Pith. sign in

REVIEW 3 major objections 4 minor 62 references

Differentiable OPLS Force Field Parameterization for Ionic Electrolytes and High-Throughput Application to Lithium-ion Batteries

T0 review · 3 major / 4 minor · reviewed 2026-07-11 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A density-tuned OPLS force field and a 10,000-formulation MD dataset open data-driven design of lithium-ion battery electrolytes.

desk verdict Solid methods-and-data paper: usable OPLS reparameterization plus a large multi-property MD library, with transferability only partial and reproducibility still open. read the letter →

arxiv 2607.04633 v2 pith:CD7QV4IV submitted 2026-07-06 physics.chem-ph physics.comp-ph

Haichao Huang , Zilin Chen , Qi Liu , Tianqi Zhao , Yunpei Liu , Guotao Qiu , Jianhui Chen , Zhen Li
show 9 more authors
Wenshuo Liang Minsung Cho Manxue Zhang Feiyu Kang Xiaolong Zou Yidan Cao Xushan Zhao Ziqi Cheng Ye Mei
This is my paper · ORCID
classification physics.chem-phphysics.comp-ph
keywords OPLS-AAforcefielddifferentiablemolecularionicelectrolyteslithium-ionbatterieshigh-throughputdynamicsconductivityLennard-Jonesparameterizationelectrolytepropertydataset
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Empirical trial-and-error cannot cover the huge solvent–salt space of lithium-ion battery electrolytes, and generic classical force fields often mispredict transport in concentrated multicomponent mixtures. This paper builds an automated, differentiable workflow that reclassifies atoms by local topology, shrinks the Lennard-Jones parameter set, and optimizes those parameters against experimental density while checking ionic conductivity independently. With a standardized large-system protocol, the resulting force field is used to simulate more than ten thousand formulations spanning 67 solvents and 15 lithium salts. The outcome is a five-property dataset (density, dielectric constant, viscosity, Li+ diffusion, and conductivity) whose internal structure is physically self-consistent. A sympathetic reader cares because the combination of a transferable force field and a broad, reproducible property map supplies the data foundation that inverse design and machine-learning screening of electrolytes have been missing.

What carries the argument

Topology-guided atom typification plus DMFF density optimization: atoms are reclassified by neighbor count, element, and hybridization so that only 136 Lennard-Jones parameters need tuning; gradients of a density-based loss are obtained by trajectory reweighting, yielding a single transferable parameter set validated on Onsager conductivity.

What would settle it

Measure density and Onsager ionic conductivity for a held-out set of multicomponent electrolytes that contain the poorly performing anions (BF4−, ClO4−) or new functional salts outside the training frequencies; systematic RMSE well above the reported ~3 mS/cm conductivity error, or clear density–transport decoupling, would falsify transferability of the density-tuned force field.

Watch

Extended reading notes

Core claim

An automated differentiable OPLS-AA parameterization—topology-guided atom typification down to 68 types (136 Lennard-Jones parameters), density-targeted optimization via DMFF, 0.7 charge scaling on salts, and conductivity as an independent check—produces a transferable electrolyte force field that, together with a ~100,000-atom / 35–40 ns protocol, enables reliable high-throughput MD of over 10,000 formulations and a self-consistent five-property dataset for data-driven electrolyte design.

Load-bearing premise

Fitting a fixed-charge OPLS model primarily to experimental density, with one global salt charge-scaling factor of 0.7, is assumed sufficient for accurate transport properties across diverse concentrated electrolytes—even though optimized parameters for some anions had to be abandoned because they produced severe aggregation.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The manuscript presents an automated workflow for refining OPLS-AA force fields for multicomponent ionic electrolytes used in lithium-ion batteries. Topology-guided reclassification reduces atom types to 68 (136 LJ parameters), which are optimized via DMFF against experimental densities under a global salt charge-scaling factor of 0.7; ionic conductivity serves as an independent check. After establishing a ~10^5-atom / 35–40 ns NVT protocol via finite-size and time-convergence tests, the authors run high-throughput MD on the Tianqiong platform for >10 000 formulations spanning 67 solvents and 15 lithium salts, producing a five-property dataset (density, dielectric constant, viscosity, Li+ diffusivity, Onsager conductivity). t-SNE embeddings are used to illustrate salt clustering, concentration/temperature gradients, and internal physical consistency.

