Research Spotlight: How Computer Simulation Can Be Used to Predict Behaviour in Corrosion Systems
By Brianna Rector, PhD Candidate in the Wren Group
Aqueous corrosion occurs at the interface where metal and solution are in contact. The corrosion process has many individual steps, that when combined, eventually lead to the production of oxides, such as rust.
Transition metals such as iron, copper, and nickel are found in a wide range of alloys commonly used in industry. One concern with the use of these alloys in industrial settings is the long-term potential for corrosion to weaken them and shorten their lifespans. Aqueous corrosion involves the transfer of both electrons and metal atoms at the interface where metal and solution are in contact. Determining how fast or to what extent a material will corrode at this interface is not a simple process. This is because corrosion is a complex and dynamic process, which can be influenced by many factors specific to a given corrosion system, such as pH or flow rate.
Corrosion consists of many elementary electrochemical, chemical, and transport steps that produce various solvated corrosion products and solid oxides (such as rust). The rate of each of these steps will depend in specific ways on solution properties such as pH, ionic strength, volume, flow rate, temperature, presence of radiation, and type and concentration of oxidant(s). Further complicating the process is the potential for some of these steps to couple and form feedback loops, which occur when the product of one step can affect the rate of a preceding step. These feedback loops can result in periodic patterns in observable chemical phenomena, which we refer to as chemical waves.
Electrochemical tests and optical images demonstrating chemical waves, which occur due to the coupling of elementary steps.
Studying corrosion is difficult as it is a slow process that can occur over many years, whereas corrosion experiments are typically restricted by laboratory time scales (e.g., we cannot easily study a single corrosion case over 100 years). However, a precise and detailed understanding of how corrosion changes with time is critical for the prediction of how a material will corrode in a particular environment. Many existing corrosion mechanisms and corresponding models assume that the system will quickly reach a single steady state, the behaviour of which can be extrapolated to long timescales. Recent research in the RAMPS group, however, has shown that corrosion is a dynamic process that progresses through a number of different stages and steady states with time. Being able to determine how solution properties affect the progression through these different steady states is critical for the development of reliable predictive models.
Computer simulations (bottom) of electrochemical data (top) in carbon steel corrosion at pH 7.0. The modelling results demonstrate that we can simulate key behaviours observed experimentally, including current vs. time initial slopes and limiting current behaviours, using the modelling approach described.
Computer simulation is a powerful tool that we can use to address the difficulty in predicting corrosion over long timescales. Developing a dynamic model to better simulate how corrosion progresses with time under various solution conditions and environments is a current focus of ongoing research in the RAMPS group.
While the use of computer simulation is not new in corrosion science, many existing rate models cannot predict the coupling between steps which leads to chemical wave phenomena, and are often limited to a narrow range of parameters or timescales. Moreover, many of these models are empirical data-fitting models involving many parameters, used when the physical and chemical processes are difficult to identify. These types of models are difficult to use in a predictive manner and cannot be applied to conditions outside the experimentally studied ranges.
Research within the RAMPS group has focused on developing a modelling approach that balances the need for mechanistic detail with the need for practicality (i.e. minimizing complexity and computer processing time). To develop a corrosion model that can be used to simulate corrosion under a wide range of solution and environmental conditions, we must:
(1) Identify the key steps that contribute to the overall corrosion process.
(2) Formulate rate and flux equations for these elementary steps as a function of solution parameters.
(3) Couple the rate equations for individual elementary steps to form the overall rate equation.
(4) Solve the overall rate equation.