Thermal Cycling Dwell Time on Reliability of Pb-free Solder Alloys

Richard Coyle, Chloe Feng, Isaac Becker
Nokia Bell Laboratories, NJ, USA

Dave Hillman
Hillman Electronic Assembly Solutions, IA, USA

Michael Osterman
CALCE, MD, USA

Tim Pearson
Collins Aerospace, IA, USA

Joe Smetana
Nokia, TX, USA

Keith Howell
Nihon Superior Co., Ltd., Osaka, JP

Julie Silk
Keysight Technologies, CA, USA

Hongwen Zhang, Jie Geng
Indium Corp., NY, USA

Jayse McLean, Ranjit Pandher
John Deere, ND, USA

Derek Daily
Senju Comtek Corp., CA, USA

Anna Lifton
MacDermid Alpha Electronics Solutions, NJ, USA

Morgana Ribas
MacDermid Alpha Electronics Solutions, Bangalore, IND

Raiyo Aspandiar
Intel, OR, USA

James Wertin
Heraeus Electronics, PA, USA

Jean-Christophe Riou
Safran Electronics and Defense, Valence, FR

Madan Jagernauth
HDP, TX, USA

Grace O’Malley
iNEMI, NC, USA

Martin Anselm
Rochester Institute of Technology, NY, USA

Dennis Fritz
Fritz Consulting, OH, USA

Shantanu Joshi, Jasbir Bath
Koki Solder America, OH, USA

The past decade has seen the development and introduction of commercial, third-generation, high-performance Pb-free solder alloys designed to meet the requirements of higher temperature use environments. Most of these offerings are based on the Sn-Ag-Cu (SAC) system, with major alloying additions of bismuth (Bi), antimony (Sb), or indium (In). These elements, individually or in combination, promote additional precipitate, solid solution, or dispersion strengthening that can enhance resistance to degradation at elevated temperature or during aggressive thermal cycling.

Results from the literature show that an increase in thermal cycling dwell time can decrease the thermal cycling reliability of SAC solders. Because high-performance alloys are designed for extended operation at higher temperatures, it is important to understand their behavior and characterize their reliability at extended thermal cycling dwell times.

This paper presents the initial results from an experimental program designed to compare thermal cycling results for high-performance solder alloys using an extended dwell of 60 minutes to a typical short dwell time of 10 minutes. The 10-minute dwell data were generated in the initial phase oftesting and published previously. The data reported here arefrom a thermal cycling profile of -55/125 °C (TC7 in IPC-9701B) and the test vehicle is a 192-pin chip array ball gridarray (192CABGA). Contrary to the results for SAC solders,the 60 minute dwell time did not reduce the reliabilityconsistently for all the high-performance alloys in the testmatrix.

Based on evaluation criteria of characteristic lifetime and 1% cumulative failure rate from a 2-parameter Weibull plot, the high-performance alloys had comparable reliability performance with 60-minute and 10-minute dwell times. Although all the alloys exhibited fatigue failures in the bulk solder, many of the alloys also exhibited interfacial and mixed mode failures, which complicates interpretation of the data. Multiple failure modes for these solder alloys also were reported for the 10-minute dwell testing.

This paper presented the initial findings from an experimental program designed to compare thermal cycling results for high-performance solder alloys using an extended dwell of 60 minutes to a typical short dwell time of 10 minutes. Results are reported for a daisy chained, 192-pin chip array ball grid array (192CABGA) component test vehicle tested with an accelerated thermal cycling profile of -55/125 °C. The alloy reliability test matrix consisted of six Pb-free, high performance solder alloys that were developed for use in aggressive service environments. These alloys are based on the Sn-Ag-Cu (SAC) system, but modified with combinations of major alloy additions of Bi, Sb, and In. The SAC305 hypoeutectic alloy was used as the performance baseline.

In contrast to the published results for SAC solders and the current results for the SAC305 baseline, the 60-minute dwell time did not reduce the reliability consistently for all the high-performance alloys in the test matrix. Based on evaluation criteria of characteristic lifetime and 1% cumulative failure rate from a 2-parameter Weibull plot, the high-performance alloys had comparable reliability performance with 60-minute and 10-minute dwell times. Although all the alloys exhibited fatigue failures in the bulk solder, many of the alloys also exhibited interfacial and mixed mode failures, which complicated interpretation of the data. The Weibull slopes for the 60-minute dwell test were higher than those for the 10-minute dwell test, except for SAC305 and 279, the only alloys that did not have interfacial failures.

Multiple failure modes for these solder alloys had also been reported previously for the 10-minute dwell testing. Despite the complications introduced by the occurrence of multiple failure modes, the high-performance alloys outperformed SAC305 by a significant amount. These high-performance alloys perform well with the 60-minute dwell time, at least with the 192CABGA component and the -55/125 °C thermal cycling profile. Results for the 84CTBGA component and two other thermal cycling profiles will be presented later.