Optimization Of Bonding Process Research Articles

An interpretable machine learning model for predicting bond strength of CFRP-steel epoxy-bonded interface

This study develops an interpretable machine learning model for predicting the bond strength of CFRP-steel epoxy bonding interfaces and reveals key bond parameters. A total of 302 sets of experimental data were collected from the existing literature, including 16 influencing factors (i.e. the “features”) for the bond strength of the CFRP-steel bonded interface. Firstly, the sequential backward selection algorithms based on random forest (RF), maximum information coefficient (MIC), and distance coefficient (DC) were assessed. Then, Catboost and RF-based models were optimized and evaluated to determine the best prediction model. Finally, Shapley Additive Explanation (SHAP) was used to interpret the Catboost-based model. The results showed that the average Coefficient of Determination (R2) of the prediction for the test set by the Catboost and RF-based models was 95.5 % and 91.5 %, respectively, indicating the Catboost-based model has a higher prediction accuracy. The SHAP analysis indicates that the bond length, bond width, and the elastic modulus of the adhesive are the critical bonding parameters governing the bond strength of CFRP-steel epoxy-bonded interfaces. The Pareto diagram shows that the effective bond length is about 120 mm. This study provides valuable guidance for the design of bond performance and process optimization of CFRP-steel epoxy bonding interfaces.

Laser Soldering Process Optimization of MEMS Probe of Probe Card for Semiconductor Wafer Test

In this study, the laser soldering process of probe-multilayer ceramic (MLC) substrates was optimized using Type 4 (T4) and Type 7 (T7) Sn-3.0Ag-0.5Cu (SAC305), T7 Sn-0.7Cu solders to improve the high-temperature durability of probe solder joints and satisfy the fine pitch requirements of the soldering process. Thermal cycle tests (TCTs) were performed to compare the heat resistance characteristics of the probe solder joint. The results showed that the bonding strength degradation rates of the SAC305 and Sn-0.7Cu solders after the TCT were within 40% compared to that of the as-soldered state. Because the probe surface and MLC were plated with Ni/Au, various intermetallic compounds (IMCs) were observed at the solder joint. The total IMC thickness of the T7 solder joint was thinner than that of the T4 solder because of the difference in redox reactions according to the particle size of the solder paste. The fracture surface of the probe-MLC solder joint after the shear strength test exhibited a mixed ductile and brittle fracture, and that of the Sn-0.7Cu solder joint exhibited a ductile fracture with numerous dimples.

Laser Soldering Process Optimization of MEMS Probe of Probe Card for Semiconductor Wafer Test

Laser Soldering Process Optimization of MEMS Probe of Probe Card for Semiconductor Wafer Test

Co-simulation of MATLAB and ANSYS for ultrasonic wire bonding process optimization

Ultrasonic wire bonding is a solid-state joining process, used in the electronics industry to form electrical connections, e.g. to connect electrical terminals within semiconductor modules. Many process parameters affect the bond strength, such like the bond normal force, ultrasonic power, wire material and bonding frequency. Today, process design, development, and optimization is most likely based on the knowledge of process engineers and is mainly performed by experimental testing. In this contribution, a newly developed simulation tool is presented, to reduce time and costs and efficiently determine optimized process parameter. Based on a co-simulation of MATLAB and ANSYS, the different physical phenomena of the wire bonding process are considered using finite element simulation for the complex plastic deformation of the wire and reduced order models for the transient dynamics of the transducer, wire, substrate and bond formation. The model parameters such as the coefficients of friction between bond tool and wire and between wire and substrate were determined for aluminium and copper wire in experiments with a test rig specially developed for the requirements of heavy wire bonding. To reduce simulation time, for the finite element simulation a restart analysis and high performance computing is utilized. Detailed analysis of the bond formation showed, that the normal pressure distribution in the contact between wire and substrate has high impact on bond formation and distribution of welded areas in the contact area.

200 mm Ge Wafer Production for Oxide-Free Si-Ge Direct Wafer Bonding

Large diameter, dislocation-free Ge single crystalline wafers are relatively new in the semiconductors industry. In the past a few groups reported on the use of 200 mm Ge wafers for manufacturing of germanium-on-insulator (GeOI) substrates [1-3], with an oxide bonding interface, while oxide-free, electrically conductive Si-Ge wafer bonding was not addressed so far. Finally, by using Si wafers this type of bond can be used e.g. for high-efficient multi-junction solar cells and high-electron-mobility transistor (HEMT) [4,5].First the manufacturing of high quality 200 mm diameter Ge wafers was addressed. For the wafer bonding processes standard silicon wafers were bonded to germanium wafers in order to fabricate oxide-free bonds. The wafer bonding process was performed using the EVG ComBond® system: first the native oxides from both wafer surfaces were removed using ion beam sputtering process, followed by the transfer of both wafers in ultra-high vacuum (UHV) to the bonding process station. In order to prevent thermally induced stress due to the different thermal expansion coefficients the bonding process is performed at room temperature (RT).The goal of this work was to demonstrate the feasibility of the bonding process, so main focus was on mechanical and morphological analysis of the surfaces and of the bonded interfaces. Atomic force microscopy (AFM) and spectroscopic ellipsometry (SE) were used to determine the surface roughness and oxide thickness, respectively. Both parameters are crucial for successful RT Si-Ge wafer bonding. The high quality results and the bonding energy were revealed using scanning acoustic microscopy (SAM) measurements as well as the Maszara blade test. Furthermore, cross-section transmission electron microscopy (X-TEM) showed a bonding interface with an about 2.2 nm thin amorphous layer and no additional oxygen concentration (see Figure 1). In order to enhance the visibility of the oxygen and argon distributions measured with energy dispersive x-ray spectroscopy (EDXS) the detected element intensities are expressed qualitatively. Figure 1. X-TEM (left) and qualitative EDXS line scan (right) of the Si-Ge bonding interface. Here 100% represents the maximum detected intensity of each element. The work reported here is a first step in developing wafer bonding processes e.g. for solar cells fabrication using large diameter Si and Ge wafers. Future work will focus on bonding process optimization and electrical characterization of the bonded interfaces.[1] C. Deguet et al ., “Fabrication and characterisation of 200 mm germanium-on-insulator (GeOl) substrates made from bulk germanium,” in El. Let., vol. 42, no. 7, pp. 51-52 (2006).[2] T. Akatsu et al., “200mm germanium-on-insulator (GeOI) by Smart CutTM technology and recent GeOI pMOSFETs achievements,” in 2005 IEEE Int. SOI Conf. Proc. (2005): 137-138.[3] C.J. Tracy et al., “Germanium-on-insulator substrates by wafer bonding,” in Journal of El. Mat. 33, 886–892 (2004).[4] O. Höhn et al., ”Development of Germanium-Based Wafer-Bonded Four-Junction Solar Cells,” in IEEE Journal of Photovoltaics, vol. 9, no. 6, pp. 1625-1630 (2019).[5] W. Rachmady et al., “300mm Heterogeneous 3D Integration of Record Performance Layer Transfer Germanium PMOS with Silicon NMOS for Low Power High Performance Logic Applications,” in 2019 IEEE International Electron Devices Meeting (IEDM), San Francisco, CA, USA, pp. 29.7.1-29.7.4 (2019). Figure 1

Optimization of Bonding and Sizing Process for Automotive Wet Clutch Discs applying "Robust Design"

Many parts for transmitting power are incorporated in an automatic transmission of an automobile. The technical problems of the wet clutch disc (DISC) having important functions related to gear shifting are to improve the automatic gear shifting performance and to ensure the stability of the power transmission torque in the contact action between the DISCs. DISC is manufactured by adhesive bonding of friction material (Fa) and iron core plate (C/ P). In particular, the stabilization of the DISC adhesive strength and thickness accuracy suppresses fluctuations in the surface friction coefficient and is effective in reducing the shock phenomenon during shifting. In this study, the robustness of the DISC bonding sizing method by applying robust design was aimed at stabilizing the bonding strength and thickness accuracy, and the optimum processing conditions were extracted to optimize the mass production method.

Consumable and Process Improvement for Large Copper Wire Bonding

Abstract Aluminum (Al) wire has certain limitations and cannot satisfy the increased demands for power electronics interconnects. Copper (Cu) wire has many benefits and is replacing Al wire in high power, high temperature applications. Cu wire is much harder and wears the bond tool much faster than Al wire. Consumable lifetime and process stability have to be improved so that Cu wire interconnect can be accepted for mass production. The bond tool wear mechanism is investigated. Bond process parameter, bond tool tip geometry, and bond tool material affect bond tool lifetime and bond performance. By combining the bond tool tip design and bond process optimization, an ultra-shallow groove bond tool that has a groove depth (GD) of 40% wire diameter achieved producing 260,000 Cu wire bonds. The groove opening angle (GOA) affects wire confine capability of a bond tool and a larger GOA is preferred for bond tools with a shallower GD. Alternative bond tool material shows benefits of improving bond appearance and further extending bond tool lifetime.

