Comparative performance of gold wire bonding on rigid and flexible substrates

This paper reports comparative performance of wire bondability of electrolytically plated Au/Ni/Cu bond pads on rigid FR-4 and bismaleimide trazine (BT) PCBs, as well as flexible polyimide (PI) substrate. The metallization surfaces were treated with plasma to study the effect of bond pad surface cleanliness on wire bondability. Process windows were constructed as a function of bonding temperature and bond power for the individual substrate materials. Significant improvements of wire pull strength and process window were noted after plasma treatment with a substantial reduction in minimum bonding temperature from 120°C to 60°C for both the rigid and flexible substrates. The minimum bond power required to produce successful bonds decreased with increasing bonding temperature. At a bonding temperature of 120°C, the process window for the flexible substrate was wider than the rigid substrates. The wire bondability and wire pull strength of rigid substrates decreased with increasing bonding temperature above 120°C due to softening of the substrate which adversely affected the effective bond force and the transmission of ultrasonic energy. In contrast, the wirebonding performance of the flexible substrate remained stable at 120°C or above because the thermo-mechanical properties of flexible PI substrate were rather insensitive to temperature. The process windows of flexible substrates with and without stiffener showed similar bondability.     

Fundamentals of thermo-sonic copper wire bonding in microelectronics packaging

Fine copper wire bonding is capable of making reliable electrical interconnections in microelectronic packages. Copper wires of 0.8–6 mil diameter have been successfully bonded to different bond pad metallized and plated substrate materials such as Al, Cu, Ag, Au and Pd. The three metallurgical related factors; solid-solubility and diffusion of dissimilar contact metals, oxide film breakage and plastic deformation of asperities play a critical role in the bonding. Plastic deformation of an asperity is the most significant factor one has to consider to attain good bonding. Soft aluminum metal (30–40 VHN), with a lower % asperity threshold deformation is easier to wire bond than harder metallic surfaces (Ni, W, Mo, Cr, Co, Ta) of 150–500 VHN. Good adhesion of wire bonding is achieved for the surface roughness (Ra) of 0.01–0.15 μm and 0.02–0.6 μm of bare and plated surfaces respectively. It is rationalized that the application of ultrasonic energy principally breaks the oxide film and deform the asperities, while a compressive force increases the proximity of asperities. Hence wire welds to bond pad surface by molecular attraction and inter diffusion. Storage of copper ball bonds at 175 °C for 100–1,000 h forms copper aluminide at the interface. EDAX and Auger analysis reveal 22 at% Al + 78 at% Cu composition of the aluminides and Cu₃Al₂ empirical formula is calculated, which, does not match with any of the reported copper aluminides. Hardness of the copper ball bonds and stitch bonds are higher than wire exhibiting work hardening of the bonds on processing.     

Further characterization of Pd-coated copper wire on wirebonding process: from a nanoscale perspective

The objective of this research is to investigate the nanomechanical properties of ultra-thin Pd-coated copper (Cu) wire (ψ = 0.6 mil) and nanotribology along the interfacial between free air ball (FAB) and aluminum (Al) bond pad during wirebonding process. Two major analyses are conducted in the present paper. In the first, the characteristic of heat affected zone and FAB for Cu wire has been carefully experimental measured. Nanoscale interfacial tribology behavior between Cu FAB and Al pad is examined by atomic force microscopy. Secondary, the dynamic response on Al bond pad and beneath the pad during wirebonding process has been successfully predicted by finite element analysis. Micro-tensile mechanical properties of Cu wire before and after electric flame-off (EFO) process have been investigated by self-design pull test fixture. Experimental obtained hardening constant in Hall–Petch equation has significantly influence on the localize stressed area on Al pad. This would result in Al pad squeezing (large plastic deformation) around the smashed FAB during impact stage and the consequent thermosonic vibration stage. Microstructure of FAB is also carefully investigated by nanoindentation instruments. A real-time secondary EFO scheme has been conducted to reduce the strength of Cu wire and increase the bondability. All the measured data serve as material inputs for the finite element model based on explicit software ANSYS/LS-DYNA. In addition, nanoscale bondability on Cu-Al intermetallic compound is simulated by molecular dynamics. A series of comprehensive parametric studies were conducted in this research.     

