Atomic Scale Material Design and Manufacturing

Amphiform builds materials from the atom up. Our Amphimaterials enable programmable structuring of functional groups at the atomic level – giving direct control over real-world properties.

| The Problem:

Material performance has plateaued. Not because of physics, but because of architecture.

In virtually every high-performance material (batteries, fuel cells, semiconductors, extreme-environment components) only a small fraction of atoms do meaningful work. The rest – the “deep” atoms – are structurally necessary but functionally inert. This is the defining constraint in modern material science.

The clearest example is catalysis in fuel cells – this is how our research started. The active sites that drive the core chemical reaction exist only at the material’s surface. Every atom beneath it – often over 99% of the expensive catalytic material – contributes nothing. The famous result of decades of research in this area is nanoparticles – physically increasing the percentage of surface area. But a deeper constraint remains: the reaction requires simultaneous access to electrons, protons, and fuel at the same point in space. It is impossible for classical materials to deliver all three through their bulk. The surface will always be the bottleneck.

The same structural limitation is observed across energy storage, power electronics, extreme-environment materials and so on. The market has worked around the constraint rather than eliminated it.

The real constraint isn’t elements (we’re not even close to the limit of what’s available on Earth). It’s that no one has designed the architecture.

If you could place every atom with a defined function in a specific place (as opposed to the complete randomness of current approaches) and allow access to it, you could convert a surface phenomenon into a volumetric one. A material where the entire bulk participates, not just its skin.

This requires two things working together: precise placement of functional groups within the building molecular blocks, and atomic-scale structural control of those blocks. Each has been achieved in isolation. Nobody has done both – the challenge (and our breakthrough) is making them work together without one destroying the other. Together, they enable materials where properties become programmable outputs – not something that emerges from bulk composition.

| The Breakthrough:

Programmable Matter

Amphiform moves material science from mixing chemicals to atomically building structures. Organic chemistry provides enormous flexibility in designing functional groups and consequently the “functions” – properties of the respective groups. However, it lacks any structural control at the nanoscale.

Inorganic techniques provide atomic-level structural precision, but lack chemical flexibility.

By combining these approaches, Amphiform enables:

• precise placement of functional groups defining properties
• controlled spacing and geometry at the atomic scale
• structurally defined transport pathways

This allows us to design materials where both structure and function are engineered together, rather than optimised independently.

The result changes how we develop materials:

Materials can be designed with predetermined properties, rather than discovered through trial and error.

| What we’re building:

The same platform can be directed at semiconductors, extreme-environment aerospace components, or any application where material properties are the limiting variable.

|Closed-loop learning:

Our second critical advantage is the speed of iteration.

Traditional materials development is slow, expensive and requires massive trial runs.

Amphiform’s approach allows a closed-loop development system:

Rapid synthesis (days, using agents and robotics) → real-world testing → iterative optimisation.

As a result, materials development begins to resemble modern software or drug discovery workflows.

| Why now:

The window is open now. It won’t be in 5 years.

Manufacturing:

First, the underlying manufacturing capability has finally matured. The manufacturing techniques required for this architecture have reached a point where they’re compatible with real production pipelines, not just research demonstrations. This wasn’t true earlier.

Second, the giants of the world can’t follow. Toyota and Honda have invested decades and billions into hydrogen fuel cell architectures. Accepting that the architecture is wrong means writing off the R&D spend, IP portfolio, supplier base etc.

Geopolitics:

Global energy demand is entering a new phase, driven by exponential growth in data centres and AI compute, electrification of transport and industry and increasing need for distributed, high-density power sources.

This demand is increasing and the scaling of classical energy production ways will not satisfy it – especially with the conflicts in Middle East. We need higher power density, faster development times and flexibility.

This creates a strong pull for fundamentally new material architectures that unlock higher performance without proportional increases in cost or complexity.

| Why us:

My research into functional materials beyond Li-ion architectures began in 2021. It started as part of the National Chemistry Olympiad in Ukraine, which I won twice. That work has then evolved into a multi-year deep focus on the hardest-to-develop use-case of functional materials: atomically engineered fuel cell materials.

The scientific foundation that followed was intentional:

A degree in Computational Quantum Chemistry at Oxford helped with the precise understanding of how materials behave at the atomic level. And most importantly, how to develop them from first principles rather than by trial and error. 

A post-doc research project at AstraZeneca (during undergrad) introduced experience with large-scale industrial research and ML-driven approaches to organic chemistry. Work at Enamine (the largest organic chemistry lab in Europe) helped with practical experience in synthesis at scale.

This is unusual: instead of background in research or the industry, I spent years at the intersection of both – with both being directly useful for Amphiform.

On execution: co-founder and CTO of Lon: (consumer tech, hardware), where I was responsible for manufacturing, engineering, and sales. The company scaled from zero to $1M ARR in 2025, $6M projected for 2026 and a successful exit underway. 

The science founding team: professors from Oxford and Cambridge in nanocatalysis and nanoparticle synthesis. These are world-leading researchers in their respective fields.

| The closest approach now:

See: https://www.forgenano.com/

Won a $100M U.S. Department of Energy grant to build classical 2D/surface ALD materials – this is state of the art in ALD applications with a lot of published research.

Their platform applies the atomically precise inorganic coatings to particle surfaces: a 2D intervention that marginally improves performance. It does not (and cannot) reach the bulk of the material – the fundamental constraint of the surface approach, and the current standard. Amphiform on the other hand operates through the full volume of the material and offers the step change.

The fact that $100M in government capital is flowing into the limited 2D approach is the signal of how large the demand for new energy and defence materials is. None of the current spending is pointed at the volume.