Wedge Bond Process Optimization Method for 1.5 mil Al Wire on Al Bond Pad

Modern wedge bonders have evolved since their early inception in 1957. This paper will review the common challenges process engineers face when selecting a wedge bond machine configuration and developing robust processes. Wedge bond cases presented will show the tradeoff between process inputs and the resulting bond shapes, bond appearance of black ring, burrs, pull results, etc. The purpose of this work was to optimize the process outputs: bond shape, black ring, burrs, and pulls on a die with aluminum bond pads. Process inputs included Force, Time, and Ultrasonic Level. An aluminum wafer was used to understand the basic relationship between process parameter inputs and outputs. The learning was then applied to a die with aluminum bond pads. Examples of non-compliance and compliance will be shown to help process engineers evaluate wedge bonds and make refinements. The case studied was for an aluminum bond pad/Al wafer and 1.5 mil aluminum wire interaction that creates burrs around the bond (wire to pad interface), black ring on the bond periphery (wedge tool to wire interface) and the resulting pulls. Both the graphical and numerical results of the case study have clearly demonstrated the relationship between the typical process inputs and outputs, particularly bond shape, burrs, black ring and pulls. The findings in this study will provide a general guideline and a troubleshooting reference for wedge bonding process development.

Design and Solder Process Optimization in MID Technology for High Power Applications

Molded interconnect device (MID) technology is a key enabling technology with growing markets in automotive, communications, consumer electronics through integration with lighting and sensor technologies. The MID technology is yet to be explored for high temperature applications in automotive or consumer lighting. One of the hindering factors for such implementation in the serial production and time to market is the improper electronic and thermal packaging of the light emitting diodes (LEDs) on the MID substrates. This paper addresses the optimization of mold design, surface metallization and soldering process for the effective thermal management of the high-power LED systems. By using a simulation model, the thermal distribution and the resultant decrease in temperatures for varying forward electrical currents in high-power LEDs by design optimization is demonstrated. In addition, the optimization of solder process with respect to solder profile in context of vacuum vapor-phase soldering for void-free solder connections is discussed.

Impact Factors on Low Temperature Cu-Cu Wafer Bonding

Physical mechanism of Cu-Cu wafer bonding was studied as a base for low temperature (<200°C) wafer bonding process optimization. It was found out that for the successful bonding of two copper surfaces the following (classes of) impact factors have to be considered (I) contacting, even at atomic scale, (II) cleanliness of the surface, (III) diffusion enhancing materials properties and (IV) process conditions. From these fundamental findings subsequently low temperature metal thermo-compression Cu-Cu wafer bonding and even room temperature direct bonding was facilitated.

Influence of the fluxes properties on quality and the microstructure of lead-free solder joints executed by selective soldering

Purpose – The purpose of this paper is to investigate the influence of the properties of new compositions of fluxes for selective soldering on lead-free solder joints quality and microstructures as well as showing which flux properties are the most important. Design/methodology/approach – The three different types of fluxes were tested, which differed in composition, solids content, amount and type of activators added. The selective soldering process was done with the use of lead-free solder SAC 305. The test boards had two coatings SnCu (HASL) or Au/Ni. Basic chemical and physical properties of fluxes were examined according to the relevant standards. Different types of components from the bulky ones, demanding more heat, to the smaller ones were used during the assembly process. AOI and X-ray analyses as well as cross-sections and SEM analyses were utilized to deeply assess the quality and microstructure of the investigated solder joints. Findings – The results showed that information about density or static activity of flux is not enough for correct flux assessment. The dynamic activity of flux measured by wetting balance method is the best for this, especially in the case when there is short soldering time and heat transfer is hindered. The quality and the microstructure of lead-free solder joints are related not only with wetting properties of the flux used for soldering but also with other properties like solids content in a flux. Research limitations/implications – It is suggested that further studies are necessary for the confirmation of the practical application, especially of the reliability properties of the joints obtained with the use of the elaborated fluxes. Originality/value – The results showed that type of flux (ORL or ROL) as well as minor changes in their dynamic activity and solids content might have significant influence on quality of solder joints and their microstructure. It was noted that selective soldering is demanding technique and optimization of soldering process for different type of components and fluxes is important.

