Papers

Mamun, O., Yang, C., & Yue, S. (2025). Deep Graph Kernel Learning for Material & Atomic Level Uncertainty Quantification in Adsorption Energy Prediction.
```
@article{2025-mamun-dgkl,
  title = {Deep Graph Kernel Learning for Material & Atomic Level Uncertainty Quantification in Adsorption Energy Prediction},
  author = {Mamun, Osman and Yang, Chenlu and Yue, Shuwen},
  preprint = {10.26434/chemrxiv-2025-pfng2-v2},
  journal = {},
  volume = {},
  number = {},
  pages = {},
  year = {2025},
  date = {2025-03-21},
  publisher = {}
}
```
```
Graph Neural Networks (GNNs) provide an efficient surrogate for computationally intensive density functional theory calculations in catalytic material discovery, yet they often struggle with out-of-domain generalization. This limitation necessitates reliable prediction uncertainty quantification for informed catalyst discovery. While Gaussian Processes (GPs) offer principled Bayesian uncertainty quantification, their cubic time complexity, high memory requirements, and inability to learn from graph structures limit their application in high-throughput discovery. We introduce Deep Graph Kernel Learning (DGKL), a scalable framework that combines a GNN backbone with a sparse variational Gaussian Process (SVGP) for uncertainty quantification in adsorption energy prediction. We benchmark DGKL against state-of-the-art methods, including ensemble/query-by-committee, evidential, and Monte-Carlo dropout approaches. DGKL consistently outperforms existing methods across ranking-based metrics (negative log-likelihood, expected normalized calibration error, miscalibration area) and error-based metrics (RMSE vs. RMV and error vs. standard deviation plots) while maintaining computational efficiency. Specifically, DGKL achieves the lowest expected normalized calibration error (0.06-0.10), lowest miscalibration area (0.04-0.07), and highest Spearman correlation coefficient (0.34-0.51) across diverse datasets and GNN backbone combinations. Qualitatively, DGKL’s RMSE vs. RMV plots demonstrate superior calibration compared to competing methods. Additionally, we propose a DGKL variation capable of predicting atomic-level uncertainty - a feature absent in existing methods — offering finegrained insights into out-of-domain data. DGKL can be incorporated into active learning frameworks to efficiently explore catalytic material space, accelerating the discovery of novel catalysts.
```
Schwindt, N. S., Epsztein, R., Straub, A. P., Yue, S., & Shirts, M. R. (2025). Molecular Details and Free Energy Barriers of Ion De–Coordination at Elevated Salinity and Pressure and Their Consequences for Membrane Separation.
```
@article{2025-schwindt-membrane,
  title = {Molecular Details and Free Energy Barriers of Ion De--Coordination at Elevated Salinity and Pressure and Their Consequences for Membrane Separation},
  author = {Schwindt, Nathanael S and Epsztein, Razi and Straub, Anthony P and Yue, Shuwen and Shirts, Michael R},
  preprint = {10.48550/arXiv.2501.19344},
  journal = {},
  volume = {},
  number = {},
  pages = {},
  year = {2025},
  date = {2025-03-05},
  publisher = {}
}
```
```
Ion dehydration has been hypothesized to strongly influence separation performance in membrane systems and ion transport in nanoscale channels. However, the molecular details of ion dehydration in membranes are not well understood, in particular under the high pressures and concentrations required for brine treatment. In this study, we define de-coordination as the process by which an ion decreases its total coordination number, including both water molecules and counterions. We estimate the de-coordination free energies in bulk solution for a range of different ions at high pressure and salinity relevant to brine treatment using molecular simulation. We also propose alternatives to the coordination number as the size constraint for traversing nanoscale constrictions, such as the maximum cross-sectional area of the complexed ion. We show that high operating pressures do not significantly change cation hydration shell stability nor the shell size, while high ionic concentrations lower the free energy barrier to reduce the cation coordination number. We find that anion de-coordination free energies are largely unaffected by elevated salinity and pressure conditions. Finally, we discuss the implications on ion-ion selectivity in separations membranes (e.g. extracting lithium from salt-lake brines) due to the effects of elevated pressure and salinity on ion de-coordination.
```
Ball, A. K., Terrones, G. G., Yue, S., & Kulik, H. J. (2025). Data-Driven Discovery of Water-Stable Metal-Organic Frameworks With High Water Uptake Capacity.
```
@article{2025-ball-mof-water,
  title = {Data-Driven Discovery of Water-Stable Metal-Organic Frameworks With High Water Uptake Capacity},
  author = {Ball, Akash K and Terrones, Gianmarco G and Yue, Shuwen and Kulik, Heather J},
  preprint = {10.26434/chemrxiv-2025-mz223},
  journal = {},
  volume = {},
  number = {},
  pages = {},
  year = {2025},
  date = {2025-03-19},
  publisher = {}
}
```
```
Metal-organic frameworks (MOFs) are promising candidate materials for applications that would benefit from precise chemical patterning, such as desalination and atmospheric water harvesting, but many MOFs suffer from poor stability in water. In addition to water stability, high water uptake capacity in ambient conditions is expected to be necessary for water-related practical applications of MOFs, motivating large-scale search that can only be achieved computationally. Here, we take a combined machine learning and high-throughput screening approach to identifying water-stable MOFs with high water uptake capacities. Starting from a subset of previously curated MOFs with experimentally known exceptionally high stability in water, we explore the effect of linker functionalization with twelve diverse hydrophilic functional groups expected to further tune water uptake. For these 736 MOFs, we use grand canonical Monte Carlo (GCMC) simulations to compute their water uptake capacity. We observe strong positive correlations between MOF pore features (e.g., the largest cavity diameter and volumetric pore volume) and water uptake capacity, although we notice breakdowns of such correlations in MOFs with extremely hydrophobic linkers that repel water molecules despite having large pores. Finally, we develop machine learning models to screen new MOFs simultaneously for water stability and water uptake capacity. From a pool of hypothetical and experimental MOFs, we identify 74 promising materials within the domain of applicability of the machine learning models that are predicted to be both water-stable and have high water uptake.
