Chemists explore climate solutions with Qiskit at 2023 Deloitte Quantum Climate Challenge
By Daniella Garcia Almeida
Some researchers believe quantum computers could one day play a vital role in the fight against climate change, enabling simulations of novel sustainable materials, or solving complex optimization problems to minimize greenhouse gas emissions. But when will quantum computers become mature enough to actually tackle these problems at a useful scale? How do we assess which applications and methods will provide the most value?
The annual Deloitte Quantum Climate Challenge aims to answer these questions and more by inviting researchers and developers from around the world to try their hand at using quantum computers to solve a scaled-down version of a challenging climate change problem. This year’s challenge centered on simulating materials designed to capture carbon dioxide from the atmosphere, a method known as direct air capture (DAC). Some climate scientists believe DAC will be essential to achieving the goals laid out in the international Paris Agreement, which calls for limiting the increase in average global temperature to no more than 2 degrees Celsius between the years 1900 and 2100.
The winner of the 2023 Deloitte Quantum Climate Challenge was EcoQult, a team of three researchers — chemical engineer Kourosh Sayar Dogahe, quantum chemist Tamara Sarac, and data scientist Delphine De Smedt — hailing from the Belgian quantum computing startup QBee. Their winning project involved calculating complex properties of atmospheric gas molecules, an important step in developing effective DAC materials. In a recent interview, we sat down with team leader Tamara Sarac to gain insight into their work. A more detailed article describing the specifics of their research is available on arXiv, here.
Responses edited for clarity and length.
Daniella Garcia Almeida: How did your team come across the 2023 Deloitte Quantum Climate Challenge, and why did you want to get involved?
Tamara Sarac: We found it accidentally! One of our team members found a LinkedIn post about the challenge and our team thought it posed a very interesting use case in the chemistry domain. We all work for a start-up called QBee, and we are always very keen on bridging quantum computational methods with real world applications. It’s especially nice to work on problems that are important to industry and society, where the world is saying “yes, we need help solving this computational problem.” The Deloitte Quantum Challenge met all of those criteria.
DGA: This year’s challenge centered on investigating how quantum computers may help to improve materials used in direct air capture of carbon dioxide. Why is quantum computing well-suited to this application?
TS: Every time you have a complex problem and you want to have a realistic representation of the system, it becomes very intensive for classical computers. We need more powerful methods and solutions that can extend the computational capabilities we already have. I remember when I first started learning about quantum computing and its potential. That was when I started to really believe in this technology and how much we can do with it. But we still have a lot of work to do to fulfill this potential. Material simulations, especially when you have these complex systems, are so computationally intensive that you need to give the classical computer a helping hand. And I do believe quantum computing will offer that in the future.
DGA: Can you briefly summarize the work you did in this project?
TS: We approached the problem from the perspective of our domain expertise — as chemists. Going into the challenge, Deloitte offered a few different suggestions for how we might approach solving this problem. Some people were dealing with advancing the algorithm side of things — mostly focusing on VQE. However, our team was mostly focused on working at it from the chemical perspective. We spent a lot of time thinking about how to represent the molecule in the most suitable way. We had to figure out how to focus on the most active part of the molecule. In chemistry, we also tend to think in terms of molecular orbitals, so we wanted to figure out the best way of representing these orbitals.
We performed our energy calculations using VQE. And when it comes to the energy calculation and VQE, it was interesting because internally we tend to work with qubit simulations that are hardware agnostic. We focus on developing computational methods and on how to mathematically represent the system in a way that’s suitable for quantum. Our founder at QBee even has a name for this — “PISQ” or “Perfect Intermediate Scale Quantum.” We hardly ever worked with hardware implementation, and initially it was a very slow process. However, we really improved over time, and using the hardware helped us learn even more about quantum and take our skills a step further.
DGA: What is your team’s background in either quantum computing, the materials’ sciences and/or other relevant fields, and how did that background inform your work here?
TS: My personal background is in material science, I did my PhD in Belgium focusing on modeling the behavior of polymer material under gamma radiation. Kourosh Sayar Dogahe has a similar background to me focusing on modeling behavior of biomedical materials. Delphine De Smedt’s background is in data science, coming from a computer engineering environment with a sustainability focus. She supported our team a lot with data visualization.
DGA: How did your team use Qiskit in this project?
TS: Qiskit was the most logical choice for us. It was our number one tool for our previous simulation projects, it was really beginner friendly. We had amazing help from the IBM support team on our implementations and how to improve them. The challenge helped us grow in the way that we used Qiskit and also grew our mindset on how we can achieve usable quantum-based computation of models.
DGA: Which IBM Quantum computer did your team use (over the cloud)? How did you use simulators vs. real quantum hardware in this work?
TS: We used three different IBM systems: Manila, Lima, and Quito. We have a vision where we think quantum computing should become more available for domain experts. Domain experts don’t always have a background in physics or computing, which makes simulators a natural step to dive into the world of quantum. With time, we expect their next step to be implementation on hardware.
DGA: What would you say is the key insight or accomplishment that made this a winning submission?
TS: I think it was because we answered the question, or at least we tried to address the question, in the right way. The question was “can we use QC to advance simulations of materials that will be used to capture CO2?” So, how you answer this question is really about benchmarking several different methods — specifically the methods that are classically available, methods you get when you assume perfect qubit behavior (i.e. simulation methods), and the methods you have available today in terms of existing quantum hardware. It wasn’t the most complete analysis, I must say. We did this in a very short period of time, and someone will have to do this work at a much larger scale to provide really useful answers. But still I think this benchmarking approach helped a lot.
The other thing is that these simulations really require you to approach them from a chemistry perspective, so having two chemists on our team definitely helped. I think our domain expertise in chemistry gave us a significant advantage.
DGA: How do you think this work will impact the quantum community?
TS: We want to motivate people from different backgrounds to look into quantum computing. It’s easy to say quantum computing will solve a complex problem but it’s not always straightforward even to define a use case in a certain field — let alone to formulate an efficient quantum-based solution and build a fault tolerant quantum accelerator. We hope to invite people from end-user domains, like professors or software developers, to look into quantum computing and see what they can contribute and improve.
To learn more about team EcoQult’s award-winning research, be sure to read their article summarizing their work on arXiv.
