Automated identification of pesticide mixtures via machine learning analysis of TLC-SERS spectra.

Identification of components in pesticide mixtures has been a major challenge in spectral analysis. In this paper, we assembled monolayer Ag nanoparticles on Thin-layer chromatography (TLC) plates to prepare TLC-Ag substrates with mixture separation and surface-enhanced Raman scattering (SERS) detection. Spectral scans were performed along the longitudinal direction of the TLC-Ag substrate to generate SERS spectra of all target analytes on the TLC plate. Convolutional neural network classification and spectral angle similarity machine learning algorithms were used to identify pesticide information from the TLC-SERS spectra. It was shown that the proposed automated spectral analysis method successfully classified five categories, including four pesticides (thiram, triadimefon, benzimidazole, thiamethoxam) as well as a blank TLC-Ag data control. The location of each pesticide on the TLC plate was determined by the intersection of the information curves of the two algorithms with 100 % accuracy. Therefore, this method is expected to help regulators understand the residues of mixed pesticides in agricultural products and reduce the potential risk of agricultural products to human health and the environment.

Interfacial Self-Assembly of Surfactant-Free Au Nanoparticles as a Clean Surface-Enhanced Raman Scattering Substrate for Quantitative Detection of As5+ in Combination with Convolutional Neural Networks.

Surface-enhanced Raman scattering (SERS) is a highly sensitive tool in the field of environmental testing. However, the detection and accurate quantification of weakly adsorbed molecules (such as heavy metal ions) remain a challenge. Herein, we combine clean SERS substrates capable of capturing heavy metal ions with convolutional neural network (CNN) algorithm models for quantitative detection of heavy metal ions in solution. The SERS substrate consists of surfactant-free Au nanoparticles (NPs) and l-cysteine molecules. As plasmonic nanobuilt blocks, surfactant-free Au NPs without physical or chemical barriers are more accessible to target molecules. The amino and carboxyl groups in the l-cysteine molecule can chelate As5+ ions. The CNN algorithm model is applied to quantify and predict the concentration of As5+ ions in samples. The results demonstrated that this strategy allows for fast and accurate prediction of As5+ ion concentrations, and the determination coefficient between the predicted and actual values is as high as 0.991.

Study on the classification and identification of various carbonate and sulfate mineral medicines based on Raman spectroscopy combined with PCA-SVM algorithm.

The efficacy of mineral medicines varies greatly between different origins. Therefore, investigating a method to quickly identify similar mineral medicines is meaningful. In this paper, a visual classification and identification model of Raman spectroscopy combined with principal component analysis (PCA) and support vector machine (SVM) algorithms was developed to rapidly classify and identify carbonate and sulfate mineral medicines. The results reveal that although the Raman spectra are too similar to distinguish by naked eye, the PCA-SVM algorithm can perform accurate classification and identification, and its accuracy, precision, recall and F1-score parameters all reach 100%. The proposed method is rapid, accurate, nondestructive, convenient, portable, and low cost, and has important application value for the classification, identification and quality supervision of various carbonate and sulfate mineral medicines.

Qualitative and quantitative detection and identification of two benzodiazepines based on SERS and convolutional neural network technology.

Drug abuse is a global social issue of concern. As the drug market expands, there is an urgent need for technological methods to rapidly detect drug abuse to meet the needs of different situations. Here, we present a strategy for the rapid identification of benzodiazepines (midazolam and diazepam) using surface-enhanced Raman scattering (SERS) combined with neural networks (CNN). The method uses a self-assembled silver nanoparticle paper-based SERS substrate for detection. Then, a SERS spectrum intelligent recognition model based on deep learning technology was constructed to realize the rapid and sensitive distinction between the two drugs. In this work, a total of 560 SERS spectra were collected, and the qualitative and quantitative identification of the two drugs in water and a beverage (Sprite) was realized by a trained convolutional neural network (CNN). The predicted concentrations for each scenario could reach 0.1-50 ppm (midazolam in water), 0.5-50 ppm (midazolam in water and diazepam in Sprite), and 5-150 ppm (diazepam in Sprite), with a strong coefficient of determination (R2) larger than 0.9662. The advantage of this method is that the neural network can extract data features from the entire SERS spectrum, which makes up for information loss when manually identifying the spectrum and selecting a limited number of characteristic peaks. This work clearly clarifies that the combination of SERS and deep learning technology has become an inevitable development trend, and also demonstrates the great potential of this strategy in the practical application of SERS.

