First-hand information for everyone
Researchers used the sharp tip of a scanning tunneling microscope to press very slightly on a single layer of iron selenide (FeSe) grown on strontium titanate (SrTiO3). That tiny, controlled strain pushed an electron band called dz2 closer to the energy level where electrons can form Cooper pairs. As a result, the superconducting energy gap increased in two clear steps, from 17.8 milli‑electronvolts to 23.6 meV for the main gap — an increase of about 32 percent.
The team changed the tip-sample distance by varying the tunneling current in the microscope and read out the local electronic spectrum with scanning tunneling spectroscopy (STS). At large tip distance (low current) they saw the usual U-shaped superconducting gap. As the tip moved closer (current increased from about 100 picoamperes up to several nanoamperes), the gap became larger. The gap evolution showed two stages. Below about 1.25 nanoamperes the gap rose slowly. Above that current there was a sharper, stronger increase in gap size. The inner and outer gap features rose by roughly 39% and 32% overall, respectively.
To understand why the gap changed, the authors tracked features in the STS spectra that they associate with specific iron d-orbital bands. Orbitals are simply different shapes that an electron’s probability cloud can take around an atom; the labels dz2, dxy and dxz/yz name those shapes. The dz2 feature normally lies far below the energy that matters for pairing. Under tip-induced tensile strain the dz2 feature moved up in energy toward the Fermi level. At larger strain it mixed, or hybridized, with a relatively flat dxy band near −80 meV. That band reconstruction coincided with the second, stronger jump in the superconducting gap. The authors also report that Fermi wave vectors stayed the same, which argues the effect comes from reshaping and renormalizing bands and correlations, not from adding or removing charge carriers.
This work matters because it challenges a common assumption that the dz2 orbital does not take part in pairing in iron-based superconductors. It shows a practical way to tune superconductivity without chemical doping: apply local strain to change electronic correlations and the ordering of bands. The result suggests a new, orbital-selective route to enhance superconductivity in multiband, correlated materials. That could point to strategies for designing or optimizing superconductors in other layered systems.
Important caveats remain. The experiment uses a local, dynamic mechanical perturbation from an STM tip, so the effects are measured at the nanometer scale and may not immediately translate to bulk samples or devices. The interpretation relies on assigning spectral features to orbitals and on comparisons to prior angle-resolved photoemission spectroscopy (ARPES) data; such assignments carry some uncertainty. The authors tested multiple samples and tips and reproduced the two-stage behavior, but further work will be needed to confirm how broadly this mechanism applies and to fold it into theoretical models of pairing in correlated multiband superconductors.