Table of Contents III IV 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

May 2024_EDFA_Digital

A RESOURCE FOR TECHNICAL INFORMATION AND INDUSTRY DEVELOPMENTS MAY 2024 | VOLUME 26 | ISSUE 2 ELECTRONIC DEVICE FAILURE ANALYSIS edfas.org SThM FOR LOCALIZING AND MONITORING DEFECTS ELECTRO-THERMAL SIMULATION AND RELIABILITY OF A BGA DIFFERENTIAL LASER VOLTAGE PROBE OVERVIEW MICROSTRUCTURAL HIERARCHY DESCRIPTOR FOR MFA 4 22 10 32

A RESOURCE FOR TECHNICAL INFORMATION AND INDUSTRY DEVELOPMENTS MAY 2024 | VOLUME 26 | ISSUE 2 ELECTRONIC DEVICE FAILURE ANALYSIS edfas.org SThM FOR LOCALIZING AND MONITORING DEFECTS ELECTRO-THERMAL SIMULATION AND RELIABILITY OF A BGA DIFFERENTIAL LASER VOLTAGE PROBE OVERVIEW MICROSTRUCTURAL HIERARCHY DESCRIPTOR FOR MFA 4 22 10 32

Meet the 2024 Keynote Speaker: Dr. James Chambers Vice President of Silicon Engineering and Sourcing Nvidia Corp. Wednesday, October 30 The International Symposium for Testing and Failure Analysis (ISTFA) is the premier event for the microelectronics failure analysis community. Join us at the only North American event devoted to the semiconductor, electronic sample preparation, and imaging markets. This is the best venue for failure analysts and the FA community to share challenges and acquire the technical knowledge and resources needed to take them on. OCTOBER 28–NOVEMBER 1, 2024 | SAN DIEGO, CA save the date Riding the wave of artificial intelligence 2024 Registration opens July 2024! Mark your calendars now and prepare to network with the best in the FA industry. ORGANIZED BY: istfaevent.org

edfas.org 1 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 A RESOURCE FOR TECHNICAL INFORMATION AND INDUSTRY DEVELOPMENTS MAY 2024 | VOLUME 26 | ISSUE 2 edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS DEPARTMENTS MicroStructural Hierarchy Descriptor Enabling Interpretative AI for Microelectronic Failure Analysis Zhiheng Huang, Ziyan Liao, Kaiwen Zheng, Xin Zeng, Yuezhong Meng, Hui Yan, and Yang Liu The MicroStructural Hierarchy Descriptor is proposed as a systematic and quantitative approach to spectra and image data in microelectronic failure analysis. Author Guidelines Author guidelines and a sample article are available at edfas.org. Potential authors should consult the guidelines for useful information prior to manuscript preparation. 4 10 2 GUEST EDITORIAL Felix Beaudoin 40 2024 PHOTO CONTEST 41 2024 VIDEO CONTEST 42 EDUCATION NEWS Bhanu Sood 44 DIRECTORY OF FA PROVIDERS Rosalinda Ring 46 LITERATURE REVIEW Michael R. Bruce 47 PRODUCT NEWS Ted Kolasa 49 TRAINING CALENDAR Rosalinda Ring 51 GUEST COLUMN Sarah Poehlmann, Renee Parente, Joseph Caroselli, and Tom Schamp 52 ADVERTISERS INDEX Electro-thermal Simulation and Reliability of a Ball Grid Array Norelislam El Hami, Aicha Koulou, Maria Zemzami, and Abdelkhalak El Hami The influence of electric current flow and electrically induced Joule heat on thermal stress for weld joint cracks at both interfaces is still not fully comprehended. This article investigates the effect of subjecting the BGA package to a cyclic current input. 22 For the digital edition, log in to edfas.org, click on the “News & Magazines” tab, and select “EDFA Magazine.” Scanning Thermal Microscopy for Localizing and Monitoring Defects in Electronics Séverine Gomès This article presents the principle of SThM instruments and their potential uses for the local thermal analysis of passive and active electronic components and devices. 10 4 22 32 Differential Laser Voltage Probe: A Brief Overview and Thoughts on What Could Come Next Kristofor Dickson Differential laser voltage probe simultaneously acquires waveform data from a single target while the device under test fluctuates between passing and failing test outcomes. 32 ABOUT THE COVER Micro potted plant. Photo by Susanne Hübner, Fraunhofer Institute for Microstructure of Materials and Systems IMWS. First Place Winner in Black and White Images, 2023 EDFAS Photo Contest.

