Interfacial reaction and failure mode analysis of the solder joints for flip-chip LED on ENIG and Cu-OSP surface finishes

The interfacial reactions and failure modes of the solder joints for flip-chip light emitting diode (LED) on electroless nickel/immersion gold (ENIG) and Cu with organic solderability preservatives (Cu-OSP) surface finishes were investigated in this study. The experimental results demonstrate that the interfacial reactions in the Au/Sn–Ag–Cu(SAC)/ENIG and Au/SAC/Cu systems are different but the failure mechanisms of the two types of solder joints are similar during the shear test. For the Au/SAC/ENIG system, the Au layer on the surface finish of the diodes dissolved into the molten solder and transformed into a continuous (Au, Ni)Sn₄ IMC layer at the diode/solder interface during reflow and the interfacial IMC at the solder/ENIG interface is dendritic Ni₃Sn₄ IMC grains which are surrounded by (Au, Ni)Sn₄. For the Au/SAC/Cu system, however, no IMC layers can be observed at the diode/solder interface. The interfacial IMC at the solder/Cu interface is (Cu, Au)₆Sn₅ and a Cu₃Sn IMC layer at the (Cu, Au)₆Sn₅/Cu interface. Tiny (Au, Cu)Sn₄ IMC grains distribute in the solder layer and surround the (Cu, Au)₆Sn₅ grains. For the two types of systems, the primary failure mode for the cathode is due to the broken of the Si-based insulation layer which led to a high residue stress and poor connection between the Si-based layer and the solder layer. Meanwhile, the failure of the solder joint for the anode is mainly because of the failure of the solder layer under the conductive via. The crack generally forms at this area and then propagated along the diode or the diode/solder interface.     

Influence of Cu content on compound formation near the chip side for the flip-chip Sn-3.0Ag-(0.5 or 1.5)Cu solder bump during aging

Sn-Ag-Cu solder is a promising candidate to replace conventional Sn-Pb solder. Interfacial reactions for the flip-chip Sn-3.0Ag-(0.5 or 1.5)Cu solder joints were investigated after aging at 150°C. The under bump metallization (UBM) for the Sn-3.0Ag-(0.5 or 1.5)Cu solders on the chip side was an Al/Ni(V)/Cu thin film, while the bond pad for the Sn-3.0Ag-0.5Cu solder on the plastic substrate side was Cu/electroless Ni/immersion Au. In the Sn-3.0Ag-0.5Cu joint, the Cu layer at the chip side dissolved completely into the solder, and the Ni(V) layer dissolved and reacted with the solder to form a (Cu_1−y,Ni_y)₆Sn₅ intermetallic compound (IMC). For the Sn-3.0Ag-1.5Cu joint, only a portion of the Cu layer dissolved, and the remaining Cu layer reacted with solder to form Cu₆Sn₅ IMC. The Ni in Ni(V) layer was incorporated into the Cu₆Sn₅ IMC through slow solid-state diffusion, with most of the Ni(V) layer preserved. At the plastic substrate side, three interfacial products, (Cu_1−y,Ni_y)₆Sn₅, (Ni_1−x,Cu_x)₃Sn₄, and a P-rich layer, were observed between the solder and the EN layer in both Sn-Ag-Cu joints. The interfacial reaction near the chip side could be related to the Cu concentration in the solder joint. In addition, evolution of the diffusion path near the chip side in Sn-Ag-Cu joints during aging is also discussed herein.     

