Skip to main content
Physical Sciences and Engineering

Molecular Materials

The research carried out in the Molecular Materials Group focuses on the AI-accelerated design and discovery of new materials for energy storage and conversion.
Illustration of down-selection of candidate molecules for electrical energy storage electrolytes based on high-throughput quantum chemical calculations.

The Molecular Materials Group research focuses on the design and discovery of new materials and interfaces for advanced energy storage and the frontiers of catalysis. The Group is involved in closely integrated theoretical and experimental research that encompasses computations, machine learning, characterization, and synthesis.

State-of-the-art computational approaches are used to understand materials and interfacial properties leading to predictive design of new materials with desired properties. The approaches we use include electronic structure methods, such as density functional theory and highly accurate wave-function-based methods, quantum Monte Carlo methods, and classical molecular dynamics. We utilize and develop state-of-the-art machine learning techniques, such as delta learning, active learning, graph neural networks, and foundational language models  for materials design and discovery. We are also interested in high-throughput computations and autonomous materials design of new energy storage and catalytic materials. In the area of energy storage materials, we are carrying out experimentation on advanced energy storage concepts, including Li-ion and Li-oxygen.

Current research topics on energy storage and conversion include:
Designing interfaces

Fundamental research on electrochemical interfaces is carried out to advance materials that are employed in electrochemical systems for energy conversion and storage, namely, batteries, fuel cells, and electrolyzes. The goal is understanding electrochemical phenomena to develop materials that are active, stable, selective, and conductive. We employ computational methods, such as density functional theory and ab initio molecular dynamics, to provide understanding of interfaces that are being investigated experimentally. The joint work is being used to identify key physical parameters that determine functional properties of interfaces at the atomic scale. As an example, we carry out studies of solid-liquid interfaces in Li-ion batteries to find key reactions contributing to their long-time behavior.

Liquid electrolytes  

One of the main challenges in developing advanced batteries is the selection and discovery of suitable liquid electrolyte materials. Generally, the minimal requirement for the solvents and salts that make up the electrolyte is that (i) they need to be stable within the electrochemical window of the battery to not degrade performance or safety and (ii) the solvent must provide good solubility for the salt to ensure sufficient ionic conductivity. Furthermore, the specific energy storage technologies can present additional requirements for the electrolytes. We are using a high fidelity computations-enabled machine learning approach for understanding and predicting various aspects of liquid electrolytes, including solvation structure, transport properties, redox windows, and stability. 

Solid state electrolytes

Replacement of the liquid electrolyte with a solid-state electrolyte is a goal of advanced battery systems to decrease the fire hazard and improve safety associated with present-day Li-ion batteries. Thus, the use of solid electrolytes enables use of lithium anodes, which can lead to batteries with significantly higher energy density.  We are carrying out computational studies to investigate the structure, stability, and transport at interfaces between potential solid-state electrolytes and lithium metal or cathode materials. The computational studies are integrated with experimental characterization of model and real systems and continuum modeling. 

Next generation electrochemistry

Beyond lithium-ion batteries are considered as a potential alternative to Li-ion batteries for transportation applications due to their high theoretical specific energy. We are contributing to research on various types of these advanced systems for transportation, such as multivalent batteries where the Li ion is replaced by a divalent ion, such as Mg or Ca, and chemical transformation batteries such as Li-oxygen and Li-sulfur. For grid storage our research involves flow batteries, which  use liquid solutions of redox molecules instead of the solid electrodes of conventional batteries. We are using a comprehensive computational approach combined with experiment to gain understanding of the new processes involved in these systems and helping in the discovery of new materials to enable cycle life and efficiency.  

Advanced materials synthesis

Synthesis of the predicted materials with novel functionality is often complicated due to their metastability and insufficient control of kinetic pathways. We are using computational methods to study and predict reaction pathways of the precursors to understand the kinetics of growth processes and inform kinetic modeling and experimental studies. These studies also include surface reactions occurring in metal organic chemical vapor deposition, atomic layer deposition, and molecular beam epitaxy. The knowledge of synthesis pathways leading to desired composition and phase can accelerate development of new materials for energy applications.

Carbon utilization via catalysis

The creation of a bioenergy industry is an important pathway for providing sustainable and renewable energy alternatives. To overcome this, there are catalysis challenges for the conversion of biomass and waste resources into fuels, chemicals, and materials. We are using predictive simulations and catalyst informatics  to discover improved catalysts by computing properties such as adsorption energies, vibrational spectra, equilibrium surface compositions under relevant conditions, and kinetic parameters along proposed reaction mechanisms, and we are elucidating trends in reactivity. The work is part of a consortium of five national laboratories combining molecular-scale chemistry to process-scale operations, integrated with experiment.  

Machine learning applications to energy storage and catalysis

Machine learning enables predictions based on patterns in large amounts of data, and it has become increasingly important to scientific discovery. We are integrating machine learning with our computational studies to accelerate electrolyte discovery, redox active flow battery molecules, liquid organic hydrogen carriers, and catalytic materials. Machine learning is being used to help fill fundamental science gaps and enable discovery of transformative material