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 2 I had the fortunate opportunity to attend the first CHIPS Metrology Workshop on Failure Analysis (FA) and Reliability Testing, which was held February 22, 2024, at the NIST Campus in Boulder, Colorado. It was attended by approximately 100 individuals from a wide variety of sectors: government, including NIST and CHIPS act representatives, aerospace and defense, software and hardware tool vendors, service providers, and semiconductor fabless and foundry companies. The workshop was opened by Paul D. Hale, Deputy Director, CHIPS Metrology program. He indicated the workshop was planned thanks to the CHIPS Metrology Industrial Advisory Committee (IAC) recommendation to prioritize FA and reliability. This recognition resulted from efforts made by the EDFAS FA Technology Roadmap’s leadership team who had multiple engagements with the IAC in 2023 to emphasize the need for a focus on FA. Hale also acknowledged that FA was identified as a central component linking all CHIPS for America programs. He stated that the workshop goals were twofold: Give NIST researchers an industry perspective, and evaluate the CHIPS Act Metrology R&D program industry relevance to plan for future projects. The keynote was given by Eric Hadland, director of technology policy at Semiconductor Industry Association (SIA). He emphasized the industry’s unprecedented growth forecast and the associated challenges such as the transition from 2D to 3D, the introduction of new materials, and the increased need for local insight at wafer scale such as TEM automation and wider use of AI/ ML models. The workshop consisted of two panel discussions followed by breakout room discussions for reliability and FA, and a one-hour CHIPS Metrology NIST research project showcase. Overall, the panel discussion’s aim was to introduce the high-level challenges of FA and reliability. The follow-up breakout room discussions mostly consisted of elaborating on the concepts introduced by the panelists through specific questions from NIST moderators. Bob Keller, a NIST group leader and CHIPS metrology program manager, moderated the reliability panel with Brendan Foran, The Aerospace Corp.; MAY 2024 | VOLUME 26 | ISSUE 2 A RESOURCE FOR TECHNICAL INFORMATION AND INDUSTRY DEVELOPMENTS ELECTRONIC DEVICE FAILURE ANALYSIS GUEST EDITORIAL A SHORT SUMMARY OF THE FIRST CHIPS METROLOGY WORKSHOP ON FAILURE ANALYSIS AND RELIABILITY TESTING Felix Beaudoin, GlobalFoundries felix.beaudoin@globalfoundries.com edfas.org Beaudoin (continued on page 43) PURPOSE: To provide a technical condensation of information of interest to electronic device failure analysis technicians, engineers, and managers. Nicholas Antoniou Editor/PrimeNano nicholas@primenanoinc.com Mary Anne Fleming Director, Journals, Magazines & Digital Media Joanne Miller Senior Editor Victoria Burt Managing Editor Allison Freeman Production Supervisor ASSOCIATE EDITORS Navid Asadi University of Florida Guillaume Bascoul CNES France Felix Beaudoin GlobalFoundries Michael R. Bruce Consultant Jiann Min Chin Advanced Micro Devices Singapore Edward I. Cole, Jr. Sandia National Labs Michael DiBattista Varioscale Inc. Rosine Coq Germanicus Universitié de Caen Normandie Szu Huat Goh Qualcomm Ted Kolasa Northrop Grumman Space Systems Rosalinda M. Ring Thermo Fisher Scientific Tom Schamp Materials Analytical Services LLC David Su Yi-Xiang Investment Co. Martin Versen University of Applied Sciences Rosenheim, Germany FOUNDING EDITORS Edward I. Cole, Jr. Sandia National Labs Lawrence C. Wagner LWSN Consulting Inc. GRAPHIC DESIGN Jan Nejedlik, jan@designbyj.com PRESS RELEASE SUBMISSIONS magazines@asminternational.org Electronic Device Failure Analysis™ (ISSN 1537-0755) is published quarterly by ASM International®, 9639 Kinsman Road, Materials Park, OH 44073; tel: 800.336.5152; website: edfas. org. Copyright © 2024 by ASM International. Receive Electronic Device Failure Analysis as part of your EDFAS membership. Non-member subscription rate is $175 U.S. per year. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by ASM International for libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that the base fee of $19 per article is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923, USA. Electronic Device Failure Analysis is indexed or abstracted by Compendex, EBSCO, Gale, and ProQuest. Failure analysis panelists at the CHIPS Metrology Workshop on FA.

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 4 EDFAAO (2024) 2:4-8 1537-0755/$19.00 ©ASM International® SCANNING THERMAL MICROSCOPY FOR LOCALIZING AND MONITORING DEFECTS IN ELECTRONICS Séverine Gomès CNRS, Centre d’Energétique et de Thermique de Lyon (CETHIL) Villeurbanne, France severine.gomes@insa-lyon.fr INTRODUCTION The constant downscaling of electronic components for the integration of electronic devices has increased the need for thermal measurements at micro and nano scales. Heat flow management at these scales is crucial to meet the challenges of thermal management of complete integrated systems, detecting defects and failures in devices, and improving the design and performance of microelectronic technologies. This need has motivated the development of various thermal techniques, such as thermoreflectance, Raman thermometry, fluorescence, and luminescence. However, these optical methods suffer from a spatial resolution limited by optical diffraction, which is not sufficient to char- acterize nanoscale electronic components. To overcome the lack of versatile, high-resolution thermometry techniques, scanning thermal microscopy (SThM),[1] based on scanning probe microscopy, has been developed and is today one of the most effective techniques used for thermal characterization of materials and systems at small scales. SThM can detect and localize heat sources generated in active devices and analyze the thermal conductance of systems with a sub-100 nm lateral spatial resolution. Nevertheless, the correct analysis and evaluation of heat transfer within individual nanostructures, which have recently been successfully implemented in miniaturized devices, remains an experimental challenge. To meet this challenge, an innovative hybrid instrument based on the combination of a SThM and a scanning electron microscope (SEM) has been built at CETHIL.[2] This article presents the principle of SThM instruments and their potential uses for the local thermal analysis of passive and active electronic components and devices. The hybrid SEM-SThM instrument recently developed at CETHIL is then described and demonstrated through a quantitative analysis of heat transport in an indivi- dual nanowire. SET UP The SThM instruments are based on atomic force microscopy (AFM) systems and use AFM cantilevers equipped with a resistive element located on the tip (Fig. 1). Piezoelectric scanners are used to move the sample vertically and laterally. As the sample surface is scanned by the tip, deviations of the cantilever, corresponding to variations in the force of interaction between the tip and the sample, are detected optically. The thermal sensor on the tip is used to analyze the sample’s thermal properties. The sensor’s temperature (Tp) can be determined because its metallic electrical resistance (Rp) depends linearly on temperature: ΔRp = Rp0 · αp · (Tp-Tp0), with αp the temperature coefficient of the metal’s electrical resistivity and Rp0 its electrical resistance at a reference temperature T0. ΔRp is measured using a control unit based on a Wheatstone bridge. In imaging mode, the SThM probe is generally used in the AFM’s contact and constant force modes. The SThM system can provide a topography image and “thermal” mapping simultaneously. The contrast of the thermal mapping reflects the variation in heat flux exchanged between the probe and the sample. Any variation in this heat flux that is linked to the local sample’s properties during the sample’s scanning will induce a variation in the probe temperature Tp and its electrical resistance Rp. DIFFERENT MODES The probe can be operated in both passive and active modes. In passive mode, a very low electrical current is used to measure the electrical resistance, Rp, of the thermal sensor without heating the sample, and the probe acts as a resistive thermometer. SThM in this mode has mainly been applied to imaging the temperature field at the surface of self-heated micro devices. In active mode, the probe is Joule-heated and used simultaneously as an ultra-local heat source for the sample

