BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Exploring amorphous graphene with machine-learned atomic energies 
 - Zakariya El-Machachi\, Inorganic Chemistry Laboratory\, University of Ox
 ford\, 
DTSTART:20230522T130000Z
DTEND:20230522T133000Z
UID:TALK201250@talks.cam.ac.uk
CONTACT:Dr M. Simoncelli
DESCRIPTION:While experimental realization of crystalline two–dimensiona
 l materials has been common in recent years\, there are limited examples o
 f amorphous counterparts. Amorphous graphene\, a two-dimensional extended 
 carbon system with disorder\, is a prototype for studying 2D disorder due 
 to its complex and rich configurational space\, which is not yet fully und
 erstood.\nHere we report on an atomistic modelling study of amorphous grap
 hene using a machine learning (ML) based force field. ML force fields are 
 typically “trained” on data from computationally expensive density fun
 ctional theory (DFT) calculations but can achieve near DFT accuracy with s
 ignificantly reduced computational cost. One key assumption in many of the
 se methods is that the global energy can be separated into sums of local c
 ontributions. The extent to which these “machine-learned” local energi
 es are physically meaningful is an interesting research question.\nWe crea
 te structural models by introducing defects into ordered graphene through 
 Monte-Carlo bond switching\, defining acceptance criteria using the machin
 e-learned local\, atomic energies associated with a defect\, as well as th
 e nearest-neighbour (NN) environments.\nWe find that physically meaningful
  structural models arise from ML atomic energies\, ranging from continuous
  random networks to paracrystalline structures. Our results show that ML a
 tomic energies can be used to guide Monte-Carlo structural searches in pri
 nciple\, and that their predictions of local stability can be linked to sh
 ort- and medium-range order in amorphous graphene. We expect that the form
 er point will be relevant more generally to the study of amorphous materia
 ls\, and that the latter has wider implications for the interpretation of 
 ML potential models.\n\nAcknowledgements:\nThe authors would like to ackno
 wledge the use of the University of Oxford Advanced Research Computing (AR
 C) facility in carrying out this work. This work was supported by the Engi
 neering and Physical Sciences Research Council [grant number EP/L015722/1]
 .\n\n
LOCATION:Cavendish Laboratory\, Small Lecture Theatre
END:VEVENT
END:VCALENDAR