Cracking chemistry using quantum computers? First, master collaboration
The computing technology will enable radical new approaches to solving chemistry problems — but you're going to need a diverse team to pull it off.
If you want your business to get the most out of quantum computing, you're going to need a passion for collaboration.
Jeanette Garcia is IBM's senior manager for quantum applications, algorithms and theory. Her team, which includes chemists, computer scientists, algorithm developers and mathematicians, is based in New York, Tokyo and Zurich, among other locations. To make matters even more difficult, some of them work for different companies.
"One of my favorite challenges as a manager is dealing with teams as diverse as this," she says.
Her latest collaboration, which she and colleagues presented at CES in January, involves pioneering quantum computation in collaboration with Daimler AG, the parent company of Mercedes-Benz.
"One of the first applications of quantum computers will be in the fields of materials science and chemistry,"says Christoph Sedlmayr, who manages vehicle R&D and sustainable mobility communications for Mercedes-Benz. "So Daimler has been looking into the field of quantum computing for some years now,"
Their focus: batteries, which are a major bottleneck in the road toward solving problems such as climate change. Whether it's electric car performance or storage of renewable energy, better batteries would be a game changer. And this is one of the areas where quantum computing could be transformative.
Most current batteries are based on lithium hydride technologies, but these store too little energy per kilogram. The IBM-Daimler collaboration is concerned with computing the chemistry inside future lithium-sulphur batteries, which have the potential to give a much higher energy density. "Essentially, you get more bang for your buck," Garcia says. "For a car, that means a lighter battery that enables you to go farther."
Making lithium sulphide batteries work well means mastering the chemical processes inside them. Essentially, batteries charge and discharge through controlled chemical reactions. In a lithium sulphide battery, lithium and sulphur combine in complex ways inside a space filled with a liquid called an electrolyte. The sulphur goes through a stage of being composed of eight atoms, then seven, then six, step-wise all the way down to one before it is able to combine with the lithium. The intermediates can drastically slow — or even stall — the battery chemistry.
Preventing this from happening depends on understanding how something called the "dipole moment" changes as the sulphur atoms pull apart from each other. This depends on a quantum property called entanglement.
There is no equivalent of entanglement outside of the quantum world, and when anything more than a small number of particles is involved, calculating its effects quickly escalates into a problem that is beyond classical computers. At the moment, the classical limits kick in with systems that contain around 22 electrons — fairly small molecules, in other words. Beyond this level of complexity, even supercomputers must approximate their answers.
But quantum computers, being composed of quantum entities, have entanglement built in. There is no need to add extra processing power to compute the entanglement effects. And so quantum computation of chemical reactions is far more efficient than anything possible by any other means.
The work presented at CES — and published online — presents, among other breakthroughs, the first ever calculation of lithium hydride's "dipole moment" using quantum hardware. The computation was carried out on IBM's Valencia quantum processor, which has five superconducting quantum bits, or qubits.
This doesn't go beyond what can be done with classical machines — and it's still not clear when that will happen — but it is an important step, Garcia says. In her view, we are currently in what could be seen as a calibration phase that is checking the technology as it matures. "Quantum computing is a really new technology, and this allows us to compare and contrast against known results," she says."Someday we may not have that, so we rely on the fact that these methods were developed correctly in the first place."
The other important part of our progression toward full implementation of quantum computing, she adds, is gaining experience in getting everyone involved to work together despite their very different backgrounds. Going from automotive engineering, through theoretical chemistry, and into quantum algorithm design makes for a long journey, with a lot of potential for communication breakdowns.
"You run into a lot of 'Lost in Translation' moments," Garcia says. "Our team is extremely diverse, and all of these things have to come together."