edfas.org 5 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 and as a heat flux meter. During tip-sample contact, the hot probe gives up a quantity of heat, Q, to the cold sample and cools down (Tp and Rp decrease). As Q depends on the thermal conductance of the system under study, SThM can be used to analyze the thermal properties of components at the subsurface of the system. Today, the active mode is preferred to the passive mode for sample surface temperature analysis. As with large-scale contact thermal sensors, measurement uncertainty is linked to the thermal resistance of the probe, tip-sample contact, and sample itself. The contact thermal resistance depends on many parameters, including topography, local variations in sample surface roughness and the sample’s thermal properties. These parameters can be determined in active mode using methods based on the principle of active probes with compensated thermal flux,[1] which is not possible in passive mode. For both passive and active modes, the sensor’s temperature can be determined from experimental raw data. Careful calibration of the probe and modeling of the probesample system are required to quantitatively analyze the temperature or thermal properties of a sample. The current state of the art is based on the post-processing of the probe’s electrothermal response with tridimensional models that reproduce the best experiments. This article does not fully address this aspect of the SThM technique but the reader can refer to references 1-3 for more details. DETECTION OF LOCALIZED INHOMOGENEITY The active-mode SThM technique can be useful for defect de- tection and coating/thin-film evaluation. The first experiment to demonstrate the technique for these applications involves an analysis of a polymer matrix filled with expanded graphite particles. As shown in Fig. 2, the contrast of the nanocomposite’s thermal image does not match the topography of the sample. SThM can therefore be used to locate particle agglomerations in the matrix.[4] The second experiment involved a sample consisting of nine silicon dioxide (SiO2) steps ranging in thickness from 3 to 1000 nm on a silicon (Si) substrate[3] (Fig. 3, left). The results presented in Fig. 3, right, show that the SThM technique can detect the variation in SiO2 thickness over the entire thickness range studied, remaining sensitive to the presence of the Si substrate up to a SiO2 layer thickness greater than 1 µm. Heat conduction in oxide layers, inhomogeneity in oxide thickness, layer/substrate thermal resistance as well as details on the sample’s surface can then be analyzed by means of SThM. DETECTION OF LOCALIZED HEATING SThM can also allow detecting and locating electronic component overheating and hot spots, which often reflect component malfunction or impending device failure. Fig. 1 Top, schematic of the SThM instruments; bottom SThM probe with a resistive metallic element deposited on the AFM tip. Fig. 2 Left, topography image; right, thermal image of 10% weight fraction of expanded graphite particle, high-density polyethylene sample.[4]

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 6 As an example, Fig. 4 shows the results obtained using the SThM method in active mode on a fully flat, thermally active test device (manufactured at IEMN in Lille, France). The device is based on the implantation of boron ions into the top layer of a silicon-on-insulator substrate to create three heat-generating resistive channels ranging in width from 20 µm to 100 nm. The two sides of each channel are connected by two platinum (Pt) wires for four-point electrical measurements. Experiments were carried out on unheated channels and on lines heated with an electrical current of a few mA, while they had a separate ground and a common ground to create a current leak in the device. As Si and ion-implanted Si have virtually the same thermal conductivity, implanted lines are not detected when they are not heated (Fig. 4, right top). Platinum wires, on the other hand, are detected because their height is around 40 nm, and the thermal conductivity of Pt is lower than that of Si. When one of the three lines is heated, it is clearly detected in both measurement configurations (different grounds and common ground in Fig. 4, right, middle, and bottom, respectively). The other two lines, however, appear heated when all three lines have a common ground, indicating that in this case, a current leak occurs in the upper Si layer of the device. Using a higher current led to a high voltage short circuit. SThM can then be used for failure analysis of electronic devices, helping to explain why a device fails and, more importantly, to define the safety margin of device operation. All of these results were obtained in air. Under such conditions, the spatial resolution of the technique can be degraded due to thermal conduction in the air between the probe and the sample. Thermal conduction through the water meniscus at the tip-sample contact can contribute to the measurement. Because modeling these heat transfers Fig. 3 SiO2 step sample: top left, cross-sectional diagram of a sample; bottom left, sample topography; right, thermal signal as a function of SiO2 thickness. The inset shows a thermal image of the sample.[3] Fig. 4 Left, top view and cross section of the sample’s active zone. Middle, topography, and right, thermal images of the active zone with three lines of ion implanted silicon.

edfas.org 7 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 remains a challenge at the nanoscale, it is recommended to use the SThM method under vacuum conditions when quantitative measurements are expected. Moreover, it becomes difficult to characterize quantitatively, with low uncertainty, a physical phenomenon such as heat dissipation in a nanodevice or the thermal transport properties of nanostructures.[1] COMBINED SThM-SEM INSTRUMENT The combined SThM-SEM instrument (Fig. 5) developed at CETHIL is based on a compact AFM system designed to be inserted into a SEM for handling and metrological scanning.[5] The advantages of this instrument coupling is the ability to conduct experiments in vacuum conditions and perform correlative AFM, SThM, and SEM analysis. The equipment can simultaneously image the sample with high resolution, accurately measure heights, distances and thermophysical properties of the material or temperature field at the sample surface when active, while retaining the SEM’s large field of view to position the cantilever exactly where needed. The user can also observe in real time the shape and size of the tip, as well as the sample surface during thermal image acquisition and force spectroscopy measurements. The system improves workflow by saving time without having to move the sample between instruments. In addition, the combined instrument allows better knowledge of the geometry, size (radius of curvature) and surface condition of the tip, as well as better knowledge of physical contact (materials at the apex of the tip and at tip-sample contact). All these parameters are valuable inputs for modeling and quantitative thermal analysis at the nanoscale. ANALYSIS OF SUSPENDED NANOSTRUCTURES To demonstrate the capabilities of the combined setup, the heat transport within a 100 nm in diameter and 20 µm in length rough Si nanowire (NW) was analyzed.[2] The nanowire was suspended between two platforms. This work was carried out in collaboration with the Catalonia Institute for Energy Research. The experiments involved a series of approach curves measurements (in AFM’s force spectroscopy mode) along the suspended nanowire. A three-dimensional electrothermal model of the probe interacting with the nanowire, established by a finite element method, was used to calibrate the probe and to determine, from the experimental results, the local thermal conductance along the nanowire Gtip-NW(x) (Fig. 6, top). Fitting Gtip-NW(x) with an analytical model of heat dissipation in the nanowire enabled estimating the nanowire’s thermal conductivity at 13.7 ± 1.6 Wm-1K-1 (Fig. 6, bottom), which is in very good agreement with its values calculated elsewhere. This work highlights the capabilities of the combined SThM-SEM instrument, in particular the value of highly controlled probe positioning at the nanoscale. CONCLUSION This article presents the SThM technique, showing that this microscopy can be used not only to study the Fig. 5 Combined SThM-SEM instruments.

