Dr. Peter J. Hatton (Pete)
Computational Materials Science
and Advanced Scientific Computing
I am an accomplished materials scientist with extensive and varied experience in
computational modeling of materials, complemented by a strong track record of
publishing high-quality science. My strengths lie in my robust mathematical and
computational background as well as my passion for conducting impactful,
cutting-edge science focused on sustainable and environmentally conscious applications.
My research focuses on using quantum, classical, and long-timescale atomistic
modelling techniques to study defect transport in chemically and structurally
complex materials. I work primarily within multidisciplinary and multiscale
research teams to integrate atomistic modelling, higher length-scale thermodynamic
models and experimental observations. Generally, my goal is to use these techniques
alongside high-performance computing resources to understand the microstructural
evolution of materials in extreme environments and predict their macroscopically
observable features.
Plasma-Material Interactions (PMI)
The accumulation of gas atoms in a fusion-reactors divertor material is a topic of great and
long-standing interest to the plasma-facing materials community. I use state-of-the-art
atomistic modelling techniques alongside a multi-scale modelling framework
to elucidate complex mechanisms which can occur in these environments.
Defect Transport in High/Medium Entropy Alloys (HEAs)
The chemical complexity intrinsic to HEAs can lead to large variations in local
environments and therefore defect transport.
Here we integrate atomistics, machine learning and a KMC algorithm in a closed loop
that automatically generates and refines defect
diffusion tensors.
Defect Transport in Chemically Complex Oxides
Many complex oxides have the ability to accommodate
a large proportion of cation antisites before amorphising leading to exceptional
radiation damage tolerence.
Using atomistic modelling techniques I study how the chemical complexity of
these materials in radiation damage environments manifest within their microstructure.
My work has outlined how these features may impact their application
to the nuclear materials and alternative battery material communities.
CdTe Thin Film Solar Cells
Cadmium Telluride (CdTe)-based solar cells have poor efficiency unless treated with Chlorine (Cl),
however, this treatment leads to surface rupture and failure in cells grown with magnetron sputtering.
The atomistic mechanisms behind these experimental processes remained a mystery for ~30 years.
A high-throughput Molecular Dynamics workflow for generating atomistic
databases of defect transport data in chemically complex materials.
Developed at Los Alamos National Lab (LANL) as part of a massively-parallel
automated workflow for approximating diffusion tensors of defects in Medium and
High Entropy Alloys with a focus on exploiting exascale computing resources.
Hop-Decorate is Open-Source under the BSD-3 License.
These are a sample of my personal interests and projects which I
contribute to in my free time.
Batching Molecular Dynamics Simulations for Increased Efficiency on NVIDIA GPUs
Atomistic simulations with small computational footprints pose a serious problem
for efficiency when executed on GPUs - as is becoming the norm. I have proposed
a method of batching simulations together to increase combined efficiency and
throughput on High-Performance Computers which feature NVIDIA A100 GPUs.
python-BlackJack (pyBJ)
A large scale time-dependent BlackJack simulator for evaluating
static decision and betting strategies. This code is part of an ongoing
personal project to dispell commonly held superstition and false
intuition about games of chance through the use of rigorous theory
and simulation - something that is severly lacking in the community.