Metallurgical reaction of the Sn-3.5Ag solder and Sn-37Pb solder with Ni/Cu under-bump metallization in a flip-chip package

Several international legislations recently banned the use of Pb because of environmental concerns. The eutectic Sn-Ag solder is one of the promising candidates to replace the conventional Sn-Pb solder primarily because of its excellent mechanical properties. In this study, interfacial reaction of the eutectic Sn-Ag and Sn-Pb solders with Ni/Cu under-bump metallization (UBM) was investigated with a joint assembly of solder/Ni/Cu/Ti/Si₃N₄/Si multilayer structures. After reflows, only one (Ni,Cu)₃Sn₄ intermetallic compound (IMC) with faceted and particlelike grain feature was found between the solder and Ni. The thickness and grain size of the IMC increased with reflow times. Another (Cu,Ni)₆Sn₅ IMC with a rod-type grain formed on (Ni,Cu)₃Sn₄ in the interface between the Sn-Pb solder and the Ni/Cu UBM after more than three reflow times. The thickness of the (Ni,Cu)₃Sn₄ layer formed in the Sn-Pb system remained almost identical despite the numbers of reflow; however, the amounts of (Cu,Ni)₆Sn₅ IMC increased with reflow times. Correlations between the IMC morphologies, Cu diffusion behavior, and IMC transformation in these two solder systems will be investigated with respect to the microstructural evolution between the solders and the Ni/Cu UBM. The morphologies and grain-size distributions of the (Ni,Cu)₃Sn₄ IMC formed in the initial stage of reflow are crucial for the subsequent phase transformation of the other IMC.     

Optimal phosphorous content selection for the soldering reaction of Ni-P under bump metallization with Sn-Ag-Cu solder

Nickel plating has been used as the under bump metallization (UBM) in the microelectronics industry. The electroplated Ni-P UBM with different phosphorous contents (7 wt.%, 10 wt.%, and 13 wt.%) was used to evaluate the interfacial reaction between Ni-P UBM and Sn-3Ag-0.5Cu solder paste during multiple reflow. (Cu,Ni)₆Sn₅ intermetallic compounds (IMC) formed in the SnAgCu solder/Ni-P UBM interface after the first reflow. For three times reflow, (Ni,Cu)₃Sn₄ IMC formed, while (Cu,Ni)₆Sn₅ IMC spalled into the solder matrix. With further increasing cycles of reflow, the Ni-Sn-P layer formed between (Ni,Cu)₃Sn₄ IMC and Ni-P UBM for Ni-10wt.%P and Ni-13wt.%P UBM. However, almost no Ni-Sn-P layer was revealed for the Ni-7wt.%P UBM even after ten cycles of reflow. In consideration of the wettability of Ni-P UBM, the interfacial reaction of SnAgCu/Ni-P, and dissolution of Ni-P UBM, the optimal phosphorous selection in Ni-P UBM was proposed and also discussed.     

Investigation of interfacial reaction between Sn-Ag eutectic solder and Au/Ni/Cu/Ti thin film metallization

This paper reports the formation of intermetallic compounds in Au/Ni/Cu/Ti under-bump-metallization (UBM) structure reacted with Ag-Sn eutectic solder. In this study, UBM is prepared by evaporating Au(500 Å)/Ni(1000 Å)/Cu(7500 Å) /Ti (700 Å) thin films on top of Si substrates. It is then reacted with Ag-Sn eutectic solder at 260 C for various times to induce different stages of the interfacial reaction. Microstructural examination of the interface, using both chemical and crystallographic analysis, indicates that two types of intermetallic compounds are formed during the interfacial reaction. The first phase, formed at the intial stage of the reaction, is predominantly Ni₃Sn₄. At longer times the Ni₃Sn₄ phase transforms into (Cu, Ni)₆Sn₆, probably induced by interdiffusion of Cu and Ni. At this stage, the underlying Cu layer also reacts with Sn and forms the same phase, (Cu,Ni)₆Sn₅. As a result, the fully reacted interface is found to consist of two intermetallic layers with the same phase but different morphologies.     

