The DLR Institute for Future Fuels wants to make renewable resources available to replace fossil resources: Air + water + a huge amount of renewable energy that can be used to produce such future fuels cheaply and on an industrial scale. The development of new materials for extraction and storage is a major challenge that quantum computers can help with. In this interview, Ulrich Biedermann, head of the QCMineral work package QUADRANT, explains the development goals and how the QML start-up Tensor AI(recently awarded an R&D contract) is supporting research and development.
Why is your approach to developing new materials for energy generation and storage in QCMineral exciting?
We are specifically looking for alternatives to established processes such as photovoltaics in combination with electrochemical water splitting or CO₂ reduction. Efficiency is a key driver. Thermochemical cycles are particularly promising here because they can theoretically achieve efficiencies of up to 40 per cent.
What are thermochemical cycles and how do they achieve high levels of efficiency?
Thermochemical cycles are processes in which heat – usually from solar energy or industrial waste heat – is utilised to allow chemical reactions to take place in several steps. Substances are cyclically converted and regenerated again at the end. In this way, heat can be efficiently converted into chemical energy, for example to produce hydrogen or other fuels, without consuming the materials used.
The key advantage is that solar radiation can be completely converted into heat. This heat can then be converted relatively efficiently into chemical energy – provided that the optimum material is found. This is precisely the challenge that we are currently working on intensively.
What makes the search for materials so challenging?
In order to evaluate suitable materials, we need to know both the thermodynamics, i.e. the achievable efficiency, and the kinetics, i.e. reaction barriers, very precisely. To do this, we need highly precise energy calculations – and this is precisely where today’s standard methods reach their limits. The method commonly used for this, density functional theory (DFT), is not precise enough for many systems. Systematic errors occur particularly in the case of transition metals with strongly correlated electrons. The necessary chemical accuracy cannot be achieved with this method.
Are there no more precise classical methods?
These exist, for example the more precise wave function-based methods. However, they are too computationally complex for realistically large and complex systems. In practice, they cannot be applied to the material systems that are relevant for real processes.
And this is where quantum computers come into play?
Exactly, quantum computers are particularly well suited to describing the quantum mechanical properties of electrons, especially correlations. It is precisely these effects that are decisive for material properties. In principle, these phenomena can be modelled more naturally and efficiently with qubits than with classical computers.
What is the significance of tensor networks?
The mathematical formalism of tensor networks, which our contractor Tensor AI Solutions contributes, is particularly powerful. It allows quantum properties such as entanglement and superposition to be described efficiently and utilised in a targeted manner. This is a key building block for making progress with quantum computing in material simulation.
What are the exciting material properties that can be developed in this way?
In addition to efficiency, we also optimise the service life of the materials. To do this, we calculate properties such as thermal and chemical volume expansion as well as mechanical stability. At the same time, the materials must be environmentally friendly, i.e. non-toxic, without critical raw materials and economical to produce.
However, the QCMineral project is not just about extracting energy, but also about storing it.
Efficiency and, of course, energy density also play a key role here. The decisive factor is how much high-temperature heat can be stored per volume or mass. If we store chemical energy as well as sensible heat, the energy density increases significantly and the heat losses through insulation decrease.
What is the challenge here?
The reaction or redox energy of the material must be precisely matched to the respective working temperature. Each application has its own temperature range. We therefore have to optimise the active material in a targeted manner – and in turn calculate the energies very precisely in order to be able to predict the best material composition. Quantum computers can be a great help here.
