Accelerating the discovery of organic lasers with self-driving labs
A self-driving lab (SDL) at the University of Toronto has discovered organic lasers with state-of-the-art performance—and it only took 2 days. Published in Advanced Materials, this research is led by Acceleration Consortium (AC) members Alán Aspuru-Guzik, Jason Hein, Martin Burke, and collaborators from around the world.
A self-driving lab (SDL) at the University of Toronto has discovered organic lasers with state-of-the-art performance—and it only took 2 days. SDLs work by combining artificial intelligence (AI), automation, and advanced computing to reduce the time and cost of bringing materials, like organic lasers, to market.
With unique light-emitting properties, lasers have become critical parts of our everyday life, such as communication, data storage, medicine, and industrial manufacturing. Lasers play an increasingly important function in many current and emerging technologies, from internet communications and the navigational sensors in self-driving cars to eye surgery and lifesaving cancer treatment.
Unlike most automation efforts in chemistry which focus on making the molecules, this platform also successfully integrates materials evaluation, a key step in determining efficacy.
“This is an exciting demonstration of the power of self-driving labs to accelerate discovery, as showcased here with high-performance organic laser molecules,” says Dr. Tony C Wu, the lead author of the paper and a former postdoc in Aspuru-Guzik's lab. “While the application potential is huge, organic lasers are still challenging to make and are not yet used in any commercial devices on the market. To overcome the challenges, our SDL optimizes the development process to include automated lego-like synthesis, purification and characterization—all essential aspects to organic laser exploration. ”
“Using our self-driving labs, we were able to explore the laser efficacy of 40 molecules in just two days,” says Prof. Andrés Aguilar Granda, a co-author of this work and a former postdoc in Aspuru-Guzik's lab. “Currently, the lab is continuing to explore more laser molecules, searching through more than 200 candidates within one month. Without the support of SDLs, a predecessor in the field published fewer than 10 laser molecule candidates in five years.”
Beyond lasers, what is especially exciting is that as a platform technology, SDLs can also be applied to a wide range of industries and products, from renewable energy and biodegradable plastics to resistance evasive drugs and low-carbon cement. Without AI and automation, scientists can only explore a limited area of chemical space using conventional trial and error methods. Not only do SDLs make materials innovation cheaper and faster, but they also expand our capacity to search for novel materials and molecules exponentially.
“Whether it is climate change or a pandemic, many of the world’s most pressing challenges require immediacy,” says Alán Aspuru-Guzik, director of the Acceleration Consortium (AC). “Self-driving labs will allow us to respond with material solutions in a timescale where these solutions can have real impact. Companies and stakeholders will also reap the benefits of bringing better and greener products to the market at a faster pace.”
Related research was recently published in Science where AI, “building-block” chemistry and a molecule-making machine teamed up to find the best general reaction conditions for synthesizing chemicals important to biomedical and materials research—a finding that could speed innovation and drug discovery as well as make complex chemistry automated and accessible. This work, also part of the multi-year project, represents another collaboration of AC members Martin Burke, Bartosz Grzybowski, and Alán Aspuru-Guzik. These reaction conditions are used now in the self-driving lab at the University of Toronto.
As Burke told the University of Illinois Urbana-Champaign news bureau, “Generality is critical for automation, and thus making molecular innovation accessible even to nonchemists. Traditionally, chemists customize the reaction conditions for each product they are trying to make. The problem is that this is a slow and very specialist-dependent process, and very hard to automate because the machine would have to be optimized every time. What we really want are conditions that work almost every time, no matter what two things you’re trying to snap together.”
As the research under the DARPA AMD project continues, the team at large, which includes contributions from AC member Lee Cronin, is expected to soon share their findings on a large chemical space exploration using the tool described in this paper. “Stay tuned,” says Alán Aspuru-Guzik, “and as Marty Burke suggests, watch out!”