Detection of Soluble Mercury in Cinnabar Using a CV-Ag NPs SERS Probe.

In the current work a uniform morphological Ag nanoparticles (Ag NPs) were prepared with ascorbic acid as a reducing agent and citrate as a stabilizer. The surface of Ag NPs modified by crystal violet (CV) and potassium iodide (KI) was used as an aggregation agent to obtain CV modified Ag NPs (CV-Ag NPs) probes for detecting mercury ions. The mercury ions could be reduced to mercury molecules by citrate, and then deposited on the surface of Ag NPs, leading to the separation of CV molecules from the surface of Ag NPs. Therefore, the SERS signal intensity of CV decreased with the increase of the Hg2+ concentration and the concentration of Hg2+ was in the range of 1 × 10-11 to 1 × 10-5 M. Taking the change of the characteristic peak intensity of CV at 913 cm-1 as a reference, the SERS spectrum intensity of CV has a linear relationship with the Hg2+ concentration. The equation is y = -333.55x + 1343.05, where the linear correlation coefficient is R2 = 0.980, and the recovery rate is between 84.20 to 105.60%. Finally, the CV-Ag NPs probe was used to quickly detect soluble mercury in cinnabar. Compared with the conventional large-scale instrument detection method, this simple and fast method, can be applied for rapid detection of soluble mercury, and has a certain significance for concerning the research of mineral medicine processing mechanism.

Sensitive and reliable identification of fentanyl citrate in urine and serum using chloride ion-treated paper-based SERS substrate.

Recently, the phenomenon of fentanyls overdose leading to death is emerging in an endless stream. There is an urgent requirement to quickly identify fentanyl content in body fluids for medical and judicial purposes. With this in mind, we present a paper-based SERS substrate decorated with uniform gold nanospheres treated by chloride ion for the detection of fentanyl citrate in urine and serum. In particular, the paper-based SERS sensor was prepared by liquid/liquid self-assembly technique and chloride ion was introduced to clean and modify the substrate surface, which improved the sensitivity of the solid substrate with an enhancement factor (EF) as high as 1.64 × 105. Moreover, the uniformity of each paper-based substrate and the repeatability on different batches of substrate were excellent, and there was no obvious change in the intensity response of Raman spectra within a month. As a result, the quantitative analysis of fentanyl citrate in artificial urine and rat serum were performed based on the modified paper-based substrate with the limit of detection as low as 0.59 µg/mL and 2.78 µg/mL, respectively. Both the concentrations of the two biological samples with the Raman signal intensity were linearly plotted and the recovery of the spiked samples with different concentrations was collected to verify the accuracy of the quantitative curves. All the results suggest that this work makes SERS method available for the rapid identification and quantitative analysis of illicit drug in the real biological samples.

Rapid Detection of Dezocine in Biological Fluids Based on SERS Technology.

This paper describes a method based on surface enhanced Raman spectroscopy (SERS) technology for rapid detection of dezocine in urine and serum. Firstly, an Ag colloid substrate was prepared and characterized. Then the Raman characteristic peaks of dezocine were assigned from both theoretical and experimental aspects. Finally, the Raman peak at 661 cm-1 was selected as its characteristic peak to perform SERS detection on dezocine in urine and serum, and the detection limits of dezocine in urine and serum were determined. The relationships between the characteristic peak intensity and the concentration of dezocine in urine and serum were fitted and the recovery rates were calculated. This rapid, accurate and non-destructive method establishes a good foundation for rapid on-site detection of dezocine in biological samples.

