Why Scanning Tunneling Microscopy on Sr${}_2$RuO${}_4$ sometimes doesn't see the superconducting gap

Scanning tunneling microscopy (STM) is perhaps the most promising way to detect the superconducting gap size and structure in the canonical unconventional superconductor Sr${}_2$RuO${}_4$ directly. However, in many cases, researchers have reported being unable to detect the gap at all in simple STM conductance measurements. Recently, an investigation of this issue on various local topographic structures on a Sr-terminated surface found that superconducting spectra appeared only in the region of small nanoscale canyons, corresponding to the removal of one RuO surface layer. Here, we analyze the electronic structure of various possible surface structures using first principles methods, and argue that bulk conditions favorable for superconductivity can be achieved when removal of the RuO layer suppresses the RuO${}_4$ octahedral rotation locally. We further propose alternative terminations to the most frequently reported Sr termination where superconductivity surfaces should be observed.