SuperCSPTM

SuperCSP is fabricated by building up the interposer with high reliability encapsulant on the chip by wafer level packaging technology. New encapsulation technology enables real chip-sized package from a package perspective. It is also a known good encapsulated die (KGED) from a die perspective. The reasons why board level reliability of SuperCSP is good regardless of extremely low bump-standoff height are as follows. (1) The C.T.E of encapsulant for SuperCSP is close to that of motherboard, so that the encapsulant layer effectively reduces stress occurring in the solder interconnecting portion. (2) Encapsulant with high adhesive strength reinforces and fixes the delicate connecting portion of chip and post, and also does not allow its deformation. (3) Connecting portion of solder ball and post has a strong structure and can tolerate the stress because solder balls catch hold of the whole surface of metal posts, which stick out from the encapsulant and have a mound like structure     

WLCSP中微焊球结构尺寸对其热应力的影响

晶圆级芯片尺寸封装(WLCSP)微焊球结构尺寸对其热机械可靠性有重要的影响。通过二维有限元模拟筛选出对WLCSP微焊球及其凸点下金属层(UBM)中热应力影响显著的参数,采用完全因子实验和多因子方差统计分析定量评估各种因子影响的显著性,最后建立三维模型,用子模型技术研究关键尺寸因子对热应力变化的影响。研究发现,焊球半径是影响焊球热应力的最关键尺寸因子,电镀铜开口和铜焊盘厚度对焊球热应力的影响也较显著;钝化层开口和焊球半径是影响UBM热应力的最关键尺寸因子。随着焊球半径增大,焊球热应力减小。     

Improved thermal fatigue reliability for flip chip assemblies usingredistribution techniques

The bond pad design on a chip can be reconfigured to a new pad design using a redistribution layer, based on multichip module-deposited (MCM-D) technology. The new pad configuration can be used for flip chip mounting. The thermo-mechanical reliability of these redistributed flip chip structures is in particular determined by the visco-plastic deformations of the solder joints and by the stresses in the photosensitive BCB redistribution layers. In this paper, the influence of this redistribution layer on the solder joint reliability is investigated. Also the induced stresses in this redistribution layer may not exceed the ultimate stress level. Three different redistribution processes are considered. Finite element simulations and Coffin-Manson based reliability models are used to compare the thermal cycling reliability of redistributed and standard flip chip assemblies. The existence of a photosensitive BCB redistribution layer on the chip influences the thermal fatigue of solder joints. The largest reliability improvement using redistributed chips is achieved by moving the solder joints from the perimeter to the interior of the die resulting in an area array flip chip     

3-D Structure Design and Reliability Analysis of Wafer Level Package With Stress Buffer Mechanism

With the present trend of multifunction and minimizing of size, the conventional electronic package type no longer meets the requirement of the new-generation products. Consequently, new type packaging, based on the wafer level packages (WLPs) and chip scale packages (CSPs) technology are being developed to achieve these requirements, as well as long term reliability. Novel wafer-level chip scale packages (WLCSP) with a stress buffer layer and bubble-like plate (Fig. 1) are proposed in this research to improve the solder joint fatigue life. The thermal stress caused by the coefficient of thermal expansion mismatch can be significantly reduced, and the reliability of the WLP could be substantially enhanced by this new design. In order to realize the relationship of the solder joint fatigue life, stress buffer layer and bubble-like plate, a finite element parametric analysis applying software ANSYS is utilized. In additions, the methodology based on the finite element method (FEM) used in the study has been verified by the relative experiments in our previous researches. The design parameters include the thickness of the stress buffer layer, thickness, bending angle and standoff height of the different types of bubble-like plate. The results of the FEM analysis reveal that the stress buffer layer and bubble-like plate can relax the thermal stresses of solder joints and enhance the package reliability. Besides, the peeling stress between stress buffer layer and two different types of bubble-like plates is discussed, and the stress state of the leadframe is also analyzed in this research. Furthermore, the findings of this research can be used as the guideline for advanced WLCSP design     

Impact Testing of Sn-3.0Ag-0.5Cu Solder with Ti/Ni(V)/Cu Under Bump Metallization After Aging at 150°C

Nonmagnetic Ni(V) metal and low consumption rate with solders are the advantages of sputtered Ti/Ni(V)/Cu under bump metallization (UBM). However, a Sn-rich phase (“Sn-patch” herein) can form in the Ni(V) layer after reflow and aging. In lead-free solder, Sn-patches form and grow more quickly than in Sn-Pb solder. Thus, the effect of Sn-patches on solder joint reliability becomes critical. In this study, Sn-3.0Ag-0.5Cu solder was reflowed with Ti/Ni(V)/Cu UBM at 250°C for 60 s, and then aged at 150°C for various durations. A high-speed impact test was introduced to evaluate solder joint reliability. After impact testing, it was found that, the larger the Sn-patch, the greater the propensity of the solder joint to suffer brittle fracture. The correlation between Sn-patch and solder joint reliability is discussed.     

