Ag Solder Balls Research Articles

Microstructure evolution of Au/SnSb-CuNiAg/(Au)Ni during high temperature aging

PurposeThe purpose of this paper is to investigate the morphology evolution and the composition transformation of Au-Sn intermetallic compounds (IMCs) of the new Au/Sn-5Sb-1Cu-0.1Ni-0.1Ag/(Au)Ni solder joint during the high temperature aging.Design/methodology/approachSn-5Sb-1Cu-0.1Ni-0.1Ag solder balls (500 µm in diameter), heat sink with structure of 7.4 µm Au layer on 5 µm Ni-plated Cu alloy and Si chip with 5.16 µm plated Au were used to fabricate micro-solder joints. The joints were performed in a furnace at 150°C for 150, 250 and 350 h aging. The samples were polished and deep etched before analyzed by metallographic microscope and scanning electron microscopy, respectively. Energy dispersive x-ray spectroscopy was used to identify the composition of the IMCs.Findingsß-(Au,Ni,Cu)10Sn phase is formed during the soldering process. The IMCs evolution has two periods during the aging. The first is theξ-(Au,Ni,Cu)5Sn,ξ-(Au,Cu)5Sn andδ-AuSn were formed and grew to form a full-compound joint after about 150 h aging. The second is the conversion of the full-compound joint. The IMCs converted toξ′ phase when the aging time extends to 250 h, and transformed toε-(Au,Ni,Cu)Sn2andη-(Au,Ni,Cu)Sn4after 350 h aging. The thicker gold layer and thinner solder joint can promote the growth of the IMCs.ß-(Au,Ni,Cu)10Sn emerged in Au/SnSb-CuNiAg/(Au)Ni in this research, which is not usually found.Originality/valueThe results in this study have a significant meaning for the application of the new Sn-5Sb-1Cu-0.1Ni-0.1Ag in harsh conditions.

Size effect model on kinetics of interfacial reaction between Sn-xAg-yCu solders and Cu substrate.

The downsizing of solder balls results in larger interfacial intermetallic compound (IMC) grains and less Cu substrate consumption in lead-free soldering on Cu substrates. This size effect on the interfacial reaction is experimentally demonstrated and theoretically analyzed using Sn-3.0Ag-0.5Cu and Sn-3.5Ag solder balls. The interfacial reaction between the Sn-xAg-yCu solders and Cu substrates is a dynamic response to a combination of effects of interfacial IMC growth, Cu substrate consumption and composition variation in the interface zone. A concentration gradient controlled (CGC) kinetics model is proposed to explain the combined effects. The concentration gradient of Cu at the interface, which is a function of solder volume, initial Cu concentration and reaction time, is the root cause of the size effect. We found that a larger Cu concentration gradient results in smaller Cu6Sn5 grains and more consumption of Cu substrate. According to our model, the growth kinetics of interfacial Cu6Sn5 obeys a t1/3 law when the molten solder has approached the solution saturation, and will be slower otherwise due to the interfering dissolution mechanism. The size effect introduced in this model is supported by a good agreement between theoretical and experimental results. Finally, the scope of application of this model is discussed.

Fabrication and interfacial reaction of carbon nanotube-embedded Sn–3.5Ag solder balls for ball grid arrays

Abstract Carbon nanotubes (CNTs) that incorporated Sn–3.5Ag solder balls were fabricated by a surface impact mixing (SIM) process. The SIM process consisted of a ball milling process that provided enough impact energy for CNT embedding on the solder surface and at the same time, maintained the size and shape of the solder balls. As the SIM time increased, the density of the embedded CNTs on the solder surface increased and the entire solder surface was covered with CNTs after 24 h of the SIM process. The shear strength of the CNT-embedded Sn-3.5Ag solder balls increased by 10% after the 24 h-SIM process. The thickness of the intermetallic compound (IMC) layer between the solder and Ni decreased when CNTs were included in the Sn–3.5Ag solder. As aging time increased, the IMC thickness difference between CNT-embedded and non-CNT samples remarkably increased as well.

