Quinolone medicine SERS (Surface Enhanced Raman Scattering) detection method based on improved silver nano-substrate and spectral angle matching model

Abstract

The invention belongs to the field of antibiotic analysis and detection, and discloses a quinolone drug SERS (Surface Enhanced Raman Scattering) detection method based on an improved silver nano-substrate and a spectral angle matching model. An improved Lee-Meisel method is adopted to synthesize silver nano sol with precisely controlled particle size, molecular adsorption is realized by utilizing electrostatic interaction of a positively charged quinolone drug and negatively charged silver nano particles, and a high-density Raman enhanced hot spot is formed after an agglomeration agent is added; the SERS spectrum is collected at the excitation wavelength of 785 nm and the integral time of 10 s, qualitative distinguishing of drugs is achieved by combining a spectrum angle matching model, and quantitative detection is achieved by constructing a standard curve through the linear relation between the characteristic peak intensity and the drug concentration logarithm. The detection limit of the four drugs is as low as 1.03-1.16 ppb, the classification accuracy rate reaches 100%, the method is obviously superior to existing similar methods, the detection matrix range is wide, such as animal food and dairy products, and universality and high sensitivity are both considered.

Description

Technical Field

[0001] This invention belongs to the technical field and specifically discloses a method for detecting quinolone drugs based on an improved silver nanosubstrate and spectral angle matching model. Background Technology

[0002] Quinolone antibiotics, with 4-quinolone as their core, exert their antibacterial effects by specifically inhibiting bacterial DNA gyrase and topoisomerase IV, thereby blocking DNA replication and transcription. Due to their broad antibacterial spectrum, high activity, high oral bioavailability, and good tissue penetration, they are widely used in clinical treatment of respiratory, urinary tract, and gastrointestinal infections, as well as in agriculture and animal husbandry as feed additives or therapeutic drugs for livestock and aquaculture, playing a significant role in preventing animal diseases and improving farming efficiency. However, the extensive use and even abuse of these drugs have led to serious safety issues: they exhibit chondrotoxicity, affecting the skeletal development of minors and young animals; long-term exposure can easily cause gastrointestinal reactions, central nervous system abnormalities, and skin allergies; some drugs also pose risks of genotoxicity, cardiotoxicity, and liver and kidney damage; furthermore, residual quinolone antibiotics in the environment can accumulate through the food chain, continuously threatening the ecological environment and human health.

[0003] Currently, there are many detection methods available both domestically and internationally, mainly including high-performance liquid chromatography (HPLC), liquid chromatography-mass spectrometry (LC-MS), gas chromatography (GC), microbial detection methods, and immunoassay. Among these, HPLC offers high separation efficiency and good accuracy, making it the mainstream method for routine detection; however, sample pretreatment is complex and the detection time is long. LC-MS offers high sensitivity and specificity and can detect multiple components simultaneously, but the instrument cost is high and the operation requirements are strict, making it difficult to popularize in grassroots laboratories. GC requires derivatization of quinolone drugs with poor thermal stability, and the steps are cumbersome and prone to errors. Microbial detection methods are simple to operate and low in cost, but have poor specificity and long detection cycles. Immunoassay is rapid and sensitive, but antibody preparation is difficult and cross-reactivity is prone to occur, limiting its applicability.

[0004] Therefore, while these methods can meet some detection needs, they have significant limitations in rapid on-site detection, trace residue analysis, and low-cost applications. Summary of the Invention

[0005] To address the problems of poor substrate particle size control, low matrix versatility, and complex algorithms in existing quinolone SERS detection methods, this paper proposes a quinolone drug SERS detection method based on an improved silver nanosubstrate and spectral angle matching model. This method achieves accurate qualitative identification (100% accuracy) and highly sensitive quantitative identification (detection limit ≤1.16 ppb) of four specific quinolones and is applicable to a variety of matrices.

[0006] To achieve the above technical objectives, the technical solution adopted by this invention is as follows:

[0007] A method for detecting SERS of quinolone drugs based on an improved silver nanosubstrate and spectral angle matching model includes the following steps:

[0008] (1) Preparation of reinforced substrate: Sodium citrate was added to boiling silver nitrate solution, the heating temperature was controlled at 120~150℃ and the stirring speed was 600~900 rpm. The solution was heated and stirred until it turned gray-green. After natural cooling, it was stored in the dark to obtain silver nanosol with a particle size of 50±10 nm.

