Files

Résumé

Solders are widely used as interconnect material for the electronic industry. The solder interconnect provides both electrical contact and mechanical connection between the integrated circuit devices and their substrate. Thus solders with good reliability are required to facilitate successful functioning of electronic devices and components that we encounter in daily life. However traditionally used Sn-Pb solders have to be replaced by Sn rich lead-free solders because of the toxicity concerns surrounding the use of lead. Many Sn-rich lead-free solders do exists, however they are not drop-in substitutes for Sn-Pb solders as the knowledge concerning their mechanical behavior and reliability are not well understood. A particularly critical aspect is the lead-free solders' reliability at high service temperatures required in several microelectronics applications. Such applications require solders possessing good microstructural stability, strength and creep resistance. In this regard, literature reports the alloying and the composite approach to improve the properties of the lead-free solders. Among these two methods, the composite approach wherein suitable particles are added in to the solder matrix seems to be promising. However, the scientific understanding between property improvement and characteristic of the particle reinforcement is too limited. Also the processing of the composite solder is limited by stringent process parameters that require significant improvement to increase its industrial applicability. Another important issue arises from the fact that solders as such exhibit complex elasto-plastic properties. Furthermore the solder's mechanical response is greatly influenced by specimen geometry and process parameters. Therefore novel localized strain measurement techniques that could assess and understand the mechanical response of the solder without much effect from substrate are required. The goals of the current PhD thesis are identified within the framework of the issues previously mentioned. The composite approach is demonstrated to improve the microstructural and the mechanical properties of existing lead-free solders. As reinforcing particles either Cu or Ni is used with volume fraction from 0.8-4.2 vol. %, while Sn-4.0Ag-0.5Cu (SAC405) alloy is chosen as matrix. Throughout the study the results of the composite solder is compared against the un-reinforced SAC405 reference solder. The solder material in bulk form (reflowed in ceramic crucible) is used for microstructural analysis and also as joining material between Cu substrate to investigate the mechanical response in tension and shear. The current research demonstrates a new processing method developed for producing composite lead-free solders utilizing micro and nanometer scale metallic particles. Generally the preparation of nano-composite solders requires several process parameters that restrict their industrial use. In literature even the exact process route is still not disclosed due to technical secrecy. This processing issue could be overcome by adopting the mechanical mixing method reported in this research. A significant improvement has been achieved in reducing porosity within the newly developed composite solders. This reduced porosity improves the consistency of mechanical properties in many folds. Till date no standardized mechanical test method for solder alloys has been reported. In this regard, the design of a shear test specimen proposed in this PhD thesis is novel. Extensive finite element (FE) simulations accompanied by series of experimental tests is performed to optimize the specimen geometry. This proposed specimen geometry minimizes the effect of plastic strain localization which ensures almost uniform plastic strain at the center of the joining layer. This design leads to "cleaner" shear tests minimally affected by stress concentration and with a low risk of premature failure at the substrate/solder interface. As a consequence, reliable mechanical properties can be identified from experimental results. In case of tensile specimens, the gap width is chosen such that effects of plastic constraints from rigid substrates (Cu) are kept at minimum. In-situ optical extensometry combined with Digital Image Correlation (DIC) is used to determine the stress-strain response of the solder joint. Finally, the constitutive properties of solder materials are numerically identified from the global stress-strain response of the solder joint by inverse FE modeling of the shear test. For the first time, a 3D FE homogenization model for particle reinforced composites is employed to predict the solder elastoplastic behavior as a function of the particle volume fraction. This objective has been demonstrated on the example of composite solder prepared with micrometer sized Ni particles. The homogenization results are then compared with the constitutive model of the composite solder identified from the experimental data. This comparison serves as a basis for the discussion of the effects of particle reinforcement in the composite solder. Another objective of the current research is to investigate the effect of adding reactive particles to the solder alloy on the subsequent growth of the interfacial intermetallic layer and the ensuing tensile strength of the joint. To address this objective, two composite solders (solder alloy with additional particles) were fabricated for comparison against the un-reinforced alloy through aging at 150 °C for up to one week. Either micrometer scale or nanometer scale Cu particles are employed as the compositing particles to subsequently produce additional Cu6Sn5 intermetallic reinforcing phases. The expectation is that the large difference in initial particle size of this additional source of Cu will help to elucidate the rate limiting mechanism growth of the interfacial intermetallics. It is demonstrated that while a similarly reduced steady-state of intermetallic growth is obtained for the two composite samples, it is achieved sooner in the one fabricated with the nanometer scale particles. Mechanical tests indicate that the composite specimen prepared with nanometer scale particles also provides the highest tensile strength both before and after aging. As a final objective, creep tests were conducted at room temperature to assess the steady state creep strain rate of composite solders. Though many publications report creep tests conducted at various temperatures, the current study is unique with respect to the attempt made to clarify the effect of applied stress on the creep strain rate of the composite solders. Subsequent effects of particle reinforcement on the creep strain rate are discussed. It is demonstrated that composite solder prepared with Cu nanoparticles exhibit the best creep properties (lowest creep strain rate) among the solders considered in this study. The mechanisms responsible for strength increase, ductility loss and creep properties observed for various composite solders are rationalized based on the current understanding of the microstructure-mechanical response relationship. Differential Scanning Calorimetric (DSC) studies were conducted to investigate the solidification temperature of the particle reinforced lead-free solders. The distinct microstructural features obtained for composite solder are shown to be significantly influenced by disparity in solidification kinetics caused by particle addition. In this context a key result achieved during this research was the solidification mechanism proposed for the composite solders. Overall the results of this research lead to: (i) development of a viable processing method for nano composite solders with excellent creep properties and (ii) demonstration of a novel approach (experimental and modeling) to better understand the effect of particle volume fraction on the elasto-plastic response of the composite solder. This approach will provide some valuable guidelines for tailoring the properties of composite lead-free solder based on the target application.

Détails

PDF