1
|
Zhu S, Ramsundar B, Annevelink E, Lin H, Dave A, Guan PW, Gering K, Viswanathan V. Differentiable modeling and optimization of non-aqueous Li-based battery electrolyte solutions using geometric deep learning. Nat Commun 2024; 15:8649. [PMID: 39369004 PMCID: PMC11455955 DOI: 10.1038/s41467-024-51653-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/15/2023] [Accepted: 08/12/2024] [Indexed: 10/07/2024] Open
Abstract
Electrolytes play a critical role in designing next-generation battery systems, by allowing efficient ion transfer, preventing charge transfer, and stabilizing electrode-electrolyte interfaces. In this work, we develop a differentiable geometric deep learning (GDL) model for chemical mixtures, DiffMix, which is applied in guiding robotic experimentation and optimization towards fast-charging battery electrolytes. In particular, we extend mixture thermodynamic and transport laws by creating GDL-learnable physical coefficients. We evaluate our model with mixture thermodynamics and ion transport properties, where we show improved prediction accuracy and model robustness of DiffMix than its purely data-driven variants. Furthermore, with a robotic experimentation setup, Clio, we improve ionic conductivity of electrolytes by over 18.8% within 10 experimental steps, via differentiable optimization built on DiffMix gradients. By combining GDL, mixture physics laws, and robotic experimentation, DiffMix expands the predictive modeling methods for chemical mixtures and enables efficient optimization in large chemical spaces.
Collapse
Affiliation(s)
- Shang Zhu
- Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, USA
- Department of Mechanical Engineering, University of Michigan, Ann Arbor, USA
| | | | - Emil Annevelink
- Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, USA
| | - Hongyi Lin
- Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, USA
- Department of Mechanical Engineering, University of Michigan, Ann Arbor, USA
| | - Adarsh Dave
- Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, USA
| | - Pin-Wen Guan
- Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, USA
- Sandia National Laboratories, Livermore, USA
| | - Kevin Gering
- Energy Storage & Technology, Idaho National Laboratory, Idaho Falls, USA
| | - Venkatasubramanian Viswanathan
- Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, USA.
- Department of Mechanical Engineering, University of Michigan, Ann Arbor, USA.
- Department of Aerospace Engineering, University of Michigan, Ann Arbor, USA.
| |
Collapse
|
2
|
Carrer M, Cezar HM, Bore SL, Ledum M, Cascella M. Learning Force Field Parameters from Differentiable Particle-Field Molecular Dynamics. J Chem Inf Model 2024; 64:5510-5520. [PMID: 38963184 PMCID: PMC11267579 DOI: 10.1021/acs.jcim.4c00564] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/01/2024] [Revised: 05/15/2024] [Accepted: 06/12/2024] [Indexed: 07/05/2024]
Abstract
We develop ∂-HylleraasMD (∂-HyMD), a fully end-to-end differentiable molecular dynamics software based on the Hamiltonian hybrid particle-field formalism, and use it to establish a protocol for automated optimization of force field parameters. ∂-HyMD is templated on the recently released HylleraaasMD software, while using the JAX autodiff framework as the main engine for the differentiable dynamics. ∂-HyMD exploits an embarrassingly parallel optimization algorithm by spawning independent simulations, whose trajectories are simultaneously processed by reverse mode automatic differentiation to calculate the gradient of the loss function, which is in turn used for iterative optimization of the force-field parameters. We show that parallel organization facilitates the convergence of the minimization procedure, avoiding the known memory and numerical stability issues of differentiable molecular dynamics approaches. We showcase the effectiveness of our implementation by producing a library of force field parameters for standard phospholipids, with either zwitterionic or anionic heads and with saturated or unsaturated tails. Compared to the all-atom reference, the force field obtained by ∂-HyMD yields better density profiles than the parameters derived from previously utilized gradient-free optimization procedures. Moreover, ∂-HyMD models can predict with good accuracy properties not included in the learning objective, such as lateral pressure profiles, and are transferable to other systems, including triglycerides.