An analysis of intermetallics formation of gold and copper ball bonding on thermal aging

In IC packages, thermosonic wire bonding is the preferred method for making electrical connections between the die pad and lead frame. These interconnect are made using fine metal wires. On thermal aging (under 175 °C for 5 h) gold aluminide easily forms in gold ball bonds while formation of intermetallic compound is absent in case of copper ball bonds. An analysis of the atomic property of the elements bonded explains that atomic radii and electronegativity factors favor gold aluminide formation.     

Illumination system for wire bonding inspection

A novel lighting system was devised for 3D defect inspection in the wire bonding process. Gold wires of 20 microm in diameter were bonded to connect the integrated circuit (IC) chip with the substrate. Bonding wire defects can be classified as 2D type and 3D type. The 2D-type defects include missed, shifted, or shorted wires. These defects can be inspected from a 2D top-view image of the wire. The 3D-type bonding wire defects are sagging wires, and are difficult to inspect from a 2D top-view image. A structured lighting system was designed and developed to facilitate all 2D-type and 3D-type defect inspection. The devised lighting system can be programmed to turn the structured LEDs on or off independently. Experiments show that the devised illumination system is effective for wire bonding inspection and will be valuable for further applications.     

Irregular Characteristics of Bond Interface Formation in Ultrasonic Wire Wedge Bonding

The mechanism of ultrasonic wire wedge bonding, one of the die/chip interconnection methods, was investigated based on the characteristics of the ultrasonic wire bonding joints. The Al-1%Si wire of 25 μm in diameter was bonded on Au/Ni/Cu pad and the joint cross-section was analyzed by scanning electron microscopy （SEM） and energy-dispersive X-ray spectroscopy （EDX）. The results indicated that it is irregular for the ultrasonic bond formation, non-welded at the centre but joining well at the periphery, especially at the heel and toe of the joint. Furthermore, the diffusion and/or reaction at the cross-section interface are not clear at C-zone, while there exists a strip layer microstructure at P-zone, and the composition is 78.96 at. pct Al and 14.88 at. pct Ni, close to the Al3Ni intermetallic compound. All these observations are tentatively ascribed to the plastic flow enhanced by ultrasonic vibration and repeated cold deformation driving interdiffusion between AI and Ni at bond interface.     

Fabrication and mechanical characterization of 3D woven Cu lattice materials

3D metallic lattices designed to have two distinctly different material architectures have been woven with metallic Cu wires. A vacuum soldering technique was employed to metallurgically bond the wire nodes and form stiff 3D lattice materials. The structures and mechanical properties of the as-woven and soldered lattices were characterized by optical microscopy and micro-scale mechanical property experiments. The measured in-plane shear stiffness shows good agreement with predictions from finite element (FE) models that account for variations in the manufacturing and solder bonding. The study indicates that stiffness is influenced by the percentage of bonded nodes and the location of bonding. The 3D woven lattice materials manufactured in this study exhibited a very high percentage (80%) of bonded nodes and a unique combination of stiffness and density as compared to that typically reported for ultra lightweight lattice materials.     

TEM microstructural analysis of As-Bonded Al–Au wire-bonds

In this study the interface morphology of a model 99.999% (5N) Au wire bonded to Al pads in the as-bonded state was examined by scanning/transmission electron microscopy with energy dispersive spectroscopy. Specimens for transmission electron microscopy were prepared using the lift-out method in a dual-beam focused ion beam system. Analysis of the bond microstructure was conducted as a function of the Al pad content and as a function of the bonding temperature. Additions of Si and Cu to the Al pad affect the morphology and the uniformity of the interface. A characteristic-void line is formed between two intermetallic regions with different morphologies in the as-bonded samples. According to the morphological analysis it was concluded that a liquid phase forms during the bonding stage, and the void-line formed in the intermetallic region is the result of shrinkage upon solidification and not the Kirkendall effect.     

