Raman spectroscopy reflects bonding information of molecules. Your finding could possibly mean different molecular structures formed in different approaches, which contained the same composition of elements and the same repeated units of crystalite grains.

If the Raman modes are collective (i.e. non-localized to a functional group), they are sensitive with respect to the crystallite size. I don't know how MOF modes look like, but you can check that out.

In principle, XRD should also be somewhat sensitive to that; do the reflexes there have the same width for all samples?

Raman spectroscopy is extremely sensitive and is said to detect the finger-print modes of vibration for a material. Even for close crystal structures, the nature of molecular bonds and associated vibrational modes can be different, which can only be understood from a detailed group theoretical analysis. There are some softwares and DFT-based techniques, that can calculate the theoretical Raman-active and IR-active phonon modes for a given structure. You can compare this data to match your experimental results and also find if any mode is optically silent.

Muhammad Usman

MOF is obtained by multi-step synthesis and it is difficult to imagine that elemental analysis and XRD are the same for different samples. EDX gives an error of 5-10% for a standard sample. MOFs are synthesized to produce small pores of 2-10 nm, but the samples themselves can be of different sizes. In addition, it is not clear that XRD shows a crystalline sample or amorphous. Peaks can overlap. If there are metal nanoparticles, then in the Raman spectrum there may be peaks SERS, and if there are aromatic fragments in the ligands, then it can cover the peaks of luminescence.

Read attached articles