Cold fusion? if its even possible then we don't know how to do it.
I could certainly achieve hot fusion, and build what is basically a neutron flux generator, and a miniature star in a bottle (not a glass bottle, a wire grid) for a type of electrostatic confinement fusion known as a Farnsworth fusor. Fusing deuterium-tritium fuel mixtures (deuterium and tritium are by far the most common/well known isotopes of hydrogen, regular hydrogen being known as 'protium' when distinguishing it as an isotope is needed, the other, less well known would be an electron orbiting a positive muon, called muonic hydrogen, an element in its own right, that for the ortho-configuration of hydrogen (H has two different configurations, IIRC its related to the nuclear spin of the electron orbiting the proton which forms the nucleus, these being ortho- and para-hydrogen), has a different half life compared to para-muonic hydrogen.
Although its still damn short, about 2 microseconds, I would bet it has interesting chemical properties though with the muon evaporating, and leaving a bare electron behind, I bet it could be used to embed charges in electrets via hydrogen embrittlement, only for the muon to decay, vanish, and leave the electret with a huge excess charge.
And of course, reductions with solvated electrons, just by firing a muonic hydrogen beam into say, anhydrous ammonia, with a particle accelerator (I say 'just', the concept is simple enough but putting it into practice is rather more common.) and neatly, muons can also catalyze nuclear fusion, called, surprisingly enough, muon-catalyzed fusion)
While with current technology as has been applied in the construction and design of the Farnsworth-Hirsch fusor, break even, or positive net gain of power has never been realized due to large conductive losses on the grids, its certainly possible to initiate a fusion reaction with a deuterium-tritium fuel mixture, and confine it with an electrostatic potential across the grids of the fusor, giving you essentially, a star in a jar, spitting out hordes of neutrons every second, which can be used to irradiate things, after passing the neutrons through a moderator to slow them down, and this neutron irradiation can transmute elements into other elements via neutron capture. For example depleted uranium (238U) if irradiated with moderated, slowed down neutrons, would end up being transmuted to a mixture of plutonium 239 and neptunium, mostly as 237Np along with some of the isotope 239Np, that then decay after a short time depending on the isotope, back into isotopes of plutonium.
I wouldn't mind building a Farnsworth fusor sometime actually. And hey, what isn't to like about having a literal small star in a bottle of your very own, as long as you turn on the grid currents and supply the fuel for the plasma.
I'd fucking LOVE to see what the results would be like once we eventually get to not just room temperature but damn high temperature superconductors. Especially magnetic ones (magnetism is rare in superconductors, usually destroying the superconductivity, but they can exist. There is an alloy of neptunium which, although a low temperature (cryogenic, -hundreds of degrees) superconductor, it at least proves superconductivity in a magnetic material is possible.
Which if a high temperature one could be developed, eventually I reckon it probably will as we understand superconductors better and better, would allow for the addition of magnetic confinement by the grids themselves if made of such a magnetic superconductor, able to superconduct at high temperatures such as faced by the grids close to the plasma ball where nuclear fusion takes place, but the conductive losses....eventually the superconductor ought to saturate, and the conductive losses due to the grids be nullified, I think, at least.
Which would mean just having to run it for a long enough time for the superconductor to saturate it, or else artificially saturate it before turning it on, using an exterior implantation source to inject current into the grids, like a small LINAC running as an electron accelerator. Then hopefully that problem with the lossy grids would be solved.