On the magnetic nature of quantum point contacts

– We present results for a model that describes a quantum point contact. We show how electron-electron correlations, within the unrestricted Hartree-Fock approximation, generate a magnetic moment in the point contact. Having characterized the magnetic structure of the contact, we map the problem onto a simple one-channel model and calculate the temperature dependence of the conductance for different gate voltages. Our results are in good agreement with experimental results obtained in GaAs devices and support the idea of Kondo effect in these systems. Ballistic transport in quantum wires with ideal contacts leads naturally to conductance quantization in units of 2e/h [1–3]. This notable result has become one of the paradigms of mesoscopic physics. By a quantum wire we understand conductor that is so thin that the quantization of the transverse modes becomes important. In short quantum wires we may also expect some quantization effects along the wire axis [4]. In some of these short wires and in quantum point contacts, as the gate voltage is varied, together with the 2e/h plateaus the conductance shows some structure near 0.7(2e/h) [5–8]. This structure has been systematically observed GaAs devices and is known as the 0.7 anomaly. Recent experiments presented evidence on the magnetic nature of this structure: for gate voltages at the 0.7 anomaly the conductance increases as the temperature is lowered, an effect that may be interpreted as a signature of Kondo effect [7]. The Kondo picture is supported by the scaling of the experimental data, that close to the 0.7 anomaly shows a universal temperature dependence. Moreover the low temperature differential conductance presents a narrow structure at zero bias voltage, which splits with an in-plane magnetic field as due to a Kondo resonance. The conventional Kondo effect is the magnetic screening of a localized moment [9], if this effect is taking place in the quantum point contact (QPC), then the main question is: where and how is the magnetic moment generated? Recent spin dependent functional density calculations suggest the formation of a magnetic moment localized at the QPC for a device built on GaAs [10–12]. This scenario in which a magnetic moment is spontaneously generated by the electron-electron correlations is appealing since it naturally leads to the Kondo physics observed experimentally, however there are still many open questions related to the magnetic nature of the QPC and its relation to the 0.7 anomaly.