Surface-enhanced Raman spectroscopy for rapid identification and quantification of Flibanserin in different kinds of wine.

Wine has always been a popular carrier for psychedelic drugs, with the rapid identification and quantification of psychedelic drugs in wine being the focus of regulating illegal behavior. In this study, surface-enhanced Raman spectroscopy (SERS) is used for the rapid detection of Flibanserin in liquor, beer and grape wine. First, the theoretical Raman spectrum with characteristic Flibanserin peaks was calculated and identified, and the limit of detection of 1 µg mL-1 for Flibanserin in liquor was determined. The curve equation was obtained by fitting using the least squares method, and the correlation coefficient was 0.995. The recovery range of the Flibanserin liquor solution ranged from 93.70% to 108.32%, and the relative standard deviation (RSD) range was 2.77% to 7.81%. Identification and quantification of Flibanserin in liquor, beer and grape wine were done by principal component analysis (PCA) and support vector machine (SVM). Machine learning algorithms were used to reduce the workload and the possibility of manual misjudgements. The classification accuracies of the Flibanserin liquor, beer and grape wine spectra were 100.00%, 95.80% and 92.00%, respectively. The quantitative classification accuracies of the Flibanserin liquor, beer and grape wine spectra were 92.30%, 91.70% and 92.00%, respectively. The machine learning algorithms were used to verify the advantages and feasibility of this method. This study fully demonstrates the huge application potential of combining SERS technology and machine learning in the rapid on-site detection of psychedelic drugs.

A dual-functional PDMS-assisted paper-based SERS platform for the reliable detection of thiram residue both on fruit surfaces and in juice.

In this work, a dual-functional SERS platform was developed via a paper-based SERS substrate with the aid of hydrophobic polydimethylsiloxane (PDMS) for effective and reliable measurements of thiram on fruit surfaces and in juice. Specifically, the paper-based SERS substrate was fabricated by coating with a core-shell Au@Ag nanorod monolayer, and the SERS performance was optimized compared with multilayer adsorption. Moreover, a versatile SERS platform was constructed by simply pasting the paper-based substrate in reverse onto PDMS using polymethyl methacrylate (PMMA) tape. On the one hand, this detection platform was able to realize sample enrichment due to the excellent hydrophobicity of PDMS, thereby increasing the sensitivity of measurements, and its function was successfully displayed through the identification of thiram in orange juice. On the other hand, PDMS could also play a supporting role and the paper-based substrate reversely stuck on PDMS was able to extract samples on the side without nanoparticles, which greatly avoids damage to nanoparticles on the substrate. Therefore, the accuracy of analysis was significantly improved and the inspection of thiram on the surface of an orange was demonstrated based on this function. As a result, this proposed SERS platform provides a new strategy for preparing multifunctional SERS sensors for the on-site monitoring of chemical contaminants in the food-safety field.

Lab-On-Capillary Platform for On-Site Quantitative SERS Analysis of Surface Contaminants Based on Au@4-MBA@Ag Core-Shell Nanorods.

A portable and highly reproducible lab-on-capillary surface-enhanced Raman scattering (SERS) platform was developed using a specially designed homemade device for rapid on-site SERS measurement. In particular, this platform was composed of a capillary with a tiny orifice, which allows an effective and lossless sample extraction, resulting in high SERS performance. The capillary-based plasmonic substrate was prepared by compactly assembling Au@Ag core-shell nanorods (NRs) embedded with the 4-mercaptobenzoic acid (4-MBA) molecule as an internal standard onto the inner wall of a capillary tube. The fabrication process is facile and convenient with no requirement for complicated procedures. The exclusively prepared nanoparticles were able to significantly improve the signal consistency and overcome the limitations of reliable quantitative SERS analysis compared with conventional methods. Importantly, it was found that this capillary-based substrate with higher sensitivity was essentially attributed to more valid nanoparticles in the effective laser excitation region derived from the unique structure of the capillary. Furthermore, the applicability of the Au@4-MBA@Ag nanorod-decorated capillary for the quantitative identification of fungicides (malachite green and crystal violet) on the shell was demonstrated. As a result, this proposed lab-on-capillary sensor holds promising practical potential for rapid on-site analysis, especially for various contaminants on an uneven surface.