Shear and Pull Testing of Sn-3.0Ag-0.5Cu Solder with Ti/Ni(V)/Cu Underbump Metallization During Aging

Ti/Ni(V)/Cu underbump metallization (UBM) is widely used in flip-chip technology today. The advantages of Ti/Ni(V)/Cu UBM are a low reaction rate with solder and the lack of a magnetic effect during sputtering. Sn atoms diffuse into the Ni(V) layer to form a Sn-rich phase, the so-called Sn-patch, during reflow and aging. In this study, the relationship between interfacial reaction and mechanical properties of the solder joints with Ti/Ni(V)/Cu UBM was evaluated. Sn-3.0Ag-0.5Cu solder was reflowed on sputtered Ti/Ni(V)/Cu UBM, and then the reflowed samples were aged at 125°C and 200°C, respectively. (Cu,Ni)₆Sn₅ was formed and grew gradually at the interface of the solder joints during aging at 125°C. The Sn-patch replaced the Ni(V) layer, and (Ni,Cu)₃Sn₄ was thus formed between (Cu,Ni)₆Sn₅ and the Sn-patch at 200°C. The Sn-patch, composed of Ni and V₂Sn₃ after reflow, was transformed to V₂Sn₃ and amorphous Sn during aging. Shear and pull tests were applied to evaluate the solder joints under various heat treatments. The shear force of the solder joints remained at 421 mN, yet the pull force decreased after aging at 125°C. Both the shear and pull forces of the solder joints decreased during aging at 200°C. The effects of aging temperature on the mechanical properties of solder joint were investigated and discussed.     

Solid-state annealing behavior of two high-Pb solders, 95Pb5Sn and 90Pb10Sn,on Cu under bump metallurgy

The solid-state annealing behavior of two high-lead solders, 95Pb5Sn and 90Pb10Sn (in wt.%), was examined. After reflow, Cu₃Sn intermetallics formed on the Cu under bump metallurgy (UBM) for both solder alloys. However, solidstate annealing produced significantly different reaction morphologies for the two solder compositions. The Cu₃Sn intermetallics spalled off faster at higher temperatures in the 95Pb5Sn solder. In the case of 90Pb10Sn solder, the Cu₃Sn intermetallics continued to grow even after 1500 h at 170°C. The difference was explained by a two-step phenomenon—Sn diffusion from the bulk solder region to the solder/Cu₃Sn interface (J_Sn), and subsequent intermetallic formation (I_Cu3Sn) by interdiffusion of Cu and Sn. For 95Pb5Sn, the relation, J_Sn < I_Cu3Sn was postulated because of insufficient supply of Sn. The relation, J_Sn > I_Cu3Sn was suggested for the continuous intermetallic growth of the 90Pb10Sn solder. Although a small difference was expected between the two quantities in both solder alloys, the difference in the solid-state annealing behavior was dramatic.     

Interfacial reaction and joint reliability of fine-pitch flip-chip solder bump using stencil printing method

This study was focused on the formation and reliability evaluation of solder joints with different diameters and pitches for flip chip applications. We investigated the interfacial reaction and shear strength between two different solders (Sn-37Pb and Sn-3.0Ag-0.5Cu, in wt.%) and ENIG (Electroless Nickel Immersion Gold) UBM (Under Bump Metallurgy) during multiple reflow. Firstly, we formed the flip chip solder bumps on the Ti/Cu/ENIG metallized Si wafer using a stencil printing method. After reflow, the average solder bump diameters were about 130, 160 and 190 μm, respectively. After multiple reflows, Ni₃Sn₄ intermetallic compound (IMC) layer formed at the Sn-37Pb solder/ENIG UBM interface. On the other hand, in the case of Sn-3.0Ag-0.5Cu solder, (Cu,Ni)₆Sn₅ and (Ni,Cu)₃Sn₄ IMCs were formed at the interface. The shear force of the Pb-free Sn-3.0Ag-0.5Cu flip chip solder bump was higher than that of the conventional Sn-37Pb flip chip solder bump.     

