A5093427169- Machine Learning in Materials Science
- nanoparticles nucleation surface interactions
- Chemical Thermodynamics and Molecular Structure
- Thermal and Kinetic Analysis
- Material Dynamics and Properties
- Analytical Chemistry and Sensors
- Methane Hydrates and Related Phenomena
- thermodynamics and calorimetric analyses
- Scientific Computing and Data Management
- Advanced NMR Techniques and Applications
- Hydrogen Storage and Materials
- Electrochemical Analysis and Applications
- Electrostatics and Colloid Interactions
- Radioactive element chemistry and processing
- Graphene research and applications
- Electrohydrodynamics and Fluid Dynamics
- Electrochemical sensors and biosensors
- Icing and De-icing Technologies
- Quantum, superfluid, helium dynamics
- Spectroscopy and Quantum Chemical Studies
University of Cambridge
2024
Thomas Young Centre
2023
University College London
2023
Machine-learned force fields have transformed the atomistic modelling of materials by enabling simulations ab initio quality on unprecedented time and length scales. However, they are currently limited by: (i) significant computational human effort that must go into development validation potentials for each particular system interest; (ii) a general lack transferability from one chemical to next. Here, using state-of-the-art MACE architecture we introduce single general-purpose ML model,...
Abstract The Bernal-Fowler ice rules stipulate that each water molecule in an crystal should form four hydrogen bonds. However, extreme or constrained conditions, the arrangement of molecules deviates from conventional rules, resulting properties significantly different bulk water. In this study, we employ machine learning-driven first-principles simulations to identify a new stabilization mechanism nanoconfined phases. Instead forming bonds, crystalline can quasi-one-dimensional...
The efficient calculation of nucleation collective variables (CVs) is one the main limitations to application enhanced sampling methods investigation processes in realistic environments. Here we discuss development a graph-based model for approximation CVs that enables orders-of-magnitude gains computational efficiency on-the-fly evaluation CVs. By performing simulations on nucleating colloidal system mimicking multistep process from solution, assess model's both postprocessing and biasing...
Malina Seyffertitz opened a general discussion of the paper by Adam P. Willard: Your simulations are based on flat electrode geometry. How difficult would it be to introduce curved surfaces, i.e. porous electrodes? Willard responded: It is certainly not impossible si
Calculating sublimation enthalpies of molecular crystal polymorphs is relevant to a wide range technological applications. However, predicting these quantities at first-principles accuracy -- even with the aid machine learning potentials challenge that requires sub-kJ/mol in potential energy surface and finite-temperature sampling. We present an accurate data-efficient protocol based on fine-tuning foundational MACE-MP-0 model showcase its capabilities physical properties ice polymorphs. Our...
We present an accurate and data-efficient protocol for fine-tuning the MACE-MP-0 foundational model a given system. Our achieves kJ/mol in predicting sublimation enthalpies below 1% error density of ice polymorphs.
Water's ability to autoionize into hydroxide and hydronium ions profoundly influences surface properties, rendering interfaces either basic or acidic. While it is well-established that the water-air interface acidic, a critical knowledge gap exists in technologically relevant surfaces like graphene-water interface. Here we use machine learning-based simulations with first-principles accuracy unravel behavior of at Our findings reveal ion predominantly residing first contact layer water. In...
Recent work has suggested that nanoconfined water may exhibit superionic proton transport at lower temperatures and pressures than bulk water. Using first-principles-level simulations, we study the role of nuclear quantum effects in inducing this superionicity We show increase ionic conductivity hexatic water, leading to behaviour previously thought possible. Our suggests be accessible graphene nanocapillary experiments.
Paul Ryan opened the discussion of introductory Spiers Memorial Lecture by Richard J. Saykally communicating: You mentioned importance contaminants. In this case are your spectroscopy measurements done under ambient conditions where contaminants known to play a significant rol
The efficient calculation of nucleation collective variables (CVs) is one the main limitations to application enhanced sampling methods investigation processes in realistic environments. Here we discuss development a graph-based model for approximation CVs, which enables orders-of-magnitude gains computational efficiency on-the-fly evaluation CVs. By performing simulations on nucleating colloidal system mimicking multistep process from solution, assess model's both postprocessing and biasing...
The efficient calculation of nucleation collective variables (CVs) is one the main limitations to ap- plication enhanced sampling methods investigation processes in realistic environments. Here we discuss development a graph-based model for approximation CVs, which en- ables orders-of-magnitude gains computational efficiency on-the-fly evaluation CVs. By performing simulations on nucleating colloidal system, assess model’s both postprocessing and biasing trajectories, thereby mimicking...
The efficient calculation of nucleation collective variables (CVs) is one the main limitations to application enhanced sampling methods investigation processes in realistic environments. Here we discuss development a graph-based model for approximation CVs, which enables orders-of-magnitude gains computational efficiency on-the-fly evaluation CVs. By performing simulations on nucleating colloidal system mimicking multistep process from solution, assess model's both postprocessing and biasing...
Bernal-Fowler ice rules govern the phase behaviors of crystalline bulk water by stipulating that each molecule forms four hydrogen bonds. However, in extreme or constrained conditions, arrangement molecules deviates from conventional rules, resulting properties significantly different water. In this study, we employ machine learning-driven first-principles simulations to observe a unique violation monolayer confined within hydrophobic channel. We quasi-one-dimensional hydrogen-bonded...