Received date: 2024-05-06
Revised date: 2024-07-30
Online published: 2024-08-26
Supported by
University of Hong Kong via Seed Fund(2201100392)
Although first-principles calculations offer high precision, they are prohibitively expensive. Conversely, molecular dynamics simulations employing classical interatomic potentials, or force fields, offer quicker but less precise outcomes. To balance between computational speed and accuracy, machine learning (ML) potential functions have been developed and have gained widespread application. The deep potential (DP) method, a type of ML potential, has attracted considerable attention recently. This paper provides a comprehensive review of DP methods in materials science. It begins with an introduction to the theoretical foundation of DP, followed by a detailed exposition of the DP model development and usage. Additionally, the application of DP in various material systems is briefly reviewed. AIS-Square contributes training databases and workflows essential for developing DP models. The paper concludes by assessing the performance of DP models relative to both first-principles calculations and classical potentials in terms of accuracy and efficiency. Finally, a brief outlook on future developments trends is provided.
WEN Tongqi , LIU Huaiyi , GONG Xiaoguo , YE Beilin , LIU Siyu , LI Zhuoyuan . Deep Potentials for Materials Science[J]. Acta Metall Sin, 2024 , 60(10) : 1299 -1311 . DOI: 10.11900/0412.1961.2024.00141
| 1 | Hafner J. Atomic-scale computational materials science [J]. Acta Mater., 2000, 48: 71 |
| 2 | Born M, Oppenheimer R. Zur quantentheorie der molekeln [J]. Ann. Phys., 1927, 389: 457 |
| 3 | Dirac P A M. Quantum mechanics of many-electron systems [J]. Proc. Roy. Soc., 1929, 123A: 714 |
| 4 | Kohn W, Sham L J. Self-consistent equations including exchange and correlation effects [J]. Phys. Rev., 1965, 140: A1133 |
| 5 | Verlet L. Computer “experiments” on classical fluids. I. Thermodynamical properties of Lennard-Jones molecules [J]. Phys. Rev., 1967, 159: 98 |
| 6 | Zwanzig R W. High-temperature equation of state by a perturbation method. I. Nonpolar gases [J]. J. Chem. Phys., 1954, 22: 1420 |
| 7 | Tersoff J. Modeling solid-state chemistry: Interatomic potentials for multicomponent systems [J]. Phys. Rev., 1989, 39B: 5566 |
| 8 | Vink R L C, Barkema G T, van der Weg W F, et al. Fitting the Stillinger-Weber potential to amorphous silicon [J]. J. Non-Cryst. Solids, 2001, 282: 248 |
| 9 | Daw M S, Baskes M I. Embedded-atom method: Derivation and application to impurities, surfaces, and other defects in metals [J]. Phys. Rev., 1984, 29B: 6443 |
| 10 | Baskes M I. Modified embedded-atom potentials for cubic materials and impurities [J]. Phys. Rev., 1992, 46B: 2727 |
| 11 | Prentice J C A, Aarons J, Womack J C, et al. The ONETEP linear-scaling density functional theory program [J]. J. Chem. Phys., 2020, 152: 174111 |
| 12 | Hacene M, Anciaux-Sedrakian A, Rozanska X, et al. Accelerating VASP electronic structure calculations using graphic processing units [J]. J. Comput. Chem., 2012, 33: 2581 |
| 13 | Hutchinson M, Widom M. VASP on a GPU: Application to exact-exchange calculations of the stability of elemental boron [J]. Comput. Phys. Commun., 2012, 183: 1422 |