```
Terrones, G. G., Huang, S.-P., Rivera, M. P., Yue, S., Hernandez, A., & Kulik, H. J. (2024). Metal–Organic Framework Stability in Water and Harsh Environments from Data-Driven Models Trained on the Diverse WS24 Data Set. Journal of the American Chemical Society, 146(29), 20333–20348.
```
@article{2024-terrones-water-stability,
  title = {Metal--Organic Framework Stability in Water and Harsh Environments from Data-Driven Models Trained on the Diverse WS24 Data Set},
  author = {Terrones, Gianmarco G and Huang, Shih-Peng and Rivera, Matthew P and Yue, Shuwen and Hernandez, Alondra and Kulik, Heather J},
  journal = {Journal of the American Chemical Society},
  year = {2024},
  date = {2024-07-10},
  volume = {146},
  number = {29},
  pages = {20333--20348},
  doi = {10.1021/jacs.4c05879},
  data = {https://zenodo.org/records/12110918},
  publisher = {ACS Publications}
}
```
```
Metal–organic frameworks (MOFs) are porous materials with applications in gas separations and catalysis, but a lack of water stability often limits their practical use given the ubiquity of water. Consequently, it is useful to predict whether a MOF is water-stable before investing time and resources into synthesis. Existing heuristics for designing water-stable MOFs lack generality and limit the diversity of explored chemistry due to narrowly defined criteria. Machine learning (ML) models offer the promise to improve the generality of predictions but require data. In an improvement on previous efforts, we enlarge the available training data for MOF water stability prediction by over 400%, adding 911 MOFs with water stability labels assigned through semiautomated manuscript analysis to curate the new data set WS24. The additional data are shown to improve ML model performance (test ROC-AUC > 0.8) over diverse chemistry for the prediction of both water stability and stability in harsher acidic conditions. We illustrate how the expanded data set and models can be used with a previously developed activation stability model in combination with genetic algorithms to quickly screen ∼10,000 MOFs from a space of hundreds of thousands for candidates with multivariate stability (upon activation, in water, and in acid). We uncover metal- and geometry-specific design rules for robust MOFs. The data set and ML models developed in this work, which we disseminate through an easy-to-use web interface, are expected to contribute toward the accelerated discovery of novel, water-stable MOFs for applications such as direct air gas capture and water treatment.
```
Burton, H. G. A., Dong, S. S., Ghosh, S., Gu, B., Jackson, N. E., Keefer, D., Lu, Y., Monroe, J. I., Peng, B., Pieri, E., Spackman, P. R., Vacher, M., Vuckovic, S., Williams–Young, D., Yang, Z., Yue, S., Zerze, G., & Zhu, T. (2024). JCTC Early Career Board Selects. Journal of Chemical Theory and Computation, 20(14), 5785–5787.
```
@article{2024-JCTC-ECB-Selects,
  title = {JCTC Early Career Board Selects},
  author = {Burton, Hugh GA and Dong, Sijia S and Ghosh, Soumen and Gu, Bing and Jackson, Nicholas E and Keefer, Daniel and Lu, Yangyi and Monroe, Jacob I and Peng, Bo and Pieri, Elisa and Spackman, Peter R and Vacher, Morgane and Vuckovic, Stefn and Williams--Young, David and Yang, Zhongyue and Yue, Shuwen and Zerze, Gul and Zhu, Tianyu},
  journal = {Journal of Chemical Theory and Computation},
  volume = {20},
  number = {14},
  pages = {5785--5787},
  year = {2024},
  date = {2024-07-23},
  doi = {10.1021/acs.jctc.4c00787},
  publisher = {ACS Publications}
}
```
```
Early career scientists continue to make pivotal contributions to theoretical and computational chemistry. The recent establishment of the Early Career Board (ECB) of the Journal of Chemical Theory and Computation (JCTC) reflects JCTC’s recognition of early career scientists and their vision. To broaden such recognition, the ECB, in collaboration with the editorial board of JCTC, has organized this collection, “JCTC Early Career Board Selects”, to highlight recent work by early career scientists who have recently achieved independence as demonstrated by being a corresponding author of their JCTC article. Most corresponding authors are within the first six years of their independent position.
```
Oh, C., Nandy, A., Yue, S., & Kulik, H. J. (2024). MOFs with the Stability for Practical Gas Adsorption Applications Require New Design Rules. ACS Applied Materials and Interfaces, 16(41), 55541–55554.
