Jeong-Tak Moon

Low-Stress Thermosonic Copper Ball Bonding

Thermosonic ball bonding processes on test chips with Al metallized bonding pads are optimized with one Au and two Cu wire types, all 25 mum diameter, obtaining average shear strengths of more than 120 MPa. The process temperature is ~110degC. Ball bonds made with Cu wire show at least 15% higher shear strength than those made with Au wire. The estimated maximum shear strength cpk value determined for Cu ball bonding (cpk = 3.7 plusmn 1.2) is almost 1.5 times as large as that of the Au ball bonding process (cpk = 2.3 plusmn 0.9), where LSL is 65.2 MPa. However, the ultrasound level required for Cu is approximately 1.3 times than that required for Au. Consequently, about 30% higher ultrasonic forces induced to the bonding pad are measured using integrated real-time microsensors. The accompanying higher stresses increase the risk of under-pad damage. One way to reduce ultrasonic bonding stresses is by choosing the softer of the two Cu wire types, resulting in a measured ultrasonic force reduction of about 5%. A second way is to reduce the ultrasound level. While this causes the average shear strength to fall by 15%, the ultrasonic force falls by 9%. The cpk value does not change significantly, suggesting that a successful Cu ball bonding operation can be run with about 0.9 times the conventionally optimized ultrasound level. The process adjusted in this way reduces the extra stress observed with Cu wire compared to that observed with Au wire by 42%.

Reliability study of low cost alternative Ag bonding wire with various bond pad materials

There have been many studies on replacing Au bonding wire with other bonding wire materials, because the cost of Au has dramatically increased by approximately 2~3 times in recent years. When replacing part of the bond pad with a noble metal, Ag bonding wire is of particular interest due to its superior electrical properties, lower cost and similar mechanical properties as compared with Au. Ag bonding wire is thermosonically bonded to 3 kinds of bond pad (Al, Au and Pd) and aged at high temperature (175°C). Then, the bondability and interface reactions are characterized at each bond pad. In the case of Ag-Al bonding, 2 kinds of intermediate phases were observed and the composition ratios of Ag and Al in these phases were 4:1 and 2:1, respectively. After 300 hrs of aging, cracks were formed in these intermediate phases and ball-lift failure occurred. However, in the case of the noble metal bond pad, a solid solution was formed between the Ag wire and bond pad and no voids or cracks were formed. This shows the robust bonding characteristics. The diffusion layer was observed and the diffraction pattern was analyzed by TEM (Transmission Electron microscopy). The Au-Al bond reliability was also characterized by a comparative study. In this study, Ag bonding wire is proposed as an alternative to Au bonding wire for noble metal pads. Also, the thermal reliability is reviewed and the failure mechanisms are verified with various bond pads.

Effects of fine size lead-free solder ball on the interfacial reactions and joint reliability

To satisfy the requirements of the newest mobile product, smaller solder ball is needed for electronic packages such as chip scale package (CSP). The purpose of this study was to investigate the effect of solder ball size on the interfacial reaction. In this study, 400, 300, and, 200 μm diameter solder balls of 98.5 wt.% Sn, 1 wt.% Ag, and 0.5 wt.% Cu(SAC105) were formed on the Cu/OSP and electroplated Ni/Au pads at 250°C. In the case of the reaction with Cu/OSP, as solder ball size decreased, the thickness of Cu<inf>6</inf>Sn<inf>5</inf> intermetallic compounds (IMCs) increased and Cu concentrations of solder balls increased. In the case of fine solder ball, Cu dissolution rate into molten solder was faster than large solder ball due to its high pad area/solder volume ratio (A/V ratio). Fast Cu dissolution into molten solder promoted the IMC growth. In the case of the reaction with electroplated Ni/Au, (Cu,Ni)<inf>6</inf>Sn<inf>5</inf> IMCs were only formed in the interface of 400 μm solder ball. Both (Cu,Ni)<inf>6</inf>Sn<inf>5</inf> and (Ni,Cu)<inf>3</inf>Sn<inf>4</inf> IMCs were formed at the interface of 300 μm solder ball. (Ni,Cu)<inf>3</inf>Sn<inf>4</inf> IMCs were only formed at the interface of 200 μm solder ball. The absolute Cu contents of 200 μm solder ball was smaller than 300 and 400 μm solder ball due to its smaller volume. (Ni,Cu)<inf>3</inf>Sn<inf>4</inf> IMCs were formed by the insufficient Cu contents in 200 μm solder ball. The substrate design with smaller A/V ratio and the SAC alloy with higher Cu contents (>0.5 wt.%) were recommended to reduce the IMC growth.

