Absorbing the sun’s rays with artificial intelligence

The sun constantly transmits trillions of watts of energy to the Earth. It will do so for another billions of years. However, we are only just beginning to benefit from this abundant and renewable source of energy at an affordable cost.

Solar absorbers are a material used to convert this energy into heat or electricity. Maria Chan, a scientist at the US Department of Energy’s (DOE) Argonne National Laboratory, has developed a machine-learning method to screen several thousand compounds as solar absorbers. Her co-author on this project was Arun Mannodi-Kanakkithodi, a former Argonne postdoc who is now an assistant professor at Purdue University.

“We are really in a new era of applying artificial intelligence and high-performance computing to discover materials.” – Maria Chan, Scientist, Center for Nanomaterials

Machine learning methods are being developed at Argonne to advance solar energy research with perovskites.

” data-image-caption=”

Image courtesy of Maria Chan/ Argonne National Laboratory.

” data-medium-file=”https://cleantechnica.com/files/2022/09/image-400×225.jpg” data-large-file=”https://cleantechnica.com/files/2022/09/image.jpg” loading=”lazy” class=”size-full wp-image-276007″ alt=”” width=”502″ height=”282″ srcset=”https://cleantechnica.com/files/2022/09/image.jpg 502w, https://cleantechnica.com/files/2022/09/image-400×225.jpg 400w” sizes=”(max-width: 502px) 100vw, 502px”/>

Argonne’s machine learning methods are being developed to advance solar energy research using perovskite. (Photo by Maria Chan/Argonne National Laboratory).

“According to a recent study by the Department of Energy, by 2035, solar energy could power 40% of the country’s electricity,” Chan said. “And it could help remove carbon from the grid and provide many new jobs.”

Chan and Mannodi-Kanakkithodi are betting that machine learning will play a vital role in achieving this lofty goal. Machine learning, a form of artificial intelligence (AI), uses a combination of large data sets and algorithms to imitate the way humans learn. It learns from training using sample data and past experience to make better predictions than ever before.

In the days of Thomas Edison, scientists discovered new materials through a laborious process of trial and error with many different candidates until one worked. Over the past several decades, they have also relied on labor-intensive calculations that require up to a thousand hours to predict the properties of a material. Now, they can shorten the two discovery processes by invoking machine learning.

At present, the primary absorber in solar cells is either silicon or cadmium telluride. Such cells are now common. But its manufacture remains rather expensive and energy-intensive.

The team used their machine learning method to evaluate the solar properties of a class of materials called halide perovskite. Over the past decade, several researchers have been studying perovskite because of its remarkable efficiency in converting sunlight into electricity. It also offers the potential for lower cost and energy input to prepare materials and build cells.

“Unlike silicon or cadmium telluride, the potential differences for halides combined with perovskites are essentially unlimited,” Chan said. Thus, there is an urgent need to develop a method by which the most promising candidates can be narrowed down to a manageable number. To this end, machine learning is an ideal tool.”

The team trained their method on data for a few hundred halide perovskite formulations, then applied it to more than 18,000 combinations as a test case. The method evaluated these formulations for key properties such as stability, ability to absorb sunlight, structure that does not break easily due to defects, and more. The calculations agreed well with the relevant data in the scientific literature. Also, the results reduced the number of formulations that deserve further study to about 400.

“Our candidate list contains compounds that have already been studied, compounds that no one has studied before, and even compounds that were not among the original 18,000,” Chan said. “So we’re very excited about that.”

The next step will be to test the predictions using experiments. The ideal scenario would be to use an independent discovery lab, such as Polypot in Argonne Center for Nanomaterials (CNM), a user facility of the Office of Science in the Department of Energy. Polybot combines the power of robotics and artificial intelligence to drive scientific discovery with little or no human intervention.

By using independent experiments to pool, characterize, and test the best out of a few hundred primary candidates, Chan and her team anticipate that they can also improve the current machine learning method.

“We are really in a new era of applying artificial intelligence and high-performance computing for material discovery,” Chan said. “Besides solar cells, our design methodology can apply to LEDs and infrared sensors.”

This research was mentioned in an article in Energy and Environmental Sciencesentitled “Data-driven design of a new perovskite halide alloy.

Support for the research came from the Department of Energy’s Office of Science. The researchers used the computing resources of the National Center for Scientific Computing for Energy Research, a DOE Office of Science user facility, and Bebop, operated by the Laboratory Computing Resource Center in Argonne.

About the Argonne Center for Nanomaterials: The Center for Nanomaterials is one of the five DOE Centers for Nanoscience Research, which are leading national user facilities for interdisciplinary research at the nanoscale with support from the Department of Energy’s Office of Science. Together, the national research centers constitute a suite of complementary facilities that provide researchers with the latest capabilities to manufacture, process, characterize and model nanomaterials, and constitute the largest investment in infrastructure for the National Nano Initiative. NSRC centers are located in the Department of Energy’s Argonne, Brookhaven, Lawrence Berkeley, Oak Ridge, Sandia, and Los Alamos National Laboratories. For more information about the Department of Energy’s NSRCs, please visit https://sci ence .osti.gov / U ser – F acilities / U ser – F acilitis – at – a – G lance.

Argonne National Laboratory It seeks to find solutions to pressing national problems in science and technology. Argonne, the country’s first national laboratory, conducts groundbreaking basic and applied scientific research in nearly every scientific discipline. Argonne researchers work closely with researchers from hundreds of companies, universities, and federal, state, and municipal agencies to help them solve their specific problems, advance American scientific leadership, and prepare the nation for a better future. With employees from more than 60 countries, Argonne is managed by UChicago Argonne, LLC to US Department of Energy Office of Science.

US Department of Energy Office of Science It is the largest supporter of basic research in the physical sciences in the United States and is working to address some of the most pressing challenges of our time. For more information visit https://energy.gov/sc ience.

courtesy Argonne National Laboratory. Written by Joseph E. Harmon


Do we appreciate CleanTechnica’s originality and clean tech news coverage? Think of becoming a CleanTechnica Member, Supporter, Technician, or Ambassador – or patron on Patreon.

Don’t want to miss the clean tech story? sign for Daily news updates from CleanTechnica on email. Or follow us on Google News!

Do you have a tip for CleanTechnica, want to advertise, or want to suggest a guest on our CleanTech Talk podcast? Contact us here.