```
@article{2024-oh-mof-adsorption,
  title = {MOFs with the Stability for Practical Gas Adsorption Applications Require New Design Rules},
  author = {Oh, Changhwan and Nandy, Aditya and Yue, Shuwen and Kulik, Heather J},
  journal = {ACS Applied Materials and Interfaces},
  volume = {16},
  number = {41},
  pages = {55541--55554},
  year = {2024},
  date = {2024-10-04},
  doi = {10.1021/acsami.4c13250},
  publisher = {ACS Publications}
}
```
```
Metal–organic frameworks (MOFs) have been widely studied for their ability to capture and store greenhouse gases. However, most computational discovery efforts study hypothetical MOFs without consideration of their stability, limiting the practical application of novel materials. We overcome this limitation by screening hypothetical ultrastable MOFs that have predicted high thermal and activation stability, as judged by machine learning (ML) models trained on experimental measures of stability. We enhance this set by computing the bulk modulus as a measure of mechanical stability and filter 1102 mechanically robust hypothetical MOFs from a database of ultrastable MOFs (USMOF DB). Grand Canonical Monte Carlo simulations are then employed to predict the gas adsorption properties of these hypothetical MOFs, alongside a database of experimental MOFs. We identify privileged building blocks that lead MOFs in USMOF DB to show exceptional working capacities compared to the experimental MOFs. We interpret these differences by training ML models on CO2 and CH4 adsorption in these databases, showing how poor model transferability between data sets indicates that novel design rules can be derived from USMOF DB that would not have been gathered through assessment of structurally characterized MOFs. We identify geometric features and node chemistry that will enable the rational design of MOFs with enhanced gas adsorption properties in synthetically realizable MOFs.
```
Yue, S., Nandy, A., & Kulik, H. J. (2023). Discovering Molecular Coordination Environment Trends for Selective Ion Binding to Molecular Complexes Using Machine Learning. The Journal of Physical Chemistry B, 127(49), 10592–10600.

— JPC-B Machine Learning in Physical Chemistry Virtual Special Issue.
```
@article{2023-yue-ion-coordinating,
  title = {Discovering Molecular Coordination Environment Trends for Selective Ion Binding to Molecular Complexes Using Machine Learning},
  author = {Yue, Shuwen and Nandy, Aditya and Kulik, Heather J},
  journal = {The Journal of Physical Chemistry B},
  volume = {127},
  number = {49},
  pages = {10592–-10600},
  year = {2023},
  month = dec,
  date = {2023-12-01},
  publisher = {ACS Publications},
  doi = {10.1021/acs.jpcb.3c06416}
}
```
```
The design of ion-selective materials with improved separation efficacy and efficiency is paramount, as current technologies fail to meet real-world deployment challenges. Selectivity in these materials can be informed by local ion binding in confined membrane ion channels. In this study, we utilize a data-driven approach to investigate design features in small molecular complexes coordinating ions as simplified models of ion channels. We curate a data set of 563 alkali metal coordinating molecular complexes (i.e., with Li+, Na+, or K+) from the Cambridge Structural Database and calculate differential ion binding energies using density functional theory. Using this information, we probe when and why structures favor exchange with alternate ions. Our analysis reveals that energetic preferences are related to ion size but are largely due to chemical interactions rather than structural reorganization. We identify unique trends in the selectivity for Li+ over other alkali ions, including the presence of N coordination atoms, planar coordination geometry, and small coordinating ring sizes. We use machine learning models to identify the key contributions of both geometric and electronic features in predicting selective ion binding. These physical insights offer preliminary guidance into the design of optimal membranes for ion selectivity.
```
Zhang, C., Yue, S., Panagiotopoulos, A. Z., Klein, M. L., & Wu, X. (2023). Why Dissolving Salt in Water Decreases Its Dielectric Permittivity. Physical Review Letters, 131(7), 076801.

— Featured in Science Magazine News
```
@article{2023-zhang-salt-dielectric,
  title = {Why Dissolving Salt in Water Decreases Its Dielectric Permittivity},
  author = {Zhang, Chunyi and Yue, Shuwen and Panagiotopoulos, Athanassios Z and Klein, Michael L and Wu, Xifan},
  journal = {Physical Review Letters},
  volume = {131},
  number = {7},
  pages = {076801},
  year = {2023},
  month = aug,
  date = {2023-08-16},
  publisher = {APS},
  doi = {10.1103/PhysRevLett.131.076801}
}
```
```
The dielectric permittivity of salt water decreases on dissolving more salt. For nearly a century, this phenomenon has been explained by invoking saturation in the dielectric response of the solvent water molecules. Herein, we employ an advanced deep neural network (DNN), built using data from density functional theory, to study the dielectric permittivity of sodium chloride solutions. Notably, the decrease in the dielectric permittivity as a function of concentration, computed using the DNN approach, agrees well with experiments. Detailed analysis of the computations reveals that the dominant effect, caused by the intrusion of ionic hydration shells into the solvent hydrogen-bond network, is the disruption of dipolar correlations among water molecules. Accordingly, the observed decrease in the dielectric permittivity is mostly due to increasing suppression of the collective response of solvent waters.
```
Roh, H., Yue, S., Hu, H., Chen, K., Kulik, H. J., & Gumyuseng, A. (2023). Unraveling Polymer–Ion Interactions in Electrochromic Polymers for their Implementation in Organic Electrochemical Synaptic Devices. Advanced Functional Materials, 2304893.
