Abstract
Carbon materials and their unique properties have been extensively studied by molecular dynamics, thanks to the wide range of available carbon bond order potentials (CBOPs). Recently, with the increase in popularity of machine learning (ML), potentials such as Gaussian approximation potential (GAP), trained using ML, can accurately predict results for carbon. However, selecting the right potential is crucial as each performs differently for different carbon allotropes, and these differences can lead to inaccurate results. This work compares the widely used CBOPs and the GAP-20 ML potential with density functional theory results, including lattice constants, cohesive energies, defect formation energies, van der Waals interactions, thermal stabilities, and mechanical properties for different carbon allotropes. We find that GAP-20 can more accurately predict the structure, defect properties, and formation energies for a variety of crystalline phase carbon compared to CBOPs. Importantly, GAP-20 can simulate the thermal stability of C60 and the fracture of carbon nanotubes and graphene accurately, where CBOPs struggle. However, similar to CBOPs, GAP-20 is unable to accurately account for van der Waals interactions. Despite this, we find that GAP-20 outperforms all CBOPs assessed here and is at present the most suitable potential for studying thermal and mechanical properties for pristine and defective carbon.
Highlights
Carbon is one of the most widely studied and important elements across many branches of science
carbon bond order potentials (CBOPs) (Tersoff, mo-Tersoff, adaptive intermolecular reactive bond order (AIREBO), moAIREBO, LCBOP, environment-dependent interatomic potential (EDIP), and ReaxFFC2013) and Gaussian approximation potential (GAP)-20 machine learning (ML)-IAP are tested with respect to lattice constants, cohesive energy, defect properties, thermal stability, van der Waals interaction, and mechanical properties for crystalline phase carbon materials
We find that all potentials can predict the crystal structure, but only GAP-20 is accurate within 95% of the DFTcalculated cohesive energy
Summary
Carbon is one of the most widely studied and important elements across many branches of science. Carbon bond order potentials (CBOPs) can simulate large-scale systems over time, by using empirical potentials, but with lower accuracy. In this work, we compare GAP-20 and widely used CBOPs (Tersoff, mo-Tersoff, AIREBO, mo-AIREBO, ReaxFFC2013, EDIP, and LCBOP) with DFT in terms of lattice constants, cohesive energies, defect properties, van der Waals (vdW) interaction, thermal stability, and mechanical properties for the carbon allotropes diamond, graphite, graphene, and CNTs. As a result of our comprehensive testing, we propose the most suitable potential for simulating crystalline phases of carbon. We select several widely used and representative potentials for testing, summarized in Table I. mo-AIREBO is the modified AIREBO where the C–C bond cutoff is set to 2.0 Å, and this form is mostly used to simulate the fracture behavior of carbon materials as mentioned above. A 10 ps long ab initio molecular dynamics simulation, with a time step of 1.0 fs, in the NVT ensemble was employed to test the thermal stability of C60
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