Effects of Cu contents in Pb-free solder alloys on interfacial reactions and bump reliability of Pb-free solder bumps on electroless Ni-P under-bump metallurgy

Using the screen-printed solder-bumping technique on the electroless plated Ni-P under-bump metallurgy (UBM) is potentially a good method because of cost effectiveness. As SnAgCu Pb-free solders become popular, demands for understanding of interfacial reactions between electroless Ni-P UBMs and Cu-containing Pb-free solder bumps are increasing. It was found that typical Ni-Sn reactions between the electroless Ni-P UBM and Sn-based solders were substantially changed by adding small amounts of Cu in Sn-based Pb-free solder alloys. In Cu-containing solder bumps, the (Cu,Ni)₆Sn₅ phase formed during initial reflow, followed by (Ni,Cu)₃Sn₄ phase formation during further reflow and aging. The Sn3.5Ag solder bumps showed a much faster electroless Ni-P UBM consumption rate than Cu-containing solder bumps: Sn4.0Ag0.5Cu and Sn0.7Cu. The initial formation of the (Cu,Ni)₆Sn₅ phase in SnAgCu and SnCu solders significantly reduced the consumption of the Ni-P UBM. The more Cu-containing solder showed slower consumption rate of the Ni-P UBM than the less Cu-containing solder below 300°C heat treatments. The growth rate of the (Cu,Ni)₆Sn₅ intermetallic compound (IMC) should be determined by substitution of Ni atoms into the Cu sublattice in the solid (Cu,Ni)₆Sn₅ IMC. The Cu contents in solder alloys only affected the total amount of the (Cu,Ni)₆Sn₅ IMC. More Cu-containing solders were recommended to reduce consumption of the Ni-based UBM. In addition, bump shear strength and failure analysis were performed using bump shear test.     

Effects of substrate metallization of solder/under-bump metallization interfacial reactions in flip-chip packages during multiple reflow cycles

The effects of various elements of substrate metallization, namely, Au, Ni, and P, on the solder/under-bump metallization (UBM), (Al/Ni(V)/Cu) interfacial reactions in flip-chip packages during multiple reflow processes were systematically investigated. It was found that Au and P had negligible effects on the liquid-solid interfacial reactions. However, Ni in the substrate metallization greatly accelerated the interfacial reactions at chip side and degraded the thermal stability of the UBM through formation of a (Cu,Ni)₆Sn₅ ternary compound at the solder/UBM interface. This phenomenon can be explained in terms of enhanced grain-boundary grooving on (Cu,Ni)₆Sn₅ in the molten solder during the reflow process. This could eventually cause the rapid spalling of an intermetallic compound (IMC) from the solder/UBM interface and early failure of the packages. Our results showed that formation of multicomponent intermetallics, such as (Cu,Ni)₆Sn₅ or (Ni,Cu)₃Sn₄, at the solder/UBM interface is detrimental to the solder-joint reliability.     

The reactions between electroless Ni-Cu-P Deposit and 63Sn-37Pb flip chip solder bumps during reflow

This study investigates the interfacial reactions between electroless Ni-Cu-P deposit and 63Sn-37Pb solder bumps under various reflow conditions. The morphology of the intermetallic compounds formed at the Ni-Cu-P/Sn-Pb interface changes with respect to reflow cycle, reflow temperature, and reflow time. The (Ni,Cu)₃Sn₄ compounds with three different morphologies of fine grain, whisker, and polygonal grain form at the Ni-Cu-P/Sn-Pb interface after reflow at 220°C for 15 s. The whisker-shape and polygonal grains detach from the Ni-Cu-P deposit into the Sn-Pb solder during multiple reflows. The (Ni,Cu)₃Sn₄ compound grows rapidly when the reflow temperature is above the Ni-Sn eutectic temperature, 231°C. A continuous (Ni,Cu)₃Sn₄ layer forms after reflow at 220°C for 10 min. A 4.5 μm Ni-Cu-P deposit prevents the interdiffusion of Sn and Al atoms across the Ni-Cu-P deposit after 10 reflow cycles at 220°C for 15 s and after reflow at 220°C for 10 min.     

