Azimuthal super-pupil beam engineering for improved fluorescence depletion microscopy

Abstract

Fluorescence depletion microscopy techniques such as STED and RESOLFT require optical fields with a well-defined and spatially confined central intensity minimum to achieve sub-diffraction lateral resolution. Here, we present the design and experimental implementation of an azimuthally polarized, doughnut-shaped depletion beam based on super-pupil engineering principles. By tailoring the radial amplitude distribution at the entrance pupil to approximate a Bessel-type target function, the resulting focal field exhibits a tighter central doughnut compared to conventional azimuthally polarized beams. The designed pupil field distribution is implemented using a phase-only spatial light modulator operated in a double pass configuration, enabling independent modulation of orthogonal polarization components via complex-field holographic encoding. Experimental characterization using sub-diffraction fluorescent beads demonstrates a reduction of the peak-to-peak distance of the central doughnut by approximately 16% relative to a nominal azimuthally polarized reference beam. Although the engineered field exhibits pronounced sidelobes, these do not preclude its use as a depletion beam, since lateral resolution is strongly influenced by the spatial confinement and effective suppression of the central intensity minimum for a given depletion intensity. This suggests that the proposed approach can enable improved lateral resolution at comparable depletion powers, providing a flexible and experimentally accessible route for engineering depletion fields in reconfigurable super-resolution microscopy systems.

Figures (10)

• Figure 1: Reference coordinate systems and the variables involved in the process (adapted from martinez2020uncertainty).
• Figure 2: (a) Profiles of the target beam $|\mathrm{J}_1(s n r)|^2$ with $s=6$ (magenta curve), the designed beam (cyan curve), and the azimuthally polarized one (dark blue curve). The vertical gray dotted lines and the labels inside the figure indicate the positions of the respective maxima. (b) 2D irradiance distributions of the top-hat azimuthally polarized beam and the designed beam in the focal plane. (c) Representation of the functions $\epsilon(\rho)$ and $\rho\,g(\rho)$. The vertical line indicates the value of a (= NA$/n$). (d) Cartesian representation of the corresponding $\rho\,g(\rho)$ hologram.
• Figure 3: Profiles of $I(r,0)$ for max(n) = 1, 7, 11, and 17, the azimuthally polarized top-hat beam (dark blue) and the target function (magenta). As in Fig. \ref{['fig:azvsown']}, the vertical, black, discontinuous lines indicates the position of the maxima of the azimuthally polarized Gaussian beam and the target function.
• Figure 4: Dependence of the beam profile with the number of terms and the windows size. (a) Window size $L = 4 \lambda$; (b) window size $L = 12 \lambda$. In both subfigures, $I(r,0)$ for $\mathrm{max}(n) = 7$ and $\mathrm{max}(n) = 17$ are depicted in cyan and blue, respectively. The horizontal, yellow line indicates the domain $L$ where the optimization takes place.
• Figure 5: Irradiances $I(r, z)$ for the azimuthally polarized top-hat beam (top left) and the designed beams using $\mathrm{max\{n\}=}$ 1, 3, 7, 11 and 17.