The objective of this study is to understand the effect of surface roughness of the Cu substrate on the wetting of molten solder alloys. Eutectic Sn-Pb, pure Sn and eutectic Sn-Cu solder alloys and Cu substrates with different surface finish viz., highly polished surface, polished surface and unpolished surface were used in this work. Highly polished surface was prepared in Metallography lab, University of Kentucky while other two substrates were obtained from a vendor. Surface roughness properties of each substrate were measured using an optical profilometer. Highly polished surface was found to be of least surface roughness, while unpolished surface was the roughest. Hot-stage microscopy experiments were conducted to promote the wetting behavior of each solder on different Cu substrates. Still digital images extracted from the movies of spreading recorded during hot-stage experiments were analyzed and data was used to generate the plots of relative area of spread of solder versus time. The study of plots showed that surface roughness of the Cu substrate had major influence on spreading characteristics of eutectic Sn-Pb solder alloy. Solder showed better spreading on the Cu substrate with least surface roughness than the substrates with more roughness. No significant influence of surface roughness was observed on the wetting behavior of lead free solders (pure Sn and eutectic Sn-Cu).

Dissertation (Ph.D.)--Auburn University, 2007. / Abstract. Vita. Includes bibliographic references.

abstract: Pb-free solders are used as interconnects in various levels of micro-electronic packaging. Reliability of these interconnects is very critical for the performance of the package. One of the main factors affecting the reliability of solder joints is the presence of porosity which is introduced during processing of the joints. In this thesis, the effect of such porosity on the deformation behavior and eventual failure of the joints is studied using Finite Element (FE) modeling technique. A 3D model obtained by reconstruction of x-ray tomographic image data is used as input for FE analysis to simulate shear deformation and eventual failure of the joint using ductile damage model. The modeling was done in ABAQUS (v 6.10). The FE model predictions are validated with experimental results by comparing the deformation of the pores and the crack path as predicted by the model with the experimentally observed deformation and failure pattern. To understand the influence of size, shape, and distribution of pores on the mechanical behavior of the joint four different solder joints with varying degrees of porosity are modeled using the validated FE model. The validation technique mentioned above enables comparison of the simulated and actual deformation only. A more robust way of validating the FE model would be to compare the strain distribution in the joint as predicted by the model and as observed experimentally. In this study, to enable visualization of the experimental strain for the 3D microstructure obtained from tomography, a three dimensional digital image correlation (3D DIC) code has been implemented in MATLAB (MathWorks Inc). This developed 3D DIC code can be used as another tool to verify the numerical model predictions. The capability of the developed code in measuring local displacement and strain is demonstrated by considering a test case. / Dissertation/Thesis / M.S. Mechanical Engineering 2011

The constitutive mechanical behaviour with a focus on creep of Sn-Pb and various Sn-Ag-Cu based Pb-free solders in the 25-150°C temperature range has been studied using nanoindentation and various new meso-scale tests. All alloys have been studied as bulk wave soldering bars, as-received solder balls and solder joints. Ball Grid Array (BGA) solder joints in a typical electronic configuration were manufactured in-house using both Cu and Pd-Ag metallizations. Microstructural characterisation of all configurations used various types of optical and electron microscopy and showed that the solder pad metallization type played a major role in intermetallic compound (IMC) formation. There were comparatively fine and coarse-grained microstructures in both as-received solder balls and BGA solder joints depending on ball diameter. Nanoindentation creep measurements in the stress range 20-500MPa showed that grain boundary sliding occurred together with dislocation glide and dislocation climb in the low temperature (25-50°C) and high temperature (100-150°C) regimes respectively. Smaller grain sizes (<20µm) encouraged grain boundary sliding that enhanced creep. New elevated temperature mechanical tests were developed using the nanoindentation platform to enable testing of entire solder joints in shear and compression, with stresses in the 1E-2 - 3MPa range, more relevant to in-service conditions than those in nanoindentation. Meso-scale spherical indentation creep behaviour in compression on as-reflowed solder balls showed good agreement with that obtained by conventional nanoindentation. However, when BGAs were tested in shear, the solder microstructure had relatively little influence on the creep response, which was significantly less creep resistant than individual phases in the ball obtained by nanoindentation or the ball itself obtained by meso-scale spherical indentation. In shear, the creep conformed to diffusion controlled behaviour and interfacial microstructure was suggested to now control creep response, with the microstructure of the majority of the solder joint playing only a minor role.

