Fig. 1. Schematic of polarization detection approach via metasurface-based vector beam generator. Polarization detection can be achieved by constructing a vector beam with modulated intensity profiles (obtained from the beam passing through an analyzer) that vary sensitively and correspond one-to-one with the incident polarization. The inset shows the change in the intensity profiles at three incident polarization states. White ellipses and arrows indicate the incident polarization states.

Fig. 2. Principle of GPVVB-based polarization detection. (a) Intensity profiles of GPVVB and PVVB under orthogonal circularly polarized light incidence. GPVVB exhibits intensity fluctuations at the grafting joints depending on the incident circular polarization state, while PVVB shows nearly identical intensity profiles. The relation between the intensity profile and the angle is extracted according to the gray arrow labeling. (b) By extracting the intensity-angle profile of the modulated GPVVB light ring and using the valleys (1–4) and grafting joints (5 and 6) as the characteristic points, polarization detection can be achieved.

Fig. 3. Characterization of the designed geometric phase metasurface GPVVB generator. (a) The phase profile design of the GPVB. (b) The phase profile of the GPVVB generator. (c) The measured fast-axis angle profile of the fabricated metasurface. Scale bar: 500 μm. (d) The SEM images of femtosecond-laser-induced nanopores. The white arrows indicate the incident polarization directions of the femtosecond laser, which are consistently perpendicular to the nanopores’ major-axis directions. Scale bar: 200 nm. (e) Experimentally measured transmittance of the fabricated metasurface exceeds 95% in the wavelength range of 450–1100 nm. The inset shows a photograph of the fabricated device. Scale bar: 10 mm.

Fig. 4. High-accuracy polarization detection of different polarized lights at 532 nm. (a) Schematic of the experimental setup. LP, line polarizer; QWP, quarter-wave plate; BE, beam expander; MS, metasurface. (b) Simulated and experimental light intensity profiles corresponding to different polarization states, including two linear polarizations, two circular polarizations, and two elliptical polarizations. Scale bar: 1 mm. (c) Intensity-angle profiles extracted from the light intensity profiles, demonstrating a high degree of consistency between experimental and simulated results. (d) Calculated Stokes parameters derived from experimental results, with measurements from a commercial polarimeter used as a reference.

Fig. 5. Broadband polarization detection. (a) Experimental intensity profiles corresponding to different incident polarized lights at wavelengths of 473 nm, 633 nm, 808 nm, and 1064 nm. The intensity profiles at different wavelengths are well consistent, demonstrating highly robust broadband polarization detection. Scale bar: 1 mm. (b) Intensity-angle profile results at different wavelengths for two polarization states selected from (a). The results in the green dashed box correspond to y polarization while those in the purple box correspond to elliptical polarization.

Fig. 6. Radial vector beam polarization profile detection. (a), (b) Simulated and experimental intensity profiles of a 5×5 detection pixel array, respectively. The white dashed lines indicate the angular positions of intensity valleys 1–4. Scale bar: 100 μm. (c) Polarization profiles. The black arrows correspond to the theoretically calculated local polarization vectors, while the red arrows represent the measured ones. The black lines highlight the individual pixels of the metasurface detection array.

Fig. 7. (a) Light field profiles and (b) intensity-angle profiles corresponding to different topological charge grafting simulated under LCP incidence.

Fig. 8. Theoretical and experimental transmittance.

Fig. 9. Measured light field profile results for 25 different incident polarization states.