Supply Chain Concentration is a Materials Problem

Sovereign capability discussions focus on manufacturing. The deeper vulnerability is materials.

Advanced systems that are in aircraft, submarines, satellites and semiconductors all depend on materials with specific properties that cannot be substituted without consequence. Many are sourced from a small number of geographies. Rare earth processing capacity is more concentrated still.

This is a strategic risk that sits upstream of every manufacturing concern. A nation that cannot access the materials its advanced systems require cannot build those systems, regardless of its manufacturing sophistication. The constraint is not assembly. Conventional responses, such as stockpiling and supply diversification are valid short-term measures but they do not address the underlying dependency.

The more durable response is to expand the range of materials that can perform the required function. If the properties that make a specific rare earth element valuable can be replicated by an alternative material - one that is more accessible or more easily synthesised - the strategic dependency is reduced.

Identifying those alternatives is a materials discovery problem. At the scale and speed sovereign capability requires, it is a computational infrastructure problem.

The organisations that solve this computationally will not be waiting for supply chains to fail before they act.

The challenge in rare earth substitution is property matching under constraint. Rare earth elements derive their functional value from specific electronic configurations - particularly the 4f electron shell - which produce magnetic, optical and catalytic properties that are difficult to replicate through compositional substitution alone.

Computational screening for rare earth alternatives requires multi-constraint optimisation across a property space that typically includes magnetic moment, magnetocrystalline anisotropy, Curie temperature, structural stability and synthesisability. No single candidate is likely to match all properties simultaneously. The practical goal is to identify Pareto-optimal candidates: materials representing the best available trade-off across the relevant constraint set, for subsequent physical validation.

High-throughput DFT-based screening of intermetallic compounds, combined with MLIP-accelerated workflows for stability assessment, provides a tractable pathway to this candidate set. The search space for transition metal and lanthanide-free magnetic materials spans tens of thousands of compositions. Physical experimentation cannot cover this space within operationally relevant timescales.

The infrastructure required to conduct this screening at scale includes validated MLIPs trained on physical structures, high-throughput structure generation, property prediction pipelines and secure compute environments. This is what Atomic Tessellator is building.

Sovereign capability in advanced materials does not begin with a policy decision or a procurement contract. It begins with the ability to discover what you need before you need it.