Creep fatigue models of solder joints: A critical review

The goal of creep fatigue modelling is the compounding of the damage caused by creep and fatigue mechanisms. The different approaches for compounding these damage mechanisms have led to several different creep fatigue models: (i) ignore fatigue damage — the creep ductility (energy density) exhaustion models; (ii) lumping plastic and creep strain (energy) into inelastic strain (energy) — the model of Dauvearx's crack initiation and propagation; (iii) linearly sum fatigue and creep damage — the model of linear damage summation; (iv) model creep and fatigue damage using a common parameter — the models of fracture mechanics; (v) partition damage into fatigue, cyclic creep, and cyclic creep-fatigue interaction — strain range/energy partitioning models; (vi) model creep and fatigue damage using a common parameter at rates that are dependent on the current state of damage — the model unified damage; (vii) model creep and fatigue damage using separate damage parameters — the mechanism based model; and (viii) integrate creep damage into the fatigue equation — creep modified strain-life equations. The rigour of the approaches increases from (i) to (vii). The creep modified strain-life equation requires no evaluation of creep strain and facilitates design analysis and evaluation of acceleration factors; however, its rigour depends on the choice of the creep functions. The unified equation is capable of covering the full spectrum of creep-fatigue from pure fatigue to pure creep rupture.     

Creep deformation of Sn-3.5Ag-xCu and Sn-3.5Ag-xBi solder joints

Creep properties of lead-free Sn-3.5Ag-based alloys with varying amounts of Cu or Bi were studied by single lap-shear test. Solder balls with five different compositions of Cu (0 wt.%, 0.75 wt.%, 1.5 wt.%) and Bi (2.5 wt.%, 7.5 wt.%) were reflowed on Cu. The Cu-containing alloy had a lower creep rate than the Bi-containing alloy. The Sn-3.5Ag alloy showed the lowest creep rate on Cu, implying that the Cu element already dissolved in the Sn-3.5Ag alloy during reflow. The Cu-containing alloy was strengthened by dispersed small precipitates of Cu₆Sn₅. As the Cu content increased up to 1.5 wt.%, the Cu₆Sn₅ coarsened and platelike Ag₃Sn intermetallics were found, which deteriorated the creep resistance.     

Creep deformation behavior in eutectic Sn-Ag solder joints using a novel mapping technique

Creep deformation behavior was measured for 60–100 μm thick solder joints. The solder joints investigated consisted of: (a) non-composite solder joints made with eutectic Sn-Ag solder, and (b) composite solder joints with eutectic Sn-Ag solder containing 20 vol. %, 5 μm diameter in-situ Cu₆Sn₅ intermetallic reinforcements. All creep testing in this study was carried out at room temperature. Qualitative and quantitative assessment of creep deformation was characterized on the solder joints. Creep deformation was analyzed using a novel mapping technique where a geometrical-regular line pattern was etched over the entire solder joint using excimer laser ablation. During creep, the laser-ablation (LA) pattern becomes distorted due to deformation in the solder joint. By imaging the distortion of laser-ablation patterns using the SEM, actual deformation mapping for the entire solder joint is revealed. The technique involves sequential optical/digital imaging of the deformation versus time history during creep. By tracing and recording the deformation of the LA patterns on the solder over intervals of time, local creep data are obtained in many locations in the joint. This analysis enables global and localized creep shear strains and strain rate to be determined.     

Creep properties of eutectic Sn-3.5Ag solder joints reinforced with mechanically incorporated Ni particles

The creep deformation behavior of eutectic Sn-3.5Ag based Ni particle rein forced composite solder joints was investigated. The Ni particle reinforced composite solder was prepared by mechanically dispersing 15 vol.% of Ni particles into eutectic Sn-3.5Ag solder paste. Static-loading creep tests were carried out on solder joint specimens at 25 C, 65 C, and 105 C, representing homologous temperatures ranging from 0.6 to 0.78. A novel-design, miniature creep-testing frame was utilized in this study. Various creep parameters such as the global and localized creep strain, steady-state creep rate, onset of tertiary creep and the activation energy for creep were quantified by mapping the distorted laser ablation pattern imprinted on the solder joint prior to testing. The Ni-reinforced composite solder joint showed improved creep resistance compared to the results previously reported for eutectic Sn-3.5Ag solder, Sn-4.0Ag-0.5Cu solder alloys, and for eutectic Sn-3.5Ag solder reinforced with Cu or Ag particle reinforcements. The activation energy for creep was ∼0.52 eV for Sn-3.5Ag and Sn-4Ag-0.5Cu solder alloys. The activation energies ranged from 0.55–0.64 eV for Cu, Ag, and Ni reinforced composite solder joints, respectively. Most often, creep fracture occurred closer to one side of the solder joint within the solder matrix.     

