Energy

Storing it, burning it, and getting it back out of waste heat.

Four strands that share the substrate rather than a market: a battery whose case does structural work, a molecule-first approach to fuels, converting waste heat into something useful, and a fusion first-wall built around being repairable.

Storage and conversion
Every figure below is a design target. Nothing has been built or measured.
Silixon Battery
Structural pack

A battery whose enclosure is load-bearing, thermally instrumented and chemically active in its own right. Graded filler zones next to the cell faces form in-plane thermal rails, and thin interlayers are tuned to produce a mechanically robust solid-electrolyte interphase that suppresses dendrite growth. The structural variant carries load and current at once, so the pack stops being dead weight bolted into a frame.

290–380 Wh/L · 210–280 Wh/kg · 9,000–16,000 cycles · under 0.7% per day self-discharge · targets
Thermophotonic conversion
Waste heat to power

Modules that take waste heat and either radiate it in a controlled spectral band or convert it into usable power, rather than moving it somewhere else and calling that cooling. It is the same phonon-to-photon mechanism the thermal work rests on, aimed at recovery instead of rejection.

On-package integration intended · shares physics with the thermal programme · concept stage
Silixon PV backskin
Solar veneer

A radiant-glaze veneer behind a photovoltaic cell, passing sub-bandgap energy through to a thermophotovoltaic layer instead of letting it become the heat that ages the cell. Solar panels lose efficiency and lifetime to their own waste heat; this points the substrate’s emission mechanism at that problem.

Concentrator-PV efficiency target 40% · no gate scheduled
Embedded power rails
On-panel distribution

Low-loss conductive rails and harvesting elements formed into structural panels, so power can be routed, harvested and serviced on the same part that carries the load. It follows directly from the board work — once a substrate can hold sealed conduits, a wall can be a bus.

Corrosion-resistant metallisation · concept stage
Fuels and working fluids
Molecule-first: engineered microbes and enzymes for assembly, catalytic chemistry for finishing, and target molecules chosen for how they burn and what they drop into rather than for headline energy density.
Lab Gas
Drop-in hydrocarbon

An engineered hydrocarbon designed as a drop-in replacement for gasoline — preserving compatibility with the engines and infrastructure that already exist, while changing where the molecule comes from. The bet is that upstream origin is easier to change than the entire vehicle fleet.

Engineered volatility · lifecycle carbon as a design constraint · molecule selection stage
Hardin Gas
Synthetic family

A synthetic hydrocarbon family emphasising tunable combustion and industrial synthesis routes compatible with the laboratory’s own process assumptions. Where Lab Gas targets compatibility, this targets designed reactivity.

Designed molecules, engineered reactivity profiles · concept stage
Methixon
Working gas

A circulatory working gas for thermal transfer and reuse in closed loops — chosen for heat capacity, chemical stability and how many cycles it survives rather than for combustion behaviour. It is the fluid the rest of the thermal architecture would move.

High heat capacity · closed-loop reuse · concept stage
Trifuel
ISRU-compatible family

A flexible fuel family designed to be synthesised off-world from local resources, with Mars in-situ resource utilisation as a constraint from the beginning rather than a retrofit. Compatible with reusable vehicles and conventional engine architectures.

Local synthesis, multi-mode combustion · concept stage
SIFR
Silixon Icarus Fusion Reactor

A fusion first-wall concept built around maintainability rather than plasma performance. The wall is a set of petal cartridges a robot can swap in hours, which shifts the operating economics from month-scale replacement to days-scale. The most speculative entry on this page, and it is listed rather than promoted.

Concept stage · no simulation posterior published