```
@article{2023-roh-electrochromic-ion-binding,
  title = {Unraveling Polymer--Ion Interactions in Electrochromic Polymers for their Implementation in Organic Electrochemical Synaptic Devices},
  author = {Roh, Heejung and Yue, Shuwen and Hu, Hang and Chen, Ke and Kulik, Heather J. and Gumyuseng, Aristide},
  journal = {Advanced Functional Materials},
  pages = {2304893},
  year = {2023},
  date = {2023-07-12},
  month = jul,
  publisher = {Wiley Online Library},
  doi = {10.1002/adfm.202304893}
}
```
```
Owing to low-power, fast and highly adaptive operability, as well as scalability, electrochemical random-access memory (ECRAM) technology is one of the most promising approaches for neuromorphic computing based on artificial neural networks. Despite recent advances, practical implementation of ECRAMs remains challenging due to several limitations including high write noise, asymmetric weight updates, and insufficient dynamic ranges. Here, inspired by similarities in structural and functional requirements between electrochromic devices and ECRAMs, high-performance, single-transistor and neuromorphic devices based on electrochromic polymers (ECPs) are demonstrated. To effectively translate electrochromism into electrochemical ion memory in polymers, this study systematically investigates polymer–ion interactions, redox activity, mixed ionic–electronic conduction, and stability of ECPs both experimentally and computationally using select electrolytes. The best-performing ECP-electrolyte combination is then implemented into an ECRAM device to further explore synaptic plasticity behaviors. The resulting ECRAM exhibits high linearity and symmetric conductance modulation, high dynamic range (≈1 mS or ≈6x), and high training accuracy (>84% within five training cycles on a standard image recognition dataset), comparable to existing state-of-the-art ECRAMs. This study offers a promising approach to discover and design novel polymer materials for organic ECRAMs and demonstrates potential applications, taking advantage of mature knowledge basis on electrochromic materials and devices.
```
Mathur, R., Muniz, M. C., Yue, S., Car, R., & Panagiotopoulos, A. Z. (2023). First-principles-based Machine Learning Models for Phase Behavior and Transport Properties of CO2. The Journal of Physical Chemistry B, 127(20), 4562–4569.
```
@article{2023-methur-ml-potential-co2,
  title = {First-principles-based Machine Learning Models for Phase Behavior and Transport Properties of CO2},
  author = {Mathur, Reha and Muniz, Maria Carolina and Yue, Shuwen and Car, Roberto and Panagiotopoulos, Athanassios Z},
  journal = {The Journal of Physical Chemistry B},
  publisher = {ACS Publications},
  volume = {127},
  number = {20},
  pages = {4562--4569},
  year = {2023},
  month = may,
  date = {2023-05-17},
  doi = {10.1021/acs.jpcb.3c00610}
}
```
```
In this work, we construct distinct first-principles-based machine-learning models of CO2, reproducing the potential energy surface of the PBE-D3, BLYP-D3, SCAN, and SCAN-rvv10 approximations of density functional theory. We employ the Deep Potential methodology to develop the models and consequently achieve a significant computational efficiency over ab initio molecular dynamics (AIMD) that allows for larger system sizes and time scales to be explored. Although our models are trained only with liquid-phase configurations, they are able to simulate a stable interfacial system and predict vapor–liquid equilibrium properties, in good agreement with results from the literature. Because of the computational efficiency of the models, we are also able to obtain transport properties, such as viscosity and diffusion coefficients. We find that the SCAN-based model presents a temperature shift in the position of the critical point, while the SCAN-rvv10-based model shows improvement but still exhibits a temperature shift that remains approximately constant for all properties investigated in this work. We find that the BLYP-D3-based model generally performs better for the liquid phase and vapor–liquid equilibrium properties, but the PBE-D3-based model is better suited for predicting transport properties.
```
Nandy, A., Yue, S., Oh, C., Duan, C., Terrones, G. G., Chung, Y. G., & Kulik, H. J. (2023). A Database of Ultrastable MOFs Reassembled from Stable Fragments with Machine Learning Models. Matter, 6(5), 1585–1603.

— Featured in MIT News
```
@article{2023-nandy-mof-ultrastable,
  author = {Nandy, Aditya and Yue, Shuwen and Oh, Changhwan and Duan, Chenru and Terrones, Gianmarco G and Chung, Yongchul G and Kulik, Heather J},
  title = {A Database of Ultrastable MOFs Reassembled from Stable Fragments with Machine Learning Models},
  journal = {Matter},
  volume = {6},
  number = {5},
  pages = {1585--1603},
  year = {2023},
  month = may,
  date = {2023-05-03},
  publisher = {Elsevier},
  doi = {10.1016/j.matt.2023.03.009},
  data = {https://zenodo.org/records/7091192}
}
```
```
High-throughput screening of hypothetical metal-organic framework (MOF) databases can uncover new materials, but their stability in real-world applications is often unknown. We leverage community knowledge and machine learning (ML) models to identify MOFs that are thermally stable and stable upon activation. We separate these MOFs into their building blocks and recombine them to make a new hypothetical MOF database of over 50,000 structures with orders of magnitude more (1) connectivity nets and (2) inorganic building blocks than were present in prior databases. This database shows a 10-fold enrichment of ultrastable MOF structures that are stable upon activation and more than 1 standard deviation more thermally stable than the average experimentally characterized MOF. For nearly 10,000 ultrastable MOFs, we compute elastic moduli to confirm that these materials have good mechanical stability, and we report methane deliverable capacities. We identify privileged metal nodes in ultrastable MOFs that optimize gas storage and mechanical stability simultaneously.
```
Panagiotopoulos, A. Z., & Yue, S. (2023). Dynamics of aqueous electrolyte solutions - Challenges for simulations. The Journal of Physical Chemistry B, 127(2), 430–437.
