Electrochemical analysis has become a new method for plant analysis in recent years. It can not only collect signals of electrochemically active substances in plant tissues, but can also be used to identify plant species. At the same time, the signals of electrochemically active substances in plant tissues can also be used to investigate plant phylogeny. In this work, we collected electrochemical finger patterns in Malvaceae leaves based on the established methodological strategy. After the second derivative treatment, the collected electrochemical fingerprints can show more obvious differences. Three different recognition models were used to attempt electrochemical fingerprinting. The results show that linear support vector classification can be used to identify species with high accuracy by combining the electrochemical fingerprint signals collected in the phosphoric acid buffer solution and acetic acid buffer solution. In addition, the fingerprint information collected by the electrochemical sensor is further used for phylogenetic investigation. The 18 species were divided into three clusters. Species of the same genus have been clustered together. Dendrogram obtained by electrochemical fingerprinting was used to compare previously reported results deduced from morphological and complete chloroplast genomes.
Keywords: electrochemical fingerprint; electrochemical sensor; pattern recognition; phylogenetics; phytochemistry.