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 55 56 57 58

Feb_EDFA_Digital

A RESOURCE FOR TECHNICAL INFORMATION AND INDUSTRY DEVELOPMENTS FEBRUARY 2024 | VOLUME 26 | ISSUE 1 ELECTRONIC DEVICE FAILURE ANALYSIS edfas.org FOUR-DIMENSIONAL STEM: PART II: APPLICATIONS THINNING AND POLISHING HIGHLY WARPED DIE: PART III HIGHLIGHTS FROM ISTFA 2023 ADVANCED CHARACTERIZATION USING ATOM PROBE TOMOGRAPHY 4 24 14 32

A RESOURCE FOR TECHNICAL INFORMATION AND INDUSTRY DEVELOPMENTS FEBRUARY 2024 | VOLUME 26 | ISSUE 1 ELECTRONIC DEVICE FAILURE ANALYSIS edfas.org FOUR-DIMENSIONAL STEM: PART II: APPLICATIONS THINNING AND POLISHING HIGHLY WARPED DIE: PART III HIGHLIGHTS FROM ISTFA 2023 ADVANCED CHARACTERIZATION USING ATOM PROBE TOMOGRAPHY 4 24 14 32

edfas.org 1 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 DEPARTMENTS Advanced Characterization of Materials using Atom Probe Tomography Jacob M. Garcia and Ann N. Chiaramonti New materials integration and improved design can be promoted by using atom probe tomography as an analysis technique, shown through several diverse examples. 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 14 A RESOURCE FOR TECHNICAL INFORMATION AND INDUSTRY DEVELOPMENTS FEBRUARY 2024 | VOLUME 26 | ISSUE 1 edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS 2 GUEST EDITORIAL Michael DiBattista 36 CALL FOR PAPERS 37 PANEL & USER GROUP SUMMARY 47 2024 EDFAS AWARDS 48 2023 EDFAS AWARD WINNERS 49 BOARD OF DIRECTORS NEWS Chris Richardson 51 CALL FOR NOMINATIONS James Demarest 52 LITERATURE REVIEW Michael R. Bruce 53 PRODUCT NEWS Ted Kolasa 55 TRAINING CALENDAR Rosalinda Ring 56 ADVERTISERS INDEX Processes for Thinning and Polishing Highly Warped Die to a Nearly Consistent Thickness: Part III Kirk A. Martin The processes and considerations for both global and area of interest are discussed and reference process recipes are given in Part III of this series. 24 For the digital edition, log in to edfas.org, click on the “News/Magazines” tab, and select “EDFA Magazine.” Four-dimensional Scanning Transmission Electron Microscopy: Part II, Crystal Orientation and Phase, Short and Medium Range Order, and Electromagnetic Fields Aaron C. Johnston-Peck and Andrew A. Herzing The second part of this series presents applications of 4D-STEM, including measurements of crystal orientation and phase, short- and medium-range order, and internal electromagnetic fields. 14 4 ISTFA 2023 Highlights A recap of the ISTFA 2023 event includes General Chair Frank Altmann’s wrap-up as well as a list of the winning ISTFA papers and posters. 32 ABOUT THE COVER Sintering-induced crack formation in an inkjet-printed silver nanoparticle metallization line. This was observed during the development of a novel sample preparation method aimed at rerouting probe pads to enhance the imaging resolution of microscopy-based debug techniques. The colorful background stems from a thin film interference effect in the deposited dielectric film. Photo by Kristof J.P. Jacobs, IMEC, First Place Winner in Color Images, 2023 EDFAS Photo Contest. 24 32

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 2 The Defense Advanced Research Project Agency (DARPA) launched the second Electronics Resurgence Initiative (ERI 2.0) Summit on August 22-24, 2023, in Seattle as a response to the United States government’s nationwide microelectronic manufacturing concerns. This is one of many U.S. government initiatives to stimulate domestic semiconductor manufacturing including: SHIP – State of the art Heterogeneous Integrated Packaging RESHAPE – Re-Shore Ecosystem for Secure Heterogeneous Advanced Packaging Electronics RAMP-C – Rapid Assured Microelectronics Prototypes NSTC – National Semiconductor Technology Center NAPMP – National Advanced Packaging Manufacturing Program ME Commons – Microelectronics Commons NGMM – Next Generation Microelectronics Manufacturing ERI – Electronics Resurgence Initiative The ERI 2.0 DARPA-based initiative from the Microsystems Technology Office (MTO) is focused on driving next generation dual use microelectronics for national security and domestic needs by creating U.S. capability for threedimensional heterogeneous integration (3DHI) manufacturing and pursuing focused research for the manufacture of complex 3D microsystems. This effort represents an inflection point and brings forward significant challenges and opportunities for the failure analysis community to contribute successfully to 3DHI manufacturing. Semiconductor, defense industry, academia, and government leaders from across the U.S. presented their vision and insight for next generation 3DHI semiconductors manufacturing during the three-day conference with the keynote addresses by Patrick Gelsinger, CEO, Intel Corp. and Stefanie Tompkins, DARPA Director. Semiconductor industry leaders and expert talks throughout the conference included Pooya Tadayon, Fellow and Director of Assembly and Test Pathfinding, Intel Corp., who spoke on Challenges & Opportunities in 3D IC Test, Ahmad Bahai, Senior Vice President and Chief Technology Officer, Texas Instruments, talked about the Impact of 3D Heterogeneously Integrated Microsystems on Analog, Mixed-Signal, and RF Applications and Capabilities. Joshua Fryman, Fellow, Intel Corp. talked about New Applications Enabled by Complex 3D Microsystems, and Bill Dally, Chief Scientist and Senior Vice President of Research, Nvidia Corp. also presented. The slides and presentations for all the plenary talks are available online at https://eri-summit.darpa.mil/2023-agenda. FEBRUARY 2024 | VOLUME 26 | ISSUE 1 A RESOURCE FOR TECHNICAL INFORMATION AND INDUSTRY DEVELOPMENTS ELECTRONIC DEVICE FAILURE ANALYSIS GUEST EDITORIAL THE ELECTRONICS RESURGENCE INITIATIVE 2.0 FOR U.S. SEMICONDUCTOR MANUFACTURING Michael DiBattista, Varioscale miked@varioscale.com edfas.org DiBattista (continued on page 50) 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.

