![Sample our Physical Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110819/TOC-Subject-Sample-2021-Physical-Sciences.jpg"})
Abstract
Towards understanding the effect of solid supports on the catalytic activity of supported metal nanoparticles, all-atom molecular dynamics (MD) simulations are often conducted. However, these calculations are hampered by the uncertainty related to describing metal–support interactions (typically described as Lennard-Jones (LJ) potentials) at the atomic length scale. Ab initio electron-structure calculations are expected to refine such calculations by providing better estimates for the metal–support pair interaction potential. In the case of platinum nanoparticles supported on graphite, recent ab initio results suggest the correct energetic LJ parameter should be about four times that used in previous simulation studies from our group, as well as from others. Stimulated by these findings, MD simulations have been used here to investigate the effect of the magnitude of the metal–carbon interaction on the structure of supported metal nanoparticles. The LJ potential was used to model the metal–carbon interactions, and the embedded-atom method was used to model the metal–metal interactions. The morphology of platinum nanoparticles of 130, 249 and 498 atoms supported on graphite and various bundles of carbon nanotubes (CNTs) was studied. For the larger nanoparticle it was found that, although the details of platinum–carbon interactions are important for correctly capturing the morphological details, the morphology of the support is the primary factor that determines such features. Platinum–carbon interactions affect more significantly the results obtained for metal nanoparticles supported by CNT bundles. In this case, we found that the deviations become significant for small supported nanoparticles, as well as for nanoparticles of any size supported on CNTs of small diameter.
Keywords:
Acknowledgements
The authors acknowledge financial support from the US DOE under contract #FG02-06ER64239, and from the Vice President for Research at the University of Oklahoma through a Junior Faculty Research Program award. Generous allocations of computing time were provided by the OU Supercomputing Center for Education and Research (OSCER) at the University of Oklahoma and by the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory.