An odyssey begins with a question, long before the next shore has a name.
I set sail with curiosity, without knowing where every question will lead. In the search for truth, I am willing to change course—to follow unexpected evidence, learn another discipline’s language, or find a fellow traveller whose questions reshape my own.
These are the crossings still unfolding: ideas to test, paths to discover, and shores not yet on the map.
Beyond Ithaca
01 · Chemistry & catalysis
World (Action) Models for Chemistry/Catalysis
Towards a mechanism-guided, self-driving laboratory.
I am exploring models that represent a reacting system and predict how it responds to intervention, with catalysis as a starting point. The aim is to connect mechanistic kinetic Monte Carlo, model-guided decisions, and experimental feedback in a closed information loop.
A model proposes a calculation or experiment; its outcome updates the model’s account of the chemistry and informs the next action.
The research direction
The world model represents the chemical state and its evolution; the action model predicts the consequences of changing conditions or choosing an experiment. Mechanistic KMC would connect candidate reaction pathways to observable kinetics and product distributions, helping test which explanations remain consistent with the results.
The first step is to test whether this loop can distinguish competing mechanisms in simulation, before testing it in a physical laboratory.
02
Ontologies for Evidence-linked Kinetic Modelling
Advised by Prof. Aidong Yang.
Can every modelling choice be traced back to the evidence behind it?
This direction connects ontologies with kinetic modelling: making the relationships between chemical entities, reaction and transport events, parameters, observations, and evidence explicit.
Follow this thread
I am interested in using this structure to support the construction of executable kinetic models, distinguish assumptions from calculated or measured evidence, and identify which missing information is worth obtaining next. Polymer upcycling and diffusion-aware modelling provide a motivating context.
03
From Polymer Reaction Engineering to Fast Kinetic Monte Carlo
Following the shape of a reaction
How can a fast simulation retain the molecular structure that makes each reaction matter?
With Prof. Kim B. McAuley and Prof. Pieter Iedema, I am working on connecting Fast KMC with explicit polymer topology. The project translates chemical knowledge from polymer reaction engineering into computational rules: which reactions can occur, how they change molecular connectivity, and how those changes can be simulated efficiently.
Follow this thread
The distinction matters: chain length and branch counts summarise a molecule, while topology records how its parts are connected. We are working towards a framework in which reaction kinetics and structural changes remain consistent, with validation based on molecular-weight and branching statistics.