The effects of solder deformation on the wetting characteristics and interfacial reaction of Sn–3.5Ag solders on Cu substrates in fluxless soldering

The effects of solder deformation on the wetting characteristics during fluxless soldering were studied when deformed Sn–3.5Ag solder balls were reacted with Cu or oxidized Cu substrates. The Cu surfaces were oxidized at 100 °C for 2 or 4 h in air. After the 760 μm diameter solder balls were deformed on the substrates under 0–30 N, they were then reflowed at 300 °C for 30 s without flux. An optical microscope and a scanning electron microscope equipped with energy dispersive spectroscopy were used to measure the wetting angles and to characterize interfacial microstructures. As solder deformation increased, the wetting angle of solder bumps on the Cu or oxidized Cu substrates decreased and the spreading area increased. The oxide layer on the Cu surface decreased the wettability of the solders. Intermetallic compound (IMC) growth was suppressed in the solder interface when the solder reacted with oxidized Cu, while the IMC thickness increased with solder deformation. Solder deformation exposed a fresh Sn surface and improved contact between the solder and Cu substrate, thereby increasing the wettability of the solders.

Finite element based fatigue life estimation of the solder joints with effect of intermetallic compound growth

This paper develops an analysis procedure to study the effects of intermetallic compound (IMC) growth on the fatigue life of 63Sn–37Pb (lead-rich)/96.5Sn–3.5Ag (lead-free) solder balls for flip-chip plastic ball grid array packages under thermal cycling test conditions. In this analysis procedure, the thickness of the IMC increased with the number of thermal cycles, and was determined using the growth rate equation. A series of non-linear finite element analyses was conducted to simulate the stress/strain history at the critical locations of the solder balls with various IMC thicknesses in thermal cycling tests. The simulated stress/strain results were then employed in a fatigue life prediction model to determine the relationship between the predicted fatigue life of the solder ball and the IMC thickness. Based on the concept of continuous damage accumulation and incorporated with the linear damage rule, this study defines the damage of each thermal cycle as the reciprocal of the predicted fatigue life of the solder joints with the corresponding IMC thickness. The final fatigue failure of the solder ball was determined as the number of cycles corresponding to the cumulative damage equal to unity. Results show that the solder joint fatigue life decreased as the IMC thickness increased. Moreover, the predicted thermal fatigue life of lead-rich solders based on the effects of IMC growth is apparently smaller than that without considering the IMC growth in the reliability analysis. Results also show that the influence of the IMC thickness on the fatigue life prediction of the lead-free solder joint can be ignored.

Effects of Joining Sequence on the Interfacial Reactions and Substrate Dissolution Behaviors in Ni/Solder/Cu Joints

The effects of the joining sequence on the interfacial reactions and substrate dissolution behaviors in Ni/solder/Cu joints were studied by using 500-μm (diameter) Sn-3.5Ag solder balls and substrates with a 375-μm (diameter) opening. Three distinct paths for the joining sequence were studied. In path I, a solder ball was first joined to the Cu substrate and then to the Ni substrate. In path II, a solder ball was joined to both the Cu and Ni substrates simultaneously. Path III had the opposite joining sequence to path I. The results of this study indicated that (Cu,Ni)6Sn5 was the predominant reaction product at both the Ni/solder and solder/Cu interfaces regardless of the joining sequence. However, the composition, morphology, and thickness of the (Cu,Ni)6Sn5 varied considerably with the different paths. The dissolution behaviors of Cu and Ni were also different. Dybkov’s dissolution kinetics and Cu-Ni-Sn isotherm data were utilized to rationalize these differences.