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 8 thermal conductance of components, but also to locate hot spots on a microscopic scale. In addition to the installation of test benches, the measurement concept is presented. Three examples demonstrate the SThM’s ability to perform thermal analysis on a microscopic scale. Thanks to its ability to detect defects embedded in a matrix and hot spots localized on the sample surface, and its sensitivity to the thickness of thin layers on the substrate, SThM could be used as a powerful tool for analyzing printed circuit boards and electronic devices with high spatial resolution, during the development cycle, failure analysis during and after manufacture, and during operation. Complementing other instruments such as the transmission x-ray microscope, scanner, and electron probe micro-analyzer, it could help identify and understand failure points, inspect the appearance of defective areas, and correctly understand the physical properties of microscopic structures. This article also demonstrated a new combined SThM-SEM instrument by measuring a suspended nanostructure. Further work will focus on measuring temperature along a self-heated nanowire. REFERENCES 1. S. Gomès, A. Assy, and P.-O. Chapuis: “Scanning Thermal Microscopy: A Review,” Phys. Status Solidi A, 2015, 212, p. 477-494. 2. J.M. Sojo-Gordillo, et al.: “Local Heat Dissipation and Elasticity of Suspended Silicon Nanowires Revealed by Dual Scanning Electron and Thermal Microscopies,” Small, 2023, p. 2305831. 3. E. Guen, et al.: “Scanning Thermal Microscopy on Samples of Varying Effective Thermal Conductivities and Identical Flat Surfaces,” Journal of Applied Physics, 2020, 128(23), p. 235301. 4. W. Sun, et al.: “Investigation of the Thermal Conductivity En- hancement Mechanism of Polymer Composites with Carbon-based Fillers by Scanning Thermal Microscopy,” AIP Advances, 2022, 12(10), p. 105303. 5. D. Renahy, A. Assy, and S. Gomès: “A Combined SThM/SEM Instrument for the Investigation of Influent Parameters in Nano-scale Thermal Contact,” International Workshop on THERMAL INVESTIGATIONS of ICs and Systems, 2015, THERMINIC, 30 September 2015. Fig. 6 SThM analysis of a suspended Si nanowire (NW).[2] Top, approach curves along the NW in terms of thermal conductance of the probe as a function of distance between the probe-apex and NW. Bottom, identification of the Si NW’s thermal conductivity. ABOUT THE AUTHOR Séverine Gomès obtained her European doctorate in physics from the University of Reims Champagne-Ardenne in France in 1999. She began her career as a CNRS researcher at the Centre d’énergie et de thermique de Lyon (CETHIL) at the French National Institute of Applied Sciences (INSA) in France in 2000. She has received two awards: The CNRS Bronze Medal in 2015 for her promising early work in scanning probe thermal microscopy and the European Star from the French Ministry of Higher Education, Research and Innovation in 2018 for her successful management of the four-year European collaborative project “QUANTItative scanning probe microscopy techniques for HEAT transfer management in nanomaterials and nano-devices (QUANTIHEAT).” In 2018, she became a CNRS professor at CETHIL. Since 2020, she has been coordinator of a French scientific working group on thermal nano-metrology. In cooperation with national and international universities, she regularly supervises master and doctoral students.

edfas.org 9 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 10 MICROSTRUCTURAL HIERARCHY DESCRIPTOR ENABLING INTERPRETATIVE AI FOR MICROELECTRONIC FAILURE ANALYSIS Zhiheng Huang1, Ziyan Liao1, Kaiwen Zheng1, Xin Zeng1, Yuezhong Meng1, Hui Yan2, and Yang Liu3 1The Key Laboratory of Low-carbon Chemistry & Energy Conservation of Guangdong Province, and School of Materials Science and Engineering, Sun Yat-sen University, Guangzhou, China 2School of Computer Science, Sun Yat-sen University, Guangzhou, China 3School of Electronics and Information Technology, Sun Yat-sen University, Guangzhou, China hzh29@mail.sysu.edu.cn EDFAAO (2024) 2:10-18 1537-0755/$19.00 ©ASM International® INTRODUCTION The MicroStructural Hierarchy Descriptor (µSHD)[1–3] was initially proposed under the background of materials genome engineering (MGE),[4] where a quantitative description on material microstructure is critical to find the so called “materials genomes.”[5] To achieve this goal, quantitative relationships must be established between the processing, microstructure, and properties/performance of different materials. Therefore, the quantification of a microstructure, which to date is based only on crude low-level statistics, is an essential step, and µSHD has been proposed in this context. In the field of microelectronic failure analysis (MFA), characterization techniques such as scanning acoustic microscopy (SAM), magnetic field imaging (MFI), scanning electron microscopy (SEM), energy-dispersive x-ray spectroscopy (EDS), transmission electron microscopy (TEM), and others with spatial resolutions that span multiple scales, are routinely used to obtain spectra and images for fault localization and isolation. For diagnostics at the integrated circuit (IC) level, atomic-scale or even electron-scale characterization techniques are common. In contrast, it involves mainly mesoto-macroscopic scales at the packaging level. However, the bottleneck switches to packaging and interconnects in the 3D IC era. Recent years have seen the significant miniaturization of packaging and interconnects along with the IC processing node to enhance the performance and efficiency of electronic devices. Therefore, comprehensive characterization techniques that span from the electron scale to the macroscopic scale are required for MFA. In addition, the task of fault localization and isolation is mainly experience-based, and the data from characterization are only qualitatively utilized according to the operator’s experience and background. However, artificial intelligence (AI) may automatically control and perform characterization equipment and MFA in the foreseeable future.[6–8] With this vision in mind, a systematic and data-driven approach will be necessary. To this end, the convergence of the MGE and MFA objectives, and the µSHD approach will play a role in facilitating the adoption of AI in these two fields. It is noted that common AI tools have already been explored in the field of MFA for higher resolution imaging, increased throughput, enhanced image contrast, and faster scan times.[9] Deep convolutional neural network (DCNN)- based computing models have also been utilized for superior image segmentation and object classification.[10] While satisfactory results may have been obtained, the working principles behind DCNNs remain mysterious. It is an interpretative descriptor, i.e., the µSHD, rather than the mysterious one generated and hidden in the DCNNs, that is the focus of this work. The following sections introduce the concept of µSHD and benchmark its behavior for images captured with common characterization techniques to stimulate further and systematic application scenarios. MICROSTRUCTURAL HIERARCHY DESCRIPTOR As mentioned in the previous section, there are abundant spectra and images with resolutions of several orders

