Finite Element Simulation of Localized Flip-Chip Thermo-Compression Bonding for GaN-Based Micro-LEDs

Journal article, 2026

Micro light-emitting diode (Micro-LED) has attracted significant research interest owing to its outstanding optoelectronic properties. Flip-chip thermo-compression bonding is a critical step in its manufacturing process, wherein a systematic investigation of process parameters is essential for ensuring bonding quality. Conventional process development heavily relies on experimental trial-and-error approaches, which are time-consuming and costly, thus necessitating the integration of simulation technologies for efficient optimization. Existing simulation studies predominantly focus on single-pixel device bonding; however, this approach faces practical challenges such as high-precision alignment requirements, mechanical damage, particle contamination, and positional inaccuracies introduced during mass transfer. In contrast, directly bonding an unpatterned, full Micro-LED epitaxial wafer to a driver substrate, followed by pixel definition through photolithography and etching, circumvents the mass transfer procedure, thereby enhancing process yield and robustness. Nevertheless, simulation research on this direct bonding strategy remains scarce. To address this gap, this paper develops a localized representative finite element model to simulate the thermo-compression bonding process between a Micro-LED epitaxial chip and a driver substrate incorporating a representative 10 × 10 Au bump array. The study focuses on the thermomechanical coupling behavior within the bump array, with systematic analysis of the influences exerted by structural dimensions, bump geometry, and temperature on stress and deformation distributions. To alleviate the issue of stress concentration at the array periphery, a mitigation approach using stress buffer layers filled between bumps is proposed. Results show that adopting polyimide as a buffer layer markedly improves stress distribution uniformity, thereby reducing potential failure risks.