High Performance Computing Could Help Scale Up Graphene Production

Graphene has become one of the most interesting materials discovered by mankind. Creating this material involves great precision and cost, limiting the volume in which it can be manufactured - and a new approach using HPC could work around this problem.

A two-dimensional material, graphene sheets exist in atomically thin forms that are just one atom thick. However, the bond between carbon atoms makes it incredibly strong while being light and flexible. Structurally, graphene is an allotrope of carbon: it is basically the same as graphite (such as the lead in pencils) and diamonds, only that the carbon atoms are arranged differently. In this 2D material, carbon atoms are arranged in a hexagonal lattice arrangement.

ALSO READ : Ice Formation on Graphene Shows the Need for Heat

The Interest In Graphene and its Limits

The material has been the subject of multiple studies from different fields of study. Graphene has also been studied as a potential material for use in electronics, industrial tools and instruments, and biomedicine.

With its potential in different applications, there comes the challenge with graphene: producing it in commercially viable volumes, especially in the face of its increasing demand. Studies on the material suggest the use of liquid copper catalysts to encourage the "nucleation" or formation of graphene. This includes a 2013 study published in the journal Nano Letters and a 2013 study on Chemical Science. It has since created a fast and efficient way of producing 2D material. However, the mechanisms at work at the atomic level are not yet fully understood. This means that the procedure is not yet fully usable to reliably produce graphene in commercially feasible quantities.

Graphene's impressive qualities are derived from its perfectly uniform structure. The presence of imperfections in this honeycomb lattice of carbon atoms severely affects the material's qualities. In laboratory tests and studies where researchers only need small amounts of the material, they use an adhesive - like scotch tape - and "peel" layers of graphite away.

The Use of HPC for Graphene Simulations

Researchers often experiment with different atoms to see how they behave and interact with each other under varying conditions. Through advancements in technology have allowed for more controlled and more precise testing of these interactions, there are unknown variables that account for unexpected results. This led other researchers to employ computer simulations, which better controls the "environment" and show expected behavior based on a number of preprogrammed parameters.

"The problem describing anything like this is you need to apply molecular dynamics (MD) simulations to get the right sampling," says Mie Andersen, one of the authors of the new HPC study and the Chair for Theoretical Chemistry and Catalysis Research Center at TUM, in a university press release. "Then, of course, there is the system size-you need to have a large enough system to accurately simulate the behavior of the liquid."

With molecular dynamics simulations, researchers could observe individual events happening on an atomic scale from different angles. They can even pause the simulation to better examine events that occur in fractions of a second.

This led researchers to look into other potential methods for faster, reliable production methods. A team from the Technical University of Munich (TUM) in Germany used high-performance computing (HPC), with the JUWELS and SuperMUC-NG computers, at the Jülich Supercomputing Centre (JSC) and Leibniz Supercomputing Centre (LRZ). They ran high-resolution simulations of graphene formation on a liquid copper catalyst layer.

Details and results of their study are published in the report "Real-Time Multiscale Monitoring and Tailoring of Graphene Growth on Liquid Copper," appearing in the latest ACS Nano.