Search

Transient liquid phase (TLP) bonding of Cu substrates was conducted with interlayer systems with the stacking sequences Ag–Sn–Ag (samples A), Ni–Sn–Ni (samples B), and combined Ag–Sn–Ni (samples C). Because of the low mismatch of the coefficients of thermal expansion, characteristics of the TLP process and mechanical and thermal behavior of TLP-bonded samples could be investigated without interference from thermally induced residual stresses. An ideal process temperature of 300°C, at which the number of pores was lowest, was identified for all three layer systems. It was verified experimentally that formation of pores resulted from volume contraction during isothermal solidification of liquid Sn into intermetallic compounds (IMC). Temperature and interlayer-dependent growth characteristics of IMC accounted for the increasing size and number of defects with increasing process temperature and for different defect positions. The shear strength was measured to be 60.4 MPa, 27.4 MPa, and 40.7 MPa for samples A, B, and C, respectively, and ductile fracture features were observed for Ag3Sn IMC compared with the purely brittle behavior of Ni3Sn4 IMC. Excellent thermal stability for all three layer systems was confirmed during long-term annealing at 200°C for up to 1200 h, whereas at 300°C the microstructure was driven toward Ag–Sn solid solution, accompanied by Cu diffusion from the substrate along grain boundaries and Cu3Sn IMC formation (A), and toward Ni-rich IMC phases (B). Combined IMC interlayers (C) tended to be transformed into Ni-based IMC when held at 300°C; intermixing into (Ni,Cu)3Sn was accompanied by pore formation.

Transient liquid-phase sintering (TLPS) using a Cu-solder-resin composite for the die-attach application of high-temperature silicon carbide (SiC) power modules was evaluated with the goal of controlling the joint stiffness. The Cu-solder-resin composite mainly contains Cu particles, Sn–3Ag–0.5Cu solder particles, and polyimide-type thermosetting resin. Microstructural observations, shear strength tests, and thermal cycling tests of the SiC die-attached specimens bonded through the pressureless reflow process at 250 °C for 1 min were carried out after preheating at 100 °C for 60 min in a nitrogen atmosphere. A skeleton-shaped microstructure, consisting of Cu and Cu–Sn intermetallic compounds (IMCs) partially filled with polyimide resin, was observed in the joints, which exhibited shear strengths exceeding 12 MPa. The thermal cycling tests revealed that the TLPS specimens showed a superior reliability through 1200 thermal cycles in the temperature range from −55 °C to 175 °C. These experimental results were supported by mechanical finite-element (FE) simulations of simplified models that consist of the skeleton-shaped microstructure between the SiC chip and the substrate against an external force to mimic the shear test condition. The FE analyses proved that the stiffness of the composite joints strongly depends on the interconnection density that results from the formation of IMCs between the particles.

Thin metallic Cu and Ni meshes were successfully embedded within SAC305 solder joints. Defects (residues) within the joint were removed by a vacuum step in the bonding sequence. The metallic mesh inlets were metallurgically bonded by the formation of scallop-like Cu6Sn5 and faceted Ni3Sn4 intermetallic compounds (IMCs) at the Cu mesh-SAC305 solder (CMS) and Ni mesh-SAC305 solder (NMS) interfaces, respectively. After process optimization, the averaged shear strengths were found to be 44.1 (reference), 44.7 (+1%), and 51.4 MPa (+16%) for SAC305, NMS, and CMS composite joints, respectively. The solder-substrate-mesh interaction resulted in a characteristic fracture pattern, with the cracking path partly within the solder and partly at the solder-mesh interface. Finite element (FE) simulations suggested a mesh-induced stiffening effect of the solder joint by 5%. The additional reinforcement (11%) introduced by the Cu mesh inlets was attributed to a locally enhanced fracture resistance. Keywords: Solder-mesh composite joints, Microstructure, Shear strength, Finite element modeling, Fracture

High-frequency linear friction welding (HFLFW) with a 250 Hz oscillation frequency was applied on similar joining of AA5052 and AA6063, i.e., work- and precipitation-strengthened Al alloys, with 11 × 11 mm2 friction surface. Sound joints were achieved under a 30 MPa friction pressure and with processing times longer than 1 s. The thermomechanically affected zone (TMAZ) of AA6063 joints suffered from softening at a distance of 2–3 mm from the joined interface in contrast to the TMAZ of AA5052 joints. The pressure reduction to 7 MPa narrowed down the thickness of the softened TMAZ to 1 mm from the interface. A thermal finite element (FE) model was calibrated against in situ temperature measurements. Significant temperature differences of up to 300 °C at the welding interface were assigned to the process parameters as well as to the different thermal and mechanical properties of the Al alloys. Linear hardening and softening effects in AA5052 and AA6063, respectively, were quantified in hardness vs. temperature diagrams, and a pressure-hardening phenomenon was identified. A user subroutine was fed with these correlations, which visualized the hardness distribution over the near-interface volume for a low- and a high-pressure process.