Comparison of Sn-Ag-Cu Solder Alloy Intermetallic Compound Growth Under Different Thermal Excursions for Fine-Pitch Flip-Chip Assemblies

The intermetallic compound (IMC) evolution in Cu pad/Sn-Ag-Cu solder interface and Sn-Ag-Cu solder/Ni pad interface was investigated using thermal shock experiments with 100-μm-pitch flip-chip assemblies. The experiments show that low standoff height of solder joints and high thermomechanical stress play a great role in the interfacial IMC microstructure evolution under thermal shock, and strong cross-reaction of pad metallurgies is evident in the intermetallic growth. Furthermore, by comparing the IMC growth during thermal aging and thermal shock, it was found that thermal shock accelerates IMC growth and that kinetic models based on thermal aging experiments underpredict IMC growth in thermal shock experiments. Therefore, new diffusion kinetic parameters were determined for the growth of (Cu,Ni)₆Sn₅ using thermal shock experiments, and the Cu diffusion coefficient through the IMC layer was calculated to be 0.2028 μm²/h under thermal shock. Finite-element models also show that the solder stresses are higher under thermal shock, which could explain why the IMC growth is faster and greater under thermal shock cycling as opposed to thermal aging.     

Effects of Surface Finishes and Current Stressing on Interfacial Reaction Characteristics of Sn-3.0Ag-0.5Cu Solder Bumps

The effects of surface finishes on the in situ interfacial reaction characteristics of ball grid array (BGA) Sn-3.0Ag-0.5Cu lead-free solder bumps were investigated under annealing and electromigration (EM) test conditions of 130°C to 175°C with 5.0 × 10³ A/cm². During reflow and annealing, (Cu,Ni)₆Sn₅ intermetallic compound (IMC) formed at the interface of electroless nickel immersion gold (ENIG) finish. In the case of both immersion Sn and organic solderability preservative (OSP) finishes, Cu₆Sn₅ and Cu₃Sn IMCs formed. Overall, the IMC growth velocity of ENIG was much lower than that of the other finishes. The activation energies of total IMCs were found to be 0.52 eV for ENIG, 0.78 eV for immersion Sn, and 0.72 eV for OSP. The ENIG finish appeared to present an effective diffusion barrier between the Cu substrate and the solder, which leads to better EM reliability in comparison with Cu-based pad systems. The failure mechanisms were explored in detail via in situ EM tests.     

Isothermal and thermal cycling aging on IMC growth rate in lead-free and lead-based solder interface

The growth of interfacial intermetallic compounds (IMC) between Pb-free and Pb-based solders with different surface finish (Cu and Ni/Au) metallization is a major concern for long-term solder joint reliability performance in electronic assemblies. The growth rate of the IMC layer can affect the solder joint reliability. Analysis of solid-state diffusion mechanism for the growth of IMC between solder-to-substrate interface for Pb-free and Pb-based solders subject to isothermal and thermal cycling aging were conducted. Experimental study of IMC layer growth between Sn3.8Ag0.7Cu and Ni/Au surface finish by isothermal aging versus thermal cycling (TC) aging was investigated to develop a framework for correlating IMC layer growth behavior. An integrated model for IMC growth was derived to describe the Ni-Cu-Sn IMC growth behavior subject to TC aging. Comparison of modeling and test results showed that IMC layer growth rate under TC aging was accelerated. It is noted that IMC layer growth study from various references showed different experimental data and growth kinetic parameters for both liquid-state and solid-state reactions.     

Effect of additions of ZrO2 nano-particles on the microstructure and shear strength of Sn-Ag-Cu solder on Au/Ni metallized Cu pads

Nano-sized, nonreacting, noncoarsening ZrO₂ particles reinforced Sn-3.0 wt%Ag-0.5 wt%Cu composite solders were prepared by mechanically dispersing ZrO₂ nano-particles into Sn-Ag-Cu solder. The interfacial morphology of unreinforced Sn-Ag-Cu solder and solder joints containing ZrO₂ nano-particles with Au/Ni metallized Cu pads on ball grid array (BGA) substrates and the distribution of reinforcing particles were characterized metallographically. At their interfaces, a Sn-Ni-Cu intermetallic compound (IMC) layer was found in both unreinforced Sn-Ag-Cu and Sn-Ag-Cu solder joints containing ZrO₂ nano-particles and the IMC layer thickness increased with the number of reflow cycles. In the solder ball region, AuSn₄, Ag₃Sn, Cu₆Sn₅ IMC particles and ZrO₂ nano-particles were found to be uniformly distributed in the β-Sn matrix of Sn-Ag-Cu solder joints containing ZrO₂ nano-particles, which resulted in an increase in the shear strength, due to a second phase dispersion strengthening mechanism. The fracture surface of unreinforced Sn-Ag-Cu solder joints exhibited a brittle fracture mode with a smooth surface while Sn-Ag-Cu solder joints containing ZrO₂ nano-particles ductile failure characteristics with rough dimpled surfaces.     

