We consider a single-layer graphene nanoribbon with armchair edges and with a longitudinally constant external potential, pointing out that it can be described by means of an effective non-Hermitian Hamiltonian. We show that this system has some features typical of dissipative systems, namely, the presence of exceptional points and of PT-symmetry breaking, although it is not dissipative.
We predict a huge interference effect contributing to the conductance through large ultraclean quantum dots of chaotic shape. When a double-dot structure is made such that the dots are the mirror image of each other, constructive interference can make a tunnel barrier located on the symmetry axis effectively transparent. We show (via theoretical analysis and numerical simulation) that this effect can be orders of magnitude larger than the well-known universal conductance fluctuations and weak localization (both less than a conductance quantum). A small magnetic field destroys the effect, massively reducing the double-dot conductance; thus a magnetic field detector is obtained, with a similar sensitivity to a SQUID, but requiring no superconductors.
In this paper we show how numerical and analytical modeling of graphene-based devices is used to consider possible approaches to engineer a gap in graphene and to evaluate the perspectives of different technological options towards graphene nanoelectronics.
The study of the transport and noise properties of graphene‐based devices requires the computation of the potential profile as a function of the applied bias voltages. However, an exact solution for the potential profile involves a complete self‐consistent treatment of the electrostatic and transport equations, which is computationally very expensive. Here, generalising the approach proposed by Das , et al. , the authors describe an approximate method that allows the evaluation of the potential profile by properly modifying, as a function of the bias voltages applied to the gates, the profile for a reference bias point, which is supposed to be known. The proposed approach is not very demanding from the computational point of view and can be useful for simulations aimed at the interpretation of experimental results or at device design.