Interfacial reactions and shear strengths between Sn-Ag-based Pb-free solder balls and Au/EN/Cu metallization

The morphological and compositional evolutions of intermetallic compounds (IMCs) formed at three Pb-free solder/electroless Ni-P interface were investigated with respect to the solder compositions and reflow times. The three Pb-free solder alloys were Sn3.5Ag, Sn3.5Ag0.75Cu, and Sn3Ag6Bi2In (in wt.%). After reflow reaction, three distinctive layers, Ni₃Sn₄ (or Ni-Cu-Sn for Sn3.5Ag0.75Cu solder), NiSnP, and Ni₃P, were formed on the electroless Ni-P layer in all the solder alloys. For the Sn3.5Ag0.75Cu solder, with increasing reflow time, the interfacial intermetallics switched from (Cu,Ni)₆Sn₅ to (Cu,Ni)₆Sn₅+(Ni,Cu)₃Sn₄, and then to (Ni,Cu)₃Sn₄ IMCs. The degree of IMC spalling for the Sn3.5Ag0.75Cu solder joint was more than that of other solders. In the cases of the Sn3.5Ag and Sn3Ag6Bi2In solder joints, the growth rate of the Ni₃P layer was similar because these two type solder joints had a similar interfacial reaction. On the other hand, for the Sn3.5Ag0.75Cu solder, the thickness of the Ni₃P and Ni-Sn-P layers depended on the degree of IMC spalling. Also, the shear strength showed various characteristics depending on the solder alloys and reflow times. The fractures mainly occurred at the interfaces of Ni₃Sn₄/Ni-Sn-P and solder/Ni₃Sn₄.     

Effects of Ni thickness and reflow times on interfacial reactions between Ni/Cu under-bump metallization and eutectic Sn-Pb solder in flip-chip technology

Flip-chip interconnection technology plays a key role in today’s electronics packaging. Understanding the interfacial reactions between the solder and under-bump metallization (UBM) is, thus, essential. In this study, different thicknesses of electroplated Ni were used to evaluate the phase transformation between Ni/Cu under-bump metallurgy and eutectic Sn-Pb solder in the 63Sn-37Pb/Ni/Cu/Ti/Si₃N₄/Si multilayer structure for the flip-chip technology. Interfacial reaction products varied with reflow times. After the first reflow, layered (Ni_1−x,Cu_x)₃Sn₄ was found between solder and Ni. However, there were two interfacial reaction products formed between solders and the UBM after three or more times reflow. The layered (Ni_1−x,Cu_x)₃Sn₄ was next to the Ni/Cu UBM. The islandlike (Cu_1−y,Ni_y)₆Sn₅ intermetallic compound (IMC) could be related to the Ni thickness and reflow times. In addition, the influence of Cu contents on phase transformation during reflow was also studied.     

Failure Morphology after the Drop Impact Test of the Ball Grid Array Package with Lead-Free Sn-3.8Ag-0.7Cu on Cu and Ni Under-Bump Metallurgies

High strain-rate drop impact tests were performed on ball grid array (BGA) packages with lead-free Sn-3.8Ag-0.7Cu solder (in wt.%). Plated Ni and Cu under-bump metallurgies (UBMs) were used on the device side, and their drop test performances were compared. Failure occurred at the device side, exhibiting brittle interfacial fracture. For Ni UBM, failure occurred along the Ni/(Cu,Ni)₆Sn₅ interface, while the Cu UBM case showed failure along the interface between two intermetallics, Cu₆Sn₅/Cu₃Sn. However, the damage across the package varied. For Cu UBM, only a few solder balls failed at the device edge, whereas on Ni UBM, many more solder bumps failed. The difference in the failure behavior is due to the adhesion of the UBM and intermetallics rather than the intermetallic thickness. The better adhesion of Cu UBM is due to a more active soldering reaction than Ni, leading to stronger chemical bonding between intermetallics and UBM. In our reflow condition, the soldering reaction rate was about 4 times faster on Cu UBM than on Ni UBM.     