Characteristics of thermally expanded core fiber

Thermally expanded core (TEC) fiber is expected to reduce fiber-to-fiber and fiber-to-laser diode connection loss. This paper describes the characteristics of TEC fiber theoretically and experimentally. We reveal theoretically that when fabricating TEC fiber the mode field diameter (MFD) is enlarged more effectively by increasing the heating temperature rather than the heating time. In the 1300-1600°C temperature range with heating times between 0 and 60 min, it is necessary to control the temperature accurately so that no deviation from the target temperature is more than ±30°C. This is in order to ensure that any connection loss caused by MFD mismatch is less than 0.1 dB. We show experimentally that the propagation loss of TEC fiber is dependent on the heating region and wavelength by using a micro burner with a propane/oxygen flame. Based on the relationship between the loss characteristics and the expanded MFD, we suggest a method for nondestructively measuring the MFD in TEC fibers     

平行平面热稳定谐振腔

利用费涅耳标量衍射理论和光学模成象法则,讨论了具有内透镜的谐振腔;导出了等效谐振腔的基本结构.计算与分析表明,具有内透镜的平行平面腔可以等效为平球型谐振腔.当腔镜与透镜的配置满足本文所推导的公式时,该腔是热稳腔.讨论了这类谐振腔的选模特性和一些设计实例.     

Dynamics of thermally induced optical bistability

The steady-state model of thermally induced optical bistability, presented in a previous paper, is extended to describe transient phenomena. A nonlinear integral equation is derived which can be readily solved for arbitrary time dependence of the input power. We apply the model to switching processes in bistable ZnSe interference filters. The corresponding experimental results are found to be in good agreement with the theoretical predictions.     

Micromachined thermally based CMOS microsensors

An integrated circuit (IC) approach to thermal microsensors is presented. The focus is on thermal sensors with on-chip bias and signal conditioning circuits made by industrial complementary metal-oxide-semiconductor (CMOS) IC technology in combination with post-CMOS micromachining or deposition techniques. CMOS materials and physical effects pertinent to thermal sensors are summarized together with basic structures used for microheaters, thermistors, thermocouples, thermal isolation, and heat sinks. As examples of sensors using temperature measurement, we present micromachined CMOS radiation sensors and thermal converters. Examples for sensors based on thermal actuation include thermal flow and pressure sensors, as well as thermally excited microresonators for position and chemical sensing. We also address sensors for the characterization of process-dependent thermal properties of CMOS materials, such as thermal conductivity, Seebeck coefficient, and heat capacity, whose knowledge is indispensable for thermal sensor design. Last, two complete packaged microsystems-a thermoelectric air-flow sensor and a thermoelectric infrared intrusion detector-are reported as demonstrators     

电动式扬声器单元支撑系统的蠕变效应

电动式扬声器单元支撑系统的蠕变效应表现在扬声器单元的位移在共振频率以下会有所上升.因此在扬声器单元线性集中参数模型中仅仅用线性弹簧并联简单的粘滞阻尼来代表支撑系统是不够的.在此引入对数蠕变模型,通过测量实际的扬声器单元与传统的线性模型进行对比,结果表明对数蠕变模型与实际测量结果更吻合.     

Creep phenomena in lead-free solders

A critical review of data on microstructure and creep process activation energy values for a number of lead-free solder alloys like Sn-Ag; Sn-Bi; Sn-In was conducted. The review revealed a scatter in experimental data, which could not be explained by the dislocation creep mechanism only, even after the published data was corrected for Young’s modulus temperature dependence. An analysis of the data implies that possible origin of such a scatter is nucleation, accumulation and further growth of such internal defects as pores and microcracks during creep. It is shown that these processes may affect the measured steady-state creep rates, and may be one of the major reasons for the observed scatter in experimental data, and, therefore, must be taken into consideration in lead-free solder alloys’ creep studies.     