Dynamic Filling Characteristics of a Capillary Driven Underfill Process in Flip-Chip Packaging

This study investigates the dynamic flow characteristics of capillary-driven underfill flows in a flip chip package. In the present study, we used two different bump arrays using Sn-2.5Ag solder balls with 80 μm and 100 μm diameters on commercially available flip chips, which have different pitches of 150 μm and 180 μm. First, we measured surface tension and viscosity with a rheometer and a tensiometer, respectively, and conducted an experimental visualization of the dynamic filling behavior of the underfill flows. From the captured images, we estimated the filling times, which can be affected by two important factors: bump arrangements and resin viscosities. In addition, we conducted a FVM (finite volume method)-based numerical simulation using commercial CFD code (Fluent v. 6.3.26), and compared its numerical results to both the experimental data and the analytical solutions given by the previous model described by Wan et al. (2005). The numerical predictions and analytical solutions estimating filling time were in good agreement with the experimental data, and the increase in spatial density of solder bumps allowed the flow to fill more slowly due to the increase in flow resistance. We conclude that the non-Newtonian characteristics and bump arrangement are very important factors in the design of flip-chip packaging.

Secondary IMC formation induced by Kirkendall voiding in Cu/Sn–3.5Ag solder joints

In this investigation on the formation of multiple-layered Kirkendall voids at Cu/Sn–3.5Ag solder joints, Sn–3.5Ag solder balls were reacted with Cu under bump metallurgy (UBM), which was electroplated using bis-sodium sulfopropyl–disulfide, C6H12O6S4Na2 (SPS) additive. The sequence of multilayer Kirkendall voids and Cu–Sn IMC (intermetallic compounds) formations are explained with the aid of cross-sectional scanning electron microscopy (SEM) micrographs and schematic diagrams. During the aging treatment at 150 °C, layers of Cu6Sn5/Cu3Sn formed at the solder joints and Kirkendall voids nucleated at the Cu3Sn/Cu interface as a result of the segregation of residual S originating from SPS. However, with Kirkendall void growth, the net section area of the Cu/Cu3Sn interface decreased and the Cu flux into Cu3Sn was inhibited. As the atomic ratio of Cu against Sn in the Cu3Sn dropped, transformation of Cu3Sn into Cu6Sn5 ensued. Subsequent diffusion of Sn atoms into the remaining Cu UBM through the remaining ligament of the Cu6Sn5/Cu interface precipitated secondary Cu3Sn beneath the primary Cu3Sn/Cu interface, and the secondary Kirkendall voids formed at the new Cu3Sn/Cu interface and so on.

Activities during melting and reflowing behaviour of solders

PurposeThe purpose of this paper is to visualise the activities of three solders; Sn‐37Pb, Sn‐9Zn and Sn‐3.5Ag on Cu substrates during reflow near their melting points and to relate them with reflow reactions between solder and substrate.Design/methodology/approachMelting activities of three solders near their melting points on copper substrates are visualised in an infrared reflow furnace.FindingsSolder balls demonstrate different ways of melting and reflowing behaviours in dissimilar times and temperature intervals. Melting of Sn‐9Zn solder balls is initiated simultaneously at the surface and joint between solder balls. This is followed by the melting at the joint between solder balls and the Cu substrate. During melting, solder balls are first merged into each other and then reflow on the substrate from top to bottom. Opposite to Sn‐9Zn, Sn‐3.5Ag solder balls start to melt at the surface and the joint between the solder and substrate, simultaneously. Balls are first reflowed from top to bottom and, in the process, liquid solder is merged. Unlike Sn‐9Zn and Sn‐3.5 Ag, melting of Sn‐37Pb solder balls is initially commenced at the surface only. This is followed by simultaneous melting at both joints. Variation in melting activities of these solders is found to be closely related to the coalescence mechanism of solder balls and the reflow reactions between the solders and the Cu substrate.Originality/valueThe elementary melting activities of different solders on Cu substrates is related with their reflow behaviours. This provides better understanding of solder behaviour and selection of good lead‐free solder for applications in the electronic industry.

