材料科学
制作
激光器
光电子学
电子包装
脉冲激光器
光学
复合材料
医学
物理
病理
替代医学
作者
Daniel Franz,David Schuster,Simon Schwarz,Stefan Rung,Cemal Esen,Ralf Hellmann
标识
DOI:10.1016/j.optlaseng.2025.109106
摘要
We report on a comprehensive study of through-glass via (TGV) drilling for glass-based core packaging using a multi-path scanning strategy with a rotating start point and an ultrashort pulsed laser emitting in the green spectral range. In order to optimize the laser drilling quality in a 200 μm thick Borofloat (BF) 33 glass substrate, processing parameters such as laser pulse duration, laser pulse energy, number of passes and TGV hole diameter are studied in detail. After laser drilling, through-holes are evaluated for the hole diameter, taper angle and crack formation using optical microscopy and scanning electron microscopy. Although small taper angles down to 6 ± 0.3° are obtained using a femtosecond pulse duration, microcracks and backside ablation of the glass substrate are observed. Using a laser pulse duration of 5 ps, a laser pulse energy of 25 μJ and 600 passes for TGV fabrication with a hole diameter of 125 μm, a minimum taper angle of down to 7.3 ± 0.1° is obtained without the formation of cracks. Minimum hole pitches of 180 μm are realized for a hole diameter of 125 μm. To meet industrial requirements for temperature resistance in electronics manufacturing, laser-drilled TGV arrays are tested for thermal sensitivity using a furnace for both thermal cycling and shock at temperatures up to 950°. After thermal tests, no crack formation is observed for a minimum hole pitch of 180 μm and maximal laser-induced stresses is reduced by 37.4% down to 8.2 MPa on average, demonstrating the great potential of the applied scanning strategy for glass-based core packaging. • USP laser drilling of TGVs with a multi-path scanning strategy in Borofloat 33 glass. • Influence of laser parameters on the taper angle, TGV diameter and induced stress. • High packaging density and resistance of TGVs to temperature cycling and shocks.
科研通智能强力驱动
Strongly Powered by AbleSci AI