Flexible fabrication of a paper-fluidic SERS sensor coated with a monolayer of core-shell nanospheres for reliable quantitative SERS measurements.

In recent years, paper-based Surface-enhanced Raman spectroscopy (SERS) substrates have received extensive attention in the field of rapid analysis. However, obtaining quantitative SERS results is still challenging because of the inferior uniformity originating from the irregular morphology of the filter paper. In this work, a novel paper-fluidic SERS sensor was developed and its in-depth applications in the real-word quantitative analysis of contaminants in complex matrices were demonstrated. In particular, the Au@Ag core-shell nanospheres were labeled with an internal standard molecule to successfully normalize the fluctuation of the SERS signal caused by the microstructure of the filter paper, which could significantly improve the detection accuracy and accomplish the SERS quantitative analysis. In addition, a facile and robust strategy for the fabrication of a paper-based SERS sensor, which uses a dropper and mask to transfer the nanoparticle monolayers, was developed. This convenient and flexible approach effectively achieved a precise patterned assembly of nanoparticles on the filter paper. Furthermore, the paper-fluidic SERS sensor was fabricated by cutting and packaging for two detection modes, i.e., lateral-flow and vertical-flow, which generates the functionalization of the paper-based SERS substrate. Both detection modes integrated sample pretreatment and sample enrichment with SERS detection were applied to accurately detect the pesticide thiram in a complex sample of orange juice with pulp. In summary, this paper-fluidic SERS sensor with a simple preparation process and integrated functions is an ideal candidate for real sample analysis without pretreatment.

Detection of Chlortetracycline Hydrochloride in Milk with a Solid SERS Substrate Based on Self-assembled Gold Nanobipyramids.

This paper described how a high-yield, monodisperse Au nanobipyramids (Au NBs) sol was prepared by a seed-mediated method, and gold nanoparticles were assembled on the surface of a silicon wafer by self-assembly technology to obtain a solid SERS substrate. Scanning electron microscopy (SEM) showed that the average length of Au NBs was 34.31 nm, and the analysis enhancement factor (AEF) was approximately 7.3 × 105 with rhodamine 6G (R6G) used as a probe. SERS detection of chlortetracycline hydrochloride (CCH) in milk was performed utilizing the prepared Au NBs substrate, and the limit of detection was 0.01 mg/mL. In the range of 0.01 - 1 mg/mL, the mass concentration of CCH and the SERS signal intensity satisfied the linear relationship y = 258.467x + 150.501; the value of the correlation coefficient was 0.9785. In addition, the recovery of spiked samples fluctuated between 96.80 to 111.38%. These results proved that the method is simple and fast, and it is promising to be applied to the field detection of antibiotics in milk.

A Rapid Detection Method for On-site Screening of Estazolam in Beverages with Au@Ag Core-shell Nanoparticles Paper-based SERS Substrate.

Estazolam (EST) is a common sedative-hypnotic drug with a risk of abuse. Therefore, rapid on-site detection of EST is necessary to control the abuse of EST. In this paper, a fast, simple, and sensitive method is demonstrated for the detection of EST in both water and beverages, using surface-enhanced Raman spectroscopy (SERS) techniques. Au@Ag core-shell nanoparticles (NPs) assembled on the filter paper as a SERS substrate exhibit good applicability and practicality. At the same time, density functional theory (DFT) is used to assign the vibration mode of the EST molecules, which can be used as a guide for subsequent experiments. The lowest detectable concentration of EST in aqueous solution can be as low as 5 mg/L, and signal uniformity is excellent (RSD687 = 5.56%, RSD1000 = 4.35%). In addition, EST components artificially added to orange juice and pomegranate juice can be effectively detected by simple pretreatment with a minimum detection concentration as low as 10 mg/L. Therefore, this study found that the use of Au@Ag core-shell nanoparticles paper-based SERS substrate provides a quick and easy method for the detection of illegally added drugs in beverages.