Intermetallic Compounds Formed in Sn-20In-2.8Ag Solder BGA Packages with Ag/Cu Pads

The interfacial reactions in a Sn-20In-2.8Ag solder ball grid array (BGA) package with immersion Ag surface finish are investigated. After reflow, the Ag thin film dissolves quickly into the solder matrix, and scallop-shaped intermetallic layers, with compositions of (Cu0.98Ag0.02)6(In0.59Sn0.41)5, appear at the interfaces between Sn-20In-2.8Ag solder ball and Cu pad. No evident growth of the (Cu0.98Ag0.02)6(Sn0.59In0.41)5 intermetallic compounds was observed after prolonged aging at 100 °C. However, the growth accelerated at 150 °C, with more intermetallic scallops floating into the solder matrix. The intermetallic thickness versus the square root of reaction time (t 1/2) shows a linear relation, indicating that the growth of intermetallic compounds is diffusion-controlled. Ball shear tests show that the strength of Sn-20In-2.8Ag solder joints after reflow is 4.4 N, which increases to 5.18 N and 5.14 N after aging at 100 and 150 °C, respectively.

Effect of Laser Parameter on the Bond Characteristics of Sn-3.5%Ag Solder Ball

Pulsed Nd;YAG laser was applied to investigate the bond characteristics of Sn- 3.5wt%Ag (hereafter, Sn-3.5Ag) solder ball. The ball diameter is 500µm, and the UBM (Under Bump Metallurgy) on a FR4-PCB consisted of Cu/Ni/Au from bottom to top with thickness of 50µm/25µm/0.05µm, respectively. A RMA (rosin mildly activated) flux for BGA (Ball Grid Array) was applied before laser soldering to set up the solder ball on the UBM. Laser power was 250W, and its beam diameter at the focal point was around 250µm. Experimental results reveal that the solder ball showed a good bonded state when the laser energy was in the range between 0.8 and 1.0 J. Laser energy lower than 0.8J caused non-wetting of the solder ball, and energy higher than 1.2J showed crack along the UBM. Microstructure of the laser soldered Sn-3.5Ag was examined with the number of pulse, when the frequency, pulse width and laser energy were fixed at 5 Hz, 20 ms and 0.8 J, respectively. The IMC (Intermetallic Compound) of Ag3Sn in the solder changed from acicular to rod type with increasing laser pulses. Ni3Sn4 was produced between the solder and the UBM, and its thickness increased from 0.65 to 1.80µm with increasing number of pulse from 2 to 4. These are thinner than those of general air reflow. Shear strength of the laser soldered ball with output energy of 1 J showed 652.3 gf, and it was similar value compared to the result from hot plate soldering, 654.5 gf.

Intermetallic reactions in Sn-3.5Ag solder ball grid array packages with Ag/Cu and Au/Ni/Cu pads

During the reflow process of Sn-3.5Ag solder ball grid array (BGA) packages with Ag/Cu and Au/Ni/Cu pads, Ag and Au thin films dissolve rapidly into the liquid solder, and the Cu and Ni layers react with the Sn-3.5Ag solder to form Cu6Sn5 and Ni3Sn4 intermetallic compounds at the solder/pad interfaces, respectively. The Cu6Sn5 intermetallic compounds also appear as clusters in the solder matrix of Ag surface-finished packages accompanied by Ag3Sn dispersions. In the solder matrix of Au/Ni surface-finished specimens, Ag3Sn and AuSn4 intermetallics can be observed, and their coarsening coincides progressively with the aging process. The interfacial Cu6Sn5 and Ni3Sn4 intermetallic layers grow by a diffusion-controlled mechanism after aging at 100 and 150°C. Ball shear strengths of the reflowed Sn-3.5Ag packages with both surface finishes are similar, displaying the same degradation tendencies as a result of the aging effect.

