Exploration of “chemical space” with Professor Anatole von Lilienfeld

Teacher Anatole of Lilienfeld (Chemistry, MSE) navigates space – but rather than exploring the depths of the universe, his work is here on Earth in “chemical space”.

And instead of chasing unknown stars, galaxies and other celestial objects, it focuses on the untapped potential of undiscovered chemical combinations. To do this work, he is not equipped with a powerful telescope – his tool of choice is artificial intelligence (AI).

Von Lilienfeld is the inaugural Clark Chair in Advanced Materials at the Vector Institute and the University of Toronto, and a key member of the University of Toronto Acceleration Consortium (AC). Jointly appointed to the Department of Chemistry in the Faculty of Arts and Science and the Department of Materials Science and Engineering at U of T Engineering, he is a leading expert in using computers to understand the vastness of chemical space.

Von Lilienfeld, who was recently named the Canada CIFAR Chair in AI, was a speaker at CA’s first annual Accelerate conference last month at the U of T.

This four-day program focused on the power of autonomous laboratories (SDL), an emerging technology that combines AI, automation and advanced computing to accelerate the discovery of materials and molecules. The Accelerate conference brought together over 200 people and featured talks and panels with over 60 experts from academia, industry and government who are shaping the emerging field of accelerated science.

Erin Warner, communications specialist at the Acceleration Consortium, recently spoke with von Lilienfeld about the conference and the digitalization of chemistry.

How big is the “chemical” space?

We are surrounded by materials and molecules. Consider the chemical compounds that make up our clothes, the sidewalk we walk on, and the batteries in our electric cars. Now think about possible new combinations waiting to be discovered, such as catalysts for efficient capture and use of atmospheric CO2, low carbon cement, lightweight biodegradable composites, membranes for water filtration and powerful molecules for the treatment of cancer and bacteria. – resistant disease.

In practice, the chemical space is infinite and traversing it is not an easy task. A low estimate indicates that it contains 1060 compounds – more than the number of atoms in our solar system.

Why do we need to accelerate the search for new materials?

Many of the most widely used materials no longer serve us. Most of the plastic waste generated in the world to date has not yet been recycled. But the materials that will power the future will hopefully be sustainable, circular and inexpensive.

Conventional chemistry is a slow, often tedious series of trial and error that limits our ability to explore beyond a small subset of possibilities. However, AI can speed up the process by predicting which combinations might result in a material with the desired set of characteristics we are looking for (eg conductive, biodegradable, etc.).

It’s just one step in Autonomous Labs, an emerging technology that combines AI, automation and advanced computing to reduce discovery and development time and cost by up to 90% of materials.

How can human chemists and AI work together effectively?

AI is a tool that humans can use to speed up and improve their own research. It can be considered as the fourth pillar of science. The pillars, which build on each other, include experimentation, theory, computer simulation and AI.

Experimentation is the basis. We experiment for the purpose of improving the physical world for humans. Next comes theory to give shape and direction to your experiences. But the theory has its limits. Without computer simulation, the amount of calculations needed to support scientific research would take far longer than a lifetime. But even computers have constraints.

With difficult equations comes the need for high performance computing, which can be quite expensive. This is where the AI ​​comes in. AI is a cheaper alternative. It can help scientists to predict both experimental and computational outcome. And the more theory we put into the AI ​​model, the better the prediction. AI can also be used to power a robotic lab, allowing the lab to operate 24/7. Human chemists will not be replaced; instead, they can spend tedious hours of trial and error to focus more on goal design and other higher-level analysis.

Professor Anatole von Lilienfeld at the Accelerate conference at the University of Toronto. (Photo: Clifton Li, Acceleration Consortium)

Are there any limitations to AI, like the ones you’ve described in the other pillars of science?

Yes, it is important to note that AI is not a silver bullet and there is an associated cost that can be measured in data acquisition. You can’t use AI without data. And acquiring data requires experimenting and recording the results in a way that can be processed by computers. Like a human, the AI ​​then learns by examining the data and making an extrapolation or prediction.

Data acquisition is costly, both financially and in terms of carbon footprint. To remedy this, the goal is to improve the AI. If you can encode our understanding of physics into AI, it becomes more efficient and requires less data to learn but offers the same predictive qualities. If less data is needed for training, the AI ​​model becomes smaller.

Rather than just using AI as a tool, the chemist can also interrogate it to see how well its data captures theory, perhaps leading to the discovery of a new relative law for chemistry. Although this interactive relationship is not as common, it may be on the horizon and could improve our theoretical understanding of the world.

How can we make AI for discovery more accessible?

The first way is open-source research. In the emerging field of accelerated science, there are many proponents of open source access. Not only do journals provide access to research articles, but also, in many cases, to data, which is a major element in making the field more accessible. There are also repositories for templates and code like GitHub. Providing access could lead to scientific advances that will ultimately benefit all of humanity.

A second way to expand AI for discovery is to include more students. We need to teach basic computer and coding skills as part of a chemistry or materials science background. Schools around the world are starting to update their curricula for this purpose, but we have yet to see more incorporate this essential training. The future of science is digital.

How do initiatives like the Acceleration Consortium and a conference like Accelerate help advance the field?

We are at the dawn of a real digitization of the chemical sciences. Coordinated and joint efforts, such as the Acceleration Consortium, will play a crucial role in synchronizing efforts not only at the technical level but also at the societal level, thus enabling the global implementation of an “updated version “chemical engineering with unprecedented benefits for mankind at large.

The consortium also serves to connect academia and industry, two worlds that could benefit from a closer relationship. Commercial visionaries can imagine opportunities, and the consortium will be there to help make science work. The revolutionary nature of AI is that it can be applied to any industry. AI is on track to have an even greater impact than the advent of computers.

Accelerate, the consortium’s first annual conference, was a great rallying event for the community and a reminder that great things can come from a gathering of bright minds. While Zoom has done a lot for us during the pandemic, it cannot easily replicate the excitement and enthusiasm often cultivated during an in-person conference that is needed to guide research and encourage a group to pursue a goal. complex.

Which area of ​​“chemical space” fascinates you the most?

Catalysts, which allow a certain chemical reaction to occur but remain unchanged in the process. A century ago, Haber and Bosch developed a catalytic process that would convert nitrogen – the dominant substance in the air we breathe – into ammonia. Ammonia is an essential raw material for chemical industries, but also for fertilizers. It made mass production of fertilizer possible and saved millions from starvation. Major fractions of humanity would not exist at this time without this catalyst.

From a physical perspective, what defines and controls catalyst activity and components are fascinating questions. They could also be essential in helping us meet some of our most pressing challenges. If we were to find a catalyst that could use sunlight to quickly and efficiently turn nitrogen into ammonia, we might be able to solve our energy problem by using ammonia as fuel. You can think of the reactions that catalysts enable as ways to travel through chemical space and connect different states of matter.