescence sensor fluorescence spectroscopy

[0050] (1) Quantitative detection by fluorescence spectroscopy

[0051] 300 μL of 0.4 mg / mL Tb-MOF and 700 μL of enoxacin solutions of different concentrations were added to EP tubes and mixed thoroughly to obtain mixed solutions that interacted with enoxacin solutions of different concentrations (enoxacin concentrations in the system were 0, 0.05, 0.08, 0.10, 0.30, 0.50, 0.70, 0.90, 1.00, 2.00, 3.00, 4.00, 5.00, 7.00, 10.00, 15.00, and 20.00 μmol / L). After reacting for 15 min, the fluorescence spectra of the mixed solutions were collected at room temperature and atmospheric pressure. A standard curve for detecting enoxacin was established based on the concentration of the enoxacin solution and the fluorescence enhancement of the corresponding mixed solutions.

[0052] like Figure 3 As shown in Figure A, with the increase of enoxacin concentration in the system (0-20.00 μmol / L), the fluorescence emission peak at 545 nm is enhanced, while the fluorescence emission peak at 417 nm changes slightly. Figure 3 The results from Figure B indicate that within the range of 0-20.00 μmol / L, there is a good linear relationship between the fluorescence intensity of the system and the concentration of enoxacin, with the standard curve being lgY = 0.229 + 0.598lgC. Eno (R 2 =0.992), where Y represents the ratio of fluorescence intensity at 545 nm to fluorescence intensity at 417 nm, and C EnoThis indicates the concentration of enoxacin. According to Y... L =Y B +ks, the calculated limit of detection (LOD) for enoxacin is 0.09 μmol / L, where Y L Y represents the smallest detectable analytical signal. B denoted as σ, where s is the average fluorescence intensity of the blank sample, s is the standard deviation of the blank signal, and k is the confidence level selection coefficient; in this study, k = 3 was used. As shown in Table 1, compared with other literature, the sensing system developed in this invention has a lower detection limit. More importantly, as... Figure 3 As shown in C (black arrows indicate concentration change trends) and D, the Tb-MOF-based sensor exhibits rich color shifts (blue and green) and significant linearity.

[0053] Table 1 Comparison of the present invention with existing enoxacin residue detection methods

[0054]

[0055] References

[0056] [1]Y.Li,

[0057] [2]A.Kundu,B.Maity,S.Basu,Rice Husk-Derived Carbon Quantum Dots-BasedDual-Mode Nanoprobe for Selective and Sensitive Detection of Fe 3+ andFluoroquinolones,ACS Biomater.Sci.Eng.8(2022)4764–4776.

[0058] [3] S.Yin,

[0059] [4]T.Garg,Renu,J.Kaur,P.Kaur,Nitansh,V.Kumar,K.Tikoo,A.Kaushik,S.Singhal,An innovative Z-scheme g-C3N4 / ZnO / NiFe2O4 heterostructure for theconcomitant photocatalytic removal and real-time monitoring of noxiousfluoroquinolones,Chem.Eng.J.443(2022)136441.

[0060] [5] X.Wang, Q.Li, B.Zong, X.Fang, M.Liu, Z.Li, S.Mao, K.(Ken)Ostrikov, Discriminative and quantitative color-coding analysis of fluoroquinolones with dual-emitting lanthanide metal-organic frameworks, Sens. Actuator. B-Chem. 373(2022)132701. (2) The mechanism of Tb-MOF fluorescence sensor was further investigated. The interaction mechanism between enoxacin and Tb-MOF was further explored. Figure 4 As shown in Figure A, at an excitation wavelength of 365 nm, the enoxacin solution exhibits a characteristic emission peak at 417 nm. The Tb-MOF solution also shows extremely weak characteristic emission peaks at 417 nm, 490 nm, and 545 nm. This may be due to the Tb... 3+ The unsaturated coordination of Tb-MOFs leads to the occupancy of coordination sites by water molecules. However, after the addition of enoxacin, Tb-MOFs exhibit improved performance at the same excitation wavelength of 490 nm. 5 D4→ 7 F6), 545nm 5 D4→7 F5), 585nm 5 D4→ 7 F4) and 623nm 5 D4→ 7 Four distinct emission peaks were observed at F3, accompanied by strong yellow-green fluorescence. Figure 4 (Illustration of A in the image). Among these peaks, the emission intensity is strongest at 545 nm. This is likely because the enoxacin molecule contains a β-diketone structure that can interact with Tb. 3+ Coordination, acting as an "antenna", enables Tb 3+ Sensitive. This triggered Tb 3+ from 5 D4 energy level to 7 Electron transitions in the Fj (j = 3–6) energy level transfer energy to Tb. 3+ This ultimately led to Tb 3+ The appearance of characteristic emission peaks.

