MixSmart: A Framework for AI-Powered Binary Solvent Selection in Reaction and Process Design
Project ID: 2531bd1631
(You will need this ID for your application)
Research Theme: Engineering
Research Area(s):
Engineering
Manufacturing the Future
UCL Lead department: Chemical Engineering
Lead Supervisor: Lauren Ye Seol Lee
Project Summary:
Solvents are the hidden heroes of modern manufacturing. They control how reactions run, how products are separated, and how sustainable a process can be. Yet most research still focuses on single solvents, overlooking the vast potential of mixtures. This project aims to transform how binary solvents are designed by focusing on their effect on transition states—fleeting molecular configurations that govern reactivity and selectivity. You will build MixSmart, an AI-driven platform that integrates predictive models and optimisation tools to design solvent mixtures based on their influence on transition states. Here, you will:
- Benchmark solvent models to test their ability to describe transition-state stabilisation.
- Develop machine-learning approaches that capture mixture-specific effects and non-ideal interactions.
- Integrate these models into a design framework that can propose solvent blends tuned for reactivity, selectivity, and separation properties.
Depending on your interests, you may also explore process-level optimisation or contribute to experimental validation with robotic platforms at UCL East. By bridging molecular physics, machine learning, and chemical engineering, the project will pioneer new ways to use solvent mixtures as tools to shape transition states—and with them, the outcomes of chemical reactions.
Who you will be working with: You will be supervised by Dr Lauren Lee (AI-enabled molecular and reaction design) and Dr Marcello Sega (molecular simulation of complex fluids) at UCL Chemical Engineering Department. With strong connections to the Sargent Centre and industrial partners, you will receive interdisciplinary training in machine learning, computational chemistry, and process systems engineering, and have access to global experts and cutting-edge research facilities.
Who we are looking for: We seek a motivated student with a 2:1 MEng/MSc or equivalent in Chemical Engineering, Chemistry, Physics, or related disciplines. Experience in molecular modelling, optimisation, or machine learning is desirable. Above all, we value curiosity, independence, and enthusiasm for interdisciplinary research.