Jonathan K. Whitmer

Synthetic polyelectrolyte complex coacervates are a ubiquitous class of liquid–liquid phase-separated (LLPS) systems that closely mimic the dynamic condensates often seen in cellular environments, where macromolecular crowding fundamentally shapes their physicochemical properties and biological function. For synthetic systems, neutral crowding agents, such as poly(ethylene glycol) (PEG), enhance coacervation, although the direct impact of osmotic stress on the structural and thermodynamic stability of the dense coacervate phase remains incompletely understood. Here, we utilize coarse-grained molecular simulations alongside systematic experimental studies to reveal the intrinsic compressibility and hydration dynamics of the coacervate phase under controlled osmotic perturbations. This framework establishes quantitative design principles for engineering synthetic condensates with programmable mechanical stability in macro-molecular crowded environments, with direct implications for developing biomimetic organelles that maintain structural integrity within the crowded cellular environment.

Glucose recognition plays a central role in the development of glucose-responsive materials such as sensing platforms, and in therapeutic applications, such as drug delivery systems. This motivates a deeper understanding of the structural and chemical features that most strongly influence binding affinity between glucose and synthetic host molecules. We present a machine learning framework that integrates molecular dynamics (MD) simulations, chemical and structural feature extraction, principal component analysis (PCA), and Gaussian process (GP) regression to investigate glucose binding to synthetic host molecules. Structural and chemical descriptors were systematically assembled to capture geometric, electronic, and topological characteristics of each host, and subsequently reduced in dimensionality using PCA. The resulting principal components were used to train GP models that learn the relationship between feature variations and MD-derived free energy profiles. Representative potentials of mean force (PMFs) illustrate the diversity of binding stability across the host series, with the full dataset reported in the Supporting Information. Although the developed framework does not propose new host features, the combined PCA–GP analysis identifies significant design parameters—such as cavity size, heteroatom distribution, and hydrogen-bonding capacitythat most strongly govern glucose affinity, providing mechanistic insight and guiding the design of next-generation glucose binders.

Complex coacervation, a liquid–liquid phase separation phenomenon driven by electrostatic interactions between oppositely charged polyelectrolytes (PEs), has attracted widespread attention because of its relevance in biological systems and potential applications in materials science. Although many theoretical models, experimental investigations, and computational studies have investigated the thermodynamics, phase coexistence behavior, and rheological properties in great detail, a molecular-level understanding of the internal structure of the complex coacervate phase is still lacking. In this study, we investigate the effects of the degree of polymerization of the polyelectrolytes (N) on the phase behavior and internal structure of the resulting coacervate phase using molecular dynamics simulations employing a simplistic bead–spring model of polyelectrolytes and explicit nonpolar solvents. Our simulations show an increase in coacervate phase stability with N, elevating the critical temperature in agreement with existing theoretical predictions and experimental observations. The polyelectrolytes inside the dense phase maintain a homogeneous overlapping distribution without collapsing into globules. The compactness of the dense phase increases with N in agreement with prior experimental observations, despite a concomitant increase in the polymer’s effective size as quantified by its radius of gyration (Rg). We discuss the implications of this model for a fundamental understanding of the coacervation process and as a first step toward the systematic examination of the mutual role of electrostatics and chemistry in the behavior of solvated polyions.

This work utilizes molecular simulations and machine learning to identify key interactions underlying favorable binding of molecules to glucose so they may be used in the development of novel materials and therapeutics.

This work explores liquid-liquid phase separation in biologically relevant soft matter systems, and how it is sensitively controlled by charge strength, enabling narrow pockets of stability which respond to the local environment.

This study presents a novel method for constructing free-energy profiles and steady-state distributions from either equilibrium or non-equilibrium trajectories along a defined reaction coordinate. The method works by tracking the final states of a swarm of short simulations launched under different initial conditions with no prior knowledge of the free energy landscape. Subsequently, this trajectory information is used to build a transition matrix whose primary eigenvector captures the steady-state occupation probability for each value of the reaction coordinate, yielding the free energy profile in equilibrium. This innovative method holds potential for many new materials and engineering applications where it is desired to know the free energy of rate-limiting configurations as may be relevant for transport processes (in, e.g., battery electrolytes and nano-filtration membranes), complexation (in, e.g., self-assembly and ligand-binding interactions), or in tuning properties such as adsorption. We illustrate the effectiveness of the method by capturing the free energy associated with a one-dimensional barrier potential modeling a separation membrane and the particle distribution associated with thermophoresis under a temperature gradient. Further extensions and applications of the method are also discussed.

