Betavoltaic (Nuclear) Batteries: Decade-Scale Power for Sensors and Space

Figure 1. An in-pocket nuclear battery developed at NUST MISiS.
Credit: National University of Science and Technology MISiS

From lab curiosity to micro-power products

The selling point is disarmingly simple: a battery you don’t replace. Betavoltaic cells turn the gentle trickle of particles from a radioisotope into electricity inside a semiconductor, much like a solar cell turns light into power—but here the “sun” is a tiny, sealed source that fades over decades rather than hours. That makes them natural fits for sensors you bury in bridges, spacecraft that nap through deep space, and memory keep-alives in hardware that must never lose its brain. Recent demonstrations and products have moved this category from curiosity to procurement line item.

Figure 1 shows the in-pocket nuclear battery from Russia’s National University of Science and Technology (NUST) MISiS and its pitch is: shrink the package, raise the energy density, and make the form factor friendly to designers. MISiS’ prototypes used nickel-63 as the source and diamond-based semiconductors to harvest its electrons—an example of how materials science is nudging output upward while keeping radiation tightly contained. The power is micro- to microwatt-scale, but the life stretches to decades because nickel-63 decays so slowly.

How does it work in practice? In a betavoltaic device, a beta emitter (often tritium or nickel-63) releases high-speed electrons. Those electrons strike a semiconductor junction and create charge carriers that a circuit can harvest. The physics mirrors photovoltaics; the constraints are different. Output is tiny but continuous, and it declines predictably with the radioisotope’s half-life. Tritium’s half-life is about 12.3 years, which is why vendors quote 20+ years of usable power before you fall below design thresholds; nickel-63’s century-scale half-life stretches the window further but with different power trade-offs.

The business question is where that trickle is worth more than a rechargeable alternative. Three markets stand out.

Industrial and critical-infrastructure sensors. Think strain gauges inside concrete, corrosion monitors in pipelines, or seismic beacons along fault lines—places where swapping a coin cell means a lane closure, a dive team or a helicopter. For these, decade-scale trickle beats high bursts and short lives. A U.S. maker, City Labs, has shipped tritium betavoltaics for years into memory backup and sensor roles, emphasising stability across extreme temperatures—useful where lithium cells falter or where service calls dominate lifetime cost.

Space and high-temperature environments. Traditional RTGs (radioisotope thermoelectric generators) make watts to kilowatts by turning heat into electricity; betavoltaics aim lower, powering electronics directly without heat cycles. That matters for small spacecraft, long-sleep probes or instruments that sit near hot surfaces where chemical batteries degrade. NASA has flagged betavoltaics as candidates for unattended, long-duration operation, and its own history books carry a quirky precedent: promethium-147 betavoltaics once powered early pacemakers, delivering tens to hundreds of microwatts for years. The modern stack replaces those legacy semiconductors with higher-efficiency materials and better shielding, and it packages the devices for aerospace qualification.

Medical and security niches. Not every implant needs continuous watts; some need guaranteed micro-power without surgery for a decade. Likewise, secure modules in defence assets value a non-replaceable, tamper-evident source that never goes flat. These are small markets, but they pay for predictability and certification, not headline wattage.

Safety and perception are the gating items with non-specialist buyers, so the narrative must be plain. The beta radiation used here is weak and stopped by millimetres of material; in tritium devices, the particles can be blocked by a sheet of paper. The source is sealed and embedded behind multiple barriers; the risk is not external exposure but leakage, which is the same engineering concern you already manage for lithium or lead systems—containment, certification, end-of-life take-back. The devices are regulated accordingly, and reputable vendors publish handling and disposal procedures alongside the usual electrical specs.

On performance, the key to credible adoption is honesty about power levels. Betavoltaics supply microwatts, sometimes less, at a few volts. That is plenty for a memory keep-alive, a timer, or a low-duty sensor that wakes briefly, writes a number, and sleeps. Designers built for bursty radios will pair a betavoltaic with a small capacitor or supercap: the nuclear cell trickle-charges the buffer, the system transmits, then it recharges at leisure. IEEE Spectrum’s recent review puts it succinctly: use a semiconductor to turn each emitted particle into thousands of carriers, run at usable voltages, and design the load profile around a steady trickle rather than a sprint.

Cost and supply chain are where this turns into a real industry. Tritium is produced at modest scale for self-luminous signs and research; nickel-63 requires reactor time and processing. Neither is on supermarket shelves, and both demand chain-of-custody and licensing that many electronics firms have never touched. That is a barrier—and a moat. Companies that master source procurement, encapsulation, and paperwork will look more like regulated component vendors than trendy startups. The upside is pricing power and sticky, long-lived contracts; the downside is lead times counted in quarters. The right posture is to sell modules and multi-year support, not just bare cells, and to own end-of-life logistics so customers can hand the problem back when devices retire.

Technology is moving on two fronts at once. Materials researchers are experimenting with diamond and wide-bandgap semiconductors that shrug off radiation and heat while converting more of each particle’s energy into current. Others are exploring architectures that embed the source inside the semiconductor to lift efficiency, especially for nickel-63 and carbon-14. Meanwhile, commercial houses continue to refine tritium-on-semiconductor stacks that are simple, rugged and certifiable today. The R&D and the shipping product are converging, but they are not the same thing; design teams should pick what can be bought, qualified and insured now, and keep an eye on the papers for the next design cycle.

Where does this leave buyers in 2025? With a clean brief: treat betavoltaics as always-on power companions. Use them to keep memory alive, clocks honest and sensors patient; add a buffer for bursts; and write a disposal clause into the contract on day one. If your assets live where trucks struggle—underground, underwater, on orbit—decade-scale micro-power is worth more than headline watts. The MISiS “in-pocket” prototype is a useful symbol: the category is shrinking, hardening and moving toward designer-friendly packages. What turns that symbol into business is the boring excellence of qualification, supply, and service.

Fusion, satellites, pipelines, medical devices—these systems all fail more from neglected components than from glamorous cores. A battery that refuses to quit is not a headline; it is a habit. Betavoltaic vendors who make that habit easy to buy—clear data sheets, predictable lead times, take-back plans—will own a quiet but durable corner of the power market for decades.