Significance. If the force field and dataset prove sufficiently reliable, the work supplies both a practical, transferable classical model and a large, multi-property computational resource that can accelerate data-driven electrolyte screening beyond the sparse experimental conductivity tables currently available. Strengths that support this potential include the dual-property (density-fit / conductivity-validate) strategy, systematic charge-scaling and Yeh–Hummer-style size-dependence analyses, explicit platform cross-checks against GPU results, and transparent reporting of poorer performance for BF4−/ClO4− systems. The scale of the high-throughput campaign and the closed-loop coupling of parameterization to dataset construction are genuine contributions to the field.

major comments (3)
  1. [Results: Predictive Performance of Ionic Conductivity] Results, “Predictive Performance of Ionic Conductivity” and Fig. 5: the optimized LJ parameters for BF4− and ClO4− produced severe aggregation and numerical instability in large-scale runs, forcing reversion to original OPLS-AA parameters and yielding R as low as 0.548 (Onsager) / 0.354 (NE). Because these anions remain inside the claimed 15-salt scope, the assertion of a “general” / “transferable” force field for ionic electrolytes is only partially secured and must be qualified more carefully in the abstract, introduction and conclusion.
  2. [Models: Force Field Model / Optimization] Models and Computational Methods, “Force Field Model” / “Optimization” and Fig. 2: density is only weakly sensitive to the charge-scaling factor (R ≈ 0.89–0.90), while conductivity is highly sensitive. The dual-property strategy therefore rests almost entirely on a single hand-chosen global scale of 0.7 plus post-hoc conductivity checks rather than joint optimization of structure and transport. Given that density-matched LJ parameters failed for two anions, the claim that density is a sufficient primary target for transferable transport accuracy across diverse multicomponent electrolytes requires stronger justification or additional validation metrics (e.g., radial distribution functions, ion-pairing free energies).
  3. [Results / Fig. 5 and Fig. 6 caption] Fig. 5a and overall statistics: Onsager conductivity achieves only moderate agreement (R = 0.720, RMSE = 3.94 mS cm−1) across the single-salt benchmark; even the better-performing LiPF6 subset remains at RMSE ≈ 2.8–3.2 mS cm−1. Combined with frequent non-convergence of Green–Kubo viscosity at low T / high concentration (Fig. 6 caption), the language of an “accurate” force field that “ensures reliable quantification of the transport properties” overstates the quantitative fidelity for the full chemical space.
minor comments (4)
  1. [Abstract / Conclusion] Abstract and Conclusion repeatedly call the force field “accurate and transferable” while the body text already documents clear limitations for certain anions; the wording should be aligned with the actual performance metrics.
  2. [Fig. 6 caption] Fig. 6 caption notes that non-convergent viscosity trajectories were excluded, reducing the single-salt set from 8077 to 7868; the dual-salt set is similarly pruned. The fraction and chemical identity of excluded points should be quantified more explicitly so readers can judge coverage bias.
  3. [Models: Electrolyte Screening] Supporting Information screening criteria (e.g., exclusion of most boron- and phosphorus-containing salts) are reasonable but should be cross-referenced in the main-text “Electrolyte Screening” subsection for completeness.
  4. [Throughout] Occasional typographical inconsistencies appear (e.g., “seleted GPU nodes”, “duale-salt”, missing spaces around units). A careful proof-reading pass is needed.

Circularity Check

1 steps flagged · score 3.0 of 10

Density accuracy gains are partly by construction from the LJ fit; conductivity is held out as independent validation, so circularity remains moderate.

