No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Microstructural variation and high-speed impact responses of Sn-3.0Ag-0.5Cu/ENEPIG solder joints with ultra-thin Ni-P deposit
Abstract
Electroless Ni-P/electroless Pd/immersion Au (ENEPIG) with ultra-thin Ni-P deposit serve as a potential replacement of traditional ENEPIG surface finish because of its superior electrical performance in flip chip solder joints interconnection. However, the interfacial reaction and mechanical reliability of solder joints in ENEPIG with ultra-thin Ni-P layer is not yet well evaluated. In this study, we investigated the characteristic microstructure of interfacial intermetallic compounds and high-speed impact responses of Sn-3.0Ag-0.5Cu/ENEPIG attachments with 4.8, 0.3, and 0.05 mu m Ni-P deposit. ENEPIG with Ni-P layer of 0.3 mu m exhibited the eutectic structure dispreading in the solder alloys and layer-type P-rich IMCs at solder/metallization interface, while there was (Cu,Ni)(6)Sn-5 precipitation in the solder but no P-rich IMCs layer formed in ENEPIG with 0.05 mu m Ni-P layer. Slower interfacial reaction rate in ENEPIG with 0.3 mu m Ni-P layer was attributed to the effect of electroless Ni-P diffusion barrier layer, which would further provide better impact resistivity than that of ENEPIG with 0.05 mu m Ni-P deposit. Moreover, breach in P-rich IMCs and underneath (Cu,Ni)(6)Sn-5 patch were observed in ENEPIG with 0.3 mu m Ni-P layer. The growth mechanism was closely related to the Ni diffusion from surface finish and element redistribution.
... Some of the solder joints created in the 1st reflow might further be involved in the subsequent reflows, and 3-5 reflow cycles are generally experienced for the advanced packaging process. The solderability of the ultrathin-Ni(P)-type ENEPIG after multiple reflows followed by solid-state aging had been investigated in the literature [6][7][8][9][10][11]. The previous studies showed that an ultrathin Ni(P) film can be completely depleted in soldering reaction, and the Ni(P) depletion depended on the reaction time (or reflow number), the Ni(P) thickness [6,7], and the Pd(P) thickness [9]. ...
... The depletion of the Ni(P) (or Ni 3 resulted in a series of nanovoids between (Cu,Ni) 6 Sn 5 (high Ni) and (Cu,Ni) 6 Sn 5 (low Ni), particularly in the interior of Ni 2 SnP [6][7][8][9]. This Ni(P)-depletion-induced microstructure transition might greatly affect the mechanical properties of solder joints [6,7,10,11]. Ho et al. [6,7] further pointed out that the existence of the IMC multilayer along with the nanovoid formation was very detrimental to the mechanical reliability of the solder joint, resulting in the shear fracture of solder joints in the vicinity of Ni 2 SnP in HSBS testing. ...
... Ho et al. [6,7] further pointed out that the existence of the IMC multilayer along with the nanovoid formation was very detrimental to the mechanical reliability of the solder joint, resulting in the shear fracture of solder joints in the vicinity of Ni 2 SnP in HSBS testing. A similar result was also obtained by Duh et al. [10,11] via high-impact testing, where joint cracks tended to develop within the Ni 2 SnP layer in testing. These previous studies [6][7][8][9][10][11] confirmed that the interfacial reaction between solder and Au/Pd(P)/Ni(P)/Cu (ultrathin-Ni(P)-type) and the resulting mechanical properties are strongly dependent on the Ni(P) depletion or the Ni(P) thickness. ...
Recently, a Au/Pd(P)/Ni(P) surface finish with an ultrathin Ni(P) thickness (<1 μm) has received a great deal of attention from microelectronic industry because of the requirements of the fine-pitch packaging and the signal performance for high-frequency applications. The effect of Ni(P) and Pd(P) thickness on the solderability had been previously investigated in the literature; however, information regarding the P content effect of the Pd(P) film on this reliability issue is still seriously lacking to date. The focus of this study was to examine the interfacial reaction and mechanical reliability between a Sn-3Ag-0.5Cu alloy and an ultrathin-Ni(P)-type Au/Pd(xP)/Ni(P)/Cu metallization pad, where the thicknesses of Au/Pd(xP)/Ni(P) were 0.08 μm/0.13 μm/0.13 μm and x were 0 wt%, 1–2 wt%, and 4–6 wt%. We found that the growth morphologies of intermetallic compounds (IMCs) at the Sn-3Ag-0.5Cu/Au/Pd(xP)/Ni(P)/Cu joints strongly depended on x and reflow number. The shear resistance of the solder joints could be greatly enhanced by reducing x, especially for one reflow; however, this difference was nearly alleviated after multiple reflows. This investigation provided valuable information about the role of P in the Sn-3Ag-0.5Cu/Au/Pd(P)/Ni(P)/Cu reaction system and the strategy to enhance the solderability of the Au/Pd(P)/Ni(P) surface finish.
... In order to retard the rapid interaction between Cu substrate and Sn-based solders, it would be of significance to seek a reliable surface finish technology for the application of Cu substrate. Ni-based metallization has been regarded as a potential surface finish for Cu substrates and diffusion barrier against rapid reaction between the solder and Cu [13]. Due to its low cost, simplicity, strong adhesion and high resistance to corrosion, electroless Ni-P plating of the Cu substrate has been widely used as a diffusion barrier layer to prevent a rapid reaction between the solders and Cu [12][13][14][15]. ...
... Ni-based metallization has been regarded as a potential surface finish for Cu substrates and diffusion barrier against rapid reaction between the solder and Cu [13]. Due to its low cost, simplicity, strong adhesion and high resistance to corrosion, electroless Ni-P plating of the Cu substrate has been widely used as a diffusion barrier layer to prevent a rapid reaction between the solders and Cu [12][13][14][15]. Generally, the electroless Ni-P plating layer contains about 15 at.% ...
... The EDS result presents in Fig. 5b confirms that the P-rich Ni layer is Ni 3 P. The presence of Ni 3 P was ascribed to the consumption of Ni and the accumulation of P in the electroless Ni-P plating layer as a result of the formation of Ni 3 Sn 4 [11,13,35]. The Ni 3 Sn 4 IMC and Ni 3 P layer grow with the prolongation of aging times, as shown in Fig. 4. ...
In this work, the interfacial reactions and IMC growth of the Sn–35Bi–1Ag on Cu, Ni–P/Cu and Ni–Co–P/Cu were studied during reflowing at 220 °C for 10 min and solid–state treatment at 150 °C with various aging times. For the solder joints of Sn–35Bi–1Ag/Cu, Cu6Sn5 IMC was formed at the interfacial layer after soldering, while Cu3Sn appeared after aging treatment. In the case of the electroless Ni–P plating of Cu, the IMC formed at the interface during isothermal aging was mainly Ni3Sn4 and a small amount of P-rich Ni (Ni3P) layer between the Ni3Sn4 IMC and the electroless Ni–P plating layer. In the case of the electroless Ni–Co–P plating of Cu, the IMC formed at the interface during isothermal aging was mainly (Ni, Co)3Sn4 and a small amount of P-rich Ni ((Ni, Co)3P) layer between (Ni, Co)3Sn4 IMC and the electroless Ni–Co–P plating layer. The addition of Co atoms could effectively inhibit the IMC growth rate, which inducing a lower growth rate of (Ni, Co)3Sn4 than that of Ni3Sn4. The thickness of Ni3P and (Ni, Co)3P layer reached about 2.31 µm and 1.25 µm after aging for 360 h, respectively. Also, the growth kinetics of the Ni3P and (Ni, Co)3P layer was found to be a diffusion–controlled process, which followed a parabolic relationship in the thickness increase. During the aging, the consumption of the electroless Ni–Co–P plating layer was greatly reduced compared with that of the electroless Ni–P plating layer. Moreover, there were no voids observed in the electroless Ni–Co–P plating layer while some defects formed in that of the electroless Ni–P plating layer. The electroless Ni–Co–P plating layer will be a good diffusion barrier for the lead-free soldering.
... The Au/Pd(P)/ Ni(P) structure (ultrathin-Ni(P)-type) was customarily deposited with a thickness combination of 0.05-0.1 μm/0.05-0.3 μm/0.1-0.5 μm, and such a trilayer might be destroyed within a typical reflow process [1,[4][5][6][7]. The effect of δ Ni(P) on the solderability was recently investigated in several studies [1,[4][5][6][7]. ...
... μm/0.1-0.5 μm, and such a trilayer might be destroyed within a typical reflow process [1,[4][5][6][7]. The effect of δ Ni(P) on the solderability was recently investigated in several studies [1,[4][5][6][7]. These works confirmed that the interfacial microstructure and the resulting mechanical properties of the Sn-Ag-Cu/Au/Pd(P)/Ni(P)/Cu joint system strongly depend on δ Ni(P) [1,[4][5][6][7], even though the difference in δ Ni(P) is only on the 0.1-μm scale [5]. ...
... The effect of δ Ni(P) on the solderability was recently investigated in several studies [1,[4][5][6][7]. These works confirmed that the interfacial microstructure and the resulting mechanical properties of the Sn-Ag-Cu/Au/Pd(P)/Ni(P)/Cu joint system strongly depend on δ Ni(P) [1,[4][5][6][7], even though the difference in δ Ni(P) is only on the 0.1-μm scale [5]. It was observed that the Au/Pd(P)/Ni(P) trilayer can be readily and completely exhausted, particularly for a trilayer with a small δ Ni(P) , and subsequently, the interaction between the solder and the underlying Cu substrate occurs with the breakdown of the Ni(P) diffusion barrier. ...
The strong effect of the Pd(P) thickness (δPd(P)) on the interfacial reaction between Sn-3Ag-0.5Cu solder and ultrathin-Ni(P)-type Au/Pd(P)/Ni(P)/Cu pad was revealed. We observed that the P of the Pd(P) film would redeposit over the Ni(P) platings and formed layered Ni2SnP in the early stage of soldering reaction. An excessive growth of Ni2SnP inhibited the formation of layered (Cu,Ni)6Sn5, which is argued to be an effective diffusion barrier of Ni. Thus, an extremely fast depletion of the Ni(P) film accompanied by a substantial dissolution of the Cu pad might be induced for large δPd(P). These observations demonstrated that δPd(P) plays a dominant role in determining the interfacial microstructure of solder joints and that a precise control on the δPd(P) value is required for microelectronic packaging reliability.
... The combination of fine-pitch interconnection and high-frequency signal results in signal loss and degradation in electronic devices. In particular, signal losses in conductors mainly occur in the surface finish of Cu pads due to the skin effect caused by highfrequency signals [4][5][6]. The most commonly used surface finish in semiconductor package substrates is electroless nickel electroless palladium immersion gold (ENEPIG) due to its long shelf life and excellent multiple reflow properties [3][4][5]. In particular, the Pd layer in the ENEPIG prevents corrosion of the Ni(P) layer during the displacement reaction of gold plating, which improves solder wettability. ...
... To suppress the ferromagnetic Ni(P)-related impedance increase, a thin ENEPIG surface finish with a significantly reduced Ni(P) thickness of 0.1-0.3 lm has been introduced [6]. A recent study by Ho et al. [7] suggested that Ni(P) thickness reduction was necessary in high-frequency package substrates. ...
The interfacial microstructure and brittle fracture reliability of solder joints on direct electroless gold (DEG) and electroless palladium immersion gold (EPIG), which are novel surface finishes for high-frequency package substrates, were evaluated in this study. A Cu6Sn5 intermetallic compound (IMC) was formed at the interface of Sn-3.0Ag-0.5Cu (SAC305)/DEG or SAC305/EPIG, while (Cu,Ni)6Sn5 and Ni3P were formed at the interface of SAC305/ENEPIG. After 1000 h of thermal aging, the IMC thickness of the SAC305/DEG and SAC305/EPIG samples increased by 217% and by 181%, respectively, while that of SAC305/ENEPIG increased by only 51%. Although ENEPIG had the lowest IMC thickness, its brittleness was higher than that of DEG and EPIG because of the different fracture paths. Fractures occurred between Cu6Sn5/Cu pad and Cu6Sn5/Cu3Sn interface for SAC305/DEG and SAC305/EPIG, while fractures mainly occurred in the Ni3P layer of SAC305/ENEPIG.
... The formation of Ni 3 Sn 4 intermetallic layer is characteristic for the interface between the Sn-based solder and nickel [9], and its growth rate is slower. Therefore, nickel can serve as an excellent reaction barrier that restricts the excessive dissolution of Cu into the solder joints [10,11]. The structure of the solder pad (such as its roughness) also influences the IMC, though not primarily in its growth rate, but rather in the IMC ratio between various intermetallic phases, such as Ag 3 Sn, Cu 3 Sn, and Cu 6 Sn 5 phases [12], where their presence is both inevitable and necessary. ...
