Obviously fluorescent tubes have mercury. And this isn't really even a problem that needs to be solved since the amount of mercury is so small (and fluorescent tubes are rapidly falling out of favor). But if we were to design a mercury-free fluorescent tube, would xenon be our best bet? I heard somewhere that xenon is used in modern-made real neon signs to excite the phosphors, but apparently it doesn't work quite as well as good old mercury metal. Obviously we would have to design entirely new phosphors and all of that jazz, but assuming we got that all figured out would this even be a viable solution? How much worse would it be efficiency-wise?

Some scanners and digital copiers used mercury free CCFL and electrodeless fluorescent lamps that used xenon.
Osram also made the Planon, which was a flat mercury free fluorescent lamp based on dielectric barriers.

Efficiency-wise I would imagine quite bad. Xenon has very strong IR spectral lines, this limits efficiency in even the highest wattage xenon short-arc lamps to about 40 lm/W or lower. In a low pressure discharge, the efficiency without a phosphor would probably be even worse. Xenon does have great color rendering and a color temperature very similar to sunlight. But for general purpose lighting a high CRI CMH lamp would work a lot better, as CRI can be very similar with much greater efficiency than a xenon lamp. Although better electrode design might permit higher efficiency.

The idea of making mercury free fluorescent tubes was very lucrative and actually was researched quite intensively. Low pressure xenon in fact emits quite strong UV, but at very short vacuum UV wavelengths, roughly 140-170nm. The problem with this is very high losses due to stokes shift when you convert say 150nm photon to 600mn photon you lose 3/4 of the energy. So xenon fluorescents ended up being ~2 times less efficient than mercury variants. Making hypothetical two-photon phosphor absorbing one photon at short wavelength and emitting two at longer was not fruitful.  And as we all know the proliferation of LEDs killed that all.

I did some xenon tubes in cold cathode sign tube format, generally they were somewhat dim in comparison with mercury filled variants, but with nice clean colors due to very low visible radiation of Xe. On a good side, Xe tubes are practically insensitive to environmental temperature and have impressive lifetime.

There are mass produced mercury free HPS lamps, and actual mercury free MH in the form of automotive D3/D4 lamps.

Ironically, damn EU bureaucrats actually *killed* mercury-free HPS on the grounds of poor efficiency, but allowed regular Na+Hg variants to survive!

@dor123
I did not know that, that is definitely interesting to know.

@NeXe Lights
Only 40 lm/W? That is pretty low.

@RRK
Well that explains it. Were these ever produced for the market or are there any engineering samples? I would love to see one of these. I did not know about mercury-free HPS, I would think removing mercury wouldn't make that much of an efficiency difference in HPS due to sodium being the main emitter, but apparently I am wrong.