(4hj) Development and Implementation of Enhanced Sampling Approaches:Applications to Ion-pairing in Battery Electrolytes and Nucleation ofNano-porous Materials

Fast and accurate simulations of rare phenomena and their energetics are limited by the timescale accessible to conventional molecular dynamics (MD). The development of novel enhanced sampling approaches have provided significant thrusts to overcome those limitations. In this presentation, I will discuss two such examples from my research: one that utilizes an unconventional implementation of an existing method and another that develops a new statistical mechanical method for analyzing complex free energy landscapes.

In the first work, we utilize a biased-sampling (metadynamics) approach on a cluster model consisting of an ion-pair with a single shell of solvent. By sampling transitions between the ion-paired state and solvent separated ion-pair state, we demonstrate that quick estimates of the ion-pairing free energy can be obtained from cluster models that agree well with bulk solvent calculations [1]. However, the cluster model is 1 to 2 orders of magnitude faster than bulk solvent calculations. This has implications for the design of battery electrolytes because ion-pairing can significantly impact conductivities and performance of batteries. Our method provides a novel basis for the screening and assessment of candidates for electrolyte designs.

In the second work, we developed a novel graph theory-based sampling approach [2] for modeling the nucleation of weak electrolytes that overcomes limitations of existing approaches for bulk solvent systems. Our method seeks to exploit the property of materials whose crystal structure exhibit directional bonding and thus can be described as a “graph” of connected monomers. By utilizing a rigorous statistical-mechanics approach, we generate an ensemble of representative nuclei and their corresponding free energies via a “bootstrapping” approach in the nucleus size. Starting with a simple system of LiF(aq), we elucidate the factors governing the nucleation and growth of LiF clusters. Subsequently, this work will be extended to complex materials such as zeolites and metal-organic frameworks.

[1] Ajay Muralidharan, Tyler Lytle, and Arun Yethiraj. “why lithium ions stick to some anions and not others”. The Journal of Physical Chemistry B, 2021.
[2] Ajay Muralidharan, Xinyi Li, and J.R. Schmidt. “a hierarchical graph theory-based sampling approach to study the nucleation of weak electrolytes”. (In preparation), 2021.

Research Interests
My primary research interest is to develop modern theoretical methods that can take advantage of state-of-the-art simulation capability to address the big challenges, specifically, in the area of materials design for energy, separations and catalysis. Recently, the broad interest in the development of biomaterials (e.g. therapeutics based on peptides) have also drawn my attention due to my recent
ventures into that arena.

Question of chemical detail vs accuracy
A significant question that my current research addresses is the following: “How much chemical detail should you include for accurate modelling of ionic species in practical systems?” (e.g. batteries, supercapacitors, complex membrane ion channels). My work has shown that the answer depends on the nature of the system, however, it has led to fundamental theoretical developments that are applicable universally. For instance, an important contribution of my dissertation work was the extension [1] of Quasi Chemical theory, a “van der Waals like” theory that splits the free energy of solution species into components based on the definition of artificial boundaries. This theory provides an exact statistical mechanics framework within which approximations such as the level of chemical detail can be prescribed for desired accuracy. Applications of those ideas led to fruitful research in the areas of energy storage [2, 3] and the interpretation of selectivity in biological ion channels [1, 4].

Hunt for materials
My aspiration is to use my theoretical and modeling skills for the design of materials for energy and biological applications. Thus, I accepted a fellowship at the University of Wisconsin-Madison (UW) where I am now focused on cultivating a blend of independent and collaborative research. My initial collaboration was based on the discovery [5] that fluorinated prolines can lead to faster folding and stable proteins. We introduced a theoretical scheme to model this effect in classical force field simulation models [6]. In the future, this can be extended to design peptide and poly-electrolyte systems with enhanced thermal stability to use as rafts for drug delivery.

Another collaborative project I am working on is the theoretical development [7] and implementation of enhanced sampling approaches for the study of rare events like ion-pairing and nucleation phenomena. Specifically, we combine a graph theory-based approach coupled with thermodynamic integration to address limitations of current approaches for the study of nucleation of weak electrolytes. Starting with simple systems such as aqueous lithium fluoride, we plan to extend future work to understand nucleation principles and guide the synthesis of zeolites and metal-organic frameworks.

Future directions

I am currently developing novel design principles to supplement computational and experimental screening efforts for soft materials in energy applications. A key direction that I am exploring is the combination of enhanced sampling simulations with machine learning and data science techniques. This approach will enable rapid screening by guiding the exploration of high-dimensional design spaces. In conclusion, my future research lab will develop a combination of cutting-edge theory, modelling and machine learning capability with an end goal of understanding, designing and screening materials for commercial applications.

References
[1] Ajay Muralidharan, LR Pratt, MI Chaudhari, and SB Rempe. “Quasi-Chemical Theory for Anion Hydration and Specific Ion Effects: Cl−(aq) vs. F−(aq)”. In: Chemical Physics Letters (2019).
[2] Ajay Muralidharan, Tyler Lytle, and Arun Yethiraj. ““Why Lithium Ions Stick to Some Anions and not Others””. In: The Journal of Physical Chemistry B (Accepted) (2021).
[3] Ajay Muralidharan, Tyler Lytle, and Arun Yethiraj. “Tuning Ion Correlations for the Rational Design of (Poly)electrolytes”. In: Bulletin of the American Physical Society (2021).
[4] Mangesh I Chaudhari, Juan M Vanegas, LR Pratt, Ajay Muralidharan, and Susan B Rempe.“Hydration mimicry by membrane ion channels”. In: Annual review of physical chemistry 71 (2020), pp. 461–484.
[5] Ulrich Arnold and Ronald T Raines. “Replacing a single atom accelerates the folding of a protein and increases its thermostability”. In: Organic & biomolecular chemistry 14.28 (2016), pp. 6780–6785.
[6] Ajay Muralidharan, J.R. Schmidt, and Arun Yethiraj. “Solvation Induced Ring Puckering Effect in Fluorinated Prolines and Its Inclusion in Classical Force Fields”. In: The Journal of Physical Chemistry B 124.28 (2020), pp. 5899–5906.
[7] Ajay Muralidharan, Xinyi Li, and J.R. Schmidt. “A hierarchical graph theory-based sampling approach to study the nucleation of weak electrolytes”. In: (In preparation) (2021).