[0009] The treatment in step (1) ensures uniform particle size of silver nanoparticles and an absolute value of ≥30 mV for the Zeta potential, thus ensuring solution stability. The preparation of this substrate only takes 1-2 hours, reducing costs by 60% and enhancing performance to be more stable.

[0010] (2) Centrifuge the sample solution containing quinolone drugs to remove impurities, then mix it with the silver nanosol synthesized in step (1), add a conventional agglomerating agent to aggregate the silver nanoparticles, and form Raman-enhanced hot spots in the interparticle gaps;

[0011] (3) Signal acquisition and analysis: The mixed system was irradiated with a laser beam with an excitation wavelength of 785 nm, an integration time of 10 s, and 3 integration times to obtain SERS spectral data; the quinolone drugs were qualitatively distinguished by combining the spectral angle matching model, and quantitative detection was achieved by using the standard curve of characteristic peak intensity versus the logarithm of drug concentration.

[0012] Preferably, the spectral angle matching model calculates the spectral angle after normalizing the unknown spectrum and the standard spectrum. A spectral similarity of ≥98% indicates that the drug is the same. Qualitatively, the spectral angle matching model eliminates the need for PCA dimensionality reduction, reducing computation time to within 10 seconds. By comparing the normalized matching degree of the unknown spectrum with the four standard quinolone spectra, 100% classification accuracy is achieved.

[0013] Preferably, the modeling process of the spectral angle matching model is as follows:

[0014] 1) Import reference spectral database: Import standard spectral data of quinolone drugs into the model;

[0015] 2) Preprocessing the reference spectrum: The reference spectral data is preprocessed for noise reduction using conventional methods;

[0016] 3) Importing unknown spectral data and preprocessing unknown spectra: Same as importing and preprocessing the reference spectra in steps 1) and 2) above;

[0017] 4) Execute the spectral angle matching model algorithm: First, set the matching threshold to 0.05~0.15, then initialize the result matrix, perform spectral angle matching for each unknown spectrum, and calculate the spectral angle between the unknown spectrum and each corresponding reference spectrum. Calculate the dot product, magnitude, and cosine value of the two spectral vectors, set the cosine value range to -1 to 1, and calculate the angle to obtain the matching result.

[0018] More preferably, the formula for the spectral angle matching model is as follows:

[0019] Spectral similarity = cos(θ) = Σ(xi×yi) / (√Σ(xi) 2 )×√Σ(yi 2 ));

[0020] Where xi is the i-th band value of the reference spectrum, and yi is the i-th band value of the spectrum to be matched.

[0021] θ is the angle between the two spectral vectors.

[0022] Preferably, the preprocessing method for the reference spectrum is at least one or more of smoothing (SG), multivariate scattering correction (MSC), baseline correction (airPLS), and normalization.

[0023] Preferably, a linear fit is performed with the logarithm of drug concentration as the abscissa and the corresponding characteristic peak intensity as the ordinate to obtain a standard curve equation, which is used to calculate the concentration of unknown samples. Quantitative analysis using the standard curve of characteristic peak intensity versus logarithm of concentration can achieve a detection limit as low as 1.03~1.16 ppb.

[0024] The standard curve equation and correlation coefficient of the drug are as follows:

[0025] Enrofloxacin: y = 5935.3x - 921.35, R 2 =0.995;

[0026] Saladin: y = 3446.16x - 1975.39, R 2 =0.991;

[0027] Lomefloxacin: y = -1276.47x + 2753.66, R 2 =0.997;

[0028] Norfloxacin: y = 3692.19x + 45.26, R 2 =0.988.

[0029] Preferably, in step (1), the concentration of silver nitrate solution is 0.001~0.005 mol / L, the volume of sodium citrate solution added is 1~4 mL, and the mass fraction of sodium citrate solution is 1%~3%.

[0030] Preferably, the quinolone drug is one or more of enrofloxacin, sarafloxacin, lomefloxacin, or norfloxacin.