[0061] We speculate that the enoxacin molecule may have replaced Tb in the Tb-MOF. 3+ Water molecules are coordinated around the ions, and thus further coordinate with the metal center. Figure 4 The UV-Vis absorption spectrum of enoxacin (Tb-MOF) showed absorption peaks at 267 nm and 340 nm. After mixing enoxacin with Tb-MOF solution, its absorption peak red-shifted from 267 nm to 273 nm, and the absorbance increased significantly. These changes in spectral characteristics provide further evidence for the coordination interaction between enoxacin and Tb-MOF. Furthermore, we investigated the changes in fluorescence lifetime before and after the addition of enoxacin. Figure 4 The C-ray discrepancy showed that the fluorescence lifetime of Tb-MOF at 545 nm significantly increased from 467.76 μs to 858.02 μs after the addition of enoxacin. This indicates that the enhanced fluorescence emission of Tb-MOF at 545 nm is attributed to the efficient energy transfer between Tb-MOF and enoxacin. Figure 4 The D diagram illustrates the sensing mechanism for enoxacin detection. The enoxacin molecule, through its β-diketone structure, can bind with Tb in the Tb-MOF. 3+ To achieve coordination binding and thus replace Tb 3+ The surrounding water molecules are already coordinated. Under photoexcitation, the enoxacin molecule is excited from the ground state (S0) to the first excited singlet state (S1), and then transforms into the first excited triplet state (T1) through an intersystem crossing (ISC) process. Energy is efficiently transferred from the T1 state of enoxacin to Tb. 3+ of 5 D4 excited state, ultimately through Tb 3+ of 5 D4→ 7The characteristic energy level transition of Fj (j=3-6) produces a characteristic yellow-green fluorescence emission. The antenna effect requires a specific energy difference between the donor molecule and the lanthanide ion. According to Latva's rule, when the T1 state of the ligand is different from the Tb state... 3+ of 5 The energy level difference (ΔE) between the D4 states is at 2000 cm⁻¹ -1 Up to 4000cm -1 Within this range, energy transfer efficiency is highest, and energy transfer is easy. Enoxacin's T1 state (22820 cm⁻¹) -1 ) and Tb 3+ excited state ( 5 D4, 20400cm -1 The band gap between them is 2420 cm. -1 The results conform to Latva's rule, indicating that the energy transfer process is feasible and efficient. Therefore, enoxacin effectively sensitizes the yellow-green fluorescence of Tb-MOF through the "antenna effect," providing a basis for the concentration fluorescence detection of enoxacin.

[0062] Step (3): Quantitative analysis of enoxacin based on machine learning

[0063] To expand the applicability of Tb-MOF fluorescence sensors in detecting enoxacin in complex samples, a fluorescence visual detection method combined with machine learning algorithms was developed. Figure 5Since complex matrices often introduce significant interference affecting detection accuracy, machine learning algorithms can be used to eliminate this interference, thereby enabling rapid and accurate quantification of enoxacin in complex samples. Real samples (honey, milk, beef) were pretreated, centrifuged, filtered, and the supernatant was mixed with enoxacin standard solution to prepare enoxacin solutions of different concentrations (0, 0.10, 0.30, 0.70, 1.00, 2.00, 3.00, 5.00, 10.00, 20.00 μmol / L). 700 μL of each concentration of enoxacin solution was mixed with 300 μL of Tb-MOF solution (0.4 mg / mL) in a 2 mL EP tube and reacted for 15 min. 400 μL of the mixture was then transferred to a 96-well plate. Fluorescence images of the mixture were collected using a mobile phone camera and imported into MATLAB. The images were converted to RGB values, and quantitative analysis was performed using an RF algorithm. The fluorescence image acquisition was repeated three times in parallel. The milk was processed as follows: First, 1 mL of milk was diluted to 10 mL with deionized water. Then, 100 μL of chloroform and 100 μL of 10% trichloroacetic acid (TCA) were added and mixed thoroughly. The mixture was sonicated for 10 minutes and then centrifuged at 13,000 rpm for 10 minutes. Next, the resulting mixture was filtered through a 0.22 μm microporous membrane to obtain the supernatant. Then, the supernatant was diluted to 50 mL with deionized water. The beef was processed as follows: 1 g of beef was placed in a centrifuge tube containing 30 mL of acetone solution. The mixture was sonicated for 20 minutes and then centrifuged for 10 minutes (10,000 rpm). The supernatant was collected. Finally, the obtained supernatant was filtered through a 0.22 μm microporous membrane and diluted 20 times with deionized water. The honey processing method is as follows: Dilute commercially available honey with deionized water at a ratio of 1:20 to reduce viscosity, then filter it through a 0.22-micron filter membrane to remove particulate impurities from the sample, and collect the filtrate to obtain the supernatant for later use.