| 14 | Jia W L, Cao Z Y, Wang L, et al. The analysis of a plane wave pseudopotential density functional theory code on a GPU machine [J]. Comput. Phys. Commun., 2013, 184: 9 |
| 15 | Jia W L, Fu J Y, Cao Z Y, et al. Fast plane wave density functional theory molecular dynamics calculations on multi-GPU machines [J]. J. Comput. Phys., 2013, 251: 102 |
| 16 | Bishop C M. Pattern Recognition and Machine Learning [M]. New York: Springer, 2006: 1122 |
| 17 | Jordan M I, Mitchell T M. Machine learning: Trends, perspectives, and prospects [J]. Science, 2015, 349: 255 |
| 18 | Mahesh B. Machine learning algorithms-A review [J]. Int. J. Sci. Res., 2020, 9: 381 |
| 19 | Blank T B, Brown S D, Calhoun A W, et al. Neural network models of potential energy surfaces [J]. J. Chem. Phys., 1995, 103: 4129 |
| 20 | Behler J, Parrinello M. Generalized neural-network representation of high-dimensional potential-energy surfaces [J]. Phys. Rev. Lett., 2007, 98: 146401 |
| 21 | Schütt K T, Sauceda H E, Kindermans P J, et al. SchNet—A deep learning architecture for molecules and materials [J]. J. Chem. Phys., 2018, 148: 241722 |
| 22 | Schütt K T, Kessel P, Gastegger M, et al. SchNetPack: A deep learning toolbox for atomistic systems [J]. J. Chem. Theory Comput., 2019, 15: 448 |
| 23 | Ghasemi S A, Hofstetter A, Saha S, et al. Interatomic potentials for ionic systems with density functional accuracy based on charge densities obtained by a neural network [J]. Phys. Rev., 2015, 92B: 045131 |
| 24 | Hy T S, Trivedi S, Pan H, et al. Predicting molecular properties with covariant compositional networks [J]. J. Chem. Phys., 2018, 148: 241745 |
| 25 | Unke O T, Meuwly M. PhysNet: A neural network for predicting energies, forces, dipole moments, and partial charges [J]. J. Chem. Theory Comput., 2019, 15: 3678 |
| 26 | Pun G P, Batra R, Ramprasad R, et al. Physically informed artificial neural networks for atomistic modeling of materials [J]. Nat. Commun., 2019, 10: 2339 |
| 27 | Zuo Y X, Chen C, Li X G, et al. Performance and cost assessment of machine learning interatomic potentials [J]. J. Phys. Chem., 2020, 124A: 731 |
| 28 | Han J Q, Zhang L F, Car R, et al. Deep potential: A general representation of a many-body potential energy surface [J]. Commun. Comput. Phys., 2018, 23: 629 |
| 29 | Zhang L F, Han J Q, Wang H, et al. Deep potential molecular dynamics: A scalable model with the accuracy of quantum mechanics [J]. Phys. Rev. Lett., 2018, 120: 143001 |
| 30 | Wang H, Zhang L F, Han J Q, et al. DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics [J]. Comput. Phys. Commun., 2018, 228: 178 |
| 31 | Jia W L, Wang H, Chen M H, et al. Pushing the limit of molecular dynamics with ab initio accuracy to 100 million atoms with machine learning [A]. Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis [C]. Atlanta: IEEE, 2020: 1 |
| 32 | Zhang D, Bi H R, Dai F Z, et al. Pretraining of attention-based deep learning potential model for molecular simulation [J]. npj Comput. Mater., 2024, 10: 94 |
| 33 | Wang H, Zhang D, Liu X Z J, et al. DPA-2: Towards a universal large atomic model for molecular and materials simulation [DB/OL]. arXiv: 2312. 15492, 2024 |
| 34 | Zhang L F, Han J Q, Wang H, et al. End-to-end symmetry preserving inter-atomic potential energy model for finite and extended systems [A]. Advances in Neural Information Processing Systems [C]. Montreal, Canada, Dec.2-8, 2018 |
