Neural network potentials for metals and oxides – First applications to copper clusters at zinc oxide

The development of reliable interatomic potentials for large-scale molecular dynamics (MD) simulations of chemical processes at surfaces and interfaces is a formidable challenge because a wide range of atomic environments and very different types of bonding can be present. In recent years interatomic potentials based on artificial neural networks (NNs) have emerged offering an unbiased approach to the construction of potential energy surfaces (PESs) for systems that are difficult to describe by conventional potentials. Here, we review the basic properties of NN potentials and describe their construction for materials like metals and oxides. The accuracy and efficiency are demonstrated using copper and zinc oxide as benchmark systems. First results for a potential of the combined ternary CuZnO system aiming at the description of oxide-supported copper clusters are reported. [GRAPHICS] Model of a copper cluster at the ZnO($10\overline {1} 0$) surface.

[1]  Campbell,et al.  Growth model for metal films on oxide surfaces: Cu on ZnO(0001)-O. , 1993, Physical review. B, Condensed matter.

[2]  Steven Hobday,et al.  Applications of neural networks to fitting interatomic potential functions , 1999 .

[3]  D. Marx,et al.  F centers versus dimer vacancies on ZnO surfaces: characterization by STM and STS calculations. , 2007, Angewandte Chemie.

[4]  J. Grunwaldt,et al.  In Situ Investigations of Structural Changes in Cu/ZnO Catalysts , 2000 .

[5]  Donald L. Thompson,et al.  Interpolating moving least-squares methods for fitting potential energy surfaces: Detailed analysis of one-dimensional applications , 2003 .

[6]  Paul L. A. Popelier,et al.  A polarizable high-rank quantum topological electrostatic potential developed using neural networks: Molecular dynamics simulations on the hydrogen fluoride dimer , 2007 .

[7]  D. Marx,et al.  Methanol synthesis on ZnO(0001). III. Free energy landscapes, reaction pathways, and mechanistic insights. , 2011, Journal of Chemical Physics.

[8]  Burke,et al.  Generalized Gradient Approximation Made Simple. , 1996, Physical review letters.

[9]  Paul L. A. Popelier,et al.  Polarisable multipolar electrostatics from the machine learning method Kriging: an application to alanine , 2012, Theoretical Chemistry Accounts.

[10]  Ajeevsing Bholoa,et al.  A new approach to potential fitting using neural networks , 2007 .

[11]  Robert Schlögl,et al.  Role of lattice strain and defects in copper particles on the activity of Cu/ZnO/Al(2)O(3) catalysts for methanol synthesis. , 2007, Angewandte Chemie.

[12]  Michele Parrinello,et al.  Generalized neural-network representation of high-dimensional potential-energy surfaces. , 2007, Physical review letters.

[13]  P. Popelier,et al.  Potential energy surfaces fitted by artificial neural networks. , 2010, The journal of physical chemistry. A.

[14]  Density-functional study of Cu atoms, monolayers, films, and coadsorbates on polar ZnO surfaces , 2003, cond-mat/0303143.

[15]  Rustam Z. Khaliullin,et al.  Microscopic origins of the anomalous melting behavior of sodium under high pressure. , 2011, Physical review letters.

[16]  J. Tersoff,et al.  New empirical model for the structural properties of silicon. , 1986, Physical review letters.

[17]  Roger M. Smith,et al.  Applications of genetic algorithms and neural networks to interatomic potentials , 1999 .

[18]  Michele Parrinello,et al.  Nucleation mechanism for the direct graphite-to-diamond phase transition. , 2011, Nature materials.

[19]  Nongnuch Artrith,et al.  High-dimensional neural network potentials for metal surfaces: A prototype study for copper , 2012 .

[20]  Conyers Herring,et al.  Some Theorems on the Free Energies of Crystal Surfaces , 1951 .

[21]  B. Delley,et al.  Cu/ZnO(0001) under oxidating and reducing conditions: A first-principles survey of surface structures , 2011 .