[0031] Preferably, the sample solution is one or more of an aqueous solution and a dairy product.

[0032] Preferably, the agglomerating agent is one or a mixture of two of potassium chloride and sodium chloride, and the volume ratio of silver nanosol to agglomerating agent is 1:1.

[0033] Compared with the prior art, the beneficial effects of the present invention are:

[0034] 1. Superior substrate performance: The improved Lee-Meisel method precisely controls the size of silver nanoparticles, resulting in high stability and no precipitation after 30 days of continuous storage. The Raman enhancement efficiency is 3 to 5 times higher than that of existing silver sols with undefined particle sizes.

[0035] 2. High qualitative accuracy: The spectral angle matching model does not require dimensionality reduction, has a fast calculation speed, and can classify unknown samples within 10 seconds with a classification accuracy of 100%, which is better than existing algorithms;

[0036] 3. Higher sensitivity: The detection limit is as low as 1.03~1.16 ppb, which meets the needs of trace residue detection. Attached Figure Description

[0037] Figure 1 Transmission electron microscopy image of silver nanoparticles;

[0038] Figure 2 Zeta potential diagrams for quinolone drugs, silver nanosols, and mixtures of the two;

[0039] Figure 3 A represents the SERS spectra of 4-mercaptobenzoic acid standard solution with the addition of silver nanoparticles at different concentration ratios; Figure 3 B shows the SERS spectra of silver nanosol substrates and agglomerating agents with different amounts of added silver nanosol in 4-mercaptobenzoic acid standard solution;

[0040] Figure 4 Characteristic SERS spectra of four quinolone drugs: enrofloxacin, sarafloxacin, lomefloxacin, and norfloxacin;

[0041] Figure 5 Enrofloxacin, sarafloxacin, lomefloxacin, and norfloxacin, four quinolone drugs, were administered at 737 cm⁻¹. -1 531 cm-1 809 cm -1 735 cm -1 A standard curve is plotted with the characteristic peak intensity as the ordinate and the logarithm of the concentration as the abscissa.

[0042] Figure 6 A shows the SERS spectra of enrofloxacin collected 20 times consecutively; Figure 6 B is a bar chart with the intensity of the 20 characteristic peaks of enrofloxacin SERS spectrum as the vertical axis;

[0043] Figure 7 A is the original quinolone drug spectrum; Figure 7 B is the spectrum after normalization; Figure 7 C is a comparison diagram of the unknown spectrum and the reference spectrum in the spectral angle matching model; Figure 7 D is a confusion matrix diagram showing the accuracy of the spectral angle matching model in identifying unknown spectra. Detailed Implementation

[0044] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0045] Unless otherwise specified, the experimental methods used in the embodiments of this invention are conventional methods; unless otherwise specified, the materials and reagents used are commercially available.

[0046] This invention first prepares a silver nanosol-reinforced substrate, then optimizes the concentration of the reinforcement substrate and the amount of reinforcement substrate and agglomerating agent added. By establishing a standard curve of the characteristic peak intensity of each quinolone drug versus the logarithm of the drug concentration, the purpose of quantitative detection is achieved. A spectral angle matching model is constructed to normalize the spectral data and qualitatively identify the types of drugs in unknown samples.

[0047] Example 1 This example includes the following steps:

[0048] 1) Synthesis of silver nanosols

[0049] Add 100 mL of silver nitrate solution to a round-bottom flask and shake well. Place the round-bottom flask in an oil bath and stir magnetically until the solution boils. Slowly and continuously add freshly prepared trisodium citrate solution while continuing to heat and stir until the solution turns grayish-green. Stop heating, allow to cool naturally, and then store away from light.

[0050] The concentration of the silver nitrate solution is 0.001~0.005 mol / L, the volume of trisodium citrate is 1~4 mL, the mass fraction is 1%~3%, the heating temperature is 120~150°C, and the magnetic stirring speed is 600~900 rpm.

