This is what we presume is beauty of TEM. We have nice pictures showing hexagonal structures of graphene and carbon nanotubes where we position carbon atoms.
Of course, at the Angstrom length scale, there are no "particles" - rather the electronic wave functions of atoms. And yes, this is what High-Resolution TEM is aiming to successfully resolve - since mid 1980s I guess.
Here a small movie in HAADF STEM mode of a Cs corrected tool showing aglomeration of Pt clusters dynamically under the e-beam. Sometimes even single atoms are vislible in the surounding matrix, but they are highly dynamic and move faster than the scan speed.
Metal atoms are in sub Angstrom in size. If it is possible visualise through TEM, one can see their position in solid Inorganic framework viz. zeolites. Beside one can see the position of non metallic elments also in solid or liquid molecular structure.