```
@article{2023-panagiotopoulos-yue-perspective,
  author = {Panagiotopoulos, Athanassios Z and Yue, Shuwen},
  journal = {The Journal of Physical Chemistry B},
  publisher = {ACS Publications},
  title = {Dynamics of aqueous electrolyte solutions - Challenges for simulations},
  volume = {127},
  number = {2},
  pages = {430--437},
  year = {2023},
  month = jan,
  date = {2023-01-06},
  doi = {10.1021/acs.jpcb.2c07477}
}
```
```
This Perspective article focuses on recent simulation work on the dynamics of aqueous electrolytes. It is well-established that full-charge, nonpolarizable models for water and ions generally predict solution dynamics that are too slow in comparison to experiments. Models with reduced (scaled) charges do better for solution diffusivities and viscosities but encounter issues describing other dynamic phenomena such as nucleation rates of crystals from solution. Polarizable models show promise, especially when appropriately parametrized, but may still miss important physical effects such as charge transfer. First-principles calculations are starting to emerge for these properties that are in principle able to capture polarization, charge transfer, and chemical transformations in solution. While direct ab initio simulations are still too slow for simulations of large systems over long time scales, machine-learning models trained on appropriate first-principles data show significant promise for accurate and transferable modeling of electrolyte solution dynamics.
```
Yue, S., Oh, C., Nandy, A., Terrones, G. G., & Kulik, H. J. (2023). Effects of MOF linker rotation and functionalization on methane uptake and diffusion. Molecular Systems Design & Engineering, 8(4), 527–537.

— Selected as MSDE HOT Article
```
@article{2023-yue-mof-linker,
  title = {Effects of MOF linker rotation and functionalization on methane uptake and diffusion},
  author = {Yue, Shuwen and Oh, Changhwan and Nandy, Aditya and Terrones, Gianmarco G and Kulik, Heather J},
  journal = {Molecular Systems Design \& Engineering},
  year = {2023},
  date = {2023-01-02},
  month = jan,
  volume = {8},
  number = {4},
  pages = {527--537},
  publisher = {Royal Society of Chemistry},
  doi = {10.1039/D2ME00237J}
}
```
```
The flexible degrees of freedom in metal–organic frameworks (MOFs) can have significant effects on guest molecule behavior. However, in the majority of studies applying molecular simulations to MOFs, the framework is assumed to be rigid in order to minimize computational cost. Here we assess the significance of this assumption on a representative example of methane uptake and diffusion in UiO-66. We introduce an open-source code to modify MOFs through functionalization and linker rotation and we perform Grand Canonical Monte Carlo and molecular dynamics simulations of methane in each of the functionalized and linker-rotated derivatives of UiO-66. We find that linker rotation moderately influences methane uptake and significantly influences methane diffusion. Our assessment provides ranges of property values that serve as measures of uncertainty of these two properties associated with linker rotation. We further determine that void volume fraction and minimum pore size are the features that govern methane uptake and diffusion, respectively. These findings illustrate the impact of linker rotation on MOFs and provide design principles to guide future investigations.
```
Mondal, A., Kussainova, D., Yue, S., & Panagiotopoulos, A. Z. (2022). Modeling Chemical Reactions in Alkali Carbonate–Hydroxide Electrolytes with Deep Learning Potentials. Journal of Chemical Theory and Computation, 19(14), 4584–4595.

— JCTC Machine Learning for Molecular Simulation Special Issue
```
@article{2022-mondal-dp-molten-salts,
  author = {Mondal, Anirban and Kussainova, Dina and Yue, Shuwen and Panagiotopoulos, Athanassios Z},
  journal = {Journal of Chemical Theory and Computation},
  publisher = {ACS Publications},
  title = {Modeling Chemical Reactions in Alkali Carbonate--Hydroxide Electrolytes with Deep Learning Potentials},
  volume = {19},
  number = {14},
  pages = {4584–4595},
  year = {2022},
  month = oct,
  date = {2022-10-14},
  doi = {10.1021/acs.jctc.2c00816}
}
```
```
We developed a deep potential machine learning model for simulations of chemical reactions in molten alkali carbonate-hydroxide electrolyte containing dissolved CO2, using an active learning procedure. We tested the deep neural network (DNN) potential and training procedure against reaction kinetics, chemical composition, and diffusion coefficients obtained from density functional theory (DFT) molecular dynamics calculations. The DNN potential was found to match DFT results for the structural, transport, and short-time chemical reactions in the melt. Using the DNN potential, we extended the time scales of observation to 2 ns in systems containing thousands of atoms, while preserving quantum chemical accuracy. This allowed us to reach chemical equilibrium with respect to several chemical species in the melt. The approach can be generalized for a broad spectrum of chemically reactive systems.
```
Yue, S., Riera, M., Ghosh, R., Panagiotopoulos, A. Z., & Paesani, F. (2022). Transferability of data-driven, many-body models for CO2 simulations in the vapor and liquid phases. The Journal of Chemical Physics, 156(10), 104503.
```
@article{2022-yue-mb-co2,
  author = {Yue, Shuwen and Riera, Marc and Ghosh, Raja and Panagiotopoulos, Athanassios Z and Paesani, Francesco},
  journal = {The Journal of Chemical Physics},
  number = {10},
  pages = {104503},
  publisher = {AIP Publishing LLC},
  title = {Transferability of data-driven, many-body models for CO2 simulations in the vapor and liquid phases},
  volume = {156},
  year = {2022},
  month = mar,
  date = {2022-03-09},
  doi = {10.1063/5.0080061},
  data = {https://dataspace.princeton.edu/handle/88435/dsp019g54xm81w}
}
```
```
Extending on the previous work by Riera et al. [J. Chem. Theory Comput. 16, 2246–2257 (2020)], we introduce a second generation family of data-driven many-body MB-nrg models for CO2 and systematically assess how the strength and anisotropy of the CO2–CO2 interactions affect the models’ ability to predict vapor, liquid, and vapor–liquid equilibrium properties. Building upon the many-body expansion formalism, we construct a series of MB-nrg models by fitting one-body and two-body reference energies calculated at the coupled cluster level of theory for large monomer and dimer training sets. Advancing from the first generation models, we employ the charge model 5 scheme to determine the atomic charges and systematically scale the two-body energies to obtain more accurate descriptions of vapor, liquid, and vapor–liquid equilibrium properties. Challenges in model construction arise due to the anisotropic nature and small magnitude of the interaction energies in CO2, calling for the necessity of highly accurate descriptions of the multidimensional energy landscape of liquid CO2. These findings emphasize the key role played by the training set quality in the development of transferable, data-driven models, which, accurately representing high-dimensional many-body effects, can enable predictive computer simulations of molecular fluids across the entire phase diagram.