Reduction of underpad stress in thermosonic copper ball bonding

Ball bonding processes on test chips with Al metallized bonding pads are optimized with one Au and two Cu wire types, all 25 mum diameter, obtaining average shear strengths of more than 120 MPa. The process temperature is about 110degC. The results demonstrate that ball bonds made with Cu wire show at least 15% higher shear strength than those made with Au wire. The estimated maximum shear strength cpk value determined for Cu ball bonding (cpk = 3.7 plusmn 1.2) is almost 1.5 times as large than that of the Au ball bonding process (cpk =2.3 plusmn0.9), where LSL is 65.2 MPa. However, the ultrasound level required for Cu is approximately 1.3 times that required for Au. Consequently, about 30% higher ultrasonic forces induced to the bonding pad are measured using integrated real-time microsensors. The higher stresses increases the risk of underpad damage. One way to reduce ultrasonic bonding stresses is by choosing the softer of the two Cu wire types, resulting in a measured ultrasonic force reduction of about 5%. A second way is to reduce the ultrasound level. While this causes the average shear strength to fall by 15%, the ultrasonic force falls by 9%. The cpk value does not change significantly, suggesting that a successful Cu ball bonding operation can be run with about 0.9 times the conventionally optimized ultrasound level. The process adjusted in this way reduces the extra stress observed with Cu wire compared to that observed with Au wire by 39%. Hence, significantly lower than optimized ultrasound levels can be used in a Cu wire bonding process to obtain cpk values higher than that of a comparable Au wire bonding process while reducing the risk for underpad damage.

Explaining Nondestructive Bond Stress Data From High-Temperature Testing of Au-Al Wire Bonds

The application of an alternative method of bond monitoring during high-temperature aging is reported using a custom made test chip with piezoresistive integrated CMOS microsensors located around test bond pads. The sensor detects radial stresses originating from the bond pad and can resolve changes because of intermetallic compound (IMC) formation, voiding, or crack formation at the bond interface. Optimized Au ball bonds are aged for over 2000 h at 175 °C. It is found that stress sensors next to the bonds are capable of showing the stages of IMC growth, consumption of pad Al layers, and monitoring the formation of low-density and Al-rich IMC (AuAl2) which shows an advanced stage of aging. In particular, a first stress signal increase corresponds to the conversion of all Al above the diffusion barrier into IMCs. The second increase in stress signal after a period of stability corresponds to conversion of all Al below the barrier into IMCs. The IMC formation in these periods causes shear strength increase. After complete bond Al consumption, the bond, however, reaches maximum strength. As bond degradation starts, e.g., by lateral IMC formation, voiding, and oxide formation, as well as because of lateral pad Al transformation to IMC, the signal exhibits a strong decrease. The findings are corroborated by results obtained from classical methods such as interruptive or destructive testing including visual inspection, shear testing, cross sectioning, and by bond resistance monitoring.

Effect of Sb addition in Sn-Ag-Cu solder balls on the drop test reliability of BGA packages with electroless nickel immersion gold (ENIG) surface finish

Recently, Sn-Ag-Cu solders have been widely used as lead-free candidates for the Ball-Grid-Array (BGA) interconnection in the microelectronic packaging industry. However, widely used Sn-Ag-Cu solders such as with 3.0-4.0 wt% Ag in microelectronics exhibit significantly poorer drop test reliability than SnPb solder due to the low ductility of Sn-Ag-Cu solder bulk. The brittle failure of solder joints occurs at intermetallic compound (IMC) layer after drop test. Because the brittle nature of IMC or defects around IMC transfers a stress to the interfaces as a result of the low ductility of solder bulk. For the improvement of the drop test reliability by solder alloys, the low ductility of solder bulk and the IMC control at the interface are needed. In this paper, the bulk property of solder alloys and interfacial reactions with ENIG of Sb-added Sn-Ag-Cu solder were studied and finally, drop test was performed. Low Ag solder such as Sn1.0Ag0.5Cu and Sn1.2Ag0.5Cu0.5Sb showed higher ductility than high Ag solder such as Sn3.0Ag0.5Cu. In the interfacial reaction, all of the solders had (Cu,Ni)6Sn5 IMCs and P-rich Ni layer, however, Sn1.2Ag0.5Cu0.5Sb solder showed the lowest P-rich Ni layer thickness, because less Ni participated in the formation of (Cu,Ni)6Sn5 IMCs. In the drop test, the longer lifetime was in order of Sn1.2Ag0.5Cu0.5Sb, Sn1.0Ag0.5Cu, and Sn3.0Ag0.5Cu. Sn1.2Ag0.5Cu0.5Sb solder showed the best drop test reliability compared with other two solders due to the thinnest P-rich Ni layer. The failures of all packages occurred along P-rich Ni layer which is the most brittle phase at the solder/ENIG interface.

The behavior of FAB (free air ball) and HAZ (heat affected zone) in fine gold wire

The trend of high integration and miniaturization of semiconductors has accelerated the development of gold bonding wire with smaller diameter. For the stable bonding of fine wire, it is important to characterize the wire with various diameters during the bonding process. To investigate this relationship, the experiments were done for the various sizes of wire diameter and FAB. The wire size and the FAB size were chosen for the test from 15 /spl mu/m to 25 /spl mu/m and from 1.4 WD (wire diameter) to 2.0 WD, respectively. The results showed that as the size of FAB became smaller, the size deviation of FAB increased and FAB itself was tilted to one side. When FAB was formed at the same parameter, the length of HAZ became shorter for the wire with the high temperature of recrystallization. It is also revealed that the length of HAZ decreased for the smaller size of FAB. This phenomenon is considered to be related to the beat generated during the FAB formation.