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 4 EDFAAO (2024) 1:4-13 1537-0755/$19.00 ©ASM International® FOUR-DIMENSIONAL SCANNING TRANSMISSION ELECTRON MICROSCOPY: PART II, CRYSTAL ORIENTATION AND PHASE, SHORT AND MEDIUM RANGE ORDER, AND ELECTROMAGNETIC FIELDS Aaron C. Johnston-Peck and Andrew A. Herzing National Institute of Standards and Technology, Gaithersburg, Maryland aaron.johnston-peck@nist.gov INTRODUCTION The second part of this series presents applications of four-dimensional scanning transmission electron microscopy (4D-STEM). The first article of this series, published in the August 2023 issue of EDFA, described a multitude of ways to characterize a sample’s structure. As a refresher, 4D-STEM is a spatially resolved electron diffraction technique that records the electron scattering distribution (kx, ky) at each point of the electron beam raster (rx, ry), thereby producing a four-dimensional dataset (rx, ry, kx, ky). This article covers measurements of crystal orientation and phase, short- and medium-range order (SRO/MRO), and internal electromagnetic fields. CRYSTAL ORIENTATION AND PHASE MAPPING Orientation and phase mapping of materials is commonly carried out using electron backscatter diffraction (EBSD) in the scanning electron microscope (SEM). This involves measurement of the position and angles between lines in the Kikuchi patterns formed at each position in an image scan. More recently, the spatial resolution of this method has been improved by collecting the forward- scattered electrons in a transmission Kikuchi diffraction (TKD) experiment.[1] For example, TKD was recently used to map the phases present in Zr-doped HfO2 films used in nonvolatile memory applications.[2] The development of high-speed direct electron detection cameras has enabled the acquisition of such data over very large fields of view, which would have been prohibitively slow using CCDbased cameras. The utility of this increased analytical area was recently demonstrated by using TKD to characterize a variety of 2D materials such as graphene and MoS2. [3] The resulting maps showed the orientation and phase over several millimeters, in some cases, while retaining spatial resolution below 10 nanometers. The spatial resolution of this type of measurement can be improved still further using a 4D-STEM approach. A reduction in interaction volume, due to the higher beam energies available in STEM combined with the thin specimens employed, further improves the spatial resolution and signal above background for the diffraction peaks. As in EBSD/TKD, STEM-based orientation mapping can be performed by analyzing the Kikuchi lines present in the patterns. This approach is typically only an option for thicker specimens since the Kikuchi lines are formed by diffracted electrons that had previously undergone diffuse scattering. In thinner samples, Kikuchi lines are much weaker or often invisible and therefore cannot be used for phase and orientation mapping by 4D-STEM. In this case, the measurement is based on indexing individual diffraction peaks.[4] Both commercial[5] and open-source software packages (references 6 and 7, for example,) are available for this task. Several approaches have been demonstrated to improve the fidelity of analyzing the Bragg reflections to determine sample orientation. As discussed in Part I, the ability to precess the incident beam about the optic axis during 4D-STEM data acquisition can be invaluable. In the case of orientation and phase mapping, the primary benefit of precession is that the resulting patterns can be more readily fit to models due to the reduction in dynamic scattering effects.[5] However, precession requires additional specialized hardware and poses some experimental constraints. A recently proposed alternative is multibeam diffraction. This technique utilizes a diaphragm with multiple beam forming apertures, thereby generating multiple electron probes on the sample and hence multiple diffraction patterns are simultaneously

edfas.org 5 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 generated at each point of the electron beam raster.[8] Multibeam diffraction was reported to improve the orienta- tion mapping of a small-grained sample compared to a map generated using a single beam measurement.[8] Template matching is frequently used to index diffraction patterns, where the patterns for all possible orientations of each phase are pre-calculated and compared to the experimental data by cross-correlation. This means the angular resolution depends on the density of templates simulated and thereby creates a tradeoff in angular resolution versus the time to generate the initial library plus comparing it to experimental results.[9,10] Angular resolutions as small as ≈ 0.3° have been reported using template matching, but is commonly on the order of 1 degree.[11] Improvements to angular resolution have been realized using artificial neural networks (ANNs). ANNs were trained using diffraction patterns simulated using the Bloch wave algorithm, meaning it incorporated dynamical scattering effects.[12] The angular resolution of the ANN-based analysis was reportedly 0.009°, over an order of magnitude improvement compared to previous reports using tem- plate matching. Henry et al.[13] used a combination of precession 4D-STEM and elemental mapping via energy dispersive x-ray spectroscopy (EDX) to determine the RESET failure mechanism in Ge-Sb-Te, a non-volatile phase-change memory (Fig. 1, Panel I). By using the 4D-STEM phase mapping results to segment the EDX maps, they were able to demonstrate that failure occurred due to depletion of the Ge content in the amorphous volume of the cell while the composition of the crystalline regions was unchanged. Another semiconductor application of 4D-STEM was reported by daSilva et al.[14] who investigated the formation of superlattices in PbS nanocrystal assemblies. By precisely measuring the position and orientation of nanocrystals in the transient region of these assemblies, they were able to determine that the transformation and subsequent superlattice orientation is dominated by the correlated alignment of the individual nanocrystals. Another interesting application reported by Londoño-Calderon et al.[15] quantified the orientation variation along the length of semiconducting Te nanowires and therefore was able to measure the overall twist rate and chirality (Fig. 1, Panel II). Finally, the speed and sensitivity of modern detectors has enabled phase and orientation mapping of semiconducting materials prone to electron beam damage. Panova et al.[16,17] used 4D-STEM to characterize conjugated semiconducting polymers and copolymers. By characterizing the response of the materials to the electron dose and staying below the damage threshold for a given material, they were able to map the crystal location and orientation in these technologically important soft materials. Finally, this approach was extended by Wu et al.[18] who employed a modified electron optical approach to improve the signal intensity and lower the required electron dose even further. In doing so, they were able to monitor the evolution of crystal size, distribution, and orientation as a function of heat treatment by serial collection of 4D datasets while remaining below the dose threshold. Beyond using zero order Laue zone (ZOLZ) Bragg reflections, higher order Laue zone (HOLZ) features also provide unique measurement opportunities. Where the ZOLZ includes the origin of the reciprocal lattice and the transmitted beam, HOLZ reflections originate from diffraction vectors that are not perpendicular to the electron beam direction. As a result, the HOLZ reflection contains information about the crystal parallel to the beam direction and, as such, can provide information about the threedimensional ordering of the crystal. Analysis of HOLZ diffraction features were used to map a displacive transformation in a perovskite heterostructure of La0.7Sr0.3MnO3/ LaFeO3/SrTiO3, where cell tilting and La A-site modulation in the LaFeO3 is suppressed at both interfaces. [19] DISORDERED MATERIALS So far, this article has discussed crystalline materials, that is, materials with long range order (translational symmetry) and by doing so, ignored an extensive category of technologically important materials, i.e., disordered or amorphous materials. The lack of long-range order poses unique challenges to characterization. For example, highresolution TEM (HRTEM) images of amorphous materials can depict fringes which may be tempting to interpret as crystal lattice planes and the presence of local ordering. However, these fringes can be an artifact of the band filtering action of the image forming (objective) lens.[20] Additionally, a small local region of order may not be detected above the background when embedded in a thicker disordered matrix. In response to such challenges, several techniques have been developed providing insight into the structure of amorphous materials. While many of these techniques predated modern day 4D-STEM, the large, low-noise datasets provided by modern pixelated STEM detectors are advantageous because many of these techniques are sensitive to noise and also rely on statistical methods which benefit from greater sampling. The pair distribution function (PDF) and its variants (e.g., radial distribution function), describe the probability distribution of interatomic distances.[21] A benefit of using STEM to generate PDFs is the small interaction volume;

