PhD Studentship: Smart Crystals: Computer Vision and Machine Learning Assisted Industrial Crystallization Modeling and Control

Department: Chemical Engineering
Title: Smart Crystals: Computer Vision and Machine Learning Assisted Industrial Crystallization Modeling and Control
Application deadline: 31/01/2025

How to apply: please click on the 'Apply' button above.

No. of ads: one 

This 3.5 year PhD project is funded by AstraZeneca and includes a 3-month placement in AstraZeneca's facility in Macclesfield. The tuition fees will be paid and a tax free stipend based at the standard UKRI rate will be paid (£19,237 for 2024/25). The start date is October 2025. This is project is for home students only, i.e. UK nationals or Europeans with pre-settled status.

This AstraZeneca-funded project (Macclesfield, U.K.) addresses key challenges in automating crystallization processes, a crucial separation and purification technique widely used in producing fine chemicals, including pharmaceuticals. By leveraging state-of-the-art advances in computer vision and machine learning, we aim to manipulate Particle Size and Shape Distribution (PSSD) under industrially relevant conditions.

Crystallization processes typically yield powders with varying particle properties. Research has shown that equant-shaped crystals with narrow size distributions and minimal fine particles are ideal, as nonequant shapes (e.g., needles, platelets) complicate downstream processes like filtration, drying, and formulation. Recent breakthroughs in process monitoring and computational techniques, including foundational work by Dr. Rajagopalan’s group, have shifted the focus toward directly manipulating PSSD, especially for challenging elongated and plate-like particles, commonly encountered in the pharmaceutical and agrochemical sector.

A cornerstone of this project is DISCO, a stereoscopic imaging device enabling real-time PSSD monitoring during crystallization. This tool, combined with advanced population balance modeling aided by scientific machine learning, will enable real-time feedback control, achieving precise particle size and shape outcomes. These innovations will empower the broader crystallization community to overcome longstanding challenges in process automation, advancing the field toward Pharma 4.0 standards.

Applicants should have or expect to achieve a first-class honors degree in Chemical Engineering. 

Please contact Dr. Ashwin Kumar Rajagopalan with a cover letter and a copy of their CV at a.rajagopalan@manchester.ac.uk for informal enquiries. Applicants can also visit ash23win.github.io

for further information regarding Dr. Rajagopalan’s research group. 

If you have any queries regarding making an application, please contact our admissions team FSE.doctoralacademy.admissions@manchester.ac.uk

Advert information