Sessions Index

Bio:
Akihiro Shimizu received his B.E. and M.E. degrees in electrical and electronics engineering in 2005 and 2007 from Okayama University respectively, and his M.S. degree in electrical engineering from University of Colorado at Boulder in 2013. He is currently enrolled in the doctoral program at Gifu University as a working professional and is expected to receive his Ph.D. in September 2025. He joined Ushio Inc. in 2007 and is currently working as a chief engineer, focuing on the surface treatment techniques of polymer materials using xenon excimer lamps and atmospheric pressure plasma, with applications in the packaging and automotive industries.

Abstract:
INTRODUCTION
The wet desmear [1]—comprising swelling, permanganate, and reduction treatments—remains the industry standard for effective smear removal in printed circuit board (PCB) manufacturing. Nevertheless, challenges persist regarding efficiency and environmental impact, as permanganate treatment produces manganese dioxide sludge that requires frequent solution replacement, increasing costs, waste, and safety risks due to its strong oxidizing nature.
To address these issues, the author previously reported at IMPACT 2024 that vacuum ultraviolet (VUV) irradiation using an excimer lamp as pretreatment reduced permanganate time from 15 to 10 min at a VUV irradiation dose of 0.3 J/cm2 or greater [2]. This improvement was attributed to VUV-induced photochemical decomposition of smear, enhancing swelling solution permeability and smear removal efficiency during permanganate treatment. However, excessive VUV irradiation may partially degrade the epoxy resin surface, leading to increased surface roughness during subsequent permanganate treatment.
This study presents a novel approach that balances permanganate time reduction with the preservation of epoxy resin surface quality by optimizing the VUV irradiation dose—a critical trade-off optimization of VUV pretreatment conditions not explored in the IMPACT 2024 study.

EXPERIMENT
A 30 µm-thick glass epoxy resin film (ABF GX-T31) was vacuum-laminated onto both sides of a copper-clad laminate and thermally cured. Blind via holes (80 µm diameter) were formed by CO2 laser drilling. Samples were subjected to VUV irradiation using an excimer transfer device, with the VUV irradiation dose varied from 0.3 to 3 J/cm2. Following VUV irradiation, the wet desmear was performed under the conditions of swelling (60 °C, 5 min), permanganate (80 °C, 10 min), and reduction (40 °C, 5 min) treatments. Surface roughness (Sa) over a 10 × 10 µm area was measured using an atomic force microscope after VUV irradiation followed by permanganate treatment (including reduction treatment) (Fig. 1).

RESULTS AND DISCUSSION
Fig. 2 shows the relationship between the VUV irradiation dose and the Sa, following 10 min of permanganate treatment after VUV irradiation. The Sa increased markedly with increasing VUV irradiation dose, rising from 42.9 nm at 0.3 J/cm2 to 52.9 nm at 0.5 J/cm2, 72.4 nm at 1 J/cm2, and reaching 110 nm at 3 J/cm2. This significant increase in surface roughness at higher VUV doses is attributed to VUV-induced epoxy degradation.
The solid line represents an Sa of 50.7 nm after 15 min of permanganate treatment alone. Remarkably, at VUV irradiation doses of 0.5 J/cm2 or higher, the Sa exceeded this reference, suggesting that the surface may have been roughened more than necessary.
The appropriate VUV irradiation dose should be determined based on the desired Sa, which varies depending on subsequent process requirements. Assuming an acceptable Sa range of 80–120% relative to that from 15 min permanganate treatment alone (41–61 nm, shown by dashed lines), the suitable VUV irradiation dose lies between 0.3 and 0.7 J/cm2.

CONCLUSION
This study demonstrated that excessive VUV irradiation leads to significant surface roughening. Therefore, precise optimization of the VUV irradiation dose—for example, between 0.3 and 0.7 J/cm2—is essential for achieving an effective and balanced pretreatment in the wet desmear PCB manufacturing.

 

Bio:
Dr. Frank Xu, Global Product Manager at MacDermid Alpha Electronics Solutions, brings 18 years of experience to the field of PCB final finishes product development and management. Throughout his tenure, he has been responsible for overseeing research and development of new products, providing crucial on-site support for beta testing and applications, and facilitating direct technical and strategic roadmap discussions with both direct and OEM clients. His expertise is further highlighted by his presentations at numerous technical conferences and his holding of multiple patents.

