Assessing Queue Time in D2w Hybrid Bonding Through Precise Bond Strength Measurements

The process development of direct placement (or sequential bonding) in die-level hybrid bonding has not yet reached full maturity. A key distinction between die-level and wafer-level hybrid bonding is the queue time (Q-time) between activation/deionized water (DIW) rinse and bonding. In wafer-towafer (W2W) bonding, the bonding sequence has been done immediately and simultaneously after DIW cleaning within a cluster bonding tool. In contrast, die-to-wafer (D2W) bonding introduces a time lag due to tool-to-tool transfer and individual die handling/bonding, resulting in an extended Q-time. In this study, the Q-time in D2W hybrid bonding has been comprehensively investigated using a novel method for accurately measuring bond strength. During the extended Q-time, water evaporation at the bonding interface occurs, leading to insufficient water retention which is critical for direct bonding. This results in an increased microscopic interface gap, ultimately weakening the bond strength in direct bonding. The extent of water evaporation and the hydrophilicity of the interface were analyzed using water contact angle (WCA), thermal desorption spectroscopy (TDS) and zeta potential measurements, all of which corroborated this phenomenon. These findings provide critical insights into the challenges and optimization strategies for the D2W bonding process, paving the way for an improved D2W stacking methodology.