Optimized Cu-Sn Wafer-Level Bonding Using Intermetallic Phase Characterization

The objective of this study is to optimize the Cu/Sn solid–liquid interdiffusion process for wafer-level bonding applications. To optimize the temperature profile of the bonding process, the formation of intermetallic compounds (IMCs) which takes place during the bonding process needs to be well understood and characterized. In this study, a simulation model for the development of IMCs and the unreacted remaining Sn thickness as a function of the bonding temperature profile was developed. With this accurate simulation model, we are able to predict the parameters which are critical for bonding process optimization. The initial characterization focuses on a kinetics model of the Cu3Sn thickness growth and the amount of Sn thickness that reacts with Cu to form IMCs. As-plated Cu/Sn samples were annealed using different temperatures (150°C to 300°C) and durations (0 min to 320 min). The kinetics model is then extracted from the measured thickness of IMCs of the annealed samples.

Distortion Free Wafer Bonding Technology for Backside Illumination Image Sensors

Significant distortion can originate during the bonding process, resulting in overlay issues for final backside illumination image sensor wafer processing, impacting final wafer yield and device reliability. This paper presents a distortion free bonding process where the bonding initiation is controlled by low gas pressure. No external force is needed to initiate the bonding. The resulting overlay capability enables the backside illumination image sensor technical roadmap. In addition, a wafer shape analysis methodology developed for early distortion detection is discussed, allowing bonding process optimization and diagnostic for bonding rework.

Ball Bond Process Optimization with Cu and Pd-Coated Cu Wire

Cu and Pd-coated Cu (PdCu) wire have been replacing Au wire in electronic packaging. The biggest reason for this transition is the high cost of gold. Methods for process characterization and optimization in Cu and Pd-coated Cu wire differ substantially from those typically used for Au wire. This paper will examine how modifications and newly developed metrology tools are used in the characterization of PdCu and Cu wire bonding. Shear strength, traditionally the most important measure of bond quality in Au wire bonding, is of only limited use in Cu wire bonding. Ball pull has become a much more useful measure of bond quality since it is a more direct measure of too little bonding (ball lifts) and too much bonding (pad peeling). Other important measurements for Cu wire bonding include measurement of Al pad splash and bonded ball cross-sectioning to inspect the ball pad interface and the remaining Al thickness. For Au bonding, the process recipe can be fairly easily optimized and transferred to another bonder. Process windows are much narrower in the Cu and PdCu 1st bond process. It is very critical to develop a robust process with an acceptable process window for Cu wire bonding and to have good control for equipment and material during production. Due to the high demand of process optimization, equipment repeatability, material and production control, collaborations between companies in the electronic packaging supply chain is important to ensure advances in Cu wire bonding technology.

Sn–Cu–Ni Soldering Process Optimization Using Multivariate Analysis

One lead free solder candidate, Sn-Ag-Cu305 (Sn96.5/Ag3.0/Cu0.5), has come into widespread use as a replacement for traditional tin-lead solder (Sn63/Pb37). However, the price of silver has increased dramatically in recent years. This has increased manufacturing costs, impacting firm competitiveness. This paper evaluates the performance of the low cost Sn-Cu-Ni solder alloy used for wave soldering. Sample products for several applications are used to assess critical factors in the wave soldering process. A principal component analysis was performed in this paper to integrate multiple quality characteristics, namely assembly yield and solder joint strength, for process development. As a result, a parameter combination of 270 soldering temperature and 8 s dwell time is recommended. A test vehicle with daisy-chain circuitry design is used to investigate the performance of samples prepared by the optimal process. Samples are subject to the thermal cycling test. The performance of the recommended process is therefore verified.

PBGA Cost Reduction by Reducing Plated Au Thickness of Substrate Pads

In the past decade, we have witnessed the significant increase of gold price. Coupled with continuous price decrease of electronic products, the reduction of gold consumption has been explored in every possible area. This paper is related to reducing gold consumption on wire bonding pads of low density interconnect substrates. A layer of gold is plated on the substrate pads with typical minimum thickness of 0.5um to facilitate wire bonding process using gold wire. The objective of this study was to reduce gold thickness on the substrate pad to 0.3um minimum without compromising wire bond integrity. DOE considering different substrate sources wire bonded by different assembly suppliers were conducted. Wire bonding process optimization was followed by package level reliability stress testing, heat aging, and board level solder joint reliability testing was carried out for validation. The results showed plated gold thickness on the wire bond substrate pads can be reduced to a minimum of 0.3um without compromising package reliability performance. With the success of thinner gold on the substrate pads, we can use the same natural resources - gold to build more units, 'better for environment, if we consume less'