Interfacial microstructure of Pb-free and Pb-Sn solder balls in the ball-grid array package

Ball-grid array (BGA) samples were aged at 155°C up to 45 days. The formation and the growth of the intermetallic phases at the solder joints were investigated. The alloy compositions of solder balls included Sn-3.5Ag-0.7Cu, Sn-1.0Ag-0.7Cu, and 63Sn-37Pb. The solder-ball pads were a copper substrate with an Au/Ni surface finish. Microstructural analysis was carried out by electron microprobe. The results show that a ternary phase, (Au,Ni)Sn₄, formed with Ni₃Sn₄ in the 63Sn-37Pb solder alloy and that a quaternary intermetallic phase, (Au,Ni)₂Cu₃Sn₅, formed in the Sn-Ag-Cu solder alloys. The formation mechanism of intermetallic phases was associated with the driving force for Au and Cu atoms to migrate toward the interface during aging.     

Investigation of interfacial reaction between Au-Sn solder and Kovar for hermetic sealing application

The microstructural evolution and interfacial reactions of Au/Sn/Au/Au/Ni/Kovar joint were investigated during aging at 180 and 250 °C for up to 1000 h. The Au/Sn combination formed a rapid diffusion system. Even in non-annealed joint, three phases such as AuSn, AuSn₂ and AuSn₄ were formed. After initial aging at 180 °C, the AuSn, AuSn₂, AuSn₄, Au and Sn phases, which were formed after plating, were fully transformed into ζ-phase and δ-phase, and (Ni, Au)₃Sn₂ intermetallic compound (IMC) layer was observed between the ζ-phase and Kovar. As a whole, the microstructure of the joint was stable during aging at 180 °C. On the other hand, the solid-state interfacial reaction was much faster at 250 °C than at 180 °C. During aging at 250 °C, the Ni layer on the Kovar reacted primarily with the δ-phase in the solder, resulting in the formation and growth of the (Au, Ni)Sn IMC layer at the interface. After aging for 48 h, the Fe-Co-Ni-Au-Sn phase was formed underneath the (Au, Ni)Sn IMC layer. Furthermore, cracks were observed inside the interfacial layers after complete consumption of the Ni layer. The study results clearly demonstrate the need for either a thicker Ni layer or an alternative surface finish on Ni, in order to ensure the high temperature reliability of the Au/Sn/Au/Au/Ni/Kovar joint above 250 °C.     

Additive Effect of Kirkendall Void Formation in Sn-3.5Ag Solder Joints on Common Substrates

The Sn-3.5Ag and Sn-3.5Ag-0.2Co-0.1Ni lead-free solders were investigated on common electronics substrates, namely, organic solderability preservative (OSP) and electroless Ni/immersion Au (ENIG) surface finishes. The formation of Kirkendall voids at the interfacial region during isothermal solid aging was explored. For the Sn-3.5Ag-0.2Co-0.1Ni/OSP solder joint, the Kirkendall voids were present after isothermal solid-state aging at higher temperature (e.g., 150°C); however, the size of voids did not change remarkably with prolonged aging time due to the depressed Cu₃Sn layer growth. For ENIG surface finishes, the 0.2Co-0.1Ni additions seemed to enhance the longitudinal groove-shaped voids at the Ni₃P layer; however, void formation at the solder/Ni₃Sn₄ interface was effectively reduced. This might be attributed to the reduced Sn activity in the solder matrix and the suppressed Ni-P-Sn layer formation.     

