Interesting!
As the art points out, additional work will be needed to scale the tech up to make it useful.
How did they figure out it will hold power for 18 years? I guess when the package arrived in China, they measured the remaining charge and calculated the rate of decay, which turned out to be quite low. They probably calculated that it would still have some agreed-upon useable level, say 80%, 18 years hence. You can barely keep unstabilized ethanol-laced gasoline on hand for half a year (ethanol-free can keep a little longer), and long-term storage of diesel fuel can be problematic too, esp. if it gets moist!
If they were at all accurate characterizing it as a "thermoelectric" device, that would mean that
it is in effect a solid-state heat engine, and therefore requires not only a heat source, but also a heat sink, to work. In space, that shouldn't be a problem so long as the sink-side can be aimed away from the Sun and insulated from the heat-source side. (Are there any satellites powered by the thermoelectric effect?)
Does being based on thermoelectric principles mean
the method can be used with other heat sources besides the Sun, such as waste heat from boilers and furnaces or from nuclear power plants or the like? That would seem to have potential uses. Auxiliary and instrumentation/control power for a boiler or furnace might be supplied in this way?
How fast would they charge and how many cycles of useable life (big issues with conventional rechargeable batteries, as anyone with a cell phone or laptop could attest!) would they have?
I could imagine that near-Sun orbiting banks of "batteries" based on this principle could be swapped in and out to power enterprises in the outer Solar system where solar arrays become impractical. This could be a use for
Kerbal Cycler-type orbiting transports, ferrying these modules back and forth between near orbit charging locations and the outer Solar system. Land-based arrays could also find uses, though again in many high-power applications they'd have to be physically transported from sun-exposed charging station to end-use, and the discharged devices collected and brought back. They might become useful on Earth oceanic islands, for example, where conventional energy generation can be expensive and problematic. Perhaps they could be exposed to sun on big rafts, using the ocean as the heat sink. Likewise, there might be possibilities for powering colonies on rocky space bodies, and large satellites including the aforementioned cyclers.