Cheng-Han Li

Graph-based electronic structure theory offers a scalable approach to study large, complex atomistic systems using distributed and hybrid computational platforms. We demonstrate the coupling of graph-based linear scaling electronic structure theory, as implemented in the Scalable Ecosystem, Driver, and Analyzer for Complex Chemistry Simulations (SEDACS), with semiempirical quantum chemistry methods as implemented in the PySEQM code, with Graphics Processing Unit (GPU) acceleration. This powerful combination enables efficient, scalable electronic structure calculations over many nodes, significantly reducing computational cost while naturally harnessing parallelism. Detailed analyses of parallelization efficiency, computational accuracy, and communication overheads are provided, highlighting an order-of-magnitude speedup for systems of up to 10,000 atoms.

We present an extended Lagrangian shadow molecular dynamics scheme with an interatomic Born-Oppenheimer potential determined by the relaxed atomic charges of a second-order charge equilibration model. To parametrize the charge equilibration model, we use machine learning with neural networks to determine the environment-dependent electronegativities and chemical hardness parameters for each atom, in addition to the charge-independent energy and force terms. The approximate shadow molecular dynamics potential in combination with the extended Lagrangian formulation improves the numerical stability and reduces the number of Coulomb potential calculations required to evaluate accurate conservative forces. We demonstrate efficient and accurate simulations with excellent long-term stability of the molecular dynamics trajectories. The significance of choosing fixed or environment-dependent electronegativities and chemical hardness parameters is evaluated. Finally, we compute the infrared spectrum of molecules via the dipole autocorrelation function and compare to experiments to highlight the accuracy of the shadow molecular dynamics scheme with a machine learned flexible charge potential.

We present an extended Lagrangian shadow molecular dynamics scheme with an interatomic Born-Oppenheimer potential determined by the relaxed atomic charges of a second-order charge equilibration model. To parameterize the charge equilibration model, we use machine learning with neural networks to determine the environment-dependent electronegativities and chemical hardness parameters for each atom, in addition to the charge-independent energy and force terms. The approximate shadow molecular dynamics potential in combination with the extended Lagrangian formulation improves the numerical stability and reduces the number of Coulomb potential calculations required to evaluate accurate conservative forces. We demonstrate efficient and accurate simulations with excellent long-term stability of the molecular dynamics trajectories. The significance of the choosing fixed or environment-dependent electronegativities and chemical hardness parameters is evaluated. Finally, we compute the infrared spectrum of molecules via the dipole auto-correlation function and compare to experiments to highlight the accuracy of the shadow molecular dynamics scheme with a machine learned flexible charge potential.

Studies investigating the influence of the length of linkers between redox-active moieties and peptide-based polymer backbones were conducted to advance fundamental knowledge toward the design and development of sustainably-sourced, recyclable, and degradable materials for energy applications. In this work, precursor polypeptides were synthesized through the ring-opening polymerizations of N-carboxyanhydrides decorated with varying lengths of alkylchloride side chain groups, followed by post-polymerization installation of the viologen moieties. Electrochemical interrogation of the viologen-based polypeptides provided estimates of the electron transfer rate constants, both heterogeneous (k⁰) and electron self-exchange (k_ex), the apparent diffusion coefficient (D_ap), and their device-based energy storage performance. For the first redox couple (viologen dication state to viologen radical-cation state), it was found that the rate of electron transfer among the pendant groups in all viologen-based polypeptides, k_ex, was not significantly impacted by linker length. In contrast, for the second redox couple (viologen radical-cation state to the neutral viologen), k_ex varied with linker length and was fastest during reduction from the viologen radical-cation state to the neutral viologen. Most interestingly, a linear relationship was identified between log(k⁰) and log(k_ex) with a slope of 1.85, indicating that electron transport in the viologen-based polypeptides followed most closely to Marcus–Hush theory with diffusion limitations or Laviron–Andrieux–Savéant (LAS) theory. Finally, the polypeptides were studied in lithium metal half cells to determine the relationship between k_ex and energy storage performance. The viologen-based polypeptide with the moderate length linker exhibited the highest capacity and lowest degree of swelling, but only moderate k_ex, demonstrating that the device performance was primarily influenced electrode swelling. Taken together, the viologen-polypeptide backbone dictated the mechanism of electron transfer, whereas the linker length could be used to alter the rate of electron transfer (k_ex). Balancing the rate of electron transfer (k_ex) and degree of swelling will be a major challenge to identify polymers for high performance energy storage devices.

Redox-active polymers serving as the active materials in solid-state electrodes offer a promising path toward realizing all-organic batteries. While both cathodic and anodic redox-active polymers are needed, the diversity of the available anodic materials is limited. Here, we predict solid-state structural, ionic, and electronic properties of anodic, phthalimide-containing polymers using a multiscale approach that combines atomistic molecular dynamics, electronic structure calculations, and machine learning surrogate models. Importantly, by combining information from each of these scales, we are able to bridge the gap between bottom-up molecular characteristics and macroscopic properties such as apparent diffusion coefficients of electron transport (D _app). We investigate the impact of different polymer backbones and of two critical factors during battery operation: state of charge and polymer swelling. Our findings reveal that the state of charge significantly influences solid-state packing and the thermophysical properties of the polymers, which, in turn, affect ionic and electronic transport. A combination of molecular-level properties (such as the reorganization energy) and condensed-phase properties (such as effective electron hopping distances) determine the predicted ranking of electron transport capabilities of the polymers. We predict D _app for the phthalimide-based polymers and for a reference nitroxide radical-based polymer, finding a 3 orders of magnitude increase in D _app (≈10^-6 cm² s^-1) with respect to the reference. This study underscores the promise of phthalimide-containing polymers as highly capable redox-active polymers for anodic materials in all-organic batteries, due to their exceptional predicted electron transport capabilities.