```
Zhang, C., Yue, S., Panagiotopoulos, A. Z., Klein, M. L., & Wu, X. (2022). Dissolving salt is not equivalent to applying a pressure on water. Nature Communications, 13(1), 1–6.

— Featured in Springer Nature Research Communities
— Computation and Machine Learning for Chemistry Collection
```
@article{2022-zhang-salt-water,
  author = {Zhang, Chunyi and Yue, Shuwen and Panagiotopoulos, Athanassios Z and Klein, Michael L and Wu, Xifan},
  journal = {Nature communications},
  number = {1},
  pages = {1--6},
  publisher = {Nature Publishing Group},
  title = {Dissolving salt is not equivalent to applying a pressure on water},
  volume = {13},
  year = {2022},
  month = feb,
  date = {2022-02-10},
  doi = {10.1038/s41467-022-28538-8},
  data = {https://figshare.com/articles/dataset/Data_from_Dissolving_salt_is_not_equivalent_to_applying_a_pressure_on_water_/17193023/1}
}
```
```
Salt water is ubiquitous, playing crucial roles in geological and physiological processes. Despite centuries of investigations, whether or not water’s structure is drastically changed by dissolved ions is still debated. Based on density functional theory, we employ machine learning based molecular dynamics to model sodium chloride, potassium chloride, and sodium bromide solutions at different concentrations. The resulting reciprocal-space structure factors agree quantitatively with neutron diffraction data. Here we provide clear evidence that the ions in salt water do not distort the structure of water in the same way as neat water responds to elevated pressure. Rather, the computed structural changes are restricted to the ionic first solvation shells intruding into the hydrogen bond network, beyond which the oxygen radial-distribution function does not undergo major change relative to neat water. Our findings suggest that the widely cited pressure-like effect on the solvent in Hofmeister series ionic solutions should be carefully revisited.
```
Muniz, M. C., Gartner III, T. E., Riera, M., Knight, C., Yue, S., Paesani, F., & Panagiotopoulos, A. Z. (2021). Vapor–liquid equilibrium of water with the MB-pol many-body potential. The Journal of Chemical Physics, 154(21), 211103.

— Featured in JCP Scilight
— Selected as JCP Featured Article
```
@article{2021-muniz-mbpol-vle,
  author = {Muniz, Maria Carolina and Gartner III, Thomas E and Riera, Marc and Knight, Christopher and Yue, Shuwen and Paesani, Francesco and Panagiotopoulos, Athanassios Z},
  journal = {The Journal of Chemical Physics},
  number = {21},
  pages = {211103},
  publisher = {AIP Publishing LLC},
  title = {Vapor--liquid equilibrium of water with the MB-pol many-body potential},
  volume = {154},
  year = {2021},
  month = jun,
  date = {2021-06-02},
  doi = {10.1063/5.0050068},
  data = {https://dataspace.princeton.edu/handle/88435/dsp01br86b664p}
}
```
```
Among the many existing molecular models of water, the MB-pol many-body potential has emerged as a remarkably accurate model, capable of reproducing thermodynamic, structural, and dynamic properties across water’s solid, liquid, and vapor phases. In this work, we assessed the performance of MB-pol with respect to an important set of properties related to vapor–liquid coexistence and interfacial behavior. Through direct coexistence classical molecular dynamics simulations at temperatures of 400 K < T < 600 K, we calculated properties such as equilibrium coexistence densities, vapor–liquid interfacial tension, vapor pressure, and enthalpy of vaporization and compared the MB-pol results to experimental data. We also compared rigid vs fully flexible variants of the MB-pol model and evaluated system size effects for the properties studied. We found that the MB-pol model predictions are in good agreement with experimental data, even for temperatures approaching the vapor–liquid critical point; this agreement was largely insensitive to system sizes or the rigid vs flexible treatment of the intramolecular degrees of freedom. These results attest to the chemical accuracy of MB-pol and its high degree of transferability, thus enabling MB-pol’s application across a large swath of water’s phase diagram.
```
Yue, S., Muniz, M. C., Calegari Andrade, M. F., Zhang, L., Car, R., & Panagiotopoulos, A. Z. (2021). When do short-range atomistic machine-learning models fall short? The Journal of Chemical Physics, 154(3), 034111.