A Novel Design Structure for WLCSP With High Reliability, Low Cost, and Ease of Fabrication

Wafer level chip scale packaging (WLCSP) has some advantages, such as real die size packaging, high electrical performance, and low manufacturing cost. However, because the mechanical reliability of a large die can not be guaranteed due to the coefficient of thermal expansion (CTE) mismatch between silicon and organic printed circuit board (PCB), WLCSP technology is still not fully accepted. We have developed a new solder joint protection-WLCSP (SJP-WLCSP) structure with a delamination layer interposed between the top layer of the chip and the bottom insulating layer of the metal redistribution traces. The stress on the solder joints can be released by the cracks forming in the delamination layer, which protects the solder joints from cracking. Since the cracking of the delamination layer is irrelevant to the electrical circuits of the packaging, the packaged integrated circuits (IC) device remains functional. One of the possibilities for processing the SJP-WLCSP was implemented and validated successfully in the SiLK-wafer samples. The board level packaging samples, using the daisy chain resistance measurement passed 1000 cycles of the temperature cycling testing.     

Design and optimization of thermo-mechanical reliability in wafer level packaging

In this paper, a variety of wafer level packaging (WLP) structures, including both fan-in and fan-out WLPs, are investigated for solder joint thermo-mechanical reliability performance, from a structural design point of view. The effects of redistribution layer (RDL), bump structural design/material selection, polymer-cored ball application, and PCB design/material selection are studied. The investigation focuses on four different WLP technologies: standard WLP (ball on I/O WLP), ball on polymer WLP without under bump metallurgy (UBM) layer, ball on polymer WLP with UBM layer, and encapsulated copper post WLP. Ball on I/O WLP, in which solder balls are directly attached to the metal pads on silicon wafer, is used as a benchmark for the analysis. 3-D finite element modeling is performed to investigate the effects of WLP structures, UBM layer, polymer film material properties (in ball on polymer WLP), and encapsulated epoxy material properties (in copper post WLP). Both ball on polymer and copper post WLPs have shown great reliability improvement in thermal cycling. For ball on polymer WLP structures, polymer film between silicon and solder balls creates a ‘cushion’ effect to reduce the stresses in solder joints. Such cushion effect can be achieved either by an extremely compliant film or a ‘hard’ film with a large coefficient of thermal expansion. Encapsulated copper post WLP shows the best thermo-mechanical performance among the four WLP structures. Furthermore, for a fan-out WLP, it has been found that the critical solder balls are the outermost solder balls under die-area, where the maximum thermal mismatch takes place. In a fan-out WLP package, chip size, other than package size, determines the limit of solder joint reliability. This paper also discusses the polymer-cored solder ball applications to enhance thermo-mechanical reliability of solder joints. Finally, both experimental and finite element analysis have demonstrated that making corner balls non-electrically connected can greatly improve the WLP thermo-mechanical reliability.     

Solder joints layout design and reliability enhancements of wafer level packaging using response surface methodology

As the industry keeps moving towards further miniaturization of electronic devices, even smaller sizes, a lower economical cost, and higher reliability are not only convenient but have become a necessity of the design. A well-designed package structure can effectively restrain the solder joint fatigue failure induced by material coefficient of thermal expansion (CTE) mismatch. Wafer level chip scaling package (WLCSP) has a high potential for future advanced packaging. However, the solder joint reliability for large chip sizes of up to 100 mm² without underfill is still an issue that needs to be resolved. For solving this problem, a double-layer WLCSP (DL-WLCSP) with both a stress compliant layer and dummy solder joints is proposed in this research to enhance the solder joint fatigue life. Moreover, a hybrid method is employed to predict the profile of solder joint after reflow process. To ensure the correctness of the methodology of the analysis, a Rambus DRAM layout is implemented as the test vehicle to demonstrate the applicability and reliability of the DL-WLCSP. The results of the thermal cycling experimental test show good agreement with the simulated analysis. In addition, besides the geometrical design parameters of the silicon die thickness and the thickness of the stress compliant layer, the reliability impact for the arrangement of die-side and substrate-side pad diameter is investigated by means of the design of experiment (DOE). In addition, the Response Surface Methodology (RSM) with central composite designs (CCD) is adopted to obtain the parameter sensitivity information by the three-dimensional nonlinear finite element analysis (FEA). Analysis of variance (ANOVA) is conducted to determine the significance of the fitted regression model. The analytic results reveal that the stress compliant layer and the dummy joints can effectively reduce the stress concentration phenomenon, which occurs around the outer-corner of the solder joint. The smaller thermal strains can be controlled through better size combination between die-side and substrate-side pad diameter.     