Rapid and Sensitive Surface-enhanced Raman Spectroscopy Method for Determination of Ketamine in Urine.

Surface-enhanced Raman spectroscopy (SERS) has gradually proved to be a powerful tool with wide applications in various fields. Here, a simple and rapid SERS method was developed for the determination of ketamine in urine based on silver aggregates as a SERS substrate. Ketamine in urine were demonstrated by the SERS technique with silver sol aggregated by a 0.5 M NaBr solution. The limit of detection for ketamine in urine could be obtained as low as 7.5 ppm, and a linear relationship for ketamine in urine between the Raman intensity and the concentrations was achieved in the range from 7.5 to 150 ppm (R2 = 0.977). Additionally, the recovery of this method ranged from 95.7 to 104.9%, which laid a favorable foundation for the rapid and reliable quantitative detection of ketamine in urine. Therefore, this SERS approach with high sensitivity and simplicity has a great prospect for the real-world application of ketamine in urine.

Highly monodisperse Au@Ag nanospheres: synthesis by controlled etching route and size-dependent SERS performance of their surperlattices.

For accurate experimental and theoretical research, the preparation of nanocrystrals with regular morphology is of significant importance. In this work, we investigated a facile and effective route for generating highly spherical Au@Ag nanospheres (NSs) with tuneable size and uniform morphology at room temperature. The aqueous synthesis mainly involved seed-mediated growth method together with oxidation etching employing sodium hypochlorite (NaClO) as etching agent. The termination of the etching reaction with NaClO as etching agent was simply related to the amount of NaClO and had no connection with time. Thus Au@Ag NSs with controllable diameters in range from 24 to 87 nm were prepared only by varying the amount of NaClO added into the solutions of Au@Ag nanocubes. Additionally, combined with interface self-assembly technique, Au@Ag NSs were assembled into densely arranged two-dimensional (2D) monolayer film. Moreover, the SERS performance of these monolayers were evaluated by calculating the analytical enhancement factor using crystal violet as probe molecule. The AEF increased obviously as the diameter of Au@Ag NSs went up, and the maximum AEF could reach to 0.94 × 107 at the laser excitation wavelength of 785 nm. Besides, the electromagnetic field distribution for Au@Ag NSs array were also confirmed by Mie theory and FDTD solutions software and the results revealed the similar trend with the experimental results. In general, this 2D assemblies in term of high quality Au@Ag NSs have broad prospects to act as promising candidates for SERS analytical sensor and other applications.

Detection and Quantification of Bucinnazine Hydrochloride Injection Based on SERS Technology.

In this investigation, surface-enhanced Raman spectroscopy (SERS) technology was performed to detect bucinnazine hydrochloride (BH) injection in water and urine. The theoretical Raman spectrum of BH with characteristic peaks was calculated and identified by density functional theory (DFT). Employing an improved silver sol as a SERS active substrate, the SERS spectra of a BH solution with different concentrations were acquired with a 0.5 mol/L KI solution as an aggregation agent. It was determined that the limit of detection (LOD) was low, to 0.01 µg/mL. A good linear relationship of BH between the Raman intensity and the concentrations was obtained in water at a concentration range from 0.5 to 6 µg/mL (R2 = 0.9914), which laid a favorable foundation for quantitative analysis. In addition, the recovery rate of spiked samples from 95.13 to 120.54% were calculated. Finally, the detection of BH injection in artificial urine was completed and the detection limit could reach 0.5 µg/mL, which met the requirements of a rapid on-site detection of drugs in urine. As a result, it indicates that the inspection of BH by the SERS method is with simplicity and high sensitivity, having a great potential for real-time detection.

Detection of Scopolamine Hydrobromide via Surface-enhanced Raman Spectroscopy.

