Figure 1

Magnetoresistance and layout of focusing devices. Voltage across the detector (contacts 3 and 4) is measured as a function of magnetic field perpendicular to the surface of the sample ($B_{\perp }$). Lithographical separation between point contacts is $0.8\mu \mathrm{m}$. Current of 1 nA is flowing through the injector (contacts 1 and 2). Positions of the magnetic focusing peaks are marked with arrows. Inset: AFM micrograph of a sample ($5\mu \mathrm{m}\times 5\mu \mathrm{m}$). Light lines are the oxide which separates different regions of 2D hole gas. Conductance of quantum point contacts is controlled via voltages applied to the detector $G_{\mathrm{d}\mathrm{e}\mathrm{t}}$, injector $G_{\mathrm{i}\mathrm{n}\mathrm{j}}$ and the central $G_C$ gates. Semicircles show schematically the trajectories for two spin orientations.Reuse & Permissions

Figure 2

Magnetoresistance is measured with magnetic filed oriented (a) perpendicular or (c) at $\sim 3°$ to the surface. Series resistance due to 2DHG is subtracted. In-plane field $B_{||}=3.3\mathrm{T}$ corresponds to the center of the peak and is aligned with the direction of the carriers injection. Black (solid) curves are measured with injector and detector QPCs gated to pass both spin orientations (conductance $G=2e^2/h$). Red (dashed) curves are measured with injector QPC gated at $G\approx 0.5e^2/h$. In (b) conductance of the injector QPC is plotted as a function of the gate voltage at $B_{||}=3.3\mathrm{T}$. (d) Height of the low- (magenta dots) and high-field (blue triangles) peaks in (b) is plotted as a function of the injector gate voltage.Reuse & Permissions