Cross-Interaction Study of Cu/Sn/Pd and Ni/Sn/Pd Sandwich Solder Joint Structures

Cross-interactions between Cu/Sn/Pd and Ni/Sn/Pd sandwich structures were investigated in this work. For the Cu/Sn/Pd case, the growth behavior and morphology of the interfacial (Pd,Cu)Sn₄ compound layer was very similar to that of the single Pd/Sn interfacial reaction. This indicates that the growth of the (Pd,Cu)Sn₄ layer at the Sn/Pd interface would not be affected by the opposite Cu/Sn interfacial reaction. We can conclude that there is no cross-interaction effect between the two interfacial reactions in the Cu/Sn/Pd sandwich structure. For the Ni/Sn/Pd case, we observed that: (1) after 300 s of reflow time, the (Pd,Ni)Sn₄ compound heterogeneously nucleated on the Ni₃Sn₄ compound layer at the Sn/Ni interface; (2) the growth of the interfacial PdSn₄ compound layer was greatly suppressed by the formation of the (Pd,Ni)Sn₄ compound at the Sn/Ni interface. We believe that this suppression of PdSn₄ growth is caused by heterogeneous nucleation of the (Pd,Ni)Sn₄ compound in the Ni₃Sn₄ compound layer, which decreases the free energy of the entire sandwich reaction system. The difference in the chemical potential of Pd in the PdSn₄ phase at the Pd/Sn interface and in the (Pd,Ni)Sn₄ phase at the Sn/Ni interface is the driving force for the Pd atomic flux across the molten Sn. The diffusion of Ni into the ternary (Pd,Ni)Sn₄ compound layer controls the Pd atomic flux across the molten Sn and the growth of the ternary (Pd,Ni)Sn₄ compound at the Sn/Ni interface.     

Characterizing metallurgical reaction of Sn3.0Ag0.5Cu composite solder by mechanical alloying with electroless Ni-P/Cu under-bump metallization after various reflow cycles

Electroless Ni-P/Cu under-bump metallization (UBM) is widely used in electronics packaging. The Sn3.0Ag0.5Cu lead-free composite solder pastes were produced by a mechanical alloying (MA) process doped with Cu₆Sn₅ nanoparticles. In this study, the detailed interfacial reaction of Sn3.0Ag0.5Cu composite solders with EN(P)/Cu UBM was investigated after reflow. A field-emission scanning electron microscope (FESEM) was employed to analyze the interfacial morphology and microstructure evolution. The intermetallic compounds (IMCs) formed at the interface between the Sn3.0Ag0.5Cu composite solders and EN(P)/Cu UBM after one and three reflows were mainly (Ni_1−x,Cu_x)₃Sn₄ and (Cu_1−y,Ni_y)₆Sn₅. However, only (Ni_1−x,Cu_x)₃Sn₄ IMC was observed after five reflows. The elemental distribution near the interfacial region was evaluated by an electron probe microanalyzer (EPMA) as well as field-emission electron probe microanalyzer (FE-EPMA). Based on the observation and characterization by FESEM, a EPMA, and an FE-EPMA, the reaction mechanism of interfacial phase transformation between Sn3.0Ag0.5Cu composite solders and EN(P)/Cu UBM after various reflow cycles was discussed and proposed.     

Intermetallic reactions in Sn3Ag0.5Cu and Sn3Ag0.5Cu0.06Ni0.01Ge solder BGA packages with Au/Ni surface finishes

The intermetallic compounds (IMCs) formed during the reflow and aging of Sn3Ag0.5Cu and Sn3Ag0.5Cu0.06Ni0.01Ge solder BGA packages with Au/Ni surface finishes were investigated. After reflow, the thickness of (Cu, Ni, Au)₆Sn₅ interfacial IMCs in Sn3Ag0.5Cu0.06Ni0.01Ge was similar to that in the Sn3Ag0.5Cu specimen. The interiors of the solder balls in both packages contained Ag₃Sn precipitates and brick-shaped AuSn₄ IMCs. After aging at 150°C, the growth thickness of the interfacial (Ni, Cu, Au)₃Sn₄ intermetallic layers and the consumption of the Ni surface-finished layer on Cu the pads in Sn3Ag0.5Cu0.06Ni0.01Ge solder joints were both slightly less than those in Sn3Ag0.5Cu. In addition, a coarsening phenomenon for AuSn₄ IMCs could be observed in the solder matrix of Sn3Ag0.5Cu, yet this phenomenon did not occur in the case of Sn3Ag0.5Cu0.06Ni0.01Ge. Ball shear tests revealed that the reflowed Sn3Ag0.5Cu0.06Ni0.01Ge packages possessed bonding strengths similar to those of the Sn3Ag0.5Cu. However, aging treatment caused the ball shear strength in the Sn3Ag0.5Cu packages to degrade more than that in the Sn3Ag0.5Cu0.06Ni0.01Ge packages.     