Joining of electronic components onto the circuit boards is done by soldering operation, during production of all electronic devices. In many countries, including Turkey, traditionally used tin-lead (Sn-Pb) solder alloys have been restricted to be used in consumer electronics appliances because of the toxic effects of lead (Pb) within these alloys. Tin-silver-copper (Sn-Ag-Cu) based alloys have been developed as the most promising candidate that can replace the Sn-Pb alloys. However, various problems have emerged with the increasing trend in use of Sn-Ag-Cu solder alloys in electronics industry, namely large intermetallic compound formation, low wettability and thermal shock resistance. Many researches have been done in the past decade to overcome these problems. The solutions are based on changing the undercooling of the solder alloy / which was determined to be done by either changing the composition of the solder alloy by micro-alloying or changing the cooling rate during soldering operation. In this thesis study Sn-3.5Ag-0.9Cu (wt. %) lead-free solder having the eutectic composition, was micro-alloyed with additions of aluminum (Al), iron (Fe) and titanium (Ti). Experimental results were compared with commercially available near-eutectic Sn-40Pb (wt. %) solder, a commercially available Sn-3.0Ag-0.5Cu (wt. %) solder and also eutectic Sn-3.5Ag.0.9Cu (wt. %) and near-eutectic Sn-3.7Ag-0.9Cu (wt. %) solders that were produced for this thesis study. In the first stage of the study, the effects of 0.05 wt. % of Al, Fe and Ti micro-alloying were investigated. When preliminary results of mechanical and thermal test were compared, Fe was found to make positive effect on shear strength and undercooling. Further research was carried out to establish a relationship between the Fe compositions and solder properties. Therefore, 0.01, 0.03, 0.07 and 0.1 wt. % Fe additions were also studied and results were reported. 0.01 wt. % and 0.07 wt. % Fe added solders were found to have a smaller undercooling, resulting with dispersed intermetallic compound (IMC) and thus has highest shear strength. Different cooling rates / 0.017, 0.17 and 1.7 &deg / C/sec were applied to solder-copper joints and microstructures were investigated. Large IMC-free microstructure was achieved by 0.01 wt. % Fe micro-alloyed solder, which was cooled with 1.7 &deg / C/sec rate. Wetting of copper substrate was found to be improved by additions of Al, Fe and Ti compared to alloy with eutectic composition of Sn-Ag-Cu alloy. Selected SAC+X alloys have been subjected to thermal shock experiments for crack formation analysis on the copper substrate-solder joints. The results showed that SAC+0.05Al solder has the higher thermal shock resistance, which no cracks were observed after 1500 cycles of thermal shock. In order to understand the insights of SAC performance, some of the lead-free solders were applied onto printed circuit boards for thermal shock resistance test. These results have indicate that the cracking may occur after thermal shock cycles due to process conditions of soldering operation (i.e. cooing rate), independent of the solder alloy composition.

This thesis explores mechanical behavior of microelectronic devices and lithium-ion batteries. We first examine electromigration-induced void formation in solder bumps by constructing a theory that couples electromigration and creep. The theory can predict the critical current density below which voids do not form. Due to the effects of creep, this quantity is found to be independent of the solder size and decrease exponentially with increasing temperature, different from existing theories. / Engineering and Applied Sciences

The disturbed state concept (DSC) based constitutive model is the focus of this research. It is applied for characterizing two problems; thermomechanical reliability analysis of electronic packages, and prediction of glacial motion. A new procedure for construction of static yield surface for materials is proposed. Further, a modified DSC model to include effect of rate of loading on material behavior is proposed.The DSC is applied to characterize the behavior of Sn-3.9Ag-0.6Cu (SAC) lead free solder alloy used in electronic packages. Proposed procedure of construction of static curve and rate dependent DSC model is applied for prediction of creep and rate dependent behavior of the SAC alloy. Laboratory test data is adopted from the literature and material parameters are determined. The DSC model is validated using the derived material parameters. A finite element analysis of the BGA 225 package is performed under cyclic thermomechanical loading. Analysis results are compared with available test data. A failure criterion for prediction of number of cycles to failure for BGA 225 is then derived.The second application of DSC discussed in this work is prediction of glacial motion. Mechanical behavior of glacial till and its contribution to overall ice movement is characterized using DSC. Two regionally significant tills are chosen and samples are collected from field. A series of laboratory tests are conducted on samples. Tests results are used for model calibration and validation. A numerical simulation of an idealized ice - till system under gravity loading is performed. Two such analyses are performed with DSC and Mohr Coulomb models and the results are compared.The DSC predicts failure when a significant portion of the material reaches a critical disturbance whereas the Mohr Coulomb model predicts failure based on peak stress. DSC predicts a gradual progression to failure whereas the Mohr Coulomb model predicts early catastrophic failure. According to DSC, the material undergoes considerable plastic strains before it reaches failure whereas the Mohr Coulomb predicts failure at very low elastic strains. In general the DSC is considered to provide a more realistic and general constitutive model for glacial tills.

abstract: Microelectronic industry is continuously moving in a trend requiring smaller and smaller devices and reduced form factors with time, resulting in new challenges. Reduction in device and interconnect solder bump sizes has led to increased current density in these small solders. Higher level of electromigration occurring due to increased current density is of great concern affecting the reliability of the entire microelectronics systems. This paper reviews electromigration in Pb- free solders, focusing specifically on Sn0.7wt.% Cu solder joints. Effect of texture, grain orientation, and grain-boundary misorientation angle on electromigration and intermetallic compound (IMC) formation is studied through EBSD analysis performed on actual C4 bumps. / Dissertation/Thesis / M.S. English 2013