Membranes that selectively enhance target solute permeation while rejecting competing species are essential for precision separations. This study introduces charge-patterned mosaic membranes (CMMs) that selectively transport divalent asymmetric salts by leveraging a net-neutral membrane–solution interface. This mechanism, dictated by the charge ratio of positive and negative domains on the membrane surface and the balance of cations and anions in the salt, is supported by analytical, numerical, and experimental results. Analytical solutions identified cationic domain coverages ( f ⁺ ) of 33%, 50%, and 66% as optimal for the selective transport of +2:−1 salts, +1:−1 salts, and +1:−2 salts, respectively, under conditions where the pattern size ( L ) is significantly larger than the Debye length. Numerical simulations and experiments using CMMs with alternating charged-stripes inkjet-printed onto nanostructure copolymer substrates confirmed these findings. By varying stripe widths to control f ⁺ , pressure-driven filtration experiments demonstrated selective enrichment of MgCl ₂ and K ₂ SO ₄ at the predicted f ⁺ values, with deviations from these values leading to salt rejection. These results highlight the pivotal role of a net-neutral interface in enabling asymmetric salt enrichment. This study positions CMMs as a versatile platform for tuning ion selectivity, addressing challenges in resource recovery, water treatment, and precision separations.

Machine learning force fields (MLFFs) promise to accurately describe the potential energy surface of molecules at the ab initio level of theory with improved computational efficiency. Within MLFFs, equivariant graph neural networks (EQNNs) have shown great promise in accuracy and performance and are the focus of this work. The capability of EQNNs to recover free energy surfaces (FES) remains to be thoroughly investigated. In this work, we investigate the impact of collective variables (CVs) distribution within the training data on the accuracy of EQNNs predicting the FES of butane and alanine dipeptide. A generalizable workflow is presented in which training configurations are generated with classical molecular dynamics simulations, and energies and forces are obtained with ab initio calculations. We evaluate how bond and angle constraints in the training data influence the accuracy of EQNN force fields in reproducing the FES of the molecules at both classical and ab initio levels of theory. Results indicate that the model’s accuracy is unaffected by the distribution of sampled CVs during training, given that the training data includes configurations from characteristic regions of the system’s FES. However, when the training data is obtained from classical simulations, the EQNN struggles to extrapolate the free energy for configurations with high free energy. In contrast, models trained with the same configurations on ab initio data show improved extrapolation accuracy. The findings underscore the difficulties in creating a comprehensive training dataset for EQNNs to predict FESs and highlight the importance of prior knowledge of the system’s FES.

Solid-state electrolytes are currently being explored as a safe material capable of addressing consumer energy-storage demands. Solid polymer electrolytes, in particular, offer a high energy density and improved safety when compared to liquid-based electrolytes, but tend to have a significantly lower ionic conductivity. We hypothesize structured ionic liquids can enhance conductivity compared to polymer electrolytes. Here, we explore the performance of these materials through coarse-grained molecular dynamics simulation. While we observe similar phase behavior (incorporating solid, smectic, and liquid phases) to that seen in experiments, we also observe significantly more mobility in the cationic species compared to the anionic species before the system reaches an arrest transition. We further discuss how the general results within this paper can guide further studies and target the design of new highly conductive solid electrolytes with the potential to enable the use of multivalent ionic species as ion conductors.

We explore an idealized theoretical model for ion transport within highly asymmetric ionic liquid mixtures. A primitive model-inspired system serves as a representative for asymmetric ionic materials (such as liquid crystalline salts) which quench to form disordered, partially arrested phases. Self-consistent generalized Langevin equation theory is applied to understand the connection between the size ratio of charge-matched salts and their average mobility. Within this model, we identify novel glassy states where one of the two charged species (without loss of generality, either the macro-cation or the micro-anion) is arrested, while the other retains liquid-like mobility. We discuss how this result is useful in the development of novel single-ion conducting phases in ionic liquid-based materials, for instance, conductors operating at low temperature or the technology associated with ionic liquid crystals.