Ultralow-Power Pseudospintronic Devices via Exciton Condensation in Coupled Two-Dimensional Material Systems

Abstract

“Pseudospintronic” device concepts, novel “beyond-CMOS” device proposals targeted toward revolutionizing the current semiconductor technology based on MOSFETs and CMOS logic, are addressed in detail. These pseudospin devices include the voltage-controlled Bilayer pseudoSpin Field-Effect Transistor (BiSFET) and the current-controlled Bilayer pseudoSpin Junction Transistor (BiSJT). MOSFETs are confronted by the intractable physics of thermionic emission and resulting source-to-drain leakage that limits voltage scaling. As a result, CMOS faces an “energy crisis” much as the one faced by bipolar junction transistor-based logic that led to CMOS. As for many other beyond-CMOS concepts, these pseudospintronic devices are based on a completely different physics of switching, potentially allowing much lower voltage and power operation. These pseudospintronic device concepts employ possible room-temperature interlayer electron–hole exciton condensates between two dielectrically separated layers of two-dimensional (2D) materials for subthermal voltage (sub-k _B T/q) switching, specifically from a nearly shorted interlayer conductance state to a highly resistive interlayer conductance state with increasing interlayer voltage. These collective exciton states with their “which-layer” degree of freedom are somewhat analogous to collective spin states in magnets, which is the origin of the “pseudospintronics” moniker. Device performance in the presence of such condensates is the primary focus of this work; the possibility of room-temperature condensates, itself, is addressed by other research still in progress. We begin with a discussion of the underlying physics. Graphene-based pseudospintronic systems then are analyzed using quantum transport simulations incorporating the nonlocal exchange interaction. However, the essential transport physics should be much the same for other 2D material systems, including transition metal dichalcogenides for which the realization of the condensate may be easier. The BiSFET and BiSJT device concepts are presented in detail, and basic logic gate designs are illustrated for each. Compact device models are developed and SPICE-level circuit simulations are performed to demonstrate possible switching energies on the scale of or below a tenth of an attojoule, well below even end-of-the-road-map CMOS. However, like many other beyond-CMOS concepts, these devices remain concepts without solid experimental embodiments. The fabrication concerns of such novel devices are also discussed along with recent experimental progress.