Abstract:
Printed Circuit Boards (PCBs) are essential components in modern electronics, serving as the backbone for electrical connections. The final finish of a PCB is crucial for reliable solder joint formation, manufacturing yields, and performance of the PCBA. As the industry has moved towards lead-free technology, fabricators have largely abandoned Hot-Air Solder Leveling (HASL). Instead, several alternative finishes - such as Immersion Silver (ImAg), Immersion Tin (ImSn), Organic Solderability Preservative (OSP), Electroless Nickel Immersion Gold (ENIG), and Electroless Nickel Electroless Palladium Immersion Gold (ENEPIG) - have become well established, each meeting fabricators' specific cost and performance requirements.

Both electroless nickel and traditional immersion gold processes operate at high temperatures (~80 °C) and require long plating times (>10 minutes), making them among the least sustainable of the alternative surface finishes. ENIG and ENEPIG are also the most expensive and complex finishes to implement. Despite this, they remain the most widely used surface finishes by sales revenue, thanks to their excellent shelf life, reliable solderability, wire bond capability, and the rising value of precious metals.

This paper explores strategies to improve the sustainability of ENIG and ENEPIG processes. Each step of the process is examined to identify opportunities for optimization. For electroless nickel, sustainability enhancements include the elimination of dummy plating and a reduction in minimum bath loading requirements. For electroless palladium and immersion gold, the focus is on optimizing precious metal concentrations and controlling thickness distribution. These improvements aim to enhance environmental sustainability while also lowering the cost of ownership. The investigation encompasses both flexible and rigid PCB applications of ENIG and ENEPIG.

Key words: Sustainability, ENIG, ENEPIG, electroless nickel, electroless palladium, immersion gold, cyanide free gold.

 

Bio:
Professor Tzu-Chien Wei is a Chair Professor in the Department of Chemical Engineeringat National Tsing Hua University (Hsinchu, Taiwan). He received his Ph.D. degree in2007, followed by four years of industrial experience (2007–2011), and completed hispostdoctoral training in 2012. Since then, he has been with National Tsing Hua University,focusing on renewable energy and advanced electroless plating technologies. Professor Wei’s research interests include perovskite solar cells, dye-sensitized solarcells, and advanced electroless plating processes. He has published more than 100 peer-reviewed articles in these fields and has received numerous academic honors, such as theProfessor Kao Kwang Memorial Award, the Ministry of Science and Technology Outstanding Research Award, the Environmental Protection Administration Green Chemistry Education Award, the National Tsing Hua University Outstanding Teaching Award, the Future Technology Award, the Wu Ta-You Memorial Award, and the NTHU New Faculty Research Award. In addition to his academic achievements, Professor Wei is also an entrepreneur, having founded three start-up companies: Taiwan Perovskite Technology, Chinghui Perovskite Consulting, and Nihe Advanced Materials.

Abstract:
Palladium (Pd)-based catalysts are indispensable in the electroless plating process of printed circuit board (PCB) fabrication, particularly for initiating copper deposition on non-conductive substrates. However, the deactivation of these catalysts generates Pd-containing wastewater, which is typically outsourced to third-party recyclers. This open-loop approach offers limited material traceability, suboptimal recovery efficiency, and incurs significant cost burdens. To address these challenges, this study presents a simple, in-house method for recovering Pd from electroless plating wastewater and converting it into reusable, active catalysts. Using sodium borohydride reduction followed by nitric acid dissolution and polyvinyl alcohol (PVA) stabilization, we synthesized polyvinyl alcohol-stabilized recycled Pd nanoparticles (R-PVA-Pd). The optimal PVA to recycled Pd (PVA: R-pd) mass ratio of 11.4:1 yielded stable colloidal nanoparticles suitable for catalytic applications. Characterization revealed that R-PVA-Pd derived from two different waste sources maintained consistent particle size distribution and catalytic performance.
Catalytic activity tests demonstrated that R-PVA-Pd exhibited slightly lower activity than laboratory-synthesized pure PVA-Pd but significantly outperformed commercial Pd-based catalysts. Scanning transmission electron microscopy coupled with energy-dispersive spectroscopy (STEM-EDS) confirmed the presence of Pd–Cu bimetallic nanoparticles. Despite a slightly larger average particle size, the R-PVA-Pd catalysts remained highly effective. Their long-term stability was enhanced through ultrasonic dispersion, low-temperature storage, and dilution prior to use. Importantly, the recovered catalyst successfully enabled the plated through-hole (PTH) activation process, validating its industrial applicability. Economic analysis showed that the production cost of R-PVA-Pd was only 1.16% of the market price of equivalent Pd, indicating substantial cost savings and a viable path to circular resource utilization within the PCB manufacturing workflow. This approach not only reduces raw material consumption and waste handling costs but also enhances supply chain transparency and process control—key enablers of sustainable and circular electronics production.