  1. fitted input called prediction [Abstract; Models and Computational Methods (Optimization); Result and Discussion (Performance of Fine-tuned Force Field, Fig. 2c)]
    "optimizes Lennard-Jones parameters via the DMFF framework, with experimental density as the fitting target and ionic conductivity as an independent validation metric. ... For density prediction (Figure 2c), the fine-tuning process significantly improves the accuracy: the Pearson correlation coefficient increases from 0.917 to 0.959, and the RMSE decreases from 0.051 to 0.035 g/cm^{3}."

    The 136 LJ parameters are updated by gradient descent on a loss that is the (weighted) mean-squared error between simulated and experimental densities. The subsequent claim of 'significantly improved' density accuracy is therefore the direct numerical consequence of that minimization rather than an independent prediction of a held-out structural observable.

full rationale

The paper is transparent that Lennard-Jones parameters are optimized against experimental density via DMFF (with a hand-chosen global charge scale of 0.7 selected by inspecting initial-FF conductivity). Consequently the reported post-optimization density metrics (R rising from 0.917 to 0.959, RMSE falling from 0.051 to 0.035 g cm^{-3}) are improved by construction of the loss and should not be read as pure out-of-sample prediction. Ionic conductivity is explicitly reserved as an independent validation metric and is never part of the training loss; the Onsager R = 0.720 / RMSE = 3.94 mS cm^{-1} (and the poorer BF4-/ClO4- subset) therefore constitute genuine external checks. Topology-guided typification, the standardized 10^5-atom / 35-40 ns protocol, the high-throughput dataset, and the t-SNE physical-consistency observations do not reduce to the fitted inputs. No load-bearing self-citation uniqueness theorems or ansatz smuggling appear. The only circular step is the density fit itself, yielding a modest overall score of 3.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The central claim rests on classical fixed-charge MD being adequate after LJ reparameterization and charge scaling; on density as the primary experimental anchor; on a large hand-chosen parameter set (136 LJ values plus scaling); and on computational protocols (system size, time, Onsager estimators) rather than on new physical entities. No new particles or forces are postulated—only a retyped, refitted OPLS model and a simulation library.

free parameters (5)
  • 136 Lennard-Jones parameters (σ, ε for 68 atom types)
    Primary optimized degrees of freedom; updated by gradient descent on density MSE via DMFF trajectory reweighting across formulation batches.
  • Salt charge scaling factor
    Chosen as 0.7 after comparing 1.0, 0.8, and 0.7 on density and conductivity; not derived from first principles.
  • Adam learning rate and optimization schedule
    Fixed LR=0.001, batches of eight formulations, ≥20 cyclic iterations until loss drift <5%; hand-set hyperparameters of the fit.
  • Loss formulation weights ω_i
    Default weight 1, elevated to 4 or 8 for rare-component formulations to balance the fit.
  • Manual atom-type merges (e.g., fluorinated carbons → alkane quaternary carbons; Si retained as original OPLS)
    Ad hoc reductions from 78 to 68 types based on chemical similarity and experimental accessibility.
assumptions (6)
  • domain assumption Classical fixed-charge OPLS-AA functional form with PME electrostatics and LJ nonbonded terms can represent multicomponent ionic electrolytes after parameter adjustment.
    Stated throughout Models and Computational Methods; Limitations later note fixed-charge constraints.
  • domain assumption Experimental density is an adequate primary target for structural/thermodynamic accuracy of the force field.
    Force Field Model / Optimization: density is the loss; conductivity is validation only.
  • domain assumption Onsager transport formalism (including cross-correlations) is the appropriate conductivity estimator for concentrated electrolytes; NE is a limited baseline.
    Property Prediction section equations (2)–(4).
  • domain assumption Finite-size diffusion scales as 1/L (Yeh–Hummer); ~1e5 atoms and 35–40 ns NVT suffice for production transport properties.
    System-Size and Time Dependence section and standardized protocol choice.
  • standard math Trajectory reweighting in DMFF yields usable gradients of density w.r.t. LJ parameters without full re-simulation each step.
    Relies on the published DMFF framework (Wang et al.).
  • ad hoc to paper High-frequency EDB-1 solvents/salts define a representative commercial electrolyte chemical space after stated exclusion rules.
    Electrolyte Screening and Supporting Information screening criteria.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Differentiable OPLS Force Field Parameterization for Ionic Electrolytes and High-Throughput Application to Lithium-ion Batteries." pith.science (2026). https://pith.science/paper/CD7QV4IV