... The formation of Ni3Sn4 intermetallic layer is characteristic for the interface between the Sn-based solder and nickel [9], and its growth rate is slower. Therefore, nickel can serve as an excellent reaction barrier that restricts the excessive dissolution of Cu into the solder joints [10,11]. The structure of the solder pad (such as its roughness) also influences the IMC, though not primarily in its growth rate, but rather in the IMC ratio between various intermetallic phases, such as Ag3Sn, Cu3Sn, and Cu6Sn5 phases [12], where their presence is both inevitable and necessary. ...
Flux contained in solder paste significantly affects the process of solder joint creation during reflow soldering, including the creation of an intermetallic layer (IML). This work investigates the dependence of intermetallic layer thickness on ROL0/ROL1 flux classification, glossy or matt solder mask, and OSP/HASL/ENIG soldering pad surface finish. Two original SAC305 solder pastes differing only in the used flux were chosen for the experiment. The influence of multiple reflows was also observed. The intermetallic layer thicknesses were obtained by the image analysis of micro-section images. The flux type proved to have a significant impact on the intermetallic layer thickness. The solder paste with ROL1 caused an increase in IML thickness by up to 40% in comparison to an identical paste with ROL0 flux. Furthermore, doubling the roughness of the solder mask has increased the resulting IML thickness by 37% at HASL surface finish and by an average of 22%.
... Lately, electroless Ni-electroless Pd-immersion Au (ENEPIG) surface finish is being increasingly employed in electronics packaging [1][2][3][4]. However, the ENEPIG surface finish has some drawbacks such as its unsuitability for fine-pitch packaging applications such as 3D IC interconnection packages and high electrical resistance of the solder joint package due to a relatively thick Ni layer (generally 4-6 lm) [5][6][7][8][9][10]. These problems will be more prominent in future 5G/6G communication systems and high-radiofrequency (RF) module packages. ...
... These problems will be more prominent in future 5G/6G communication systems and high-radiofrequency (RF) module packages. To reduce the Ni thickness in the normal ENEPIG surface finish, various thin ENEPIG finishes were proposed and evaluated [5][6][7][8][9][10][11][12][13]. We studied the interfacial reactions and mechanical strengths of the Sn-3.0Ag-0.5Cu ...
A comparison study of the interfacial reactions and mechanical shear strengths of normal and thin electroless-nickel electroless-palladium immersion gold (ENEPIG) with Sn–3.0Ag–0.5Cu (SAC305) solder during isothermal aging at 150 °C is presented. The thicknesses of the Ni layers of the normal and thin ENEPIG were 6 and 0.1 μm, respectively. In the normal ENEPIG substrate, small and thin needle-shaped (Cu,Ni)6Sn5 intermetallic compounds (IMCs) were formed at the interface after reflowing, and a relatively thin IMC layer remained despite a long aging period of 1000 h. On the other hand, in the thin ENEPIG substrate, the thin Ni layer as well as the Au and Pd layers were completely reacted, and a relatively thick (Cu,Ni)6Sn5 IMC was formed at the interface after reflowing. In addition, the interfacial IMC layer grew continuously with increasing aging time. During aging at 150 °C, the interfacial IMC layers at the thin ENEPIG joints were consistently thicker than those at the normal ENEPIG joints. During a low-speed shear test, the shear strength did not significantly change depending on the aging time and surface finish, and all the fractures occurred in the ductile mode. On the other hand, in the high-speed shear test, the thick (Cu,Ni)6Sn5 IMC layer significantly deteriorated the shear strength of the thin ENEPIG joints. The thin ENEPIG joints showed inferior mechanical reliability (especially, high impact reliability) than the normal ENEPIG joints during solid-state isothermal aging at 150 °C.
... Meanwhile, P atoms in the Ni(P) plating accumulated at the interface between Ni-Sn compound and substrate to form a P-rich layer (Ni 3 P or Ni 2 SnP) [14]. Besides, it has been reported that the thickness of Ni(P) plating had a significant influence on the reliability of solder joints [15][16][17][18]. ...
... Indeed, a thick Ni(P) plating could effectively inhibit atom diffusion between the solder and substrate. However, excessive thick Ni(P) plating not only hindered the fine-pitch packaging application, but also reduced reliability of electrical signals [15][16][17]. Besides, the excessive thickness also increased the cost and created unnecessary waste. Thus it was suggested that the 1.5 μm Ni(P) plating was over-thick in this study. ...
Effects of the Ni(P) plating thickness on interfacial reaction in the Sn–58Bi/Ni(P)/Cu joint system were revealed. It was found that the interfacial reaction was significantly influenced by the thickness of Ni(P) plating, and 0.1 μm Ni(P) plating completely transformed into Ni2SnP layer after soldering. This Ni2SnP layer not only provided a large number of diffusion channels but also reduced the solder joint reliability, demonstrating that 0.1 μm Ni(P) plating was not efficient in inhibiting the diffusion process between solder and substrate. However, the Ni(P) plating with thickness more than 0.5 μm could effectively inhibit atomic diffusion, and the Sn–Ni interaction would dominate the interfacial reaction instead of Cu–Sn phases. Although the Ni(P) plating with thickness of 0.5 μm partly transformed into Ni2SnP layer, the growth rate of compound layer was suppressed. In addition, the Ni3Sn4 would transform into (Cu,Ni)6Sn5 since Ni2SnP layer provided channels for Cu diffusing toward the solder/Ni3Sn4 interface. The Ni(P) plating with thickness of 1.5 μm remained integrated even after aging for 240 h, nonetheless, the excessive thickness of this barrier was unnecessary. Thus, it could be concluded that the appropriate thickness of Ni(P) plating should be controlled at 0.5–1.5 μm.
... In order to correlate the modeling TTF to actual experiment results, the vacancy concentration threshold needs to be tuned. Moreover, it can be used as the parameter to characterize the diffusion barrier effect [13] when the test structure has different surface finishes. In the modeling we describe the critical concentration as: ...
With smaller and denser transistors, the physical flow of electrons may inhibit the performance of the device
over time by forming voids and cracks at interconnects due to Electro-Migration (EM). Circuit designs that
fail to meet EM specifications may lead to catastrophic failures and SI/PI performance degradation. One way
of mitigating EM is to use multiple vias between layers of copper traces to reduce the current crowding effect.
However, the quantities of vias may affect the current density and current redistribution inside critical joints.
Current studies mainly focus on predicting the EM time-to-failure (TTF) based on the empirical Black’s
equation. However, this method may not give enough insights about void formation and crack propagation
and reflects the current redistribution that could impact the TTF. In this study, we compared the EM lifetime
of Ball Grid Array (BGA) test vehicles with different structural designs and developed a methodology to
consider the diffusion of atoms in solder joints based on Multiphysics field migration to study the current
redistribution influence of vias. Moreover, crack propagation was also simulated to understand the failure
mechanism. BGA traces without vias and with 8 vias are stressed under 5A, 7A, and 9A at 150C to compare
the EM performance. Moreover, each test structure is manufactured with two different surface finishes: A
and B. Based on the experimental results, Finite Element Analysis (FEA) simulations based on Atom Flux
Divergence (AFD) were performed to compare with the experiment results. It was found that the current
crowding effect could be significantly reduced by 8 vias compared to daisy chained traces. The study shows
better EM resistance with 8 and 4 vias than no-via traces and helps predict the EM life of different structures
to provide guidance for design optimization
... Recent 2010s, the trends in electronic packaging technologies have evolved to include high-performance, multi-functionality, large input/output (I/O) density, high-capability, and 3-D integration with downsizing packages [1][2][3][4]. In addition, these performances should ensure high mechanical and electrical reliabilities. ...
To analyze the effects of Ni(P) layer thickness and Pd layer composition on interfacial reactions and the mechanical reliabilities of Sn–58Bi solder joints, we evaluated a phosphorous-contained Ni (Ni(P)) layer thicknesses ranging from 0.3 to 1.0 μm with Au/Pd/Ni(P) or a phosphorous-contained Pd [Au/Pd(P)/Ni(P)] layer in thin-electroless-nickel electroless-palladium immersion gold (ENEPIG) with Sn–58Bi solder joint after aging test. (Pd, Au)Sn4 and Ni3Sn4 intermetallic compounds (IMCs) were dominantly formed at the interfaces of the 0.3 µm to 1.0 μm Ni(P) layers in the thin-Au/Pd/Ni(P) or thin-Au/Pd(P)/Ni(P) joints after aging at 85 °C and 95 °C for 100 h. However, the Ni3Sn4 IMC layer changed to the (Cu, Ni)6Sn5 IMC layer in the 0.3 μm Ni(P) layer contained the Au/Pd/Ni(P) joint after aging at 85 °C for 300 h, because the Cu elements in a Cu pad penetrated through the P-rich Ni layer. Otherwise, the Ni3Sn4 IMC of the 0.3 μm Ni(P) layer in the Au/Pd(P)/Ni(P) joint changed to (Ni, Cu)3Sn4 IMC after aging at 105 °C and 115 °C for 1000 h, due to the P in the Pd layer, which affects the IMC growth rate. The 0.7 µm and 1.0 μm Ni(P) layers in the Au/Pd/Ni(P) or Au/Pd(P)/Ni(P) joints were attributed to the Ni3Sn4 IMC layer for whole aging conditions because the thick P-rich Ni layer suppress Sn and Cu diffusion during aging. In a high-speed shear tests, the shear strength of the 0.3 μm Ni(P) layer in the Au/Pd/Ni(P) joints was relatively low than that of the Au/Pd(P)/Ni(P) joints after aging at 105 °C and 115 °C for 100 h. Ni3Sn4 IMC was observed at the fracture surfaces of the 0.3 μm Ni(P) layer in the Au/Pd(P)/Ni(P) joints after aging at 115 °C for 1000 h, whereas the fracture surface of the Au/Pd/Ni(P) joint was Cu substrate. Therefore, Ni(P) layer thicknesses in excess of 0.7 μm and the P-contain Pd layer in the thin-ENEPIG surface finish with Sn–58Bi solder joints are expected to be highly reliable after long-term aging treatment.
... Furthermore, the ENEPIG surface finish increases electrical resistance and noise problems due to the high electrical resistance of the Ni layer [26]. For these requirements, thin-ENEPIG surface finishes have been researched for the past few years for application in the fine-pitch process and for improving the electrical characteristics to reduce the resistance and impedance of the solder joints [27,28]. ...
The effects of Ni(P) layer thickness (5 μm and 0.7 μm) on the microstructural behavior and electrical reliability of electroless-nickel electroless-palladium immersion gold (ENEPIG) (substrate-side) surface-finished printed circuit boards (PCBs) with Sn–3.0Ag–0.5Cu (SAC305) solder joints under current stressing of 9000 A/cm² have been investigated. An organic solderability preservative (OSP) surface finish was applied on the chip-side. (Cu,Ni)6Sn5 and Cu6Sn5 intermetallic compound (IMC) layers were formed on the chip-side (the interface between SAC305 and the OSP surface finish) and substrate-side (the interface between the ENEPIG surface finish and SAC305) solder joints after reflow. The thicknesses of the (Cu,Ni)6Sn5 and Cu6Sn5 IMC layers of the chip- and substrate-side of the normal- and thin-ENEPIG/SAC305/OSP solder joints increased with increasing current stressing time, regardless of the nickel phosphorous (Ni(P)) layer thickness. The total IMC thicknesses of normal-ENEPIG/SAC305/OSP solder joints were relatively thinner than those of thin-ENEPIG/SAC305/OSP solder joints under current stressing for 50–120 h. This is the reason why only a P-rich Ni layer was formed at the interface between SAC305 solder and the Cu pad in the thin-ENEPIG/SAC305 solder joints under current stressing. Otherwise, the P-rich Ni and Ni(P) layers remained at the interface of the normal-ENEPIG/SAC305 solder joint under current stressing. The Ni(P) layer of the ENEPIG surface finish played an important diffusion barrier role by suppressing IMC growth and movement toward the SAC305 solder under current stressing. In the electrical evaluation, the time to failure at the normal-ENEPIG solder joint was relatively longer (approximately 2.2 times) than that of the thin-ENEPIG solder joint. Therefore, the relatively thick Ni(P) layer contained in the ENEPIG/SAC305/OSP solder joint is expected to attain higher electrical reliability under the electromigration test.