[0064] Specifically, the acquired fluorescence images were imported into MATLAB, RGB values ​​were extracted, and 30 pixels were extracted for each concentration to construct a 300(RGB)×30 (sample) single-concentration data matrix. Finally, the images of 10 different concentrations were aggregated into a 300(RGB)×300 (sample) matrix. To establish the optimal quantitative model, the 300(RGB) data of the 300 samples at 10 different concentrations were randomly divided into 300(RGB)×210 (sample) training samples and 300(RGB)×90 (sample) prediction samples. Subsequently, based on the training set, the RF algorithm was used to construct an enoxacin concentration quantitative model. Figure 6 Data from the study showed that within the concentration range of 0-20 μmol / L, there was a good linear relationship between the predicted and actual concentrations of enoxacin.

[0065] Furthermore, Table 2 shows the performance metrics of the RF regression model: R2c (training set) and R2p (prediction set) are used to evaluate the quantitative correlation between predicted and actual concentrations, with values ​​closer to 1 indicating better model prediction performance; RMSEC (training set) and RMSEP (prediction set) are both metrics based on root mean square error (RMSE), used to measure the deviation between predicted and true values, with smaller values ​​indicating higher model accuracy. These results further confirm that the model combining RGB values ​​with the RF algorithm has excellent predictive performance for enoxacin concentrations in real samples.

[0066] Table 2. Predictive model parameters for enoxacin concentration in different real samples based on RGB values.

[0067]

[0068] Step (4): Visual inspection of enoxacin using a portable smartphone sensing platform

[0069] To achieve portable on-site detection, we developed a portable fluorescence detection platform by integrating the Tb-MOF probe with a smartphone application (building an RGB value conversion program and the trained RF model on the smart terminal). Figure 5 This system is capable of acquiring fluorescence images, extracting RGB values, and quantitatively analyzing enoxacin in complex samples. Specifically, after thoroughly mixing Tb-MOF solution and enoxacin solution in a 96-well plate, images are acquired using a smartphone under 365nm UV light irradiation. These images are then processed and RGB values ​​are extracted via a WeChat mini-program. After uploading these values ​​to the detection interface, an integrated cloud-based machine learning model automatically generates a quantitative prediction of enoxacin concentration. We selected honey as a representative composite sample to validate the accuracy of the portable sensing platform. A series of concentrations of enoxacin standards (0.30, 1.00, 10.00 μM) were added to the honey sample, with three parallel concentrations for each concentration, and then thoroughly mixed with Tb-MOF solution. As shown in Table 3, the recoveries ranged from 98.00% to 110.00%, and the RSDs ranged from 1.0% to 4.0%. These findings demonstrate that the developed mobile platform successfully converts chromofluorescence signals into rapid and accurate quantitative data, providing a simple and portable method for on-site detection.

[0070] Table 3 Comparison of predicted and actual values ​​of honey matrix from portable sensing platform

[0071]

[0072] In summary, this invention utilizes the antenna effect between a Tb-MOF modified with 2,6-pyridinedicarboxylic acid and 2-aminoterephthalic acid and enoxacin, resulting in characteristic changes in the fluorescence intensity and color of the Tb-MOF, thereby achieving ultrasensitive detection of enoxacin. This invention also designs a portable device integrated with a smartphone and incorporates a cloud server data analysis system with machine learning algorithms, enabling efficient acquisition, correction, analysis, and display of fluorescence color data. This innovative method fully leverages the fluorescence color signal capture and processing capabilities of modern smartphones, providing a user-friendly on-site detection solution for enoxacin. Compared to other detection methods, the Tb-MOF sensor can generate a richer color gamut, and when combined with a portable fluorescence visual sensing platform, it can further achieve sensitive on-site detection of enoxacin. The detection method of this invention features low cost, low detection limit, high portability, and high selectivity.

[0073] The embodiments described above merely illustrate precise implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. A dual-ligand lanthanide metal organic framework, Tb-MOF, for rapid detection of enoxacin, characterized in that, The ligand of the Tb-MOF is 2,6-pyridine dicarboxylic acid and 2-amino terephthalic acid.

2. A method for preparing the dual-ligand lanthanide metal-organic framework Tb-MOF for rapid detection of enoxacin as described in claim 1, characterized in that, The preparation method comprises the following steps: 2,6-pyridine dicarboxylic acid, 2-amino terephthalic acid and triethylamine are dissolved in anhydrous ethanol to obtain solution A; terbium nitrate hexahydrate is dissolved in water to obtain solution B; then, solution A and solution B are mixed, and then the reaction is stirred in a 30-50°C water bath, and finally, the double-ligand lanthanide metal organic framework Tb-MOF is obtained after washing and drying.

