Available CCD image sensors is restricted to 400-1100 nm range. Is it possible to get a this for 2 um range. What difficulties it encountered to develop such sensor ?
Mercury Cadmium Telluride is probably the go-to material at this wavelength. InGaAs imaging arrays are also available, though MCT are "state of the art," I suppose. The ordinary ones, with the spectral sensitivity in the 400-1100 nm are based on Silicon.
The difficulties with both MCT and InGaAs are much higher noise than Silicon, and sensitivity to background thermal radiation. They generally require cooling (down into cryogenic range) and shielding. They are also difficult materials to work with, so processing is much harder than for Silicon.
Another established manufacturer in this area is Teledyne.