[0051] 2) Quantitative analysis of the sample on the silver nanosol substrate

[0052] The synthesized silver nanosol was centrifuged and concentrated, then added to the sample solution to be tested. An agglomerating agent was added, and the mixture was thoroughly mixed. The SERS spectrum of the sample was then acquired under optimal testing conditions. The intensity of the characteristic peaks in the SERS spectrum changes depending on the drug concentration in the sample. A standard curve was established using the logarithmic correlation between the intensity of the sample's SERS characteristic peaks and the drug concentration for quantitative analysis of unknown samples. Specifically, the optimal testing conditions were: excitation wavelength of 785 nm, integration time of 10 s, and integration times of 3.

[0053] Figure 1 The image in the middle is a transmission electron microscope image of silver nanoparticles. The particle size range of the silver nanoparticles under the conditions of Example 1 is 50 ± 10 nm. Figure 2 The Zeta potential analysis results of silver nanosol, quinolone drug, and their mixture are presented. The figures show that the quinolone drug exhibits a positive potential, while the synthesized silver nanosol shows a large negative potential. This indicates that the silver nanoparticles synthesized using sodium citrate as a reducing agent are surrounded by a negative charge, resulting in better stability. After adding the silver nanosol to the drug, the negatively charged silver nanoparticles attract the positively charged quinolone drug to adhere to their surface, neutralizing the negative charge of the silver nanoparticles. Due to the change in surface charge, the silver nanoparticles aggregate, creating SERS hotspot regions in the intermaterial gaps. Drug molecules then generate SERS signals under Raman laser excitation.

[0054] Example 2

[0055] Optimization of silver nanosol concentration and reagent dosage: SERS spectra of 4-mercaptobenzoic acid in the presence of silver nanosol substrates at different concentration factors were collected to determine the optimal concentration of silver nanosol. SERS spectra of 4-mercaptobenzoic acid with different volume ratios of reinforcing substrate and agglomerant were collected to determine the optimal ratio of reinforcing agent to agglomerant.

[0056] Specifically, a fixed volume of 200 μL was used to prepare silver nanoparticles with concentration factors of 0, 1, 4, 8, 16, and 32 times. These were then added to the 4-mercaptobenzoic acid solution, and their SERS spectra were collected. Figure 3A. Silver nanosol, concentrated 8 times, exhibits the best signal enhancement effect. With fixed volumes of 4-mercaptobenzoic acid and silver nanosol concentrations, and with added volumes of the enhancer and agglomerant set at 20 ± 50 μL, 50 ± 50 μL, and 50 ± 20 μL, SERS spectra were collected, as shown below. Figure 3 As shown in Figure B, the sample exhibits the best SERS signal when the amount of reinforcing agent and agglomerator added is 50 ± 50 μL.

[0057] Characteristic SERS spectra of four quinolone drugs: Concentrated silver nanoparticles were added to 200 μL of a quinolone standard solution at the optimal excitation wavelength of 785 nm, followed by 50 μL of agglomerating agent. The SERS spectra of the solution were then collected. Figure 4 As shown, the characteristic peaks of enrofloxacin, sarafloxacin, lomefloxacin, and norfloxacin are located at 737 cm⁻¹. -1 531 cm -1 809cm -1 735 cm -1 Place.

[0058] Test Results

[0059] Quantitative SERS detection of four quinolone drugs: Under optimal testing conditions, SERS spectra of enrofloxacin, sarafloxacin, lomefloxacin, and norfloxacin at gradient concentrations were acquired. A standard curve was established by comparing the signal intensity at the characteristic peaks with the drug concentration. Figure 5 As shown, the characteristic peak intensity exhibits a good linear relationship with the logarithm of the drug concentration. Based on the standard curve, the detection limits for the four drugs were calculated to be 1.03 ppb, 1.16 ppb, 1.06 ppb, and 1.05 ppb, respectively, with linear ranges of 10-1000 ppb, 10-200 ppb, 10-500 ppb, and 5-200 ppb, respectively.

[0060] Based on Example 1, different treatments of silver nanosol were compared. The following comparative examples provide different treatments and results for silver nanosol:

[0061] Comparative Example 1 (refer to CN116678864A): 25 mg of Fe3O4@COF and 20 mg of silver nitrate were added to 200 mL of ultrapure water, stirred for 10 minutes, and then heated with sodium citrate (1%, 4 mL) as a reducing agent. After adding the reducing agent, the mixture was boiled for 30 minutes to obtain the Fe3O4@COF@Ag substrate. The detection limit of the SERS substrate synthesized in this comparative example for norfloxacin was 6.2 ppb, while that of this invention was 1.05 ppb.