Bond pad cratering study by reliability tests

Silicon cratering is one of the major obstacles to turn thermosonic copper wire bonding technology into a mass-production mode. The effects of reliability tests, i.e. aging test, temperature cycle test, and autoclave test on silicon craters of copper and gold wire-bonded Si substrate are discussed in this paper. Prior to reliability tests, wire-bonded specimens were examined by initial bond etch test. Results showed that there was no crack or silicon cratering observed in the selected samples. It was found that reliability tests, unlike improper bonding parameters, did not induce silicon craters at the Si substrate. However, reliability tests degraded the Si substrate and bonded interfaces, which were exposed later during destructive tests, i.e. wire pull test and ball shear test.     

Parametric investigation of laser diode bonding using eutectic AuSn solder

Based on an optimized bonding technique for semiconductor lasers, parametric investigations of different bonding configurations were performed using AuSn solder. No abrupt electrical degradation was observed for both face-up and face-down bonding configurations. The typical optical output achieved for face-up and face-down bonding approach was about 100 mW/facet and 150 mW/facet, respectively. Face-up bonded LDs exhibited a steady differential quantum efficiency η_D of ∼ 0.425 mW/mA before gradual degradation at 200 mA, while face-down bonded LDs improved its performance beyond 350 mA. The characteristic temperature T₀ also improved to as much as 643 K for face-down bonded LDs. Spectrally-resolved emission measurement showed that the temperature rise in unbonded, face-up bonded and face-down bonded LDs were approximately 11, 7-8 and 2-3 °C, respectively. These investigations showed that the improved performances for face-down bonding approach compared to face-up approach were due to better thermal management.     

Neuartige Faserverbunddrähte für die Leistungselektronik

Novel fibre reinforced wires for power electronics The use of power electronics within the scope of mechatronic applications as well as the increasing integration of components lead to increased requirements concerning their mechanical and thermal reliability. Today contact making in power electronics is mostly done by aluminum thick wire bonding. This process is highly productive, however the life time of power electronic components is meanwhile predominantly limited by the durability of these wire bonds. The thermal mismatch between the wire material and the connected components is one cause. A new starting point, in order to improve the reliability, is the application of new fibre reinforced metal matrix composite (MMC) wires with increased reliability under thermo‐mechanical stress. In the context of a research project MMC bond wires of different material combinations and arrangements were manufactured. Aluminum wires with copper fiber reinforcement as well as Copper wires containing FeNi36 fibre reinforcement have successfully be drawn to a final diameter of 300 μm. The fibre reinforcements should reduce the coefficient of thermal expansion and improve the mechanical strength. By aluminium copper MMC the electrical conductivity is increased as well. Measurements of the produced MMC wires confirmed these expectations. The manufacturing of the MMC took place on the basis of wire material of different diameters. These wires were stacked in capsules in different arrangements and material combinations. Subsequently, the capsules were either hot‐isostatically pressed or directly extruded. In such a way produced composites have been manufactured by rotary swaging and wire drawing into bond wires and after that tested.     

Fabrication and mechanical properties of steel–steel composites

Metal matrix composites based on a low carbon steel matrix reinforced with high carbon steel wires have been fabricated by a combined cold and hot rolling process. Both continuously and discontinuously aligned composites have been produced. A subsequent heat treatment allowed the formation of martenisitc, bainitic or pearlitic wires in a ferrite predominantly matrix. The optimum wire microstructure giving a composite with high strength and reasonable ductility was found to be bainitic — martensitic wires were found to contain microcracks that gave poor composite strengths and ductilities. The discontinuous wire composites produced similar strengths to the continuous composites only when they were deformed to give a wire aspect ratio greater than 20. The strengths of both types of composites showed a good fit to the rule of mixtures as the volume fraction of fibers was increased.     