A thermally tunable terahertz metamaterial absorber

A thermally tunable terahertz metamaterial absorber (MA) with InSb embedded in a metal-dielectric-metal structure is proposed. The transmission and tuning properties of the proposed metamaterial absorber are analyzed for the temperature ranging from 160 K to 350 K. The simulated results show that the maximum absorption of the absorber is nearly 99.8% at a full-width at half-maximum (FWHM) of 38 GHz, and the absorption frequency can be dynamically tuned from 0.82 THz to 1.02 THz.     

IR luminescence in thermally treated silicon

Near-edge IR luminescence with peak emission at E = 1.084 eV has been studied in n- and p-type silicon samples with different contents of interstitial oxygen under excitation with a Nd: YAG laser diode. Thermal treatments of the samples demonstrated that the luminescence nearly completely disappears upon thermal treatments at T = 1050°C and is partly restored in two stages in subsequent thermal treatments in the temperature range 550–800°C. The temperature intervals of luminescence quenching and restoration (500–600 and 700–800°C) correlate with the temperature ranges of dissolution for shallow oxygen precipitates (1000°C) and the generation of oxygen-containing thermal defects, the so-called thermal donors of types I and II. The data obtained suggest that the electronic states related to thermal donors are traps for nonequilibrium carriers and the emptying of these traps contributes to the near-edge emission.     

Radiation damage of thermally oxidized MOS capacitors

The influence of preparation parameters and the effect of X-rays (150 keV, 10⁴rad (Si)) on oxide charge Qoxand interface state density Nssin thermally oxidized MOS varactors under different biasing conditions during irradiation has been investigated. The interface state density was determined by the ac conductance method before and after irradiation. The oxide charge has been evaluated with regard to the charge Qssof the interface states. Qsshas beeu discussed with the aid of simple models concerning the energetic distribution and recharge character of the interface states. The results indicate a similar dependence between flatband voltage, interface state density, and normalized oxide charge density as a function of gate bias during irradiation. Furthermore, the so-called "oxidation triangle" of oxide charge before irradiation exists for interface states as well. Calculations on the basis of the Schottky barrier model of the irradiated MOS structure show that the radiation-induced charge exists at both interfaces in the oxide layer. Radiation tolerance of the MOS capacitors as a function of technological parameters is discussed.     

A Method to Identify Steady Creep Strain from Indentation Creep Using a New Reference Area of Indentation

For the design of high-density electronic packages, finite element method (FEM) analyses to evaluate strength reliabilities of solder joints should be conducted by employing the material parameters which can precisely reflect the creep properties of solder joints in actual electronic equipment. To obtain accurate results of the structural analyses of the solder joints, a method to evaluate the steady-state creep deformation in situ must be developed. The indentation creep test is an effective method to evaluate the creep properties of the solder joints in situ; however, the creep properties obtained by this method do not give the same results as those obtained by tensile creep tests using bulk specimens. In this paper, the indentation creep test at 1 N loading for 9,000 s duration was experimentally conducted to confirm that the steady-state creep deformation obtained by the indentation creep test did not coincide with that by the tensile creep tests using bulk specimens. To identify the reason, the indentation creep simulation was conducted by FEM analysis. As a result, it was found that the reference area used to obtain the creep strain from the indentation creep test should be modified. A method to obtain the new reference area is proposed from comparisons of experiments with simulations. Finally, this paper shows that the creep properties obtained by the indentation creep test using the new reference area coincided with those obtained by tensile creep tests using bulk specimens.     

A quasi-equilibrium thermally stimulated current process

A TSC spectrum in which the peak current occurs at a temperature independent of heating rate has been observed in polycrystalline semiconductor TFTs. A model is proposed in which a band of states in the grain boundary region pins the Fermi level. The peak is due to emptying of these states to the conduction band when the TFT is depleted of mobile carriers.     

Off-state instabilities in thermally nitrided-oxide n-MOSFETs

The significant off-stage gate current of nitrided-oxide n-MOSFETs can be attributed to severe hot-hole injection into the gate oxide during band-to-band (B-B) tunneling due to a nitridation-induced lowering of the barrier height for hole injection. Some of the injected holes are even trapped in the gate oxide above the deep-depletion layer of the drain and thus decrease the gate-induced drain leakage (GIDL) current. A subsequent hot-electron injection into the gate oxide can neutralize these trapped holes and make the reduced GIDL current recover, even increase beyond the original value. The proposed mechanism of the GIDL degradation and recovery behaviors can be confirmed by the observed change in the ratio of the substrate to source currents, as well as by the field-distribution analysis of the gate oxide under stressing     