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 6 therefore, small fluctuations in atomic correlations due to local order are not lost due to spatial averaging.[22] When implemented as a 4D-STEM technique, variations of the PDF can be tracked as a function of position. This has been done in metallic glasses,[23,24] semiconductor polymer blends,[25] and metal-organic framework (MOF) Fig. 1 Panel I: Correlated 4D-STEM and EDX analysis of Ge-Sb-Te phase-change memory material. Results from RESET-Pass (a-d) and RESET-Fail (e-h) cells are shown. Diffraction correlation index maps (a,e) and phase maps (b,f) from the TiN, Ge, and Ge2Sb2Te5 phases. Also shown are the EDX spectral abundance maps for the Ge2Sb2Te5 (c,g) and Ge (d,h) extracted via vertex component analysis. Reproduced with permission from Ref 13. Panel II: Structural phase map of a single tellurium nanowire (a) where each color represents one of the real space regions (b) with a common diffraction pattern signature. Also shown are the average diffraction pattern for each class (c) and the class patterns with disc intensity weights overlaid in red (d). Reproduced with permission from Ref 15. (a) (e) (b) (f) (c) (g) (d) (h) (a) (b) (c) (d) I. II.

edfas.org 7 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 composites,[26] where data dimensionality reduction techniques, including principal or independent component analysis, can aid in identifying variations in the PDF.[26,27] Because the PDF is a two-body distribution function it is sensitive to SRO; but it does not elucidate MRO well,[28,29] where MRO are correlations on length scales of approximately 1.0 to 2.5 nm. Whereas fluctuation electron microscopy (FEM) is dependent on higher order (two, three, and four body) pair-pair combinations and thereby gains sensitivity to MRO.[29] Intensity fluctuations in diffraction patterns due to coherent interference, sometimes called speckle, can result from regions of local order. From a series of diffraction patterns the variance about the mean can be computed and it is this operation that creates the dependence on higher order pair-pair combinations. In a more intuitive sense, the variance identifies the scattering angles, hence length scales, exhibiting the greatest fluctuation (Fig. 2, Panel I). Furthermore, the size of the electron probe can then be varied through control of the convergence angle as an additional variable to interrogate MRO because the sensitivity of FEM to any ordering will depend on the relative length scales of the probe size and ordered domain.[29] FEM has been used to characterize amorphous semiconductors,[30,31] dielectrics,[32,33] and metallic glasses.[34] It has also been used to elucidate structure-property relationships (Fig. 2, Panel II).[33,35,36] In amorphous silicon, FEM has been used to test the validity of different theoretical models of the structure (e.g., continuous random network versus paracrystalline) and it was concluded that paracrystalline models match the data best.[37,38] Other analysis routines have been used to identify specific rotational symmetries, e.g., twofold up to tenfold, within diffraction patterns.[33,39,40] This would include symmetry elements (e.g., five-fold) that are prohibited in translational symmetry, but which are highly relevant for glasses which contain quasicrystals. When combined with spatially resolved measurements, one can map the spatial extent of different symmetry elements (Fig. 2, Panel III). In a Zr36Cu64 metallic glass Liu et al. [39] used this approach to identify icosahedral clusters as the dominant feature and from spatial distribution maps made inferences into the nature of MRO. While Huang et al.[41] was able to identify the coexistence of SRO with icosahedral, bicapped square antiprisms, and tricapped trigonal prism geometries in Pd77.5Cu6Si16.5. INTERNAL FIELDS Beyond crystallography and structure, STEM can also directly detect other physical properties of a material. For instance, it is possible to map and measure electromagnetic fields distributed within materials, due to the interaction between electromagnetic fields and the electron beam. The origin of the signal can be understood in a classical physics context as the Lorentz force acting upon the electron beam, or through quantum mechanics and the Aharonov-Bohm effect, which describes how an electromagnetic potential introduces a phase shift to the electron wave.[42] Using STEM to detect magnetic fields can be traced back to the 1970s. Shortly after Rose[43] proposed that new modes of phase contrast could be achieved using two detectors in parallel, Dekkers and de Lang[44] described an implementation of differential phase contrast (DPC) in the scanning transmission electron microscope relying on a split detector where the DPC signal is the difference of the signal from the two individual segments. In 1978, Chapman and coworkers[45] then demonstrated that DPCSTEM could map magnetic domains in permalloy and iron. These studies using segmented detectors do not qualify as 4D-STEM, because they do not use a pixelated detector that enables fine angular sampling of the diffraction plane. Yet through coarse angular sampling of reciprocal space, segmented detectors are gaining sensitivity to the distribution of electrons in the diffraction plane as a function of probe and can therefore be considered as an approximate of recording the full diffraction pattern.[46] Even though this article focuses on 4D-STEM measurements, some of the articles cited will be those using segmented detectors. These articles are included because many salient points for the theory and measurement artifacts of mapping electromagnetic fields are relevant regardless of detector type. A sampling of electromagnetic field mapping applications using 4D-STEM includes: mapping magnetic domains in various materials including permalloy,[47,48] Fe 60Al40, [49,50] FeRh,[51] NiFe,[52] plus identifying skyrmions in FeGe thin films[53,54] and the local magnetization in Fe 2As, [55] an antiferromagnet. Electric fields have been measured in Si[56,57] and GaAs[58] p-n junctions, along with polarization-induced internal electric fields in AlN/GaN nanowire heterostructures.[59] This set of applications extends even further if one considers studies using segmented detectors.[60] While the discussion here only covers long-range fields, there is an active body of research which studies the measurement of atomic electric fields as well.[61,62] If we consider the action of the Lorentz force on the electron beam in a sample, the angular deflection (β) due to the projected in-plane magnetic field is described as β = eB0lt/h, where B0 is the local magnetic field, e is the electron charge, λ is the wavelength of the electron beam,