 

Bio:
I am from the Department of Materials Science and Technology of National Yang Ming Chao Tung University and am currently in the first year of my doctoral program. My main research areas are advanced packaging materials and hybrid bonding technology. In this report, I will explain how to use the CSR system platform to optimize the process parameters to produce high-strength copper foils and demonstrate its potential as a collector for lithium batteries in the future.

Abstract:
This study presents the development of high-strength nanotwinned copper (NT-Cu) foils through the optimization of electroplating parameters using the Complex System Response (CSR) platform. By systematically tuning four key parameters—copper ion concentration (0.5 M), chloride ion concentration (49.13 ppm), electrolyte temperature (10°C), and current density (7.11 ASD)—an NT-Cu foil with an ultimate tensile strength (UTS) of 826 MPa, electrical conductivity of ~70% IACS, and ductility of 2.46% was fabricated. The CSR approach significantly reduced the number of experimental iterations required to achieve optimal results. Advanced microstructural characterization using FIB revealed a refined grain structure and high twin boundary density, both of which contributed to the enhanced mechanical performance. Subsequent low-temperature annealing further improved ductility without compromising tensile strength, achieving a peak UTS of 828 MPa. These results demonstrate the efficacy of the CSR methodology in optimizing multi-parameter electroplating systems and underscore the suitability of NT-Cu foils for next-generation lithium-ion battery current collectors.

 

Bio:
I am currently a student at the Academy of Circular Economy, National Chung Hsing University (NCHU), under the supervision of Professor Jenn-Ming Song. I conduct my research in the Advanced Interconnect Materials Lab. My research focuses on nanotechnology, Cu–Cu direct bonding, surface treatment, copper oxide control, and electrochemical analysis. In particular, I am dedicated to developing low-temperature, high-reliability Cu–Cu bonding processes for advanced packaging applications through surface modification and interfacial oxide control strategies.

Abstract:
To address the critical challenge of copper oxidation in Cu–Cu direct bonding for 2.5D/3D semiconductor packaging, this study proposes a novel temporary surface protection approach using molecular nanolayers (MNLs), combined with an effective removal strategy prior to bonding. Three MNLs featuring different chain lengths and functional groups, were chemically grafted onto Cu surfaces to suppress oxidation. Systematic surface characterization revealed that all three MNLs were successfully grafted onto the Cu surface, as indicated by significant increases in water contact angle (from 15° to 98°), confirming uniform molecular coverage. Subsequent electrochemical and XPS analyses demonstrated their effectiveness in delaying surface oxidation. However, residual organics from MNLs can compromise bonding quality. To overcome this, we developed an innovative, contact-free removal technique using VUV irradiation under organic acid atmosphere. This method leverages high-energy photon-assisted decomposition, where VUV irradiation cleaves molecular bonds, while organic acid vapor not only facilitates the generation of hydroxyl radicals (OH•), but also lowers the bond dissociation energy of the molecular nanolayers through chemical interaction, thereby enhancing their photodegradability. The resulting OH• species react with the Cu surface to form Cu(OH)₂, which upon heating dehydrates to Cu–O–Cu linkages, serving as an interfacial bridge that promotes atomic diffusion and strengthens metallurgical bonding. This approach offers a reliable and environmentally friendly solution for enhancing bonding performance and interconnect integrity in advanced packaging.