— Selected as JCP Featured Article
```
@article{2021-yue-mlp-long-range,
  author = {Yue, Shuwen and Muniz, Maria Carolina and Calegari Andrade, Marcos F and Zhang, Linfeng and Car, Roberto and Panagiotopoulos, Athanassios Z},
  journal = {The Journal of Chemical Physics},
  number = {3},
  pages = {034111},
  publisher = {AIP Publishing LLC},
  title = {When do short-range atomistic machine-learning models fall short?},
  volume = {154},
  year = {2021},
  month = jan,
  date = {2021-01-21},
  doi = {10.1063/5.0031215}
}
```
```
We explore the role of long-range interactions in atomistic machine-learning models by analyzing the effects on fitting accuracy, isolated cluster properties, and bulk thermodynamic properties. Such models have become increasingly popular in molecular simulations given their ability to learn highly complex and multi-dimensional interactions within a local environment; however, many of them fundamentally lack a description of explicit long-range interactions. In order to provide a well-defined benchmark system with precisely known pairwise interactions, we chose as the reference model a flexible version of the Extended Simple Point Charge (SPC/E) water model. Our analysis shows that while local representations are sufficient for predictions of the condensed liquid phase, the short-range nature of machine-learning models falls short in representing cluster and vapor phase properties. These findings provide an improved understanding of the role of long-range interactions in machine learning models and the regimes where they are necessary.
```
Kussainova, D., Mondal, A., Young, J. M., Yue, S., & Panagiotopoulos, A. Z. (2020). Molecular simulation of liquid–vapor coexistence for NaCl: Full-charge vs scaled-charge interaction models. The Journal of Chemical Physics, 153(2), 024501.
```
@article{2020-kussainova-molten-salt-vle,
  author = {Kussainova, Dina and Mondal, Anirban and Young, Jeffrey M and Yue, Shuwen and Panagiotopoulos, Athanassios Z},
  journal = {The Journal of Chemical Physics},
  number = {2},
  pages = {024501},
  publisher = {AIP Publishing LLC},
  title = {Molecular simulation of liquid--vapor coexistence for NaCl: Full-charge vs scaled-charge interaction models},
  volume = {153},
  year = {2020},
  month = jul,
  date = {2020-07-08},
  doi = {10.1063/5.0012065}
}
```
```
Scaled-charge models have been recently introduced for molecular simulations of electrolyte solutions and molten salts to attempt to implicitly represent polarizability. Although these models have been found to accurately predict electrolyte solution dynamic properties, they have not been tested for coexistence properties, such as the vapor pressure of the melt. In this work, we evaluate the vapor pressure of a scaled-charge sodium chloride (NaCl) force field and compare the results against experiments and a non-polarizable full-charge force field. The scaled-charge force field predicts a higher vapor pressure than found in experiments, due to its overprediction of the liquid-phase chemical potential. Reanalyzing the trajectories generated from the scaled-charge model with full charges improves the estimation of the liquid-phase chemical potential but not the vapor pressure.
```
Yue, S., & Panagiotopoulos, A. Z. (2019). Dynamic properties of aqueous electrolyte solutions from non-polarisable, polarisable, and scaled-charge models. Molecular Physics, 117(23-24), 3538–3549.
```
@article{2019-yue-electrolyte-dynamics,
  author = {Yue, Shuwen and Panagiotopoulos, Athanassios Z},
  journal = {Molecular Physics},
  number = {23-24},
  pages = {3538--3549},
  publisher = {Taylor \& Francis},
  title = {Dynamic properties of aqueous electrolyte solutions from non-polarisable, polarisable, and scaled-charge models},
  volume = {117},
  year = {2019},
  month = aug,
  date = {2019-08-14},
  doi = {10.1080/00268976.2019.1645901}
}
```
```
We investigated the dynamic properties of alkali halide solutions (NaCl, NaF, NaBr, NaI, LiCl, and KCl) using molecular dynamics simulations and several non-polarisable, polarisable, and scaled-charge models. The concentration dependence of shear viscosity was obtained with low statistical uncertainties to allow for calculation of the viscosity Jones-Dole B-coefficients. No prior values are available for the B-coefficients from molecular simulations of fully atomistic models for electrolyte solutions. In addition, we obtained diffusion coefficients with rigorous finite-size corrections to access ion mobilities; these provide insights on single ion hydration behaviour. We find that all models studied, even polarisable and scaled-charge models, quantitatively over-predict water structuring but qualitatively follow the experimentally determined Hofmeister series. All ion models considered are kosmotropes based on their calculated B-coefficient and diffusion coefficients, even for ions experimentally found to be chaotropes. These observations indicate that the water-ion interactions in these models are not adequately represented; additional interactions such as charge transfer must be incorporated in future models in order to better represent electrolyte solution properties.
```
Whitley, J. W., Jeffrey Horne, W., Shannon, M. S., Andrews, M. A., Terrell, K. L., Hayward, S. S., Yue, S., Mittenthal, M. S., O’Harra, K. E., & Bara, J. E. (2018). Systematic Investigation of the Photopolymerization of Imidazolium-Based Ionic Liquid Styrene and Vinyl Monomers. Journal of Polymer Science Part A: Polymer Chemistry, 56(20), 2364–2375.
```
@article{2018-whitley-il-polymerization,
  author = {Whitley, John W and Jeffrey Horne, William and Shannon, Matthew S and Andrews, Mary A and Terrell, Kelsey L and Hayward, Spenser S and Yue, Shuwen and Mittenthal, Max S and O'Harra, Kathryn E and Bara, Jason E},
  journal = {Journal of Polymer Science Part A: Polymer Chemistry},
  number = {20},
  pages = {2364--2375},
  publisher = {Wiley Online Library},
  title = {Systematic Investigation of the Photopolymerization of Imidazolium-Based Ionic Liquid Styrene and Vinyl Monomers},
  volume = {56},
  year = {2018},
  month = sep,
  date = {2018-09-09},
  doi = {10.1002/pola.29211}
}
```
```
The use of ionic liquids (ILs) as media in radical polymerizations has demonstrated the ability of these unique solvents to improve both reaction kinetics and polymer product properties. However, the bulk of these studies have examined the polymerization behavior of common organic monomers (e.g., methyl methacrylate, styrene) dissolved in conventional ILs. There is increasing interest in polymerized ILs (poly(ILs)), which are ionomers produced from the direct polymerization of styrene-, vinyl-, and acrylate-functionalized ILs. Here, the photopolymerization kinetics of IL monomers are investigated for systems in which styrene or vinyl functionalities are pendant from the imidazolium cation. Styrene-functionalized IL monomers typically polymerized rapidly (full conversion ≤1 min) in both neat compositions or when diluted with a nonpolymerizable IL, [C2mim][Tf2N]. However, monomer conversion in vinyl-functionalized IL monomers is much more dependent on the nature of the nonpolymerizable group. ATR-FTIR analysis and molecular simulations of these monomers and monomer mixtures identified the presence of multiple intermolecular interactions (e.g., π–π stacking, IL aggregation) that contribute to the polymerization behaviors of these systems.
