Fig. 1. Schematic diagram of the designed spin-decoupling and multi-modal modulation device. (a) Structural sketch of the dual-channel device. The inset in the middle shows a schematic representation of the meta-atom with a period (P) of 360 nm, height (H) of 550 nm, and a range of length (L) and width (W) from 60 nm to 300 nm in simulation. The inset on the right depicts the global coordinate system (x,y) for the entire metasurfaces and the local coordinate system (u,v) for each nanopillar. (b) The phases delay of the meta-atoms under excitation in the u and v directions, denoted as φu (red circle, solid line) and φv (red circle, dashed line), along with their phase difference |φu−φv| (black square, solid line). The transmission of the meta-atoms under u and v directional excitation, denoted as tu (blue circle, solid line) and tv (blue circle, dashed line). (c) Transmittance of the co-polarized component tll (solid blue line) and cross-polarized component trl (solid red line) of the meta-atoms under LCP incidence, along with the corresponding polarization conversion efficiency (PCE, solid black line).

Fig. 2. Simulation results of the far-field pattern of Device 1 for collimation and deterministic modal modulation functions. (a) Far-field plots of the electric field intensity distributions of the total output beams. The white dotted lines represent the area of θ=40°, corresponding to an NA of 0.65. (b) Far-field phase distributions of the total output beams. (c) Mode purities of single photons carrying distinct topological charges. Each column in (a)–(c) shows the cases of l=0–3.

Fig. 3. Simulation results of Device 2 with different polarization states of QD emission. (a) Far-field plots of the electric field for QD emissions with a 0° linear polarization state (H). (b), (c) The corresponding extracted far-field electric field intensity distributions for LCP and RCP. (d), (e) The corresponding far-field spiral phase distributions for LCP and RCP. The orange and green dashed circles in (a) indicate the LCP and RCP regions shown in (b)–(e). (f) Collection efficiencies, (g) divergence angles, and (h) mode purities of the corresponding extracted spin-components for different polarization states (H, D, V for 0°, 45°, and 90° linear polarization, and L and R for LCP and RCP) of QD emission.

Fig. 4. Simulation results of the far-field pattern of Device 3 as an eight-channel photon emitter with determinate states. (a) Far-field plots of the electric field intensity distributions. (b), (c) The corresponding extracted far-field electric field intensity distributions for LCP and RCP. (d)–(g) The corresponding far-field phase distributions, divergence angles, and mode purities for LCP and RCP. The orange and green dashed circles in (a) indicate the LCP and RCP regions shown in (d) and (e).

Fig. 5. Extraction efficiency of a slab photonic structure. (a) Schematic diagram of the flat structure. (b) Extraction efficiency of upward radiation as the thickness of silica layer tSiO2 independently changes when tGaAs=160  nm and dQD-GaAs=30  nm. (c) Extraction efficiency of upward radiation as the thickness of GaAs layer tGaAs independently changes when tSiO2=40  nm and dQD-GaAs=30  nm. (d) Extraction efficiency of upward radiation as the distance from QD to the lower surface of the GaAs film dQD-GaAs independently changes when tSiO2=40  nm and tGaAs=160  nm.

Fig. 6. Relationship between the spin states of the output beams and polarization states of QD emission. (a) Far-field plots of the electric field intensity distributions with different QD emission polarization states (D, V for 45° and 90° linear polarization, while L and R for LCP and RCP). (b) The corresponding extracted far-field electric field intensity distributions for LCP and RCP. (c) The corresponding far-field spiral phase distributions for LCP and RCP. The orange and green dashed circles in (a) indicate the LCP and RCP regions shown in (b) and (c).

Fig. 7. Fabrication tolerance of Device 1 (l=1). Far-field plots of the electric field intensity distributions (within a zenith angle of 10°) for a single variable. (a), (b) Varying the thickness of the middle silica layer tSiO2. (c), (d) Varying the thickness of GaAs layer tGaAs. (e), (f) Shifting the QD position by a distance dx.

Fig. 8. Performance of QD emissions from Device 1 with different silica thicknesses between the QD and metasurfaces. (a) Distribution of the electric field amplitude at different thicknesses of the top silica layer without metasurfaces. (b), (c) Far-field intensity distributions at focal lengths of 2 μm and 6 μm with metalens, respectively. Here, the focal lengths correspond to the thickness of the top silica layer.

Fig. 9. Performance of QD emissions from Device 1 at different working wavelengths. (a)–(f) Far-field electric field intensity distributions (within a zenith angle of 10°) for a single OAM beam with a topological charge 1, at working wavelengths of 880 nm, 890 nm, 900 nm, 920 nm, 930 nm, and 940 nm, respectively. (g) Mode purities, (h) divergence angles, and (i) collection efficiencies of the output OAM beams for the corresponding working wavelengths.