M A T E R I A L S F O R T O M O R R O W ' S I N N O V A T I V E A N D S U S T A I N A B L E I N D U S T R Y

S C I E N T I F I C H I G H L I G H T S

6 6 H I G H L I G H T S 2 0 2 3 I

Mapping the minutiae: Insights into the growth of graphene with liquid metal catalyst

Growing graphene on a molten metal surface may be the best way to get a high-quality, defect- free product in commodity quantities. X-ray techniques were combined with machine learning to characterise the interface between the 2D material and a liquid copper catalyst with unprecedented sub-angstrom precision.

Graphene is one of the strongest materials on earth, even more so than diamond. It has a wide range of applications in transportation, electronics and medication, owing to its unique mechanical, electric and optical properties, which can be attributed to its equally unique structure. The creation of a one-atom-layer thickness 2D graphene film was recognised with a Nobel Prize in 2010. Despite its widespread popularity and applications, its large-scale commercialisation has been hindered due to the defects formed while synthesising it.

Recently, liquid metal catalysts, such as copper (Cu), have been used to synthesise extremely high-quality graphene, almost without any defects. The liquid surface of the catalyst is perfectly flat, which is thought to be the reason behind graphene s highly crystalline structure. The fundamental problem with understanding liquid metal catalysts in detail is the microscopic processes occurring at the surface, which are not detectable by conventional experiments and computational approaches. This is because of the harsh conditions needed for graphene s growth and the dynamic liquid interface.

Set against this backdrop, molecular dynamics simulations based on machine-learning (ML) algorithms and synchrotron X-ray reflectivity were used to study this system with unprecedented accuracy. Studying the interface between liquid Cu and graphene is an important but challenging task that helps to understand the process of growing high-quality 2D materials. The researchers were interested in finding out if simulations assisted by machine learning could accurately predict the interface and match the results of synchrotron experiments. Both in-situ and in-silico techniques were employed to determine the height at which the graphene monolayer was adsorbed above liquid Cu.

The experimental results were obtained using in- situ synchrotron X-ray reflectivity at beamline ID10 (Figure 48a). The beamline is equipped with a dedicated reactor for studying liquid metals at high temperatures and vapour rates, allowing X-ray experiments combined with simultaneous radiation-mode optical microscopy and Raman microspectrometry. Moreover, the necessity to work with the curved liquid surface imposes an additional degree of experimental complexity, which can be overcome by developing the methodology and analysis methods to study such surfaces.

The team trained the ML potentials, precisely moment tensor potentials, using density functional theory reference data to build the computational model. This helped to obtain large-scale molecular dynamics simulations while leveraging the efficiency of ML algorithms. The team then