PNNL

The Science

Nitrate and nitrite ions are found in significant concentrations in legacy nuclear waste and can form complex precipitates that complicate waste remediation. However, the mechanisms that underly the behavior of the concentrated electrolyte solutions, including ion clustering, remain poorly understood. Researchers performed molecular dynamics simulations of sodium nitrite, sodium nitrate, and mixed solutions of the two and analyzed the ion clustering behavior with new graph-theoretic and machine learning workflows. The team found that the mixed electrolyte exhibited faster aggregate growth and dissolution behavior but slower interconversion between organizational patterns of the clusters. This work provides new insights into the molecular-level behavior of concentrated two-component electrolyte solutions, with implications for predicting changes in solubility relative to the single electrolyte systems.

The Impact

Concentrated aqueous electrolytes are found in industrial processes and legacy nuclear waste, including at the Department of Energy’s Hanford Site. As these electrolytes become increasingly concentrated, the ions can aggregate into large clusters that alter solution viscosity, dynamically evolve because of background radiation, and can lead to precipitation. Modeling these clusters has remained challenging, particularly for mixtures of multiple electrolytes. This work presents a new and transferable machine learning workflow for modeling electrolyte cluster dynamics and supports learning how mixing electrolytes alters solution properties and precipitation processes.

Summary

Mixed aqueous electrolytes are increasingly valuable in industrial applications and often behave very differently from their single-salt counterparts. However, the underlying molecular mechanisms are elusive and not well understood. Researchers performed a comprehensive molecular dynamics simulation study of ion networks in aqueous sodium nitrite (NaNO₂(aq)), sodium nitrate (NaNO₃(aq)), and their mixtures. Combining new graph-theoretic workflows with machine learning enabled the team to identify chemical states on the free energy landscape, composed of ion aggregate macro-states based on the number of cations in the cluster and microstates based on the inter-ion connectivity patterns. Detailed investigation of state transitions in mixed versus single-salt electrolyte systems revealed changes to the cluster free energy landscape upon electrolyte mixing. Specific ion effects were observed, where nitrite ions (NO₂⁻) preferentially partitioned into clusters, which influenced both the growth/dissolution pathways of clusters and their interconversion between microstates. Relative to the single-salt solutions, the mixed electrolyte exhibited faster macro-state aggregate growth and dissolution but slower interconversion between microstates. These behaviors are attributed to substantial changes in the cluster free energy landscape, which may help explain reported solubility shifts in the NaNO₂(aq)/NaNO₃(aq) two-salt solution. More broadly, these data provide molecular-level evidence that anion competition in two-salt electrolytes directly influences concentration-dependent ion cluster behavior, relevant to properties ranging from conductivity in battery and electrochemical applications to solubility in industrial wastes. The graph-theoretic workflow introduced here for identifying microstates and their transitions is extensible and represents a significant computational advancement for studying electrolyte systems.

Contact

Aurora Clark, University of Utah, aurora.clark@utah.edu 

Carolyn Pearce, Pacific Northwest National Laboratory, carolyn.pearce@pnnl.gov

Yihui Wei, University of Utah, u1448045@utah.edu

Funding

This research was supported by Ion Dynamics in Radioactive Environments and Materials (IDREAM), an Energy Frontier Research Center funded by the Department of Energy, Office of Science, Basic Energy Sciences program (FWP 68932).