Multimaterial DLP printing with centrifugal-assisted rinsing and inverse design of voxel structures

Voxel structures provide precise control over material distribution at the microscale. The integrated material–structure design paradigm offers a broader design space than traditional approaches based solely on geometry or homogeneous materials. A fundamental question in voxel structures is how to accurately print each voxel with designed sizes while avoiding interfacial contamination. To address it, we develop a multimaterial DLP 3D printer that integrates centrifugal forces and liquid rinsing, significantly reducing interfacial contamination, especially for high-viscosity resins. Additionally, we establish a multiphysics model considering chemical reactions, diffusion effects, and Gaussian light fields to predict voxel sizes accurately. This model is applicable to multiple resin types and is used to optimize printing parameters for individual voxels. By combining the fabrication method and theoretical model, voxel structures can be fabricated with high precision, facilitating the design of diverse functional behaviors. Hence, a machine learning-evolution algorithm method is developed to inversely design voxel distribution across a vast design space, and voxel structures with tailored buckling behaviors are demonstrated. The proposed fabrication, modeling, and design framework pave the way for developing voxel structures.