Comparison of electroplated eutectic Bi/Sn and Pb/Sn solder bumpson various UBM systems

The effect of a reflow process and under bump metallurgy (UBM) systems on the growth of intermetallic compounds (IMC) of the 57Bi/43Sn and 37Pb/63Sn solder bump/UBM interfaces was investigated. The selected UBM systems were sputtered Al/Ti/Cu, sputtered Al/NiV/Cu, Al/electroless Ni/immersion Au, and Al/Ti/electroless Cu. An alloy electroplating method was used for the solder bumping process. The microstructure and composition of intermetallic compound (IMC) phases and their morphologies were examined using scanning electron microscopy and X-ray diffraction. The Cu₆Sn₅ η'-phase IMC appeared on all Cu containing UBM cases with Pb/Sn and Bi/Sn solders and the Cu ₃Sn ϵ-phase was detected only with Pb/Sn solder bumps. The Ni₃Sn₄ IMC was found to be the main IMC phase between Ni and solder. The Ni₃Sn secondary IMC was also detected on the electroless Ni UBM with PbSn solder after ten times reflow. Through the bump shear test, Al/NiV/Cu, Al/elNi/Au, and Al/Ti/elCu UBMs showed good stability with Bi/Sn and Pb/Sn solder in terms of metallurgical aspects     

Interfacial reactions and compound formation of Sn-Ag-Cu solders by mechanical alloying on electroless Ni-P/Cu under bump metallization

Electroless Ni-P under bump metallization (UBM) has been widely used in electronic interconnections due to the good diffusion barrier between Cu and solder. In this study, the mechanical alloying (MA) process was applied to produce the SnAgCu lead-free solder pastes. Solder joints after annealing at 240°C for 15 min were employed to investigate the evolution of interfacial reaction between electroless Ni-P/Cu UBM and SnAgCu solder with various Cu concentrations ranging from 0.2 to 1.0 wt.%. After detailed quantitative analysis with an electron probe microanalyzer, the effect of Cu content on the formation of intermetallic compounds (IMCs) at SnAgCu solder/electroless Ni-P interface was evaluated. When the Cu concentration in the solder was 0.2 wt.%, only one (Ni, Cu)₃Sn₄ layer was observed at the solder/electroless Ni-P interface. As the Cu content increased to 0.5 wt.%, (Cu, Ni)₆Sn₅ formed along with (Ni, Cu)₃Sn₄. However, only one (Cu, Ni)₆Sn₅ layer was revealed, if the Cu content was up to 1 wt.%. With the aid of microstructure evolution, quantitative analysis, and elemental distribution by x-ray color mapping, the presence of the Ni-Sn-P phase and P-rich layer was evidenced.     

Studies of electroless nickel under bump metallurgy—Solder interfacial reactions and their effects on flip chip solder joint reliability

The electroless-deposited Ni-P under bump metallurgy (UBM) layer was fabricated on Al pads for Sn containing solder bumps. The amount of P in the electroless Ni film was optimized by controlling complexing agents and the pH of plating solution. The interfacial reaction at the electroless Ni UBM/solder interface was investigated in this study. The intermetallic compound (IMC) formed at the interface during solder reflowing was mainly Ni₃Sn₄, and a P-rich Ni layer was also formed as a by-product of Ni-Sn reaction between the Ni-Sn IMC and the electroless Ni layer. One to four microns of Ni₃Sn₄ IMC and a 1800–5000 Å of P-rich Ni layer were formed in less than 10 min of solder reflowing depending on solder materials and reflow temperatures. It was found that the P-rich Ni layer contains Ni, P, and a small amount of Sn (～7 at.%). Further cross-sectional transmission electron microscopy (TEM) analysis confirmed that the composition of the P-rich Ni layer was 75 at.% Ni, 20at.%P, and 5at.%Sn by energy-dispersive x-ray spectroscopy (EDS) and the phase transformation occurred in the P-rich Ni layer by observing grain size. Kirkendall voids were also found in the Ni₃Sn₄ IMC, just above the P-rich Ni layer after extensive solder reflow. The Kirkendall voids are considered a primary cause of the brittle fracture; restriction of the growth of of the P-rich Ni layer by optimizing proper processing conditions is recommended. The growth kinetics of Ni-Sn IMC and P-rich Ni layer follows three steps: a rapid initial growth during the first 1 min of solder reflow, followed by a reduced growth step, and finally a diffusion-controlled growth. During the diffusion-controlled growth, there was a linear dependence between the layer thickness and time^1/2. Flip chip bump shear testing was performed to measure the effects of the IMC and the P-rich Ni layers on bump adhesion property. Most failures occurred in the solder and at the Ni₃Sn₄ IMC. The brittle characteristics of the Ni-Sn IMC and the Kirkendall voids at the electroless Ni UBM-Sn containing solder system cause brittle bump failure, which results in a decreased bump adhesion strength.     