Improved reliability in small multichip ball grid arrays

The subject of this paper is a 14×22 mm ball grid array (BGA) integrated circuit assembly containing two or three chips. Three failure modes came to light in the reliability testing of this BGA package: delamination of solder resist from the top copper layer occurred in moisture resistance testing; cracking of the top layer solder resist and consequent cracking of the top copper layer occurred in temperature cycling; and cracking of the bottom layer solder resist which propagated into the bottom copper layer occurred in thermal shock. The failure analysis techniques used to disclose these failures are presented. Finite element analysis of thermomechanical stress within the multichip structure was carried out. The purpose was to find the root cause of one of the failure modes and to explore possible means of overcoming the stress damage. The characteristics of the original and modified substrate layout designs are detailed. The improved performance in reliability testing is compared with the original. All failure modes were eliminated in the final design, and the product was qualified to greatly improved reliability standards.     

Ball-grid-array solder joint model for assembly-level impact reliability prediction

It has been well established that lead-free solder underperforms conventional leaded solder in reliability under dynamic impact. Common failures observed on ball-grid-array (BGA) solder balls on chip under board level impact include bulk solder ductile failure, intermetallic (IMC) layer crack and pad-lift. In this work, a finite element modeling approach was proposed to model bulk solder ductile failure and intermetallic layer crack. The use of beam elements and connector elements to represent the bulk solders and board/component side intermetallic layers, respectively, offers the advantage of simplicity over the use of continuum elements and cohesive elements for solder joints. This approach enables the modeling of assembly level impact with significantly less computational resources. The model was verified by comparing its prediction of BGA solder reliability against actual test results in a dynamic four-point bend test. The physical tests consist of ball impact at varying heights on a board with a mounted chip, and the subsequent analysis of the failure modes of the BGA solder joints. Simulation results were in good agreement with test results. The study shows that it is feasible to model BGA solder joint ductile failure and intermetallic layer crack under impact with simple elements with reasonable accuracy.     

Sensitivity Design of DL-WLCSP Using DOE With Factorial Analysis Technology

Newer, faster, and smaller electronic packaging approaches with high I/O counts and more complex semiconductor devices are emerging steadily and rapidly. Wafer-level chip scaling package (WLCSP) has a high potential for future electronic packaging. However, the solder joint reliability for a large chip size of about 100 mm² without underfill remains a troubling issue that urgently requires a solution. To this end, a double-layer WLCSP (DL-WLCSP) with stress compliant layers and dummy solder joint is adopted in this research in order to study the design parameters of enhancing the solder joint fatigue life. To ensure the validity of the analysis methodology, a test vehicle of Rambus DRAM is implemented to demonstrate the applicability and reliability of the proposed DL-WLCSP. The results of the thermal cycling in the experimental test show good agreement with the simulated analysis. Furthermore, to investigate the reliability impact of the design parameters, including solder volume, the arrangement of the die-side and substrate-side pad diameter, second compliant layer thickness, die thickness, and the printed circuit board (PCB) thickness, a design of experiment (DOE) with factorial analysis is adopted to obtain the sensitivity information of each parameter by the three-dimensional nonlinear finite-element models (FEMs). The statistics results of the analysis of variance reveal that the thickness of the second stress compliant layer and the volume of the solder joint can effectively reduce the stress concentration phenomenon, which occurs around the outer corner of the solder joint. In addition, the evident interaction between design parameters can also be obtained. The smaller thermal strains can be achieved through a better combination of design parameters of the geometry so as to provide the actual requirement of the physical information prior to manufacturing     

Effects of build-up printed circuit board thickness on the solderjoint reliability of a wafer level chip scale package (WLCSP)