edfas.org 11 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 of magnitude involved in MFA. A systematic approach must be developed to overcome this challenge on multiple scales. C.S. Smith first proposed the concept of structural hierarchy to describe the multiscale nature of material structures.[11] He pointed out that what matters is the interlocking of the smaller scales that generate the larger overarching structures.[12] Unfortunately, no quantitative descriptor for the hierarchy of structures has been developed. S.M. Mallat studied the mathematics behind AI for image recognition,[13] that is, DCNNs, and found that the features learned from DCNNs resemble wavelet systems, which inherently maintain a structural hierarchy. Wavelet-like physical structures exist in our ears and brains, as physiology research has identified.[14] Based on these connections, the Mallat scattering transform (MST)[15] was proposed and used to dramatically reduce computing costs but maintain AI power and accuracy for image recognition. In short, the invariants from the MST are good descriptors, like the ones learned from the DCNNs. Therefore, one can naturally connect this MST-invariant system with a quantitative descriptor of the structural hierarchy, that is, µSHD, which is generally applicable to digital signals such as spectra and images. Figures 1a and b illustrate the secondary and backscattered electron images of an aluminum line with a tungsten cap layer after electromigration, and Figs. 1c and d show the analysis Morlet wavelets and the µSHD curves. Note that all input images for µSHD analysis in this work are trimmed to the same size of 512 × 512 pixels2. The first eight columns in Fig. 1c show the real part, while the rest show the imaginary part of the analysis wavelets. Each column from top to bottom corresponds to an increasing scale from J = 1 to J = 8 with the same orientation, L. Similarly, each row from left to right corresponds to an increasing orientation from L = 1 to L = 8 on the same scale J. It is noted that multiresolution wavelet analysis in the images is the first step toward the µSHD curve. To make the descriptor more robust, e.g., invariant under small image distortion or rotation, a series of invariants must be obtained following the MST. Mathematics is heavily involved in this step, and details can be referred to Fig. 1 SEM images and corresponding µSHD curves of an aluminum line with a tungsten cap layer after electromigration. (a) Secondary electron image, (b) backscattered electron image, (c) real and imaginary parts of the Morlet wavelet system for analysis, and (d) µSHD curves. (a) and (b) are adapted from Vanderlinde.[16] (a) (b) (c) (d)

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 12 Mallat’s original work[13,15] and previous publications.[1–3] The first-order invariants S1 consist of 64 numbers, which are constructed according to the analysis results by a system of wavelets with eight scales and each scale with eight orientations. The invariants are then plotted according to the scale index J and the orientation index L, that is, J = 1, L = 1,..., J = 1, L = 8, J = 2, L = 1,..., J = 8, L = 1,..., J = 8, L = 8. A curve connecting those data points according to this order is called a µSHD curve, and the invariant on a certain scale J and orientation L after taking the logarithm (base 10) is called a µSHD value or simply a µSHD. The µSHD curves in Fig. 1d represent, systematically, quantitatively, and for the first time, what the differences are between the secondary and backscattered images of the same object. In terms of features on different scales, the µSHDs on scales J = 1 and J = 2 and for all orientations on curve b from the backscattered electron image are larger than those on curve a from the secondary electron image. However, the µSHDs on scales from J = 3 to J = 8 on curve a grow increasingly larger with scales and become larger than their counterparts on curve b. The features in different orientations of both curves seem to match very well, that is, both curves show peaks at L = 3 and L = 7 on smaller scales, i.e., J = 1 and J = 2, but consistently at L = 5, that is, the horizontal direction, from J = 3 to J = 8. Reexamining Figs. 1a and b can find that the majority of the information from the aluminum line is indeed horizontal. Note that the µSHD approach differs significantly from the currently available AI approach in MFA. First, the µSHD approach is based on the mathematical principles behind AI and is interpretative and can eliminate the deep learning process intensive with computation. In other words, machine learning and AI-based models can only become more efficient and interpretative if µSHD is used directly as a quantitative descriptor. Furthermore, µSHD treats spectra and images as systems rather than individual signals and can utilize the concept of structural hierarchy to derive deeper insights behind the data. In summary, all the spectra and images in MFA can be systematically quantified within the same framework using the µSHD. With the maturity of data collection and benchmarking for different applications, µSHD enabled interpretative AI framework for MFA can be expected. The following section provides a few examples that highlight such efforts. CASE STUDIES SCANNING ACOUSTIC MICROSCOPY Figures 2a, b, and c show the SAM images of an IC package with an in- crease in measured frequency com- Fig. 2 SAM images of an IC package measured with a frequency and focal length setting of (a) 75 MHz and 10 mm, (b) 120 MHz and 5 mm, and (c) 180 MHz and 5 mm. (d) Redraws of (a) with the wavelets of the analysis superimposed. (e) Plots of the µSHD curves of the SAM images. (a), (b), and (c) are adapted from Hartfield et al.[17] (a) (b) (d) (e) (c)