 

Bio:
2023 – Present Master student, National Cheng Kung University, Tainan, Taiwan 2019 – 2023 bachelor's degree, Feng Chia University, Taichung, Taiwan

Abstract:
Moore's Law claims that transistor density doubles approximately every two years. As transistor sizes shrink and density increases, the performance of electronic devices continues to improve. To meet these demands, optical lithography has become a key technology in semiconductor manufacturing. Early generations of lithography employed mercury lamps with g-line (436 nm) and i-line (365 nm) wavelengths. As feature sizes continued to decrease, these sources were replaced by Deep Ultraviolet (DUV) lithography using 248 nm (KrF) and 193 nm (ArF) excimer lasers. To further extend the resolution limit, immersion DUV lithography was introduced, in which a high-refractive-index fluid is inserted between the lens and the wafer to enhance the numerical aperture. However, even with immersion techniques and resolution enhancement strategies, DUV systems have approached their physical limitations.
As semiconductor devices continue to evolve with shrinking dimensions and narrower line widths, Extreme Ultraviolet Lithography (EUVL) has become a critical process in advanced manufacturing. In this process, tin is used as the light source material. It is excited by a CO2 laser into a plasma state, emitting EUV light with a wavelength of 13.5 nm. Ruthenium is commonly deposited as a capping layer on the EUV mask to protect the Mo/Si reflective layer from contamination and oxidation. The Ru capping layer is easily contaminated by Sn droplets from the light source, which can react with Ru at the interface, leading to Sn contamination and deteriorating the quality of the exposure mask pattern. This interfacial reaction not only reduces the reflectivity of the multilayer mirror but also complicates Sn removal due to the formation of intermetallic compounds. Therefore, understanding the interfacial behavior between Sn and Ru is essential for improving mask durability and ensuring stable EUV lithography performance.
However, the interfacial reaction between ruthenium and tin has not been studied yet. Therefore, this study will investigate the morphology, growth rates, and evolution process of intermetallic compounds (IMCs) of this system at different temperatures and reflow times, and the reaction mechanism under thermodynamics and kinetics will be clarified. This study investigates the reflow process under a vacuum environment of 1 Pa at reflow temperatures of 300°C and 350°C for different reaction times: 24 h, 18h, 12 h, 6 h, and 3 h. The research finds that at the Sn/Ru interface, the Ru3Sn7 intermetallic compound was formed. As the reflow time increases, the IMC thickness increases linearly, exhibiting reaction-controlled behavior. Within this IMC layer, three distinct microstructures are observed: a dense layer, a loose layer, and a solidified precipitate layer, respectively.
This study provides a fundamental understanding of the interfacial reactions between Sn and Ru. Through phase diagrams, thermodynamics, and kinetic analysis, the mechanisms of the interfacial reactions and the growth kinetics of the intermetallic compounds (IMC) will be revealed. This research can help to develop effective processes for removing Sn contamination from the Ru substrate and dealing with the Sn contamination issue.

 

Bio:
Yu Sheng received his B.S. degree in Materials Engineering from Ming Chi University of Technology, New Taipei City, Taiwan, in 2024, and is currently an M.S. student in Materials of Science and Engineering at National Taiwan University of Science and Technology, Taipei, Taiwan. His research interests include electroplating and interface reactions.