The creep analyses of solder-bumped wafer level chip scale package (WLCSP) on build-up printed circuit board (PCB) with microvias subjected to thermal cyclic loading are presented. The emphasis of this study is placed on the effects of the thickness of the PCB with a microvia build-up layer on the solder joint reliability of the WLCSP assembly. The 62Sn-2Ag-36Pb solder joints are assumed to follow the Garofalo-Arrhenius creep constitutive law. The shear stress and creep shear strain hysteresis loops, shear stress range, creep shear strain range, and creep strain energy density range at different locations in the corner solder joint are presented for a better understanding of the thermal-mechanical behaviors of the solder-bumped WLCSP on build-up PCB with microvias. It is found that, due to the large coefficient of thermal expansion of the build-up resin, the effects of thickness of the PCB with microvia build-up layer become much more significant than that without the microvia build-up layer     

Life prediction of HCPV under thermal cycling test condition

High-concentration photovoltaic (HCPV) module is subject to larger thermal stress due to its more severe temperature fluctuation in real operating conditions. In the thermal cycling test, excessive thermal stress might occur at the peripheral solder layer. For the large area bonding structure, thermal-induced stress is the main cause for cracks. Crack growth is expected to start from the edges of the solder layer and progress to the center. The shrinkage of the bonding area increases the junction temperature of solar cells and reduces the energy-conversion efficiency of the HCPV module. In this study, the stress/strain behavior of the HCPV module under thermal cycling test is analyzed using finite element analysis software, ANSYS®. Results indicate that the von Mises creep strain distribution at the solder layer’s edge is independent of the package’s dimensions. The lifetime of HCPV with uniform solder layer could be predicted by assuming that the crack propagation rate is constant during solder layer degradation. Furthermore, lifetime of tilted HCPV module could be predicted by compensating the variation of thickness of solder layer during crack propagation.     

Board level solder joint reliability analysis of stacked die mixed flip-chip and wirebond BGA

Stacked die BGA has recently gained popularity in telecommunication applications. However, its board level solder joint reliability during the thermal cycling test is not as well-studied as common single die BGA. In this paper, solder joint fatigue of lead-free stacked die BGA with mixed flip-chip (FC) and wirebond (WB) interconnect is analyzed in detail. 3D fatigue model is established for stacked die BGA with considerations of detailed pad design, realistic shape of solder ball, and non-linear material properties. The fatigue model applied is based on a modified Darveaux’s approach with non-linear viscoplastic analysis of solder joints. Based on the FC–WB stack die configuration, the critical solder ball is observed located between the top and bottom dice corner, and failure interface is along the top solder/pad interface. The modeling predicted fatigue life is first correlated to the thermal cycling test results using modified correlation constants, curve-fitted from in-house lead-free TFBGA46 (thin-profile fine-pitch BGA) thermal cycling test data. Subsequently, design analyzes are performed to study the effects of 20 key design variations in package dimensions, material properties, and thermal cycling test conditions. In general, thinner PCB and mold compound, thicker substrate, larger top or bottom dice sizes, thicker top die, higher solder ball standoff, larger solder mask opening, smaller PCB pad size, smaller thermal cycling temperature range, longer ramp time, and shorter dwell time contribute to longer fatigue life. SnAgCu is a common lead-free solder, and it has much better board level reliability performance than eutectic solder based on modeling results, especially low stress packages.     

Three-Dimensional Thermomechanical Simulation of Fine-Pitch High-Count Ball Grid Array Flip-Chip Assemblies

Flip-chip technology is increasingly prevalent in electronics assembly [three-dimensional (3D) system-in-package] and is mainly used at fine pitch for manufacture of megapixel large focal-plane detector arrays. To estimate the reliability of these assemblies, numerical simulations based on finite-element methods appear to be the cheapest approach. However, very large assemblies contain more than one million solder bumps, and the optimization process of such structures through numerical simulations turns out to be a very time-consuming task. In many applications, the interconnection layer of such flip-chip assemblies consists of solder bumps embedded in epoxy filler. For such configurations, we propose an alternative approach, which consists in replacing this heterogeneous interconnection layer by a homogeneous equivalent material (HEM). A micromechanical model for the estimation of its equivalent thermoelastic properties has been developed. The obtained constitutive law of the HEM was then implemented in finite-element software (Abaqus^®). Thermomechanical responses of tested assemblies submitted to loads corresponding to manufacturing conditions have been analyzed. The homogenization–localization process allowed estimation of the mean values of stresses and strains in each phase of the interconnection layer. To access more precisely the stress and strain fields in these phases, two models of structural zoom, taking into account the real solder bump geometry, have been tested. The obtained local stress and strain fields corroborate the experimentally observed damage initiation of the solder bumps.     

