Electrical tuning of the $g$ factor of single self-assembled quantum dots

The $g$ factor of a single self-assembled quantum dot is tuned by applying an electrical bias voltage. Individual InGaAs quantum dots embedded in a stripe mesa structure sandwiched by Schottky electrodes are studied by photoluminescence measurements. We find that under applied magnetic field a dot with an asymmetric shift for applied bias voltage shows an increase of the exciton $g$-factor absolute value by about 8%, while most of the dots with relatively symmetric shifts show no significant change. The anomalous $g$-value increase is discussed in terms of the Kondo effect.

Figure 1

(Color online) (a) Schematic view and (b) the scanning tunneling microsocope (SEM) picture of our device. A mesa stripe structure in which InGaAs quantum dots are embedded is sandwiched by $\mathrm{Ti}∕\mathrm{Au}$ metal electrodes. (c) Energy and position dependences of the lower conduction band edge under applied a bias voltage when a dot is embedded near an electrode. (d) Spectral shift of the emission energy of various single quantum dots as a function of bias voltage. (e) Number of quantum dots as a function of emission energy difference between the bias voltages of $+0.5\mathrm{V}$ and $-0.5\mathrm{V}$, which represents the asymmetry of the Stark shifts for the bias voltage.

Figure 2

(Color online) (a) Photoluminescence for various applied bias from a dot located near the center of the mesa stripe under applied magnetic field of $5\mathrm{T}$ in Faraday configuration. (b) Energy shift of the Zeeman-split peaks. Solid lines give the calculated emission energy (shifted vertically by about $10\mathrm{meV}$ for comparison). (c) Experimental and calculated energy splitting or corresponding exciton $g$ value as a function of applied voltage.

Figure 3

(Color online) (a) Emission energy shift and (b) the splitting magnitude of the Zeeman-split peaks at $5\mathrm{T}$ in a dot located near the electrode. (c) Photoluminescence spectra of the Zeeman-split peaks, plotted as a function of the energy separation measured from the lower peak. The spectra are shifted to line up the lower peak to show the splitting increase clearly. The PL spectra at $10\mathrm{K}$ were measured at excitation density of $4\mathrm{W}∕{\mathrm{cm}}^2$ because no essential power dependence was observed at the temperature.

Figure 4

(Color online) Experimental and theoretical $g$-value variations as a function of the corresponding Stark shift. The experimental $g$-value variations are shown for quantum dots with the symmetric (QD A) and the asymmetric (QD B) energy shifts for the external bias voltage. The experimental Stark shift is measured from the emission energy at the bias voltage which corresponds to the zero effective electric field (Ref. 15).