[22]  J. J. Soares Neto,et al.  The fitting of potential energy surfaces using neural networks: Application to the study of vibrational levels of H3+ , 1998 .

[23]  J. Behler,et al.  Dissociation of O2 at Al(111): the role of spin selection rules. , 2004, Physical Review Letters.

[24]  U. Köhler,et al.  Small Cu-clusters on ZnO(0 0 0 1) Zn: Nucleation and annealing behavior , 2007 .

[25]  Dmitrii E. Makarov,et al.  FITTING POTENTIAL-ENERGY SURFACES : A SEARCH IN THE FUNCTION SPACE BY DIRECTED GENETIC PROGRAMMING , 1998 .

[26]  Mark N. Gibbs,et al.  Combining ab initio computations, neural networks, and diffusion Monte Carlo: An efficient method to treat weakly bound molecules , 1996 .

[27]  Jens R. Rostrup-Nielsen,et al.  Atom-Resolved Imaging of Dynamic Shape Changes in Supported Copper Nanocrystals , 2002, Science.

[28]  C. Hättig,et al.  Prediction of vibrational frequencies of possible intermediates and side products of the methanol synthesis on ZnO(0001) by ab initio calculations. , 2012, The Journal of chemical physics.

[29]  Rustam Z. Khaliullin,et al.  Graphite-diamond phase coexistence study employing a neural-network mapping of the ab initio potential energy surface , 2010 .

[30]  George Cybenko,et al.  Approximation by superpositions of a sigmoidal function , 1992, Math. Control. Signals Syst..

[31]  Sergei Manzhos,et al.  Using neural networks to represent potential surfaces as sums of products. , 2006, The Journal of chemical physics.

[32]  Foiles,et al.  Embedded-atom-method functions for the fcc metals Cu, Ag, Au, Ni, Pd, Pt, and their alloys. , 1986, Physical review. B, Condensed matter.

[33]  G. Ertl,et al.  Handbook of Heterogeneous Catalysis , 1997 .

[34]  Matthias Scheffler,et al.  Ab initio molecular simulations with numeric atom-centered orbitals , 2009, Comput. Phys. Commun..

[35]  J. Behler,et al.  Construction of high-dimensional neural network potentials using environment-dependent atom pairs. , 2012, The Journal of chemical physics.

[36]  B. Delley,et al.  Oxygen adsorption and stability of surface oxides on Cu(111) : A first-principles investigation , 2006 .

[37]  R Komanduri,et al.  Ab initio potential-energy surfaces for complex, multichannel systems using modified novelty sampling and feedforward neural networks. , 2005, The Journal of chemical physics.

[38]  Ajeevsing Bholoa,et al.  Silicon potentials investigated using density functional theory fitted neural networks , 2008 .

[39]  Hung M. Le,et al.  Modified feed-forward neural network structures and combined-function-derivative approximations incorporating exchange symmetry for potential energy surface fitting. , 2012, The journal of physical chemistry. A.

[40]  M. A. Collins,et al.  Molecular potential energy surfaces by interpolation , 1994 .

[41]  J. Nørskov,et al.  The Active Site of Methanol Synthesis over Cu/ZnO/Al2O3 Industrial Catalysts , 2012, Science.

[42]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[43]  D. Pettifor,et al.  Atomistic modelling of materials with bond-order potentials , 2009 .

[44]  Paul L A Popelier,et al.  Beyond Point Charges: Dynamic Polarization from Neural Net Predicted Multipole Moments. , 2008, Journal of chemical theory and computation.

[45]  Jörg Behler,et al.  A neural network potential-energy surface for the water dimer based on environment-dependent atomic energies and charges. , 2012, The Journal of chemical physics.

[46]  Jeffery K Ludwig,et al.  Ab initio molecular dynamics of hydrogen dissociation on metal surfaces using neural networks and novelty sampling. , 2007, The Journal of chemical physics.