Procedure for design optimization of a T-cap flip chip package

In this study, a 1/4 three-dimensional finite element model of a T-cap flip chip package containing the substrate, underfill, solder bump, silicon die, metal cap and cap attachment was established to conduct thermo-mechanical reliability study during the flip chip fabrication processes. The applied thermal load was cooled from 183 °C to ambience 25 °C to determine the thermal stress and warpage during the curing period of solder ball mounting process. Under fixed geometry, two levels of underfill, metal caps and cap attachments were used to conduct the 2³ factorial design for determining reliable material combinations. The statistical tests revealed that the significant effects affecting the thermal stress were the underfill, metal cap, cap attachment and the interaction between the underfill and cap attachment. The metal cap, cap attachment and their interaction significantly affected the warpage. The proposed regression models were used to perform the surface response simulations and were useful in selecting suitable materials for constructing the package. This study provides a powerful strategy to help the designer to easily determine reliable packaging structures under various reliability considerations.     

Reliability evaluations of under bump metallurgy in two soldersystems

Under bump metallurgy (UBM) reliability is one of the critical issues in the total reliability of a flip-chip bumping technology. Since the UBM materials and structures vary for different bumping technologies, the UBM strength and reliability need to be determined for each design and process. In addition, the stress that a UBM experiences during thermal cycles depends on the solder alloy used in the interconnect. Different solder alloys require different UBM structures and strengths to achieve good reliability in thermal cycling. In this study, a simplified stress model is developed to determine the UBM stress during thermal cycling. A simplified stress model for the UBM strength is also developed. These models are used to predict the stress and strength of the UBM under the die pull test and the thermal cycle conditions for both eutectic and high lead solder systems. A methodology for using the pull test results to evaluate UBM reliability is also discussed. This methodology can be extended to the studies of UBM's with other solder systems such as lead free solder systems     

Thermo-mechanical finite element analysis in a multichip build up substrate based package design

This paper presents a thermo-mechanical analysis of a multichip module (MCM) package design, with emphasis on the package warpage, thermally induced stress and the second level solder joint reliability. The MCM package contains four flip chips which are mounted on a build up substrate. First, the effect of the positioning of four silicon dice within the MCM package on the warpage of the package is studied. Second, the effect of package dimensions (the heat spreader thickness, the structural adhesive thickness and the substrate thickness) on the maximum residual stress as well as the warpage of the package is performed. Finally, this paper presents a 3D sliced model for solder joint reliability of the MCM assembly. A creep constitutive relation is adopted for the 63Sn/37Pb solder to account for its time and temperature dependence in thermal cycling. The fatigue life of solder joint is estimated by the Darveaux's approach. A series of parametric study is performed by changing the package dimensions. The results show that the largest die tends to experience highest stresses at its corner and has more influence on the warpage of the package than smaller dice. The results also show the most sensitivity factors that affect the package warpage and the second level solder joint reliability are the substrate thickness and the heat spreader thickness. The structural adhesive thickness has no major effect on the package warpage, the maximum von Mises stress of the package and the second level solder joint reliability.     

Effect of randomness of Cu-Sn intermetallic compound layerthickness on reliability of surface mount solder joints

A statistical reliability analysis on thermal fatigue lifetime of surface mount solder joints, considering randomness of Cu-Sn intermetallic compound (IMC) layer thickness, is presented. Based on published thermal fatigue life test data, the two-parameter Weibull distribution of the thermal fatigue lifetime for a fixed IMC layer thickness is found, and a K-S goodness-of-fit test is conducted to examine the goodness of fit of the assumed Weibull distribution. Then, the Weibull parameters as functions of IMC layer thickness are obtained. Considering the randomness of IMC layer thickness, the MTTF and reliability of surface mount solder joints on thermal cycles are analyzed. For surface mount solder joints formed under the same conditions and loaded during the same thermal cycling as stated in the publication, numerical results of the MTTF and reliability are presented. The results show that when the mean value of MC layer thickness is low (e.g., smaller than 1.5 μm), the effect of randomness of IMC layer thickness is significant; i.e., the MTTF has strong dependence on IMC layer thickness distribution; and the reliability is significantly different at high thermal cycles. When the mean value of IMC layer thickness is high (e.g., greater than 2.0 μm), the effect of randomness of IMC layer thickness is negligible. Therefore, the presented results are important to the reliability study of surface mount solder joints. Even though the validity of the presented results based on the test data remains to be verified from other sources of data, the proposed statistical method is generally applicable for thermal fatigue reliability analysis of surface mount solder joints. By combining the proposed method with the forming mechanism of IMC layer under varying manufacturing and loading conditions, a comprehensive reliability analysis on thermal fatigue lifetime of surface mount solder joints can be expected