Test Method Development to Quantify the In Situ Elastic and Plastic Behavior of 62%Sn–36%Pb–2%Ag Solder Ball Arrays in Commercial Area Array Packages at −40 °C , 23 °C, and 125 °C

The objective of this study is to describe and evaluate test methods developed to experimentally characterize the in situ mechanical behavior of solder ball arrays connecting printed wiring boards to area array packages under tensile, compressive, and shear loading at −40, 23, and 125 °C. The solder ball arrays tested were composed of 62%Sn–36%Pb–2%Ag solder alloy. Finite element modeling was performed. The results indicated that the test fixture should be geometrically equivalent to the projected shape of the ball grid array to achieve uniform loading. Tension, compression, and shear tests were conducted. For tensile loading the interfaces and the solder balls are loaded in series resulting in a large apparent strain (13%). Various interfacial failure modes are observed. Under compression and shear loading the effect of the interfaces are negligible and therefore a significant deformation and a remarkable yielding behavior of solder ball arrays can be observed. Furthermore, the specimens tested under shear loading showed different failure modes such as cohesive or adhesive failure modes depending on the test temperature. From the overall results, it has been determined that shear loading is the most representative test to measure the actual mechanical behavior of solder in ball grid arrays.

Plasma Reflow Bumping of Sn-3.5 Ag Solder for Flux-Free Flip Chip Package Application

A new flux-free reflow process using Ar+10%H/sub 2/ plasma was investigated for application to solder bump flip chip packaging. The 100-/spl mu/m diameter Sn-3.5wt%Ag solder balls were bonded to 250-/spl mu/m pitch Cu/Ni under bump metallurgy (UBM) pattern by laser solder ball bonding method. Then, the Sn-Ag solder balls were reflowed in Ar+H/sub 2/ plasma. Without flux, the wetting between solder and UBM occurred in Ar+H/sub 2/ plasma. During plasma reflow, the solder bump reshaped and the crater on the top of bump disappeared. The bump shear strength increased as the Ni/sub 3/Sn/sub 4/ intermetallic compounds formed in the initial reflow stage but began to decrease as coarse (Cu,Ni)/sub 6/Sn/sub 5/ grew at the solder/UBM interface. As the plasma reflow time increased, the fracture mode changed from ductile fracture within the solder to brittle fracture at the solder/UBM interface. The off-centered bumps self-aligned to patterned UBM pad during plasma reflow. The micro-solder ball defects occurred at high power prolonged plasma reflow.

Sn-Pb系はんだボール接合体の組織と強度に及ぼすAg添加とリフロー冷却速度の影響

Sn-37 mass%Pb and Sn-36 mass%Pb-2 mass%Ag solder balls have been bonded on Cu pads at the cooling rates of reflow. The microstructure and the mechanical properties of the ball bonding have been investigated. The eutectic lamellar structure (Sn phase/Pb phase) of both ball was finer. The lamellar spacing of Sn-36 mass%Pb-2 mass%Ag solder ball was much smaller than that of Sn-37 mass%Pb one at the faster cooling rates. The lamellar structure became finer since Ag may interrupt the diffusion of Sn and Pb. The ball hardness depended on the cooling rate and increased with the smaller lamellar spacing at the faster cooling rate. Since the eutectic Ag3Sn intermetallic compound formed in Sn-36 mass%Pb-2 mass%Ag solder ball, the hardness of this alloy ball was higher than that of Sn-37 mass%Pb one. The shear strength of both ball bonding dropped remarkably at 10 K/min since Ni3Sn4 reaction layer at the bonding interface was thicker. The shear strength of Sn-36 mass%Pb-2 mass%Ag ball was lower than that of Sn-37 mass%Pb one at each cooling rate. The primary needle-shape Ag3Sn was observed in Sn-36 mass%Pb-2 mass%Ag ball and near the bonding interface. The lower shear strength of Sn-36 mass%Pb-2 mass%Ag ball bonding was due to the primary needle-shape Ag3Sn near the bonding interface and the brittle Ni3Sn4 reaction layer. Since the fracture face of the Sn-36 mass%Pb-2 mass%Ag ball was flatter than that of the Sn-37 mass%Pb ball, the brittle fracture occurred at the Sn-36 mass%Pb-2 mass%Ag ball bonding interface.