edfas.org 13 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 bined with a decrease in focal length.[17] In contrast to Fig. 2a measured at 75 MHz and 10 mm focal length, Hartfield et al. claimed that better resolution and more details were obtained with an increase in frequency from 120 to 180 MHz and a focal length of 5 mm, as shown in Figs. 2b and c, respectively. Figure 2d redraws Fig. 2a with the wavelets of analysis superimposed. The real parts of the wavelets on the first to fifth scales are plotted with labels from J = 1 to J = 5, and wavelets with different orientations, labeled L = 1 to L = 7, are also shown on the scale J = 5. The feeling of better resolution and more details becomes concrete and quantified by referring to the µSHD curves of the SAM images shown in Fig. 2e. It is obvious that the µSHDs of the best image, that is, Fig. 2c, are the largest on smaller scales from J = 1 to J = 3, but become almost the smallest from J = 5 to J = 8, with the µSHDs of the three images being almost the same at J = 4. Features on larger scales, for example, from J = 5 to J = 8, sketch the outline of the image, which are already available in the images captured with lower frequencies, i.e., Figs. 2a and b. However, the features on smaller scales enrich the fine details of the images. Carefully examining the µSHD curves, one can find that the µSHDs on smaller scales in Figs. 2a and b are very similar, with levels obviously below 2c. Therefore, information on smaller scales provides better resolution and more detail in SAM images. Furthermore, the peaks on the µSHD curves appear to coincide very well in general, regardless of the measurement frequency and focal length settings. This means that the directional sensitivity under different measurement settings is similar, but fine details may differ. For example, it is the value of µSHD in L = 1, that is, the vertical direction, the maximum under the 75 MHz and 10 mm settings, but L = 5, that is, the horizontal direction, under the other two settings, on scale J = 8. With observation and understanding of the features presented in the µSHD curves of SAM images, we can establish, based on sufficient data, quantitative relationships between the µSHDs of SAM images and image quality in one direction and the measurement frequency and focal length settings in another direction. The supervised learning approach, which requires a labeled image quality score, by training a shallow neural network, can lead to a quantitative relationship between µSHDs, 64 values with a system of J = 8 and L = 8, and the image score. Similarly, a quantitative relationship between the µSHDs of an image or the low-level statistics of the µSHD curve, such as the mean, standard deviation on smaller and larger scales, and the measurement parameters, can also be obtained. Using directly the 64 µSHD values as a quantitative descriptor, unsupervised learning, such as self-organizing maps (SOM), can be used to automatically classify SAM images into different clusters. SOM results can also reveal how the measuring parameters, µSHD features, and image quality are mutually related. It should be noted that the number of features on a µSHD curve can be adjusted by varying J and L, or by using the second-order invariant S2, which generates substantially more features. After data insights are obtained, one can, for example, from the µSHDs of the SAM image, derive the measurement frequency and focal length setting, or judge whether or not the image quality is good enough under the current measurement parameters. MAGNETIC FIELD IMAGING Using the MFI technique, Hechtl[18] obtained current distribution maps in a flip-chip packaged die with top and bottom interposer metallization as shown in Figs. 3a to d, in which the current path detected by the superconducting quantum interference device (SQUID) sensor at a constant current of 1 mA for the same die with a thickness of 200, 75, 40, and 5 µm, respectively. The current image is blurred more for the case where the die is the thickest, that is, Fig. 3a. With the die becoming thinner to 75 µm and below, the image resolution improves, and the current paths on the die or even the interposer, located below the die, can be clearly identified. Being objects of interest for MFI, the current paths are better characterized by sharp lines, which correspond to µSHD curves with low energy (or low information). Therefore, it is reasonable to find that the image with the best resolution, that is, Fig. 3d, corresponds to the µSHD curve with the lowest energy, that is, the curve d in Fig. 3i. Note that the thickness of the die gradually decreases from 200 µm to 5 µm for Figs. 3a to d, but the µSHD curves do not drop accordingly. Although the µSHDs of Figs. 3a, b, and c vary little on scales from J = 1 to J = 4, the µSHDs of Fig. 3c are located between Fig. 3a and b on scale J = 5 and onward. The reason why the µSHDs on larger scales of Fig. 3c do not drop from the level of Fig. 3b lies in that larger scale features exist on top of Fig. 3c, which indeed appear to be larger than the counterpart in Fig. 3b. Hechtl further studied the effect of current, ranging from 1 mA to 25 µA with the die thickness fixed at 5 µm, on image quality, as shown in Figs. 3e to h. It is obvious that the information captured by the sensor decreases with decreasing current, although the electrical fault can still be located around the brightest area of the image with the lowest current of 25 µA. Among the µSHD curves shown in Fig. 3j, the best image, that is, Fig. 3e, maintains the lowest energy, while other images exhibit higher energy in the µSHD curves as their resolution power on current paths weakens. An obvious decrease in energy

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 14 on a larger scale in the µSHD curve of Fig. 3h signals only an information loss as the current drops to the lowest level. From Figs. 3i and j, it is clearly observed that the range within which the µSHD vary is between 0.03 and 1.54 for the effect of separation distance, while that is between 0.05 and 1.40 for the effect of current. As Orozco qualitatively noted in the original reference, the impact of reducing the current on the quality of the images is evident, but clearly less than the effect of distance.[18] Similar approaches and procedures as discussed in the SAM section can be followed to obtain data insight on the interrelationships between the measurement distance, current, and µSHD curves of the MFI images. SCANNING ELECTRON MICROSCOPY Due to some technique differences between SEM and the focused ion beam (FIB), Vanderlinde[16] pointed out that the adjustment of the focus and astigmatism corrections is the most difficult for a novice electron microscopist. Fortunately, such experience and skills can now be systematically quantified and demystified with the tool of µSHD curves. Figures 4a to d are the examples provided by Vanderlinde representing the same image under focus, over focus, true focus, and true focus after correction for astigmatism, respectively. The µSHD curves of the under- and overfocus images, i.e., curves a and b in Fig. 4e, are found to oscillate with increasing amplitude Fig. 3 MFI current density image from backside scanning under 1 mA with the separation distance decreasing from 200 µm (a) to 75 µm (b), 40 µm (c), 5 µm (d), and with a fixed separation distance of 5 µm under 1 mA (e), 500 µA (f), 100 µA (g), 25 µA (h). Images courtesy of M. Hechtl.[18] µSHD curves of (a) to (d) plotted in (i) and (e) to (h) in (j). (a) (b) (i) (j) (c) (d) (e) (f) (g) (h)

