Here's some details on the plutonium alternatve, although the fuels are still radioactive, so Luddite moonbats will still object. No moving parts makes this an ideal mechanism for powering ion thrust spacecraft on missions to the ocean worlds, Europs, Enceladus and Titan as well as Kuiper belt and other journeys. There's a pdf with fuel details for the curious.

The development of the new MEMS devices and micro electronics in the 40 nm technologies provides an excellent background for the production of the electric power harvesting and conversion devices embedded in the fuel. The new nano-structured materials may be produced as radiation energy harvesting tiles that are free of actinides, using them for harvesting the energy of radioactive sources and controlled fusion devices, or may include actinides in the structure achieving critical or sub-critical accelerator driven nuclear reactor assemblies. Another predictable advantage of the nano-structure is the property of self-repairing and self-organizing structure to compensate the radiation damage and improve the lifetime. Due to the direct conversion the power density of the new materials may increase ... driving to miniaturization of nuclear power sources and reductions of the shield weight. At these dimensions and power densities of few thousands horse power per liter the nuclear power source becomes suitable for mobile applications as powering trains, strategic airplanes, etc. These new developments may drive to the production of high power solid-state compact nuclear battery for space applications....

A prototype of the heat engine, called the Johnson Thermoelectromechanical Energy Conversion System, or JTEC, will be ready in a few months. It will convert heat to electricity at rates reaching just under 40 percent of the maximum theoretical efficiency available in an engine operating between two temperatures—the Carnot efficiency. The former U.S. Air Force and NASA Jet Propulsion Lab engineer says his group’s aim is to produce a commercial version whose efficiency can approach 85 percent of the Carnot ideal. Such a device would be capable of converting 66 percent of the available thermal energy into electrical energy.

In contrast, photovoltaic devices have net conversion efficiencies in the teens and thermionic (or thermoelectric) chips reach only a little higher than 20 percent

Like all wicked pissah stuff, commercialization is a decade away.