“Sustainability is the only thing worth pursuing,” says chemical synthesis innovator

As a co-founder of Chemical Data Intelligence (CDI) opens in new tab/window and Professor of Sustainable Reaction Engineering at the University of Cambridge opens in new tab/window, Prof Alexei Lapkin opens in new tab/window is transforming how scientists approach chemical synthesis. We spoke with him about his innovative technology and vision for a more sustainable chemical industry. Responses were edited for clarity and concision.

In simple terms, what does your work at Chemical Data Intelligence involve?

We’re developing tools to help scientists creating new medicines or other functional molecules to work more efficiently and sustainably. Chemistry is incredibly complex and has many factors to consider. Our tools organize and present this information systematically so scientists have all the facts available when making decisions. Essentially, we’re making it easier to develop advanced molecules faster and with cleaner and less harmful processes to the environment.

Can you expand on the role of sustainability in this work?

Sustainability is at the core of everything we do. We focus on helping chemists develop cleaner syntheses, more efficient manufacturing processes, and prevent pollution. In today’s world, for me personally, sustainability is the only thing worth pursuing.

Computer-aided synthesis planning (CASP) has a long history. What does CDI-CASP bring to the workbench that’s different?

CDI-CASP accelerates and expands the process using data science, chemoinformatics, machine learning and physical models-based predictive tools. It can perform retrosynthesis (planning how to make a specific molecule) and forward synthesis (exploring what can be made from a particular reactant) and help explore alternative starting materials, such as bio-based feedstocks or carbon dioxide. Our tool can also evaluate different synthesis plans using metrics like overall yield and environmental impact, and it provides chemists with a ranked list of possible routes to choose from.

A recent white paper opens in new tab/window has received considerable attention. It offers a case study of how CDI-CASP has improved a chemical process. Can you explain why it’s important?

We recently worked with Shionogi opens in new tab/window pharmaceutical company on a case study of S-Zanubrutinib, a drug used for lymphoma treatment. It highlights how using Reaxys data with our technology makes it possible to have a computer-aided green synthesis tool. The original synthesis route filed in a patent consisted of 12 steps and involved many hazardous molecules. Using our CDI-CASP tools, we achieved three key improvements: reducing the number of synthesis steps, finding routes with higher overall yield, and avoiding hazardous reagents, solvents and intermediates. The result was a shorter, greener and more efficient synthesis route.

How do you convince chemists to adopt your tools and change their workflows?

Chemists are a rather shrewd audience and need to see concrete results. They recognize the value when we show them what questions we can answer and the chemistry insights our tools generate. For example, the benefit is evident when we demonstrate how our synthesis planning tools can help them accomplish a task in 20 minutes, which might take two weeks of manual work. It’s about enabling them to be more productive, avoid biases of habit, and focus on more exciting work rather than tedious manual tasks.

You have a varied background. Did that help in preparing you for what you are doing today?

Certainly. My undergraduate university education was comprehensive. As a chemistry student, I studied mathematics and physics intensively alongside biology and chemistry. This breadth helped me work with mathematical tools and collaborate with mathematicians and computer scientists.

Later, I worked in different areas. My research group developed concepts in intensifying chemical processes, some of which have gone into industry. We were among the first groups to work on machine learning applications in chemical process engineering — though we didn’t call it that then. We just did these interesting things with mathematics, which grew into a significant area. At the time, we were having fun. Now, it’s probably the way many sciences will be done in the future.

What inspired you to be a CDI cofounder and enter the business world?

I already had a business when I was a student: I ran a student magazine for a couple of years — let’s say it was not a runaway success [laughter]. But it gave me some ideas of what business involves: working with all these different people — investors, contractors, manufacturers, etc. — and bringing their skills and ideas together.

But then, academic life seduced me by offering independence in creating new things. However, after developing several ideas and seeing them in publications, I wanted also to see them work in practice. I had a PhD student and a postdoc interested in co-founding a company with me, which gave me the courage to move forward. I wouldn’t have done it alone, but with a team, it felt possible.

How did your collaboration with Elsevier and Reaxys begin?

It started over 10 years ago when I was looking for a large dataset to explore hidden patterns in chemistry data. We became one of the first research groups to gain access to Reaxys data, which was the beginning of the Elsevier Research Network — a very inspiring group to share ideas with. After working with Elsevier data through several PhD students and postdocs, we developed a portfolio of potential products. Elsevier encouraged us to set up a company, making it possible through legal agreements and support.

What are some of the other sustainability tools you’re developing for different clients?

We’re building several tools focused on environmental aspects. One is a complex tool to predict life cycle assessment impacts in pharmaceutical manufacturing, funded by a consortium of pharma companies. It will enable rapid evaluation of different production routes and their environmental impacts. We’re also developing tools to use AI to predict new performance chemicals with considerations for ease of manufacturing, environmental sustainability and biodegradability. We have also built our own “Dry Lab” — an environment for our scientists to work in.

What’s next for CDI?

The following year or two are crucial as we launch several products. We need to scale our operations. This includes building teams for maintenance and support, which is a new challenge for us. Until now, we've grown organically without external investment, expanding from three people to six, with more hires coming.

Transitioning from contract R&D to creating software-as-a-service products involves market adoption and pricing uncertainty. Managing this transition while ensuring financial stability is our current focus – and causing me the most sleep loss [laughter]. These are new challenges for us, but we’ve found very good people to work with CDI and drive key areas of development. Despite these challenges, we’re excited about bringing our sustainability-focused tools to market.

What do you wish people understood better about AI in chemistry?

There’s a lot of hype around AI and some false promises. This happens with every emerging technology — we’ve seen similar cycles with nanotechnology and hydrogen. Organizations must build the capacity to deal with these innovations effectively rather than sitting on the sidelines waiting to see what happens, or jumping in ill-prepared and wasting money. The key is to train people, learn what's happening and figure out how to extract the best value from developments in data science, AI and robotics.

In terms of the big picture, how do you see AI changing the field of chemistry over the longer term?

AI makes complex things easier to do. It’s not about replacing jobs but enhancing capabilities. With the right tools, you can turn an average engineer into an excellent one. And while AI can bring about this democratization of science and technology, we also need to learn how to use these tools properly, understand what they can and cannot do, what we can trust, and what requires verification. In this way, AI is changing the practice of science and engineering; it’s not changing the essence of these fields.