Energy harvesting from the invisible spectrum

What "invisible radiation" means

Every square metre of the Earth's surface, at all times of day, is traversed by a continuous flux of particles and fields that carry no visible light and leave no track detectable by the human eye. The largest components are:

Solar neutrinos — approximately 6.5 × 10¹⁰ neutrinos per square centimetre per second, streaming from the Sun's fusion core and passing through the Earth in eight minutes
Cosmic neutrino background — the relic population of roughly 336 neutrinos per cubic centimetre, filling space at temperatures of about 1.95 K, a fossil of the first second after the Big Bang
Reactor and geological antineutrinos — emitted by nuclear power plants and by radioactive decay in the Earth's crust and mantle
Atmospheric muons — secondary particles produced when cosmic rays strike the upper atmosphere, with sea-level flux of about 170 per square metre per second
Ambient electromagnetic fluctuations — natural (atmospheric sferics, solar radio, galactic synchrotron) and anthropogenic (broadcast, telecommunications, power-line harmonics) — spanning frequencies from DC to hundreds of gigahertz
Thermal infrared background — blackbody emission from surfaces at ambient temperature, at wavelengths between 3 and 100 micrometres

Together these constitute a radiation environment that is dense in the thermodynamic sense but dispersed in the engineering sense: there is energy everywhere, but extracting any meaningful fraction of it requires either a very large target volume or a physical coupling mechanism that aggregates contributions across multiple channels.

Why it is an engineering frontier

Harvesting energy from the visible and near-infrared bands — photovoltaics, concentrated solar thermal, solar water heaters — is a mature industry with well-characterised physics and established supply chains. The invisible-spectrum analogue has no equivalent maturity. Each of the components above carries, per unit area, much less energy than daylight, and each couples to matter through a different physical mechanism. A successful invisible-spectrum harvester therefore cannot simply be a scaled-down photovoltaic cell; it must be a heterogeneous device that integrates multiple coupling channels into a single charge-collection architecture.

Comparing available power densities

To a first approximation, the energy flux crossing one square metre at sea level in each channel is:

Sunlight (for comparison): ~1000 W/m² at solar noon, ~200 W/m² 24-hour average, ~100 W/m² annual average at mid-latitudes
Atmospheric muons (kinetic energy flux): ~0.1 W/m² — low compared to sunlight, but present day and night, at all latitudes, and largely independent of weather
Ambient RF (0.1 – 3 GHz band): ~10⁻⁴ – 10⁻² W/m², strongly location-dependent — higher near urban transmitters, negligible in remote regions
Thermal gradient (across device stack): Can be 1 – 100 mW/m² for a 1 K gradient across 1 cm, scales linearly with gradient and thermal conductivity
Neutrino kinetic energy flux (solar): ~10⁻³ W/m² total kinetic energy; with coherent-scattering cross-section the deposited fraction is many orders of magnitude smaller

The key engineering observation is that while no single invisible channel is competitive with solar photovoltaics on a per-square-metre basis, the channels are complementary in their time profiles — muons at night, RF on demand, thermal gradient whenever temperature differences exist, neutrinos continuously. A device that couples to all of them can operate continuously, in locations and conditions where photovoltaics cannot.

Four channels, one equation

The Schubart Master Equation P(t) = η · ∫_V Φ_eff(r,t) · σ_eff(E) dV is a single integral expression that aggregates all four channels. Each channel enters Φ_eff as one term of the sum and brings its own σ_eff(E). The efficiency η captures the device-specific question of how much of the deposited energy is actually collected as electrical output.

The equation is agnostic about which channel dominates at a given location. At a remote rural site with weak RF activity, the neutrino and muon channels dominate. In an urban environment with strong broadcast activity, the RF channel contributes substantially. In a location with a large thermal gradient (near a thermal power plant, for example), the Seebeck contribution may dominate. The value of a unified framework is that the same device can operate across all these environments, with the dominant channel self-adjusting to whatever ambient conditions are present.

Historical context

The idea of harvesting ambient energy from sources other than sunlight is not new. Research on radiofrequency energy harvesting dates to the 1960s, with growing contemporary interest for powering low-power Internet-of-Things devices. Thermoelectric generators have been deployed in radioisotope power sources since the Apollo era. Cosmic-ray-based energy extraction has been periodically discussed in the literature since the 1970s, though no practical device has emerged from those efforts.

What distinguishes the current generation of work is the coupling of these disparate channels into a single integrated nanostructure — made possible by the availability of large-area graphene deposition since roughly 2010, and by the engineering maturity of low-threshold charge-collection electronics developed for direct-dark-matter-detection experiments. The combination of these two capabilities — active-area material fabrication at scale, and charge readout at the keV-recoil level — is new in the last decade.

Economic niches

Photovoltaic power at utility scale now competes with fossil generation in levelised cost terms. Invisible-spectrum harvesting, at current conversion efficiencies, is not yet at that threshold for large-scale grid supply. But there are niches where it is already interesting:

Continuous low-power applications — remote sensors, indoor IoT nodes, devices that cannot rely on daylight
Base-load complement to photovoltaics — a device that operates 24 hours per day can reduce battery requirements in hybrid installations
Deployment in environments hostile to photovoltaics — underground, underwater, under dense cloud cover, polar winter
Space applications — where RF and cosmic-ray fluxes are higher than at sea level and where the sun may not always be available

Open questions

Like any emerging energy-technology field, invisible-spectrum harvesting has a wide gap between laboratory demonstration and industrial deployment. The questions to be resolved over the next decade include:

Independent cross-laboratory reproduction of reported output levels
Quantitative per-channel attribution — how much is neutrino, how much muon, how much RF, how much thermal
Scalability of η as active volume V increases
Long-term stability of the multilayer stack under real-world deployment conditions
Integration with conventional storage, power-conditioning, and grid architecture

None of these is a physics question — the underlying processes are all established in the standard particle-physics and condensed-matter literature. They are engineering questions, and the answers will come from the same kind of iterative device-characterisation work that drove photovoltaics from the 1970s onward.