Semiconductor of spinons: from Ising band insulator to orthogonal band insulator

Abstract: Within the ionic Hubbard model, electron correlations transmute the single-particle gap of a band insulator into a Mott gap in the strong correlation limit. However understanding the nature of possible phases in between these two extreme insulating phases remains an outstanding challenge. We find two strongly correlated insulating phases in between the above extremes: (i) The insulating phase just before the Mott phase can be viewed as gapping a non-Fermi liquid state of spinons through staggered ionic potential. The quasi-particles of underlying spinons are orthogonal to physical electrons and hence they do not couple to photoemission probes, giving rise to "ARPES-dark" state due to which the ARPES gap will be larger than optical and thermal gap. (ii) The correlated insulating phase just after the normal band insulator corresponds to the ordered phase of slave Ising spins (Ising insulator) where charge configuration is controlled by an underlying Ising variable which indirectly couples to external magnetic field and hence gives rise to additional temperature and field dependence in semiconducting properties. In the absence of tunability for the Hubbard $U$, such a temperature and field dependence can be conveniently employed to achieve further control on the transport properties of Ising-based semiconductors. The rare earth monochalcogenide semiconductors where the magneto-resistance is anomalously large can be a candidate system for the ordered phase of Ising variable where pairs of charge bosons are condensed in the background. Combining present results with our previous dynamical mean field theory study, we argue that the present picture holds if the ionic potential is strong enough to survive the downward renormalization of the ionic potential caused by Hubbard $U$.