(501c) Computational Modeling of Aluminum Covetics for Improving Manufacturing Yields | AIChE

(501c) Computational Modeling of Aluminum Covetics for Improving Manufacturing Yields


Lustig, S., Northeastern University
Introduction: Manufacturing of a unique covalently bonded graphene-aluminum composite, referred to as aluminum covetics, are inhibited by inhomogeneous distribution of carbon and carbide formation within liquid aluminum; this has resulted in poor manufacturing yields of aluminum covetics. A significant drawback to fabrication is the limited understanding of graphene and graphene-aluminum formation in liquid aluminum. Therefore, the objective of this project is to determine the reaction mechanism pathway for the formation of graphene-aluminum covetics. The ultimate goal is to apply thermodynamic and kinetic properties, ascertained from ab initio molecular dynamic (AIMD) and density functional theory (DFT), to design and fabricate an efficient reactor for improved manufacturing yields of covetics.

Methods: To tackle this challenge, we optimize carbon-aluminum species using a COductor like Screening MOdel (COSMO) in conjunction with DFT, in the presence of an electric field and added charges. Thermochemical analysis is used to calculate thermodynamic properties of the various species. These thermodynamic properties are verified for accuracy by comparing to the thermodynamic properties of aluminum-carbon species formed from explicit AIMD simulations. AIMD simulations are performed with 64 atom aluminum liquid, at 1100°C, and various carbon concentrations (C1-C6), with periodic boundary conditions. The two-phase thermodynamic (2PT) method is used to calculate thermochemical properties from AIMD and the radial distribution function (RDF) is used to calculate the bond lengths. The optimized carbon-aluminum species are then used as building blocks to generate larger graphene-aluminum structures. The reaction pathway is then generated from the base case of AlXCX to graphene-aluminum covetics using transition state theory to ascertain the intermediate species.

Results: Gibbs energies between AIMD and COSMO-DFT optimized species vary from 0.46% to 3.57%. Our results suggest that COSMO-DFT can accurately mimic AIMD derived thermodynamic properties of aluminum covetics to complement AIMD simulations. Diffusion of carbon-aluminum species increase with increasing external potential, and further increase in the presence of added electrons. The added electrons decrease molecular size of carbon-aluminum species, leading to increased diffusion. Finally, a reaction mechanism pathway and reaction rates are presented from Al4C1 to graphene-aluminum covetic formation.