The physical origins of high-temperature superconductivity in copper-oxide based materials are believed to be intimately related to the physics of doped Mott insulators. A Mott insulator is a material where “double occupancy” of the atomic valence orbitals is prohibited due to strong “on-site” electron-electron repulsions. Hence, the conduction electrons stay “at home.” They cannot visit neighboring atoms as this would imply temporary double occupancy. Such a system is electrically insulating and often exhibits antiferromagnetic order.
So how can such an insulator be made superconducting? The trick is to introduce chemical dopants, i.e., electron acceptors, such that some of the relevant orbitals are emptied. Now, electrons are free to move, as double occupancy during transport can be avoided. A doped Mott insulator becomes metallic as more dopants are added, which in the case of doped cuprate compounds led to the spectacular Nobel Prize-winning discovery of high-temperature superconductivity in 1986.
The cuprate compounds are not easy to use for device applications. They are chemically complex and very brittle. They are also electronically inhomogeneous, which presumably has to do with the random placement of dopant impurities in these materials. This begs the question if it would be possible to recreate doped Mott physics on a much simpler materials platform and hopefully establish superconductivity.
In the December 29, 2017, issue of Physical Review Letters, Postdoctoral Researcher Fangfei Ming and coworkers took advantage of the strong electron-electron repulsion in the dangling-bond orbitals of a monatomic tin layer on a silicon surface, and created a well-ordered two-dimensional Mott insulator. Using heavily-doped silicon substrates, they were able to empty some of these dangling bonds, as shown in the figure below. Using scanning tunneling microscopy and spectroscopy, Ming found that the electronic structure of the resulting surface exhibits the tell-tale characteristics of a hole-doped Mott insulator. This interpretation was firmly corroborated by theoretical calculations of Assistant Professor Steven Johnston. Moreover, contrasting with known doped Mott insulators, the structure and electronic structure of the two-dimensional tin layer are perfectly homogeneous. Superconductivity has yet to be established, though a recent theory paper indicated that the hole doping levels needs to be increased well beyond the current levels. The UT group is now pursuing other doping schemes in an attempt to double the hole concentration.
Sn atoms (blue) on the Si(111) surface exhibits dangling bond orbitals (lobes). If all of them contain one electron, as marked by the dots, the layer will be insulating. Electrons are forbidden to hop from one atom to the next as this would imply double occupancy. However, the introduction of boron dopants in the substrate underneath (red atoms) vacates some of those dangling bonds so that electrons can move freely.
The realization of homogeneous doped Mott insulators on a silicon platform suggests a new avenue for exploiting correlated electron physics in search for novel materials, including unconventional superconductors. The fact that these phenomena can be realized on the technologically important silicon platform offers a significant advantage.
The experimental work was done in Professor Hanno Weitering’s laboratory at the Joint Institute for Advanced Materials. The theory component was completed in collaboration with Dr. Thomas Maier at Oak Ridge National Laboratory. Other co-authors include UT student Tyler Smith, former postdoc Daniel Mulugeta, Research Scientist Paolo Vilmercati, Joint Faculty Assistant Professor Paul Snijders, and Dr. Geunseop Lee from Inha University, South Korea.