Answers are grounded in this paper and cite their sources.
With an in situ observation during flip-chip bonding under vacuum conditions, we demonstrate That accurate passive alignment can only be achieved when the oxide on the solder is removed. For this purpose, we used molecular hydrogen (H2) under vacuum conditions, in order to prevent the negative side effects when using fluxes or fluxlike chemicals. With this new, fluxless technology, we achieved at moderate bonding temperatures [e.g., 250/spl deg/C for SnPb(60/40)] and heating durations (<2 min) self-aligned FC-bonds well within 2 μm residual offset. Using evaporated SnPb(60/40) and AuSn(80/20) solder bumps, FC-bonding experiments in vacuum (10/sup -4/ Pa) showed spontaneous alignment within 5 s but the accuracy of alignment for both solders was nevertheless always insufficient to meet the required tolerances for single-mode fiber-to-chip coupling.
No takes yet. Share an insight, caveat, or question.
Kuhmann et al. (1996) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: