Effect of dwell times on thermomechanical fatigue behavior of Sn-Ag-based solder joints

Thermomechanical fatigue (TMF) caused by the mismatch in the coefficient of thermal expansion (CTE) between solder and substrate gradually degrades the mechanical properties of solder joints during service. Solder joints fabricated with eutectic Sn-Ag and Sn-Ag solder with Cu or Ni were subjected to TMF between −15°C and +150°C with dwell times of 115 min at high-temperature extreme and 20 min at low-temperature extreme. Characterization of surface damage and residual-mechanical strength of these solder joints were carried out after 0, 250, 500, and 1,000 TMF cycles. Results obtained from this study were compared with those obtained with longer dwell time at lower temperature extreme. The solder joints that experienced longer dwell times at high-temperature extreme exhibited less surface-damage accumulation and less decrease in simple-shear strength as compared to those that experienced longer dwell times at low-temperature extreme. Quaternary alloys containing small amounts of Cu and Ni exhibit better TMF performance than binary and ternary alloys under TMF cycling with longer dwell times at high-temperature extreme.     

Temperature profile effects in accelerated thermal cycling of SnPb and Pb-free solder joints

Accelerated thermal cycling (ATC) has been widely used in the microelectronics industry for reliability assessment. ATC testing decreases life cycle test time by one or more of the following means: increasing the heating and cooling rate, decreasing the hold time, or increasing the range of the applied temperature. The relative effect of each of these cycle parameters and the failure mechanisms they induce has been the subject of many studies; however uncertainty remains, particularly regarding the role of the heating and cooling rate. In this research, three conditions with two ramp rates (14 °C/min and 95 °C/min) and two temperature ranges (ΔT = 0–100 °C and −40 to 125 °C) were applied to resistor 2512 and PBGA 256 test vehicles assembled with SnPb and Pb-free solders. The test results showed that the higher ramp rate reduced the testing time while retaining the same failure modes, and that the damage per cycle increased with the temperature difference. For the resistors, the Pb-free solder joints lasted longer than the SnPb joints at the smaller ΔT, but were inferior at the larger ΔT. In contrast, the Pb-free solder joints in the PBGA test vehicles lasted longer than the SnPb solder under both conditions.     

Study of micro-BGA solder joint reliability

A new reflow parameter, heating factor (Q_η), which is defined as the integral of the measured temperature over the dwell time above liquidus, has been proposed in this report. It can suitably represent the combined effect of both temperature and time in usual reflow process. Relationship between reliability of the micro-ball grid array (micro-BGA) package and heating factor has been discussed . The fatigue failure of micro-BGA solder joints reflowed with different heating factor in nitrogen ambient has been investigated using the bending cycle test. The fatigue lifetime of the micro-BGA assemblies firstly increases and then decreases with increasing heating factor. The greatest lifetime happens at Q_η near 500 s °C. The optimal Q_η range is between 300 and 750 s °C. In this range, the lifetime of the micro-BGA assemblies is greater than 4500 cycles. SEM micrographs reveal that cracks always initiate at the point of the acute angle where the solder joint joins the PCB pad.     

Residual-mechanical behavior of thermomechanically fatigued Sn-Ag based solder joints

Thermomechanical fatigue (TMF) due to the mismatch in coefficients of thermal expansion between solder and substrate gradually degrades the mechanical properties of electronic solder joints during service. This study investigated the role of TMF on the residual-mechanical behavior of solder joints made with eutectic Sn-Ag solder and Sn-Ag solder with Cu or Ni additions. The TMF tests were carried out between −15°C and +150°C with a ramp rate of 25°C/min for the heating segment and 7°C/min for the cooling segment. The hold times were 20 min at the high extreme and 300 min at the low extreme. Residual shear strength was found to drop significantly during the first 250 TMF cycles, although it did remain relatively constant between 250 and 1000 cycles. Alloying elements were found to affect the residual creep strength of solder joints after TMF.     

Thermally induced deformation of solder joints in real packages: Measurement and analysis

The paper presents a hybrid experimental and analytical approach to track the deformation of solder joints in an electronic package subject to a thermal process. The solder joint strain is directly measured using a computer vision technique. The strain measurement is analyzed following an approach that is devised based on an established solder constitutive relation. The analysis leads to the determination of the solder joint stress and in turn, to the separation of the elastic, plastic and creep strain from the measured total strain. The creep strain rate and stress–strain hysteresis loop are also obtained. Two case studies are presented to illustrate the applications and to show the viability of the approach. Each case involves a resistor package with SAC (Sn95.5Ag3.8Cu0.7) solder joints, which is subjected to a temperature variation between ambient and 120 °C. The results confirm that shear is a dominant strain component in such solder joints. The shear strain varies nearly in phase with the temperature whereas the shear stress exhibits a different trend of variation due to stress relaxation. The peak shear stress of around 10 MPa to 15 MPa are found, which occur at near 70 °C in both cases, when the temperature ramps up at approximately 3 °C/min. The creep shear strain goes up to 0.02 and accounts for over 80% of the total shear strain. The creep strain rate is in the order of magnitude of 10⁻⁵ s⁻¹. Responding to the temperature cycling with such moderate rate, the creep strain shows modest ratcheting while the stress–strain hysteresis stabilizes in two cycles.     

Effect of primary creep and plasticity in the modeling of thermal fatigue of SnPb and SnAgCu solder joints

It has been conventional to simplify the thermo-mechanical modeling of solder joints by omitting the primary (transient) contributions to total creep deformation, assuming that secondary (steady-state) creep strain is dominant and primary creep is negligible. The error associated with this assumption has been difficult to assess because it depends on the properties of the solder joint and the temperature–time profile. This paper examines the relative contributions of plasticity, primary and secondary creep in Sn40Pb and Sn3.8Ag0.7Cu solders using the analysis of a trilayer solder joint structure with finite elements and a newly developed finite difference technique. The influences of temperature amplitude and ramp rate have been quantified. It was found that for the thermal profiles considered, the role of plasticity was negligible for trilayer assemblies with SnPb and SnAgCu solder interlayers. Furthermore, when primary creep was included for SnAgCu, the temperature-dependent yield strength was not exceeded and no plastic strains resulted. Neglect of primary creep can result in errors in the predicted stress and strain of the solder joint. Damage metrics based on the stabilized stress vs. strain hysteresis loop, for symmetric 5 min upper/lower dwell periods, differ widely when primary creep is considered compared to the secondary-only creep model. Creep strain energy density differences between the secondary-only and primary plus secondary creep models for SnPb were 32% (95 °C/min–Δ165 °C thermal profile), 32% (95 °C/min–Δ100 °C) and 35% (14 °C/min–Δ100 °C); similarly for SnAgCu, the differences were 29% (95 °C/min–Δ165 °C), 46% (95 °C/min–Δ100 °C) and 58% (14 °C/min–Δ100 °C). Accumulated creep strain differences between the secondary-only and primary plus secondary creep models for SnPb were 21% (95 °C/min–Δ165 °C), 25% (95 °C/min–Δ100 °C) and 25% (14 °C/min–Δ100 °C); similarly for SnAgCu the differences were 82% (14 °C/min–Δ100 °C), 89% (95 °C/min–Δ100 °C) and 100% (95 °C/min–Δ165 °C). In turn, these discrepancies can lead to errors in the estimation of the solder thermal fatigue life due to the changing proportion of primary creep strain to total inelastic strain under different thermal profiles, particularly for SnAgCu.     

Effects of process conditions on reliability, microstructure evolution and failure modes of SnAgCu solder joints

In this study, microstructure evolution at intermetallic interfaces in SnAgCu solder joints of an area array component was investigated at various stages of a thermal cycling test. Failure modes of solder joints were analyzed to determine the effects of process conditions on crack propagation. Lead-free printed-circuit-board (PCB) assemblies were carried out using different foot print designs on PCBs, solder paste deposition volume and reflow profiles. Lead-free SnAgCu plastic-ball-grid-array (PBGA) components were assembled onto PCBs using SnAgCu solder paste. The assembled boards were subjected to the thermal cycling test (−40 °C/+125 °C), and crack initiation and crack propagation during the test were studied. Microstructure analysis and measurements of interface intermetallic growth were conducted using samples after 0, 1000, 2000 and 3000 thermal cycles. Failures were not found before 5700 thermal cycles and the characteristic lives of all solder joints produced using different process and design parameters were more than 7200 thermal cycles, indicating robust solder joints produced with a wide process window. In addition, the intermetallic interfaces were found to have Sn–Ni–Cu. The solder joints consisted of two Ag–Sn compounds exhibiting unique structures of Sn-rich and Ag-rich compounds. A crystalline star-shaped structure of Sn–Ni–Cu–P was also observed in a solder joint. The intermetallic thicknesses were less than 3 μm. The intermetallics growth was about 10% after 3000 thermal cycles. However, these compounds did not affect the reliability of the solder joints. Furthermore, findings in this study were compared with those in previous studies, and the comparison proved the validity of this study.     