... The surface finish was responsible for improved joint reliability. It is germane that surface finish could prevent oxidizing, contamination (Sona and Prabhu, 2014) and act as a diffusion barrier to prevent rapid reaction between the solder and Cu substrate (Rohan et al., 2002;Ho et al., 2013;Pun et al., 2014;Siti Rabiatul Aisha et al., 2015). From the foregoing, it is crucial to seek a reliable surface finish for the Cu substrate applications. ...
Purpose
The surface finish is an essential step in printed circuit boards design because it provides a solderable surface for electronic components. The purpose of this study to investigate the effects of different surface finishes during the soldering and ageing process.
Design/methodology/approach
The solder joints of Sn-4.0Ag-0.5Cu/Cu and Sn-4.0Ag-0.5Cu/electroless nickel/immersion silver (ENImAg) were investigated in terms of intermetallic (IMC) thickness, morphology and shear strength. The microstructure and compositions of solder joints are observed, and analyzed by using scanning electron microscopy (SEM-EDX) and optical microscope (OM).
Findings
Compounds of Cu 6 Sn 5 and (Cu, Ni) 6 Sn 5 IMC were formed in SAC405/Cu and SAC405/ENImAg, respectively, as-reflowed. When the sample was exposed to ageing, new layers of Cu 3 Sn and (Ni, Cu) 3 Sn 5 were observed at the interface. Analogous growth in the thickness of the IMC layer and increased grains size commensurate with ageing time. The results equally revealed an increase in shear strength of SAC405/ENImAg because of the thin layer of IMC and surface finish used compared to SAC405/Cu. Hence, a ductile fracture was observed at the bulk solder. Overall, the ENImAg surface finish showed excellent performance of solder joints than that of bare Cu.
Originality/value
The novel surface finish (ENImAg) has been developed and optimized. This alternative lead-free surface finish solved the challenges in electroless nickel/immersion gold and reduced cost without affecting the performance.
... Recently, Au/Pd/Ni/Cu metallization with an ultra-thin Ni deposition [16] and Au/Pd/Cu metallization [17,18] have been employed in solder joints to modify the microstructures of intermetallic compounds and improve the mechanical properties of solder joints. The removal of Ni metallization in solder joints seems to have a positive effect on functional performance. ...
The present study investigated the reaction behavior of the AuSn 4 intermetallic compound in the Ni/Sn1.8Ag+Sn3Ag0.5Cu/Au/Pd/Cu solder joint when subjected to multiple reflow for five cycles and solid-state aging at 150 °C for 1000 h. Multiple reflow induced grain refining of the AuSn 4 in the solder matrix. The grain refining was ascribed to the repeated dissolution and reprecipitation of the compound that occurred during multiple reflow. The dissolution of the AuSn 4 was driven by the chemical potential gradient at the AuSn 4/molten solder interface. The dissolution behavior enriched the solder matrix with the constituent elements that were partly involved in the formation of the Cu 6Sn 5 at the solder/Cu interface and the subsequent reprecipitation as the refined AuSn 4 particles when the solder was solidified. Solid-state aging, however, induced complete dissolution of the AuSn 4 and the migration of the constituent elements toward the solder/Cu interface to form a more stable (Cu,Ni,Pd,Au) 6Sn 5 phase due to the thermodynamic stability competition that occurred among the compounds. Consequently, no AuSn 4 reprecipitation behavior occurred during solid-state aging. An understanding of the reaction behavior of the AuSn 4 intermetallic compound is important for practical applications due to its brittle characteristics. The refining of the compound throughout the solder matrix may also strengthen the solder joint, leading to a better performance.
... Ball grid array (BGA) packaging has been widely adopted in high-performance electronic devices to achieve a higher density of I/O connections with the increasing demand of thermal and electrical requirements. Because of the features of smaller pitch spacing and electrical path length on the printed circuit board (PCB), BGA package is suitable for a wide variety of devices, such as microprocessors, attracted wide attention recently [2][3][4][5][6][7][8][9][10], mainly because of two reasons: (1) solder materials are evolving to be lead-free as legislated by RoHS [11], which leads to lower ductility and higher melting temperature; and (2) the dimensions of package sizes are miniaturizing, which may challenge the robustness of packaging structures. ...
Based on the unified creep and plasticity theory, an improved constitutive model is proposed in this study to describe the uniaxial mechanical behaviour of Sn3.0Ag0.5Cu (SAC305) solder alloy subjected to a wide range of strain rates. In the usual service condition of electronic devices, the strain rates of solder material are far less than 1.0 s−1 at which the creep deformation is dominant, especially at higher working temperatures. However, the strain rate could range from 1.0 to 300 s−1 under drop impact in electronic packaging structures, which is drawing more attention due to lack of experimental data, especially on dynamic mechanical properties of lead-free solder alloys. In extreme impact conditions, the solder material may experience even higher strain rates. As different mechanisms dominate the respective regime of strain rates, the developed constitutive model is calibrated to be applicable to most of the strain rate regimes by properly considering the coupled effect of creep and plasticity. Moreover, the parameters in the proposed model are defined with clear physical meanings and reasonably determined by regression to the published experimental studies. Lastly, the developed model is compared with other constitutive models from the literature, including the power-law equation for creep deformation at low strain rates and the Johnson–Cook model for plastic deformation at high strain rates. It is concluded that the proposed model is more generalized and capable of predicting uniaxial mechanical behaviour of SAC305 solder at low, medium and high strain rates with reasonable accuracy.
... The development of micron-scale Ni(P) surface finishes over Cu pads, such as electroless nickel/electroless palladium/immersion gold (ENEPIG), has been restrained in specific packaging fields [1,2]. This is because a thick Ni(P) deposition can degrade signal-delivery performance , especially in radio frequency applications [3,4], and can cause bridging in fine-pitch electronic components [5,6]. ...
A comparative study between Au/Pd/Cu and Au/Pd(2–5 wt% P)/Cu films in soldering applications was carried out by means of field-emission scanning electron microscopy (FE-SEM), field-emission electron probe X-ray microanalysis (FE-EPMA), ultraviolet/visible/near infrared (UV/VIS/NIR) spectrophotometry, X-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), and high-speed ball shear (HSBS) test. We investigated the interfacial microstructure via FE-SEM and FE-EPMA, the mechanical properties of solder joints via HSBS test, and the light properties and the microstructure of the exterior surfaces of solder bumps via UV/VIS/NIR spectrophotometry, XPS, and TEM. The interfacial microstructures/mechanical properties (with Cu pads) were approximately independent of the P content in the Pd(P) films. However, the discoloration in the solder bumps resulting from the formation of a SnO2 nanolayer (or Au/Pd/Cu) could be greatly improved by replacing it with a Sn-O-P phase through a minor addition of P into the Pd film. This suggested that the Au/Pd(P)/Cu pads enabled an appropriate solderability. The results of the above investigations indicated that the dissolved P from the Pd(P) film during soldering played a significant role in determining the light properties of solder joints, even though the Pd(P) film was less than one micron thick and its P content was quite limited (2–5 wt%).
The growing need to use low-power and high-frequency signals for the high-speed transmission of large-capacity data is driving the demand for suitable printed circuit boards. Most printed circuit boards use Cu as a conducting material and require a surface finish that forms an organic or a metallic protective layer on the Cu surface to prevent Cu diffusion/oxidation. Metallic protective layers typically have a higher resistivity than Cu, resulting in significant signal transmission losses in the high-frequency range. For example, electroless Ni/electroless Pd/immersion Au (ENEPIG), a surface finish that forms a metallic protective layer, sequentially forms Ni-P, Pd-P, and Au layers on a Cu surface. Among these, the Ni-P layer is the main cause of signal transmission loss owing to its high resistivity. In this study, we developed immersion Ag/immersion Pd/immersion Au (ISIPIG), a cost-effective surface finishing process that does not form a Ni-P layer. ISIPIG effectively prevented the diffusion/oxidation of Cu and improved solder wettability while exhibiting a lower insertion loss and higher antenna efficiency than ENEPIG. Our results suggest that ISIPIG is a promising surface finishing process for applications that require the high-speed transmission of large-capacity data using low-power and high-frequency signals.
Crystalline and amorphous cobalt-phosphorus (Co-P) coatings have their own advantages in interfacial diffusion resistance and mechanical ductility of solder joints, respectively, but it cannot be better at both. In this work, the hybrid crystalline-amorphous Co-P coating (Co-9.1 at.% P) was fabricated and applied on the Cu/Sn interface by controlling the crystallinity through compositional design. Combining the advantages of crystalline and amorphous Co-P, strategies for solder joints with high shear strength, ductile intermetallic compounds (IMCs) and slow interface consumption was proposed. The evolution, microstructure, phase, and mechanical properties of the Co-Sn, Co-Sn-P and P-rich layers in solid-state diffusion were systematically characterized and tested by means of SEM, EPMA, EBSD, TEM, nanoindentation, and shear test. It was found that the hybrid Co-P coating achieved a low consumption rate, 14.2 nm/h, attributing to the partially crystalline structure and the Co-Sn-P exfoliation behavior. In addition, solder joints containing large-thickness ductile Co-Sn IMCs, 3.2-3.6 GPa of hardness, was established. The shear strength after aging reached 75.4 MPa relying on the large thickness of Co-Sn IMCs. Severe mechanical loads would be able to be withstand by virtue of the plastic deformation of ductile IMCs. Hybrid crystalline-amorphous Co-P would be a promising interface material in microelectronic packaging.
The demand for flexible wearable devices/substrates with miniaturization and improved integration in microelectronic devices has intensified the research interest in low-temperature laser soldering processes as an alternative to conventional reflow soldering processes owing to their advantages, such as local heating, non-contact heating, and short bonding time. In this study, we compared and evaluated the reliability of laser soldered and conventional reflow soldered joints using representative low melting temperature eutectic SnBi solder and thin electroless Ni-electroless Pd-immersion Au (ENEPIG)-finished Cu pads. Laser soldering was performed using various laser powers (130, 150, and 170 W) and times (2 and 4 s). Furthermore, an aging test was performed at 110 °C for 2000 h to evaluate the long-term reliability of the soldered joints. The mechanical properties, including the top and cross-sectional views and fracture surfaces, of the soldered joints were analyzed by conducting shear tests after aging. During laser soldering, various intermetallic compounds (IMCs) were formed at the joints depending on the applied energy. The metallization layer and Cu reacted with Sn in the solder after different aging durations, and additional IMCs were formed and grown. After aging for 2000 h, the shear strength decreased, and the interfacial IMC thickness increased. As the aging time increased, the fracture mode changed from an initial ductile fracture to brittle fracture (between the solder and IMCs and/or between IMCs and the Cu pad). The reflow soldered joints exhibited stable shear strength, resulting in ductile fracture until aging for 500 h. However, the shear strength decreased sharply after aging for 1000 and 2000 h, and Bi-segregation was observed after aging for 1000 h, resulting in inferior long-term reliability. After laser soldering at 150 and 170 W for 4 s, the strength of the samples decreased sharply after aging for 1000 and 250 h, respectively, and Bi-segregation was observed after aging for 2000 h. The shear strength of the sample laser soldered at 170 W for 2 s gradually decreased with increasing aging time and maintained a stable shear strength until aging for 2000 h. Therefore, laser soldering at 170 W for 2 s was considered as the optimal condition for long-term reliability.
Intermetallic compound (IMC), as an inevitable part between pad and solder, has a severe effect on the strength and reliability of microelectronic interconnection. Here, an investigation was carried out on IMC growth for different devices and complex components. The device-level experiments were conducted with five factors: peak temperature, time duration above solder liquidus temperature, the thickness of solder paste, surface finish types, and package types including BGA and QFP. Meanwhile, four complex components with the same reflow profile were conducted and compared for component-level experiments. A scanning electron microscopy was used to measure the thickness and determine the spatial distribution of the elements through the intermetallic compound. The multivariate analysis of the formation and growth of IMC during reflow soldering was studied based on Nernst-Shchukarev's equation and the results of the experiments. The difference in IMC thickness between BGA and QFP with different factors was discussed and compared separately. The results showed that the peak temperature and time above liquidus played a vital role in the IMC growth and the solder paste thickness and different pad metallization could not be ignored. SEM pictures of the solder and statistical results were revealed that the surface finish type has a marked impact on the formation of the IMC. For PCB with numbers of components, the IMC thickness and uniformity of solder joints at corner and center positions showed some regularity differences. Meanwhile, the bump shape (Cu1-xNix)6Sn5 IMC was observed for small size BGA with ENIG during the reflow process. The results have a significant meaning to optimize its reflow process parameters for complex components, to improve the interconnection reliability in engineering.