| 35 | Perdew J P, Schmidt K. Jacob's ladder of density functional approximations for the exchange-correlation energy [J]. AIP Conf. Proc., 2001, 577: 1 |
| 36 | DiStasio R A, Santra B, Li Z F, et al. The individual and collective effects of exact exchange and dispersion interactions on the ab initio structure of liquid water [J]. J. Chem. Phys., 2014, 141: 084502 |
| 37 | Li Z Y, Wen T Q, Zhang Y Z, et al. An extendable cloud-native alloy property explorer [DB/OL]. arXiv: 2404. 17330, 2024 |
| 38 | Gasteiger J, Shuaibi M, Sriram A, et al. GemNet-OC: Developing graph neural networks for large and diverse molecular simulation datasets [J]. Trans. Mach. Learn. Res., 2022 |
| 39 | Lu D H, Wang H, Chen M H, et al. 86 PFLOPS deep potential molecular dynamics simulation of 100 million atoms with ab initio accuracy [J]. Comput. Phys. Commun., 2021, 259: 107624 |
| 40 | Lu D H, Jiang W R, Chen Y X, et al. DP compress: A model compression scheme for generating efficient deep potential models [J]. J. Chem. Theory Comput., 2022, 18: 5559 |
| 41 | Wen T Q, Wang R, Zhu L Y, et al. Specialising neural network potentials for accurate properties and application to the mechanical response of titanium [J]. npj Comput. Mater., 2021, 7: 206 |
| 42 | Mendelev M I, Underwood T L, Ackland G J. Development of an interatomic potential for the simulation of defects, plasticity, and phase transformations in titanium [J]. J. Chem. Phys., 2016, 145: 154102 |
| 43 | Hennig R G, Lenosky T J, Trinkle D R, et al. Classical potential describes martensitic phase transformations between the α, β, and ω titanium phases [J]. Phys. Rev., 2008, 78B: 054121 |
| 44 | Wen T Q, Zhang L F, Wang H, et al. Deep potentials for materials science [J]. Mater. Futures, 2022, 1: 022601 |
| 45 | Zhang L F, Lin D Y, Wang H, et al. Active learning of uniformly accurate interatomic potentials for materials simulation [J]. Phys. Rev. Mater., 2019, 3: 023804 |
| 46 | Wang H, Guo X, Zhang L F, et al. Deep learning inter-atomic potential model for accurate irradiation damage simulations [J]. Appl. Phys. Lett., 2019, 114: 244101 |
| 47 | Clouet E, Caillard D, Chaari N, et al. Dislocation locking versus easy glide in titanium and zirconium [J]. Nat. Mater., 2015, 14: 931 |
| 48 | Wang X Y, Wang Y N, Zhang L F, et al. A tungsten deep neural-network potential for simulating mechanical property degradation under fusion service environment [J]. Nucl. Fusion, 2022, 62: 126013 |
| 49 | Gong X G, Li Z Y, Pattamatta A S L S, et al. An accurate and transferable machine learning interatomic potential for nickel [J]. Commun. Mater., 2024, 5: 157 |
| 50 | Andolina C M, Bon M, Passerone D, et al. Robust, multi-length-scale, machine learning potential for Ag-Au bimetallic alloys from clusters to bulk materials [J]. J. Phys. Chem., 2021, 125: 17438 |
| 51 | Wang Y N, Zhang L F, Xu B, et al. A generalizable machine learning potential of Ag-Au nanoalloys and its application to surface reconstruction, segregation and diffusion [J]. Modell. Simul. Mater. Sci. Eng., 2022, 30: 025003 |
| 52 | Chen B, Zeng Q Y, Wang H, et al. Atomistic mechanism of phase transition in shock compressed gold revealed by deep potential [DB/OL]. arXiv: 2006. 13136, 2020 |
