fundamentally? yes, of course. there's no physical laws or reason why you couldn't - just observe the usual caveats, such as being unable to achieve 100% efficiency, that sort of thing.

now, the question of how to design an actual working solution? i don't know; and basically, it depends. you would then also have the problem of whether it was possible to implememnt that solution, using real matter and physical objects...

unfortunately, that was probably the question you really wanted answered, right?