Properties of low temperature Sn–Ag–Bi–In solder systems

The metallurgical and mechanical properties of Sn–3.5 wt%Ag–0.5 wt%Bi–xwt%In (x = 0–16) alloys and of their joints during 85 °C/85% relative humidity (RH) exposure and heat cycle test (−40–125 °C) were evaluated by microstructure observation, high temperature X-ray diffraction analysis, shear and peeling tests. The exposure of Sn–Ag–Bi–In joints to 85 °C/85%RH for up to 1000 h promotes In–O formation along the free surfaces of the solder fillets. The 85°C/85%RH exposure, however, does not influence the joint strength for 1000 h. Comparing with Sn–Zn–Bi solders, Sn–Ag–Bi–In solders are much stable against moisture, i.e. even at 85 °C/85%RH. Sn–Ag–Bi–In alloys with middle In content show severe deformation under a heat cycles between −40 °C and 125 °C after 2500 cycles, due to the phase transformation from β-Sn to β-Sn + γ-InSn₄ or γ-InSn₄ at 125 °C. Even though such deformation, high joint strength can be maintained for 1000 heat cycles.     

Rapid power cycling of flip-chip and CSP components on ceramic substrates

Power cycling has been done for flip-chip and CSP components solder joined onto ceramic substrates. Cycle periods as short as 1 min were applied in the experiments where the chip temperature varied between about 30°C in the power off-state and 100–150°C in the power on-state. Disconnections of the joints were found after 4000–17 000 power cycles. The flip-chip components joined onto low temperature cofired ceramic substrate showed slightly better reliability than the components joined onto alumina substrate. Most of the samples showed clear effects of deterioration of the joints seen as increasing chip temperature for power on-state. The experimental results are compared with calculations based on modified Coffin–Manson equation as well as with one-dimensional simulations.     

Board level reliability assessment of chip scale packages

A large program had been initiated to study the board level reliability of various types of chip scale package (CSP). The results on six different packages are reported here, which cover flex interposer CSP, rigid interposer CSP, wafer level assembly CSP, and lead frame CSP. The packages were assembled on FR4 PCBs of two different thicknesses. Temperature cycling tests from −40°C to +125°C with 15 min dwell time at the extremes were conducted to failure for all the package types. The failure criteria were established based on the pattern of electrical resistance change. The cycles to failure were analyzed using Weibull distribution function for each type of package. Selected packages were tested in the temperature/humidity chamber under 85°C/85%RH for 1000 h. Some assembled packages were tested in vibration condition as well. In all these tests, the electrical resistance of each package under testing was monitored continuously. Test samples were also cross-sectioned and analyzed under a Scanning Electronic Microscope (SEM). Different failure mechanisms were identified for various packages. It was noted that some packages failed at the solder joints while others failed inside the package, which was packaging design and process related.     

Thermal fatigue endurance of collapsible 95.5Sn4Ag0.5Cu spheres in LTCC/PWB assemblies

The behavior of thermomechanically loaded collapsible 95.5Sn4Ag0.5Cu spheres in LTCC/PWB assemblies with high (LTCC/FR-4; ΔCTE 10 ppm/°C) and low (LTCC/Arlon; ΔCTE < 10 ppm/°C) global thermal mismatches was studied by exposing the assemblies into two thermal cycling tests. The characteristic lifetimes of the LTCC/FR-4 assemblies, tested over the temperature ranges of 0–100 °C and −40 to 125 °C, were 1475 and 524 cycles, respectively, whereas the corresponding values of the LTCC/Arlon assemblies were 5424 and 1575 cycles. According to the typical requirements for the industrial lifetime duration of solder joints, the former values are inadequate, whereas the latter are at an acceptable level in a few cases. Furthermore, the global thermal mismatch affected the thermal fatigue behavior of the 95.5Sn4Ag0.5Cu spheres in the temperature range of −40 to 125 °C.     

Modeling thermomechanical fatigue behavior of Sn-Ag solder joints

Stresses that develop because of the coefficient of thermal expansion (CTE) mismatch between solder and substrate/components contribute to thermomechanical fatigue (TMF) of the solder joints. However, the relative importance of several processes that contribute to damage accumulation and its role in affecting the reliability of the solder joints are far from being understood. Aging, creep/stress relaxation, and stress/strain reversals are some of the important processes. These processes are affected by service conditions, such as the temperature extremes experienced, rates of heating and cooling, dwell times at the extreme temperatures, and so on. In addition, the elastic and plastic anisotropy of tin could also contribute to the damage accumulation during TMF of Sn-based solders. This preliminary effort to model TMF in Sn-Ag solder joints will consider the role of each of these parameters, with significant emphasis on the anisotropic-elastic behavior of Sn grains.     

Pb-free high temperature solders for power device packaging

Reliabilities of joints for power semiconductor devices using a Bi-based high temperature solder has been studied. The Bi-based solder whose melting point is 270 °C were prepared by mixing of the CuAlMn particles and molten Bi to overcome the brittleness of Bi. Then, joined samples using the solder were fabricated and thermal cycling tests were examined. After almost 2000 test cycles of −40/200 °C test, neither intermetallic compounds nor cracks were observed for CTE (Coefficient of Thermal Expansion) matched sample with Cu interface. On the other hand, certain amount of intermetallic compound such as Bi₃Ni was found for a sample with Ni interface. In addition, higher reliability of this solder than Sn-Cu solder was obtained after −40/250 °C test. Furthermore, an example power module structure using double high temperature solder layers was proposed.     

Fatigue analysis of miniaturized lead-free solder contacts based on a novel test concept

The paper presents the method of generating lifetime-prediction-laws on special prepared very stiff specimen. The combination of thin- and thick-film technology allows building up test samples on ceramic very similar to electronic packages including the measurement issues. Influences of pad surface metallurgy, microstructure of solder, ineutectic solder alloys and assembly process parameter are regarded now. The investigation objects provide monitoring of electrical and mechanical damage process of SnAgCu solder bump. Different thermo-mechanical loads will be applied in temperature ranges of 0 to +80 °C, −40 to +125 °C and −50 to +150 °C, where the temperature gradient and cycle frequency also vary. A Variation of four different chip sizes allows the determination of fatigue laws for each temperature profile, to be able to compare in between them. The results of these tests will give universal lifetime-prediction laws for SnAgCu base solder joints. Main goals are to find coefficients for lifetime prediction models such as Coffin–Manson- or Norris–Landzberg-relation, which are transferable in between different electronic packages.     

Creep of thermally aged SnAgCu-solder joints

The creep behaviour of Sn96.5Ag3.5- and Sn95.5Ag3.8Cu0.7-solder was studied specifically for its dependence on technological and environmental factors. The technological factors considered were typical cooling rates and pad metallizations for solder joints in electronic packaging. The environmental factors included microstructural changes as a result of thermal aging of solder joints. Creep experiments were conducted on three types of specimens—flip–chip joints, PCB solder joints and bulk specimens. flip–chip specimens were altered through the selection of various under bump metallizations (Cu vs. NiAu), cooling rates (40 K/min vs. 120 K/min), and thermal storage (24 h, 168 h, and 1176 h at 125 °C). PCB solder joints were studied by using a copper pin soldered into a thru-hole connection on a printed circuit board having a NiAu metallization. Bulk specimens contained the pure alloys. The creep behaviour of the SnAg and SnAgCu solders varied in dependence of specimen type, pad metallization and aging condition. Constitutive models for SnAg and SnAgCu solders as they depend on the reviewed factors are provided.     

Accelerated thermal fatigue of lead-free solder joints as a function of reflow cooling rate

Leadless chip resistor (LCR) assemblies were manufactured using both traditional tin-lead (Sn37Pb) and lead-free (Sn3.8Ag0.7Cu) solders. The leadfree test vehicles were assembled using three different cooling rates: 1.6°C/sec, 3.8°C/sec, and 6.8°C/sec. They were then exposed to accelerated thermalcycling (ATC) tests between 0°C and 100°C with a 10–14°C/min ramp rate and a 5-min dwell time. The test results indicated that these lead-free solder joints had better creep-fatigue performance than the tin-lead solder joints. The LCR built with the medium cooling rate showed the longest fatigue life compared with the resistors built with the normal cooling rate of 1.6°C/sec and the higher cooling rate 6.8°C/sec. The number of cycles to failure was significantly correlated to the void defect rate. Failure analyses were done using cross-sectioning methods and scanning electron microscopy (SEM). Finite-element models were built to analyze the inelastic, equivalent strain range in solder joints subjected to thermal-cycling conditions with different degrees of solder wetting. The results indicated that poor wetting increases strains throughout the joint significantly, which is in accordance with the ATC results.     

