Machine learning hydrogen adsorption on nanoclusters through structural descriptors

Catalytic activity of the hydrogen evolution reaction on nanoclusters depends on diverse adsorption site structures. Machine learning reduces the cost for modelling those sites with the aid of descriptors. We analysed the performance of state-of-the-art structural descriptors Smooth Overlap of Atomic Positions, Many-Body Tensor Representation and Atom-Centered Symmetry Functions while predicting the hydrogen adsorption (free) energy on the surface of nanoclusters. The 2D-material molybdenum disulphide and the alloy copper–gold functioned as test systems. Potential energy scans of hydrogen on the cluster surfaces were conducted to compare the accuracy of the descriptors in kernel ridge regression. By having recourse to data sets of 91 molybdenum disulphide clusters and 24 copper–gold clusters, we found that the mean absolute error could be reduced by machine learning on different clusters simultaneously rather than separately. The adsorption energy was explained by the local descriptor Smooth Overlap of Atomic Positions, combining it with the global descriptor Many-Body Tensor Representation did not improve the overall accuracy. We concluded that fitting of potential energy surfaces could be reduced significantly by merging data from different nanoclusters.Machine learning: hydrogen adsorption energy on nanocluster surfacesThe accuracy and efficiency of descriptors for machine learning are tested for hydrogen evolution reaction in nanocatalytic systems. A team led by Adam Foster at Aalto University analysed the performance of SOAP, MBTR and ACSF structural descriptors used to gain insight to the free energy of hydrogen adsorption on the surface of nanoclusters relevant to catalysis applications. Using atomically thin MoS2 and AuxCuy metallic alloys as test systems, the accuracy of the predictors was evaluated when used as features in kernel ridge regression. Analysis on a dataset of 91 MoS2 clusters and 24 AuxCuy clusters (x + y = 13) indicated that while none of the descriptors which had been designed for molecules and crystals was optimized for nanoclusters, SOAP performed significantly better, making it a more suitable candidate for adsorption energy prediction.

[1]  Mark D. Symes,et al.  Earth-abundant catalysts for electrochemical and photoelectrochemical water splitting , 2017 .

[2]  S. Grimme,et al.  A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. , 2010, The Journal of chemical physics.

[3]  J. Behler.  Perspective: Machine learning potentials for atomistic simulations. , 2016, The Journal of chemical physics.

[4]  Gábor Csányi,et al.  Comparing molecules and solids across structural and alchemical space. , 2015, Physical chemistry chemical physics : PCCP.

[5]  Yadong Li,et al.  Special Topic: Catalysis—Facing the Future Heterogeneous catalysis for green chemistry based on nanocrystals , 2015 .

[6]  Stefan Grimme,et al.  Effect of the damping function in dispersion corrected density functional theory , 2011, J. Comput. Chem..

[7]  N. Lewis,et al.  Powering the planet: Chemical challenges in solar energy utilization , 2006, Proceedings of the National Academy of Sciences.

[8]  A. Krasheninnikov,et al.  Fabrication and atomic structure of size-selected, layered MoS2 clusters for catalysis. , 2014, Nanoscale.

[9]  Thomas Bligaard,et al.  Trends in the exchange current for hydrogen evolution , 2005 .

[10]  Matthias Rupp,et al.  Unified representation of molecules and crystals for machine learning , 2017, Mach. Learn. Sci. Technol..

[11]  Teter,et al.  Separable dual-space Gaussian pseudopotentials. , 1996, Physical review. B, Condensed matter.

[12]  Xiangming He,et al.  Shape control of CoO and LiCoO2 nanocrystals , 2010 .

[13]  Yadong Li,et al.  Catalysis based on nanocrystals with well-defined facets. , 2012, Angewandte Chemie.

[14]  George E. Dahl,et al.  Prediction Errors of Molecular Machine Learning Models Lower than Hybrid DFT Error. , 2017, Journal of chemical theory and computation.

[15]  Joost VandeVondele,et al.  Gaussian basis sets for accurate calculations on molecular systems in gas and condensed phases. , 2007, The Journal of chemical physics.

[16]  O. A. von Lilienfeld,et al.  Communication: Understanding molecular representations in machine learning: The role of uniqueness and target similarity. , 2016, The Journal of chemical physics.

[17]  M. Mavrikakis,et al.  Surface and subsurface hydrogen: adsorption properties on transition metals and near-surface alloys. , 2005, The journal of physical chemistry. B.

[18]  G. Watson,et al.  Atomistic models for CeO(2)(111), (110), and (100) nanoparticles, supported on yttrium-stabilized zirconia. , 2002, Journal of the American Chemical Society.

[19]  Xianfeng Ma,et al.  Orbitalwise Coordination Number for Predicting Adsorption Properties of Metal Nanocatalysts. , 2017, Physical review letters.