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 8 Fig. 2 Panel I: (a) Example nanodiffraction patterns obtained using three different probe sizes. (b) A plot of the normalized variance, V, of the diffracted intensity as a function of scattering angle, k, for six different probe sizes. The variance is at maximum for small probes sizes suggesting a large degree of structural fluctuation in the Cu64.5Zr35.5 metallic glass at those length scales. Reproduced with permission from Ref 34. Panel II: (a) Normalized variance plots of amorphous TiO2 films grown at different temperatures. (b) Plots of the averaged angular correlation function reveals the presence of rotational symmetry by identifying angular correlation between diffraction speckles at the same k. Reproduced under the terms of a Creative Commons CC BY license.[33] Panel III: (a) Symmetry coefficients calculated from a 4D dataset acquired from Pd77.5Cu6Si16.5 along with spatial maps depicting regions exhibiting (b) 2-fold, (c) 4-fold, (d) 6-fold, (e) 8-fold, and (f) 10-fold symmetry. Reproduced with permission from Ref 41. I. II. III. (a) (b) (a) (b) (a) (b) (c) (d) (e) (f)

edfas.org 9 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 Fig. 3 Panel I: A schematic depicting the behavior of the diffraction disks as a function of potential. If there is no gradient the disk does not shift (a). The diffraction pattern shifts rigidly when experiencing a constant gradient (b), while the intensity within the diffraction disks is redistributed when the potential is smaller than the electron probe (c). α is the probe convergence semi-angle. Panel II: Competition between diffraction contrast and magnetic field contrast in a polycrystalline FeGe thin film can be observed as a function of collection angle. COM in the x direction as a function of collection angles: (0 to 430) μrad (a); (0 to 700) μrad (b); and (430 to 700) μrad (c). Grain contrast is evident in (a) while in (c) contrast due to the magnetic fields are greatly enhanced. Because both signals are present in the 4D dataset both crystallographic (d) and magnetic induction (e) orientation maps can be calculated. The scale bar in (d) is equal to 200 nm. Reproduced with permission from Ref 53. I. II. (a) (b) (c) (a) (b) (d) (e) (c)

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 10 t is the sample thickness, and h is Planck’s constant.[42] Therefore, by analyzing the 4D dataset, the strength of the electromagnetic field can be obtained by measuring the deflection angle via the shift of the transmitted beam at each scan point. Two important components of this framework are measuring the disk shift as precisely as possible and understanding the limits of this model. The “rigid-intensity-shift model” or “rigid disk shift” model assumes that the electromagnetic field shifts the diffraction disk in the far-field. While its limits have been discussed in detail elsewhere,[62-64] the basics are that the scattering behavior will depend on the width of the electron beam relative to the phase gradient induced by the electromagnetic potential. If a potential does not induce I. II. (a) (a) (b) (c) (d) (b) Fig. 4 Panel I: Electric field vector color maps of a GaN/AlGaN heterostructure obtained from the [1120] zone axis, with (a) no tilt-scan averaging and (c) tilt-scan averaged DPC images. The scale bars denote 30 nm. (b) and (d) show the line profiles of COM X and COM Y of (a) and (c), respectively. Reproduced with permission from Ref 67. Panel II: A Landau domain structure, comparing results from the same region obtained by (a) traditional COM and (b) analysis based on the circular Hough transform, where the fidelity of the analysis has been improved. Reproduced with permission from Ref 47.