edfas.org 15 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 in an antiphase manner across scales from J = 1 to J = 5. More precisely, the µSHD curve of the underfocus image is concave upward with the minimum µSHD in orientation L = 5, while the overfocus image is concave downward with the maximum µSHD in the same orientation L = 5. This result agrees well with Vanderlinde’s description that the under- and overfocus images exhibited sharp edges only in the vertical and horizontal direction, respectively. Moreover, it is reasonable to locate the µSHD curve of the true focus image between the under- and over-focus ones. Note that the under, over, and true focus µSHD curves converge at larger scales at J = 7 and J = 8. The µSHDs of the sharper image after astigmatism on smaller scales, that is, J = 1 to J = 4, are found to increase substantially, but vary less significantly between different orientations at each scale. On larger scales from J = 5 to J = 8, all µSHD curves exhibit a large peak at L = 5, and the effect of astigmatism only shows a small increase in µSHD. With the above calibration, Vanderlinde’s experience on adjusting focus and astigmatism correction can be quantitatively described and passed onto novice electron microscopists. The µSHDs of a true focus image lie right in between these of the under- and over-focus images on the smaller scales, but maintain the largest on the larger scales. An astigmatism correction makes the µSHDs of the image the largest across all scales and also varies less significantly with orientation on smaller scales. Such quantitative information can also be used to realize autonomous focus of SEM. ENERGY DISPERSIVE X-RAY SPECTROSCOPY Figures 5a to c show the EDS mapping of O, Si, and Al on a sample consisting of 1.25 µm width and 0.5 µm thickness Al metal lines in a SiO2 dielectric with a 5 keV electron beam energy, respectively. Considering the energy of the characteristic x-rays of the elements and the penetration depth of the electron beam, the choice of 5 keV electrons is adequate, and the three elemental maps are well defined compared with the SEM image provided in the original reference. Because the bright dots matter in the elemental maps, we concentrate on smaller scales for this case. Examining the µSHD curves a to c in Fig. 5g on scales from Fig. 4 Illustration of the underfocus (a), overfocus (b), true focus (c), and true focus after astigmatism (d) SEM images. Images are adapted from Vanderlinde.[16] (e) The µSHD curves of the images. (a) (b) (e) (c) (d)

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 16 J = 1 to J = 3 one can find that the µSHDs of the O and Si maps overlap with each other, but the Al map shows the lowest µSHDs. As the O and Si signals come from the SiO2 dielectrics, the overlap of these two images makes sense. Unlike the O and Si maps, the Al map in Fig. 5c shows a higher density of bright dots, which aggregate and connect well to form continuous stripes on much larger scales. This explains the characteristics of its curve µSHD in Fig. 5g, that is, lower µSHD on smaller scales but much higher µSHD on larger scales. For a direct comparison, the same EDS maps under 15 keV electrons are provided in Figs. 5d to e, which are noticeably fuzzier and have inferior lateral resolution. As explained by Vanderlinde,[19] 15 keV electrons can penetrate more than 2 µm into the sample, which weakens the Al signal relative to Si because the Al lines are only 0.5 µm thick. In contrast to Si, the O signal is very short-ranged and can escape only from the near-surface area. Therefore, there will be a gap between the Si and O signals received under 15 keV EDS mapping. The µSHD curves in Fig. 5g clearly show the curve e, that is, the Si signal, exhibits the largest µSHDs on smaller scales as expected. There is now a gap between the Si and O signals. Furthermore, all the µSHD curves are above those in the 5 keV case. One of the reasons lies in the aforementioned strengthened Si signal because of deeper penetration depth, the other is due to deteriorated lateral resolution under 15 keV, which obviously expands the regions where the dielectric and metal lines are located. The above benchmark paves the way to gain data insight regarding the interrelationship amongst the µSHDs of the elemental map, electron beam energy and its penetration depth, characteristic x-rays of elements, film thickness, and lateral resolution. For specific applications, a data-driven approach can be expected to discriminate different elements according to the µSHD curves of the EDS maps. TRANSMISSION ELECTRON MICROSCOPY FIB-deposited metal films are often used for circuit editing, and the electrical resistivity of the deposited metals is of particular concern in high-speed and radiofrequency (RF) circuits. Figures 6a to f show the morphology and elemental EDS maps of two FIB-deposited platinum (Pt) films under TEM studied by DiBattista et al. This section attempts to correlate the µSHD curves of the Pt films with their electrical resistivities. By examining the µSHD curves in Fig. 6g, one can see that curve a, representing the structure shown in Fig. 6a, gradually increases its energy from J = 1 to J = 5, and then decreases its energy from J = 6 to J = 8. It is obvious that the curve ends at an Fig. 5 EDS mapping on a sample consisted of 1.25 µm width and 0.5 µm thickness Al metal lines in a SiO2 dielectric under an electron beam energy of 5 keV (a) to (c), and 15 keV (d) to (f). O K-α map in (a) and (d), Si K-α map in (b) and (e), and Al K-α map in (c) and (f). All maps are under 10,000x magnification. Images are adapted from Vanderlinde.[19] (g) The µSHD curves of the images. (a) (b) (d) (g) (c) (d) (e) (f)

