Matter Machine - why we invested .

There’s a reason there was a Bronze age: Bronze.

The history of human progress is a history of materials. It’s why we name our eras after them: the Stone Age, the Bronze Age, the Silicon Age. Behind every civilizational leap is a step change in our use of materials. The question is what comes next. The answer matters now more than ever. Trade wars, climate change, electrification, and the race for AI compute all depend on it.

We invested in Matter Machine Technologies (MMT), an AI-guided materials discovery company spun out of the University of Liverpool, to help expand that frontier. MMT focuses on crystalline inorganic materials, complicated to predict and central to new battery chemistries, rare earth magnet replacements, solar photovoltaics and the frontier of computation.

MMT is built on a world leading team, from a world leading department

MMT is built on twenty years of computationally-enabled materials discovery, led by three-time Royal Society medallist and one of the world’s foremost materials scientists Professor Matthew Rosseinsky OBE FRS at the University of Liverpool’s globally recognised Materials Innovation Factory.

Professor Rosseinsky's team is one of a small number to publish, in leading peer-reviewed journals, the AI-enabled discovery of new materials that were not just predictions, but also physically created samples which were shown to deliver state-of-the-art performance. For example, their Li7Si2S7I (Science, 2024) is an entirely new structural family of solid lithium electrolyte, with measured conductivity that is high enough to replace the liquid electrolytes in today's lithium-ion batteries. The material also conducts in an unexpected way, overturning the established rules for designing fast-ion conductors, and creating the possibility of a whole class of new materials. These physically realised discoveries distinguish the Matter Machine approach from the many AI materials discoveries that exist only in silico, and from AI tools that create candidate materials by largely recombining what is already known. Here, AI guided by expert chemists found a material in genuinely new chemical territory, far beyond the chemical space which existing models were built on.

Visiting the £100m Materials Innovation Factory, we were blown away by the quality of the department. Liverpool’s facility is at the forefront of the nation’s novel material research. We were further impressed by the range of commercial partners that are working with Professor Rosseinsky and his colleagues in the Materials Innovation Factory. These include: one of the world’s largest consumer goods companies, a multinational leader in catalysis technologies, a pioneer in solid oxide fuel cells, one of the world’s largest glass manufacturers and one of the world’s largest energy technology companies. Together, these partnerships span fuel cells, electrolysers, catalysis for sustainable manufacturing, and functional coatings on glass for net zero buildings and solar cells. Professor Rosseinsky’s academic research extends across a broader set of areas, including solar photovoltaics and superconducting materials. MMT is not building from a standing start.

Complementing the technical team, Dr Gordon Sanghera CBE FRS has also joined MMT as launch CEO and Chair, following his time at Oxford Nanopore Technologies (ONT). As co-founder and CEO Gordon took the business from a from a lab scale university spin out to a +£bn FTSE-listed genomics platform deployed across more than 100 countries. At IP Group we have invested in ONT across the life of the company, from seed to IPO, and have followed Gordon across the entire journey. We’ve personally seen the difference he makes in building a world leading business. He is one of a handful of entrepreneurs in Europe who have taken a deeptech university spin out from foundation through to unicorn status.

A different approach: AI enabled, lab integrated, expert guided

Material discovery will be accelerated by the ongoing phase change in AI where everyday language has become interoperable with coding, autonomous agents and the world’s existing knowledge base. However, while this will unlock further progress in search, modelling and robotic lab management, it does not directly translate into new viable materials.

An in silico candidate is not a functioning material. Computation cannot yet reliably predict whether a material can be made, or how it will perform once it is: synthesis depends on reaction pathways and kinetics that models capture poorly; competing structures often differ by less than the models' own margin of error; and real-world performance is frequently dominated by defects, disorder and interfaces that idealised simulations leave out. A functioning material is something that has been synthesized, measured and confirmed to behave as predicted. Core to the foundations of Matter Machine is the integration of the full discovery stack into an iterative loop – spanning AI-guided candidate generation to automated synthesis and precision characterization. MMT’s discovery journey is led by expert chemists, not computer scientists.

In recent years more than $3 billion has been invested in platforms that generate millions of in silico candidate structures; but no AI materials platform has yet delivered a generationally transformative material at scale. We believe Matter Machine has the foundational team, setup and funding base to deliver on this.

If you are working in this space or want to learn more, please get in touch with Ben Murphy and Arielle Schacter.

IP Group exists to take the UK’s best science and scale it into lasting companies. We are proud to be part of Matter Machine’s journey alongside Northern Gritstone, Parkwalk, Norrsken and LYVA Labs.