Fingerprinting black tea: When spectroscopy meets machine learning a novel workflow for geographical origin identification.

Food fraud, along with many challenges to the integrity and sustainability, threatens the prosperity of businesses and society as a whole. Tea is the second most commonly consumed non-alcoholic beverage globally. Challenges to tea authenticity require the development of highly efficient and rapid solutions to improve supply chain transparency. This study has produced an innovative workflow for black tea geographical indications (GI) discrimination based on non-targeted spectroscopic fingerprinting techniques. A total of 360 samples originating from nine GI regions worldwide were analysed by Fourier Transform Infrared (FTIR) and Near Infrared spectroscopy. Machine learning algorithms (k-nearest neighbours and support vector machine models) applied to the test data greatly improved the GI identification achieving 100% accuracy using FTIR. This workflow will provide a low-cost and user-friendly solution for on-site and real-time determination of black tea geographical origin along supply chains.

Data fusion and multivariate analysis for food authenticity analysis.

A mid-level data fusion coupled with multivariate analysis approach is applied to dual-platform mass spectrometry data sets using Rapid Evaporative Ionization Mass Spectrometry and Inductively Coupled Plasma Mass Spectrometry to determine the correct classification of salmon origin and production methods. Salmon (n = 522) from five different regions and two production methods are used in the study. The method achieves a cross-validation classification accuracy of 100% and all test samples (n = 17) have their origins correctly determined, which is not possible with single-platform methods. Eighteen robust lipid markers and nine elemental markers are found, which provide robust evidence of the provenance of the salmon. Thus, we demonstrate that our mid-level data fusion - multivariate analysis strategy greatly improves the ability to correctly identify the geographical origin and production method of salmon, and this innovative approach can be applied to many other food authenticity applications.

Multivariate versus machine learning-based classification of rapid evaporative Ionisation mass spectrometry spectra towards industry based large-scale fish speciation.

Detection and prevention of fish food fraud are of ever-increasing importance, prompting the need for rapid, high-throughput fish speciation techniques. Rapid Evaporative Ionisation Mass Spectrometry (REIMS) has quickly established itself as a powerful technique for the instant in situ analysis of foodstuffs. In the current study, a total of 1736 samples (2015-2021) - comprising 17 different commercially valuable fish species - were analysed using iKnife-REIMS, followed by classification with various multivariate and machine learning strategies. The results demonstrated that multivariate models, i.e. PCA-LDA and (O)PLS-DA, delivered accuracies from 92.5 to 100.0%, while RF and SVM-based classification generated accuracies from 88.7 to 96.3%. Real-time recognition on a separate test set of 432 samples (2022) generated correct speciation between 89.6 and 99.5% for the multivariate models, while the ML models underperformed (22.3-95.1%), in particular for the white fish species. As such, we propose a real-time validated modelling strategy using directly amenable PCA-LDA for rapid industry-proof large-scale fish speciation.

Elementomics combined with dd-SIMCA and K-NN to identify the geographical origin of rice samples from China, India, and Vietnam.

The COVID-19 pandemic has impacted the food industry and consumers, with production gaps, shipping delays, and changes in supply and demand leading to an increased risk of food fraud. Rice has a high probability for adulteration by food fraudsters, being a staple commodity for more than half the global population, making the assessment of geographical origins of rice for authenticity important in terms of protecting businesses and consumers. In this study, we describe ICP-MS elemental profiling coupled with elementomic modelling to identify the geographical indications of Indian, Chinese, and Vietnamese rice. A PLS-DA model exhibited good discrimination (R2 = 0.8393, Q2 = 0.7673, accuracy = 1.0). Data-driven soft independent modelling of class analogy (dd-SIMCA) and K-nearest neighbours (K-NN) models have good sensitivity (98%) and specificity (100%).

Identification of milk from different animal and plant sources by desorption electrospray ionisation high-resolution mass spectrometry (DESI-MS).

This study used desorption electrospray ionisation mass spectrometry (DESI-MS) to analyse and detect and classify biomarkers in five different animal and plant sources of milk for the first time. A range of differences in terms of features was observed in the spectra of cow milk, goat milk, camel milk, soya milk, and oat milk. Chemometric modelling was then used to classify the mass spectra data, enabling unique or significant markers for each milk source to be identified. The classification of different milk sources was achieved with a cross-validation percentage rate of 100% through linear discriminate analysis (LDA) with high sensitivity to adulteration (0.1-5% v/v). The DESI-MS results from the milk samples analysed show the methodology to have high classification accuracy, and in the absence of complex sample clean-up which is often associated with authenticity testing, to be a rapid and efficient approach for milk fraud control.

Authentication of organically grown vegetables by the application of ambient mass spectrometry and inductively coupled plasma (ICP) mass spectrometry; The leek case study.

Health conscious and environmentally aware consumers are purchasing more organically produced foods. They prefer organic fruits and leafy vegetables as these are much less likely to have been exposed to contaminants such as pesticides. The detection of fraudulent activity in this area is difficult to undertake, because many chemical plant protection treatments degrade very quickly or can be washed off to remove evidence of their existence. It was found that when combining DART-MS with a compact, inexpensive and robust single quadrupole mass spectrometer, it was possible to differentiate organic from conventional leeks with 93.8% to 100% accuracy. ICP-MS results showed similar performance, with an ability to differentiate conventional from organic leeks with 92.5% to 98.1% accuracy. This study has paved the way for the certification of vegetables as being organically produced. The next step is to create data libraries to support the roll out of the methodologies described.