[0062] Comparative Example 2: The stability of the silver nanoparticles of this invention was compared with the SERS detection methods for quinolone drugs in other literature. The SERS intensity of the silver nanoparticles using the method in the references decreased to below 85% after 30 days of storage. Reference: Yuan Y, Peng X, Wang C, Bai Z: Surface-enhanced Raman scattering sensor based on MWCNTs@ZnO / Ag to detect the enrofloxacin in pork. SpectrochimicaActa Part A: Molecular and Biomolecular Spectroscopy, 2025, 332:125818.

[0063] Example 3: Stability of quinolone drugs in SERS detection:

[0064] Taking enrofloxacin as an example, 20 SERS spectra were continuously collected to examine the reproducibility and stability of the signal during detection. Figure 6 As shown in Figure A, the SERS spectra acquired in 20 sampling sessions exhibit good homogeneity. Figure 6 As shown in B, enrofloxacin in 20 tests was at 737 cm⁻¹. -1 The standard deviation of the characteristic peak intensity at the specified location is RSD=4.84%, indicating that the method has good signal stability.

[0065] Example 4 Qualitative analysis of quinolone drugs:

[0066] A spectral angle matching model was constructed. Characteristic spectra of each drug were collected as a reference database and normalized to eliminate interference from signal intensity. Spectral data of unknown samples were collected, normalized, and imported into the model. The model automatically matched the database and classified the unknown samples. The modeling process of the spectral angle matching model is as follows:

[0067] 1) Import the reference spectral database: Create a database of standard spectral data files for each quinolone drug and import it into the model. Each reference spectrum should include the drug category and the signal intensity at the corresponding spectral position.

[0068] 2) Preprocessing of the reference spectrum: Maximum value normalization and SG smoothing are used to perform conventional noise reduction preprocessing on all Raman peaks in the reference spectrum data to eliminate interference caused by excessive differences in Raman peak intensity and instrument noise. The above normalization or smoothing processes are directly referenced from existing technologies and will not be elaborated upon here.

[0069] 3) Importing unknown spectral data and preprocessing unknown spectra: Same as importing and preprocessing the reference spectra above.

[0070] 4) Execute the spectral angle matching model algorithm: First, set the matching threshold to 0.1. Then, initialize the result matrix. Perform spectral angle matching for each unknown spectrum and calculate the spectral angle between it and each reference spectrum. Calculate the dot product, magnitude, and cosine value of the two spectral vectors, setting the cosine value to between -1 and 1. Calculate the angle, then save the matching degree with the reference spectrum. Find the best matching result and generate a visualization image. The formula for the spectral angle matching model is as follows:

[0071] Formula 1: Spectral angle θ = arccos(Σ(xi×yi) / (√Σ(xi)) 2 )×√Σ(yi 2 ))),

[0072] Formula 2: Spectral similarity = cos(θ) = Σ(xi×yi) / (√Σ(xi) 2 )×√Σ(yi 2 )),

[0073] Where xi is the i-th band value of the reference spectrum, and yi is the i-th band value of the spectrum to be matched.

[0074] θ is the angle (in radians) between two spectral vectors.

[0075] like Figure 7 A and Figure 7 As shown in B, the spectrum was normalized, and then spectral matching was performed. The matching result is as follows. Figure 7 As shown in the confusion matrix in D, the accuracy rate for identifying 60 unknown samples is 100%.

[0076] The specific implementation schemes described above are merely a further detailed explanation of the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are only specific examples of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting SERS of quinolone drugs based on an improved silver nanosubstrate and spectral angle matching model, characterized in that, Includes the following steps: (1) Preparation of reinforced substrate: Sodium citrate was added to boiling silver nitrate solution, the heating temperature was controlled at 120~150℃ and the stirring speed was 600~900 rpm, and the solution was continuously heated and stirred until it turned gray-green. After cooling, it was stored to obtain silver nanosol with a particle size of 50±10 nm. (2) Centrifuge the sample solution containing quinolone drugs to remove impurities, then mix it with the silver nanosol synthesized in step (1), add a conventional agglomerating agent to aggregate the silver nanoparticles, and form Raman-enhanced hot spots in the interparticle gaps; (3) Signal acquisition and analysis: The mixed system in step (2) was irradiated with a laser beam with an excitation wavelength of 785 nm, an integration time of 10 s, and 2 to 3 integration times to obtain SERS spectral data; the quinolone drugs were qualitatively distinguished by combining the spectral angle matching model, and quantitative detection was achieved by using the standard curve of characteristic peak intensity versus the logarithm of drug concentration.

