As a battery operates, the electrolyte undergoes chemical reactions that gradually alter its original formulation. This aging process, which accelerates at high temperatures, can significantly impact the performance of the battery cell.

Studying the effects of electrolyte aging on a battery’s performance is a major challenge in R&D, as it involves a complex process with multiple interdependent reactions influenced by the numerous chemicals present in the electrolyte and environmental conditions such as temperature.  

To rationalize the process of electrolyte aging, computational tools have emerged as a time- and cost-effective alternative to traditional experimental procedures.

Computational tools based on the laws of quantum mechanics can provide unique insights into the effects of aging on the properties of the electrolyte, and therefore its overall performance. One key property of the electrolyte that impacts performance and can be affected by the aging process is viscosity.  

Generally, lower electrolyte viscosity results in higher performance as it facilitates the movement of ions through the electrolyte. Optimizing electrolyte viscosity is an essential step in the development of novel battery materials. Therefore, understanding how viscosity changes as the electrolyte ages is crucial for ensuring the long-term performance of these new materials.

Use case: Effects of aging on electrolyte viscosity

The first step to simulate the effects of aging on an electrolyte is to create a model. A previous use case detailed how QuantistryLab’s quantum nanoreactor feature can be used to simulate the reactions occurring as a commercial electrolyte goes through thermal decomposition. That simulation served as a starting point to create several models of an aged electrolyte, each containing different decomposition by-products.  

The aged version of the commercial electrolyte was created to model the altered ratios of its components — ethylene carbonate, dimethyl carbonate, and the lithium salt (lithium hexafluorophosphate) — as the original formulation ages over time. Using this aged electrolyte as a base, three additional models were developed by introducing different types of by-products of the aging process, previously identified using the quantum nanoreactor feature.

One of these models included gases created during the aging process, such as carbon dioxide and carbon monoxide, which can cause swelling of the battery. Another model contained the main decomposition products with phosphorous-fluoride bonds, known to potentially be toxic. The third model included carbon-based oligomers that can form as degradation products react with each other — a reaction that was simulated separately using the nanoreactor feature.  

Once all the models are created, viscosity can be calculated with just a few clicks by starting a workflow with QuantistryLab’s viscosity feature and setting the desired temperature and pressure for the simulation. The viscosity workflow deploys molecular dynamics simulations to measure fluctuations in the stress tensor over time. Using the Green-Kubo approach, a statistical mechanics method which employs autocorrelation functions of the stress tensor, an estimation of the viscosity at equilibrium is obtained. For very high viscosity values, the platform ensures that the long-time behavior of the fluctuations is well captured by fitting a tail to the autocorrelation function.

The results of the viscosity simulation yield an atomistic model of the electrolyte that represents the dynamical behavior of the atoms that compose it, along with the average viscosity value calculated during the simulation. The same workflow was run separately on each of the aged electrolyte formulations to compare their viscosity values and study the effects of different aging products on the properties of the electrolyte.

The results obtained using QuantistryLab, shown in the graph above, reveal that the viscosity of the electrolyte decreases when gaseous products form during the aging process, and increases when oligomeric compounds are present in the aged electrolyte. A notable conclusion is that, since higher electrolyte viscosity is associated with decreased battery cell performance, creating electrolyte formulations that minimize the production of larger oligomeric compounds can help prevent a decline in battery performance over time.  

These findings demonstrate how simulations can effectively guide the design and optimization of electrolyte formulations. In battery R&D, such insights are invaluable for understanding how electrolyte viscosity changes with aging, offering crucial input for optimizing the performance of new electrolyte materials. The key advantage of simulations is their ability to save time and resources by predicting electrolyte properties through computational models, reducing the reliance on numerous laboratory experiments.

In addition to studying electrolyte formulations under various conditions, QuantistryLab offers a comprehensive suite of computational tools to investigate and optimize other critical material properties relevant to industrial R&D, including electrodes, alloys, and polymers.

‍