In order to achive an atomic resolution, microscopy method should provide an interaction matching with the atomic dimension. (e.g. electron microscopy like TEM in which the electron wavelength i.e. ~1 nm is in the order of atomic dimension i.e. several angstroms). For light microscopy, the wavelength does not meet such requirement and thus, the main problem is Diffraction Limit as you may know already about. According to Prof. Farid's comment, super-resolved fluorescence microscopy can obviate this limitation superiorly.