Sustainability and quantum computing

“We have a history of refining and marketing petroleum products, but we’re adding a focus on carbon reduction,” says Takeshi Ibuka, Data Science Group Manager at ENEOS. “We’re researching how to produce and transport hydrogen fuel, which is a very complicated project.”

Transportation presents a challenge because of hydrogen’s light weight, so ENEOS needs to convert it to methylcyclohexane. To help achieve this, the company started catalyst research based on vibration analysis of molecules, including the transition state of a reaction. Generally speaking, vibration analysis requires a lot of computational power, but the vibration analysis of the transition state is even more demanding and requires a tremendous amount of computational power. This is due to the necessity of considering electron correlation when calculating the transition state. 

Operating on the leading edge of new paths for computation, ENEOS decided to explore the use of quantum computing. Quantum computing harnesses the unique behavior of quantum physics and uses quantum mechanics to run calculations on specialized hardware, resulting in the tantalizing potential of solving computational problems that are out of reach for standard classical computing. This concept is still almost entirely theoretical, but top researchers are beginning to make headway. 

Bringing expertise together

ENEOS had a lot of in-house expertise but needed to find the right partners for research, computation, and hardware to fully bring the project to life. In 2020, the company teamed up with QunaSys, a software startup based in Tokyo. QunaSys focuses on algorithm and software development for chemical calculation on quantum computers. Its 28 employees bring a strong technical background, including 14 PhDs. The startup offers two main solutions: joint research projects and its software tool, Qamuy. 

“Vibration analysis is key to clarifying the behavior of molecules, but its computational cost is too high to get a precise result,” says Takafumi Ishii, Chemist at ENEOS. “We’re laying the groundwork for how quantum computing may help us in the future, and we chose to partner with QunaSys because of its wonderful technologies and expert knowledge.”

Based on a recommendation from QunaSys, ENEOS adopted Microsoft Azure Quantum to perform the computation. Engineers and data scientists use Azure Quantum for experimentation, research, and testing, starting by breaking down the problem into the smallest possible computational units. The idea is that solving for those on a molecular level using Azure Quantum will lead the company to solve the larger-scale, complex issue. ENEOS decided to use the Honeywell System Model H1, a quantum computer created by Honeywell Quantum Solutions, now known as Quantinuum. 

“We work closely with ENEOS, and our chemical engineers even go into the ENEOS laboratory to do molecular calculations,” says Tennin Yan, Chief Executive Officer at QunaSys. “We use Azure Quantum to move our research forward. We write algorithms, the ENEOS engineers evaluate the algorithms, and we collaborate together to further optimize them.”   

Theoretical computing, validated

In this early era of quantum computing, the goal for ENEOS was not to find a new result—because the calculations are too small for the computation time to matter—but rather to validate that the algorithm running on the quantum system matches the team’s predictions. As these micro experiments increase in size, early adopters understand that quantum computing will eventually become more scalable. 

ENEOS used both Azure Quantum running on Honeywell H1 and the Qamuy simulator to validate molecule vibration analysis accounting for electron correlation. Because the accurate calculation cost of vibration analysis for large molecules is not within the reach of classical computers, the company wanted to confirm that quantum computing could perform vibration analysis with highly accurate initial calculations. 

“Although we started with small molecules in this research, it was a fantastic proof of concept,” says Taku Watanabe, Lead Scientist at ENEOS. “The calculations we ran using Azure Quantum returned the same results on the quantum computer as the simulator, which shows that we performed the calculations successfully.”

Adds Ibuka, “We validated the algorithms and demonstrated that we can conduct the vibration analysis on the quantum computer running Azure Quantum. It sounds simple but proving that is a really big deal. Now that we’ve proven it theoretically, we’re confident that the quantum computer will be able to handle bigger problems. We’re very happy with the results and accuracy of this project, and we learned that quantum algorithms will definitely be beneficial to us.”

Forging the future

For pioneers in any groundbreaking technology, change doesn’t happen in an instant. It’s the result of incremental learning and new understandings that move them, step by step, in the right direction. “We’re moving closer to viable hydrogen fuels, but we still need to overcome many challenges before we can build a hydrogen society,” says Watanabe. “Developing a catalyst will give us a key technology that will play a big role in hydrogen fuels, and this project’s results show that Azure Quantum will be a critical part of that development.” 

Part of the value of working with QunaSys is that ENEOS now has a clear direction for future research. QunaSys is running the Quantum Practical Application Research Consortium (QPARC), a community focused on exploring the potential of quantum computing. “Our joint research currently focuses on molecular simulation, but QunaSys also provides us with the vision of quantum computing through its own thought leadership, consulting, and the QPARC community,” says Ishii. “We’re very grateful to QunaSys for introducing us to Azure Quantum.” 

Find out more about ENEOS on Twitter, Facebook, and LinkedIn. Learn about QunaSys on Twitter and LinkedIn.