Mechanical Strength and Microstructure of BGA Joints Using Lead-Free Solders

Au–Ni plated Cu pads were reflow soldered by using lead free solder balls. The microstructure and strength of the as-reflowed solder joints were investigated. For solder joints using Cu-free Sn–Pb and Sn–Ag solder balls, a Ni3Sn4 reaction layer was formed on the boundary between solder and pads. On the other hand, a Cu–Sn based (Cu, Ni)6Sn5 reaction layer (η′) was formed in solder joints using Cu-containing solder balls. The growth rate for an η′ reaction layer during heat exposure at 423 K was much slower than that for a Ni3Sn4 reaction layer. This suppression of an η′ reaction layer growth can be attributed to the fact that the Cu in solder balls was mostly removed during the formation of the η′ layer. By ball shear test, cold bump pull and hot bump pull tests, mechanical properties of the obtained BGA joints were investigated. Fracture loads and crack propagation path changed by changing the mechanical tests, the BGA joints using Cu containing Sn–Ag–Cu solder or low P type Ni plating revealed better mechanical properties. We established the mismatch of the boundaries between reaction layers and the P-enriched Ni–P layer, which was caused by the chained voids formed due to the Kirkendall effect, led to low joint strength.

Melting and Joining Behavior of Sn/Ag and Sn-Ag/Sn-Bi Plating on Cu Core Ball

We fabricated Cu core Sn–Ag solder balls by plating pure Sn and Ag on Cu balls and clarified that Sn/Ag plating began to melt at a rather low temperature, the eutectic temperature of Sn–Ag–Cu. This early melting at the eutectic temperature was ascribed to the diffusion of Cu and Ag into the Sn plating during the heating process. We investigated the solderability of the BGA joint with the Ni/Au coated Cu pad to compare it with that of the commercial Sn–Ag and Sn–Ag–Cu balls. After reflow soldering, we observed a eutectic microstructure composed of β-Sn, Ag3Sn, and Cu6Sn5 phases in the solder, and a η′-(Au, Cu, Ni)6Sn5 reaction layer was formed at the interface between the solder and the Cu pad. The BGA joint using Cu core solder balls could prevent the degradation of joint strength during aging at 423 K because of the slower growth rate of the η′-(Au, Cu, Ni)6Sn5 reaction layer formed at the solder-pad interface. Furthermore, we were able to fabricate Cu-cored, multicomponent Sn–Ag–Bi balls by sequentially coating binary Sn–Ag and Sn–Bi solders onto Cu balls. The coated balls also exhibited almost the same melting and soldering behaviors as those of the previously alloyed Sn–2Ag–0.75Cu–3Bi solders.

Effect of Cu in Pb Free Solder Ball on the Microstructure of BGA Joints with Au/Ni Coated Cu Pads

Effect of Cu in the eutectic Sn–37 mass%Pb or Sn–3.5 mass%Ag solder ball has been investigated on the microstructure and the shear strength of BGA joints with Ni/Au electroplated Cu pad after reflow and the subsequent heat exposure at 423 K . For joints using both Cu free Sn–Pb and Sn–Ag solder balls, a metastable (Au, Ni)Sn3 reaction layer was formed on the interface between the solder and the pad after reflow soldering and the subsequent heat storage at 423 K, while stable η′-(Au, Cu, Ni)6Sn5 reaction layer was formed for the joints using Cu containing solder balls. The growth rate during heat exposure was much slower for η′ reaction layer than that for (Au, Ni)Sn3 reaction layer. All the joints after a heat exposure fractured on the lower shear load than as reflowed joints, but the shear strength of the joint using Cu containing ball was less degraded than that using Cu free solder balls, because η′ reaction layer grows slower and plays a roll as a good barrier layer to suppress the Ni–Sn interaction. Almost the same effect was confirmed for the BGA joints using Cu cored solder balls, since Cu dissolves into the solder during reflow soldering.