Nanoindentation Creep Behavior of Nanocomposite Sn-Ag-Cu Solders

High-density, ultrasmall-pitch electronic applications require miniaturized solder bumps with improved thermomechanical performance. In addition, novel techniques which are able to precisely characterize these solder bumps are needed. One approach to meeting both of these requirements is to make use of recently developed nanocomposite solders with enhanced creep resistance, and to characterize these solders using a nanoindentation technique. In the present study, the creep behavior of ceria-reinforced nanocomposite solder foils fabricated by the accumulative roll-bonding process was characterized using a depth-sensing nanoindentation technique. It was found that the creep resistance of the composites increased with increasing volume fraction of CeO₂ reinforcement, and it was deduced that the creep deformation of this nanocomposite proceeded by deformation of the matrix, with the role of the reinforcement being to increase the creep resistance by reducing the effective stress acting on the matrix. The values of the creep exponent suggested that the dominant creep deformation mechanisms involved were diffusion creep and grain boundary sliding.     

Stress in thermally annealed parylene films

Stress is a problem for thin film Multichip Modules (MCM) fabrication and reliability. Stress can cause adhesion failure, wire cracking and fracture of the polymer. Parylene can be deposited in a low compressive stress state, and here we explore its temperature-stress behavior. The stress generated by thermal coefficient of expansion (TCE) mis-match between thick parylene-n (PA-n) films (4-7 μm)and Si substrates has been studied as a function temperature. Measurements of stress were made on as-deposited films; after thermal annealing and also in-situ. The as-deposited films are under compressive stress of 20 MPa. As the films are annealed, the stress decreases and then becomes tensile, peaking at about 53 MPa. It was observed that stress was reduced in PA-n films during the crystal phase transformations at around 200 and 270°C. The stress has been compared to reported data from other sources and found to agree. Stress in PA-c (chlorinated), and co-polymerized PA-n ande (ethyl) films, and a fluorinated parylene (PA-f) are also evaluated. The effective glass transition temperatures and the linear coefficient of expansion have also been estimated.     

Creep Behavior of Bi-Containing Lead-Free Solder Alloys

The creep behavior of Sn-3.0Ag-0.5Cu (SAC305), Sn-3.4Ag-1.0Cu-3.3Bi (SAC-Bi), and Sn-3.4Ag-4.8Bi (SnAg-Bi, all wt.%) was studied in constant-stress creep tests from room temperature to 125°C. The alloys were tested in two microstructural conditions. As-cast alloys had a composite eutectic-primary Sn structure, while in aged alloys the eutectic regions were replaced by a continuous Sn matrix with coarsened intermetallic (Cu₆Sn₅ and Ag₃Sn) particles. After aging, Bi in SAC-Bi and SnAg-Bi was found as precipitates at grain boundaries and grain interiors. The creep resistance of of-cast SAC305 was higher than that of as-cast Bi-containing alloys, but after aging the SAC305 had the lowest creep resistance. The creep strain rates in SAC-Bi and SnAg-Bi were much less affected by aging. The apparent activation energy for creep was also changed more for SAC305 than for the other two alloys. The creep behavior of SAC-Bi and SnAg-Bi can be understood by considering the solubility of Bi in Sn. The difference in creep behavior between as-cast and aged SAC-Bi is greatly reduced when room-temperature test results are excluded from analysis. This suggests that the strongest influence on creep in these alloys is due to Bi solute interaction with moving dislocations during deformation.     

无铅电子钎料合金蠕变性能研究

设计制作了一种简单可靠的弯折蠕变测量装置,比较了两种无铅电子钎料合金Sn-9Zn和Sn-3.5Cu-0.7Ag与传统电子钎料合金Sn-40Pb的常温蠕变性能,以及冷却条件对其蠕变强度的影响。结果表明:两种无铅钎料的抗蠕变性能大大优于传统锡铅钎料;Sn-3.5Ag-0.7Cu合金的抗蠕变性能优于Sn-9Zn合金;冷却速率对Sn-9Zn合金和Sn-3.5Ag-0.7Cu合金组织的影响类似,然而对蠕变强度的影响却相反:水冷使两种合金的组织相对于空冷都明显细化,Sn-9Zn合金的蠕变强度因之降低,而Sn-3.5Ag-0.7Cu合金的蠕变强度却因之提高。对可能产生的原因进行了讨论。