It is not that simple.
if you want "high-end" absorbtion you need to know the wavelength you are looking at because I don't think there is a single material that works well for the whole IR part of the spectra.
If you are using thin films you will also need to use multi-layer anti-reflection coatings in order to reduce the reflectivity. If this is done you can even use various metals as the actual absorber (IR bolometers can use all sorts of materials as the absorber, even gold).
Thank you, Cthugha! Yes, I want it for a detector. I can cool it down to 4.2K too. Single photon sensitivity is also required. I need the maximum absorption for the thinnest possible film, say a few nano meter. Spectral resolution is not an issue as long as either telecom freqs i.e. 1310 or 1550 is covered. The higher the bandwidth, even the better.So you want some kind of detector?
What are your requirements? Do you need spatial resolution? Temporal resolution? Will you be able to cool it to reduce noise? Do you need spectral resolution? Do you need single photon sensitivity?
If you want something like a photo diode, InGaAs, Ge or InSb are some of the obvious choices for the IR.
They are generally made from thin (a few nm) NbN films. However, in order to increase the absolute efficienty they also use multi-layer anti-reflection coating which is what I mentioned above.A different and relatively new way are Superconducting nanowire single-photon detectors. They are typically made of NbN or something similar. You might want to google these. I am not quite up to date what the most efficient designs of these are.
I should have made this more clear, but this is my opinion, too. Unless the guy who opened this thread works in a group specialized in building detectors, this is not a do-it-yourself thing. Right now, there are many groups studying potential single photon sources at telecom wavelengths. If this is about that topic, too, the easiest way to demonstrate single photon emission lies in getting two good avalanche photo diodes in the IR and a good coincidence counting circuit and work with these. For the cost of the working time a typical PhD student will need to design something even remotely similar, you can buy a lot of these devices for sure.If you want single photon sensitivity I am afraid this is one of those "if you have to ask" questions, there is simply no way you will be able to do this yourself.
Your best bet is to get a complete system from somewhere, either a company or a university group.