Enabling the hydrogen economy through quantum computing
Hydrogen promises to be the fuel of the future – it stores energy well and is carbon neutral. However, using hydrogen to generate energy efficiently depends on a quantum mechanical understanding of hydrogen chemistry and its catalysed reaction with oxygen. These reactions happen in a device called a fuel cell, a core ingredient of the future hydrogen economy.
The fuel cell predates the internal combustion engine by 24 years but, despite being used on Apollo missions to the moon, it has never seen mainstream adoption. Though highly efficient, fuel cells, like batteries, require catalysts and electrolytes – materials that promote the controlled reaction of hydrogen and oxygen – to rapidly combine hydrogen and oxygen and convert them to energy. The high cost and low durability of these components is the biggest barrier to fuel cell commercialisation.
As an example, the vast majority of fuel cell technologies have adopted platinum as the catalyst of choice. While expensive and sensitive to fuel impurities, platinum allows fuel cells to achieve a better performance-by-weight than petrol-based solutions while emitting only water. Developing this technology to be cheap enough for large-scale commercialisation is a big ask, however, as success depends upon understanding the complex chemistry and materials properties of these systems.
Understanding the inner workings of fuel cells in the quantum regime could have a transformative effect on their adoption and the broader transition to a hydrogen economy. But there is a fundamental challenge in realising this understanding: the techniques scientists have developed over decades for predicting molecular behaviour are hindered by the difficulty that classical computers cannot efficiently simulate the quantum mechanics which underpins chemistry. Without this microscopic understanding, finding better catalysts and electrolytes is Edisonian — many hard experiments done with constant refinement, rather than a broad exploration powered by computing.
Fortunately, these quantum problems can be attacked with a new approach to computing. Quantum computers use the unique properties of the quantum states of single ions, electrons or photons (phenomena such as superposition and entanglement) to store and manipulate data. This means that complex calculations that simulate huge numbers of different chemical configurations can be solved using a quantum computer that naturally operates in this large space. A large-scale quantum computer would render these intractable problems feasible and open the door to a much more efficient search for better catalysts and electrolytes.
No-one has yet built such a large-scale quantum computer. Specifically, today’s quantum bits are error-prone, limiting the number of operations which can be performed reliably. Improvements in the stability and scale of quantum hardware are a key part of addressing this problem, but just as important is implementing techniques to allow quantum computers to detect and correct their own errors, a process known as quantum error correction. Several methods for doing quantum error correction have been demonstrated in theory, but implementing these methods on a real quantum computer requires a device to solve the complex decoding problem to identify errors billions of times per second. Overcoming these scientific and engineering challenges will get us to so-called ‘fault-tolerant’ devices, quantum computers which will enable scientists to understand properties of catalysts and other materials.
Riverlane, together with Johnson Matthey and quantum hardware manufacturer Rigetti, has been working for the past year on some important first steps that will enable quantum simulations in the near future and solutions to the decoding problem to enable large-scale simulations in the longer term. We recently shared the first results of this collaboration via a live demonstration on a Rigetti quantum computer, where a small example of the relevant chemistry was selected by an audience in London and computed in California.
To achieve this demonstration requires writing quantum code for this highly specific application, and requires substantial expertise and time to achieve. The Riverlane, Rigetti, Johnson Matthey team crosses disciplines in ways that allow us to effectively provide a solution by combining expertise across all areas of the problem. This enables chemists to write simulations in the language of a quantum computer without needing to get lost in the details of how the circuits are designed.
These simulations are at an early stage but with quantum computing following its own Moore’s Law (performance doubling every two years), the promise of a full understanding of the forces and interactions involved in catalysis of hydrogen is within our grasp over the next decade. The most critical next steps will be getting to the error corrected regime by both building better quantum bits and by solving the decoding problem.
The interaction of a hydrogen molecule on a platinum catalyst may sound esoteric, but its importance cannot be overstated. Better understanding of hydrogen-platinum chemistry would give us insight into the inner workings of fuel cells and allow us to understand how molecules interact with catalytic surfaces. This in turn could lead us into a world where we are purposely designing the materials of tomorrow rather than trying to empirically discover what works.
Furthermore, the application of quantum computers could be a breakthrough in many areas that will accelerate the net zero economy. Their use could lead to the design of new systems for capturing CO2, the invention of carbon-free fuels, alternative processes for the currently high carbon-emitting manufacture of ammonia, or a breakthrough in the efficiency of generating hydrogen through electrolysis – or even directly using light, as plants do.
The transition to the hydrogen economy will need generation, storage, and usage of hydrogen as a fuel – all areas that the quantum effects of chemistry play a role in, and all places where quantum computers can dramatically accelerate discovery. Taken together, this could enable a world where we could realistically use renewable energy to make hydrogen and fuel cells to turn it into electricity on demand. Today, that round trip is economically unviable. But tomorrow, the deep understanding enabled by quantum computing can lead us into a highly efficient and affordable hydrogen economy.