Simultaneous Detection of Eight Prohibited Flavor Compounds in Foodstuffs Using Gas Chromatography-Tandem Mass Spectrometry.

A multiflavor detection method, using gas chromatography-triple quadrupole tandem mass spectrometry (GC-MS/MS), has been developed for the simultaneous identification and quantification of eight prohibited flavor compounds in daily foods. Under the optimized extraction conditions, samples were purified directly through membrane filtration. Variables affecting the GC-MS/MS were optimized to obtain better separation. The excellent selectivity and sensitivity achieved in multiple reactions monitoring mode allowed satisfactory confirmation and quantitation. In this study, the linear ranges of the target compounds were 0.05 to 500 ng/L with good correlation coefficients ( R2 > 0.999). The limits of detection of target compounds ranged from 0.005 to 0.2 µg/kg. The average recoveries were in the range of 80.2 to 110.6% (beef jerky), 82.3 to 94.1% (cod liver oil), and 83.6 to 104.1% (candy).

[Screening of 17 cephalosporin residues in infant formula milk powder by ultra-high performance liquid chromatography-high resolution mass spectrometry].

A method for the simultaneous determination of 17 cephalosporinsin milk powder by ultra-high performance liquid chromatography-quadrupole/electrostatic field orbitrap high resolution mass spectrometry (UHPLC-Q-Orbitrap HRMS) has been developed. After dissolution in water, the homogenized milk powder samples were extracted with acetonitrile to remove impurities such as proteins and the extract was further purified by high-speed centrifugation at low temperature. A C8 chromatographic column was employed to separate the analytes. Quantitative and qualitative analyses were performed in Q Exactive Full-MS/dd-MS2 scan mode. The results revealed an excellent linear correlation over a wide range of analyte mass concentrations (5-200 µg/L) with coefficients of determination (r2) greater than 0.99 in all cases. The limits of detection (LODs) and quantitation (LOQs) were in the range of 0.0060-0.014 µg/L and 0.71-4.62 µg/kg, respectively. The overall average recoveries were 69.6%-101.4% with RSDs<10%. This method presents the advantages of simple operation, good repeatability, and high resolution, being suitable for the qualitative and quantitative screening of 17 cephalosporin residues in infant formula milk powder.

A high-throughput screening method of bisphenols, bisphenols digycidyl ethers and their derivatives in dairy products by ultra-high performance liquid chromatography-tandem mass spectrometry.

A simple and universal analytical method based on ultra-high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) for high throughput screening of 21 bisphenols, bisphenols digycidyl ethers and their derivatives in dairy products was developed. Response Surface Methodology (RSM) was used to optimize sample preparation conditions based on a quick, easy, cheap, effective, rugged and safe (QuEChERS) method. The analytes were extracted by using 15 mL acetonitrile with 1% acetic acid, and the extracts were further purified by using 190 mg of C18 and 390 mg of PSA. The extracts were analyzed by UHPLC-MS/MS with electrospray ionization (ESI) source. Linearity was assessed by using matrix-matched standard calibration and good correlation coefficients (r2 > 0.99) were obtained. The limits of quantitation (LOQs) for the analytes ranged from 0.02 to 5 µg kg-1. The extraction recoveries were in a range of 88.2%-108.2%. Good method reproducibility in terms of intra- and inter-day precision was observed, yielding relative standard deviations (RSDs) less than 8.9% and 9.9%, respectively. The validation method results revealed that the proposed method was sensitive and reliable. Finally, this method was successfully applied to dairy product analysis.

Simultaneous determination of 22 cephalosporins drug residues in pork muscle using liquid chromatography-tandem mass spectrometry.

A simple, sensitive and reliable analytical method was developed for the simultaneous determination of 22 common cephalosporins from the first generation to the fourth generation in pork muscle by liquid chromatographic-tandem mass spectrometric (LC-MS/MS) method. Under the optimized extraction conditions, samples were directly purified through membrane filtration to separate all 22 cephalosporins and the critical pairs of each parent drug were completely separated. Variables affecting the LC-MS/MS were optimized to get a better separation. The excellent selectivity and sensitivity achieved in multiple reactions monitoring (MRM) mode allowed satisfactory confirmation and quantitation for the 22 cephalosporins. The linear range of the 22 cephalosporins is 0.06-100.0µg/L with good correlation coefficients (r(2)>0.9920). The limits of detection (LODs) and limits of quantitation (LOQs) of these compounds were in the range 0.04-3.0µg/L and 0.06-10.0µg/kg, respectively. The average intra-day recoveries at 3 spiked levels (LOQ, 2LOQ, 4LOQ) were all in the range 83.6-113.0% with RSDs (n=6) lower than 6.5%. The method of LC-MS/MS developed in this study was initially applied to the research of 22 cephalosporins in 12 retail pork samples and proved to be accurate, sensitive, minimum sample pre-treatment, convenient and practical.