edfas.org 11 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 a gradient, then the disk does not shift (Fig. 3, Panel I, a). If the gradient is larger than the probe, it will displace the disk (Fig. 3, Panel I, b); however, if the potential is smaller than the probe, intensity within the diffraction disk will change (Fig. 3, Panel I, c). Moreover, because samples are heterogenous there are many phenomena that can redistribute intensity within the diffraction pattern. This can include sample bending and thickness variations[47] as well as abrupt changes in mean inner potential due to interfaces.[65] Meaning that multiple signals can be convoluted and it becomes necessary to identify data acquisition methods and analysis routines which can mitigate the contribution of these other signals. This concept of multiple signals is well illustrated in a polycrystalline FeGe film (Fig. 3, Panel II) where contrast from the magnetic fields, as well as from changes to the diffraction condition with respect to the incident beam (i.e., diffraction contrast), can be observed.[53] One ubiquitous phenomenon that is the origin of competing signals is dynamical diffraction, which causes redistribution of intensity in the diffraction discs as a function of thickness and sample tilt, for example. One method to mitigate this effect is to acquire data from multiple orientations and average the result, which suppresses rapid variations in the signal due to dynamical diffraction and changes in local orientation.[66] This may be implemented by tilting the sample[54] or incident beam[67] incrementally and acquiring multiple scans or by precession,[56,66] where the beam is tilted over an angular range at each position in the scan. As exemplified in Fig. 4, Panel I, the unwanted diffraction contrast is suppressed improving visualization of the electric fields.[67] However, for this particular result there was still not an agreement when compared to simulations predicting the electric field distribution. It was hypothesized that sample preparation may be responsible for the discrepancy, where Ga ion implantation (from focused ion beam sample preparation) are known to passivate dopants near surfaces.[68] This means additional steps may be needed during data analysis or sample preparation to ensure accurate and representative results. As for measuring the beam shift, the center of mass (COM), also known as the first moment, of a diffraction pattern is commonly used. Yet, as previously mentioned, factors such as dynamical diffraction introduce signals that do not reflect long-range electromagnetic fields. Different schemes have been implemented to reduce these unwanted effects. To suppress diffraction contrast in a polycrystalline FeGe film the COM was calculated using an annulus around the direct beam disk edge.[53] Wu and coworkers[47] applied a circular Hough transform filter and then refined the position using a nonlinear trustregion algorithm. This achieves a significant improvement over traditional COM measurements, as shown in Fig. 4, Panel II. Similar filtering and position refinement schemes have also reported improved performance over COM measurements.[56] A natural extension of field measurements is to measure these materials under stimuli. Temperature, electrical voltage or current, and magnetic field can all be readily applied using specialized holders or by tuning the remanent field of the objective lens. This can be used to access conditions where certain features, such as skyrmions, are stable,[53] as well as observing evolving features, such as magnetic domains as a function of temperature and field;[51] the evolution of internal fields at pn-junctions have been measured under different applied voltages.[57] These measurements could be extended further by increasing the time resolution to capture transient states or using tomography to discern the field in three-dimensions rather than measuring a projected, through-thickness value. SUMMARY This article covered techniques measuring structure (phase and orientation and SRO or MRO) as well as properties (electromagnetism). Relationships between processing, structure, and properties are frequently complex making it challenging to disentangle these interconnected relationships. Because 4D datasets contain an immense amount of information understanding these complex property-processing-structure relationships becomes more accessible. This possibility is represented nicely in Fig. 3, Panel II, where both the magnetic field and crystallography of the sample were extracted from the same dataset. One could envision a similar study where a processing condition is applied in situ and both the property and the structure could then be tracked as a function of time. Part III of this series will cover the topic of ptychography. REFERENCES 1. R.R. Keller and R.H. Geiss: “Transmission EBSD from 10 nm Domains in a Scanning Electron Microscope,” Journal of Microscopy, 2012, 245(3), p. 245-251. 2. M. Lederer, et al.: “Local Crystallographic Phase Detection and Texture Mapping in Ferroelectric Zr Doped HfO2 Films by Transmission-EBSD,” Applied Physics Letters, 2019, 115(22). 3. A. Orekhov, et al.: “Wide Field of View Crystal Orientation Mapping of Layered Materials,” arXiv, 2020. 4. Y. Meng and J.-M. Zuo: “Improvements in Electron Diffraction Pattern Automatic Indexing Algorithms,” Eur. Phys. J. Appl. Phys., 2017, 80(1), p. 10701.