Abstract:
This study aims to develop high-temperature lead-free solder alloys. Traditional Pb-Sn solders, with a lead content of about 85–97 wt.%, are widely used for internal electronic connections and high heat-resistant joints, operating at temperatures between 300 and 350°C. These are commonly referred to as high-temperature solders and are applied in electronics, optical devices, and automotive components. However, due to environmental concerns, the European Union has enacted the WEEE (Waste Electrical and Electronic Equipment) and RoHS (Restriction of Hazardous Substances) directives, which restrict the use of lead in electronic products. This study investigates Sn-Zn-Al-Cu base alloys with trace Ni and Ge additions to refine microstructure, control solidification phases, adjust liquidus temperature, and correlate mechanical properties with alloying elements. High-purity Zn, Sn, and Al were weighed in different proportions to prepare 18 groups of Sn-Zn-Al alloys, each with a total weight of approximately 5 grams. Each alloy composition was individually doped with 0.01 wt.% Ni, 0.002 wt.% Ge, and 5 wt.% Cu. To prevent oxidation-reduction reactions between Al in the alloy and the quartz glass, the prepared alloys were first placed in graphite crucibles, which were then inserted into quartz tubes. The tubes were evacuated to a pressure below 10-3torr and vacuum sealed. The sealed samples were placed in a high-temperature furnace at 950°C for 72 hours to ensure uniform mixing of all elements. Subsequently, the alloys were quenched in ice water and extracted. The alloys were mounted in conductive phenolic resin powder and subjected to metallographic preparation. The surface morphology of the alloys was observed using a scanning electron microscope (SEM), and the microstructural composition was analyzed using energy dispersive spectrometry (EDS) or electron probe micro-analyzer (EPMA). X-ray diffraction (XRD) analysis was employed for phase identification of the solidified precipitates, with diffraction peaks compared against the Joint Committee on Powder Diffraction Standards (JCPDS) database. Thermal analysis was conducted using differential scanning calorimetry (DSC) with a heating rate of 5°C/min. The DSC heating curves were used to determine phase transformations and liquidus temperatures. After metallographic preparation, the hardness of the prepared alloys was measured using a hardness tester with a load of 200 gf applied for 10 seconds, and the average hardness values were calculated. Tensile testing was performed using a universal testing machine with a tensile rate of 1 mm/min. Experimental results show that the Sn-Zn-Al-Cu-based alloy produces phases such as (Zn), β-Sn, and CuZn5. The liquidus temperature ranges from 275°C to 375°C, and an increase in Zn content raises the liquidus point accordingly. In addition, the increase of Zn content promotes the formation of (Zn) and CuZn5 phases, which are associated with improved hardness and tensile strength. Al also plays a significant role in mechanically strengthening the alloy. The tested hardness values range from 27 to 85 HV, while the ultimate tensile strength lies between 50 and 86 MPa. Both hardness and tensile strength increase with higher Zn and Al contents.

 

Bio:
Zhi-Chen Ai is a Master’s student in the Department of Mechanical Engineering at National Cheng Kung University. He specializes in the design of low-temperature solder alloys and the analysis of their mechanical properties. In this presentation, he will share his latest advances in applying machine learning to optimize material compositions in the Sn–Bi–X system.

Abstract:
As electronic packaging advances toward fine-pitch technologies, substrate warpage during the 230–250 °C reflow soldering process causes misalignment of solder joints, severely affecting product yield [1]. Low-temperature solders (LTS) such as Sn–Bi (eutectic melting point 139 °C) and Sn–In can significantly reduce thermal stress and have therefore attracted considerable attention [2]; however, the Bi phase in Sn–Bi systems coarsens during thermal aging, leading to decreased ductility and reliability [3]. Traditional composition optimization by adding ternary elements such as Zn, Ge, Al, Ag, and Sb [4-6] can enhance performance but requires extensive trial-and-error, which is both time-consuming and labor-intensive. In recent years, machine learning has demonstrated high efficiency in new-material development, markedly shortening research and development cycles [7]. Therefore, this study employs machine learning to design high-shear-strength alloys and validates the predictions experimentally. Sn–Bi–X (X = Zn, Ge, Al, Ag, Sb) solder alloys were prepared by rapid thermal annealing (RTA). Reflow soldering followed a preheat at 100 °C for 3 min and a reflow temperature of 170 °C for 1 min. Shear strength of the solder joints on Cu substrates was measured at a shear rate of 1 mm/min. The resulting shear-strength data were used to train machine-learning models, which were evaluated based on the coefficient of determination (R²) and root-mean-square error (RMSE). Model predictions and experimental measurements showed high concordance as Ge content increased, indicating that Sn–Bi–Ge solder alloys can potentially reach shear strengths of approximately 70 MPa. This study demonstrates that machine learning can efficiently screen and optimize the composition of Sn–Bi–X low-temperature solders, achieving high shear strength after reflow soldering on Cu substrates.