I'm afraid that the resolution of IR microscopy is not enough for this purpose. It will give composition information averaged over a larger area than the nanophase. You will have to use electron microscopy or other, high resolution techniques, sometimes coupled with diffraction techniques.
It's not easy to identify the phase of nanomaterials from FTIR spectra, but you can use micro-Raman spectroscopy (see my papers 'phases transition of nanoparticles')
The best way is to use XRD and TEM analysis. But in some special system, e.g. the syntheses and growth of CaCO3, one can use FTIR to to determine the phase of CaCO3 (calcite or amorphous CaCO3)
I mean it is available just in some special system. FTIR usually help to check the different phases of CaCO3. An example please see what is described in Journal of Crystal Growth 254 (2003) 206–218.
Obviously, the XRD technique is the best method to identifiy crystalline phases in many mateirals. But, the FTIR can help sometimes. At least, from TiO2, the anatase and rutile species can be detected by this technique.