@misc{pith2026260704633,
  author       = {Pith},
  title        = {Pith review of: Differentiable OPLS Force Field Parameterization for Ionic Electrolytes and High-Throughput Application to Lithium-ion Batteries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CD7QV4IV}},
  note         = {Machine review of arXiv:2607.04633}
}
abstract

The rational design of ionic electrolytes for lithium-ion batteries (LIBs) is severely constrained by the vast solvent-salt combinatorial space and low efficiency of empirical trial-and-error. While molecular dynamics (MD) bridges microscopic solvation structures and macroscopic physicochemical properties, classical force fields often lack sufficient accuracy for multicomponent systems. To address these challenges, we develop an automated differentiable OPLS-AA force field parameterization workflow tailored for general ionic electrolytes. It employs topology-guided atom typification to reduce parameter redundancy and optimizes Lennard-Jones parameters via the DMFF framework, with experimental density as the fitting target and ionic conductivity as an independent validation metric. Rigorous convergence tests yield a standardized simulation protocol with $\sim$100,000-atom systems and 35-40 ns NVT runs to ensure reliable transport property quantification. High-throughput MD simulations of over 10,000 formulations spanning 67 solvents and 15 lithium salts are conducted on the Tianqiong platform, generating a comprehensive dataset covering five core properties: density, dielectric constant, viscosity, diffusion coefficient, and ionic conductivity. t-SNE visualization reveals partial clustering of distinct salt chemistries, continuous property gradients with concentration and temperature, and internal physical self-consistency, with solvent composition identified as another key performance regulator. Together, the accurate transferable force field and large-scale dataset provide a solid foundation for data-driven rational design of ionic electrolytes.

Figures

Figures reproduced from arXiv: 2607.04633 by the authors.

Figure 1
Figure 1. Workflow for transferable electrolyte force field optimization and high-throughput [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Density and conductivity prediction performance of models before and after fine [PITH_FULL_IMAGE:figures/full_fig_p017_2.png] view at source ↗
Figure 3
Figure 3. System-size and time dependence of transport properties at different salt concen [PITH_FULL_IMAGE:figures/full_fig_p020_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Chemical space of screened electrolyte formulations from the EDB-1 database [PITH_FULL_IMAGE:figures/full_fig_p022_4.png]
Figure 5
Figure 5. Figure 5: Parity plots of MD-predicted vs experimental ionic conductivity for single-salt [PITH_FULL_IMAGE:figures/full_fig_p024_5.png]
Figure 6
Figure 6. Figure 6: t-SNE 2D embedding of the single-salt electrolyte property dataset. All em [PITH_FULL_IMAGE:figures/full_fig_p027_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

62 extracted references · 45 canonical work pages

  1. [1]

    Journal of the American Chemical Society , author =

    Development and. Journal of the American Chemical Society , author =. 1996 , pages =. doi:10.1021/ja9621760 , abstract =

    doi:10.1021/ja9621760 1996

  2. [2]

    Chemical Reviews , author =

    Molecular. Chemical Reviews , author =. 2019 , pages =. doi:10.1021/acs.chemrev.8b00763 , abstract =

    work page doi:10.1021/acs.chemrev.8b00763 2019

  3. [3]

    Angewandte Chemie International Edition , author =

    An. Angewandte Chemie International Edition , author =. 2021 , pages =. doi:10.1002/anie.202107657 , abstract =

    work page doi:10.1002/anie.202107657 2021

  4. [4]

