Byengkyu Kang, Yong Hun Jung, Ju-Hee Kim, Dong-Hee Choi, Jaehoon Kim, Kyungwon Park, Minseop Kim, Seok-Hyeon Kang, Kyuhwan Na, Sejoong Kim, Hyung Chul Kim, Hyunho Kim, Minsuh Kim, Seok Chung
Replicating in vivo tissues with complex, branched, and tortuous geometries remains a challenge in engineering physiologically relevant tissue models. Hydrogel compartmentalization in microfluidic chips can spatially organize cells and microenvironments, but micropillar-based confinement relies on discrete structures that generate segmented interfaces and can limit the design of continuous, nonlinear compartment boundaries. Here, we develop a weir-based microfluidic platform that forms continuous hydrogel-medium boundaries and enables stable hydrogel patterning across diverse, tissue-relevant architectures. We establish a predictive, pressure-based design framework that relates geometric parameters to the minimum pressure required to advance the gel front and the maximum pressure tolerated before interface failure. Using theoretical analysis and computational simulations, we identify intersection geometries that are susceptible to failure and provide practical layout guidelines to improve filling stability in complex networks. We validate these predictions experimentally and demonstrate reliable compartmentalization across extended branched networks, interwoven gel-medium channel architectures, and multihydrogel designs. Finally, we demonstrate biological applicability by engineering complex interconnected 3D vascular networks and continuous renal epithelial tubes that conform to complex and curved microchannels and by quantifying how local connectivity and spatial architecture drive region-specific tissue morphogenesis. Together, these results improve the predictability and expand the design versatility of weir-based microfluidic platforms for hydrogel compartmentalization in complex geometries.