Dynamics of Contact Wave in Silicon Wafer Direct Bonding

A quantitative description of the dynamics of silicon wafer direct bonding is proposed, and the contact wave for bonding front propagation is modeled as a function of time. The changes in exhaust gas power, accumulated deformation power and wafer surface energy during bond processing have been studied systematically, and the governing differential equation of the contact wave position is deduced from the law of conservation of energy. The relationship of wave front position to the height of gas gap between bonded wafers is derived. Moreover, the functional dependence of contact wave position and bonding wave velocity on time is investigated, and the average wave velocity is obtained. The model describes well the experimental data available from literatures, and provides a basis for silicon wafer direct bonding process optimization.

A Method to Quickly Measure the Thickness of Aluminum Bond Pad on Silicon Substrate

In wire bonding, the thickness of the aluminum pad under the bonds needs to be measured for bonding process optimization purposes. In Cu wire bonding, the deformation of the Al bond pad is larger than in Au wire bonding. Because of this, the uniformity and thickness of Al that remains under the ball is often used as a metric in the process optimization. The thickness of the Al pads can be inspected with manual cross sectioning or FIB (Focused Ion Beam). However, these two methods are time consuming and expensive. This paper presents a method that helps to quickly measure the aluminum pad, which involves creating a graph of penetration of electron beam at different EDS (Energy Dispersive Spectroscopy) voltages versus Al thickness. The method basically consists of: a) crafting Al layers of different thicknesses, b) obtaining EDS for each layer at different EDS voltages, c) performing cross sectioning of those layers and measuring the thickness of the layers, d) generating the metric based on the acquired data, and e) testing the acquired metric.

Concurrent Optimization of Crescent Bond Pull Force and Tail Breaking Force in a Thermosonic Cu Wire Bonding Process

In conventional wire bonding process optimization, the crescent bond is tested by destructive pulling the loop and measure the pull force (PF) required to break the bond. While this method assures the final quality, it does not necessarily minimize production stoppages which can reduce throughput significantly. Many stoppages are caused by short-tail and tail-lift errors which in turn are caused by reduced tail bond strength. The tail breaking force (TBF) is a measure for tail bond strength. A consistent tail breaking operation is needed for robust Cu wire bond production with little operator assistance required to restart stopped machines. We report the concurrent (simultaneous) optimization of PF and TBF using standard pull testing and a method that directly measures the tail bond strength in-process, respectively. The example process uses standard 25-mum-diameter Cu wire on standard Ag-plated leadframe diepads and wire loops oriented perpendicular to the ultrasonic horn. The optimization consists of (1) finding the factors to obtain a symmetrical bond shape, (2) choosing fixed values for bond time (25 ms) and heater stage temperature (220degC), and (3) adding the new optimization step of maximizing the TBF by iteratively optimizing the impact force (IF), bonding force (BF), and ultrasound (US) parameters. Among these parameters, the US and BF are the most significant. A process window (PW) is defined as the set of US/BF parameter combinations that result in a response being inside a previously defined range. PWs for PF and TBF are determined and compared with each other. There is only a partial overlap of these PWs. To increase the process capability index (cpk) for TBF, it is recommended to first carry out conventional PF optimization followed by a minimization of the bonding force parameter to the lowest value still fulfilling the PF cpk requirement.

The Destructive Wire Bond Pull Test Methodology for the Ultra Low Loop Application on Thinner Die Package

The destructive bond pull test is the most common methodology for quality control of the wire bond process optimization on assembly manufacturing. It was introduced in the 1960's and it is described in U.S military specifications. Today, some of the hottest wire bonding processes is System-In-Package (SIP), Package-On-Package (PoP) and 3D package. These processes require lower loop height on thinner die thicknesses. Ultra Low Loop (ULL, <75 mm height) on thinner die (<100mm) utilizes this process. But, the bond pull test specifications for wire pull hook position and minimum wire breaking force are not sufficient to apply for the ULL and thinner die applications necessary to meet today's criteria. This paper is presented in order to give a universal test method of the wire pull test specification and methodology for the ULL on thinner die application and optimization. It will extend the reliability and quality control of advanced packaging.