In the advanced microelectronic packaging, Cu/Sn joints require high-quality diffusion barriers to inhibit Cu–Sn intermetallic compounds (IMCs) growth and improve the interface brittleness. Cobalt-Phosphorous (Co–P) alloy coatings are ideal candidates due to their good wettability, promising diffusion resistance and low interfacial brittleness. In this work, the microstructural evolution of IMCs between crystalline/amorphous Co–P coatings and lead-free solders during the solid-state diffusion were systemically investigated. The microstructure, phase distribution, and grain characteristics of coatings and IMCs were carefully characterized by SEM, EPMA, EBSD, and TEM. It was found that the consumption rate of crystalline Co–P, 21.2 nm/h, was significantly slower than that of amorphous Co–P, 96.4 nm/h, and exhibited an excellent diffusion resistance, which was attributed to the nanocrystalline structure and the P-rich layer which performed the diffusion barriers for Sn and Cu atoms. Compared with Ni, amorphous Co–P coatings could better improve the interfacial brittleness with soft and ductile Co–Sn compounds (hardness of 3.0 GPa) growing with the rapid diffusion of Co, and the toughened Co–Sn–P attributing to the diffusion of Sn. Kinetic analysis showed that the growth rate-controlling process of Co–Sn IMCs was jointly controlled by interfacial reaction and volume diffusion, and the effect of the interfacial reaction is more pronounced in the amorphous system, which leads to coarse Co–Sn grains that are beneficial to the softness and ductility of IMCs. Crystalline and amorphous Co–P coatings can be applied to the surface treatment for die pads and print-circuit-board pads according to their own characteristics, respectively.
The increasing operation temperature of electronic devices significantly accelerates the growth of Cu-Sn intermetallic compounds (IMCs) at the interfaces of Sn-Ag solder/Cu substrate which deteriorates the device reliability. A robust diffusion barrier is thus required to suppress the interfacial reactions between solders and substrates. In this work, electroless Ni-Fe-P coatings with three different internal microstructure were prepared to compare their interfacial reactions and diffusion barrier properties in Sn-Ag solder interconnects. Ni-Fe-P coatings can effectively suppress the excessive growth of interfacial IMCs in Sn-Ag solder interconnects during reflow process. It has been found that the Ni-Fe-P coatings with crystalline structure had great influence on IMCs formation and morphology at Sn-Ag/Ni-Fe-P interfaces. Among three types of Ni-Fe-P coatings, the crystalline Ni-Fe-P exhibited the optimal diffusion barrier property owing to the presence of a uniform and thinnest FeSn2 layer at the interface, while the amorphous Ni-Fe-P coating was less effective due to the irregular and thickest Ni3Sn4 IMCs formed at the interface.
We analyzed the effects of phosphorous (P) in a palladium (Pd) layer on the interfacial reactions and mechanical properties of Sn–58Bi solder with thin electroless-nickel electroless-palladium immersion gold (ENEPIG) joints (pure Pd and Pd(P) joints) after a high-temperature storage test. Flake-type (Pd,Au)Sn4 intermetallic compounds (IMCs) mainly formed in the pure Pd and Pd(P) joints. After being aged at 85–115 °C for 300 h, top- and bottom-side (Cu,Ni)6Sn5 IMCs formed at the interface between the (Pd,Au)Sn4 IMC and partially destroyed P-rich Ni layer of the pure Pd joints. In contrast, a Ni3Sn4 IMC formed at the interface between the (Pd,Au)Sn4 IMC and P-rich Ni layer of the Pd(P) joints after most aging temperatures and durations. This is because the reaction rate of the Pd(P) joint was lower than that of the pure Pd joints. The unreacted Ni layer remained at the Pd(P) joints after aging. In a high-speed shear test, the shear strength of the pure Pd joints rapidly decreased after being aged at 115 °C for 100 h. In contrast, the shear strength of the Pd(P) joints slightly decreased after being aged at 115 °C for 1000 h. The shear strengths of the pure Pd and Pd(P) joints were quite different under same aging conditions because of the different diffusion rates and IMC compositions of the pure Pd and Pd(P) joints. Fracture surfaces of the pure Pd joints after being aged at 115 °C were observed on the Cu substrate, but those of the Pd(P) joints were on the Ni3Sn4 IMC surface. Therefore, the Sn–58Bi solder with a Pd(P) layer in thin-ENEPIG is expected to be more reliable than that with pure Pd joints under aging treatment.
The microelectronics packaging industry, although rapidly growing, faces several challenges including 3-D integration, issues with multifunctional capability, and fluctuating input/output (I/O) density, among others. Better-performing microelectronics assemblies for mitigating these challenges require alloys with superior solderability and minimal metallization layer thickness. To this end, in this study, we investigated two kinds of electroless-nickel electroless-palladium immersion gold (ENEPIG) with 0.3 μm Ni, 0.1 μm pure Pd or Pd-phosphorous (Pd(P)), and 0.1 μm Au layers plated on a printed circuit board (PCB) substrate. To analyze the effects of the pure Pd and Pd(P) layers in the thin ENEPIG, we evaluated the interfacial reactions and mechanical properties of the SAC305 solder with a pure Pd or Pd(P) layer in the thin ENEPIG joints after aging at 150 °C. Needle-type and chunky-type (Cu,Ni)6Sn5 IMCs were formed at the interfaces of the pure Pd and Pd(P) joints, respectively. The (Cu,Ni)6Sn5 IMC of the pure Pd joint was thinner than that of the Pd(P) joint after reflowing and aging for 100 h. However, the total IMC of the Pd(P) joint was thinner than that of the pure Pd joint from 250 to 1000 h. In a low-speed shear test, the shear strength of the Pd(P) joint was higher than that of the pure Pd joint for the entire aging time. Most fractures occurred at the Sn-rich surface with a ductile mode, regardless of the different substrates and aging times. After high-speed shear testing, the shear strength of the pure Pd joint was higher than that of the Pd(P) joint until aging for 100 h. After aging for 250 h, the shear strength of the Pd(P) joint was higher than that of the pure Pd joint. The results for brittle fracture rate were similar to those for high-speed shear strength. Hence, Pd(P) joints are expected to demonstrate higher reliabilities than pure Pd joints after long aging treatment.
Herein, we investigated the interfacial reactions of crystalline Pd (pure Pd) and amorphous Pd (Pd(P)) layers in thin electroless-Ni electroless-Pd immersion-Au (thin-ENEPIG) surface-finished printed circuit board with Sn-3.0Ag-0.5Cu (SAC305) solder joints reflowed at 260 °C for 10–180 s. After 20 s reflow, a thick (Pd,Au)Sn4 IMC was formed at the interface of the pure Pd joint, while thick AuSn4 and thin (Pd,Au)Sn4 IMCs were formed at the interface of the Pd(P) joint. After 30 s reflow, needle-type (Cu,Ni)6Sn5 IMC was mainly formed at the top-side interface of the pure Pd joint, while scallop-type (Cu,Ni)6Sn5 IMC was formed at the top-side interface of Pd(P) joints. Finally, the top-side (Cu,Ni)6Sn5 IMC of the pure Pd and the top- and bottom-side (Cu,Ni)6Sn5 IMCs of the Pd(P) joints coarsened with increasing reflow time up to 180 s. In the results of top-view SEM micrographs, the (Cu,Ni)6Sn5 IMCs on the Pd(P) joint were larger than those on the pure Pd joint for reflow times of 30–180 s. Therefore, P in the Pd layer was concluded to significantly affect the interfacial reactions and IMC morphology of the SAC 305 solder with ENEPIG joints during reflow reactions.
We investigated the interfacial reactions and mechanical reliabilities of different Ni layer thicknesses and P-containing Pd layers in thin electroless-Ni electroless-Pd immersion gold (thin-Au/Pd(P)/Ni(P)) surface-finished printed circuit board (PCBs) with Sn-3.0Ag-0.5Cu (SAC 305) solder joints. To analyze the optimal Ni layer thickness in the thin-Au/Pd(P)/Ni(P) for aging, we evaluated 0.3- to 1.0-μm Ni layers in thin-Au/Pd(P)/Ni(P) PCBs with SAC 305 solder joints aged at 75, 100, 125, and 150 °C for 1000 h. A scallop-type (Cu,Ni)6Sn5 intermetallic compound (IMC) dominantly formed at the bottom and top sides of the interfaces of all Ni joints under all aging conditions. Furthermore, a P-rich Ni layer formed at the interface between the (Cu,Ni)6Sn5 IMC and Cu substrate during aging regardless of Ni layer thickness. The (Cu,Ni)6Sn5 IMCs of the joints with 0.3- and 0.5-μm Ni layers aged at 125 and 150 °C for 1000 h were thicker than those of 0.7- and 1.0-μm Ni layers. In high-speed shear tests, the shear strength reduction rates of the joints with 0.3-μm Ni layer aged at 125 and 150 °C were higher than those of the 0.7- and 1.0-μm Ni layers. The brittle fracture rates of the joints with 0.3- and 0.5-μm Ni layers aged at 150 °C for 1000 h were higher than those of the 0.7- and 1.0-μm Ni layers. We determined that these trends arose from the diffusion barrier role of the relatively thick P-rich Ni layers maintained at the interfaces of the joints with 0.7- and 1.0-μm Ni layers for all aging temperatures and times. In low-speed shear tests, the shear strengths of the joints with 0.3-μm Ni layer were slightly lower than those of the 0.5- to 1.0-μm Ni layers. The low- and high-speed shear strengths of the joint with 0.7-μm Ni layer were similar to those of the 1.0-μm Ni layers for each condition. Therefore, Ni joint thicknesses of over 0.7 μm are expected to provide high reliability for aging.
This study investigates the kinetics of solid-state dissolution of Ni into Sn and Sn3.5Ag solders. At annealing temperatures of 150, 180, and 200 °C for 100–400 h, more amount of Ni was dissolved in Sn3.5Ag solder than in pure solder. The activation energies of Ni dissolution into pure Sn solder and Sn3.5Ag solder were 0.84 and 0.71 eV, respectively, indicating that the Ag content in the Sn3.5Ag solder can lower the activation energy of Ni dissolution into the Sn3.5Ag solder matrix. This study also proposes a potential enhancement mechanism for Ni dissolution using Ag solutes. Ag solutes in Sn3.5Ag solder promote a higher density of grain boundaries and a higher Sn vacancy concentration in β-Sn grains in Sn3.5Ag solder, resulting in more Ni dissolution into Sn3.5Ag solder compared to pure Sn solder. Based on the Ni dissolution results, a radial growth equation for a Ni 3 Sn 4 interfacial layer surrounding a Ni wire was established. Further, the parameters of the radial growth of this Ni 3 Sn 4 interfacial layer were deduced.
The feasibility of novel ultrathin-Electroless Ni/Electroless Pd/Immersion Au (ultrathin-ENEPIG) metallization containing ultrathin electroless Ni layer for 3D-IC packaging application was evaluated by thermal shock test in this study. Via employing ultrathin-ENEPIG pad, the joints showed higher resistance against thermomechanical stress due to minor degradation in solder and the strength enhancement of interfacial IMC. In the as-fabricated joints, larger Sn grain formed in the ultrathin-ENEPIG joints in contrast to those in the conventional-ENEPIG one, which caused the distinct failure modes between two different systems under the test. Additionally, a dual-layer structure of high/low Ni–(Cu,Ni)6Sn5 and a single layer of low Ni–(Cu,Ni)6Sn5 could be observed in the conventional-ENEPIG and the ultrathin-ENEPIG joints, respectively, after testing. The dual layer (Cu,Ni)6Sn5 degraded the thermal shock performance of conventional-ENEPIG joints as a result of the weak bonding of interface between high/low Ni–(Cu,Ni)6Sn5 layers. However, through employment of ultrathin-ENEPIG substrates, the electroless Ni layer was completely exhausted and thus only one layer of low Ni–(Cu,Ni)6Sn5 intermetallic formed in the ultrathin-ENEPIG joints. The formation of dual-layer compound was suppressed and the fracture between layers was inhibited, leading to a stronger bonding at interface. Furthermore, the overall thickness of interfacial intermetallic compound was also reduced due to the suppression of high Ni–(Cu,Ni)6Sn5 layer. Influences of Ni thickness on the related mechanisms behind dual-layer intermetallic suppression and the Sn grain structure in ultrathin-ENEPIG joints were addressed and discussed in details.