This study introduces a novel concept that uses composite structures of thin metallic meshes and Sn3.0Ag0.5Cu (SAC305) solder alloys to interconnect semiconductor chips to DBC substrates. The feasibility was proven by bonding Cu-to-Cu substrates. The averaged shear strength was measured as 44.1, 44.7 and 51.4 MPa for samples bonded with SAC305 (reference), Ni mesh/SAC305 and Cu mesh/SAC305 composites, respectively. The solder-mesh joints revealed a specific fracture behavior where crack propagation occurred partly within the solder and partly at the solder-mesh interface. Microstructural analyses confirmed that the metallic meshes were bonded by the formation of intermetallic compounds (IMC) while almost no (larger) defects were found. The solder-mesh concept was subsequently applied on Si-to-DBC assemblies. A very good resistance of Cu mesh/SAC305 composite joints against cyclic temperature between 80 and 200 °C was observed when the bonded area only reduced by 4.2 % after 8000 cycles. Thermal finite element (FE) simulations indicated that in particular Cu mesh/SAC305 composites can significantly reduce the thermal resistance of the interconnections which is equivalent to a better heat dissipation through the bonding layer. Thus, solder-mesh composite joints seem to be an attractive solution for high-temperature applications up to 200 °C.

Metal-to-ceramics direct sintering was carried out with Al2O3 and Cu / 3 μm Ni / 1 μm Au substrates. The bonding paste consisted of micron-sized Ag2O particles and a reducing solvent that provokes Ag2O-to-Ag reduction during processing accompanied by the formation of Ag nano particles. Five different sets of process parameters resulted in different joint microstructure and strength. The experimental data was used to calibrate an elasto-visco-plastic finite element model of the sintered assembly which yielded a quantitative damage function and criterion to predict the strength of direct-sintered joints. The developed ductile damage formulation introduced a parameter ξ, i.e. the product of equivalent creep strain and stress triaxiality, that controls the tolerable plastic strain at fracture. An extended numerical parameter study subsequently revealed the complex interaction between the joint strength and microstructural joint features. Thinner joints were found to provide a slightly higher strength while the amount of sinter material and costs is significantly reduced. Moreover, it is recommended to aim at a higher level of densification at the edges and corners of sintered joints since these areas apparently contribute more to the overall mechanical strength. The developed concept is capable of tailoring the microstructure of direct-sintered joints according to the requirements or vice versa.

Al2O3-to-Cu/3 μm Ni/1 μm Au assemblies were direct-sintered with a paste consisting of micron-sized Ag2O particles and a reducing solvent that provokes Ag2O-to-Ag reduction during processing accompanied by the formation of Ag nanoparticles. The variation of process temperature, pressure and time resulted in different porosity and shear strength. This experimental data set was used to calibrate a finite element model of the assembly considering the elasto-visco-plastic joint properties and the joint porosity. Local analyses within the sintered Ag layer yielded a damage function D f and a damage criterion that could assess the joint strength as a function of geometry, microstructure and process conditions. The developed damage model introduced a parameter ζ, i.e. the product of equivalent creep strain e cr,eq and stress triaxiality η, that balanced the tolerable equivalent plastic strain e pl,eq at fracture. A Fortran subroutine (UVARM) visualized the damage distribution over the sintered layer. The predominant crack initiation and propagation were identified in good agreement with the experimental fracture behavior. A numerical parameter study subsequently revealed the complex interaction between the ceramic–metal interfacial porosity, the total bulk porosity and the shear strength.

Solder joints between leads and printed circuit boards in thin small outline packages were produced with conventional Sn1.0Ag0.7Cu (SAC107) and Sn3.0Ag0.7Cu (SAC305) solders as well as various solder alloys with gradually increasing amounts of Bi (up to 3.0 wt.%) and In (up to 1.0 wt.%) within the SAC107 base solder. The reliability of soldered leads in temperature cycle (TC) tests improved most with solder alloys containing both Bi (1.6 wt.%) and In (0.5 wt.%). Microindentation and electron probe microanalysis mappings revealed that the effect originates from a combination of solution and precipitation strengthening of the initial SAC alloy. The distribution of inelastic strain accumulation (ISA), as a measure for degradation, was determined in the solder joints by finite element calculations. It was shown that defects in the solder proximal to the lead ( 60–75 μm), which was underpinned by similar cracking characteristics along the lead–solder interface. The ISA was confirmed to be lower in SAC+Bi/In alloys owing to their enhanced elasto-plastic properties. Moreover, the addition of a thin Cu coating on the leads could improve the joint reliability, as suggested by the calculation of the ISA and the acceleration factor.

