Interesting application. But I don't see wavelengths or power, of a UV source meant to drive a reaction, as a possibility. A UV lamp will send a bunch of wavelengths, but why bother to tune them, the ones you don't need shouldn't hurt anything else, and any sort of filter will rob you of some energy from the wavelengths you do want.
For analytical purposes, we use diffraction gratings or a prism, but you can't fit something like that in the space you describe.
Lokewise power, I suppose you could put the UV lamp on a rheostat, and dim or intensify it, but why? Avoid excess heat? I can't think of anything else.
Why do you need a quartz cuvette? Any old glass can hold your sample, if the sleeve that surrounds your UV source is quartz that is immersed in your sample. Or hey, just leave it open and aim the light inside. Now you're protected from short wavelengths by the regular glass and your sample gets the light, and heat can escape up.
So, building on what Borek: and Corribus: said. Did you check into the usefulness of any of those options for your application? Why specifically are they not good for your application?