The interfacial microstructures of Sn-3.0Ag-0.5Cu (SAC305) solder systems with thin Ni(P) mono-coatings and Cu-Ni(P) dual-coatings were investigated after reflowing and isothermal aging. The ultrathin mono-Ni(P) plating of the SAC305/Ni(P) solder joint was found to rapidly decompose and then transform into a Ni2SnP phase. An intermetallic compound (IMC) formed at the plating/substrate interface, indicating that the ultrathin mono-Ni(P) plating was not an effective diffusion barrier. However, only a single IMC layer ((Cu,Ni)6Sn5) formed at the solder/plating interface in the SAC305/Cu/Ni(P)/Cu system. The (Cu,Ni)6Sn5 IMC effectively suppressed atomic diffusion, protecting the Ni(P) plating and Cu substrate. Although P-Sn-O pores formed in the root of the (Cu,Ni)6Sn5 IMC layer, the dual-Cu/Ni(P) plating protected the solder system for an extended period. The IMC growth rate constants of the SAC305/Cu, SAC305/Ni(P), and SAC305/Cu/Ni(P)/Cu solder joint systems were 0.180, 0.342, and 0.068 μm/h1/2, respectively. These results indicate that the application of dual-Cu/Ni(P) plating can effectively hinder the growth of IMC.
Low-temperature soldering constitutes a promising solution in interconnect technology with the increasing trend of heat-sensitive materials in integrated circuit packaging. Experimental work was carried out to investigate the effect of electroless Ni/electroless Pd/immersion gold (ENEPIG) layer thicknesses on Sn-Bi-Ag solder joint integrity during extended reflow at peak temperatures as low as 175°C. Optimizations are proposed to obtain reliable solder joints through analysis of interfacial microstructure with the resulting joint integrity under extended reflow time. A thin Ni(P) layer with thin Pd led to diffusion of Cu onto the interface resulting in Ni3Sn4 intermetallic compound (IMC) spalling with the formation of thin interfacial (Ni,Cu)3Sn4 IMCs which enhance the robustness of the solder after extended reflow, while thick Ni(P) with thin Pd resulted in weakened solder joints with reflow time due to thick interfacial Ni3Sn4 IMCs with the entrapped brittle Bi-phase. With a suitable thin Ni(P), the Pd thickness has to be optimized to prevent excessive Ni–P consumption and early Cu outward diffusion to enhance the solder joint during extended reflow. Based on these findings, suitable Ni(P) and Pd thicknesses of ENEPIG are recommended for the formation of robust low-temperature solder joints.
Sequential interfacial reactions of Sn‐3.0Ag‐0.5Cu solder on a new multilayer metallization, ENEPIG (electroless nickel‐electroless palladium‐immersion gold) with a 0.1‐μm‐thin Ni(P) layer (thin‐ENEPIG), during reflowing were evaluated in this study. After 20‐second reflow process, layer‐shaped (Au,Cu)Sn4 intermetallic compounds formed at the interface, and thin Ni and Pd layers underneath the Au layer remained unreacted on the polycrystalline Cu substrate. On the other hand, Au, Pd, and Ni tri‐layers in the thin ENEPIG layer were exhausted and (Cu,Ni)6Sn5 intermetallic compounds formed at the interface after reflowing for 30 seconds.
A new multilayer metallization, electroless-nickel electroless-palladium immersion gold (ENEPIG) with a thin 0.1-μm-thick Ni(P) layer (thin-ENEPIG), was plated onto a Cu printed circuit board substrate for fine-pitch package applications. We evaluated the interfacial reactions and mechanical strengths of Sn–3.0Ag–0.5Cu (SAC305) solder on thin ENEPIG-coated substrates over various reflow times and compared them to those of a conventional ENEPIG-coated substrate with a 6-μm-thick Ni(P) layer. Thin Au, Pd, and Ni layers on the thin ENEPIG substrates were exhausted during the initial reflow time, which brought the underlying Cu layer in direct contact with the molten SAC305 solder, resulting in the formation of scallop-shaped (Cu,Ni)6Sn5 intermetallic compounds (IMCs) at the interface. The interfacial IMC layer for the thin ENEPIG substrate was thicker than that for the normal ENEPIG substrate due to the direct contact between the SAC305 solder and the Cu layer. In the low-speed shear test, all the fractures occurred in the bulk solder regardless of the different substrates and reflow times. In the high-speed shear test, the fracture mode was changed from ductile to brittle on increasing the reflow time. The P-rich Ni layer and thick Cu–Sn IMC formation deteriorated the shear strengths of the normal and thin ENEPIG joints, respectively. The thin ENEPIG joints showed better mechanical strength in the solder joint than the normal ENEPIG joints, despite the thick interfacial IMCs.
Hydrocarbon contamination introduced during point, line and map analyses in a field emission electron probe microanalysis (FE-EPMA) was investigated to enable reliable quantitative analysis of trace amounts of carbon in steels. The increment of contamination on pure iron in point analysis is proportional to the number of iterations of beam irradiation, but not to the accumulated irradiation time. A combination of a longer dwell time and single measurement with a liquid nitrogen (LN2) trap as an anti-contamination device (ACD) is sufficient for a quantitative point analysis. However, in line and map analyses, contamination increases with irradiation time in addition to the number of iterations, even though the LN2 trap and a plasma cleaner are used as ACDs. Thus, a shorter dwell time and single measurement are preferred for line and map analyses, although it is difficult to eliminate the influence of contamination. While ring-like contamination around the irradiation point grows during electron-beam irradiation, contamination at the irradiation point increases during blanking time after irradiation. This can explain the increment of contamination in iterative point analysis as well as in line and map analyses. Among the ACDs, which are tested in this study, specimen heating at 373 K has a significant contamination inhibition effect. This technique makes it possible to obtain line and map analysis data with minimum influence of contamination. The above-mentioned FE-EPMA data are presented and discussed in terms of the contamination-formation mechanisms and the preferable experimental conditions for the quantification of trace carbon in steels.
The mechanical properties of Sn-58Bi epoxy solder were evaluated by low-speed shear testing as functions of aging time and temperature. To determine the effects of epoxy, the interfacial reaction and mechanical properties of both Sn-58Bi and Sn-58Bi epoxy solder were investigated after aging treatment. The chemical composition and growth kinetics of the intermetallic compound (IMC) formed at the interface between Sn-58Bi solder and electroless nickel electroless palladium immersion gold (ENEPIG) surface finish were analyzed. Sn-58Bi solder paste was applied by stencil-printing on flame retardant-4 substrate, then reflowed. Reflowed samples were aged at 85°C, 95°C, 105°C, and 115°C for up to 1000 h. (Ni,Pd)3Sn4 IMC formed between Sn-58Bi solder and ENEPIG surface finish after reflow. Ni3Sn4 and Ni3P IMCs formed at the interface between (Ni,Pd)3Sn4 IMC and ENEPIG surface finish after aging at 115°C for 300 h. The overall IMC growth rate of Sn-58Bi solder joint was higher than that of Sn-58Bi epoxy solder joint during aging. The shear strength of Sn-58Bi epoxy solder was about 2.4 times higher than that of Sn-58Bi solder due to the blocking effect of epoxy, and the shear strength decreased with increasing aging time.
Silver paste sintering on bare Cu substrates can be of significant interest in power semiconductor manufacturing. However, the bare Cu substrates easily oxidized and the die attachment using the silver paste commonly applied pressure. In this study, these problems are solved by using a prepared novel silver paste, which could be pressure-less low temperature sintered completely under inert gas (99.99 wt% N2). The silver paste consists of wide size distribution particles as Ag nanoparticles, Ag sub-micro-sized particles and micro-sized Ag flakes. High content of organic system mixture in the silver paste need more evaporating time to avoid cracks in the sintered Ag film. During sintering, the Ag flakes in situ formation of highly reactive silver nanoparticles is crucial for the sintering to occur between the silver flake and particles, unlike the sintering between the silver particles. A dense silver film contributed to the sintered silver flakes, which has the nature of good densification, well sintering with the silver nanoparticle. The possible joining mechanisms of metallic bond and hydrogen bonds between bare Cu and sintered Ag were proved. Accordingly, the sandwich joints (chip/silver paste/bare Cu) include high bonding strength and low thermal impedance.
Recently Electroless Ni/Electroless Pd/Immersion Au (ENEPIG) is being offered as an alternative surface finish to established processing such as Immersion Sn, Immersion Ag, Organic solderability preservatives (OSP), Electrolytic Ni/Au (ENA), and electroless Ni/immersion Au (ENIG) surface finish. With high solder joint quality and wire bondability, it is claimed to be more cost effective as an Au layer of lower thickness can be used. The performance of solder joints upon ENA and ENEPIG surface finishes are evaluated by extended reflow tests at 245°C followed by ball shear testing. Following extended reflow, the ENEPIG/solder system provides lower shear strengths than the ENA/solder system. It is found that columnar Cu-Ni-Sn IMCs with small amount of Pd and P-rich Ni layer have formed at the interface of ENEPIG/solder system, causing brittle fracture with lower shear strength in the ball shear test. In contrast, layer-type Cu-Ni-Sn IMCs formed at the interface of ENA/solder system in the absence of Pd which caused ductile fracture in ball shear test. Therefore, it has been demonstrated that solder joint on ENA surface finish is more reliable and suitable to be used for long-term reliability of electronic products.
The long-term reliability of flip chip and boardlevel solder joint is significantly affected by the properties of the surface finish. Various surface finishes such as Immersion Sn, Immersion Ag, Organic solderability preservatives (OSP), Electrolytic Ni/Au (ENEG), and electroless Ni/immersion Au (ENIG) surface finish have been widely used on based Cu pads and can be compliant with lead free SAC solder alloys. Recently Electroless Ni/Electroless Pd/Immersion Au (ENEPIG) is being offered as an alternative surface finish with high solder joint quality and wire bondability. It is claimed to be more cost effective as an Au layer of lower thickness can be used. To evaluate the performance of solder joints upon ENEG and ENEPIG surface finishes, extended reflow tests at 245°C were conducted. In the case of ENEPIG surface finish, micro-porosities are found to be present on the Au plating surface. PdO oxide forms as a result of Pd exposure, which causes deterioration of solderability as compared with ENEG plating. Following extended reflow, it is found that columnar Cu-Ni-Sn IMCs with small amount of Pd and P-rich Ni layer have formed at the interface of ENEPIG /solder system. In contrast, layer-type Cu-Ni-Sn IMCs formed at the interface of ENEG/solder system in the absence of Pd. While Ni and Ni(P) layers act as barrier to diffusion, in some pad of the ENEPIG samples, the Ni(P) layer is found to be less than 1um after 120 minutes reflow, which is ineffective as a diffusion barrier. By comparison, ENEG surface finish remains effective with 3.5um of Ni barrier layer remaining on Cu after 120minutes reflow. This serves as a good barrier to prevent the diffusion of Cu atoms from the base Cu underneath. Therefore, it has been demonstrated that solder joint on ENEG surface finish is more reliable and suitable to be used for long-term reliability of electronic products.
The effects of Ni(P) thickness, delta(Ni(P)), on the interfacial reaction and high impact resistance of Sn-3Ag-0.5Cu/Au/Pd(P)/Ni(P)/Cu microelectronic solder joints were investigated. During soldering, the surface layers of Au (0.1 mu m) and Pd(P) (0.2 mu m) were readily eliminated from the interface, and the Ni(P)/Cu structure underneath subsequently came into contact with the solder. When delta(Ni(P)) was 0 mu m, a solder/Cu reaction occurred, and the intermetallic compounds (IMCs) Cu3Sn and Cu6Sn5 (dissolved with 1-3 at.% Pd) formed at the interface. When delta(Ni(P)) was sufficiently thick (i.e., 7 mu m), the reaction between the components transformed the system into a solder/Ni(P) system with multilayer IMCs, which included (Cu,Ni)(6)Sn-5, (Ni,Cu)(3)Sn-4, Ni2SnP, and Ni3P. On the submicron scale (e.g., delta(Ni(P)) = 0.9 mu m), the solder/Ni(P) and solder/Cu reactions can occur sequentially, indicating that a submicron-thick Ni(P) layer did not function as an efficient diffusion barrier between solder and Cu. With the aid of Cu-Ni-Sn phase diagram, diffusion path predictions can be made about the IMC translation in response of different delta(Ni(P)). High-speed ball shear testing showed that the impact resistance of the solder joints was significantly reduced with increasing delta(Ni(P)). These findings suggested that delta(Ni(P)) is an important factor in the interfacial microstructure and the resulting mechanical properties of solder joints and that the direct deposition of an Au/Pd(P) dual layer (i.e., delta(Ni(P)) = 0 mu m) on top of the Cu pads can offer an appropriate solderability for the lead-free Sn-Ag-Cu alloy. (c) 2014 Elsevier B.V. All rights reserved.
