AREA CONSORTIUM REPORTS

2021 REPORTS

Liquid Metal Thermal Interface Material Processing
Author: Xinyu Zhang

Abstract:
With the continually increasing power density and the diminishing size of the micro-chips, the heat flux density of electronic components has increased rapidly. Thermal interface materials (TIMs) are used widely in electronic packaging for thermal management, to dissipate concentrated heat generated in the microprocessors. Traditional TIMs including thermal grease, phase change materials, and silicone pads are usually based on polymer materials with their thermal conductivity being enhanced by the addition of thermally conductive particles. But now their cooling capacity is often insufficient to meet the demands of high-performance electronic components. Liquid metal (LM) with high electrical conductivity, thermal conductivity, and extraordinary fluidity has emerged as a promising class of TIMs. However, LMs feature several issues such as oxidation, intermetallic growth, dewetting and leakage.

In this project, we propose a method to prepare a sticky liquid metal paste with good plasticity by mixing gallium based liquid metal and metal particles. As a result, we found that nickel particles are the ideal and practical particle for TIMs. In addition, printing process for nickel particles mixture was designed.

Reliability Comparison of Underfilled WLCSP Assembled using No-clean Flux and No-clean Paste Dipping Processes
Author: Michael Meilunas

Abstract:
Wafer Level CSP components of various sizes were assembled to printed circuit boards using either a no-clean flux dip or a no-clean paste print process prior to underfill operations. The samples were then subjected to -40 to 125°C accelerated thermal cycling to stress the assemblies and drive solder joint fatigue failures. The failure data was compiled and the subsequent analysis indicated that the samples soldered using the no-clean flux survived longer than the samples soldered with the no-clean paste. The data was then compared to previous AREA research in which similar test vehicles had been assembled using a cleaning process prior to underfilling while other samples had been assembled without underfill. These comparisons would show that underfilling the components dramatically improved lifetime and that cleaning the samples prior to underfilling also had a positive impact on lifetime.

Edge Bonding TB2019-U Components with ZYMET UA-3307-B, ZYMET-2605-B, and Panasonic CV5797U Edgebond Materials
Author: Pericles A. Kondos

Abstract:
Earlier reports have described the TB2019-U test vehicle, how it was populated with electrically testable dummy BGAs, varying in size, thickness, pitch and number of I/Os and how it was divided into six roughly equal sets, two of which were underfilled with two capillary underfills while one was set aside to be used as a non-underfilled reference. The present report deals with the three remaining sets of test boards. They were used for edge-bonding studies which utilized three edgebond materials, namely the ZYMET UA 3307 B, the ZYMET-2605-B, and the Panasonic CV5797U. One of these materials always flowed to a small extent after dispensing, consistently reaching the first row of joints in spite of all the efforts to prevent it from doing so. The other two did not exhibit this tendency, barely wetting the underside of the edge-bonded parts. One material, marketed as reworkable, was additionally used in reworkability studies. After edge bonding, all boards were subjected to Accelerated Thermal Cycling; its results, together with failure analysis, will be presented in later reports.

Thermal Cycle Analysis of SMT Components Assembled to Low CTE PCB Laminate
Author: Michael Meilunas

Abstract:
The thermal cyclic lifetimes of surface mount SAC305 solder joints were measured on low CTE laminate printed circuit boards and compared to solder joints assembled to more traditional glass-epoxy resin laminate based printed circuit boards. Several component types including WLCSP, SMR, MLF, BGA and LGA were evaluated.

Evaluation of Reliability of Graphite Pad TIMs: In Situ Electrical Monitoring and Postmortem Failure Analysis
Authors: Peter McClure, Dylan Richmond

Abstract:
The reliability of thermal interface materials (TIMs) in large complex printed circuit board (PCB) assemblies is challenging to predict or to model so experimental study is needed. Thermal cycling, 500 cycles from 0 and 100⁰C, of four graphite pad thermal interface materials [Allied Tpad SO2001, Sekisui Manion SC, Jones 21-1500A, and Nolato Compatherm 9611] was conducted on hardware that was mechanically representative of large PCBs with common heat sinks such as server motherboards. The electrical resistance of the TIMs was monitored in situ during the thermal cycle giving insight into the mechanical response of the TIMs. In combination with postmortem failure analyses, the reliability of these TIMs were evaluated. A nominal bond line thickness of 1100µm was used for this study, representative of TIM2 pad bond lines.