Effect of Gold Content on the Microstructural Evolution of SAC305 Solder Joints Under Isothermal Aging

Au over Ni on Cu is a widely used printed circuit board (PCB) surface finish, under bump metallization (UBM), and component lead metallization. It is generally accepted that less than 3 wt.% Au in Sn-Pb solder joints inhibits formation of detrimental intermetallic compounds (IMC). However, the critical limit for Au content in Pb-free solder joints is not well established. Three surface-mount package platforms, one with a matte Sn surface finish and the others with Ni/Au finish, were soldered to Ni/Au-finished PCB using Sn-3.0Ag-0.5Cu (SAC305) solder, in a realistic manufacturing setting. The assembled boards were divided into three groups: one without any thermal treatment, one subjected to isothermal aging at 125°C for 30 days, and the third group aged at 125°C for 56 days. Representative solder joints were cross-sectioned and analyzed using scanning electron microscopy (SEM) and energy-dispersive x-ray spectroscopy (EDX) to investigate the evolution of the solder joint morphology as a function of Au content and isothermal aging. It was found that, if Cu is available to dissolve in the solder joint, the migration of AuSn₄ from the bulk to the interface as a result of thermal aging is mitigated.     

Microstructure, thermal analysis and hardness of a Sn–Ag–Cu–1 wt% nano-TiO2 composite solder on flexible ball grid array substrates

Sn-Ag-Cu composite solders reinforced with nano-sized, nonreacting, noncoarsening 1 wt% TiO₂ particles were prepared by mechanically dispersing TiO₂ nano-particles into Sn-Ag-Cu solder powder and the interfacial morphology of the solder and flexible BGA substrates were characterized metallographically. At their interfaces, different types of scallop-shaped intermetallic compound layers such as Cu₆Sn₅ for a Ag metallized Cu pad and Sn-Cu-Ni for a Au/Ni and Ni metallized Cu pad, were found in plain Sn-Ag-Cu solder joints and solder joints containing 1 wt% TiO₂ nano-particles. In addition, the intermetallic compound layer thicknesses increased substantially with the number of reflow cycles. In the solder ball region, Ag₃Sn, Cu₆Sn₅ and AuSn₄ IMC particles were found to be uniformly distributed in the β-Sn matrix. However, after the addition of TiO₂ nano-particles, Ag₃Sn, AuSn₄ and Cu₆Sn₅ IMC particles appeared with a fine microstructure and retarded the growth rate of IMC layers at their interfaces. The Sn-Ag-Cu solder joints containing 1 wt% TiO₂ nano-particles consistently displayed a higher hardness than that of the plain Sn-Ag-Cu solder joints as a function of the number of reflow cycles due to the well-controlled fine microstructure and homogeneous distribution of TiO₂ nano-particles which gave a second phase dispersion strengthening mechanism.     

Influence of Cu addition to interface microstructure between Sn-Ag solder and Au/Ni-6P plating

The formation and growth of intermetallics at the interface between Sn-Ag-(Cu) alloy balls and Au/Ni-6P plating were experimentally examined as a function of soldering period. Joint strengths were also evaluated by a ball pull test. For the joint with Sn-3.5Ag, the primary reaction product of Ni₃Sn₄ exhibits growth and shrinkage in thickness repeatedly with a passage of reaction time up to 30 min, while the Ni₃SnP reaction layer monotonously increases its thickness without fluctuation. In the cases of the joints with Cu bearing solder, Sn-3Ag-0.5Cu and Sn-3.5Ag-0.8Cu, a single η-(Cu,Ni)₆Sn₅ interface layer grows by fast Cu segregation from liquid solder to the interface layer on soldering. For all the soldered joints, a P-rich layer appears at the surface region of a Ni-6P plating layer by Ni depletion to form those intermetallic compounds at interfaces. The growth rate of a P-rich layer for Sn-3.5Ag is faster by about 4–8 times than those of the Sn-Ag-Cu. The presence of Cu in solder enhances the formation of the Cu₆Sn₅ intermetallic layer at the interface resulting in prevention of Ni diffusion to liquid solder. For all the soldered joints, coarsened reaction interfaces decrease the joint strengths.     