Collapse
Affiliation(s)
- Manuel Carrer
- Hylleraas Centre for Quantum Molecular
Sciences and Department of Chemistry, University
of Oslo, PO Box 1033, Blindern, 0315 Oslo, Norway
| | - Henrique Musseli Cezar
- Hylleraas Centre for Quantum Molecular
Sciences and Department of Chemistry, University
of Oslo, PO Box 1033, Blindern, 0315 Oslo, Norway
| | - Sigbjørn Løland Bore
- Hylleraas Centre for Quantum Molecular
Sciences and Department of Chemistry, University
of Oslo, PO Box 1033, Blindern, 0315 Oslo, Norway
| | - Morten Ledum
- Hylleraas Centre for Quantum Molecular
Sciences and Department of Chemistry, University
of Oslo, PO Box 1033, Blindern, 0315 Oslo, Norway
| | - Michele Cascella
- Hylleraas Centre for Quantum Molecular
Sciences and Department of Chemistry, University
of Oslo, PO Box 1033, Blindern, 0315 Oslo, Norway
| |
Collapse
|
3
|
Duignan TT. The Potential of Neural Network Potentials. ACS PHYSICAL CHEMISTRY AU 2024; 4:232-241. [PMID: 38800721 PMCID: PMC11117678 DOI: 10.1021/acsphyschemau.4c00004] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 01/18/2024] [Revised: 03/04/2024] [Accepted: 03/05/2024] [Indexed: 05/29/2024]
Abstract
In the next half-century, physical chemistry will likely undergo a profound transformation, driven predominantly by the combination of recent advances in quantum chemistry and machine learning (ML). Specifically, equivariant neural network potentials (NNPs) are a breakthrough new tool that are already enabling us to simulate systems at the molecular scale with unprecedented accuracy and speed, relying on nothing but fundamental physical laws. The continued development of this approach will realize Paul Dirac's 80-year-old vision of using quantum mechanics to unify physics with chemistry and providing invaluable tools for understanding materials science, biology, earth sciences, and beyond. The era of highly accurate and efficient first-principles molecular simulations will provide a wealth of training data that can be used to build automated computational methodologies, using tools such as diffusion models, for the design and optimization of systems at the molecular scale. Large language models (LLMs) will also evolve into increasingly indispensable tools for literature review, coding, idea generation, and scientific writing.
Collapse
|
4
|
Wu Z, Zhou T. Structural Coarse-Graining via Multiobjective Optimization with Differentiable Simulation. J Chem Theory Comput 2024; 20:2605-2617. [PMID: 38483262 DOI: 10.1021/acs.jctc.3c01348] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 03/27/2024]
Abstract
In the realm of multiscale molecular simulations, structure-based coarse-graining is a prominent approach for creating efficient coarse-grained (CG) representations of soft matter systems, such as polymers. This involves optimizing CG interactions by matching static correlation functions of the corresponding degrees of freedom in all-atom (AA) models. Here, we present a versatile method, namely, differentiable coarse-graining (DiffCG), which combines multiobjective optimization and differentiable simulation. The DiffCG approach is capable of constructing robust CG models by iteratively optimizing the effective potentials to simultaneously match multiple target properties. We demonstrate our approach by concurrently optimizing bonded and nonbonded potentials of a CG model of polystyrene (PS) melts. The resulting CG-PS model effectively reproduces both the structural characteristics, such as the equilibrium probability distribution of microscopic degrees of freedom and the thermodynamic pressure of the AA counterpart. More importantly, leveraging the multiobjective optimization capability, we develop a precise and efficient CG model for PS melts that is transferable across a wide range of temperatures, i.e., from 400 to 600 K. It is achieved via optimizing a pairwise potential with nonlinear temperature dependence in the CG model to simultaneously match target data from AA-MD simulations at multiple thermodynamic states. The temperature transferable CG-PS model demonstrates its ability to accurately predict the radial distribution functions and density at different temperatures, including those that are not included in the target thermodynamic states. Our work opens up a promising route for developing accurate and transferable CG models of complex soft-matter systems through multiobjective optimization with differentiable simulation.
Collapse
Affiliation(s)
- Zhenghao Wu
- Department of Chemistry, Xi'an Jiaotong-Liverpool University, Suzhou, Jiangsu 215123, P. R. China
| | - Tianhang Zhou
- College of Carbon Neutrality Future Technology, State Key Laboratory of Heavy Oil Processing, China University of Petroleum (Beijing), Beijing 102249, P. R. China
| |
Collapse
|
5
|
Matin S, Allen AEA, Smith J, Lubbers N, Jadrich RB, Messerly R, Nebgen B, Li YW, Tretiak S, Barros K. Machine Learning Potentials with the Iterative Boltzmann Inversion: Training to Experiment. J Chem Theory Comput 2024. [PMID: 38307009 DOI: 10.1021/acs.jctc.3c01051] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/04/2024]
Abstract
Methodologies for training machine learning potentials (MLPs) with quantum-mechanical simulation data have recently seen tremendous progress. Experimental data have a very different character than simulated data, and most MLP training procedures cannot be easily adapted to incorporate both types of data into the training process. We investigate a training procedure based on iterative Boltzmann inversion that produces a pair potential correction to an existing MLP using equilibrium radial distribution function data. By applying these corrections to an MLP for pure aluminum based on density functional theory, we observe that the resulting model largely addresses previous overstructuring in the melt phase. Interestingly, the corrected MLP also exhibits improved performance in predicting experimental diffusion constants, which are not included in the training procedure. The presented method does not require autodifferentiating through a molecular dynamics solver and does not make assumptions about the MLP architecture. Our results suggest a practical framework for incorporating experimental data into machine learning models to improve the accuracy of molecular dynamics simulations.