The evolution of the transient liquid-phase sintered (TLPS) Cu&ndash, Sn skeleton microstructure during thermal aging was evaluated to clarify the thermal reliability for die-attach applications. The Cu&ndash, Sn skeleton microstructure, which consists of Cu particles connected with Cu&ndash, Sn intermetallic compounds partially filled with polyimide resin, was obtained by the pressure-less TLP sintering process at 250 &deg, C for 1 min using a novel Cu-solder-resin composite as a bonding material in a nitrogen atmosphere. Experimental results indicate that the TLPS joints were mainly composed of Cu, Cu6Sn5, and Cu3Sn in the as-bonded state, where submicron voids were observed at the interface between Cu3Sn and Cu particles. After thermal aging at 150, 175, and 200 &deg, C for 1000 h, the Cu6Sn5 phase fully transformed into Cu3Sn except at the chip-side interface, where the number of the submicron voids appeared to increase. The averaged shear strengths were found to be 22.1 (reference), 22.8 (+3%), 24.0 (+9%), and 19.0 MPa (&minus, 14%) for the as-bonded state and specimens aged at 150, 175, and 200 &deg, C for 1000 h, respectively. The TLPS joints maintained a shear strength over 19 MPa after thermal aging at 200 &deg, C for 1000 h because of both the positive and negative impacts of the thermal aging, which include the transformation of Cu6Sn5 into Cu3Sn and the formation of submicron voids at the interface, respectively. These results indicate an excellent thermal reliability of the TLPS Cu&ndash, Sn skeleton microstructure.

The potential of transient liquid-phase (TLP) bonding for chip packaging applications has been evaluated, focusing on three interlayer arrangements (Ag-Sn-Ag, Ni-Sn-Ni, and Ag-Sn-Ni). Shear tests on TLP-bonded components provided the interlayer-dependent mechanical strength as well as failure mode and position. Critical local stresses, i.e., failure criteria, within the intermetallic compound (IMC) layer were derived by replicating the shear test conditions with finite-element methods. The missing coefficient of thermal expansion for Ag3Sn IMC was obtained by producing small IMC bulk samples and subjecting them to dilatometric measurements. The experimental results were implemented into a finite-element model of a representative power module architecture to provide first predictions on thermally induced residual stresses that could be classified into fail/safe, as successfully validated by TLP chip bonding experiments. A numerical parameter study then assessed thermal stresses, including failure prediction and design optimization for TLP-bonded Si chips, considering the influence of process temperature, service conditions, TLP interlayer system, and metallization layers within the TLP joint. The presented procedure serves as a guideline to choose an appropriate TLP interlayer system for predefined boundary conditions, or vice versa.

The creep and stress relaxation behavior of a Fe–17Mn–5Si–10Cr–4Ni–1(V,C) (wt%) shape memory alloy at low homologous temperatures (−45 °C

The near-isothermal growth and formation of Ni3Sn4 intermetallic compounds (IMC) in Ni-Sn interlayer systems was studied in the solid state at 473 K (200 °C) and under solid–liquid conditions at 523 and 573 K (250 °C and 300 °C) from an initial state of a few seconds. Scalloped solid-state IMC formation was mainly driven by grain boundary diffusion of Ni through the IMC layer combined with the grain coarsening of the IMC layer. Under solid–liquid conditions, the formation of faceted and needle-shaped Ni3Sn4 grains as well as an atypical IMC growth behavior with similar parabolic growth constants for 523 K and 573 K (250 °C and 300 °C) was observed within the first 180 seconds of the holding time, and IMC growth occurred as an isothermal solidification from the Ni-saturated Sn melt. Due to the progressive densification of the IMC layer and the diffusion-controlled growth, the kinetics slowed down by approximately one order of magnitude after 180 seconds of annealing. The final stage was characterized by the formation of IMC islands ahead of the interfacial Ni3Sn4 layer. Needle-like IMC growth was effectively suppressed under combined solid-state and solid–liquid conditions. Textured Ni3Sn4 IMC formation at the Ni-Sn interface was approved with pole figure measurements. The activation energy Q for solid–liquid IMC formation was calculated as 43.3 kJ/mol, and processing maps for IMC growth and Sn consumption were derived as functions of temperature and time, respectively.

In this work, cyclic loading experiments were performed to investigate the influence of the elastic–plastic properties of the substrate material as well as of artificial brazing defects on the fatigue behavior of brazed steel joints. Brazing was performed with the steel X3CrNiMo 13-4 (AISI 316 C-NM) as a base material and with AuNi18 as a filler metal. After brazing, subsequent heat treatments were performed to influence the elastic–plastic properties of the substrate material. To investigate the influence of defects, defined defect geometries were introduced in the braze layer by electrical discharge machining (EDM). The results of the experiments show that the heat treatment has a significant influence on the fatigue lifetime. Furthermore it could be shown, that the fatigue behavior of brazed components is significantly influenced by the interaction of defects and elastic–plastic properties of the substrate material. To investigate the cyclic deformation, the materials strain response was investigated. Besides extensometer measurements, Digital Image Correlations (DIC) were used to measure the strain distribution during cyclic loading with a sufficient lateral resolution. Additional SEM investigations and FE-calculations were carried out to verify the experimental results and to investigate the complex mechanisms that can lead to fatigue failure of brazed components in detail. The results of this work underline the significant influence of the elastic–plastic properties of base material on the fatigue behavior and can be used for the appropriate design of brazed components.