| 53 | Jiao J Y, Lai G M, Zhao L, et al. Self-healing mechanism of lithium in lithium metal [J]. Adv. Sci., 2022, 9: 2105574 |
| 54 | Niu H Y, Bonati L, Piaggi P M, et al. Ab initio phase diagram and nucleation of gallium [J]. Nat. Commun., 2020, 11: 2654 |
| 55 | Wang J J, Shen H, Yang R Y, et al. A deep learning interatomic potential developed for atomistic simulation of carbon materials [J]. Carbon, 2022, 186: 1 |
| 56 | Bonati L, Parrinello M. Silicon liquid structure and crystal nucleation from ab initio deep metadynamics [J]. Phys. Rev. Lett., 2018, 121: 265701 |
| 57 | Li R, Lee E, Luo T. A unified deep neural network potential capable of predicting thermal conductivity of silicon in different phases [J]. Mater. Today Phys., 2020, 12: 100181 |
| 58 | Yang M Y, Karmakar T, Parrinello M. Liquid-liquid critical point in phosphorus [J]. Phys. Rev. Lett., 2021, 127: 080603 |
| 59 | Jain A, Ong S P, Hautier G, et al. Commentary: The materials project: A materials genome approach to accelerating materials innovation [J]. APL Mater., 2013, 1: 011002 |
| 60 | Wang H D, Zhang Y Z, Zhang L F, et al. Crystal structure prediction of binary alloys via deep potential [J]. Front. Chem., 2020, 8: 589795 |
| 61 | Andolina C M, Wright J G, Das N, et al. Improved Al-Mg alloy surface segregation predictions with a machine learning atomistic potential [J]. Phys. Rev. Mater., 2021, 5: 083804 |
| 62 | Jiang W R, Zhang Y Z, Zhang L F, et al. Accurate deep potential model for the Al-Cu-Mg alloy in the full concentration space [J]. Chin. Phys., 2021, 30B: 050706 |
| 63 | Cao G H, Liang J J, Guo Z L, et al. Liquid metal for high-entropy alloy nanoparticles synthesis [J]. Nature, 2023, 619: 73 |
| 64 | Bourgeois L, Zhang Y, Zhang Z Z, et al. Transforming solid-state precipitates via excess vacancies [J]. Nat. Commun., 2020, 11: 1248 |
| 65 | Cheng B Q, Zhao X J, Zhang Y, et al. Co-segregation of Mg and Zn atoms at the planar η1-precipitate/Al matrix interface in an aged Al-Zn-Mg alloy [J]. Scr. Mater., 2020, 185: 51 |
| 66 | Wen T Q, Wang C Z, Kramer M J, et al. Development of a deep machine learning interatomic potential for metalloid-containing Pd-Si compounds [J]. Phys. Rev., 2019, 100B: 174101 |
| 67 | Tang L, Yang Z J, Wen T Q, et al. Development of interatomic potential for Al-Tb alloys using a deep neural network learning method [J]. Phys. Chem. Chem. Phys., 2020, 22: 18467 |
| 68 | Tang L, Yang Z J, Wen T Q, et al. Short- and medium-range orders in Al90Tb10 glass and their relation to the structures of competing crystalline phases [J]. Acta Mater., 2021, 204: 116513 |
| 69 | Han I, McKeown J T, Tang L, et al. Dynamic observation of dendritic quasicrystal growth upon laser-induced solid-state transformation [J]. Phys. Rev. Lett., 2020, 125: 195503 |
| 70 | Tang L, Ho K M, Wang C Z. Molecular dynamics simulation of metallic Al-Ce liquids using a neural network machine learning interatomic potential [J]. J. Chem. Phys., 2021, 155: 194503 |
| 71 | Balyakin I A, Rempel S V, Ryltsev R E, et al. Deep machine learning interatomic potential for liquid silica [J]. Phys. Rev., 2020, 102E: 052125 |
| 72 | Wan T Q, Luo C X, Sun Y, et al. Thermoelastic properties of bridgmanite using deep-potential molecular dynamics [J]. Phys. Rev., 2024, 109B: 094101 |
