An obstacle to the use of graphene as an alternative to silicon electronics has been the absence of an energy gap between its conduction and valence bands, which makes it difficult to achieve low power dissipation in the OFF state. We report a bipolar field-effect transistor that exploits the low density of states in graphene and its one-atomic-layer thickness. Our prototype devices are graphene heterostructures with atomically thin boron nitride or molybdenum disulfide acting as a vertical transport barrier. They exhibit room-temperature switching ratios of ≈50 and ≈10,000, respectively. Such devices have potential for high-frequency operation and large-scale integration.

The exceptional electronic properties of graphene and its formidable potential in various applications have ensured a rapid growth of interest in this new material [1,2]. One of the most discussed and tantalizing directions in research on graphene is its use as the base material for electronic circuitry that is envisaged to consist of nanometer-sized elements. Most attention has so far been focused on graphene nanoribbons (see [3][4][5][6][7][8][9] and references therein). In this Letter, we report quantum dot (QD) devices made entirely from graphene, including their central islands (CI), quantum barriers, source and drain contacts and side-gate electrodes. We have found three basic operational regimes for such devices, depending on their size. For relatively large (submicron) CIs, size quantization plays an insignificant role, and our devices were found to operate as orthodox singleelectron transistors (SET) exhibiting periodic Coulomb blockade (CB) oscillations. The CB regime has been extensively studied previously using metallic and semiconducting materials [10,11] and, more recently, the first SET devices made from graphite [12] and graphene [1,13,14] were also demonstrated. The all-graphene SETs reported here are technologically simple, reliable and robust and can operate above liquid-helium temperatures T, which makes them attractive candidates for use in various charge-detector schemes [10]. For intermediate CI sizes (less than ∼100nm), we enter into the quantum regime, in which the confinement energy δE >10meV exceeds the charging energy E c . Such a strong quantization for relatively modest confinement is unique to massless fermions [1,2] and related to the fact that their typical level spacing δE ≈v F h/2D in a quantum box of size D is much larger than the corresponding energy scale ≈h 2 /8mD 2 for massive carriers in other materials (v F ≈10 6 m/s is the Fermi velocity in graphene, h the Planck constant and m the effective mass). This means that level splitting in graphene-based 100-nm devices should be tens and hundreds times larger than in typical semiconducting and metal QDs, respectively. This regime is probably most interesting from the fundamental physics point of view, allowing studies of relativistic-like quantum effects in confined geometries [15][16][17][18][19][20][21]. In particular, we have observed a strong level repulsion in QDs, which is a clear signature of quantum chaos (so-called "neutrino billiards" [15]). Conductance of our smallest devices is dominated by individual constrictions with sizes down to ∼1nm, which exhibit δE ∼0.5eV and a good-quality transistor action at room T. It is remarkable that these molecular-scale structures survive microfabrication procedures, remain mechanically and chemically stable and highly conductive under ambient conditions and sustain large (nA) currents. Our devices were made from graphene crystallites prepared by micromechanical cleavage on top of an oxidized Si wafer (300nm of SiO 2 ) [22]. By using high-resolution electron-beam lithography, we de...

The celebrated electronic properties of graphene have opened the way for materials just one atom thick to be used in the post-silicon electronic era. An important milestone was the creation of heterostructures based on graphene and other two-dimensional crystals, which can be assembled into three-dimensional stacks with atomic layer precision. Such layered structures have already demonstrated a range of fascinating physical phenomena, and have also been used in demonstrating a prototype field-effect tunnelling transistor, which is regarded to be a candidate for post-CMOS (complementary metal-oxide semiconductor) technology. The range of possible materials that could be incorporated into such stacks is very large. Indeed, there are many other materials with layers linked by weak van der Waals forces that can be exfoliated and combined together to create novel highly tailored heterostructures. Here, we describe a new generation of field-effect vertical tunnelling transistors where two-dimensional tungsten disulphide serves as an atomically thin barrier between two layers of either mechanically exfoliated or chemical vapour deposition-grown graphene. The combination of tunnelling (under the barrier) and thermionic (over the barrier) transport allows for unprecedented current modulation exceeding 1 × 10(6) at room temperature and very high ON current. These devices can also operate on transparent and flexible substrates.

Lateral superlattices have attracted major interest as this may allow one to modify spectra of two dimensional (2D) electron systems and, ultimately, create materials with tailored electronic properties 1-8 . Previously, it proved difficult to realize superlattices with sufficiently short periodicity and weak disorder, and most of the observed features could be explained in terms of commensurate cyclotron orbits 1-4 . Evidence for the formation of superlattice minibands (so called Hofstadter's butterfly 9 ) has been limited to the observation of new low-field oscillations 5 and an internal structure within Landau levels 6-8 . Here we report transport properties of graphene placed on a boron nitride substrate and accurately aligned along its crystallographic directions. The substrate's moiré potential 10-12 leads to profound changes in graphene's electronic spectrum. Second-generation Dirac points 13-22 appear as pronounced peaks in resistivity accompanied by reversal of the Hall effect. The latter indicates that the sign of the effective mass changes within graphene's conduction and valence bands. Quantizing magnetic fields lead to Zak-type cloning 23 of the third generation of Dirac points that are observed as numerous neutrality points in fields where a unit fraction of the flux quantum pierces the superlattice unit cell. Graphene superlattices open a venue to study the rich physics expected for incommensurable quantum systems 7-9,22-24 and illustrate the possibility to controllably modify electronic spectra of 2D atomic crystals by using their crystallographic alignment within van der Waals heterostuctures 25 .Since the first observation of Weiss oscillations 1,2 , 2D electronic systems subjected to a periodic potential have been studied in great detail [3][4][5][6][7][8] . The advent of graphene has rapidly sparked interest in its superlattices, too [13][14][15][16][17][18][19][20][21][22] . The principal novelty in this case is the Dirac-like spectrum and the fact that charge carriers are not buried deep under the surface, allowing a relatively strong superlattice potential on a true nanometer scale. One promising avenue for making nanoscale graphene superlattices is the use of a potential induced by another crystal. For example, graphene placed on top of graphite or hexagonal boron nitride (hBN) exhibits a moiré pattern [10][11][12]26 , and graphene's tunneling density of states becomes strongly modified 12,26 indicating the formation of superlattice minibands. The spectral reconstruction occurs near the edges of superlattice's Brillouin zone (SBZ) that are characterized by wavevector G =4/ D and energy E S =v F G/2 (D is the superlattice period and v F graphene's Fermi velocity) 12,22 .To observe moiré minibands in transport properties, graphene has to be doped so that the Fermi energy reaches the reconstructed part of the spectrum. This imposes severe constraints on the misalignment angle  of graphene relatively to hBN. Indeed, D is given by  and the 1.8% difference between the two lattice constants ...