Exascale: Aurora's Search for Clean Energy Catalysts

According to Argonne National Laboratory, scientists are creating exascale software tools to facilitate the creation of novel chemicals and chemical processes for the generation of renewable energy.

One of the country's first exascale systems, Aurora, is being built by Argonne. 15 research teams are participating in the Aurora Early Science Program (ESP) through the Argonne Leadership Computing Facility (ALCF), a DOE Office of Science user facility, to build codes for the design and size of the new supercomputer. These researchers will be among the first in the world to utilize Aurora for science thanks to access to pre-production time on the system.
Industrial chemistry is centered on catalysts, and producing cleaner, more plentiful energy is centered on industrial chemistry. It should thus come as no surprise that the country places a high priority on research aimed at better understanding catalysts at the atomic and molecular level.

You might be startled to find that much of this research can be conducted without ever setting foot inside a laboratory if you have an image of individuals in white coats pouring colored chemicals into beakers. With something called the Schrödinger equation, supercomputers all across the world are being requested to conduct research in this field. Erwin Schrödinger's equation, which has become a cornerstone of modern chemistry and is incredibly helpful in chemical reaction prediction, is well known for its live-cat, dead-cat contradiction.

"A molecular understanding of how catalysts work is critical to developing new catalysts for many important applications," said David Bross, a computational chemist at the Argonne National Laboratory of the U.S. Department of Energy (DOE). Scientists may design and manufacture innovative catalysts to enhance a variety of clean energy technologies with a comprehensive atomistic understanding of the underlying chemical processes.


Bross and his team are preparing to turn on Argonne's exascale supercomputer for some intense computational chemistry as the project leader for an Aurora Early Science Program (ESP) project. When it is made accessible for use in research, the Aurora supercomputer, which was built in partnership with Intel and Hewlett Packard Enterprise, is anticipated to be among the fastest in the world.

Many things that weren't previously conceivable will become possible thanks to exascale computing, according to Bross. The promise of these computers is that you can do either extremely big simulations or a very large number of smaller simulations that were impractical to perform on earlier supercomputers.

The more comprehensive DOE Exascale Catalytic Chemistry (ECC) project is connected to Bross' Aurora ESP project. The ECC team has been working diligently since 2017 to prepare their suite of research software for the exascale era. Researchers from Argonne, DOE's Pacific Northwest National Laboratory, Brown University, and Northeastern University are a part of the ECC project, which is being led by Judit Zádor of Sandia National Laboratories.


"The variety of catalysts and operating circumstances is one of the major barriers to catalysis development. Finding interesting catalytic reactions just through experimentation is quite laborious, according to Zádor. As part of the ECC project, "our approach is to develop a software infrastructure that allows for the automatic creation of models for these systems, and study their behavior computationally."

Bross's ESP project is focused on creating and optimizing software for Argonne's Aurora system, while the ECC project is preparing for all of the country's exascale computers. Research into heterogenous catalysis, in which the catalyst and the reactant are in separate physical phases, is being advanced by the team's software tools and approaches.


Understanding the interaction between gas phase molecules and solid catalytic surfaces is our main goal. Bross said. “Many industrially relevant chemical reactions occur in these heterogeneous environments.”
Many of the technology and goods we use every day depend on catalysts. Almost every industrial processing employs a catalyst of some sort to accelerate and improve reactions. Although Bross mentioned two well-known applications for catalysts, he also said that "new catalysts are being discovered all the time." The Haber-Bosch process, which produces ammonia for fertilizing plant life, includes one typical catalyst. The catalytic converter in cars, which transforms hazardous fuel waste into safer and dispersible chemicals from your exhaust, is another well-known example.

Similar to the catalytic converter, the primary goal of catalysis research is to identify substances and procedures that may be used to lower dangerous levels of carbon or other pollutants. as in the production of fuels, new catalysts will be needed to lower the carbon footprint of vehicle exhaust.

"Catalysts can handle various carbon sources and transform them into more useful compounds. To take energy and direct it where you want, in essence," explained Bross. Our major objective is to create software that is exascale ready and can be utilized to research such systems.


In order to swiftly study the molecular energy landscapes of gas-solid surface interactions, the team's program employs innovative methodologies. These resources might lead to the identification of brand-new chemical pathways. Furthermore, calculating the Schrödinger equation will be significantly more accurate and efficient because to Aurora's potent hardware.

The team has been working on the new Sunspot testbed as well as the Theta and Polaris supercomputers from the ALCF as part of the ESP project. To remove bottlenecks and fine-tune performance difficulties, the researchers collaborate closely with the Intel team, Raymundo Hernandez Esparza, and lvaro Vázquez-Mayagoitia of the ALCF. A number of accomplishments have already been made by Bross and his ECC project partners, including the publication of academic publications. They also updated open data repositories and provided their open-source software tools on GitHub.


The project's dissemination of this software, its methods, and its publications to scholars all around the world is another crucial part. The group's work will be accessible to the public and marketed to businesses and other laboratories, opening up way for further use of exascale computing power in the design and discovery of new catalysts.

We will be able to completely and methodically examine catalytic systems by using exascale systems to run our automated processes in conjunction with precise quantum chemical computations, according to Zádor. The ability to observe the overall picture of catalytic activity would be lost if we were to use less powerful devices, which would limit our ability to examine a wider range of situations.

By discovering novel catalysts, business may produce more effective processes that use less energy and generate less chemical waste, particularly carbon, which will contribute to a future with cleaner energy sources. According to Bross, there is a great deal of promise for new discoveries in this area.


The Basic Energy Sciences division of the DOE Office of Science is funding the ECC project through computational chemical sciences.