Access required
Please enter the password to view the project descriptions.
Incorrect password. Please try again.
Programming or AI-assisted coding experience is helpful but not required. If you’re interested, you can start learning and practicing programming and computational chemistry through this project while applying it to real catalytic research questions.
具備程式撰寫經驗（包含使用 AI 輔助工具或所謂的「vibe coding」）也許會對開展下述研究有幫助，但這並非必要條件。如果你有興趣，這個專題也可以成為你開始學習程式設計與計算化學、並將其應用在真實催化工程問題上的起點。我們會提供循序漸進的指導，讓你在研究過程中逐步培養自己的能力。
Direction 1: Catalyst Performance and Reaction Kinetics
Introduction
How do catalyst properties and operating conditions control reaction rates and selectivity? Hydrogenation and dehydrogenation reactions play a central role in chemical engineering and materials processing, from energy storage and fuel production to petrochemicals and fine chemicals. In practice, these reactions operate under high temperatures and pressures, and their performance depends strongly on catalyst materials, reactor conditions, and surface chemistry.
Despite decades of industrial application, many observed behaviors, such as changes in reaction order, apparent activation energy, or selectivity, cannot be explained by simple rate equations. For example, in the dehydrogenation of methylcyclohexane (C₇H₁₄ → C₇H₈ + H₂), different catalysts show dramatically different kinetic trends under similar conditions. Understanding why requires going beyond global rate expressions and analyzing how elementary surface reactions, adsorption, and kinetics interact.
In this project, we use microkinetic modeling to connect material properties (surface composition, active sites) with engineering variables (temperature, pressure, surface coverage). The goal is to build mechanistic insight that can guide catalyst design and process optimization.
Depending on your background and interests, you may work on some of the following tasks:
The following examples illustrate typical activities in this project. You are not expected to complete all of them.
- Read engineering- and materials-oriented literature on reaction kinetics, catalytic mechanisms, and microkinetic models.
- Learn basic MATLAB programming and apply microkinetic simulations to catalytic reaction networks (example codes provided).
- Analyze experimental kinetic data, propose plausible reaction mechanisms, and evaluate how catalyst properties and operating conditions affect performance.
Reaction systems to start with (choose one)
- Thermochemical methylcyclohexane or cyclohexane dehydrogenation
- Thermochemical toluene or benzene hydrogenation
- Electrochemical hydrogenation of phenol
- Electrochemical proton reduction (hydrogen evolution reaction)
Direction 2: Machine-Learning-Accelerated Quantum Chemistry for Catalytic Materials
Introduction
How do atomic-scale structure and environment determine catalytic behavior? Catalyst performance ultimately originates from atomic-scale interactions between reactants and material surfaces. Quantum chemistry calculations, especially Density Functional Theory (DFT), are widely used to compute adsorption energies, reaction barriers, and surface stability. However, realistic catalytic systems in chemical and materials engineering often involve large unit cells, high surface coverage, and solid–liquid interfaces, making conventional DFT calculations prohibitively expensive.
Recent advances in machine learning (ML) provide a promising solution. ML-based surrogate models can learn from quantum chemistry data and rapidly predict energies and forces, enabling the study of systems closer to real operating conditions. Frameworks such as FairChem allow researchers to explore how surface coverage, co-adsorbates, and solvent environments influence reaction energetics.
This project aims to introduce you to materials modeling at scale, emphasizing how quantum chemistry computation and machine learning can support rational catalyst and process design.
Depending on your background and interests, you may work on some of the following tasks:
The following examples illustrate typical activities in this project. You are not expected to complete all of them.
- Study the fundamentals of DFT, adsorption energetics, and reaction barriers, along with machine-learning-based surrogate models.
- Learn practical skills in Linux, Python, and materials modeling tools such as FairChem and Atomic Simulation Environment (ASE).
- Build atomic models of catalyst materials and analyze trends in coverage effects, solvent effects, and structure–property relationships.
Reaction systems to start with (choose one)
- Aqueous electrochemical proton reduction (hydrogen evolution reaction) on metal surfaces
- H* + H* → H₂ on metal surfaces with high coverage of H*, CO*, or CH* (* indicates species adsorbed on the catalyst surface)
- Methylcyclohexane dehydrogenation on metal surfaces under high surface coverage of H*, C₇H₁₄*, or toluene (C₇H₈*)