```
Yue, S., Roveda, J. D., Mittenthal, M. S., Shannon, M. S., & Bara, J. E. (2018). Experimental densities and calculated fractional free volumes of ionic liquids with tri-and tetra-substituted imidazolium cations. Journal of Chemical & Engineering Data, 63(7), 2522–2532.
```
@article{2018-yue-il-ffv,
  author = {Yue, Shuwen and Roveda, John D and Mittenthal, Max S and Shannon, Matthew S and Bara, Jason E},
  journal = {Journal of Chemical & Engineering Data},
  number = {7},
  pages = {2522--2532},
  publisher = {ACS Publications},
  title = {Experimental densities and calculated fractional free volumes of ionic liquids with tri-and tetra-substituted imidazolium cations},
  volume = {63},
  year = {2018},
  month = mar,
  date = {2018-03-07},
  doi = {10.1021/acs.jced.7b01033}
}
```
```
Although it has been estimated that there are at least 1 million ionic liquids (ILs) that are accessible using commercially available starting materials, a great portion of the ILs that have been experimentally synthesized, characterized, and studied in a variety of applications are built around the relatively simple 1-n-alkyl-3-methylimidazolium ([Cnmim]) cation motif. Yet, there is no fundamental limitation or reason as to why tri- or tetra-functionalized imidazolium cations have received far less attention. Scant physical property data exist for just a few trifunctionalized imidazolium-based ILs and there is virtually no data on tetra-functionalized ILs. Thus, there are a broad experimental spaces on the “map” of ILs that are largely unexplored. We have sought to make an initial expedition into these “uncharted waters” and have synthesized imidazolium-based ILs with one more functional group(s) at the C(2), C(4), and/or C(5) positions of the imidazolium ring (as well as N(1) and N(3)). This manuscript reports the synthesis and experimental densities of these tri- and tetra-functionalized ILs as well as calculated densities and fractional free volumes from COSMOTherm. To the best of our knowledge, this is the first report of any detailed experimental measurements or computational studies relating to ILs with substitutions at the C(4) and C(5) positions.
```
Fang, Z., Both, J., Li, S., Yue, S., Apra, E., Keceli, M., Wagner, A. F., & Dixon, D. A. (2016). Benchmark calculations of Energetic properties of Groups 4 and 6 transition metal oxide nanoclusters Including comparison to density functional theory. Journal of Chemical Theory and Computation, 12(8), 3689–3710.
```
@article{2016-fang-benchmark-tm-oxide,
  author = {Fang, Zongtang and Both, Johan and Li, Shenggang and Yue, Shuwen and Apra, Edoardo and Keceli, Murat and Wagner, Albert F and Dixon, David A},
  journal = {Journal of chemical theory and computation},
  number = {8},
  pages = {3689--3710},
  publisher = {ACS Publications},
  title = {Benchmark calculations of Energetic properties of Groups 4 and 6 transition metal oxide nanoclusters Including comparison to density functional theory},
  volume = {12},
  year = {2016},
  month = jul,
  date = {2016-07-06},
  doi = {10.1021/acs.jctc.6b00464}
}
```
```
The heats of formation and the normalized clustering energies (NCEs) for the group 4 and group 6 transition metal oxide (TMO) trimers and tetramers have been calculated by the Feller–Peterson–Dixon (FPD) method. The heats of formation predicted by the FPD method do not differ much from those previously derived from the NCEs at the CCSD(T)/aT level except for the CrO3 nanoclusters. New and improved heats of formation for Cr3O9 and Cr4O12 were obtained using PW91 orbitals instead of Hartree-Fock (HF) orbitals. Diffuse functions are necessary to predict accurate heats of formation. The fluoride affinities (FAs) are calculated with the CCSD(T) method. The relative energies (REs) of different isomers, NCEs, electron affinities (EAs), and FAs of (MO2)n (M = Ti, Zr, Hf, n = 1–4) and (MO3)n (M = Cr, Mo, W, n = 1–3) clusters have been benchmarked with 55 exchange-correlation density functional theory (DFT) functionals including both pure and hybrid types. The absolute errors of the DFT results are mostly less than ±10 kcal/mol for the NCEs and the EAs and less than ±15 kcal/mol for the FAs. Hybrid functionals usually perform better than the pure functionals for the REs and NCEs. The performance of the two types of functionals in predicting EAs and FAs is comparable. The B1B95 and PBE1PBE functionals provide reliable energetic properties for most isomers. Long range corrected pure functionals usually give poor FAs. The standard deviation of the absolute error is always close to the mean errors, and the probability distributions of the DFT errors are often not Gaussian (normal). The breadth of the distribution of errors and the maximum probability are dependent on the energy property and the isomer.
```