Effect of Pd addition on ultra-fine pitch Au wire/Al pad interface

Due to increasing cost of gold and technological demand for finer pitch application, Au bonding wires with finer diameter are needed. In such demands, Au bonding wire manufacturing companies produced ultra-fine Au wires with diameter of 12 μm (0.5 mil). However, interfacial reactions between ultra-fine Au wires and Al pads may cause serious reliability problems as wire diameter get smaller. Decrease in bond reliabilities may be caused by interfacial failures such as Kirkendall voids and oxidation of Au4Al IMCs. For the application of ultra-fine Au wires in real products, studies on interfacial reactions of ultra-fine Au wires/Al pads and effects of Pd addition on Au wire/Al pad interface which is previously proven to be enhancing bond reliabilities are necessary. In this study, interfacial reactions and failures of two types of 4N ultra-fine Au wires(12 μm)/Al pads with different bonding types were analyzed, and effects of Pd addition on interfacial reactions and failures are also investigated by using 3000, 6000, and 10000 ppm of Pd added Au wires. High temperature storage test (HTST) at 175D under air-ambient was performed to simulate an accelerated thermal exposure characteristic of device operation conditions. Intermetallic compound (IMC) phases and Pd behavior at the Au wire/Al pad interface were identified using various analytical tools. Ball pull test (BPT) was conducted to evaluate bond reliabilities depending on bonding types, wire compositions, and aging times. According to experimental results, two types of ultra-fine Au wires/Al pads showed different bonding reliabilities depending on initial inter-metallic coverage. It was found that initial intermetallic coverage determined the interface bond reliability because interfacial oxidation problem depended on initial intermetallic coverage. Addition of Pd on Au wires/Al pads significantly affected bond reliabilities depending on Pd contents. 3000 ppm of Pd added Au wires/Al pads resulted the best bond reliabilities and strongest resistance against the interfacial oxidation due to the Pd accumulation at Au4Al grain boundaries. However, 10000 ppm of Pd added Au wires/Al pads showed low bond reliabilities because separated Pd-rich island grains formed among Au4Al grains resulting in less protection of Au4Al from oxidation.

Pd effect on reliability of Ag bonding wires in microelectronic devices in high-humidity environments

We investigated the effect of Pd concentration in Pd-doped Ag wires on the humidity reliability and interfacial corrosion characteristics between Ag wire and Al metallization. Additionally, we confirmed no corrosion problem between Ag wire and noble metal (Pd, Au) metallization, even after a pressure cooker test (PCT). The chemical composition of the tested Ag wires was pure Ag, Ag-1wt% Pd and Ag-3wt% Pd. These wires were bonded to Al and noble metal (Au, Pd) metallization using a thermo-sonic bonder. The interfaces were characterized by focused ion beam (FIB), high resolution transmission electron microscope (HRTEM) and energy dispersive X-ray spectroscopy (EDS). The interface corrosion of Pd doped Ag wires was significantly reduced as the Pd concentration in the Ag wires increased. Furthermore, the Ag wires on the noble metal (Au, Pd) metallization exhibited stable reliability during the PCT.

Assessing Au-Al wire bond reliability using integrated stress sensors

Wire bond reliability testing typically consists of aging bonds in a high temperature environment for long time periods, removing samples at intervals to assess bond shear strength and characterize the bond cross sections. In this way, the degradation of the bond can be monitored at discrete time intervals, and it is determined whether the bond will be reliable under long term operation at lower temperatures. This process is labour and time consuming. An alternative process is reported using piezoresistive integrated CMOS microsensors located around test bond pads. The sensors are sensitive to radial compressive or tensile stresses occurring on the bond pad due to intermetallic formation, voiding, and crack formation at the bond interface. Two sets of identical test chips are bonded with optimized Au ball bonds and aged for 2000 h at 175 °C. One set is connected to equipment which monitors signals from the stress sensors, along with the contact resistance of the bonds. The other set of chips is destructively tested by shear tests and cross sectioning. It is found that the stress sensors are capable of indicating which stage of intermetallic growth is currently being experienced, by relating the signal to the relative density of the intermetallic compounds (IMCs) which form during aging. The sensors can detect the consumption by IMCs of each Al layer in a multilayer pad, and can monitor the formation of AuAl2 which indicates an advanced stage of aging. Sensor signals combined with contact resistance measurements provide a valuable tool for preliminary reliability studies, and give real-time insight into microstructural changes. Drop in shear strength of a ball bond is detected by a change in the microsensor signal combined with a contact resistance increase.