The key thing here is that hydrogen does not contact the reactor at all.

The reactor is a container containing a mixture of gasses: uranium 235 hexafluoride, xenon difloride & helium. It is surrounded by neutron reflective material in the form of louvers that can be opened to permit neutrons to escape or closed to keep them in.

The operation of the system is as follows. When the louvers are closed, a fission chain reaction begins. Each fission produces a thick track of ionized helium from the two parts of the split nucleus. The ionized helium bounces against the xenon difluroide, boosting it to an excited state. The xenon difluoride begins stimulated emission of photons, that is it begins to lase.

A control beam of light from an electrically driven laser enters the reactor and is amplified. The optics is arranged so that the amplified light sweeps the entire reactor by means of mirrors.

The output beam of the reactor stikes a mirror and is focused upon a window at the base of the acceleration tube. Hydrogen is injected into this tube and heated to about 30,000 kelvins, from whence it expands through a nozzle, forming the propulsion jet.

No substance can exist in solid form at 30,000 kelvins. This does not constitute a major problem because the walls of the tube are water cooled, and therefore a boundary layer of relatively cool hydrogen forms. This is similar to the situation in a plasma torch which reaches comparable temperatures.

Figuring the energy in the laser beam at 10 Megawatts, such a machine could produce a high thrust (compared to alternatives such as ion engines) with a reasonable specific impulse on the order of 10 to 20 kilometers per second.

A few years ago, I submitted this idea to NASA for an evaluation. Their evaluation was that it would indeed work, but it would not be as good as other fission propulsion concepts already being investigated, in particular, using a vapor phase fission reactor to heat hydrogen by means of thermal radiation emission through a set quartz tubes.