Surface-enhanced Raman spectroscopy (SERS) was used to measure scopolamine hydrobromide. First, the Raman characteristic peaks of scopolamine hydrobromide were assigned, and the characteristic peaks were determined. The optimal aggregation agent was potassium iodide based on a comparative experimental study. Finally, the SERS spectrum of scopolamine hydrobromide was detected in aqueous solution, and the semi-quantitative analysis and the recovery rate were determined via a linear fitting. The detection limit of scopolamine hydrobromide in aqueous solution was 0.5 µg/mL. From 0 - 10 µg/mL, the curve of the intensity of the Raman characteristic peak of scopolamine hydrobromide at 1002 cm-1 is y = 4017.76 + 642.47x. The correlation coefficient was R2 = 0.983, the recovery was 98.5 - 109.7%, and the relative standard deviation (RSD) was about 5.5%. This method is fast, accurate, non-destructive and simple for the detection of scopolamine hydrobromide.

Rapid Detection of Diazepam Injection Based on Surface Enhanced Raman Spectroscopy.

Diazepam injection was detected based on a droplet surface enhanced Raman spectroscopy (SERS) platform, which overcomes the disadvantages of the poor uniformity and time-consuming sample treatment process of the conventional "drop and dry" detection strategy. The Raman peak positions of diazepam injection were determined and identified, they are mainly located in the 689, 1002, 1170 and 1598 cm-1, etc. Different concentrations of diazepam injection were detected. It was found that the intensity of 1002 cm-1 increases linearly with concentration in the range from 0.05 to 10 µg/mL and the linear correlation coefficient is 0.988. The limit of detection can reach 0.05 µg/mL. The SERS method is easy, fast and efficient. The results are accurate and reliable. It has favorable application potential in the detection of diazepam injection.

[Study on Detection of Prohibited Pigments in Drinks Using Paper-Based SERS Substrates].

In this paper, a novel paper-based Surface enhanced Raman spectroscopy (SERS) substrate with high sensitivity, good uniformity and popular price is developed via liquid/liquid interface self-assembly technique. Three pigment, rhodamine B, sunset yellow and chrysoidine were detected through paper-based SERS substrates using a portable Raman spectrometer. The structures of the three pigments were investigated and vibrational modes of characteristic peaks of three pigments were assigned. SERS spectra of rhodamine B, sunset yellow and chrysoidine in aqueous solution with different concentrations were detected respectively. Rhodamine B, sunset yellow and chrysoidine in drinks were also detected in drinks without any pretreatment. Within a certain range of concentrations, it meets certain function. For rhodamine B and sunset yellow, the relationship between concentration and Raman peak intensity is on an index curve, while for chrysoidine, the relationship is linear. In addition, high recoveries are achieved for detecting rhodamine B, sunset yellow and chrysoidine in drinks, which indicated our method is suited for semi-quantitative analysis for the concentration of rhodamine B, sunset yellow and chrysoidine in drinks. Surface-enhanced Raman spectroscopy provides an easy approach to fast and efficient detection for multiple pigments in drinks and can be used for quality control and market monitoring of drinks.

Fabrication of transparent SERS platform via interface self-assembly of gold nanorods and gel trapping technique for on-site real time detection.

A transparent SERS platform was fabricated via the gel-trapping method coupled with a liquid/liquid interface self-assembly technique. We employed gold nanorods as the building blocks for interface self-assembly because of their strong localized surface plasmons upon excitation by infrared radiation. Based on a "top cover" configuration, this transparent SERS platform endows high signal reproducibility for directly detecting liquid samples by confining the sample droplet into a regular shape. The Au NR PDMS platform was able to directly detect crystal violet in aqueous solutions down to 10 ppb level with high enhancement factor (0.87 × 10(5)) and signal uniformity (RSD = 3.9%). Furthermore, SERS-based anti-fungal agent detection on a fish scale was demonstrated by simply covering the fish scale with a tailored GNRs PDMS film. The experimental results clearly show that the Au NR PDMS SERS platform has great potential for on-site real time detection of contaminants in water as well as on curved surfaces.