Unsupervised Analysis of Optical Imaging Data for the Discovery of Reactivity Patterns in Metal Alloy

Abstract

Operando wide‐field optical microscopy imaging yields a wealth of information about the reactivity of metal interfaces, yet the data are often unstructured and challenging to process. In this study, the power of unsupervised machine learning (ML) algorithms is harnessed to analyze chemical reactivity images obtained dynamically by reflectivity microscopy in combination with ex situ scanning electron microscopy to identify and cluster the chemical reactivity of particles in Al alloy. The ML analysis uncovers three distinct clusters of reactivity from unlabeled datasets. A detailed examination of representative reactivity patterns confirms the chemical communication of generated OH− fluxes within particles, as supported by statistical analysis of size distribution and finite element modelling (FEM). The ML procedures also reveal statistically significant patterns of reactivity under dynamic conditions, such as pH acidification. The results align well with a numerical model of chemical communication, underscoring the synergy between data‐driven ML and physics‐driven FEM approaches.

Operando Surface Optical Reflectance Microscopy Study of Corrosion Film Growth on a Ni–Cr–Mo Alloy During Anodic Polarization

“…We use an emerging optical technique termed 2-Dimensional surface optical reflectance (2D-SOR) microscopy − (also known as reflectance microscopy), to follow the corrosion initiation and progression in operando conditions with high temporal and spatial resolution. 2D-SOR has previously been used to study the corrosion of duplex stainless steel, the anodization of aluminum, the local corrosion of aluminum alloys, − the electrooxidation of gold, ,, and corrosion inhibition. , Here we demonstrate the additional capabilities of this affordable and straightforward technique. During wet corrosion of the Ni–Cr–Mo alloy at high anodic potentials, we detect films of corrosion products forming on the surface.…”

Key Requirements for Advancing Machine Learning Approaches in Single Entity Electrochemistry

“…3). [62] The objective of this method is to accelerate data analysis while minimizing subjective human influence. In cases like this, such analysis becomes indispensable, as the raw, high-resolution, wide-field videos can exceed 100 gigabytes, making manual processing impractical.…”

Deciphering the Interplay between Local and Global Dynamics of Anodic Metal Oxidation

“…Recently, optical models utilizing light interference at oxidized metal interfaces have shown that variations in optical intensities can quantitatively relate to the thickness (to nanometer precision) and composition of surface films. ,,, These parameters are central to our understanding of metal passivation and corrosion protection. When augmented with machine vision methodologies, , RM evolves into an automated, robust, and flexible approach. Compared to traditional in situ methods for characterizing passive films, such as scanning tunneling microscopy (STM), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), and others, it is significantly less costly and time-consuming. ,− The versatility and efficacy of the RM approach are further demonstrated in this study through the examination of corrosion in pure polycrystalline Fe metal in a NaCl medium.…”