Fig. 1. Scheme of spatially orthogonal excitation. (a) Schematic of the spatially orthogonal excitation setup, showing the alignment of a single-mode optical fiber and the sample inside a cryostat. The optical fiber, actuated by a displacement stage with a custom mold and metal tape, is placed a few micrometers from the sample’s cleaved edge. Signal light, collected by an objective with a numerical aperture (NA) of 0.65, is directed perpendicular to the incident laser direction and coupled into a single-mode fiber. (b) A scanning electron microscope (SEM) image of the sample, consisting of seven pairs of SiO2/TiO2 top DBRs and 46 pairs of GaAs/Al0.95Ga0.05As bottom DBRs. Quantum dots (QDs) are located within a parabolic microcavity at the center. (c) A frontal EMCCD image of the sample showing the cleaved edge, microcavity with QD, and reference marker. (d) A frontal EMCCD image capturing the spatially orthogonal excitation of QDs. The incident laser, with specific polarization, couples into the sample, enabling long-distance, low-loss propagation via waveguide modes and exciting QDs along the path. Note that markers and other cavities in the sample could cause some scattered light. However, if the scattered light is sufficiently distant from the microcavity under investigation, its impact on the measurements is minimal and can be considered negligible.

Fig. 2. Spatially orthogonal resonant single-photon characterization of a QD in microcavity. (a) Photoluminescence (PL) spectra of Cavity A with a small splitting Δω=11.05  GHz. The H-polarized mode (highlighted in red) is at 932.125 nm with a linewidth of 0.098 nm (δωH=33.84  GHz), while the V-polarized mode (highlighted in blue) is at 932.157 nm with a linewidth of 0.102 nm (δωV=35.22  GHz). The inset shows that the H-polarized contribution of the QD is 57.80% while the V-polarized mode is 42.2%. (b) Cavity B exhibits a larger split of Δω=86.50  GHz. The H-polarized mode is at 913.945 nm with a linewidth of 0.111 nm (δωH=39.84  GHz), and the V-polarized mode is at 913.704 nm with a linewidth of 0.108 nm (δωH=38.78  GHz). The inset shows the highly polarized single photon, with 91% of QD’s emission in the H-mode. (c) Lifetime of resonant fluorescence of QD in microcavity under spatially orthogonal excitation. Under the “π pulse” condition, we extracted an ultra-short lifetime from the typical lifetime histogram of CX, which is ∼134  ps (blue curve) for QD A in Cavity A and ∼53  ps (red curve) for QD B in Cavity B. With the lifetime of ∼1.007  ns (black curve) in the bulk material, we obtain the Purcell factor of ∼7.5 in Cavity A and ∼19 in Cavity B. The instrument response function (IRF) is illustrated by the green line, with a τ of 35 ps. (d) The relationship between the pump strength and the count rate of the avalanche photon detector (APD), revealing a full Rabi oscillation curve. The curves represent the numerical fitting results, with the maximum count rates of QD A and QD B being ∼4.93  Mcps and ∼6.23  Mcps, respectively. The “π pulse” of QD in Cavity A and Cavity B corresponds to a power of 66 μW and 36 μW, respectively.

Fig. 3. High-performance single photon in microcavity with different excitation strategies. (a) Schematic of spatially orthogonal excitation, achieving filter-free excitation with a saturated count rate of resonant fluorescence at 6.23 Mcps. (b) Schematic of polarized-orthogonal excitation, where orthogonal polarization filtering reduces the saturation count rate to 4.64 Mcps. (c) Measurement of single-photon second-order coherence, with filter-free excitation yielding a second-order coherence of g(2)(0)=0.0472(2), and polarized-orthogonal excitation resulting in 0.0889(7). (d) Measurement of HOM interference. A clear contrast at zero-time delay of two orthogonal polarizations is observed, with values of g∥(2)(0)=0.0822(2) and g⊥(2)(0)=0.5312(12), resulting in an indistinguishability of Vraw=0.845(2) for spatially orthogonal excitation. For polarized-orthogonal excitation, we attained values of g∥(2)(0)=0.1518(3) and g⊥(2)(0)=0.5641(20), indicating Vraw=0.731(4). (e) Temperature dependence of the g(2)(0) (blue) and indistinguishability (red). As the temperature increases, the detuning between the QD and the cavity gradually increases, while g(2)(0) remains stable around 0.952, and indistinguishability decreases gradually. (f) Power dependence of g(2)(0) (blue) and indistinguishability (red). g(2)(0) increases with decreasing power, reaching a minimum of 0.0322(2), influenced by the scattered light while indistinguishability remains stable.

Fig. 4. Mollow triplet of QD in microcavity under spatially orthogonal excitation. (a) High-resolution PL spectra of the QD in a microcavity under spatially orthogonal excitation, showing a maximum Rabi splitting of 21.4 GHz at 4.1 mW laser power. (b) Linear correlation between the Rabi splitting and the square root of laser power. (c) Linear relationship between the linewidths of the side peaks and the square of the Rabi frequency.

Fig. 5. Cross-section of Fabry–Perot microcavity. Cross-sectional view of the Fabry–Perot microcavity, featuring 46-pair GaAs/Al0.95Ga0.05As bottom DBRs and seven-pair dielectric top DBRs.

Fig. 6. Experiment setup. (a) Schematic of the spatially orthogonal and polarized-orthogonal excitation setups. (b) Schematic of the Hanbury Brown and Twiss (HBT) setup for measuring second-order coherence and the Hong-Ou-Mandel (HOM) setup for assessing indistinguishability.

Fig. 7. Waveguide-cavity scheme. (a) Schematic representation of the waveguide-cavity scheme, engineered for spatially orthogonal resonant excitation, showcasing the integration of a QD within a hybrid Fabry–Perot (FP) microcavity-waveguide structure to enhance excitation and photon collection efficiency. (b) Calculated transmission mode profile within the waveguide.