Impact of Ni concentration on the intermetallic compound formation and brittle fracture strength of Sn–Cu–Ni (SCN) lead-free solder joints

Cu₆Sn₅ and Cu₃Sn are common intermetallic compounds (IMCs) found in Sn–Ag–Cu (SAC) lead-free solder joints with OSP pad finish. People typically attributed the brittle failure to excessive growth of IMCs at the interface between the solder joint and the copper pad. However, the respective role of Cu₆Sn₅ and Cu₃Sn played in the interfacial fracture still remains unclear. In the present study, various amounts of Ni were doped in the Sn–Cu based solder. The different effects of Ni concentration on the growth rate of (Cu, Ni)₆Sn₅/Cu₆Sn₅ and Cu₃Sn were characterized and compared. The results of characterization were used to evaluate different growth rates of (Cu, Ni)₆Sn₅ and Cu₃Sn under thermal aging. The thicknesses of (Cu, Ni)₆Sn₅/Cu₆Sn₅ and Cu₃Sn after different thermal aging periods were measured. High speed ball pull/shear tests were also performed. The correlation between interfacial fracture strength and IMC layer thicknesses was established.     

Cross-interaction of under-bump metallurgy and surface finish in flip-chip solder joints

The cross-interaction of the under-bump metallurgy (UBM)/solder interface and the solder/surface-finish interface in flip-chip solder joints was investigated. In this study, the UBM on the chip side was a single layer of Cu (8.5 μm), and the surface finish on the substrate side was a 0.2-μm Au layer over 5-μm Ni. It was shown that, after two reflows, the Ni layer of the surface finish had been covered with (Cu_1−xNi_x)₆Sn₅. This shows that the effect of cross-interaction of the two interfaces is important even during the reflow stage. During subsequent solid-state aging at 115°C, 135°C, and 155°C, the formation of (Cu_1−xNi_x)₆Sn₅ over the Ni layer was found to have the effect of reducing the Ni consumption rate. At the same time, the Cu consumption rate of the UBM was accelerated. The results of this study show that the selection of the UBM and the surface finish has to be considered together because the cross-interaction of the two interfaces plays an important role.     

A study on interfacial reactions between electroless Ni-P under bump metallization and 95.5Sn-4.0Ag-0.5Cu alloy

Even though electroless Ni-P and Sn-Ag-Cu solders are widely used materials in flip-chip bumping technologies, interfacial reactions of the ternary Cu-Ni-Sn system are not well understood. The growth of intermetallic compounds (IMCs) at the under bump metallization (UBM)/solder interface can affect solder-joint reliability, so analysis of IMC phases and understanding their growth kinetics are important. In this study, interfacial reactions between electroless Ni-P UBM and the 95.5Sn-4.0Ag-0.5Cu alloy were investigated, focusing on identification of IMC phases and IMC growth kinetics at various reflowing and aging temperatures and times. The stable ternary IMC initially formed at the interface after reflowing was the (Cu,Ni)₆Sn₅ phase. However, during aging, the (Cu,Ni)₆Sn₅ phase slowly changed into the quaternary IMC composed of Cu, Ni, Sn, and a small amount of Au. The Au atoms in the quaternary IMC originated from immersion Au plated on electroless Ni-P UBM. During further reflowing or aging, the (Ni,Cu)₃Sn₄ IMC started forming because of the limited Cu content in the solder. Morphology, composition, and crystal structure of each IMC were identified using transmission electron microscopy (TEM) and scanning electron microscopy (SEM). Small amounts of Cu in the solder affect the types of IMC phases and the amount of the IMC. The activation energies of (Cu,Ni)₆Sn₅ and (Ni,Cu)₃Sn₄ IMCs were used to estimate the growth kinetics of IMCs. The growth of IMCs formed in aging was very slow and temperature-dependent compared to IMCs formed in reflow because of the higher activation energies of IMCs in aging. Comparing activation energies of each IMC, growth mechanism of IMCs at electroless Ni-P/SnAgCu solder interface will be discussed.     

