Preface

This report was prepared by Jean Paul Clech through funding support from the National Institute of Standards and Technology. The need for this critical evaluation of mechanical property data of Pb-free solders was established through the NEMI-NIST-NSF-TMS "Workshop on Modeling and Data Needs for Lead-Free Solders."

Introduction

Insight into the thermo-mechanical response of solder joints is critical to the design and deployment of reliable electronic circuit board assemblies. Understanding the mechanics of lead-free soldered assemblies is also essential to the development of accelerated test plans, predictive reliability models and to their use as effective tools for product reliability assessment.

The life of SnPb or lead-free solder joints is limited by the fatigue damage that accumulates in solder materials. Accelerated testing provides distributions of failure times whose relevance to service life is determined by extrapolation to use conditions based on the appropriate Acceleration Factors (AFs). Since AFs are defined as the ratio of life under test and use conditions, their determination requires up-front predictions of solder joint lives under both sets of conditions. Such models are also of use for reliability analysis of circuit boards at the design stage. While a variety of life prediction models have been developed for near-eutectic SnPb assemblies, to this author's knowledge, such models are not currently available for lead-free soldered assemblies. The development of life prediction models requires a detailed understanding of failure modes, an adequate constitutive model that captures the thermo-mechanical behavior of lead-free solders in electronic assemblies and a reliability database that is needed for the empirical correlation of failure times under test and/or field conditions.

Some reliability data has been acquired or is in the process of being gathered by several lead-free consortia. Efforts are also underway to characterize the mechanical behavior of lead-free solders and numerous studies have been published with emphasis on secondary creep of solder. The derived creep rate equations are an important ingredient of constitutive models, however, their applications to circuit board assemblies are still the subject of validation studies.

This report provides a quantitative review of the thermo-mechanical properties of lead-free solders, with emphasis on Sn-Ag-Cu (SAC) and Sn-Ag alloys. The Sn-Ag-Cu alloys of near-eutectic composition are labeled "SAC" throughout the report, including the NEMI-selected Sn-3.9Ag-0.6Cu alloy. SAC solders are the alloys of choice for solder reflow assemblies. Eutectic SnAg is also recommended for wavesoldering applications. The microstructure of both alloys consists of a Sn matrix with finely dispersed intermetallic precipitates. We thus expect similarities in the constitutive response of SAC and eutectic SnAg solders. Given the more abundant literature on SnAg, its properties are reviewed in detail in an attempt to better understand the qualitative behavior of precipitate-strengthened solder alloys.

To put things in perspective, this report starts with a review of SnPb properties and lessons learned from the development of constitutive and life prediction models for near-eutectic SnPb assemblies. This understanding of SnPb behavior, although not fundamentally complete, has proven valuable to industry. It is also worthwhile observing that the characterization of SnPb solder and the development of reliability models for SnPb assemblies spans over three decades of research in industry and academia. For example, one of the earliest and well-known solder joint reliability models is the Norris-Landzberg (1969) model for non-underfilled flip-chip assemblies. While some of the lessons learned with SnPb will transfer to lead-free applications, a significant amount of research and development is needed for lead-free reliability models to come up to par with their SnPb equivalents.

The report then proceeds with the results of an extensive, although not exhaustive, review of material properties for SAC and SnAg alloys. Gaps in the material properties database are identified and suggestions are offered as to what additional testing and analysis are needed to develop constitutive models for engineering use.