First Principles, Explicit Interface Studies of Oxygen Vacancy and Chloride in Alumina Films for Corrosion Applications

TL;DR

This study uses first principles modeling to explore how oxygen vacancies and chloride ions influence corrosion in alumina films, revealing the conditions under which chloride insertion occurs and the role of electric potential in defect formation.

Contribution

It provides new insights into the atomic-scale mechanisms of chloride-induced corrosion at metal/oxide interfaces using explicit interface models.

Findings

Doubly charged oxygen vacancies occur only with a voltage gradient.

Chloride insertion into oxide can be energetically favorable depending on voltage.

Explicit DFT modeling of complex interfaces presents specific challenges.

Abstract

Pitting corrosion is a much-studied and technologically relevant subject. However, the fundamental mechanisms responsible for the breakdown of the passivating oxide layer are still subjects of debate. Chloride anions are known to accelerate corrosion; relevant hypotheses include Cl insertion into positively charged oxygen vacancies in the oxide film, and Cl adsorption on passivating oxide surfaces, substituting for surface hydroxyl groups. In this work, we conduct large-scale first principles modeling of explicit metal/Al(2)O(3) interfaces to investigate the energetics and electronic structures associated with these hypotheses. The explicit interface models allow electron transfer that mimics electrochemical events, and the establishment of the relation between atomic structures at different interfaces and the electronic band alignment. For multiple model interfaces, we find that doubly charged oxygen vacancies, which are key ingredients of the point defect model (PDM) often used to analyze corrosion data, can only occur in the presence of a potential gradient that raises the voltage. Cl-insertion into oxide films can be energetically favorable in some oxygen vacancy sites, depending on the voltage. We also discuss the challenges associated with explicit DFT modeling of these complex interfaces.