The solderability of an Au/Pd(P)/Cu trilayer structure (thickness 0.15 μm/0.6–1.0 μm/20 μm) was investigated by use of a scanning electron microscope equipped with an electron backscatter diffraction system, electron probe microanalysis, and high-speed ball shear (HSBS) testing. Thick Pd(P) deposition, particularly the structure with a Pd(P) thickness of 1.0 μm, resulted in a substantial amount of (Pd,Cu)Sn4 intermetallic compound in the solder matrix after reflow. This phase gradually resettled at the interface and formed a dense layer of (Pd,Cu)Sn4 over (Cu,Pd)6Sn5 during solid-state reaction. A Cu–Pd–Sn isotherm was examined to explain the formation and resettlement of (Pd,Cu)Sn4. The (Pd,Cu)Sn4/(Cu,Pd)6Sn5 dual layer at the interface was a crack initiation site in HSBS testing, and led to substantial degradation of the mechanical strength of solder joints, indicating that thick Pd(P) deposition should be avoided to prevent the Pd-induced embrittlement.
The early stage of soldering reaction between Sn-3Ag-0.5Cu solder and ultrathin-Ni(P)-type Au/Pd(P)/Ni(P)/Cu pad was investigated by field-emission scanning electron microscopy (FE-SEM) in conjunction with field-emission electron probe microanalysis (FEEPMA) and high-resolution transmission electron microscopy (HRTEM). FE-SEM, FE-EPMA, and HRTEM investigations showed that Ni2SnP and Ni3P were the predominant P-containing intermetallic compounds (IMCs) in the soldering reaction and that their growth behaviors strongly depended on the depletion of Ni(P). The growth of Ni3P dominated over that of Ni2SnP in the early stage of soldering, whereas the Ni3P gradually transformed into Ni2SnP after Ni(P) depletion. This Ni(P)-depletion-induced Ni2SnP growth behavior is different from the reaction mechanisms reported in the literature. Detailed analyses of the microstructural evolution of the IMC during Ni(P) depletion were conducted, and a two-stage reaction mechanism was proposed to rationalize the unique IMC growth behavior.
In this work, solder balls in ball grid array packaging technology with the pitch of 300 mu m were fabricated by stencil printing solder paste and then reflowed at high temperature. In order to evaluate the quality of solder ball after printing and reflowing processes, the mechanical performance of the joint between the solder balls and the pad was measured by shear test and the electrical resistance was tested after assembly of the substrate and printed circuit board. A comparative study of pad size on the interfacial reaction between solder paste and surface finish of electroless nickel-electroless palladium-immersion gold on the organic substrate was performed and then analyzed by observing the microstructure at the interface. Large discontinuous (Cu,Ni)(6)Sn-5 was found at the interface of the solder with the pad size of 120 mu m, while spalled (Pd,Ni)Sn-4 and thin (Cu,Ni)(6)Sn-5 layer appeared for a pad size of 140 mu m. The IMC (intermetallic compounds) was determined by the residual Cu concentration, the Pd concentration in the solder, and the Ni2SnP barrier layer morphology at the interface, which were significantly influenced by the pad size. A reaction model during the reflow was proposed to illustrate the growth of the IMC and the relationship between the IMC and the pad size. With Pd concentration higher than the solubility of Pd in the solder, spalled (Pd,Ni)Sn-4 took shape along the interface. The solubility of Pd was influenced by Ni concentration; however, the Ni diffusion from the substrate was largely dependent on the barrier layer Ni2SnP. Furthermore, the Ni diffusion also impacted the growth and morphology of (Cu,Ni)(6)Sn-5, which was not only limited by the Cu concentration.
The criteria of mechanical reliability in solder joints can be identified and described by comparative evaluation via drop test and high speed pendulum impact test. Systematic samples of assembly and attachment joints with various Pd additions were employed and investigated in this study. The statistical values of mechanical performances were calculated and compared. Better high speed impact performance of SAC305/ENEPIG attachment joints with 0.06 μm Pd layers was confirmed owing to the single Cu6Sn5 phase growth. However, the comparative measurement of the better performance on drop testing exhibited in ENEPIG/SAC305/immersion Sn assembly joints with 0.1 μm Pd layers deposit resulted from the thinner and layer-type IMC growth. The correlation between the cracks propagation and Pd addition was established on the basis of the elemental X-ray color mapping via Field-Emission Electron Probe Microanalyzer (FE-EPMA). It is expected that through comparison between impact and drop test in mechanical reliability, a criterion of joints reliability can be established. Besides, the optimal Pd layer deposit for the ENEPIG surface finish in the attachment and assembly solder joints was demonstrated and confirmed.
Surface finishes are used to protect exposed copper metallization in printed circuit boards from oxidation and to provide a solderable surface on which to mount electronic components. While it is true that some people have called electroless nickel electroless palladium immersion gold (ENEPIG) a “universal finish” for a wide range of applications from wire bonding to solder interconnects, this paper provides a review of the current literature on ENEPIG and assesses its overall capabilities compared to other surface finishes. Gaps in understanding the performance of ENEPIG as a printed wiring board surface finish are identified and further testing is recommended.
A new multilayer metallization, ENEPIG (Electroless Ni(P)/Electroless Pd/Immersion Au) with ultrathin Ni(P) deposit (ultrathin-ENEPIG), was designed to be used in high frequency electronic packaging in this study because of its ultra-low electrical impedance. Sequential interfacial microstructures of commercial Sn–3.0Ag–0.5Cu solders reflowed on ultarthin-ENEPIG with Ni(P) deposit thickness ranged from 4.79 μm to 0.05 μm were first investigated. Accelerated thermal aging test was then conducted to evaluate the long-term thermal stabilization of solder joints. The results showed that P-rich intermetallic compound (IMC) layer formed when the Ni(P) thickness was greater than a critical vale (about 0.18 μm). Besides, it is interesting to mention that the growth of (Cu,Ni)6Sn5 and (Cu,Ni)3Sn IMCs was suppressed with the formation of P-rich layer, i.e., Ni3P and Ni2Sn1+xP1−x phase, even though the electroless-plated Ni(P) layer was exhausted at initial stage of reflow process. The atomic Cu flux in solder joints without P-rich layer was calculated to be several times larger than that with P-rich layer formation after calculation, which implies that the P-rich layer and ultrathin Ni(P) deposit in ENEPIG served as diffusion barrier against rapid Cu diffusion.
The Sn-rich portion of the phase diagram for the Ni-Pd-Sn ternary system was preliminarily obtained by interpolation of the three constituent binary systems using the Muggianu method. Based on this proposition, 23 Ni-Pd-Sn alloys were prepared and annealed at 250°C. The annealed alloys were analyzed by scanning electron microscopy, electron probe microanalysis, electron backscatter diffraction, and x-ray diffraction. All the binaries adjacent to the Sn-rich corner (i.e., PdSn4, PdSn3, PdSn2, and Ni3Sn4) were found to have remarkable ternary solubility. The experimental results presented herein, together with a thermodynamic interpolation of the ternary system based on the results from the binary systems, were used to calculate the ternary phase diagram using the calculation of phase diagrams (CALPHAD) method. A substitution model was used to describe the Gibbs free energies of the liquid and solid solution phases, and a sublattice model was used to describe intermetallic compounds. A consistent set of thermodynamic parameters was obtained, ultimately leading to a better fit between the calculated results and the experimental data for this system.
Interfacial reactions and joint reliability of Sn-3.0Ag-0.5Cu solder with two different surface finishes, electroless nickel-immersion gold (ENIG) and electroless nickel-electroless palladium-immersion gold (ENEPIG), were evaluated during a reflow process. We first compared the interfacial reactions of the two solder joints and also successfully revealed a connection between the interfacial reaction behavior and mechanical reliability. The Sn-Ag-Cu/ENIG joint exhibited a higher intermetallic compound (IMC) growth rate and a higher consumption rate of the Ni(P) layer than the Sn-Ag-Cu/ENEPIG joint. The presence of the Pd layer in the ENEPIG suppressed the growth of the interfacial IMC layer and the consumption of the Ni(P) layer, resulting in the superior interfacial stability of the solder joint. The shear test results show that the ENIG joint fractured along the interface, exhibiting indications of brittle failure possibly due to the brittle IMC layer. In contrast, the failure of the ENEPIG joint only went through the bulk solder, supporting the idea that the interface is mechanically reliable. The results from this study confirm that the Sn-Ag-Cu/ENEPIG solder joint is mechanically robust and, thus, the combination is a viable option for a Pb-free package system.
This study examined the crystallization behavior of electroless Ni P UBM with an medium phosphorous content (similar to 15 at%) induced by a single and step heat treatment using in situ transmission electron microscopy (TEM) Different heat treatment processes affected the crystallization behavior of electroless Ni P UBM After single heat treatment at 300 degrees C for 60 min the electroless Ni P UBM contained Ni and Ni3P In addition to Ni and Ni3P more complex Ni P compounds such as Ni12P5 and Ni5P2 formed in the electroless Ni P UBM resulting from a step heat treatment at 150 C for 60 min followed by 300 C for 60 min [doi 10 2320/matertrans M2010178]
Pb(100−x)-Sn
x
solders (x=5wt.%, 10wt.%, 20wt.%, 35wt.%, 50wt.%, 60wt.%, 61.9wt.%, and 95wt.%) were directionally solidified upward with
a constant growth rate (V=37.4μm/s) in a temperature gradient (G=4.8K/mm) in a Bridgman-type growth apparatus. The variations of electrical resistivity (ρ) with temperature in the range of 323K to 423K for the directionally solidified Pb-Sn solders were measured. The present
measurements indicate that the electrical resistivity of the directionally solidified Pb-Sn solders increases with increasing
temperature, whereas the resistivity of the Pb-Sn solders decreases with increasing Sn content. The dependency of the Lorenz
number (L) on temperature and Sn content for the Pb-Sn solders was also investigated based on the Wiedemann–Franz law by using the
measured values of the thermal (K) and electrical (σ) conductivity for the same alloys.
KeywordsMetals and alloys–crystal growth–electrical resistivity–thermal conductivity
Black Pad was observed on Electroless Ni/Immersion Au (ENIG) wire bond pads. Thick immersion Au on highly corroded electroless
Ni was detected. It was determined that the pads were electrically connected to the Cu ground plane due to a Ni bridge formed
inside normally open photovias. The mechanism of the bridge formation was verified and preventative actions were taken; it
was demonstrated that formation of Black Pad could be switched on and off. The mechanism of Black Pad formation is proposed
to be defective ENIG plating involving variation of both the electroless Ni and immersion Au plating processes. The intermetallic
structures of solder joints on the above pads were studied. The study was conducted on both defective and non-defective pads
to show differences in intermetallic structure and composition. Me2Sn4 and Me2Sn2 (Me=Cu, Ni, and Au) intermetallics were formed on non-defective pads, which nucleated on the Ni layer and grew inside the
molten solder. However, only the Me3Sn intermetallic was formed on defective pads inside the corroded Ni Layer. Both mechanisms of intermetallic formation were
found on pads with mildly corroded Ni and intermediate Au thickness (4.5–7 in).
In high-speed circuits, skin effects and dielectric properties are important factors for signal degradation considerations, especially in the microwave frequency region. When surface finishes are applied to prevent traces from oxidation, the electrical properties of traces are affected. In this work, experimental study and finite element method (FEM) based full wave simulation are used to investigate the effects of hot air solder leveling (HASL) and its alternatives on signal integrity. Classical interconnect structures, microstrip line, and differential mode coupled microstrip lines subjected to different finishes are investigated. Our work reveals that the net conductor loss that results from surface finishes is the dominant factor in signal degradation when the clock frequency is within the microwave frequency region. For microstrip line, the influence of surface finishes on signal distortion is negligible; for differential mode coupled microstrip lines, however, the surface finish effects, especially those with high resistivity, can lead to significant signal distortions. These findings are expected to have strong implications when designing high-speed circuits that meet strict timing requirements.
Flip chip technologies have been used extensively for the processors of mainframe computers, servers, personal computers, notebooks, smartphones, tablets, games, etc., the application specific integrated circuits (ASICs) of networking, telecommunications, etc., and the memories of data storage devices, etc. Most of the flip chip assemblies are mass reflowed. Recently, because of the requirements of higher functionalities of the chips and shrinking the chips’ area, the number of pinout of the processors, ASICs, and memories increases and their pitch decreases. Also, because of the trends of smaller form factors for mobile and portable products, the thickness of the chips and package substrates must be as thin as possible. Higher pin counts, tighter pitches, thinner chips, and thinner package substrates lead to the necessity of the TCB or even the hybrid bonding methods of flip-chip assemblies. In this chapter, besides mass reflow, various TCB techniques are mentioned. A high throughput LPC (liquid phase contact) TCB process is mentioned. Hybrid bonding will be discussed in Chap. 2 of this book.
