Submitted by JOSIANE SANTOS DE OLIVEIRA (josianeso) on 2018-10-03T13:26:39Z No. of bitstreams: 1 Bernardo Bortolotto Ilha_.pdf: 35188571 bytes, checksum: cac488e47b7d5a9301e00e7ab4d1ad4b (MD5) / Made available in DSpace on 2018-10-03T13:26:39Z (GMT). No. of bitstreams: 1 Bernardo Bortolotto Ilha_.pdf: 35188571 bytes, checksum: cac488e47b7d5a9301e00e7ab4d1ad4b (MD5) Previous issue date: 2018-01-16 / UNISINOS - Universidade do Vale do Rio dos Sinos / The aging and electromigration (EM) effects were evaluated when up to 0.19 wt.% Zn was added to Sn-0.7 wt.% Cu solder. Currently, the Sn-0.7 wt.% Cu solder is being widely used in the electronic industries due to its advantages of low cost and high temperature applications. However, its usage is also limited by detrimental properties – for instance, when compared to SAC305, Sn-0.7 wt.% Cu solder has lower electromigration life time, shear strength and drop reliability. Minor Zn alloying to Pb-free solders reportedly enhances some of their properties, e.g.: stabilization of bulk microstructures by decreasing undercooling; formation of a thin interfacial diffusion barrier and, thereby, suppressing Cu3Sn and Cu6Sn5 interfacial IMC growth rate and retarding under bump metallurgy (UBM) diffusion through the solder; and also, compensation for Sn self-diffusion due to reverse polarity effect. In this research, the aging and EM effects are assessed when 0.09, 0.16 and 0.19 wt.% Zn were added to Sn-0.7 wt.% Cu solder. The samples underwent up to 500 h of isothermal aging at temperatures of 125, 150 and 175 °C, and EM samples underwent up to 200 h of stressing at a constant temperature of 150 °C and current of 3.25 A. Solder balls were fabricated on a BGA structure for the aging tests, and for the EM tests, a pair of solders was assembled in a daisy-chain structure with organic solderability preservative (OSP) and electroless nickel immersion gold (ENIG) surface finishes. The microstructural evolution and compositional distribution analyses were carried out using optical microscope with brightfield and cross polarized light, scanning electron microscope (SEM), energy dispersive spectrometer (EDS), electron probe micro analyzer (EPMA), and electron backscattered diffraction (EBSD). The addition of Zn suppresses the formation of Cu3Sn IMC and the total interfacial IMC thickness upon aging, and the samples with ENIG had smaller IMC thickness than OSP surface finish. In addition, the grains' microstructure becomes less interlaced and more stable, indicating lower undercooling. The electromigration effects on the microstructure is mainly governed by the relative orientation between the c-axis of Sn grains and the direction of current flow. When parallel, allows cathode UBM and alloying elements diffusion through the solder and formation of IMC, and, when transverse, inhibits this diffusion leading to failure by void formation at the anode UBM/solder interface due to depletion of the UBM and slow Sn self-diffusion.

Electronics, especially, printed circuit boards (PCBs) are a widespread technology. Metal coatings or “surface finishes” are often added to PCB board pads and component leads during manufacturing to improve their performance. Electroplated nickel/gold over copper is a popular surface finish for printed circuit boards and component leads. The presence of gold in solder joints, however, is known to have detrimental effects referred to as gold embrittlement. It is generally understood that tin-lead solder joints with less than 3 weight percent (wt%) of gold will not experience reliability issues. The acceptable level of gold in lead-free solder joints, however, is less well understood, as the technology is younger. The purpose of this study was to investigate the effect of gold content on the microstructural evolution of SnAgCu solder joints. Three integrated circuit packages with various thicknesses of gold coatings were assembled on boards that were made with thin (flash) or thick gold over nickel coatings. The boards were divided into three groups based on the isothermal aging they underwent: 0 days, 30 days, or 56 days of aging at 125 °C. Thirty four of the forty boards then underwent mechanical reliability testing. Components were cross-sectioned and polished. Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS/EDX) were used to characterize the morphology and elemental composition of the solder joints and any intermetallic compounds (IMCs) that formed. The growth of bulk and interfacial layer IMCs in each package/board system was studied. In thick gold boards, AuSn4 particles observed in the bulk solder grew larger over time, absorbed Ni, and migrated to the component and board interfaces. (Cu1-p-qAupNiq)6Sn5 and (Au1-xNix)Sn4 IMCs were found at most board and component interfaces after aging. It was observed that most fractures occurred in or along the (Au1-xNix)Sn4 IMCs. Cracks were observed within IMC particles in the bulk solder, along the boardside and component side interfaces, and in the bulk solder traveling toward voids. Components with joint Au contents higher than 10 wt% had unacceptably poor reliability. The conclusion of this work is that gold content of SAC305 solder joints on boards with Au over Ni surface finishes should be kept below 3 wt% to conservatively minimize the risk of creating a microstructure that has poor reliability.

