An alternative approach to analyze complex matrixes such as milks lies on multivariate description of the chemical composition of samples where non-specific fingerprints are analyzed using data mining methods. Portable electronic tongues (ET) are of great interest for the dairy industry [1]. However, still several problems related to the fouling of the electrodes and the preservation conditions limit their usage in milk analysis. For thus, the sensor´s development methods need to be adapted in order to reduce the fouling of electrodes caused by milk components, among which stand out fat and proteins. For that reason, these kind of matrixes need to be pre-treated to eliminate, or at least substantially diminish their presence.The benefits of using ETs in the dairy industry have been explored by different researchers [2]. These works have been used to discriminate and classify milk samples with different characteristics such as freshness, nutritional composition, adulterations, etc. However, due to its extremely complexity, the fouling of the electrodes reduces the repeatability and lifetime of the sensors. Another factor to take into account is the microbiologic process that start immediately after milking, due to it can modify the composition of milk. For that reason, it is necessary to develop techniques that not only decrease the fouling but limit also the growth of organisms without interfering in the electrochemical measurements.The necessity to optimize the measurement protocols in order to guarantee the capabilities of the sensors systems has led us to analyze and to evaluate the effect of practices such as dilution, sonication, freezing or addition of preservatives such as azidiol on the responses of electrochemical sensors. In this work, a portable system formed by an array of serigraphied sensors coupled to a pattern recognition software has been developed and used to analyze raw milk obtained from cows with different characteristics (lactation days, days of pregnancy, number of pregnancies...). In this work, it has been concluded that the performance of the sensory system is drastically improved if milk samples are diluted and sonicated previous to the analysis. In addition, it has been evidenced that the use of preservation methods such as the addition of azidiol or the freezing at -20ºC has also an important influence in the electrochemical responses. Principal Component Analysis of the signals obtained using the optimized conditions has allowed the discrimination of 12 milks with different characteristics.[1] C. Garcia-Cabezon, G.G. Teixeira, L.G. Dias, C. Salvo-Comino, C. García-Hernandez, M.L. Rodriguez-Mendez, F. Martin-Pedrosa, Analysis of phenolic content in grape seeds and skins by means of a bio-electronic tongue, Sensors (Switzerland). 20 (2020) 1–14. doi:10.3390/s20154176.[2] Lluis Pascual, Marisa Gras, Daniel Vidal-Brotóns, Miguel Alcañiz, Ramón Martínez-Máñez, Jose V.Ros-LisA voltammetric e-tongue tool for the emulation of the sensorial analysis and the discrimination of vegetal milks, Sensors and Actuators B: Chemical, Volume 270, 1 October 2018, Pages 231-238 Acknowledgments: We appreciate the financial support of MINECO-FEDER National Plan (PID2021-122365OB-100). Junta de Castilla y Leon- FEDER VA202P20. CLU-2019-04 and «Infrastructure Network of Castilla y León (INFRARED)» UVA01. Tractor Plan In Advanced Materials Focused On The Key Industrial Sectors In Castilla Y León: Agri-food, Transport, Energy And Construction (MA2TEC). We would also like to thank the Calidad Pascual farm (Aranda de Duero) for the loan of samples (ALIVAC-IDI-20211051).
Read full abstract