The effect of metal wires on the fracture of a brittle-matrix composite

The behaviour of stainless steel, work-hardened nickel and annealed nickel wires bridging a crack in a brittle-matrix has been studied as a function of the length and orientation of the wire. The pull-out stress for stainless steel wire in epoxy resin increases less than linearly with wire length, following the behaviour predicted by Takaku and Arridge [6]. Wires inclined at 20° and 40° to the tensile axis gave pull-out stresses some 30% higher than wires parallel to the tensile axis, this increase being attributed mainly to enhanced friction on the bent wire near its point of exit from the matrix. Work-hardened nickel wires fractured when their length exceeded a critical value, and the critical length was significantly shorter for inclined wires than for wires parallel to the tensile axis. In contrast, annealed nickel wires, no matter how long, did not fracture but pulled out at a limiting stress which was slightly higher for inclined wires than for wires parallel to the tensile axis. The results show that, in some cases, there does not exist a critical length above which an embedded wire will fracture rather than pull out of the matrix.     

Performance of laser bonded glass/polyimide microjoints in cerebrospinal fluid

In this paper, laser bonded microjoints between glass and polyimide is considered to examine their potential applicability in encapsulating neural implants. To facilitate bonding between polyimide and glass, a thin titanium film with a thickness of 2 μm was deposited on borosilicate glass plates by a physical vapor deposition (PVD) process. Titanium coated glass was then joined with polyimide by using a cw fiber laser emitting at a wavelength of 1.1 μm (1.0 W) to prepare several tensile samples. Some of the samples were exposed to artificial cerebrospinal fluid (aCSF) at 37^∘C for two weeks to assess long-term integrity of the joints. Both the as-received and aCSF soaked samples were subjected to uniaxial tensile loads for bond strengths measurements. The bond strengths for the as-received and aCSF soaked samples were measured to be 7.31 and 5.33 N/mm, respectively. Although the long-term exposure of the microjoints to aCSF has resulted in 26% reduction of bond strength, the samples still retain considerably high strength as compared with the titanium-polyimide samples. The failed glass/polyimide samples were also analyzed using optical microscopy, and failure mechanisms are discussed. In addition, a two dimensional finite element analysis (FEA) was conducted to understand the stress distribution within the substrate materials while the samples are in tension. The FEA results match reasonably well with the experimental load-displacement curves for as-received samples. Detailed discussion on various stress contours is presented in the paper, and the failure mechanisms observed from the experiment are shown in good agreement with the FEA predicted ones.     

镍铝复合丝和镍铝合金丝及其涂层

介绍了镍铝复合丝和镍铝合金丝及其涂层的成分、相结构 ,比较分析了涂层的结合强度 ,分析了涂层自结合作用机理 ,估算了喷涂一平方米面积所消耗的丝材及相应用丝材成本 ,介绍了两种材料适用的对象及设备     

End Effects in the Three-Omega Method to Measure Gas Thermal Conductivity

A two-dimensional analytical solution is derived for the three-omega method for measurement of thermal conductivity of materials with a fine wire. The analytical solution includes the wire heat capacity and the effect of heat losses from the ends of the wire. To derive the solution, finite Fourier transforms are applied in the direction parallel to the wire axis. The solution is compared with a one-dimensional solution and experimental data. It is found that heat losses from the wire ends have a significant effect on the 3ω components at low frequency and tend to be less important at high frequency. Moreover, it is shown that two-dimensional effects will be severe for nano-scale wires, even if the wire length-to-diameter ratio is very large.     

