Javier Argüello Luengo (Spain, 2018) Javier Argüello Luengo is taking up his PhD in Photonics at the www.icfo.eu and has just published in Nature the article“Analogue quantum chemistry simulation”, an innovative research study carried out together with a team of scientists from the Max Planck Institute of Quantum Optics, the Spanish National Research Council and the Institute for Quantum Optics and Quantum Information in Innsbruck.
The research focuses on the development of an analogue quantum simulator to solve chemistry problems that conventional computers are unable to find an answer to at present.
Javier explained a few days ago in a Twitter thread some aspects of his research, “In the past, architects did not have to wait for the latest AutoCAD update to design the Colosseum in Rome or the Sagrada Família. How did they get it right? Well, they built models that simulated the building’s physics. For example, Frei Otto used soap film to calculate the most stable surfaces that would cover the Munich Olympic Stadium”.
Is this graphic example you shared on Twitter a simile of the analogue quantum simulator?
Exactly, it's an example of how, throughout history, humans have relied on analogue simulators to tackle physics problems that could not be solved by the means of the time. To continue with the example, calculating the structural load of a building is too complex a problem to do by hand, and until computing was developed, we had no computers to help us. What these architects did was to build small models in which ropes and sandbags were used to find the best ways to distribute the loads of their building. As these models were subject to the same forces and pressures as the final building, they could see the structure they sought in them without having to calculate. Now, going back to the present, we lack computers that are efficient at solving chemistry problems, and a very desirable alternative would be to have simulators capable of reproducing the forces felt by electrons in a molecule, which determine most of the relevant chemical properties in the industry or biochemistry. How to design a simulator capable of tackling chemistry problems is something that has remained unknown up to now.
What is the simulator you have proposed like?
Well, as molecules obey the laws of quantum physics, our simulator also needs to obey those laws. A quantum system over which we have a lot of control in the laboratory is the one described by atoms at low temperatures. Using the light from a laser, we can manipulate the way these atoms move or interact with each other, making it possible to simulate some quantum problems related to the physics of materials or magnetism for more than a decade now. The challenge in applying this platform to chemistry problems is making these atoms feel the same forces affecting electrons in a molecule. In other words, for them to move, be attracted to certain nuclear positions, and repel each other in the same way as electrons in nature. What we have found in this study is an experimental scheme that is able to achieve this, opening the door to using these simulators as a tool to better understand molecular properties that are difficult to calculate using other methods.
How are new substances currently being studied in the chemical industry?
Today, there's great interest in understanding chemical structures. In the field of biomedicine, for example, predicting the most stable way in which the atoms in a protein are distributed in space makes it possible to understand the biological function they can perform. These calculations are also behind the synthesis of new drugs or the design of more efficient industrial processes. As it's a quantum problem, conventional computers are particularly inefficient at dealing with chemistry problems in a precise manner, so over the past century, increasingly refined numerical methods have been developed to find approximate solutions for some of these problems, to the extent allowed by our computing capacity. While this is enough for some tasks, other equally relevant yet complex problems are waiting to be solved by computers much more powerful than what we have at present or a radical change in the way we solve them.
In this paradigm shift, a very powerful alternative is tackling chemistry problems with the aid of computers that also obey the laws of quantum physics. This is a line of research that has been extremely active over the past few years and it’s closely linked to the development of quantum computing. In recent weeks, Google announced a specific problem that its 53-qubit quantum computer is able to solve faster than any other conventional computer. Gaining this quantum advantage in relevant chemistry problems is not only a complex and demanding challenge, but also one of the most exciting applications in this field. To do this, it is necessary for these quantum computers to continue growing in size and fidelity, for which experts have confidence in ground-breaking technological developments over the coming decades. Throughout this journey, it would then be of great interest to have quantum systems that—although they cannot be configured to solve any given problem, as a quantum computer would—are specifically designed to tackle specific chemistry problems. This is a direction towards which analogue simulators offer a very attractive opportunity.
What does your project contribute towards the future of research? Can you give us a specific example of a problem that has no solution at the moment but which could have one in the future?
One of the main advantages of analogue simulation is the great degree of control it gives. On the one hand, it allows us to study nature under more favourable conditions than those we would find in real molecules. For example, these simulated electrons are ten thousand times larger than electrons and can move up to twelve orders of magnitude more slowly, which would make it possible for us to observe how a chemical reaction takes place with a definition and precision in time difficult to achieve in a laboratory experiment. On the other hand, although the interactions felt by electrons in a molecule are determined by nature, these simulators make it possible to modify them at will by turning a laser on or off, giving rise to the exact solution to problems which are currently beyond our computing power.
Now that we know that the analogue simulation of chemistry problems is possible, the next step is to verify it through experiments. We hope that in the coming years simplified experiments using some of the strategies we have developed to solve some trivial chemistry problems will appear, paving the way for more complex simulations in the longer term. In many cases, how good the approaches used by conventional computers are remains unknown, and having a quantum system with which to compare their precision and correctness would be a very powerful tool to improve and validate the chemical models used at present.