Fabrication of 1D/2D Au nanofiber/MIL-101(Cr)–NH2 composite for selective electrochemical detection of caffeic acid: Predicting sensor performance by machine learning and investigating the porosity using AI and computer vision-based image analysis

Abstract

Artificial intelligence, including machine learning, can offer creative solutions for problems that sensors must solve to anticipate the concentrations of analyte automatically. In this article, a machine learning approach was used to predict the sensing performance of the 1D gold nanofibers decorated 2D amine-terminated chromium metal–organic framework (MIL-101(Cr)–NH2) composite for the determination of caffeic acid (CA). The MIL-101(Cr)–NH2 surface was decorated with Au nanofibers with an average diameter of 12 nm, according to the morphological examination. The composite demonstrated a good linear range of CA concentrations from 0.5 to 100 μM with a detection limit of 0.011 µM and a sensitivity of 2.53 µA/µM/cm2. The electrode's production of current for the interfering substances was incredibly low. The spiked CA in the coffee powder and red wine samples recovered exceptionally well using the composite electrode. The machine learning design forecasted the sensing efficiency of CA to support the experimental results. Linear regression, the most trivial machine learning algorithm, produced predictions that closely matched the experimental data. The composite's porosity and potential electrochemical traits were also investigated using computer vision and artificial intelligence-based algorithms and compared with the experimental results. © 2024 Elsevier B.V.