| 73 | Huang J X, Zhang L F, Wang H, et al. Deep potential generation scheme and simulation protocol for the Li10GeP2S12-type superionic conductors [J]. J. Chem. Phys., 2021, 154: 094703 |
| 74 | Ko H Y, Zhang L, Santra B, et al. Isotope effects in liquid water via deep potential molecular dynamics [J]. Mol. Phys., 2019, 117: 3269 |
| 75 | Zhang C Y, Zhang L F, Xu J H, et al. Isotope effects in x-ray absorption spectra of liquid water [J]. Phys. Rev., 2020, 102B: 115155 |
| 76 | Xu J H, Zhang C Y, Zhang L F, et al. Isotope effects in molecular structures and electronic properties of liquid water via deep potential molecular dynamics based on the SCAN functional [J]. Phys. Rev., 2020, 102B: 214113 |
| 77 | Calio P B, Li C H, Voth G A. Resolving the structural debate for the hydrated excess proton in water [J]. J. Am. Chem. Soc., 2021, 143: 18672 |
| 78 | Sommers G M, Calegari Andrade M F, Zhang L F, et al. Raman spectrum and polarizability of liquid water from deep neural networks [J]. Phys. Chem. Chem. Phys., 2020, 22: 10592 |
| 79 | Gartner T E, Zhang L F, Piaggi P M, et al. Signatures of a liquid-liquid transition in an ab initio deep neural network model for water [J]. Proc. Natl. Acad. Sci. USA, 2020, 117: 26040 |
| 80 | Andreani C, Romanelli G, Parmentier A, et al. Hydrogen dynamics in supercritical water probed by neutron scattering and computer simulations [J]. J. Phys. Chem. Lett., 2020, 11: 9461 |
| 81 | Piaggi P M, Panagiotopoulos A Z, Debenedetti P G, et al. Phase equilibrium of water with hexagonal and cubic ice using the SCAN functional [J]. J. Chem. Theory Comput., 2021, 17: 3065 |
| 82 | Zhang L F, Wang H, Car R, et al. Phase diagram of a deep potential water model [J]. Phys. Rev. Lett., 2021, 126: 236001 |
| 83 | Aragones J L, Conde M M, Noya E G, et al. The phase diagram of water at high pressures as obtained by computer simulations of the TIP4P/2005 model: The appearance of a plastic crystal phase [J]. Phys. Chem. Chem. Phys., 2009, 11: 543 |
| 84 | Tisi D, Zhang L F, Bertossa R, et al. Heat transport in liquid water from first-principles and deep neural network simulations [J]. Phys. Rev., 2021, 104B: 224202 |
| 85 | Zhang C Y, Tang F J, Chen M H, et al. Modeling liquid water by climbing up Jacob's ladder in density functional theory facilitated by using deep neural network potentials [J]. J. Phys. Chem., 2021, 125B: 11444 |
| 86 | Torres A, Pedroza L S, Fernandez-Serra M, et al. Using neural network force fields to ascertain the quality of ab initio simulations of liquid water [J]. J. Phys. Chem., 2021, 125B: 10772 |
| 87 | Shi Y, Doyle C C, Beck T L. Condensed phase water molecular multipole moments from deep neural network models trained on ab initio simulation data [J]. J. Phys. Chem. Lett., 2021, 12: 10310 |
| 88 | Chen M Y, Tan L, Wang H, et al. Imperfectly coordinated water molecules pave the way for homogeneous ice nucleation [DB/OL]. arXiv: 2304. 12665, 2023 |
| 89 | Xu M Y, Zhu T, Zhang J Z H. Molecular dynamics simulation of zinc ion in water with an ab initio based neural network potential [J]. J. Phys. Chem., 2019, 123A: 6587 |
| 90 | Niblett S P, Galib M, Limmer D T. Learning intermolecular forces at liquid-vapor interfaces [J]. J. Chem. Phys., 2021, 155: 164101 |
| 91 | Galib M, Limmer D T. Reactive uptake of N2O5 by atmospheric aerosol is dominated by interfacial processes [J]. Science, 2021, 371: 921 |
/
| 〈 |
|
〉 |
