Technical Abstracts
In Case You Missed It
Circuit Fabrication
“Fabricating Functional Circuits on 3D Freeform Surfaces Via Intense Pulsed Light-Induced Zinc Mass Transfer”

Authors: Ning Yi, et al.

Abstract: 3-D freeform surface is of significant interest to wear-able devices on curvilinear skin/tissue surfaces or smart Internet of Things with sensors on 3-D objects. Here the authors present a new fabrication strategy that can directly print functional circuits either transient or long-lasting onto freeform surfaces by intense pulsed light-induced mass transfer of zinc nanoparticles (Zn NPs). The intense pulsed light can locally raise the tem-perature of Zn NPs to cause evaporation. Lamination of a kirigami-patterned soft semi-transparent polymer film with Zn NPs conforming to a 3-D surface results in condensation of Zn NPs to form conductive yet degradable Zn patterns onto a 3-D freeform surface for constructing transient electronics. Immersing the Zn patterns into a copper sulfate or silver nitrate solu-tion can further convert the transient device to a long-lasting device with copper or silver. Functional circuits with integrated sensors and a wireless communication component on 3-D glass beakers and seashells with complex surface geometries demonstrate the viability of this manufacturing strategy. (Materials Today, Aug. 5, 2021,

Quality Assurance
“Printed Circuit Board Defect Detection Based on MobileNet-Yolo-Fast”

Authors: Guohua Liu and Haitao Wen Guohua

Abstract: Automatic detection of defects is an essential part of the PCB production process. In recent years, while great progress has been made in PCB defect detection, various problems remain in tradition-al defect detection methods: for example, overreliance on the perfect template, difficulty in achieving precise image registration, and high vulnerability to environ-mental factors such as light, noise and reflectivity. The authors propose a fast defect detection network. On one hand, this algorithm solved the problems of tra-ditional methods. On the other hand, this algorithm solved the problems of large model size and poor real-time of existing deep learning methods. First, the k-means clustering algorithm is used to obtain more reasonable anchor boxes. Second, an improved MobileNetV2 is used as the backbone network. After the feature extraction network, the spatial pyramid pooling (SPP) structure is introduced to increase the receptive field of the image. Then, the authors use complete intersection over union to optimize the loss function. Finally, the authors build an enhanced fea-ture extraction network based on the feature pyramid network for multi-scale feature fusion. The experi-mental results show this method has small model size, good real-time, and good portability, suitable for prac-tical production. (Journal of Electronic Imaging, July 2021,

Solder Reliability
“Microstructure and Damage Evolution During Ther-mal Cycling of Sn-Ag-Cu Solders Containing Anti-mony”

Authors: Tae-Kyu Lee, Weidong Xie, et al.

Abstract: The interaction between the continuous microstructure evolution during thermal cycling and the long-term reliability of wafer-level chip-scale pack-ages (WLCSPs) with Sn-1.0Ag-0.5Cu (wt%) (SAC 105), Sn-3.0Ag-0.5Cu (wt%) (SAC 305), and Sn-3.9Ag-0.6Cu (wt%) (SAC 396) solder ball intercon-nects were investigated. Three different body-sized WLCSP with three different solder alloys on three different board thicknesses were thermally cycled from 0° to 100°C with 10 min. of dwell time, and the micro-structure evolution and their impact to the lifecycle numbers were identified. Based on both experimental and calculated data, higher Ag-containing solder alloys perform better in thermal cycling. However, the com-parison between the calculated lifecycle and the experi-mental results revealed mismatch, due to the localized recrystallization areal fraction differences. Smaller die WLCSP with 4 x 4mm2 and 3.2 x 3.2mm2 exhibited a large difference in expected lifecycle numbers. The cal-culated lifecycles expected a lower cycle number with thicker boards for SAC 105 and SAC 396 WLCSPs, but the experimental data revealed an increase with SAC 105, and a similar level of lifecycle time with SAC 396 for thicker boards. A widely distributed areal fraction of damage accumulation through the solder rows was observed in SAC 105 compared with higher Ag solder alloy joints, which show localized damage accumulation at corner joints. The difference of areal recrystallization distribution explains the difference between SAC 105 and SAC 305/396 thermal cycling behavior between the calculated and experimental thermal cycling results. (IEEE Transactions on Com-ponents, Packaging and Manufacturing Technology, vol. 10, no. 10, October 2020.

This column provides abstracts from recent industry conferences and company white papers. Our goal is to provide an added opportunity for readers to keep abreast of technology and business trends.