Collapse
Affiliation(s)
- Sakib Matin
- Department of Physics, Boston University, Boston, Massachusetts 02215, United States
- Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
- Center for Nonlinear Studies, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
| | - Alice E A Allen
- Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
- Center for Nonlinear Studies, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
| | - Justin Smith
- Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
- NVIDIA Corp., Santa Clara, California 95051, United States
| | - Nicholas Lubbers
- Computer, Computational, and Statistical Sciences Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States
| | - Ryan B Jadrich
- Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
| | - Richard Messerly
- Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
| | - Benjamin Nebgen
- Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
| | - Ying Wai Li
- Computer, Computational, and Statistical Sciences Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States
| | - Sergei Tretiak
- Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
- Center for Nonlinear Studies, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
- Center for Integrated Nanotechnologies, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
| | - Kipton Barros
- Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
- Center for Nonlinear Studies, Los Alamos National Laboratory, Los Alamos, New Mexico 87546, United States
| |
Collapse
|
6
|
Nicholas TC, Stones AE, Patel A, Michel FM, Reeder RJ, Aarts DGAL, Deringer VL, Goodwin AL. Geometrically frustrated interactions drive structural complexity in amorphous calcium carbonate. Nat Chem 2024; 16:36-41. [PMID: 37749235 PMCID: PMC10774122 DOI: 10.1038/s41557-023-01339-2] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/10/2023] [Accepted: 08/24/2023] [Indexed: 09/27/2023]
Abstract
Amorphous calcium carbonate is an important precursor for biomineralization in marine organisms. Key outstanding problems include understanding the structure of amorphous calcium carbonate and rationalizing its metastability as an amorphous phase. Here we report high-quality atomistic models of amorphous calcium carbonate generated using state-of-the-art interatomic potentials to help guide fits to X-ray total scattering data. Exploiting a recently developed inversion approach, we extract from these models the effective Ca⋯Ca interaction potential governing the structure. This potential contains minima at two competing distances, corresponding to the two different ways that carbonate ions bridge Ca2+-ion pairs. We reveal an unexpected mapping to the Lennard-Jones-Gauss model normally studied in the context of computational soft matter. The empirical model parameters for amorphous calcium carbonate take values known to promote structural complexity. We thus show that both the complex structure and its resilience to crystallization are actually encoded in the geometrically frustrated effective interactions between Ca2+ ions.
Collapse
Affiliation(s)
- Thomas C Nicholas
- Inorganic Chemistry Laboratory, Department of Chemistry, University of Oxford, Oxford, UK
| | - Adam Edward Stones
- Physical and Theoretical Chemistry Laboratory, Department of Chemistry, University of Oxford, Oxford, UK
| | - Adam Patel
- Inorganic Chemistry Laboratory, Department of Chemistry, University of Oxford, Oxford, UK
| | - F Marc Michel
- Department of Geosciences, Virginia Tech, Blacksburg, VA, USA
| | - Richard J Reeder
- Department of Geosciences, Stony Brook University, Stony Brook, NY, USA
| | - Dirk G A L Aarts
- Physical and Theoretical Chemistry Laboratory, Department of Chemistry, University of Oxford, Oxford, UK
| | - Volker L Deringer
- Inorganic Chemistry Laboratory, Department of Chemistry, University of Oxford, Oxford, UK.
| | - Andrew L Goodwin
- Inorganic Chemistry Laboratory, Department of Chemistry, University of Oxford, Oxford, UK.
| |
Collapse
|
7
|
Bachs-Herrera A, York D, Stephens-Jones T, Mabbett I, Yeo J, Martin-Martinez FJ. Biomass carbon mining to develop nature-inspired materials for a circular economy. iScience 2023; 26:106549. [PMID: 37123246 PMCID: PMC10130920 DOI: 10.1016/j.isci.2023.106549] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 04/03/2023] Open
Abstract
A transition from a linear to a circular economy is the only alternative to reduce current pressures in natural resources. Our society must redefine our material sources, rethink our supply chains, improve our waste management, and redesign materials and products. Valorizing extensively available biomass wastes, as new carbon mines, and developing biobased materials that mimic nature's efficiency and wasteless procedures are the most promising avenues to achieve technical solutions for the global challenges ahead. Advances in materials processing, and characterization, as well as the rise of artificial intelligence, and machine learning, are supporting this transition to a new materials' mining. Location, cultural, and social aspects are also factors to consider. This perspective discusses new alternatives for carbon mining in biomass wastes, the valorization of biomass using available processing techniques, and the implementation of computational modeling, artificial intelligence, and machine learning to accelerate material's development and process engineering.
Collapse
Affiliation(s)
| | - Daniel York
- Department of Chemistry, Swansea University, Swansea SA2 8PP, UK
| | | | - Ian Mabbett
- Department of Chemistry, Swansea University, Swansea SA2 8PP, UK
| | - Jingjie Yeo
- Sibley School of Mechanical and Aerospace Engineering, Cornell University, Ithaca, NY 14853, USA
| | | |
Collapse
|