The master thesis analyses electrical conductivity of lead-free solder joints. The test method design for monitoring the electrical conductivity of the soldered joint is described in the practical part. Simulated BGA package with four pin (BGA4) is used for experiments. Tested PCBs are subjected to isothermal aging and current load. During isothermal aging is measured electrical conductivity and optical microscope is used for intermetallic layer (IMC) growth observation. Two types of surface finish (OSP and ENIG) are used for tests and three diameters of solder terminal balls (solder alloy SAC405). The influence of the ratio area connection / solder volume (ratio S / V) on lead-free solder joints conductivity was evaluated.

Les aciers inoxydables lean duplex sont une famille d'aciers austéno-ferritiques allégés en nickel et en molybdène, qui s'est développée à la fin des années 1990. Le compromis propriétés mécaniques, propriétés de résistance à la corrosion et coût de matière première place cette famille comme une alternative intéressante aux aciers austénitiques standards, et en particulier aux 304/304L qui représentent actuellement les deux tiers de la production d'acier inoxydable. Cependant, cette famille étant relativement récente, la stabilité en température des aciers lean duplex a été relativement peu étudiée, en particulier lors de maintiens prolongés en température. Dans le cadre de cette thèse, l'évolution microstructurale d'un acier lean duplex 2101 a été étudiée, lors de vieillissements isothermes à des températures comprises entre 20 °C et 850 °C, pour des temps s'échelonnant de quelques minutes à plusieurs mois. Les cinétiques de vieillissement ont été suivies par mesures de pouvoir thermoéléctrique (PTE), à partir desquelles des états vieillis ont été sélectionnés pour être caractérisés par microscopie électronique et par sonde atomique tomographique. A des températures intermédiaires de 350 – 450 °C, la ferrite de l'acier lean duplex 2101 est sujette à la démixtion Fe-Cr et à la formation d'amas enrichis en Ni-Mn-Si-Al-Cu, malgré les faibles teneurs en nickel de cette nuance. Ces phénomènes sont détectés par une forte augmentation du PTE. Pour des températures plus élevées, à 700 °C environ, une approche multi-techniques et multi-échelles a permis de décrire précisément les mécanismes qui régissent les différentes évolutions microstructurales : la germination et la croissance de M23C6 et de Cr2N, observés dès quelques minutes de vieillissement aux joints de phases, la précipitation de la phase σ pour des temps de vieillissement plus importants qui s'accompagne d'une transformation de la ferrite δ en austénite secondaire γ2, et la transformation partielle de l'austénite en martensite lors du refroidissement des échantillons vieillis. L'effet des différentes phases sur le PTE de l'acier lean duplex peut être décrit qualitativement lors du vieillissement par une loi des mélanges. / Lean duplex stainless steels are austeno-ferritic steels with lower nickel and molybdenum contents, developed in the late 90's. Considering mechanical properties, corrosion resistance and cost of raw material, this family is an interesting alternative to standard austenitic stainless steels, which currently represent two thirds of stainless steel production. However, lean duplex steels are relatively recent and their thermal stability has been relatively little studied, especially during long term aging. In this study, the microstructural evolution of a lean duplex steel 2101 was studied during isothermal aging at temperatures between 20 °C and 850 °C, from few minutes to several months. Aging kinetics were followed by thermoelectric power measurements (TEP), from which aged states were selected to be characterized by electron microscopy and atom probe tomography. At intermediate temperatures of 350 - 450 °C, Fe-Cr demixing and precipitation of Ni-Mn-Al-Si-Cu occur in the ferrite despite the low nickel content of this grade, leading to an increase in the TEP. For higher temperatures, at about 700 °C, the mechanisms which govern the different microstructural evolutions have been described by a multi-scale approach: the nucleation and growth of M23C6 and Cr2N, observed from few minutes of aging and the σ phase precipitation, observed for longer aging time. The latter is accompanied by a transformation of δ ferrite in γ2 secondary austenite, and by the partial transformation of austenite into martensite during cooling. The effect of different phases on the TEP of the lean duplex steel can be qualitatively described during aging by a rule of mixture.