3. The production method according to claim 2, characterized by, The ratio of the sum of the amount of substance of 2,6-pyridine dicarboxylic acid and 2-amino terephthalic acid to the amount of substance of terbium nitrate hexahydrate is 2:1; The stirring time is 1-5h.

4. The production method according to claim 3, characterized by, The ratio of the amount of substance of 2,6-pyridine dicarboxylic acid to 2-amino terephthalic acid is (20-30):

1.

5. A method for rapid detection of enoxacin concentration, characterized by, The preparation method comprises the following steps: The double-ligand lanthanide metal organic framework Tb-MOF prepared by the preparation method in any one of claims 2-4 is formulated into a Tb-MOF solution, and the Tb-MOF solution is mixed with a to-be-tested enoxacin solution; under irradiation of a 365 nm ultraviolet lamp, the concentration of the to-be-tested enoxacin solution is determined according to the change in fluorescence intensity and / or the change in fluorescence color of the mixed system. The determination method of the fluorescence intensity comprises the following steps: a standard curve of enoxacin concentration and fluorescence enhancement degree is constructed; the determination method of the change in fluorescence color comprises the following steps: image information of the mixed system is collected, the change in fluorescence color is determined through a machine learning model, and then the concentration of enoxacin is predicted.

6. The method for rapid detection of the concentration of enoxacin according to claim 5, characterized in that, The method for establishing the standard curve comprises the following steps: the Tb-MOF solution is mixed with enoxacin standard solutions with different concentrations respectively; after the reaction is completed, the standard curve of enoxacin concentration and fluorescence enhancement degree is established according to the concentration of the enoxacin solution and the corresponding change in fluorescence intensity of the mixed solution under irradiation of a 365 nm ultraviolet lamp. The standard curve is as follows: lgY = 0.229+0.598lgC Eno , R 2 = 0.992, wherein Y represents the ratio of the fluorescence intensity at 545 nm to the fluorescence intensity at 417 nm, and C Eno represents the concentration of enoxacin.

7. The method for rapid detection of the concentration of enoxacin according to claim 6, characterized in that, The reaction time of the enoxacin standard solution and the Tb-MOF solution is 10-20 min, and the concentration of enoxacin in the reaction system ranges from 0 to 20.00 µmol / L; preferably, the concentration of enoxacin in the reaction system is 0, 0.05, 0.08, 0.10, 0.30, 0.50, 0.70, 0.90, 1.00, 2.00, 3.00, 4.00, 5.00, 7.00, 10.00, 15.00 and 20.00 µmol / L respectively.

8. The method for rapid detection of the concentration of enoxacin according to claim 5, characterized in that, The method for constructing the machine learning model comprises the following steps: The Tb-MOF solution is mixed with enoxacin standard solutions with different concentrations, and after the reaction is completed, the image information of each group of mixed solutions is collected under irradiation of a 365 nm ultraviolet lamp; the RGB value digital matrix of the mixed solution region in the corresponding photo is extracted by using MATLAB software; finally, the relationship between the RGB value of the pixel point and the concentration of enoxacin is established through a random forest model RF.

9. The method for rapid detection of the concentration of enoxacin according to claim 8, characterized in that, The construction method of the machine learning model is specifically: introducing the collected photos into MATLAB software, extracting the RGB values of each pixel point in the photos by using a pixel point extraction algorithm, generating a corresponding digital matrix; taking the digital matrix as a feature input, taking the corresponding enoxacin concentration as an output, combining the RF algorithm to perform model training and construction, and finally obtaining an RF model capable of realizing quantitative prediction of the enoxacin concentration; Preferably, the specific operation mode is: in the RF model construction process, first, the RGB values and the corresponding enoxacin concentrations are used to construct an original training set; then, the independent sampling method is used to extract samples from the original training set with replacement, to generate multiple independent training subsets, and each subset is used to train a decision tree model; after all the decision tree models are constructed, the RF algorithm averages the regression prediction results of each decision tree to obtain the final prediction value, thereby completing the construction of the enoxacin concentration prediction RF model.

10. The method for rapid detection of the concentration of enoxacin according to claim 9, characterized in that, It also includes completing the concentration prediction of enoxacin after collecting the fluorescence color information through the intelligent terminal, and specifically includes the following steps: Step 1: building an RGB value conversion program and a trained RF model on the intelligent terminal; Step 2: mixing the Tb-MOF solution with the enoxacin solution to be measured to obtain a mixed solution, after the reaction is complete, under the irradiation of a 365 nm ultraviolet lamp, collecting image information of the mixed solution through the intelligent terminal, and converting the image information into standardized RGB values, then the intelligent terminal predicts the concentration of the enoxacin solution to be measured based on the RF model according to the RGB values of the enoxacin solution to be measured; Preferably, the intelligent terminal is a smartphone or a tablet computer.

Images