    Biophysical Journal , author =

    Confined lateral diffusion of membrane receptors as studied by single particle tracking (nanovid microscopy). Biophysical Journal , author =. 1993 , pages =. doi:10.1016/S0006-3495(93)81253-0 , abstract =

    work page doi:10.1016/s0006-3495(93)81253-0 1993

  5. [5]

    Materials Chemistry and Physics , author =

    Electrolytes in. Materials Chemistry and Physics , author =. 2024 , pages =. doi:10.1016/j.matchemphys.2023.128796 , language =

    work page doi:10.1016/j.matchemphys.2023.128796 2024

  6. [6]

    Crystal Growth & Design , author =

    Melting. Crystal Growth & Design , author =. 2025 , pages =. doi:10.1021/acs.cgd.4c01753 , abstract =

    work page doi:10.1021/acs.cgd.4c01753 2025

  7. [7]

    The Journal of Physical Chemistry B , author =

    Molecular. The Journal of Physical Chemistry B , author =. 2004 , pages =. doi:10.1021/jp0476545 , language =

    work page doi:10.1021/jp0476545 2004

  8. [8]

    Nanoscale , author =

    Challenges, opportunities, and roadmap for ionic liquid-based electrolytes in advancing energy storage devices , volume =. Nanoscale , author =. 2025 , pages =. doi:10.1039/D5NR00522A , language =

    work page doi:10.1039/d5nr00522a 2025

Show all 62 references
  1. [9]

    The Journal of Physical Chemistry B , author =

    System-. The Journal of Physical Chemistry B , author =. 2004 , pages =. doi:10.1021/jp0477147 , language =

    2004 doi

  2. [10]

    Molecular Simulation , author =

    Molecular dynamics simulations of concentrated aqueous electrolyte solutions , volume =. Molecular Simulation , author =. 2011 , pages =. doi:10.1080/08927022.2010.525513 , abstract =

    2011 doi

  3. [11]

    Comparing fine-tuning strategies of

    Alghamdi, Nada and Angelis, Paolo de and Asinari, Pietro and Chiavazzo, Eliodoro , month = apr, year =. Comparing fine-tuning strategies of. doi:10.48550/arXiv.2510.05020 , abstract =

    doi arXiv

  4. [12]

    J Chem Theory Comput , author =

    100. J Chem Theory Comput , author =. 2024 , pages =. doi:10.1021/acs.jctc.4c00135 , abstract =

    2024 doi

  5. [13]

    2023 , pages =

    J Chem Theory Comput , author =. 2023 , pages =. doi:10.1021/acs.jctc.2c01297 , abstract =

    2023 doi

  6. [14]

    Chemical Reviews , author =

    Applying. Chemical Reviews , author =. 2022 , pages =. doi:10.1021/acs.chemrev.1c00904 , abstract =

    2022 doi

  7. [15]

    Journal of Chemical Theory and Computation , author =

    Revisiting. Journal of Chemical Theory and Computation , author =. 2017 , pages =. doi:10.1021/acs.jctc.7b00520 , abstract =

    2017 doi

  8. [16]

    Angewandte Chemie International Edition , author =

    Probing the. Angewandte Chemie International Edition , author =. 2023 , pages =. doi:10.1002/anie.202305331 , abstract =

    2023 doi

  9. [17]

    The Journal of Physical Chemistry B , author =

    Quantum. The Journal of Physical Chemistry B , author =. 2025 , pages =. doi:10.1021/acs.jpcb.4c08461 , abstract =

    2025 doi

  10. [18]

    Journal of Computer-Aided Molecular Design , author =

    Class. Journal of Computer-Aided Molecular Design , author =. 1995 , pages =. doi:10.1007/BF00117280 , abstract =

    1995 doi

  11. [19]

    The Journal of Physical Chemistry B , author =

    1.14*. The Journal of Physical Chemistry B , author =. 2017 , pages =. doi:10.1021/acs.jpcb.7b00272 , abstract =