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 12 5. E.F. Rauch, et al.: “Automated Nanocrystal Orientation and Phase Mapping in the Transmission Electron Microscope on the Basis of Precession Electron Diffraction,” Zeitschrift für Kristallographie, 2010, 225(2-3), p. 103-109. 6. B.H. Savitzky, et al.: “Py4DSTEM: A Software Package for FourDimensional Scanning Transmission Electron Microscopy Data Analysis,” Microscopy and Microanalysis, 2021, 27(4), p. 712-743. 7. Pyxem/Kikuchipy: Kikuchipy 0.8.6., 2023. 8. X. Hong, et al.: “Multibeam Electron Diffraction,” Microscopy and Microanalysis, 2021, 27(1), p. 129-139. 9. E.F. Rauch and M. Véron: “Automated Crystal Orientation and Phase Mapping in TEM,” Materials Characterization, 2014, 98, p. 1-9. 10. A. Morawiec, et al.: “Orientation Precision of TEM-Based Orientation Mapping Techniques,” Ultramicroscopy, 2014, 136, p. 107-118. 11. J. Jeong, et al.: “Automated Crystal Orientation Mapping by Precession Electron Diffraction-Assisted Four-Dimensional Scanning Transmission Electron Microscopy using a Scintillator-Based CMOS Detector,” Microscopy and Microanalysis, 2021, 27(5), p. 1102-1112. 12. R. Yuan, et al.: “Training Artificial Neural Networks for Precision Orientation and Strain Mapping Using 4D Electron Diffraction Datasets,” Ultramicroscopy, 2021, 231, p. 113256. 13. L. Henry, et al.: “Studying Phase Change Memory Devices by Coupling Scanning Precession Electron Diffraction and Energy Dispersive X-ray Analysis,” Acta Materialia, 2020, 201, p. 72-78. 14. J.C. daSilva, et al.: “Mechanistic Insights into Superlattice Transformation at a Single Nanocrystal Level using Nanobeam Electron Diffraction,” Nano Letters, 2020, 20(7), p. 5267-5274. 15. A. Londoño-Calderon, et al.: “Intrinsic Helical Twist and Chirality in Ultrathin Tellurium Nanowires,” Nanoscale, 2021, 13(21), p. 9606-9614. 16. O. Panova, et al.: “Orientation Mapping of Semicrystalline Polymers using Scanning Electron Nanobeam Diffraction,” Micron, 2016, 88, p. 30-36. 17. O. Panova, et al.: “Diffraction Imaging of Nanocrystalline Structures in Organic Semiconductor Molecular Thin Films,” Nature Materials, 2019, 18(8), p. 860-865. 18. M. Wu, et al.: “Seeing Structural Evolution of Organic Molecular Nano-Crystallites using 4D Scanning Confocal Electron Diffraction (4D-SCED),” Nature Communications, 2022, 13(1), p. 2911. 19. M. Nord, et al.: “Three-Dimensional Subnanoscale Imaging of Unit Cell Doubling Due to Octahedral Tilting and Cation Modulation in Strained Perovskite Thin Films,” Physical Review Materials, 2019, 3(6), p. 063605. 20. O.L. Krivanek, P.H. Gaskell, and A. Howie: “Seeing Order in ‘Amorphous’ Materials,” Nature, 1976, 262(5568), p. 454-457. 21. E. Takeshi and S.J.L. Billinge: “Chapter 3 - the Method of Total Scattering and Atomic Pair Distribution Function Analysis,” in Pergamon Materials Series, Vol. 16, T. Egami and S. J. L. Billinge Eds.: Pergamon, 2012, p. 55-111. 22. D.J.H. Cockayne and D.R. McKenzie: “Electron Diffraction Analysis of Polycrystalline and Amorphous Thin Films,” Acta Crystallographica Section A, 1988, 44(6), p. 870-878. 23. X. Mu, et al.: “Radial Distribution Function Imaging by STEM Diffraction: Phase Mapping and Analysis of Heterogeneous Nanostructured Glasses,” Ultramicroscopy, 2016, 168, p. 1-6. 24. Y. Rakita, et al.: “Mapping Structural Heterogeneity at the Nanoscale with Scanning Nano-Structure Electron Microscopy (SNEM),” Acta Materialia, 2023, 242, p. 118426. 25. X. Mu, et al.: “Mapping Structure and Morphology of Amorphous Organic Thin Films by 4D-STEM Pair Distribution Function Analysis,” Microscopy, 2019, 68(4), p. 301-309. 26. J.E.M. Laulainen, et al.: “Mapping Short-Range Order at the Nanoscale in Metal-Organic Framework and Inorganic Glass Composites,” Nanoscale, 2022, 14(44), p. 16524-16535. 27. X. Mu, et al.: “Unveiling Local Atomic Bonding and Packing of Amorphous Nanophases Via Independent Component Analysis Facilitated Pair Distribution Function,” Acta Materialia, 2021, 212, p. 116932. 28. M.M.J. Treacy: “Fluctuation Microscopy: A Probe of Medium Range Order,” Reports on Progress in Physics, 2005, 68(12), p. 2899. 29. M.M.J. Treacy: “Speckles in Images and Diffraction Patterns,” Handbook of Nanoscopy, 2012, p. 405-435. 30. D. Radić, et al.: “Fluctuation Electron Microscopy on Amorphous Silicon and Amorphous Germanium,” Microscopy and Microanalysis, 2023, 29(2), p. 477-489. 31. T.T. Li, K. Darmawikarta, and J.R. Abelson: “Quantifying Nanoscale Order in Amorphous Materials Via Scattering Covariance in Fluctuation Electron Microscopy,” Ultramicroscopy, 2013, 133, p. 95-100. 32. M.J. Hart, et al.: “Medium Range Structural Order in Amorphous Tantala Spatially Resolved with Changes to Atomic Structure by Thermal Annealing,” Journal of Non-Crystalline Solids, 2016, 438, p. 10-17. 33. M. Abbasi, et al.: “In Situ Observation of Medium Range Ordering and Crystallization of Amorphous TiO2 Ultrathin Films Grown by Atomic Layer Deposition,” APL Materials, 2023, 11(1). 34. J. Hwang and P.M. Voyles: “Variable Resolution Fluctuation Electron Microscopy on Cu-Zr Metallic Glass using a Wide Range of Coherent STEM Probe Size,” Microscopy and Microanalysis, 2011, 17(1), p. 67-74. 35. P.M. Voyles, et al.: “Absence of an Abrupt Phase Change from Polycrystalline to Amorphous in Silicon with Deposition Temperature,” Physical Review Letters, 2001, 86(24), p. 5514-5517. 36. H.C. Jacks, et al.: “Structural Tunability and Origin of Two-Level Systems in Amorphous Silicon,” Physical Review Materials, 2022, 6(4), p. 045604. 37. M.M.J. Treacy and K.B. Borisenko: “The Local Structure of Amorphous Silicon,” Science, 2012, 335(6071), p. 950-953. 38. J.M. Gibson and M.M.J. Treacy: “Fluctuation Microscopy Analysis of Amorphous Silicon Models,” Ultramicroscopy, 2017, 176, p. 74-79. 39. A.C.Y. Liu, et al.: “Systematic Mapping of Icosahedral Short-Range Order in a Melt-Spun Zr36Cu64 Metallic Glass,” Physical Review Letters, 2013, 110(20), p. 205505. 40. S. Huang, et al.: “Correlation Symmetry Analysis of Electron Nanodiffraction from Amorphous Materials,” Ultramicroscopy, 2022, 232, p. 113405. 41. S. Huang, et al.: “Large Area, High Resolution Mapping of Approximate Rotational Symmetries in a Pd77.5Cu6Si16.5 Metallic Glass Thin Film,” Ultramicroscopy, 2022, 241, p. 113612. 42. J.N. Chapman: “The Investigation of Magnetic Domain Structures in Thin Foils by Electron Microscopy,” Journal of Physics D: Applied Physics, 1984, 17(4), p. 623. 43. H. Rose: “Phase Contrast in Scanning Transmission Electron Microscopy” Optik, 1974, 39(4), p. 416-436. 44. N.H. Dekkers and H. de Lang: “Differential Phase Contrast in a STEM,” Optik, 1974, 41(4), p. 452-456. 45. J.N. Chapman, et al.: “The Direct Determination of Magnetic Domain Wall Profiles by Differential Phase Contrast Electron Microscopy,” Ultramicroscopy, 1978, 3, p. 203-214. 46. E.M. Waddell and J.N. Chapman: “Linear Imaging of Strong Phase Objects Using Asymmetrical Detectors in STEM,” Optik, 1979, 54, p. 83-96. 47. L. Wu, M.-G. Han, and Y. Zhu: “Toward Accurate Measurement of Electromagnetic Field by Retrieving and Refining the Center Position of Non-Uniform Diffraction Disks in Lorentz 4D-STEM,” Ultramicroscopy, 2023, 250, p. 113745. 48. M. Krajnak, et al.: “Pixelated Detectors and Improved Efficiency for Magnetic Imaging in STEM Differential Phase Contrast,” Ultramicroscopy, 2016, 165, p. 42-50.