2. The quinolone drug SERS detection method based on an improved silver nanosubstrate and spectral angle matching model according to claim 1, characterized in that, The spectral angle matching model calculates the spectral angle after normalizing the unknown spectrum with the standard spectrum. If the spectral similarity is ≥98%, it is determined to be the same drug.

3. The quinolone drug SERS detection method based on the improved silver nanosubstrate and spectral angle matching model according to claim 2, characterized in that, The modeling process of the spectral angle matching model is as follows: 1) Import reference spectral database: Import the standard spectral data of quinolone drugs into the model; 2) Preprocessing the reference spectrum: The reference spectral data is preprocessed for noise reduction using conventional methods; 3) Importing unknown spectral data and preprocessing unknown spectra: Same as importing and preprocessing the reference spectra in steps 1) and 2) above; 4) Execute the spectral angle matching model algorithm: First, set the matching threshold to 0.05~0.15, then initialize the result matrix, perform spectral angle matching for each unknown spectrum, and calculate the spectral angle between the unknown spectrum and each corresponding reference spectrum. Calculate the dot product, magnitude, and cosine value of the two spectral vectors, set the cosine value range to -1 to 1, and calculate the angle to obtain the matching result.

4. The quinolone drug SERS detection method based on the improved silver nanosubstrate and spectral angle matching model according to claim 3, characterized in that, The formula for the spectral angle matching model is as follows: Spectral similarity = cos(θ) = Σ(xi×yi) / (√Σ(xi) 2 )×√Σ(yi 2 )); Where xi is the i-th band value of the reference spectrum, and yi is the i-th band value of the spectrum to be matched. θ is the angle between the two spectral vectors.

5. The quinolone drug SERS detection method based on the improved silver nanosubstrate and spectral angle matching model according to claim 3, characterized in that, The preprocessing method for the reference spectrum is at least one or more of the following: smoothing, multivariate scattering correction, baseline correction, and normalization.

6. The quinolone drug SERS detection method based on the improved silver nanosubstrate and spectral angle matching model according to claim 1, characterized in that, The quinolone drug is one or more of enrofloxacin, sarafloxacin, lomefloxacin, or norfloxacin.

7. The quinolone drug SERS detection method based on an improved silver nanosubstrate and spectral angle matching model according to claim 1, characterized in that, A linear fit was performed with the logarithm of drug concentration on the x-axis and the corresponding characteristic peak intensity on the y-axis. The standard curve equation and correlation coefficient for the drug are as follows: Enrofloxacin: y = 5935.3x - 921.35, R 2 =0.995; Saladin: y = 3446.16x - 1975.39, R 2 =0.991; Lomefloxacin: y = -1276.47x + 2753.66, R 2 =0.997; Norfloxacin: y = 3692.19x + 45.26, R 2 =0.

988.

8. The quinolone drug SERS detection method based on the improved silver nanosubstrate and spectral angle matching model according to claim 1, characterized in that, In step (1), the concentration of silver nitrate solution is 0.001~0.005 mol / L, the volume of silver nitrate solution added is 100mL, the volume of sodium citrate solution added is 1~4 mL, and the mass fraction of sodium citrate solution is 1%~3%.

9. The quinolone drug SERS detection method based on an improved silver nanosubstrate and spectral angle matching model according to claim 1, characterized in that, The sample solution is one or more of the following: aqueous solution and dairy product.

10. The quinolone drug SERS detection method based on an improved silver nanosubstrate and spectral angle matching model according to claim 1, characterized in that, The agglomerating agent is one or a mixture of two of potassium chloride and sodium chloride, and the volume ratio of silver nanosol to agglomerating agent is 1~3:1.

Images