[47]  Germany,et al.  Neural network interatomic potential for the phase change material GeTe , 2012, 1201.2026.

[48]  Kersti Hermansson,et al.  Representation of Intermolecular Potential Functions by Neural Networks , 1998 .

[49]  R. Kondor,et al.  Gaussian approximation potentials: the accuracy of quantum mechanics, without the electrons. , 2009, Physical review letters.

[50]  Steven D. Brown,et al.  Neural network models of potential energy surfaces , 1995 .

[51]  J. Behler,et al.  Metadynamics simulations of the high-pressure phases of silicon employing a high-dimensional neural network potential. , 2008, Physical review letters.

[52]  J. Wagner,et al.  In Situ Electron Energy Loss Spectroscopy Studies of Gas-Dependent Metal−Support Interactions in Cu/ZnO Catalysts , 2003 .

[53]  J. Neugebauer,et al.  Temperature stabilized surface reconstructions at polar ZnO(0001). , 2009, Physical review letters.

[54]  Nongnuch Artrith,et al.  High-dimensional neural-network potentials for multicomponent systems: Applications to zinc oxide , 2011 .

[55]  J. Behler,et al.  Representing molecule-surface interactions with symmetry-adapted neural networks. , 2007, The Journal of chemical physics.

[56]  J. Behler.  Atom-centered symmetry functions for constructing high-dimensional neural network potentials. , 2011, The Journal of chemical physics.

[57]  Density-functional study of the structure and stability of ZnO surfaces , 2002, cond-mat/0206549.

[58]  A. Gross,et al.  Representing high-dimensional potential-energy surfaces for reactions at surfaces by neural networks , 2004 .

[59]  B. Delley,et al.  Density functional study of oxygen on Cu(100) and Cu(110) surfaces , 2010 .

[60]  C. Wöll,et al.  Thermal behavior of MOCVD-grown Cu-clusters on ZnO(1010). , 2012, Physical chemistry chemical physics : PCCP.

[61]  Thomas B. Blank,et al.  Adaptive, global, extended Kalman filters for training feedforward neural networks , 1994 .

[62]  B. Delley,et al.  Surface oxides of the oxygen–copper system: Precursors to the bulk oxide phase? , 2007 .

[63]  Jörg Behler,et al.  Thermal transport in phase-change materials from atomistic simulations , 2012 .

[64]  D. Marx,et al.  First-principles study of CO adsorption on ZnO surfaces , 2003 .

[65]  Marco Bernasconi,et al.  Breakdown of Stokes–Einstein relation in the supercooled liquid state of phase change materials , 2012, 1207.7269.

[66]  Joel M Bowman,et al.  Ab initio potential energy and dipole moment surfaces for H5O2 +. , 2005, The Journal of chemical physics.

[67]  J. Behler,et al.  Fingerprints for spin-selection rules in the interaction dynamics of O2 at Al(111). , 2008, Physical review letters.

[68]  K. Ozawa,et al.  Oxidation of copper clusters on ZnO(101¯0): Effect of temperature and preadsorbed water , 2007 .

[69]  J. Behler,et al.  Pressure‐induced phase transitions in silicon studied by neural network‐based metadynamics simulations , 2008 .

[70]  Murray S. Daw,et al.  The embedded-atom method: a review of theory and applications , 1993 .

[71]  Thomas D. Kuhne,et al.  Ab initio quality neural-network potential for sodium , 2010, 1002.2879.

[72]  O. Dulub,et al.  STM study of Cu growth on the ZnO(1010) surface , 2002 .

[73]  H. Topsøe,et al.  On the nature of surface structural changes in Cu/ZnO methanol synthesis catalysts , 1999 .

[74]  J. Behler.  Neural network potential-energy surfaces in chemistry: a tool for large-scale simulations. , 2011, Physical chemistry chemical physics : PCCP.