    2017 doi

  12. [20]

    ACS Macro Letters , author =

    Charge. ACS Macro Letters , author =. 2024 , pages =. doi:10.1021/acsmacrolett.4c00368 , abstract =

    2024 doi

  13. [21]

    The Journal of Physical Chemistry B , author =

    Influence of. The Journal of Physical Chemistry B , author =. 2019 , pages =. doi:10.1021/acs.jpcb.9b08033 , abstract =

    2019 doi

  14. [22]

    Journal of Chemical Theory and Computation , author =

    Reliable. Journal of Chemical Theory and Computation , author =. 2015 , pages =. doi:10.1021/acs.jctc.5b00351 , abstract =

    2015 doi

  15. [23]

    The Journal of Physical Chemistry B , author =

    Charge. The Journal of Physical Chemistry B , author =. 2023 , pages =. doi:10.1021/acs.jpcb.2c08047 , abstract =

    2023 doi

  16. [24]

    Journal of Chemical Theory and Computation , author =

    Charge. Journal of Chemical Theory and Computation , author =. 2025 , pages =. doi:10.1021/acs.jctc.5c00873 , abstract =

    2025 doi

  17. [25]

    Journal of Chemical Theory and Computation , author =

    Efficient. Journal of Chemical Theory and Computation , author =. 2021 , pages =. doi:10.1021/acs.jctc.1c00268 , abstract =

    2021 doi

  18. [26]

    Physical Chemistry Chemical Physics , author =

    Competitive lithium solvation of linear and cyclic carbonates from quantum chemistry , volume =. Physical Chemistry Chemical Physics , author =. 2016 , pages =. doi:10.1039/C5CP05121E , abstract =

    2016 doi

  19. [27]

    Journal of Molecular Liquids , author =

    Simulations of dielectric constants and viscosities of organic electrolytes by quantum mechanics and molecular dynamics , volume =. Journal of Molecular Liquids , author =. 2020 , pages =. doi:10.1016/j.molliq.2020.113288 , language =

    2020 doi

  20. [28]

    The Journal of Chemical Physics , author =

    The. The Journal of Chemical Physics , author =. 1939 , pages =. doi:10.1063/1.1750343 , abstract =

    1939 doi

  21. [29]

    The Journal of Physical Chemistry B , author =

    All-. The Journal of Physical Chemistry B , author =. 1998 , pages =. doi:10.1021/jp973084f , language =

    1998 doi

  22. [30]

    Journal of the American Chemical Society , author =

    A. Journal of the American Chemical Society , author =. 1995 , pages =. doi:10.1021/ja00124a002 , abstract =

    1995 doi

  23. [31]

    Nature Machine Intelligence , author =

    A unified predictive and generative solution for liquid electrolyte formulation , volume =. Nature Machine Intelligence , author =. 2026 , pages =. doi:10.1038/s42256-025-01173-w , language =

    2026 doi

  24. [32]

    2010 , pages =

    Journal of Computational Chemistry , author =. 2010 , pages =. doi:10.1002/jcc.21367 , abstract =

    2010 doi

  25. [33]

    Molecular Simulation , author =

    On the prediction of transport properties of ionic liquid using 1-n-butylmethylpyridinium tetrafluoroborate as an example , volume =. Molecular Simulation , author =. 2017 , pages =. doi:10.1080/08927022.2017.1321760 , abstract =

    2017 doi

  26. [34]

    Industrial & Engineering Chemistry Research , author =

    Thermal and. Industrial & Engineering Chemistry Research , author =. 2012 , pages =. doi:10.1021/ie300222a , abstract =

    2012 doi

  27. [35]

    The Journal of Chemical Physics , author =

    Molecular dynamics simulation of imidazolium-based ionic liquids. The Journal of Chemical Physics , author =. 2009 , pages =. doi:10.1063/1.3042279 , abstract =

    2009 doi

  28. [36]

