Solder Paste: Fundamental Material Property / SMT Performance
Aggressive form factors, reducing pitch, thinner packages, and larger die to package ratios are leading to higher package warpage during SMT reflow.
DOWNLOAD
Authored By:
Nilesh Badwe, Shunfeng Cheng, Srinivasa Aravamudhan, Mukul Renavikar
Intel Corporation
Hillsboro, OR, USA
Summary
Aggressive form factors, reducing pitch, thinner packages, and larger die to package ratios are leading to higher package warpage during SMT reflow. It is getting more challenging to mitigate warpage driven SMT defects viz. non-wet open (NWO), head-on-pillow (HoP) and solder bridging (SB). We studied multiple paste formulations using SMT hammer tests. Lab level characterizations were also used to establish a correlation between SMT performance and fundamental properties ofsolder pastes. We found that NWO and HoP compete with each other while having a correlation with flux activity to clean OSP on the Cu surface. SB risk showed a correlation with high temperature viscosity, indicating a rheology driven defect. Printability performance also showed a good correlation with the thixotropic index. These learnings will be extremely useful to develop next generation solder pastes to mitigate warpage driven defects.
Conclusions
Package warpage-driven solder joint defects can be mitigated through solder paste material design. Solder paste ability to clean surface OSP was found to affect NWO and HoP resistance of the paste. If the solder paste activity triggers early to clean the CuOSP surface, it minimizes the NWO risk through early intermetallic compound formation on the pad surface. However, this reduces activity of the paste for further solder oxide cleaning leading to a higher risk for HoP. This aspect of the solder paste needs to be optimized to reduce both HoP and NWO defects to an acceptable level. Liu et al. proposed two different methods viz. “tiny dot paste” and “ball onto paste” to assess HoP resistance of a paste. Scalzo proposed a graping test to study heat resistance of the paste. We believe that additional tests can be used to assess the HoP / NWO risk of solder pastes. Drop in tackiness as predicted by Amir et al. [1] did not show as strong a correlation with NWO performance of the pastes as the CuOSP solder spread. Bridging defect rate for multiple pastes was observed to correlate well with high temperature viscosity. Once the viscosity drops, the flux from the paste can easily flow causing wet bridging. This eventually provides a path for molten solder to migrate from one PCB pad to its adjacent pad during reflow, thereby increasing SB risk. A high thixotropic index was observed to predict good print performance of solder pastes.
Overall, there was no correlation observed between the halogen content classification of the solder pastes and their risk of forming either NWO or HoP defects. This indicates that rosins, acids and amines in the paste have higher impact on the activity than halogen content.
The learnings from this study will be used for next generation solder paste l development at different paste suppliers. These can also be used at board assembly manufacturers (OEMs / ODMs) to assess risk level for different pastes for different defect types and as a quality check for different lots of a material.
Initially Published in the SMTA Proceedings
|
