Researchers from the University of Toronto’s Faculty of Applied Science and Engineering and Fujitsu have used quantum-inspired computing to find the promising, previously unexplored Ru-Cr-Mn-Sb-O chemical group.2 as acid oxygen evolution reaction catalysts for hydrogen production.
The best catalyst has eight times the mass activity of the state-of-the-art RuO2 and maintains performance for 180 h. An article about their work appears in the magazine Matter.
Chubisa and others.
Increasing the production of so-called green hydrogen is a priority for researchers worldwide because it offers a carbon-free way to store electricity from any source. This work provides a proof-of-concept for a new approach to overcome one of the major remaining challenges, namely the lack of highly active catalyst materials to accelerate critical reactions.
— Ted Sargent, Senior Author
Almost all commercial hydrogen is produced from natural gas. The process produces carbon dioxide as a by-product; if CO2 is released into the atmosphere, the product is known as gray hydrogen, but if CO2 is captured and stored, it is called blue hydrogen. Green hydrogen is a carbon-free method that uses an electrolyzer to split water into hydrogen and oxygen gas. The low efficiency of available electrolyzers means that most of the energy in the water splitting step is wasted as heat rather than capturing hydrogen.
Researchers around the world are trying to find better catalyst materials that could improve this efficiency. Since each potential catalyst material can be made from several different chemical elements combined in different ways, the number of possible permutations quickly becomes overwhelming.
One way to do this is through human intuition, researching what material other bands have made and trying something similar, but it’s quite slow. Another way is to use a computer model to simulate the chemical properties of all the possible materials we might try from first principles. But in this case, the calculations become very complex and the computing power required to run the model becomes enormous.
— Jehad Abed, Contributor
To find a way out, the team turned to the emerging field of quantum-inspired computing. They used the Digital Annealer, a tool developed through a long collaboration between U of T Engineering and Fujitsu Research. This collaboration has resulted in the Fujitsu Co-Creation Research Laboratory at the University of Toronto.
The Digital Annealer (DA) is a computer architecture designed to rapidly solve large-scale combinatorial optimization problems using CMOS digital technology. DA is unique in that it uses digital circuit design inspired by quantum phenomena and can solve problems that are very complex and time-consuming or even impossible for classical computers.
The Digital Annealer is inspired by quantum mechanics, but unlike quantum computers, it does not require cryogenic temperatures. DA uses a method called annealing, named after the annealing process used in metallurgy. During this procedure, the metal is heated to a high temperature before the structure stabilizes as it is slowly cooled to a lower energy, more stable state.
Using the analogy of putting blocks in a box, in the classical computing approach the blocks are placed sequentially. If no solution is found, the process restarts and repeats until a solution is found. With the annealing approach, the blocks are placed randomly and the whole system is “shaken up”. By gradually reducing the shaking, the system becomes more stable as the shapes fit together quickly.
DA is designed to solve fully coupled quadratic unconstrained binary optimization (QUBO) problems and is implemented in CMOS hardware. The second generation Digital Annealer expands the scale of problems that can be solved from the first generation’s 1024 bits launched in May 2018 to 8192 bits and increasing computational accuracy.
This results in a significant improvement in accuracy and performance for advanced problem solving and new applications, expanding the complexity that can now be tackled by the second generation Digital Annealer a hundredfold. Its algorithm is based on simulated annealing, but it also uses massive parallelization provided by custom application-specific CMOS hardware.
The Digital Annealer is a unique hardware/software hybrid designed to be highly efficient in solving combinatorial optimization problems. These problems include finding the most efficient route between multiple locations across a transportation network or selecting a pool of stocks to create a balanced portfolio. Another example is searching through different combinations of chemical elements to find a catalyst with the desired properties, which was a great challenge for our Digital Annealer.
-Hidetoshi Matsumura, Senior Researcher, Fujitsu Consulting (Canada).
In the paper, the researchers used a technique called cluster expansion to analyze a huge number of potential catalyst material structures – they estimate the total to be in the hundreds of quadrillion. For perspective, one quadrillion is about the number of seconds that would elapse in 32 million years.
Quantum hardeners and similar quantum-inspired optimizers have the potential to provide accelerated computation for certain combinatorial optimization problems. However, they have not been used for material discovery due to the lack of compatible optimization mapping methods. Here, by combining cluster expansion with a quantum-inspired superposition technique, we use quantum quenchers for the first time in the exploration of chemical space. This approach allows us to accelerate the search for materials with desired properties 10-50 times faster than genetic algorithms and Bayesian optimization, significantly improving the accuracy of ground state prediction.
— Chowbisa and others.
The results pointed to a promising group of materials consisting of ruthenium, chromium, manganese, antimony and oxygen that had not been previously explored by other research groups.
The team synthesized several of these compounds and found that the best of them exhibited mass activity about eight times greater than some of the best catalysts currently available.
The new catalyst also has other advantages: it works well in acidic conditions, which is a requirement for modern electrolyzers. Currently, these electrolyzers depend on catalysts made primarily of iridium, a rare element that is expensive to obtain. In comparison, ruthenium, the main component of the new catalyst, is more abundant and has a lower market price.
The team aims to further optimize the stability of the new catalyst before it can be tested in an electrolyser. However, the latest work serves as proof of the effectiveness of the new approach for searching chemical space.
I think what’s exciting about this project is that it shows how you can solve really complex and important problems by combining knowledge from different fields. For a long time, materials scientists have sought these more efficient catalysts, and computational scientists have developed more efficient algorithms, but the two efforts have been disconnected. When we combined them, we were able to find a promising solution very quickly. I think a lot more useful discoveries can be made this way.
-Hitarth Choubisa, Contributor
Resources
-
Hitarth Choubisa, Jehad Abed, Douglas Mendoza, Hidetoshi Matsumura, Masahiko Sugimura, Zhenpeng Yao, Ziyun Wang, Brandon R. Sutherland, Alan Aspuru-Guzik, Edward H. Sargent (2022) Accelerated search of chemical space using a quantum expansion approach, ” Matter doi: 10.1016/j.matt.2022.11.031