edfas.org 13 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 49. M. Nord, et al.: “Strain Anisotropy and Magnetic Domains in Embedded Nanomagnets,” Small, 2019, 15(52), p. 1904738. 50. G. Nordahl and M. Nord: “Improving Magnetic STEM-Differential Phase Contrast Imaging using Precession,” Microscopy and Microanalysis, 2023, 29(2), p. 574-579. 51. T.P. Almeida, et al.: “Direct Visualization of the Magnetostructural Phase Transition in Nanoscale FeRh Thin Films Using Differential Phase Contrast Imaging,” Physical Review Materials, 2020, 4(3), p. 034410. 52. V. Boureau, et al.: “High-Sensitivity Mapping of Magnetic Induction Fields with Nanometer-Scale Resolution: Comparison of Off-Axis Electron Holography and Pixelated Differential Phase Contrast,” Journal of Physics D: Applied Physics, 2021, 54(8), p. 085001. 53. K.X. Nguyen, et al.: “Disentangling Magnetic and Grain Contrast in Polycrystalline FeGe Thin Films using Four-Dimensional Lorentz Scanning Transmission Electron Microscopy,” Physical Review Applied, 2022, 17(3), p. 034066. 54. B. Wang, et al.: “Extracting Weak Magnetic Contrast from Complex Background Contrast in Plan-View FeGe Thin Films,” Ultramicroscopy, 2022, 232, p. 113395. 55. K.X. Nguyen, et al.: “Angstrom-Scale Imaging of Magnetization in Antiferromagnetic Fe2As Via 4D-STEM,” Ultramicroscopy, 2023, 247, p. 113696. 56. L. Bruas, et al.: “Improved Measurement of Electric Fields by Nanobeam Precession Electron Diffraction,” Journal of Applied Physics, 2020, 127(20). 57. B.C. da Silva et al.: “Assessment of Active Dopants and P–N Junction Abruptness Using in Situ Biased 4D-STEM,” Nano Letters, 2022, 22(23), p. 9544-9550. 58. A. Beyer, et al.: “Quantitative Characterization of Nanometer-Scale Electric Fields Via Momentum-Resolved STEM,” Nano Letters, 2021, 21(5), p. 2018-2025. 59. K. Müller-Caspary, et al.: “Electrical Polarization in AlN/GaN Nanodisks Measured by Momentum-Resolved 4D Scanning Transmission ABOUT THE AUTHORS Aaron C. Johnston-Peck is a materials research engineer at the National Institute of Standards and Technology where he characterizes materials using various electron microscopy techniques. Prior to NIST, he worked as a postdoctoral fellow at Brookhaven National Laboratory and received his Ph.D. in materials science and engineering from North Carolina State University. He has contributed to 50 peer reviewed publications. Andrew A. Herzing is a materials research engineer at the National Institute of Standards and Technology where he develops materials characterization techniques with a particular interest in tomography and 4D-STEM. He received his Ph.D. in materials science and engineering from Lehigh University in 2007 and is currently serving as the president-elect of the Microanalysis Society. Electron Microscopy,” Physical Review Letters, 2019, 122(10), p. 106102. 60. T. Seki, Y. Ikuhara, and N. Shibata: “Toward Quantitative Electromagnetic Field Imaging by Differential-Phase-Contrast Scanning Transmission Electron Microscopy,” Microscopy, 2020, 70(1), p. 148-160. 61. K. Müller, et al.: “Atomic Electric Fields Revealed by a Quantum Mechanical Approach to Electron Picodiffraction,” Nature Communications, 2014, 5(1), p. 5653. 62. K. Müller-Caspary et al.: “Measurement of Atomic Electric Fields and Charge Densities from Average Momentum Transfers using Scanning Transmission Electron Microscopy,” Ultramicroscopy, 2017, 178, p. 62-80. 63. M.C. Cao, et al.: “Theory and Practice of Electron Diffraction from Single Atoms and Extended Objects using an EMPAD,” Microscopy, 2018, 67(suppl_1), p. i150-i161. 64. L. Clark, et al.: “Probing the Limits of the Rigid-Intensity-Shift Model in Differential-Phase-Contrast Scanning Transmission Electron Microscopy,” Physical Review A, 2018, 97(4), p. 043843. 65. I. MacLaren, et al: “On the Origin of Differential Phase Contrast at a Locally Charged and Globally Charge-Compensated Domain Boundary in a Polar-Ordered Material,” Ultramicroscopy, 2015, 154, p. 57-63. 66. T. Mawson, et al.: “Suppressing Dynamical Diffraction Artefacts in Differential Phase Contrast Scanning Transmission Electron Microscopy of Long-Range Electromagnetic Fields Via Precession,” Ultramicroscopy, 2020, 219, p. 113097. 67. S. Toyama, et al.: “Quantitative Electric Field Mapping in Semiconductor Heterostructures via Tilt-Scan Averaged DPC STEM,” Ultramicroscopy, 2022, 238, p. 113538. 68. B. Haas, et al.: “Direct Comparison of Off-Axis Holography and Differential Phase Contrast for the Mapping of Electric Fields in Semiconductors by Transmission Electron Microscopy,” Ultramicroscopy, 2019, 198, p. 58-72. 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/mediakit.