    The Journal of Physical Chemistry B , author =

    Transport. The Journal of Physical Chemistry B , author =. 2006 , pages =. doi:10.1021/jp062885s , language =

    2006 doi

  29. [37]

    The Journal of Physical Chemistry B , author =

    The. The Journal of Physical Chemistry B , author =. 2015 , pages =. doi:10.1021/acs.jpcb.5b00689 , abstract =

    2015 doi

  30. [38]

    Wang, Donglin and Wang, Lei and Liu, Zijun and Wang, Tao and Pu, Zhonghua and Ding, Guangxin and Zhu, Mengchen and Yang, Lipeng and Guo, Ruoshan and Zhang, Zhiwei and Lin, Xiao and Du, Xueliang and Hao, Jie and Yang, Yongyong and Sun, Wenqin and Zhou, Fabiao and Xiao, NuoZhou ...

    2016 doi

  31. [39]

    Journal of Chemical Theory and Computation , author =

    Accelerators for. Journal of Chemical Theory and Computation , author =. 2022 , pages =. doi:10.1021/acs.jctc.1c01214 , abstract =

    2022 doi

  32. [40]

    Journal of Chemical Theory and Computation , author =

    Development of. Journal of Chemical Theory and Computation , author =. 2009 , pages =. doi:10.1021/ct900009a , abstract =

    2009 doi

  33. [41]

    and Tildesley, Dominic J

    Allen, Michael P. and Tildesley, Dominic J. , title =. 2017 , month =. doi:10.1093/oso/9780198803195.001.0001 , url =

    2017 doi

  34. [42]

    2017 , pages =

    Nucleic Acids Research , author =. 2017 , pages =. doi:10.1093/nar/gkx312 , abstract =

    2017 doi

  35. [43]

    and McGibbon, Robert T

    Eastman, Peter and Swails, Jason and Chodera, John D. and McGibbon, Robert T. and Zhao, Yutong and Beauchamp, Kyle A. and Wang, Lee-Ping and Simmonett, Andrew C. and Harrigan, Matthew P. and Stern, Chaya D. and Wiewiora, Rafal P. and Brooks, Bernard R. and Pande, Vijay S. , mo...

    doi

  36. [44]

    doi:10.5281/zenodo.3577986 , abstract =

    Lindahl and Abraham and Hess and Spoel, van der , month = dec, year =. doi:10.5281/zenodo.3577986 , abstract =

    doi

  37. [45]

    1995 , keywords =

    Computer Physics Communications , author =. 1995 , keywords =. doi:10.1016/0010-4655(95)00042-E , abstract =

    1995 doi

  38. [46]

    Proceedings of the National Academy of Sciences , author =

    Potential energy functions for atomic-level simulations of water and organic and biomolecular systems , volume =. Proceedings of the National Academy of Sciences , author =. 2005 , pages =. doi:10.1073/pnas.0408037102 , abstract =

    2005 doi

  39. [47]

    and Ba, Jimmy , month = jan, year =

    Kingma, Diederik P. and Ba, Jimmy , month = jan, year =. Adam:. doi:10.48550/arXiv.1412.6980 , abstract =

    doi arXiv

  40. [48]

    The Journal of Chemical Physics , author =

    Particle mesh. The Journal of Chemical Physics , author =. 1993 , pages =. doi:10.1063/1.464397 , abstract =

    1993 doi

  41. [49]

    Insight into physico-chemical properties of oxalatoborate-based ionic liquids through combined experimental-theoretical characterization , volume =

    Palluzzi, Matteo and Mannucci, Giorgia and Tsurumaki, Akiko and Busato, Matteo and Assunta Navarra, Maria and D’Angelo, Paola , year =. Insight into physico-chemical properties of oxalatoborate-based ionic liquids through combined experimental-theoretical characterization , vo...

    doi

  42. [50]

    The Journal of Chemical Physics , author =

    Effect of succinonitrile on ion transport in. The Journal of Chemical Physics , author =. 2022 , pages =. doi:10.1063/5.0087824 , abstract =

