Nuclear magnetic resonance probes for the Kondo scenario for the 0.7 feature in semiconductor quantum point contact devices

We discuss the expected features in nuclear relaxation and Knight shift measurements for the Kondo scenario for the '0.7 feature' of semiconductor quantum point contact ( QPC) devices defined in two-dimensional electron gases ( 2DEGs). As the conductance is more sensitive to the nuclear polarization in the centre of the QPC than that in the 2DEG leads, our analysis is focused on the region near to the centre of the QPC. We show that the exchange coupling of a bound electron in the QPC with the nuclei would lead, in the region near to the centre of the QPC, to a much higher rate of nuclear relaxation compared to that involving exchange of nuclear spin with conduction electrons. Away from the centre of the QPC, we find that the distance beyond which the latter ( conduction electron) mechanism becomes equally important is of the order of typical QPC lengths; thus, between these two electronic mechanisms, relaxation by coupling to the bound electron dominates within the QPC. Furthermore, we show that the temperature dependence of the nuclear relaxation due to coupling to the bound electron is non-monotonic in contrast to the linear-in-T relaxation from coupling with conduction electrons. Nuclear spin diffusion processes restrict the range of validity of this analysis. We present a qualitative analysis of additional relaxation due to nuclear spin diffusion (NSD), and compare the nuclear relaxation times associated with NSD and the above electronic mechanisms. We discuss circumstances in which NSD will affect our results significantly, and suggest ways in which NSD may be suppressed in the QPC so that the Kondo physics may be unearthed. Nuclear relaxation together with Knight shift measurements will help in verifying whether the '0.7' feature is indeed due to the presence of a bound electron in the QPC. While some of the results have also been discussed in the context of paramagnetic impurities in bulk conductors, our analysis is intended for application to the 0.7 effect in semiconductor systems. The qualitative and quantitative estimates that we make will allow experimental tests of the Kondo scenario for the 0.7 feature of QPCs in two-dimensional electron gas heterostructures.