Abstract Low-temperature co-fired ceramic (LTCC) based system-in-package (SiP) is an emerging multilayer module technology for wireless communication applications, mainly due to its excellent high-frequency material properties. LTCC-SiP modules are typically soldered onto an organic motherboard, but the lifetime of the 2nd-level solder joints is often poor due to the high stress level of the joints in test/field conditions. Moreover, using lead-free solders in the interconnections of LTCC modules raised new questions about the feasibility and reliability of the solder joints in LTCC applications. Therefore, the characteristic features of the 2nd-level solder joint configuration were determined in this thesis work. It was proved that collapsible Sn4Ag0.5Cu spheres are not a feasible option in LTCC/PWB assemblies with a large global thermal mismatch; a non-collapsible ball grid array (BGA) joint with a plastic core solder balls (PCSBs) was required to attain an adequate lifetime for such assemblies. To enhance the thermal fatigue endurance of the non-collapsible lead-free joints, a novel BGA joint consisting of Sn7In4.1Ag0.5Cu solder and PCSBs was developed. Moreover, this work proved that there is a relationship between the primary failure mechanisms of various Sn-based lead-free solders and thermomechanically induced stress level in the present non-collapsible BGA joint configuration. The effect of the plating material of the solder lands on the failure mechanism of the BGA joints in the LTCC/PWB assemblies was studied. The results showed that the adverse phenomena related to the sintered Ag-based metallization materials can be avoided using electroless nickel with immersion gold (ENIG) as a deposit material. On the other hand, this study also demonstrated that the inadequate adhesion strength of the commercial base metallization in the ENIG-plated modules resulted in the disadvantageous failure mechanism of the test assemblies. Therefore, the criteria for material selection and the design aspects of reliable 2nd-level interconnections are discussed thoroughly in this thesis.

With the rapid development in electronic packaging due to product miniaturisation, the size of solder joints is decreasing considerably, thus the failure of solder interconnects induced by electro-migration (EM) and thermo-migration (TM) became a reliability concern. The incorporation of foreign reinforcement can effectively improve properties of the solder alloys. However, this presents an imperative need for a further investigation to elaborate the underlying fundamentals associated with the reliability of reinforced solders. In this study, the Sn-Ag-Cu (SAC) based solder alloy powders as matrix were incorporated with Fullerene (FNS), TiC and Ni-coated graphene (NG) reinforcements to form composite solders through powder metallurgical method. These composite solders were then characterised in terms of their microstructure, physical property, solderability, followed by a systematic investigation of their performance under isothermal ageing, current stressing and large thermal gradient, respectively. The results showed that three types of reinforcements were successfully incorporated into the solder matrix; with all reinforcements added being embedded in the solder matrix or around the intermetallic compounds (IMC). The average loss of FNS and TiC particles in the solders was approximately 80% after the initial reflow, while this was only 40% for NG particles. It has been observed that β-Sn and Ag3Sn in the SAC solder alloys can be refined by adding appropriate amount of FNS and TiC, which is beneficial to the wettability with a reduced coefficient of thermal expansion (CTE) with the minimal influence on the melting point and electrical resistivity of solder alloys. For the SAC alloys without reinforcements, obvious extrusion of interfacial IMC at the anode was present after 360 hours of current (1.5×104 A/cm2) stressing, while the changes of surface profiles of all reinforced solders were unnoticeable. Under the current stressing regimes, a continuous increase of interfacial IMCs at the anode of the original SAC alloys was observed, but decreased at the cathode with stressing time. For the composite solders, both anode and cathode showed a continuous growth of interfacial IMCs; the growth rates of IMCs at the anode were greater than that at cathode. In addition, NG and TiC were found to be most effective to retard the growth of Cu3Sn IMC under current stressing. A gradient in hardness across the stressed SAC joints was present, where it was harder at anode. However, no such obvious gradient was found in SAC/FNS and SAC/NG solder joints. FNS and NG were proven to be beneficial to prolong the service life of solder joints up to approximately 7.6% and 10.4% improvements, respectively. Thermal stressing made the interfacial IMC in the original SAC joints to grow at the cold end considerably; causing serious damage at the hot end after 600 hours under temperature gradient of 1240K/cm stressing; a large number of IMCs, cracks and voids appeared in the SAC solder joints. However, a uniform increase of IMCs at both sides in the composite solders was observed without apparent damages at the interfaces under the same thermal stressing conditions, indicating an effective reduction of the elemental migration in the reinforced solders. Although there were also some voids and IMCs formed in the composite solder joints after a long-term thermal stressing, the integrity of the composite solder joints was enhanced compared with the SAC alloys. During thermal stressing, the dissolution rate of Cu atom into the SAC solder joints was estimated to be 3.1×10-6 g/h, while the values for SAC/FNS, SAC/NG and SAC/TiC were only 1.22×10-6 g/h, 1.09×10-6 g/h and 1.67×10-6 g/h, respectively.