Effect of welding parameters on diffusion bonding of type 304 stainless steel–copper bimetal

Abstract

Stainless steel AISI type 304 and electrolytic cold rolled copper were joined by diffusion bonding at temperatures ranging from 650 to 950°C, for times from 5 to 45 min, and at pressures from 2 to 12 MPa. After bonding the microstructure of the interface was investigated, including the grain size, and shear and tensile strengths of the bonded specimens were determined. From the results, it was seen that the bond shear strength was dependent on interface grain boundary migration and on grain growth during the bonding process. In addition, attempts were made to find a relationship between grain size and shear strength in the bonding area. Taking into account the results of shear testing and microstructural observation, for a sound bond, optimum bonding conditions were obtained at temperatures of 800–850°C for 15–20 min at 4–6.5 MPa. The fracture behaviour of the diffusion bonded joint was investigated by means of shear and tensile testing under different bonding conditions. It was found that both shear and tensile strengths of the bonds were sensitive to the bonding conditions, and the intermetallic phases did not affect these parameters. Furthermore, the value of shear strength of the bond surface determined by shear testing was higher than the shear strength of the fracture surface determined by tensile testing.     

Reaction chemistry at joined interfaces between silicon nitride and aluminium

Joined interfaces of HIPed additive-free silicon nitride ceramics/aluminium braze bonded at a low temperature of 1073 K for 18 ks or at a high temperature of 1473 K for 1.8 ks in vacuum of 1.3 mPa and of β silicon nitride powders/aluminium powders bonded at the low temperature for 1.8 ks or 18 ks in the same vacuum are identified by analytical transmission electron microscopy and X-ray diffraction method. Mullite, some small crystals and β′-sialon are detected at the interface of the ceramics/aluminium braze bonded at the low temperature and 15R AIN-polytype sialon, β′-sialon, aluminium nitride, mullite and silica-alumina noncrystalline are detected at that bonded at the high temperature. At the interface of the two kinds of powders, aluminium nitride and silicon are also detected besides β′-sialon and silica-alumina noncrystalline even though the bonding was conducted at the low temperature. The interfacial reactions of the joints are influenced not only by bonding temperature but also by the oxide formed at the interface before bonded.     

Pass schedule of wire drawing process to prevent delamination for high strength steel cord wire

The high-speed drawing of high carbon content steel wires is usually conducted at room temperature employing a number of passes or reductions through several dies. In the multipass drawing process, the temperature rise at each pass affects the mechanical properties of the final product (such as its bending and torsion properties, and its tensile strength). This temperature rise during deformation encourages delamination in the wire, which has a deleterious influence on the torsional properties and durability of the wire. In this study, we investigated the delamination of wires using torsion tests and evaluated the wire temperature during drawing. Our data shows that one of the main reasons for delamination was an excessive rise in wire temperature. Based on our experimental results, in order to prevent delamination due to an excessive rise in wire temperature, a new isothermal pass schedule that could control the wire temperature was designed. The pass redesign for the conventional high carbon (0.75 wt%C) steel cord wire drawing process with delamination was carried out by using the isothermal pass schedule to control the wire temperature. In order to verify the effectiveness of the proposed method, wire drawing and torsion test were conducted. From the results of experiments, it was possible to produce high carbon steel cord wire without delamination.     

Production of stainless steel wire-reinforced aluminium composite sheet by explosive compaction

Composite sheet material has been produced by explosively compacting stacks of alternately placed stainless steel wire meshes and aluminium foils. It was found that stacks could be satisfactorily bonded by using an aluminium driver plate which was prevented from bonding to the stack by interposing a polythene sheet. Stacks containing six or seven layers of mesh and having a wire volume fraction of up to 0.24 could be bonded when the driver plate kinetic energy exceeded 120 J cm^–2. It is concluded that the bonding mechanism involves cold pressure welding of the matrix metal by extrusion through the mesh apertures, and the aperture size is a controlling factor in bonding. No evidence was found of strong bonding between the wires and the matrix. In the production of larger sizes of composite sheet, (300 mm×500 mm), blistering and tearing occurred due to the presence of excess air in the stack, a consequence of bowing of the foils by the springy and curved pieces of mesh. This difficulty was overcome by enclosing the stack in a polythene envelope, which was evacuated before detonation of the charge, so that the stack was compressed by atmospheric pressure. Tests have shown that tensile and fatigue properties of the composites compare favourably with other aluminium matrix composites and with high strength aluminium alloys.