It is possible using nuclear decay to produce heat, and then converting this heat to electricity using thermoelectric generators (TEG). For 10 kg, the power output of such a nuclear generator will be rather disappointing however.
Thank you for your answers, I understand from what you say to me that the main problem with the mass is the heat/ electricity conversion, not about the nuclear reaction... is it?
That is correct, the electric output depends on the temperature difference between the decaying nuclear material and the surrounding environment, as well as the efficiency of the TEGs which is usually around 8 - 10%.
I was see somewhere that it is use somewhere in sattelite because it is more reliable and have an higher life time. 10% is not so bad, I am not sure that there is a much more efficiency in classic nuclear central...
Indeed, they are primarily used in situations where a relatively high electric power is needed for a prolonged period of time (i.e. dozens of years). Satellites and nuclear lighthouses are examples of that, but the radioactivity to make them work is prohibitive for most other applications.
That is correct, the heat is produced by decay of a radioactive isotope, so the more energy you want to produce, the faster the decay must take place and thus the more radioactive the isotope has to be. Commonly used isotopes like Pu-238 have relatively short half lives, and are extremely radioactive. Fortunately, alpha and beta radiation will be absorbed by the shielding and thus pose no risk to the environment. Isotopes emitting gamma radiation are completely unsuitable because no appropriate shielding can be provided with 10kg of mass.
Aside from the obvious radiological risks, most of those isotopes are extremely toxic too (including Pu-238) and exposure will lead to severe heavy metal poisoning. Therefore, a 10kg TEG using nuclear fuel would be possible to construct and safe to use, but any breach of the hull (i.e. after its intended life time) would immediately result in INES scale events.
It's definitely possible, but aside from security and recycling there is also a supply issue: because of their hazardous nature, most suitable isotopes are no longer actively produced. Most radio isotope thermal generators in use today (i.e. NASA spacecraft) are running off Russian stock from '80 and '90, and that stock is reportedly running out.
If you'd like to know more about the possibilities of RTGs, then look up the Russian nuclear light house articles. The aftermath (what happened to the RTGs) is an interesting story: some even ended up at scrap yards etc.