Transient liquid phase sintering (TLPS) using copper (Cu)-solder-resin composite was investigated for dieattach applications of high-temperature power modules. The copper particles were connected with Cu-Sn IMCs through transient liquid phase reaction at about 220 °C. The Cu-IMC skeleton structure was partially filled with polyimide resin. The joint strengths of die-attached specimens were tested. The die-attached specimens showed superior reliability through thermal cycling tests. These results indicate the potential of TLPS using Cu-solder-resin composite for power module dieattach applications.

Transient Liquid Phase Bonding (TLPB) makes use of the isothermal solidification of a liquid interlayer through the formation of intermetallic compounds (IMC). This work investigates the Ag3Sn IMC formation in a thin, electroplated Ag–Sn interlayer system appropriate for TLPB – while covering as-coated conditions, growth initiation up to a few minutes of holding time – with experimental and numerical methods. Both solid-state and solid–liquid reactions at 200, 250 and 300 °C, respectively, were covered under identical and near-isothermal conditions with a novel experimental setup. Morphological aspects like the lateral IMC grain size or the scallop roughness were proven to strongly depend on temperature. Growth kinetics were captured by power laws, growth exponents n were determined as 0.24, 0.36 and 0.29, parabolic growth constants k as 1.32 ⋅ 10−11, 4.04 ⋅ 10−10 and 7.24 ⋅ 10−10 cm2/s at 200, 250 and 300 °C, respectively, and the activation energy Q as 30.6 kJ/mol. Predominant diffusion mechanisms, i.e. volume or GB diffusion, as well as coarsening effects were identified as temperature and time dependent at early stages. A mathematical model was developed based on the Arrhenius relationship including a growth correction function zIMC,corr to calculate and predict IMC growth as a function of temperature and time. The study was completed by multiphase-field simulations which targeted to mimic kinetic and morphological features of Ag3Sn IMC under various conditions. Good agreement between experiments and simulations was found for a limited set of parameters which allowed drawing conclusions about the dominant transport mechanisms for the reacting elements.

To investigate the fatigue assessment of brazed steel joints, stress controlled fatigue tests were conducted with specimens of the steel AISI CA 6-NM (1.4313) and with its brazed joints. Brazing was performed in a shielding gas furnace under H2 atmosphere with Au 18wt% Ni as filler metal. Experiments were performed at a load ratio of R ¼0.1 with different specimen geometries to compare their fatigue behavior and to investigate the failure mechanism. The results of the experiments—based on a lifetime oriented approach—show the existence of two different regimes depending on the number of cycles to fracture (Nf). For Nfo10 4 the maximum tolerable loads for all specimens approach the ultimate tensile strength of the substrate material, whereas for Nf410 4 the substrate material provides the highest strength, followed by the brazed round specimens and by the brazed T-joint specimens. Investigations on the failure mechanisms revealed that for brazed specimens, fatigue and residual fracture occurred always in the interface of the brazing zone. The crack path is characterized by interfacial jumps, accompanied by ductile deformation features. The analysis of the strain evolution during the cyclic loading experiments shows that the cyclic deformation behavior is significantly influenced by cyclic creep. Furthermore, the experiments show that brazed round specimen exhibit higher strains at similar loading amplitudes, compared to the substrate material. These new findings were also confirmed by FE-calculations, showing an inhomogeneous distribution of local stresses and strains in the proximity of the braze layer. The archived results show the complex interactions of a braze layer on the cyclic deformation behavio—compared to its bulk material—and lead to a better understanding of the fatigue assessment of brazed steel joints.

Based on the combination of FE-calculations and experimental results a defect assessment procedure for brazed steel joints was developed in the present work, which reliably takes into account the effect of different steel heat treatments and defined artificial defects in the braze layer. Tensile tests with defect free joints consisting of X3CrNiMo13-4 as substrate material and of Au-18wt.-%Ni as filler metal serve as a basis to simulate the elastic-plastic material behaviour. By correlating experimental and numerical methods, failure criteria could be derived which indicate and predict failure of brazed components. These failure criteria were also presumed to calculate the maximum tolerable loads for specimens containing different defects. The results show that the simulations correlate very well with the experimental results. Furthermore, they provide a possibility to analyze the failure behaviour for different defect types and the influence of different heat treatments on the joint strength.