Effect of bump size on current density and temperature distributions in flip-chip solder joints

Three dimensional thermo-electrical analysis was employed to simulate the current density and temperature distributions for eutectic SnAg solder bumps with shrinkage bump sizes. It was found that the current crowding effects in the solder were reduced significantly for smaller solder joints. Hot-spot temperatures and thermal gradient were increased upon reducing the solder. The maximum temperature for solder joint with 144.7 μm bump height is 103.15 °C which is only 3.15 °C higher than the substrate temperature due to Joule heating effect. However, upon reducing the bump height to 28.9 μm, the maximum temperature in the solder increased to 181.26 °C. Serious Joule heating effect was found when the solder joints shrink. The higher Joule heating effect in smaller solder joints may be attributed to two reasons, first the increase in resistance of the Al trace, which is the main heating source. Second, the average and local current densities increased in smaller bumps, causing higher temperature increase in the smaller solder bumps.     

Impact of board and component metallizations on microstructure and reliability of lead-free solder joints

Aging and accelerated thermal cycling (ATC) have been performed on 2512 chip resistors assembled with Sn3.8Ag0.7Cu (wt.%) solder. The boards were finished with immersion Ag (IAg), electroless nickel/immersion gold (ENIG), and hot air solder leveling Sn–Pb eutectic solder (HASL), and the components’ terminations were finished with 100% Sn and Sn8.0Pb (wt.%). The boards were reflowed with an average cooling rate of 1.6 °C/s. It was found that the microstructure and reliability of the solder joints depended on the board surface finish. The boards containing small amounts of Pb (from board/component terminations) were the most reliable. Solder joints to copper showed a significantly higher number of cycles to first failure than the joints on nickel. Better reliability of the Sn3.8Ag0.7Cu/Cu joints was attributed to an increased copper content in the bulk due to substrate dissolution.     

A quantitative method of reliability estimation for surface mount solder joints based on heating factor Qη

A novel method of reliability analysis on thermal fatigue failure for surface mount solder joints, based on the heating factor Q_η, is presented, by which quantitative reliability estimation and prediction of solder joints suffering from cyclic thermal stress can be done. Based on the typical lifetime data of thermal cycling test, the relationship of the mean time to failure (MTTF) as well as the reliability of solder joints as an explicit function of Q_η is deduced and presented in a unified mathematic form. Numerical calculations are performed, and the result shows that the MTTF decreases quickly with the increases in heating factor and then slowly approximates to a constant value when Q_η  1500 s °C. The solder joint reliability in terms of thermal cycle degrades in an analogical fashion for different heating factors. For any given thermal cycle, calculation suggests that to obtain a higher reliability, a lower heating factor should be controlled during soldering. The presented method gives an applicable solution and can be used for online reflow control in industry. On the one hand, an ideal reflow profile can be achieved by properly controlling heating factor during soldering to meet the given reliability goal. On the other hand, the life expectancy of solder joints can be approximately estimated and predicted from a known reflow profile with a specified heating factor. Finally, for a specified reliability goal, how to properly choose and control heating factor during soldering to achieve reliable solder joints is discussed.     

Crack Development in a Low-Stress PBGA Package due to Continuous Recrystallization Leading to Formation of Orientations with [001] Parallel to the Interface

Thermal cycling was imposed on plastic ball grid array (PBGA) packages with a small die, a package design that does not impose a large strain on solder joints. Less cracking was observed after 2500 cycles from 0°C to 100°C (with 10 min dwell times and 10 min ramps) than in a prior study with a higher-stress package design, so these samples were thermally cycled (TC) to 6400 cycles to investigate the relationship between cracks, microstructure, and grain crystal orientation. Cracked joint locations within the package were identified using the dye and pry method, indicating that cracks were most often found in joints near the perimeter of the die. Using orientation imaging microscopy (OIM), cracks were observed in many joints having a variety of dominant crystal orientations where the c-axis was between 0° and ∼50° from the package interface. Continuous recrystallization processes occurred and caused gradual rotations of initial orientations that reduced the angle between the c-axis and the package interface. While cracks were observed in joints with a variety of orientations, cracks were highly correlated with recrystallized grains having the [001] c-axis nearly parallel to the interface (“red” orientations) in those joints that did not initially have this orientation.