Two self-driving labs at the Acceleration Consortium (AC) are transforming the way scientists purify chemicals, thanks to new funding from the Natural Sciences and Engineering Research Council of Canada’s Alliance program. This brings the project’s total support to over $1 million, alongside its partnership with Bristol Myers Squibb—as part of the AC’s industry-sponsored collaborations.

The funded project focuses on automating liquid-liquid extraction (LLE) — a crucial step in chemical synthesis that can impact various industries. LLE is used to separate and purify compounds based on differences in their partitioning between two liquid phases. It is widely utilized in pharmaceutical development, food and fragrance production, chemical processing to recover valuable materials, and environmental remediation to remove waste and contaminants. Despite its wide application, this purification process is often a slow, repetitive, and labour-intensive step that can delay scientific progress. Designing and optimizing the extraction conditions require extensive trial-and-error experimentation, as it is usually challenged by issues such as emulsion formation, stable dispersions, and rag layers that slow down separation and scale-up. A more rapid method of optimizing chemical synthesis can dramatically reduce the time and cost of bringing critical innovations to market — including life-saving drugs.

By building an autonomous LLE workflow, the research teams are aiming to take the guesswork out of purification. The teams are building and evolving a new system that can autonomously test, adapt, and find the best conditions for producing high-quality results while minimizing the need for manual intervention. This new system can greatly help scientists speed up chemical synthesis and better understand how to scale up chemical processes.

Since the start of the project in 2024, the teams have achieved several milestones to bring LLE automation to life. By constructing innovative hardware in-house, and combining it with industry standard equipment, they have set up proof-of-concept LLE automation platforms in two of Canada’s top research universities—the University of Toronto (U of T) and the University of British Columbia (UBC).

These two automated systems have parallel layouts—both can execute the entire extraction workflow, from process characterization to optimization. Researchers can monitor and analyze the process in real time through its computer vision system, and send instructions via IvoryOS, a user-friendly web interface powered by large language models.

To be able to fit into different lab spaces at two sites and handle various models of instrumentation, the system was designed with a modular approach for flexibility. Each unit of operation can be positioned separately, while connected by a robotic arm that can transfer the sample to subsequent workflow step.

‍

‍

The system is greatly reducing human intervention in LLE operation, with capabilities to help chemists evaluate extraction conditions and drive autonomous decision making. The teams are now working to validate the platform with diverse LLE scenarios, taking a step closer to autonomous optimization.

“The new platform shifts the role of a chemist from hands-on execution to oversight and innovation, who now also needs to align efficiency with cost reduction and resource saving. This new funding allows us to not only accelerate automation innovation, but also train the next generation of scientists,” said Dr. Jason Hein, Principal Investigator at the AC’s scale-up self-driving lab, and Professor at Department of Chemistry, UBC.

“Bringing together our academic expertise with Bristol Myers Squibb’s industry insight, we’re accelerating autonomous chemistry and preparing young talents for future careers in science and industry.” With the combined funding, nine staff scientists and trainees are gaining hands-on experience with cutting-edge automation and data-driven process design.

“Ultimately, the project is transforming how scientists approach and solve challenges in process research and development, presenting a new way forward for chemical research—one where self-driving labs handle routine experiments, giving scientists more time to focus on big-picture breakthroughs,” says Dr. Sophie Rousseaux, Principal Investigator at the AC’s organic self-driving lab and Associate Professor at Department of Chemistry, U of T.