Drop impact reliability testing for lead-free and lead-based soldered IC packages

Board-level drop impact testing is a useful way to characterize the drop durability of the different soldered assemblies onto the printed circuit board (PCB). The characterization process is critical to the lead-free (Pb-free) solders that are replacing lead-based (Pb-based) solders. In this study, drop impact solder joint reliability for plastic ball grid array (PBGA), very-thin quad flat no-lead (VQFN) and plastic quad flat pack (PQFP) packages was investigated for Pb-based (62Sn–36Pb–2Ag) and Pb-free (Sn–4Ag–0.5Cu) soldered assemblies onto different PCB surface finishes of OSP (organic solderability preservative) and ENIG (electroless nickel immersion gold). The Pb-free solder joints on ENIG finish revealed weaker drop reliability performance than the OSP finish. The formation of the brittle intermetallic compound (IMC) Cu–Ni–Sn has led to detrimental interfacial fracture of the PBGA solder joints. For both Pb-based and Pb-free solders onto OSP coated copper pad, the formation of Cu₆Sn₅ IMC resulted in different failure sites and modes. The failures migrated to the PCB copper traces and resin layers instead. The VQFN package is the most resistant to drop impact failures due to its small size and weight. The compliant leads of the PQFP are more resistant to drop failures compared to the PBGA solder joints.     

Effect of Cross-Interaction between Ni and Cu on Growth Kinetics of Intermetallic Compounds in Ni/Sn/Cu Diffusion Couples during Aging

The solid-state, cross-interaction between the Ni layer on the component side and the Cu pad on the printed circuit board (PCB) side in ball grid array (BGA) solder joints was investigated by employing Ni(15 μm)/Sn(65 μm)/Cu ternary diffusion couples. The ternary diffusion couples were prepared by sequentially electroplating Sn and Ni on a Cu foil and were aged isothermally at 150, 180, and 200°C. The growth of the intermetallic compound (IMC) layer on the Ni side was coupled with that on the Cu side by the mass flux across the Sn layer that was caused by the difference in the Ni content between the (Cu_1−x Ni _x )₆Sn₅ layer on the Ni side and the (Cu_1−y Ni _y )₆Sn₅ layer on the Cu side. As the consequence of the coupling, the growth rate of the (Cu_1−x Ni _x )₆ Sn₅ layer on the Ni side was rapidly accelerated by decreasing Sn layer thickness and increasing aging temperature. Owing to the cross-interaction with the top Ni layer, the growth rate of the (Cu_1−y Ni _y )₆Sn₅ layer on the Cu side was accelerated at 150°C and 180°C but was retarded at 200°C, while the growth rate of the Cu₃Sn layer was always retarded. The growth kinetic model proposed in an attempt to interpret the experimental results was able to reproduce qualitatively all of the important experimental observations pertaining to the growth of the IMC layers in the Ni/Sn/Cu diffusion couple.     