Use the keratin liquid which solved from wool with urea and 2-mercaptoethanol to prepare wool keratin porous membrane at different concentration and frozen temperature. Observe the morphological structure of the membrane by SEM and analyze the SEM pictures. The experimental results indicate that the liquid concentration and frozen temperature are inversely proportional to the porous membrane aperture and porosity, but proportional to the porous membrane density.
The effect of under-bump-metallization (UBM) on electromigration was investigated at temperatures ranging from to . The UBM structures were examined: 5--Cu/3--Ni and Cu. Experimental results show that the solder joint with the Cu/Ni UBM has a longer electromigration lifetime than the solder joint with the Cu UBM. Three important parameters were analyzed to explain the difference in failure time, including maximum current density, hot-spot temperature, and electromigration activation energy. The simulation and experimental results illustrate that the addition 3--Ni layer is able to reduce the maximum current density and hot-spot temperature in solder, resulting in a longer electromigration lifetime. In addition, the Ni layer changes the electromigration failure mode. With the Cu UBM, dissolution of Cu layer and formation of intermetallic compounds are responsible for the electromigration failure in the joint. Yet, the failure mode changes to void formation in the interface of and the solder for the joint with the Cu/Ni UBM. The measured activation energy is 0.85 eV and 1.06 eV for the joint with the Cu/Ni and the Cu UBM, respectively.
Interfacial reactions between Sn-57Bi-1Ag and electroless Ni-P/immersion Au were investigated following isothermal aging at 85 and 130 degrees C. A long-term aging study confirmed two intermetallics at the solder/substrate interface. With scanning electron microscopy and energy-dispersive X-ray spectroscopy (EDX) analysis, the metastable NiSn4 phase was detected coexisting with the stable Ni3Sn4 phase. The average thicknesses of both intermetallic compounds (IMCs) were graphically plotted versus the square root of aging time. The EDX showed that Ni3Sn4 nucleated first. However, after nucleation, the IMC of the NiSn4 phase grew faster than the one of Ni3Sn4 between 85 and 130 degrees C. For both temperatures, the growth constants were calculated and the corresponding activation energies were approximated. A high volume of Kirkendall voids appeared along the Ni3Sn4/NiSn4 interface at 130 degrees C, resulting in dramatic shear strength decline.
Evolution of interfacial phase formation in Sn-3.0 Ag-0.5Cu/Cu (wt%), Sn-3.0Ag-0.5Cu-0.1Ni/Cu, Sn-3.0 Ag-0.5Cu/Cu-15Zn, and Sn-3.0Ag-0.5Cu-0.1Ni/Cu-15Zn solder joints are investigated. Doping Ni in the solder joint can suppress the growth of Cu3Sn and alter the morphology of the interfacial intermetallic compounds (IMCs), however it shows rapid growth of (Cu,Ni)(6)Sn-5 at the Sn-3.0Ag-0.5Cu-0.1Ni/Cu interface. In comparison with the Cu substrates, the Cu-Zn substrates effectively suppress the formation of Cu-Sn IMCs. Among these four solder joints, the Sn-3.0Ag-0.5Cu-0.1Ni/Cu-15Zn solder joint exhibits the thinnest IMC, and only (Cu,Ni)(6)(Sn,Zn)(5) formed at the interface after aging. It is revealed that the presence of Ni acts to enhance the effect of Zn on the suppression of Cu-Sn IMCs in the SAC305-0.1Ni/Cu-15Zn solder joint. The limited formation of IMCs is related to the elemental redistribution at the joint interfaces during aging. The Sn-3.0Ag-0.5Cu-0.1Ni/Cu-15Zn joint can act as a stabilized interconnection due to the effective suppression of interfacial reaction.
The interfacial reactions of Sn/Cu-xZn (x = 15 and 30 at.%) solder joints were investigated. Before aging, [Cu 6(Sn,Zn) 5] and [Cu 6(Sn,Zn) 5/Cu-Zn-Sn] intermetallic compounds (IMCs) formed at the [Sn/Cu-15Zn] and [Sn/Cu- 30Zn] interfaces, respectively. After thermal aging at 150 °C for 80 days, [Cu 6(Sn,Zn) 5/Cu 3(Sn,Zn)/Cu(Zn,Sn)/CuZn] and [Cu 6(Sn,Zn) 5/Cu(Zn,Sn)/CuZn] IMCs, respectively, formed at the [Sn/Cu-15Zn] and [Sn/Cu-30Zn] interfaces. Increasing the amount of Zn in the Cu-Zn substrates evidently suppresses the growth of Cu 3Sn and Kirkendall voids at the solder joint interfaces. Transmission electron microscopy images show the different microstructure of CuZn and Cu-Zn-Sn phases in Sn/Cu-Zn joints. These Cu-Zn phases act to inhibit the growth of Cu 6Sn5 and Cu 3Sn IMCs. As the content of Zn increased in Cu-Zn substrates, both CuZn and Cu(Zn,Sn) grew significantly. In addition, the growth of the Cu 6(Sn,Zn) 5/ Cu 3Sn IMCs approached a reaction-controlled process. The formation mechanisms of the CuZn and Cu(Zn,Sn) phases were probed and proposed with regard to the interfacial microstructure, elemental distribution, and the compositional variation at Sn/Cu-xZn interfaces.
The mechanical properties of new lead-free Sn–Ag–Cu solder alloys containing 0.05–0.1 wt.% boron were investigated under a range of isothermal aging and reflow conditions. The boron-doped solder joints showed higher ball pull strength than the baseline Sn-1.0Ag-0.5Cu solder joint under all isothermal aging and reflow conditions examined. In particular, the high-speed ball pull strength of the 0.05 wt.% B-doped solder joint was approximately 2.5 times greater than that of the baseline Sn-1.0Ag-0.5Cu solder joint aged at 150 °C for 200 h, which is attributed mainly to the reduced rate of grain growth in the intermetallic compound (IMC) layers of B-doped solder joints under aging conditions.
Sn3.0Ag0.5Cu solder doped with 0, 100, and 500 ppm Pd was reflowed with electroless Ni/immersion Au substrate. As Pd concentration increased in the solder, formation and growth of (Cu,Ni)6Sn5 were suppressed. After thermal aging, Cu4Ni2Sn5 and Cu5NiSn5 were observed at interface of Sn3.0Ag0.5Cu–xPd/Au/Ni systems. As compared to Cu4Ni2Sn5, more Pd dissolved in Cu5NiSn5. In addition, Pd doping enhanced the growth of Cu4Ni2Sn5 and slowed the formation of Cu5NiSn5, which would stabilize the intermetallic compound. Based on quantitative analysis by field emission electron probe microanalyzer, the correlation between Pd doping and elemental redistribution in solder joints was probed and discussed. This study described a possible mechanism of the formation of different intermetallic compounds in Pd-doped lead-free solder.
Sequential wetting reactions of molten Sn–3.0Ag–0.5Cu solder on electrolytic Ni and Ni–xPd (x=4 and 10 wt%) surface finish were investigated using the in situ monitoring of force-time measurement via the wetting balance. Significantly different kinetics was documented and the correlation between the characteristic wetting stages and the formation of interfacial intermetallic compounds were addressed and discussed. Besides, the addition of Pd tends to decrease the wetting time while increase the wetting force during soldering reaction. The wetting enhancement may be attributed to the drop of activation energy as well as interfacial tension and the change in microstructure of intermetallic compounds between solder and substrates.
This work aims to investigate the interfacial reaction under liquid reactions of Sn–3.0Ag–0.5Cu (SAC305) joints with Ni–xZn films by sputtering. The surface roughness and residual stress of the Ni–xZn film, which was regarded to an under bump metallization (UBM), were evaluated using atomic force microscope (AFM) and the curvature measurement. The X-ray diffractometry (XRD) was used to further identify microstructures. Detailed morphology of the interfacial reaction in SAC305/Ni–xZn joints was performed by a field-emission scanning electron microscope (FE-SEM) with low angle backscattered electron detector (LABE). It was demonstrated that the microstructure evolution and phase formation in the SAC305/Ni–7Zn and SAC305/Ni–20Zn joints varied from reflow time and Zn content in the UBM films. The influence of the Zn concentration in Ni–7Zn and Ni–20Zn films on the different interfacial reactions was discussed and proposed.
The interfacial reactions of Sn–3.0Ag–0.5Cu solder jointed with electroless Ni–P/immersion Au (ENIG) and electroless Ni–P/electroless Pd/immersion Au (ENEPIG) were investigated. Cu6Sn5 grew rather slower in ENEPIG samples among all aging conditions as compared with ENIG. Furthermore, the second phase, Ni3Sn4, started to form in the ENIG aged joints, but not in the ENEPIG aged joints. It was demonstrated that ENEPIG could inhibit the formation of Ni3Sn4, which further decreased the growth of columnar Kirkendall voids inside the Ni3P layer. With less voids formed in the Ni3P layer, it is expected that the reliability of ENEPIG joints would be superior to that of ENIG joints. Detailed mechanisms of Ni3Sn4 suppression and void formation were discussed and proposed.
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.
The interfacial morphologies for SnAgCu/Ni–P with various phosphorous contents were investigated by electron microscopy. The initial formation of the phosphorous-rich phase in Ni–P under-bump metallizations (UBMs) was dependent on phosphorous content, i.e., Ni3P for Ni–7 wt.%P UBM and Ni12P5 for Ni–13 wt.%P UBM. Different NixPy phases significantly affected the subsequent Ni–Sn–P formation. The growth of Ni–Sn–P could be thus controlled by altering the phosphorous content of the Ni–P UBMs. Therefore, an approach to suppress Ni–Sn–P formation is proposed.
Solder reaction-assisted crystallization of electroless Ni-P under bump metallization in the Si/SiO2/Al/Ni-P/63Sn-37Pb multilayer structure was analyzed using transmission electron microscopy, scanning electron microscopy, energy dispersive x-ray, and electron probe microanalyzer. The electroless Ni-P had an amorphous structure and a composition of Ni85P15 in the as-plated condition. Upon reflow, the electroless Ni-P transformed to Ni3Sn4 and Ni3P. The crystallization of electroless Ni-P to Ni3P was induced by the depletion of Ni from electroless Ni-P to form Ni3Sn4. The interface between electroless Ni-P and Ni3P layer was planar. From the Ni3P thickness-time relationship, the kinetics of crystallization was found to be diffusion controlled. Conservation of P occurs between electroless Ni-P and Ni3P, meaning that little or no P diffuses into the molten solder. Combining the growth rates of Ni3Sn4 and Ni3P, the consumption rate of electroless Ni-P was determined. Based upon microstructural and diffusion results, a grain-boundary diffusion of the Ni or an interstitial diffusion of the P in the Ni3P layer was proposed.
Solid-state interfacial reactions between Sn–3.5Ag solder and electroless Ni–P metallization on Cu substrate were investigated for three different Ni–P thicknesses. It was found that during interfacial reactions, Ni3Sn4 intermetallic grows at the Sn–3.5Ag/Ni–P interface along with the crystallization of electroless Ni–P layer into Ni3P compound. Additional interfacial compounds (IFCs) such as Ni–Sn–P, Cu3Sn, Cu6Sn5, (Ni1−xCux)3Sn4, and (Ni1−xCux)6Sn5 were also found to grow at the Sn–3.5Ag/Ni–P/Cu interfaces depending upon the Ni–P thickness. In the sample with thin Ni–P layer, formation of these IFCs appeared at lower aging temperature and within shorter aging duration than in the samples with thicker Ni–P. The complete dissolution of electroless Ni–P layer into Ni3P and Ni–Sn–P layers was found to be the main cause for the growth of additional IFCs. Across the Ni3P and Ni–Sn–P layers, diffusion of Cu and Sn takes place resulting in the formation of Cu–Sn and Ni–Cu–Sn intermetallics. It is shown in this paper that multi-layered IFC growth at the Sn–3.5Ag/Ni–P/Cu interfaces can be avoided by the selection of proper Ni–P thickness.