    2022 doi

  43. [51]

    Journal of Molecular Liquids , author =

    Fast lithium ion mobility in organic carbonates and acetonitrile: insights from ab initio and classical. Journal of Molecular Liquids , author =. 2025 , keywords =. doi:10.1016/j.molliq.2025.127736 , abstract =

    2025 doi

  44. [52]

    Journal of Molecular Liquids , author =

    Transferable potential for phase equilibrium of trialkyl borates , volume =. Journal of Molecular Liquids , author =. 2021 , keywords =. doi:10.1016/j.molliq.2021.116740 , abstract =

    2021 doi

  45. [53]

    , month = apr, year =

    Kumar, Ritesh and Vu, Minh Canh and Ma, Peiyuan and Amanchukwu, Chibueze V. , month = apr, year =. Electrolytomics:. Chemistry of Materials , publisher =. doi:10.1021/acs.chemmater.4c03196 , abstract =

    doi

  46. [54]

    03, 2016 , author=

    Gaussian 16 Revision A. 03, 2016 , author=. Gaussian Inc. Wallingford CT , volume=

    2016

  47. [55]

    Journal of Machine Learning Research , author =

    Visualizing. Journal of Machine Learning Research , author =. 2008 , pages =

    2008

  48. [56]

    bioRxiv , elocation-id =

    Chen, Lin and Chen, Yuhan and Cheng, Ziqi and Guo, Jing and He, Ming and Li, Huan and Li, Xinyu and Li, Zhen and Ma, Jule and Ma, Shuo and Peng, Can and Qian, Cheng and Qu, Zhi and Sun, Xian-gang and Tang, Xuan and Wang, Yunfei and Yu, Bing and Zhai, Yuzhuo and Zhang, Bingyao ...

    2026

  49. [57]

    InfoMat , volume =

    Liang, Yeru and Zhao, Chen-Zi and Yuan, Hong and Chen, Yuan and Zhang, Weicai and Huang, Jia-Qi and Yu, Dingshan and Liu, Yingliang and Titirici, Maria-Magdalena and Chueh, Yu-Lun and Yu, Haijun and Zhang, Qiang , title =. InfoMat , volume =. doi:https://doi.org/10.1002/inf2.1...

    doi

  50. [58]

    Advanced Energy Materials , volume =

    Li, Yuanqin and Liu, Mingzhu and Wang, Kang and Li, Chengfeng and Lu, Ying and Choudhary, Aditya and Ottley, Taylor and Bedrov, Dmitry and Xing, Lidan and Li, Weishan , title =. Advanced Energy Materials , volume =. doi:https://doi.org/10.1002/aenm.202300918 , url =. https://a...

    doi

  51. [59]

    Nature Energy , author =

    Long cycle life and dendrite-free lithium morphology in anode-free lithium pouch cells enabled by a dual-salt liquid electrolyte , volume =. Nature Energy , author =. 2019 , pages =. doi:10.1038/s41560-019-0428-9 , abstract =

    2019 doi

  52. [60]

    Energy Storage Materials , author =

    Engineering the solvent-anion interactions of. Energy Storage Materials , author =. 2024 , keywords =. doi:10.1016/j.ensm.2024.103816 , abstract =

    2024 doi

  53. [61]

    Advanced Energy Materials , author =

    Lithium. Advanced Energy Materials , author =. 2020 , note =. doi:10.1002/aenm.202001440 , abstract =

    2020 doi

  54. [62]

    Zhao, Yumeng and Hu, Zhenglin and Zhao, Zhengfei and Chen, Xinlian and Zhang, Shu and Gao, Jun and Luo, Jiayan , month = oct, year =. Strong. Journal of the American Chemical Society , publisher =. doi:10.1021/jacs.3c08313 , abstract =

    doi

Pith tools

Reviewed July 11, 2026 · model on record in the stance chip above.