Due to its outstanding physical properties, graphene is expected to become a new generation material, able to replace or complement traditional semiconductors in device technology. Hence, many studies have been led to explore the potential of this material immediately after the successful fabrication of a single layer of graphene in 2004. However, applications of gra-phene in electronic devices are still questionable due to the gapless character of this material. In particular, regarding electronic applications, the absence of energy bandgap in the band structure makes it difficult to switch off the current in graphene devices like transistors. Re-garding thermoelectric properties, the gapless character is also a strong drawback since it pre-vents the separation of the opposite contributions of electrons and holes to the Seebeck coeffi-cient. Thus, a sizable band gap in graphene is a requirement to overcome the disadvantages of graphene and to fully benefit from its excellent conduction properties. It has been shown that many nanostructuring techniques can be used to open such a bandgap in graphene, e.g., gra-phene nanoribbons, graphene bilayer with a perpendicular electric field, graphene nanomesh lattices, channels based on vertical stack of graphene layers, mixed graphene/hexagonal boron nitride structures, nitrogen doped graphene, and so on. However, each of these methods has its own fabrication issues and/or need to be further confirmed by experiments. In this work, we focus on strain engineering, which offers a wide range of opportunities for modulating the electronic properties of graphene nanostructures. For this theoretical work, all calculations were performed using essentially two main methods, i.e., an atomistic tight binding Hamilto-nian model to describe the electronic structure and the non-equilibrium Green's function ap-proach of quantum transport. The main aim is to analyze in details the strain effects in gra-phene and to provide strategies of strain engineering to improve the performance of both elec-tronic (transistors and diodes) and thermoelectric devices.
After introducing the general context if this work and the numerical techniques developed for this purpose, we first analyze the only effect of strain. Actually, if uniformly applied, a strain of large amplitude (> 23%) is required to open a bandgap in the band structure of graphene. However, we show that with a strain of only a few percent, the strain-induced shift of the Di-rac point in k-space may be enough to open a sizable conduction gap (500 meV or more) in graphene heterojunctions made of unstrained/strained junctions, though the strained material remains gapless. After analyzing in details this property according the amplitude and direction of strain and the direction of transport, we exploit this effect using appropriate strain junctions to improve the behavior and performance of several types of devices. In particular, we show that with a strain of only 5%, it is possible to switch-off transistors efficiently, so that the ON/OFF current ratio can reach 105, which is a strong improvement with respect to pristine graphene transistors where this ratio cannot exceed 10. Then we show that by combining strain and doping engineering in such strain junctions the Seebeck coefficient can reach values higher than 1.4 mV/K, which is 17 times higher than in gapless pristine graphene. It can con-tribute to make graphene an excellent thermoelectric material. Finally, we study the effect of negative differential conductance (NDC) in graphene diodes made of either as single gate-induced strained barrier or a p-n junction. We show that appropriate strain engineering in the-se devices can lead to very strong NDC effects with peak-to-valley ratios of a few hundred at room temperature.