Reactive interdiffusion between a lead-free solder and Ti/Ni/Ag thin-film metallizations

The reactive interdiffusion between a Sn-3.0wt.%Ag-0.7wt.%Cu solder and thin-film Ti/Ni/Ag metallizations on two semiconductor devices, a diode and a metal-oxide-semiconductor field-effect transistor (MOSFET), and a Au-layer on the substrates are studied. Comprehensive microanalytical techniques, scanning electron microscopy, transmission electron microscopy (TEM), and analytical electron microscopy (AEM) are employed to identify the interdiffusion processes during fabrication and service of the devices. During the reflow process of both diode and MOSFET devices, (1) the Ag layer dissolves in the liquid solder; (2) two intermetallics, (Ni,Cu)₃Sn₄ and (Cu,Ni)₆Sn₅, form near the back metal/solder interface; and (3) the Au metallization in the substrate side dissolves in the liquid solder, resulting in precipitation of the (Au,Ni,Cu)Sn₄ intermetallic during solidification. During solid-state aging of both diode and MOSFET solder joints at 125°C and 200°C, the following atomic transport processes occur: (1) interdiffusion of Cu, Ni, and Sn, leading to the growth of a (Ni,Cu)₃Sn₄ layer until the Ni layer is completely consumed; (2) interdiffusion of Au, Cu, Ni, and Sn through the (Ni,Cu)₃Sn₄ layer and unconsumed Ni layer to the Ti layer to form a solid solution; and (3) further interdiffusion of Au, Cu, Ni, and Sn through the (Ni,Cu)₃Sn₄ layer to from an (Au,Ti,Ni,Cu)Sn₄ layer. The growth of the latter layer continues until the entire Ti layer is consumed.     

Tensile behavior of pb-sn solder/cu joints

Solders of nominal 95Pb-5Sn and 60Sn-40Pb were used to join Cu plates. The effect of ternary additions of In, Ag, Sb, and Bi to the near-eutectic solder were also investigated. Bulk solder and interfacial joint microstructures were characterized for each solder alloy. The solder joints were strained to failure in tension; joint strength and failure mode were determined. 95Pb-5Sn/Cu and 60Sn-40Pb/Cu specimens were tested both as-processed and after reflow. 95Pb-5Sn/Cu as-processed and reflow specimens failed in tension in a ductile mode. Voids initiated at β-Sn precipitates in the as-processed specimens and at the Cu₃Sn intermetallic in the reflow specimens. 60Sn-40Pb/Cu failed transgranularly through the Cu₆Sn₅ intermetallic in both the as-processed and reflow conditions. The joint tensile strength of the reflow specimens was approximately half that of the as-processed specimens for both the high-Pb and near-eutectic alloys. The Cu₆Sn{5} intermetallic dominated the tensile failure mode of the near-eutectic solder/Cu joints. The fracture path of the near-eutectic alloys with ternary additions depended on the presence of Cu₆Sn₅ rods in the solder within the Cu plates. Specimens with ternary additions of In and Ag contained only interfacial intermetallics and exhibited interfacial failure at the Cu₆Sn₅. Joints manufactured with ternary additions of Sb and Bi contained rods of Cu₆Sn₅ within the solder. Tensile failure of the Sb and Bi specimens occurred through the solder at the Cu₆Sn₅ rods.