In this work, shear strengths of the solder joints for Sn–Ag eutectic alloy with the Au/electroless Ni(P)/Cu bond pad were measured for three different electroless Ni(P) layers. Sn–Ag eutectic solder alloy was kept in molten condition (240 °C) on the Au/electroless Ni(P)/Cu bond pad for different time periods ranging from 0.5 min to 180 min to render the ultimate interfacial reaction and the consecutive shear strength. After the shear test, fracture surfaces were investigated by scanning electron microscopy equipped with energy dispersed x ray. Cross-sectional studies of the interfaces were also conducted to correlate with the fracture surfaces. It was found that formation of crystalline phosphorous-rich Ni layer at the solder interface of Au/electroless Ni(P)/Cu bond pad with Sn–Ag eutectic alloy deteriorates the mechanical strength of the joints significantly. It was also noticed that such weak P-rich Ni layer appears quickly for high-P content electroless Ni(P) layer. However, when this P-rich Ni layer disappears from a prolonged reaction, the shear strength increases again. © 2003 American Institute of Physics.
The effect of Pd concentration on the soldering reaction between Ni and Sn-xPd alloys (x = 0-0.5 wt%) was investigated in this study. When the Pd concentration was low (x ≤ 0.05 wt%), the predominant reaction product was a layer of Ni 3Sn 4. In contrast, an additional (Pd,Ni)Sn 4 layer deposited over the Ni 3Sn 4 in the case of above 0.2 wt%. This microstructure evolution significantly weakened the strength of the interface, deteriorating the reliability of solder joints. A Pd-Ni-Sn isotherm simulated by the CALPHAD method was used to rationalize the above transition in the reaction product(s). © 2010 Materials Research Society.
In this paper, the effective skin depth which provides a useful evaluation for field penetration in multilayer coated conductor is proposed. The reflection on the interface between the adjacent conductors is considered in theoretical derivation. It is found that the effective skin depths of the gold and gold-nickel coated copper rapidly vary with the thickness of the outer layer (gold) when the gold thickness is less than twice the gold skin depth and achieve stabilization as the gold thickness increases to five times the gold skin depth.
Nickel-based under bump metallization (UBM) has been widely used as a diffusion barrier to prevent the rapid reaction between
the Cu conductor and Sn-based solders. In this study, joints with and without solder after heat treatments were employed to
evaluate the diffusion behavior of Cu in the 63Sn-37Pb/Ni/Cu/Ti/Si3N4/Si multilayer structure. The atomic flux of Cu diffused through Ni was evaluated from the concentration profiles of Cu in
solder joints. During reflow, the atomic flux of Cu was on the order of 1015–1016 atoms/cm2s. However, in the assembly without solder, no Cu was detected on the surface of Ni even after ten cycles of reflow. The diffusion
behavior of Cu during heat treatments was studied, and the soldering-process-induced Cu diffusion through Ni metallization
was characterized. In addition, the effect of Cu content in the solder near the solder/intermetallic compound (IMC) interface
on interfacial reactions between the solder and the Ni/Cu UBM was also discussed. It is evident that the (Cu,Ni)6Sn5 IMC might form as the concentration of Cu in the Sn-Cu-Ni alloy exceeds 0.6 wt.%.
This study provides a comparison of the influence of Pd(P) thickness on reactions during soldering with the Sn-3Ag-0.5Cu alloy.
Soldering was carried out in an infrared-enhanced conventional reflow oven, and a multiple reflow test method (up to ten cycles)
was performed. With increasing Pd(P) thickness, the (Cu,Ni)6Sn5 grew more slowly at the solder/Ni(P) interface, while the Ni2SnP/Ni3P bilayer became predominant after the first reflow. These three intermetallics, i.e., (Cu,Ni)6Sn5, Ni2SnP, and Ni3P, gradually coarsened as the number of reflow cycles increased. Furthermore, an additional (Ni,Cu)3Sn4 layer appeared between (Cu,Ni)6Sn5 and Ni2SnP, especially for the case of a thicker Pd(P) layer (0.2μm). The attachment of the (Ni,Cu)3Sn4 to the Ni2SnP, however, was fairly poor, and a series of microcracks formed along the (Ni,Cu)3Sn4/Ni2SnP interface. To quantify the mechanical response of the interfacial microstructures, shear testing was conducted at two
different shear speeds (0.0007m/s and 2m/s). The results indicated that the interfacial strength and the Pd(P) thickness
were strongly correlated.
KeywordsAu/Pd(P)/Ni(P)-Pd(P) thickness-Sn-Ag-Cu-interfacial reaction-shear test-Ni2SnP
Nickel-based under-bump metallization (UBM) has been widely used in flip-chip technology (FCT) because of its slow reaction
rate with Sn. In this study, solder joints after reflows were employed to investigate the mechanism of interfacial reaction
between the Ni/Cu UBM and eutectic Sn-Pb solder. After deliberate quantitative analysis with an electron probe microanalyzer
(EPMA), the effect of Cu content in solders near the interface of the solder/intermetallic compound (IMC) on the interfacial
reaction could be probed. After one reflow, only one layered (Ni1−x,Cux)3Sn4 with homogeneous composition was found between the solder bump and UBM. However, after multiple reflows, another type of
IMC, (Cu1−y,Niy)6Sn5, formed between the solder and (Ni1−x,Cux)3Sn4. It was observed that if the concentration of Cu in the solders near the solder/IMC interface was higher than 0.6 wt.%, the
(Ni1−x,Cux)3Sn4 IMC would transform into the (Cu1−y,Niy)6Sn5 IMC. The Cu contents in (Ni1−x,Cux)3Sn4 were altered and not uniformly distributed anymore. With the aid of microstructure evolution, quantitative analysis, elemental
distribution by x-ray color mapping, and related phase equilibrium of Sn-Ni-Cu, the reaction mechanism of interfacial phase
transformation between the Sn-Pb solder and Ni/Cu UBM was proposed.
Experiments for measuring permeability in columnar-dendritic microstructures have provided data only up to a volume fraction
of liquid of 0.66. Hence, the permeability for flow perpendicular to the primary dendrite arms in columnar-dendritic microstructures
was calculated, extending our data base for permeability to volume fractions of liquid as high as 0.98. Analyses of the dendritic
microstructures were undertaken first by detecting the solid-liquid interfaces with a special computer program and then by
generating a mesh for a finite-element fluid flow simulation. Using a Navier-Stokes solver, the velocity and pressure at the
nodes were calculated at the microstructural level. In turn, the average pressure gradient was used to calculate the Darcy
permeability. Permeabilities calculated by this versatile technique provided data at high volume fractions of liquid that
merged with the empirical data at the lower volume fractions.
The interfacial microstructure of Sn-3Ag-0.5Cu/Ni-P with various phosphorous contents was investigated by analytical transmission
electron microscopy (TEM) and field-emission electron probe microanalysis (FE-EPMA). As the Ni-Sn-P compound was formed between
the solder matrix and Ni-P under bump metallization (UBM), the so-called phosphorous-rich layer was transformed to a series
of layer compounds, including Ni3P, Ni12P5, and Ni2P. The relationship between Ni-Sn-P formation and the evolution of P-rich layers was probed by electron microscopic characterization
with the aid of the Ni-P phase diagram. In addition, the thickness of Ni-P UBM affects the Ni-Sn-P formation. Based on TEM,
selected-area diffraction, as well as FE-EPMA, the detailed phase evolution of P-rich layers in the Sn-Ag-Cu/Ni-P joint was
revealed. Moreover, in consideration of the mechanical properties of the joint, Ni-Sn-P phase formation, and fabrication feasibility
of Ni-P UBM, the phosphorous content and suitable thickness of Ni-P UBM are discussed.
KeywordsLead-free solder-P-rich phase-Ni-Sn-P phase
The glass-passivated, face-down semiconductor chip joining technology employed in IBM's SLT (Solid Logic Technology), has become not only a fundamental element in the hybrid circuitry of System/360 but also the basis for later metallurgical designs. The “flip-chip,” copper ball terminal, solder reflow technique is comprehensively reviewed and a discussion is given of its extension, through the use of ductile, all-solder terminals, to monolithic applications.
As peripheral pads in commercial chips have a pitch in the neighbourhood of 40–50 μm, a technique that could deposit solder paste directly in such pitch would be of great interest to reduce the overall cost of flip chip.This paper describes a new technique that can considerably reduce the final pitch. The main new feature of this process is that the bump pads can be built directly onto the peripheral ones. An electroplating process allows solder bump formation with a final pitch goal of 40–50 μm and after an accurate reflow process, eutectic Sn–3.5 wt%Ag solder bumps are obtained. In fact, the typical re-routing process can be eliminated and the process cost considerably reduced.
Solder reactions between SnPb and one of the four metals, Cu, Ni, Au, and Pd have been reviewed on the basis of the available data of morphology, thermodynamics, and kinetics. The reactions on both bulk and thin film forms of these metals have been considered and compared. Also the two kinds of reactions, above and below the melting point of the solder, have been considered and compared. The rate of intermetallic compound formation in wetting reactions between the molten solder and the metals is three to four orders of magnitude faster than those between the solid state solder and the metals. The rate is controlled by the morphology of intermetallic compound formation. In the wetting reaction between molten SnPb and Cu or Ni, the intermetallic compound formation has a scallop-type morphology, but in solid state aging, it has a layer-type morphology. There are channels between the scallops, which allow rapid diffusion and rapid rate of compound formation. In the layer-type morphology, the compound layer itself becomes a diffusion barrier to slow down the reaction. Similar morphological changes occur between SnPb and Au or Pd. The stability of scallop-type morphology in wetting reaction and layer-type morphology in solid state aging have been explained by minimization of surface and interfacial energies. The unusually high rate of scallop-type intermetallic compound formation has been explained by the gain of rate of free energy change rather than free energy change. Also included in the review is the use of a stack of thin films as under-bump-metallization, such as Cr/Cu/Au, Al/Ni(V)/Cu, and Cu/Ni alloyed thin films.
Experiments using eutectic Sn–3.5% Ag solder paste were conducted with the objective of examining the conjoint influence of copper particles addition and rapid cooling on microstructural development. The composite solder mixture was made by thoroughly mixing a pre-weighed amount of copper particles with a commercial Sn–3.5% Ag solder paste. The experiments were quite similar to the heating and cooling cycle of an industrial reflow soldering process. Heating of the samples was conducted in a furnace whose temperature was carefully controlled. The cooling process was conducted on a chilled aluminum block through which coolant was circulated at 0.5 °C. When the solder temperature reached 250 °C, the circulating system would turn on automatically and the sample, which is still molten, is forced to cool rapidly. Temperature records of the solder samples revealed that addition of copper particles to the eutectic Sn–3.5% Ag did not appreciably affect the heating and melting properties when compared to the unreinforced Sn–3.5% Ag counterpart. However, copper particles did change the solidification temperature of the composite solder. Detailed observations for varying amounts of copper particle addition revealed that copper particles less than 1.0 wt.% lowered the solidification temperature of the composite solder. For copper particles greater than 1.0 wt.%, the solidification temperature increased a few degrees Celsius, indicating that some of the copper particles did not completely dissolve in the Sn-dominant solder during the melting process. Results reveal that as-solidified microstructures of the eutectic Sn–3.5% Ag solder contain columnar type dendrites of the Sn-rich phase and a eutectic mixture of the Sn3Ag and Sn-rich phase located between the dendrite columns. The addition of copper particles to the eutectic Sn–3.5% Ag solder does refine the morphology of the primary phase, which is attributed to the presence and distribution of the Cu6Sn5 intermetallic in the solder matrix.
Stresses due to thermal oxidation and film deposition processes during the fabrication of a semiconductor are analysed. The strengths of plastic encapsulants, silicon chips, bonding wires and adhesive interfaces in integrated-circuit (IC) plastic packages are evaluated using fracture mechanics and other techniques. The thermal fatigue strength of solder joints is then estimated using the fatigue test data of solder specimens and simplified modelling of IC packages.
Electroless nickel plating was investigated for bumping.
Electroless nickel bumps with a height of 265 μm have been
selectively formed on aluminium bondpads. The nickel bumps have been
inner-lead-bonded by gang bonding to a tape with a thick layer of tin.
The adhesion of the bumps was investigated as a function of the
pretreatment of the bondpads. It is noted that electroless bumping
offers the greatest advantage of reducing the costs of the bumping
process because no sputtering equipment or lithography is required. This
can be very important if only small quantities of bumped dies are
required because the process is independent of substrate size.
Electroless bumping may become an alternative to conventional wafer
bumping by electrodeposition at least for certain applications
This letter presents a method to model conductor losses in transmission lines utilizing a commercial full wave solver. The lines consist of multilayered metallization with inherent surface roughness. Metal thickness is assumed to be larger than skin depth. To validate accuracy of the modeling, the measurements of a 50-Omega microstrip line and edge-coupled microstrip filter are provided, with random errors taken into account. Good correlation between the modeled results and measurements has been demonstrated from this comparison
Fundamental of interconnect and microstrip design
- M B Steer
- T Edwards
- MB Steer