Parallel multi-focus femtosecond laser direct writing enabled by non-iterative phase regional segmentation and SVG guidance

Liquid crystal spatial light modulators (LC-SLMs) are key in femtosecond laser direct writing (LDW), enabling multifocal beam arrays for high-speed, parallel micro/nano-fabrication. Traditional iterative phase retrieval algorithms are hindered by slow convergence and high computational cost. To overcome this, we adopt a non-iterative annular phase segmentation method paired with scalable vector-based path generation using Scalable Vector Graphics (SVG). The pupil phase is divided into concentric annuli, each generating a distinct focal point. For a 16-spot configuration, intensity uniformity reaches a quality factor Q of 0.999, and phase maps are generated in under 100ms. These maps are dynamically loaded into the LC-SLM in sync with the motion stage and shutter system. Fabrication paths are derived from SVG files, parsed to reconstruct, complete, and discretize vector commands. A coordinate-adaptive filling algorithm ensures even energy distribution and prevents local overheating. For solid structures, a concentric inward-contour strategy preserves edge sharpness; for 3D hollow features, a depth-sliced gradient scan allows precise internal channels down to 127nm. Large-area fabrication is supported by dividing SVG domains into overlapping zones, with stitching errors reduced through beam-position compensation and submicron alignment verified by SEM. The system integrates beam shaping, SVG parsing, exposure control, and motion coordination into an automated workflow, eliminating manual coding. This approach enables scalable, high-resolution, and high-throughput LDW, suitable for applications in photonic integration, microfluidics, and bioinspired structures.