Intermetallics Characterization of Lead-Free Solder Joints under Isothermal Aging

Solder interconnect reliability is influenced by environmentally imposed loads, solder material properties, and the intermetallics formed within the solder and the metal surfaces to which the solder is bonded. Several lead-free metallurgies are being used for component terminal plating, board pad plating, and solder materials. These metallurgies react together and form intermetallic compounds (IMCs) that affect the metallurgical bond strength and the reliability of solder joint connections. This study evaluates the composition and extent of intermetallic growth in solder joints of ball grid array components for several printed circuit board pad finishes and solder materials. Intermetallic growth during solid state aging at 100°C and 125°C up to 1000 h for two solder alloys, Sn-3.5Ag and Sn-3.0Ag-0.5Cu, was investigated. For Sn-3.5Ag solder, the electroless nickel immersion gold (ENIG) pad finish was found to result in the lowest IMC thickness compared to immersion tin (ImSn), immersion silver (ImAg), and organic solderability preservative (OSP). Due to the brittle nature of the IMC, a lower IMC thickness is generally preferred for optimal solder joint reliability. A lower IMC thickness may make ENIG a desirable finish for long-life applications. Activation energies of IMC growth in solid-state aging were found to be 0.54 ± 0.1 eV for ENIG, 0.91 ± 0.12 eV for ImSn, and 1.03 ± 0.1 eV for ImAg. Cu₃Sn and Cu₆Sn₅ IMCs were found between the solder and the copper pad on boards with the ImSn and ImAg pad finishes. Ternary (Cu,Ni)₆Sn₅ intermetallics were found for the ENIG pad finish on the board side. On the component side, a ternary IMC layer composed of Ni-Cu-Sn was found. Along with intermetallics, microvoids were observed at the interface between the copper pad and solder, which presents some concern if devices are subject to shock and vibration loading.     

Solid-State Reactions between Cu(Ni) Alloys and Sn

Solid-state interfacial reactions between Sn and Cu(Ni) alloys have been investigated at the temperature of 125°C. The following results were obtained. Firstly, the addition of 0.1 at.% Ni to Cu decreased the total thickness of the intermetallic compound (IMC) layer to about half of that observed in the␣binary Cu/Sn diffusion couple; the Ni addition decreased especially the thickness of Cu₃Sn. Secondly, the addition of 1 to 2.5 at.% Ni to Cu further decreased the thickness of Cu₃Sn, increased that of Cu₆Sn₅ (compared to that in the binary Cu/Sn couple) and produced significant amount of voids at the Cu/Cu₃Sn interface. Thirdly, the addition of 5 at.% Ni to Cu increased the total thickness of the IMC layer to about two times that observed in the binary Cu/Sn diffusion couple and made the Cu₃Sn disappear. Fourthly, in contrast to the previous case, the addition of 10 at.% Ni to Cu decreased the total IMC (Cu₆Sn₅) thickness again close to that of the Cu/Sn couple. With this Ni content no voids were detected. The results are rationalized with the help of␣the thermodynamics of the Sn-Cu-Ni system as well as with kinetic considerations.     

Study of Electromigration-Induced Failures on Cu Pillar Bumps Joined to OSP and ENEPIG Substrates

This work studies electromigration (EM)-induced failures on Cu pillar bumps joined to organic solderability preservative (OSP) on Cu substrates (OSP–bumps) and electroless Ni(P)/electroless Pd/immersion Au (ENEPIG) under bump metallurgy (UBM) on Cu substrates (ENEPIG–bumps). Two failure modes (Cu pad consumption and gap formation) were found with OSP–bumps, but only one failure mode (gap formation) was found with ENEPIG–bumps. The main interfacial compound layer was the Cu₆Sn₅ compound, which suffered significant EM-induced dissolution, eventually resulting in severe Cu pad consumption at the cathode side for OSP–bumps. A (Cu,Ni)₆Sn₅ layer with strong resistance to EM-induced dissolution exists at the joint interface when a nickel barrier layer is incorporated at the cathode side (Ni or ENEPIG), and these imbalanced atomic fluxes result in the voids and gap formation. OSP–bumps showed better lifetime results than ENEPIG–bumps for several current stressing conditions. The inverse Cu atomic flux (J _Cu,chem) which diffuses from the Cu pad to cathode side retards the formation of voids. The driving force for J _Cu,chem comes from the difference in chemical potential between the (Cu,Ni)₆Sn₅ and Cu₆Sn₅ phases.