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 14 ADVANCED CHARACTERIZATION OF MATERIALS USING ATOM PROBE TOMOGRAPHY Jacob M. Garcia and Ann N. Chiaramonti National Institute of Standards and Technology, Applied Chemicals and Materials Division, Boulder, Colorado jacob.garcia@nist.gov EDFAAO (2024) 1:14-21 1537-0755/$19.00 ©ASM International® INTRODUCTION Electronic device failures are influenced by numerous factors, including the choice of material and associated processing methods, as well as the service conditions the device will see, e.g., temperature excursion, electric current, mechanical or thermal stress, cyclic or static electric fields. The fundamental factors that influence failure modes come down to the interplay between atoms within the material. Among the analytical techniques that provide sub-nm spatially resolved chemical information, atom probe tomography (APT) has emerged as the best compromise between analytical sensitivity and spatial resolution, as compared to other analytical methods such as secondary ion mass spectrometry (SIMS) and transmission electron microscopy (TEM). Owing to the single atom specificity, APT can provide 3D chemical maps, or tomograms, of samples comprising any element or isotope in the periodic table, with sub-nm spatial resolution in three dimensions. The ability to provide 3D reconstructions with sub-nm spatial resolution, and elemental specificity in the ppm range in some cases, has proven useful for diverse applications. To date, APT has been utilized for compositional profiling of geological minerals,[1] the direct observation of H poisoning at grain boundaries in steel,[2] cryogenically frozen biomaterial analyses,[3] and elucidating the atomic scale structure and composition of electronic materials and devices,[4] to name a few. Atom probe tomography is an evolution of field ion microscopy, which provided the first direct glimpse of atoms in 1955, but at the time worked exclusively on metals.[5] The first atom probe microscope followed in 1967 and has been iteratively improved by many groups since then, with improvements in spatial resolution, mass resolving power, and analytical sensitivity, as well as the variety of materials that can be analyzed through the addition of 3D detectors, energy compensating optics, local electrodes, and advances in pulsed laser technology. APT will be briefly reviewed here, focusing on the capabilities and advances to assist in the characterization of electronic devices. New materials integration and improved design can be promoted by using atom probe tomography as an analysis technique, shown through the following diverse examples in this article. EXPERIMENTAL OVERVIEW The sub-nm spatial resolution and analytical sensitivity of APT is obtained through a process known as field ion evaporation. Field evaporation requires applying a high standing voltage (1 to 10 kV) to a needle-shaped Fig. 1 A schematic showing the incoming laser pulse to a sample tip, resulting in field evaporation followed by ion detection on a 2D detector. Adapted from Ref 9.

edfas.org 15 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 sample, typically of diameter less than 100 nm at the apex. To encourage field ion evaporation, an additional external short high-voltage or laser pulse is used to emit the atoms already under the applied field.[6-8] The externally triggered pulse is used to selectively evaporate surface material from the tip on an atom-by-atom basis. By removing single layers of a sample, ideally one atom at a time, and collecting the evaporated ions on a positionsensitive time-of-flight (TOF) detector (Fig. 1), a mass spectrum of each pulsed event can be obtained, and the x,y coordinates where each ion was intercepted by the detector are recorded. Current APT instruments have both straight flight path and reflectron configurations, with the latter employing electrostatic lenses to increase the mass resolution. The averaged TOF mass spectrum and the ion coordinate information are used to create a 3D virtual “model,” with each voxel representing a sub-nm spatially resolved element (or isotope) from the original specimen. APT is based on a thermal process that follows an Arrhenius rate equation (Eq. 1). The rate of evaporation, k, is increased by either decreasing the potential energy barrier, Ea, of atoms on the surface by increasing the applied field using a voltage pulse, or by increasing the temperature, T, of the material using a thermal laser pulse. Eq. 1 Cryogenic temperatures are used to suppress element diffusion on the surface of the specimen needle and confine the evaporation event to coordinate with the trigger pulse, thus increasing the signal-to-noise ratio. To that end, APT samples are mounted on a specimen stage that is cryogenically cooled to temperatures as low as ~20 K using liquid He. ATOM PROBE SAMPLE PREPARATION To obtain the high field required to ideally liberate an individual atom from the specimen surface, needleshaped samples are employed. The required fields of 10 to 50 V/nm can be obtained by applying a 5 kV standing voltage, for example, to a 100 nm diameter (or less) needle-shaped specimen. To prepare APT specimens with such geometry, a focused ion beam (FIB) lift-out method is often used.[10] Figure 2 shows an example of APT tip preparation from a bulk substrate using this approach. First, the area of interest is identified, and a wedge is cut in the bulk material (Fig. 2a). The wedge is lifted out and a small section is welded to the tip of a sacrificial post (Fig. 2b) that is typically made from W, Si, or a TEM grid bar.[11] The sample welded to the post is then sculpted to the final needle-shaped geometry using the Ga+ ion beam and an annular milling scheme (Fig. 2c-e). Due to the numerous milling and imaging steps, a protective metal cap (often Ni, Cr, or FIB-deposited Pt) can be deposited first, over the sample surface to prevent Ga+ ion implantation in the top layers. FIB-scanning electron microscopy (SEM) methods also enable precise identification and extraction of site-specific Fig. 2 A summary of APT tip preparation from a bulk sample showing SEM images of the (a) FIB-prepared sample bar, (b) welded section transferred to an external post, and (c-e) the gradual sharpening of the sample tip by annular milling. Reproduced from Ref 11 under the terms of the Creative Commons Attribution License. (a) (c) (b) (d) (e)

RkJQdWJsaXNoZXIy MTYyMzk3NQ==