[20]  Klaus-Robert Müller,et al.  Assessment and Validation of Machine Learning Methods for Predicting Molecular Atomization Energies. , 2013, Journal of chemical theory and computation.

[21]  Zachary W. Ulissi,et al.  To address surface reaction network complexity using scaling relations machine learning and DFT calculations , 2017, Nature Communications.

[22]  R. Parsons.  The rate of electrolytic hydrogen evolution and the heat of adsorption of hydrogen , 1958 .

[23]  Michael K.H. Leung,et al.  Engineering stepped edge surface structures of MoS2 sheet stacks to accelerate the hydrogen evolution reaction , 2017 .

[24]  Joost VandeVondele,et al.  cp2k: atomistic simulations of condensed matter systems , 2014 .

[25]  I. Chorkendorff,et al.  Biomimetic Hydrogen Evolution: MoS2 Nanoparticles as Catalyst for Hydrogen Evolution , 2005 .

[26]  James R. McKone,et al.  Solar water splitting cells. , 2010, Chemical reviews.

[27]  Christopher J. Tassone,et al.  FROM SYNTHESIS TO PROPERTIES AND APPLICATIONS , 2013 .

[28]  Zheng Li,et al.  Feature engineering of machine-learning chemisorption models for catalyst design , 2017 .

[29]  Hua Zhang,et al.  Thin metal nanostructures: synthesis, properties and applications , 2014, Chemical science.

[30]  K. Müller,et al.  Fast and accurate modeling of molecular atomization energies with machine learning. , 2011, Physical review letters.

[31]  Colin F. Dickens,et al.  Combining theory and experiment in electrocatalysis: Insights into materials design , 2017, Science.

[32]  G. Fu,et al.  Synthesis and electrocatalytic activity of Au@Pd core-shell nanothorns for the oxygen reduction reaction , 2014, Nano Research.

[33]  J. Nørskov,et al.  Towards the computational design of solid catalysts. , 2009, Nature chemistry.

[34]  Yadong Li,et al.  Size and shape control of LiFePO4 nanocrystals for better lithium ion battery cathode materials , 2013, Nano Research.

[35]  Zhi-cheng Zhang,et al.  Engineering nanointerfaces for nanocatalysis. , 2014, Chemical Society reviews.

[36]  Koji Tsuda,et al.  Machine-learning prediction of the d-band center for metals and bimetals , 2016 .

[37]  D. Goodman,et al.  Onset of catalytic activity of gold clusters on titania with the appearance of nonmetallic properties , 1998, Science.

[38]  A. Tuxen,et al.  Structure and electronic properties of in situ synthesized single-layer MoS2 on a gold surface. , 2014, ACS nano.

[39]  Luke E K Achenie,et al.  Machine-Learning-Augmented Chemisorption Model for CO2 Electroreduction Catalyst Screening. , 2015, The journal of physical chemistry letters.

[40]  Jacob Bonde,et al.  Biomimetic hydrogen evolution: MoS2 nanoparticles as catalyst for hydrogen evolution. , 2005, Journal of the American Chemical Society.

[41]  Matthias Krack,et al.  Pseudopotentials for H to Kr optimized for gradient-corrected exchange-correlation functionals , 2005 .

[42]  Luke E K Achenie,et al.  High-throughput screening of bimetallic catalysts enabled by machine learning , 2017 .

[43]  Qiang Fu,et al.  Understanding nano effects in catalysis , 2015 .

[44]  R. Kondor,et al.  On representing chemical environments , 2012, 1209.3140.

[45]  R. Bratschitsch,et al.  Assignment of the NV 0 575-nm zero-phonon line in diamond to a 2 E- 2 A 2 transition , 2013, 1301.3542.

[46]  B. Hammer,et al.  In Situ Detection of Active Edge Sites in Single-Layer MoS2 Catalysts. , 2015, ACS nano.

[47]  J. Vybíral,et al.  Big data of materials science: critical role of the descriptor. , 2014, Physical review letters.

[48]  Felix A Faber,et al.  Crystal structure representations for machine learning models of formation energies , 2015, 1503.07406.

[49]  Noam Bernstein,et al.  Machine learning unifies the modeling of materials and molecules , 2017, Science Advances.

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

[51]  Bryan T. Yonemoto,et al.  Highly porous non-precious bimetallic electrocatalysts for efficient hydrogen evolution , 2015, Nature Communications.

[52]  S. Goedecker,et al.  Relativistic separable dual-space Gaussian pseudopotentials from H to Rn , 1998, cond-mat/9803286.

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

[54]  F. Besenbacher,et al.  MoS2 nanoparticle morphologies in hydrodesulfurization catalysis studied by scanning tunneling microscopy , 2013 .