| speaker: | Silvano De Franceschi |
| time: | Mo. 07.04.03, 11:20 - 12:10 |
A common approach to the fabrication of quantum dot structures takes advantage of semiconductor heterostructures with a high-mobility two-dimensional electron gas (2DEG) located ~100 nm below the surface of the chip. Metal gate electrodes deposited on top of the chip via e-beam lithography are used to deplete the 2DEG underneath and isolate a small puddle of electrons. The same gate electrodes can also be used to finely control the tunnel coupling between the quantum dot and the nearby electron reservoirs, which in this case consist of extended regions of the same 2DEG. In strongly coupled quantum dots (i.e. tunnel resistances comparable to the quantum resistance), a local spin-1/2, resulting from an odd number of confined electrons, can be fully screened by the anti-ferromagnetic exchange interaction with the delocalized electrons in the nearby Fermi seas. This is a manifestation of the Kondo effect, where the quantum dot itself can be regarded as a single artificial magnetic impurity. The versatility of semiconductor quantum dot structures and the high control over their physical parameters offer rich opportunities for the study of this fundamental many-body phenomenon. In the first part of the talk, I will present an experiment on the Kondo effect in a quantum dot laterally coupled to a quasi-ballistic quantum wire. A finite bias across the wire causes a splitting in the Kondo resonance which we ascribe to the formation of a double-step electron distribution function in the wire. This experiment provides the first direct evidence of voltage-induced splitting of the Kondo resonance. Kondo correlations are strongly suppressed when the bias voltage exceeds the Kondo temperature. A moderate perpendicular field enables us to selectively control the coupling between the dot and the two Fermi reservoirs at the edges of the wire. In fact, already at fields of order 100 mT only the Kondo resonance associated with the strongly coupled reservoir survives. In the second part of the talk, I shall present an alternative approach to the realization of quantum dot systems which makes use of chemically synthesized semiconductor nanowires. These recently developed nanostructures present a number of attractive properties, such as high aspect ratios (i.e., 10-100 nm diameters and lengths even larger than 100 mm), and a wide range of materials (group IV, III-V, and II-VI semiconductors), with the possibility to form nanowire heterostructures by sequential growth of different semiconductors. Nanowires can be deposited on oxidized Si substrates and contacted individually by e-beam defined metal electrodes. In Delft we have used n-doped InP nanowires grown at Philips to make single-wire field-effect devices. Low-temperature transport measurements have revealed single-electron tunneling and energy quantization resulting from the confinement in the wire. These results represent an important first step towards the development of controllable quantum-dot devices based on semiconductor nanowires. These new type of devices may provide an interesting test ground for the physics of highly correlated electron systems.