Charge transfer in nonconjugated redox-active polymers is influenced by redox site proximity and polymer flexibility, but it is challenging to observe these effects independently. In this work, spatially defined radical-containing polymers are synthesized by using acyclic diene metathesis (ADMET) polymerization of α,ω-dienes bearing a central activated ester. Postpolymerization functionalization with 4-amino-2,2,6,6-tetramethylpiperidine-1-oxyl (4-amino-TEMPO) introduces TEMPO radical groups onto the polymer backbone through amide linkages to yield spatially defined polymers with radical units every 9, 11, 15, and 21 carbons. Increased radical spacing leads to reduced spin-spin coupling and increased chain flexibility. The glass transition temperatures (T_g) range from 47.6 to -13.8 °C, depending on the radical spacing. The spatially defined TEMPO-substituted polymer with a spacing length of 15 carbons displays the lowest T_g and the shortest hopping distance, as shown through molecular dynamics simulations. Also, this polymer displays kinetics 1000 times faster than the commonly studied TEMPO-containing polymer poly(2,2,6,6-tetramethylpiperidinyloxy-4-ylacrylamide) (PTAm). Remarkably, comparison of the diffusion and kinetics attributed to the redox reaction reveals that both the apparent diffusion coefficient and the self-exchange reaction rate constant are correlated to the polymer's T_g as log[D_app] and log[k_ex,app] ∼ T_g, respectively. Critically, these data demonstrate that controlling the spacing of redox-active groups along a polymer backbone strongly influences backbone flexibility and radical packing, which leads to synergetic improvements in the charge transfer kinetics of nonconjugated redox-active polymers.

Generative molecular design strategies have emerged as promising alternatives to trial-and-error approaches for exploring and optimizing within large chemical spaces. To date, generative models with reinforcement learning approaches have frequently used low-cost methods to evaluate the quality of the generated molecules, enabling many loops through the generative model. However, for functional molecular materials tasks, such low-cost methods are either not available or would require the generation of large amounts of training data to train surrogate machine learning models. In this work, we develop a framework that connects the REINVENT reinforcement learning framework with excited state quantum chemistry calculations to discover molecules with specified molecular excited state energy levels, specifically molecules with excited state landscapes that would serve as promising singlet fission or triplet–triplet annihilation materials. We employ a two-step curriculum strategy to first find a set of diverse promising molecules, then demonstrate the framework's ability to exploit a more focused chemical space with anthracene derivatives. Under this protocol, we show that the framework can find desired molecules and improve Pareto fronts for targeted properties versus synthesizability. Moreover, we are able to find several different design principles used by chemists for the design of singlet fission and triplet–triplet annihilation molecules.

A critical bottleneck for the design of high-conductivity organic materials is finding molecules with low reorganization energy. To enable high-throughput virtual screening campaigns for many types of organic electronic materials, a fast reorganization energy prediction method compared to density functional theory is needed. However, the development of low-cost machine-learning-based models for calculating the reorganization energy has proven to be challenging. In this paper, we combine a 3D graph-based neural network (GNN) recently benchmarked for drug design applications, ChIRo, with low-cost conformational features for reorganization energy predictions. By comparing the performance of ChIRo to another 3D GNN, SchNet, we find evidence that the bond-invariant property of ChIRo enables the model to learn from low-cost conformational features more efficiently. Through an ablation study with a 2D GNN, we find that using low-cost conformational features on top of 2D features informs the model for making more accurate predictions. Our results demonstrate the feasibility of reorganization energy predictions on the benchmark QM9 data set without needing DFT-optimized geometries and demonstrate the types of features needed for robust models that work on diverse chemical spaces. Furthermore, we show that ChIRo informed with low-cost conformational features achieves comparable performance with the previously reported structure-based model on π-conjugated hydrocarbon molecules. We expect this class of methods can be applied to the high-throughput screening of high-conductivity organic electronics candidates.

Metal-free aqueous batteries can potentially address the projected shortages of strategic metals and safety issues found in lithium-ion batteries. More specifically, redox-active non-conjugated radical polymers are promising candidates for metal-free aqueous batteries because of the polymers’ high discharge voltage and fast redox kinetics. However, little is known regarding the energy storage mechanism of these polymers in an aqueous environment. The reaction itself is complex and difficult to resolve because of the simultaneous transfer of electrons, ions and water molecules. Here we demonstrate the nature of the redox reaction for poly(2,2,6,6-tetramethylpiperidinyloxy-4-yl acrylamide) by examining aqueous electrolytes of varying chao-/kosmotropic character using electrochemical quartz crystal microbalance with dissipation monitoring at a range of timescales. Surprisingly, the capacity can vary by as much as 1,000% depending on the electrolyte, in which certain ions enable better kinetics, higher capacity and higher cycling stability.