Explore the “chemical space” with Anatole von Lilienfeld

It’s not hard to be drawn in by the beauty and unprecedented detail of the photographs captured by NASA’s James Webb Space Telescope. As well as being beautiful to look at, the images have the potential to expand our understanding of the origins of the universe and reveal previously unseen aspects of the cosmos. When we can see our surroundings, we can better understand where we are going – or want to go – and how best to get there.

Anatole von Lilienfeld also navigates space, but instead of exploring the depths of the universe, his work here on Earth is in “chemical space”.

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

Von Lilienfeld is the first Clark Chair in Advanced Materials at the Vector Institute and the University of Toronto and a core member of the U of T’s Acceleration Consortium (AC). He was jointly appointed to the Department of Chemistry of the Faculty of Arts & Science and the Department of Materials Science & Engineering from the Faculty of Applied Science & Engineering and is one of the world’s brightest visionaries when it comes to using computers to understand the vastness of chemistry.

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

This four-day program focused on the power of self-driving laboratories (SDLs), an emerging technology that combines AI, automation and advanced computing to accelerate material and molecular discovery. The Accelerate conference brought together over 200 people and featured presentations and panel discussions with more than 60 experts from academia, industry and government shaping the burgeoning field of accelerated science.

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

How big is the “chemical” space?

We are surrounded by materials and molecules. Think of the chemical compounds that make up our clothes, the pavement we walk on and the batteries in our electric cars. Now think of the new possible combinations waiting to be discovered, such as: B. Catalysts for effective atmospheric CO2 capture and utilization, low-carbon cement, lightweight, biodegradable composite materials, membranes for water filtration, and potent molecules for treating cancer and bacteria-resistant disease.

In practical terms, chemical space is infinite and searching it is not an easy task. A lower estimate says it contains 10⁶⁰ Connections – more than the number of atoms in our solar system.

Why do we need to speed up the search for new materials?

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

Traditional chemistry is slow, a series of often lengthy trials and errors that limit 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 properties (e.g. conductive, biodegradable, etc.).

This is just one step toward self-driving labs, an emerging technology that combines AI, automation, and advanced computing to reduce the time and cost of materials discovery and development by up to 90 percent.

How can human chemists and AI work together effectively?

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

Experimentation is the basis. We experiment with the goal of improving the physical world for humans. Then comes theory to give shape and direction to your experiments. But the theory has its limits. Without computer simulation, the computational effort required to support scientific research would take far longer than a lifetime. But computers also have limitations.

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

Are there limitations for AI like you described in the other pillars of science?

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

Data collection is costly, both financially and in terms of carbon footprint. To counteract this, the goal is to improve the AI. If you can code 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 required for training, the AI ​​model becomes smaller.

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

How can we make AI more accessible for discovery?

The first way is open source research. In the burgeoning field of accelerated science, there are many proponents of open source access. Journals provide access not only to research, but in many cases to the data as well, which is an important component of making the field more accessible. There are also repositories for models and code like GitHub. Providing access could lead to scientific advances that ultimately benefit all of humanity.

A second way to expand the AI ​​for discovery is to involve more students. We need to teach basic computer science and programming skills as part of a chemistry or materials science education. Schools around the world are beginning to update their curricula accordingly, but we still need to see more of how this foundational education is integrated. The future of science is digital.

How are initiatives like the Acceleration Consortium and a conference like Accelerate helping to advance the field?

We are at the beginning of a real digitization of the chemical sciences. Coordinated, collaborative efforts such as the Acceleration Consortium will play a critical role in synchronizing efforts not only at the technical but also at the societal level, thereby enabling the global implementation of an “updated” version of chemical engineering with unprecedented benefits for humanity at large and whole. The consortium also serves to bridge academia and industry, two worlds that could benefit from a closer relationship. Visionaries in the commercial sector can dream up opportunities and the consortium will be there to help make the science work. The groundbreaking thing about AI is that it can be applied to any sector. AI is poised to have an even bigger impact than the advent of computers.

Accelerate, the consortium’s first annual conference, was a great rally event for the community and a reminder that remarkable things can come from a gathering of bright minds. While Zoom has done much for us during the pandemic, it cannot match the excitement and enthusiasm that is often cultivated at an in-person conference and that is required to lead research and encourage a group to pursue a complex goal just imitate.

What area of ​​chemical space fascinates you the most?

Catalysts that enable a specific chemical reaction while remaining unaltered. A century ago, Haber and Bosch developed a catalytic process that would enable the conversion of nitrogen – the dominant substance in the air we breathe – into ammonia. Ammonia is an important raw material for the chemical industry, but also for fertilizers. It enabled the mass production of fertilizers and saved millions of people from starvation. Large parts of humanity would not exist today if it were not for this catalyst.

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