Electronic Tongue Based on Composites of Metal Phthalocyanine and Carbon Nanotubes and Electrochemically Deposited Metal Nanoparticles for Metal Ions Detection Enhanced by Machine Learning
Tianqi Lu, Ammar Al‐Hamry, Anurag Adiraju, Mengze Yan, Renati Azhati, Malak Talbi, Zhenyu Wu, Ge Shi, Olfa Kanoun
Abstract
As essential elements for plant and animal growth, the precise qualitative and quantitative recognition of K <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> , Ca <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2+</sup> , and Mg <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2+</sup> becomes important for smart agriculture fields such as aquaponic, soil less cultivation to name a few.. Sensor array applying artificial gustatory, i.e., electronic tongue, plays an important role in this field and is often employed to analyze the categories and concentrations of ions in smart agriculture. This contribution investigated the detection of K <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> , Ca <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2+</sup> , and Mg <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2+</sup> by voltametric electronic tongue based on composites of COOH-functionalized multi-walled carbon nanotubes (MWCNT-COOH) and metal phthalocyanine, and electrochemically deposited gold/ silver nanoparticles. The proposed electronic tongue consists of four paper-based screen-printed carbon electrodes modified with composites. The surface of the working electrode modified with the composite were further modified by electrochemically deposited of gold or silver nanoparticles and its electrochemical behavior was characterized. By executing cyclic voltammetry, each sensor discriminates metal ions at concentrations of 100 nM to 10 μM. Reproducibility and long-term stability in 10 months are validated. The sensing properties of the sensor array are excellent and can distinguish between the categories and concentrations. The recognition results using PCA-SVM (Gaussian-kernel) are proved as the best framework and achieve total recognition as 98%. PLSR is also applied for the recognition analysis of the metal ions concentrations. The proposed electronic tongue can be well used for the quantitative recognition of K <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> , Ca <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2+</sup> , and Mg <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2+</sup> , and demonstrate its potential to be employed in environmental and medical applications.