Keywords: TIM2, ATC, Graphite Pad, Reliability, Failure Analysis, Electrical Resistance

Evaluating Mechanical Cycling of a Thermal Interface Material as a Pre-Screening for Accelerated Thermal Cycling Response
Author: Peter McClure

Abstract:
The reliability of thermal interface materials (TIMs) in electronic assemblies with a large printed circuit board (PCB) and common heat sinks is challenging to predict or model so experimental study is required. Thermal cycling of TIMs in mechanically representative structures remains a preferred way to evaluate TIM reliability. While such accelerated thermal cycling (ATC) tests are relatively fast (taking a few months), samples are physically large, require ATC environmental chambers, and allow only a few TIMs to be tested during months of ATC. This work evaluated whether accelerated mechanical cycling (AMC) could help select what TIMs should be continued onto ATC testing. If AMC could test materials on the order of hours or days, it could be used to down-select the best TIMs for ATC testing. Mechanical cycling was done with a Dage 4000+, shearing TIMs with high strain rates at room temperature. The high strain rates allowed for many cycles to be run in a small amount of time. It was found that mechanical cycling alone cannot predict TIM reliability behavior. Without the added temperature excursion and with high strain rate loadings, TIMs behaved in ways that resulted in different failure modes than did ATC, indicating that AMC and ATC did not readily correlate.

Keywords: TIM 2, Accelerated Mechanical Cycling, ATC, Failure Analysis, Capacitance

Glob Topped QFN Power Cycle Testing
Authors: Michael Gaynes, Michael Meilunas

Abstract:
Previously, power cycle testing was completed on a 12 × 12 mm QFN open cavity package where the target power-on chip temperature was 130 ⁰C that resulted in a corner solder joint temperature of 80 ⁰C. Twelve QFNs were power cycled out to > 31,000 cycles. Three parts failed with typical bulk fatigue in the corner solder joints. Only one of these fails was relevant. A runaway chip temperature on the other QFNs negated any reliability assessment. However, the one relevant failure occurred before the chip temperature increased inordinately and therefore, indicates a reliability concern and the need for continued study. An additional weakness in this testing was the use of an open cavity QFN package, which is not representative of the majority of QFN packages used that are fully encapsulated and will have higher stiffness. A second power cycle test has been completed where the open cavity QFN package was glob topped with an encapsulant that had cured properties matching the cured properties of the transfer molding compound used with the lead frame of the open cavity package. This glob topped QFN package was thermally and mechanically representative of a standard fully encapsulated QFN. Twelve QFN packages were solder attached to a 3 mm thick printed circuit board. The solder voiding between the QFN lead frame thermal pad and thermal pad on the PCB has typically > 40%, typical industry experience. Power cycle testing was performed over a period of 18 months, accumulating 32,647 cycles of 10 minutes on followed by 10 minutes off. The chips were heated to a target temperature between 120 and 130 °C. The corner solder joints were monitored electrically using both event detection and digital multimeter voltage drop. No electrical fails, instabilities nor trends were observed the entire test, demonstrating solder joint robustness and reliability for the fully encapsulated QFN package.

Key words: QFN, power cycle

Effects of Room Temperature Storage on TIM Thermal Behavior and Other Observations
Author: Peter McClure

Abstract:
The question of whether Thermal Interface Materials (TIMs) could be satisfactorily used after their listed expiration date was of general AREA member interest. The effects of long-term room temperature storage of thermal interface materials on their thermal and mechanical properties were therefore investigated. This study used thermal rod testing to evaluate the thermal performance of various TIMs when they were used within their “Best by” date as well as after years of room temperature storage. Observations associated with handling the aged TIMs were also made. Results for various types of materials were covered.

Keywords: TIM 2, Thermal rod testing, Room temperature aging