edfas.org 17 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 material system, Pt versus SiC, and the material form, film versus bulk, are both different. DiBattista et al. reported that the resistivity of the film with morphology in Fig. 6a is 7665 µΩ·cm and in Fig. 6d 666 µΩ·cm. In addition to morphology, the element distribution in the microstructure can also influence the electrical resistivity. In general, the C map, for example, curve f in Fig. 6g, correlates better with the corresponding microstructural morphology, for example, curve d, compared to the Pt map. However, the Pt map, for example, curve b, appears to overlap with the corresponding morphology, for example, curve a, only on smaller scales with J = 1 and J = 2, then deviates as J increases, and finally the discrepancy decreases on larger scales with J = 7 and J = 8. Curve e appears to correlate better with its corresponding morphology, that is, curve d, on larger scales from J = 5 to J = 8. More work is currently being carried out to identify the principal components of µSHD and establish their quantitative relationships with the properties of the material. SUMMARY This article proposes µSHD as a systematic and quantitative approach to spectra and image data in the field of MFA. The advantages of this approach lie in that it originates from the understanding of the mathematics behind AI and thus eliminates the large computation cost from DCNNs. Note that the focus of this article is only on understanding and benchmarking the role that µSHD plays in different MFA applications. Concrete routes for employing µSHD directly as the quantitative descriptor for supervised and unsupervised machine learning have been discussed. Manufacturers of MFA equipment can certainly utilize the µSHD tool to automate and improve their characterization techniques and image processing and analysis protocols. Moreover, the concept of structural hierarchy behind µSHD can help to gain deeper insights from the data from the perspective of systems. energy level lower than where it starts. In contrast, curve d, which represents the structure of the other Pt film shown in Fig. 6d, also increases its energy from J = 1 to J = 5, but the energy oscillates and remains at a high level on larger scales. A previous study on the microstructure of sintered bulk silicon carbide (SiC) ceramics, found that a rapid drop from smaller to larger scales on the µSHD curve is a characteristic of high electrical resistivity and a reverse trend, showing a gradual increase from smaller to larger scales and maintaining larger µSHDs at larger scales, characterizing low resistivity.[2] This qualitatively reconfirms this conclusion, although the Fig. 6 TEM and EDS maps of two FIB-deposited platinum films with different resistivity. TEM morphology (a) and (d), EDS Pt map (b) and (e), and EDS C map (c) and (f). The images are adapted from DiBattista et al.[20] (g) The µSHD curves of the images. (a) (b) (c) (g) (d) (e) (f)

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 2 18 ACKNOWLEDGMENTS This material is based on work supported by the National Natural Science Foundation of China under Grant Number 51832002. The first author (ZH) also thanks Chi Yung Ng, the general chair of IPFA 2023, for his invitation and support for a keynote on the conference, which lays the basis for the material presented here. Szu Huat Goh is gratefully acknowledged for his insightful discussion that initiated the writing of this article and for comments on the draft from the perspective of industry. REFERENCES 1. Z. Huang: “Structural Hierarchy from Wavelet Zoom and Invariant Construction,” Discover Materials, 2021, 1(6). 2. Z. Huang, X. Luo, D. Jia, et al.: “Microstructural Hierarchy Descriptor (µSHD)–Property Correlations of Silicon Carbide Ceramics,” J. Eur. Ceram. Soc., 2022, 42(3), p. 801–819. 3. K. Zheng, Z. Huang, D. Jia, et al.: “A Uniform Rule Between Hardness and Structure Hierarchy of Ceramics in Both Direct and Inverse Hall–petch Regimes,” J. Am. Ceram. Soc., 2023, 106(10), p. 5620–5627. 4. www.formge.cn. [Accessed 03-02-2024]. 5. www.mgi.gov. [Accessed 03-02-2024]. 6. A. Rammal, K. Ezukwoke, A. Hoayek, et al.: “Root Cause Prediction for Failures in Semiconductor Industry, A Genetic Algorithm–machine Learning Approach,” Sci. Rep., 2023, 13, p. 4934. 7. S. Bi, C. Wang, B. Wu, et al.: “A Comprehensive Survey on Applications of AI Technologies to Failure Analysis of Industrial Systems,” Engineering Failure Analysis, 2023, 148, p. 107172. 8. A. Safont-Andreu, K. Schekotihin, C. Burmer, et al.: “Artificial Intelligence Applications in Semiconductor Failure Analysis,” EDFA, 2023, 25(2), p. 16–28. 9. www.zeiss.com/microscopy/en/products/ai-microscopy-solutions. html. [Accessed 03-02-2024]. 10. M. Botifoll, I. Pinto-Huguet, and J. Arbiol: “Machine Learning in Electron Microscopy for Advanced Nanocharacterization: Current Developments, Available Tools and Future Outlook,” Nanoscale Horiz., 2022, 7, p. 1427–1477. 11. G.B. Olson: “Computational Design of Hierarchically Structured Materials,” Science, 1997, 277(5330), p. 1237–1242. 12. C.S. Smith: “Structural Hierarchy in Science, Art, and History,” Aesthetics in Science, J. Wechsler, Ed., 1978, p. 9–53. 13. S. Mallat: “Understanding Deep Convolutional Networks,” Philosophical Transactions of the Royal Society A Mathematical, Physical and Engineering Sciences, 2016, 374(2065), p. 20150203. 14. D. Hubel and T. Wiesel: Brain and Visual Perception: The Story of a 25-year Collaboration, 2005, Oxford University Press. 15. S. Mallat: “Group Invariant Scattering,” Communications on Pure and Applied Mathematics, 2012, 65(10), p. 1331–1398. 16. W. Vanderlinde: “Scanning Electron Microscopy,” Microelectronics Failure Analysis: Desk Reference, 2019, ASM International. 17. C.D. Hartfield, T.M. Moore, and S. Brand: “Acoustic Microscopy of Semiconductor Packages,” Microelectronics Failure Analysis: Desk Reference, 2019, ASM International. 18. A. Orozco: “Magnetic Field Imaging for Electrical Fault Isolation,” Micro- electronics Failure Analysis: Desk Reference, 2019, ASM International. 19. W. Vanderlinde: “Energy Dispersive X-ray Analysis,” Microelectronics Failure Analysis: Desk Reference, 2019, ASM International. 20. M. DiBattista and T. Lundquist: “Role of Advanced Circuit Edit for First Silicon Debug,” Microelectronics Failure Analysis: Desk Reference, 2019, ASM International. ABOUT THE AUTHOR Zhiheng Huang is an associate professor at the School of Materials Science and Engineering of Sun Yat-sen University. He received his B.Eng. and M.Eng. degrees in Materials from Harbin Institute of Technology in 2000 and 2002, respectively, and his Ph.D. in manufacturing engineering from Loughborough University in 2005. He is the author of more than 45 journal articles and has written five book chapters. His current research interests include materials and reliability modeling in 3D microelectronic packaging, materials genome engineering, and integrated computational materials engineering. He is a member of IEEE. Advertise in Electronic Device Failure Analysis magazine! For information about advertising in Electronic Device Failure Analysis: Mark Levis, Business Development Manager 440.671.3834, mark.levis@asminternational.org Current rate card may be viewed online at asminternational.org/advertise.

RkJQdWJsaXNoZXIy MTYyMzk3NQ==