1) Modelling and programming chemical problems (with AI support)
Step 0: Define a clear research question (in plain language).
I will provide several starter topics and help you refine the scope into a well-defined question and testable hypothesis.
Example
What problem are you interested in?
“Why is Pt commonly used for water electrolysis to generate H₂? Pt is expensive—why can’t we use other catalysts?”
What is your hypothesis?
“Maybe Pt binds hydrogen (H) particularly well.”
How will you test it?
“Compute hydrogen adsorption energies on different metal surfaces and compare them with Pt.”
Step 1: Write a model using equations.
This is the modelling part. For example:
- mass transport → Fick’s second law
- reaction rate → rate constant × concentration
- molecular energy → Schrödinger equation (conceptually)
Step 2: Translate the model into code.
This is the programming part. You will implement equations, initial conditions, and boundary conditions using MATLAB, Python, or established simulation tools. AI tools can help you learn a new language, turn equations into code, and debug more efficiently—making programming more approachable and enjoyable.
Step 3: Run, validate, and test your hypothesis.
You will check whether your scripts behave correctly, then use them to answer your research question.
2) Creating new materials and new operating methods
Step 0: Read key papers and identify the bottleneck.
You will learn how to locate and summarize relevant literature, then identify what limits the rate or selectivity. For example, in electrochemical hydrogenation (e.g., reducing CO₂ to CH₄ or CH₃OH), unwanted H₂ evolution often competes for electrons, lowering efficiency.
Step 1: Propose a new idea (materials or operating strategy).
Based on your understanding, you will design a new catalyst concept or operating approach. Bold ideas are welcome as long as they are scientifically reasoned—even if they are difficult to implement experimentally. For example, you might explore what happens if a reaction is run under rapid temperature oscillations rather than constant temperature.
Step 2: Use (and adapt) existing simulation workflows.
You will learn to use established scripts and modify them for your system.
Step 3: Evaluate performance.
Run simulations and compare predicted performance across materials and/or operating conditions.
3) Discovering new mechanisms and theories
Step 0: Identify what current models cannot explain.
For example, in hydrogen evolution (H⁺/H₂O + e⁻ → H* → H₂), the activity can depend not only on the catalyst and pH, but also on the electrolyte salt (added for conductivity). This suggests ions may influence the reaction, yet the mechanism is not always clearly explained in the literature.
Step 1: Propose a mechanism and incorporate it into a model.
You will form a hypothesis and connect it to existing pathways. For example, hydrogen evolution can proceed via two pathways:
- pathway A: H⁺ + e⁻ → H*; H* + H* → H₂
- pathway B: H⁺ + e⁻ → H*; H* + H⁺ + e⁻ → H₂
Because cations (e.g., Na⁺ in NaCl) can accumulate near a negatively charged electrode, you might hypothesize that different cations change surface site availability and shift the preferred pathway.
Step 2: Modify an existing simulation framework.
You will learn how earlier models were implemented, then extend them by adding a new variable, new adsorption terms, or new reaction steps.
Step 3: Test whether the new theory improves agreement with observations.
Compare model predictions with experimental trends and evaluate whether your added mechanism explains the “missing” phenomenon better than the original model.