To save this undefined to your undefined account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your undefined account.
Find out more about saving content to .
To save this article to your Kindle, first ensure firstname.lastname@example.org is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below.
Find out more about saving to your Kindle.
Note you can select to save to either the @free.kindle.com or @kindle.com variations. ‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
Approximately 30 years after the first use of focused ion beam (FIB) instruments to prepare atom probe tomography specimens, this technique has grown to be used by hundreds of researchers around the world. This past decade has seen tremendous advances in atom probe applications, enabled by the continued development of FIB-based specimen preparation methodologies. In this work, we provide a short review of the origin of the FIB method and the standard methods used today for lift-out and sharpening, using the annular milling method as applied to atom probe tomography specimens. Key steps for enabling correlative analysis with transmission electron-beam backscatter diffraction, transmission electron microscopy, and atom probe tomography are presented, and strategies for preparing specimens for modern microelectronic device structures are reviewed and discussed in detail. Examples are used for discussion of the steps for each of these methods. We conclude with examples of the challenges presented by complex topologies such as nanowires, nanoparticles, and organic materials.
This article reviews recent advances utilizing field-ion microscopy (FIM) to extract atomic-scale three-dimensional images of materials. This capability is not new, as the first atomic-scale reconstructions of features utilizing FIM were demonstrated decades ago. The rise of atom probe tomography, and the application of this latter technique in place of FIM has unfortunately severely limited further FIM development. Currently, the ubiquitous availability of extensive computing power makes it possible to treat and reconstruct FIM data digitally and this development allows the image sequences obtained utilizing FIM to be extremely valuable for many material science and engineering applications. This article demonstrates different applications of these capabilities, focusing on its use in physical metallurgy and semiconductor science and technology.
Atom probe has been developed for investigating materials at the atomic scale and in three dimensions by using either high-voltage (HV) pulses or laser pulses to trigger the field evaporation of surface atoms. In this paper, we propose an atom probe setup with pulsed evaporation achieved by simultaneous application of both methods. This provides a simple way to improve mass resolution without degrading the intrinsic spatial resolution of the instrument. The basic principle of this setup is the combination of both modes, but with a precise control of the delay (at a femtosecond timescale) between voltage and laser pulses. A home-made voltage pulse generator and an air-to-vacuum transmission system are discussed. The shape of the HV pulse presented at the sample apex is experimentally measured. Optimizing the delay between the voltage and the laser pulse improves the mass spectrum quality.
The local electrode atom probe (LEAP) has become the primary instrument used for atom probe tomography measurements. Recent advances in detector and laser design, together with updated hit detection algorithms, have been incorporated into the latest LEAP 5000 instrument, but the implications of these changes on measurements, particularly the size and chemistry of small clusters and elemental segregations, have not been explored. In this study, we compare data sets from a variety of materials with small-scale chemical heterogeneity using both a LEAP 3000 instrument with 37% detector efficiency and a 532-nm green laser and a new LEAP 5000 instrument with a manufacturer estimated increase to 52% detector efficiency, and a 355-nm ultraviolet laser. In general, it was found that the number of atoms within small clusters or surface segregation increased in the LEAP 5000, as would be expected by the reported increase in detector efficiency from the LEAP 3000 architecture, but subtle differences in chemistry were observed which are attributed to changes in the way multiple hit detection is calculated using the LEAP 5000.
There are two main projections used to transform, and reconstruct, field ion micrographs or atom probe tomography data into atomic coordinates at the specimen surface and, subsequently, in three dimensions. In this article, we present a perspective on the strength of the azimuthal equidistant projection in comparison with the more widely used and well-established point projection (or pseudo-stereographic projection), which underpins data reconstruction in most software packages currently in use across the community. After an overview of the reconstruction methodology, we demonstrate that the azimuthal equidistant is more robust with regards to errors on the parameters used to perform the reconstruction and is therefore more likely to yield more accurate tomographic reconstructions.
Accuracy of atom probe tomography measurements is strongly degraded by the presence of phases that have different evaporation fields. In particular, when there are perpendicular interfaces to the tip axis in the specimen, layers thicknesses are systematically biased and the resolution is degraded near the interfaces. Based on an analytical model of field evaporated emitter end-form, a new algorithm dedicated to the 3D reconstruction of multilayered samples was developed. Simulations of field evaporation of bilayer were performed to evaluate the effectiveness of the new algorithm. Compared to the standard state-of-the-art reconstruction methods, the present approach provides much more accurate analyzed volume, and the resolution is clearly improved near the interface. The ability of the algorithm to handle experimental data was also demonstrated. It is shown that the standard algorithm applied to the same data can commit an error on the layers thicknesses up to a factor 2. This new method is not constrained by the classical hemispherical specimen shape assumption.
An automated procedure has been developed for the reconstruction of field ion microscopy (FIM) data that maintains its atomistic nature. FIM characterizes individual atoms on the specimen’s surface, evolving subject to field evaporation, in a series of two-dimensional (2D) images. Its unique spatial resolution enables direct imaging of crystal defects as small as single vacancies. To fully exploit FIM’s potential, automated analysis tools are required. The reconstruction algorithm developed here relies on minimal assumptions and is sensitive to atomic coordinates of all imaged atoms. It tracks the atoms across a sequence of images, allocating each to its respective crystallographic plane. The result is a highly accurate 3D lattice-resolved reconstruction. The procedure is applied to over 2000 tungsten atoms, including ion-implanted planes. The approach is further adapted to analyze carbides in a steel matrix, demonstrating its applicability to a range of materials. A vast amount of information is collected during the experiment that can underpin advanced analyses such as automated detection of “out of sequence” events, subangstrom surface displacements and defects effects on neighboring atoms. These analyses have the potential to reveal new insights into the field evaporation process and contribute to improving accuracy and scope of 3D FIM and atom probe characterization.
Accurately identifying and extracting clusters from atom probe tomography (APT) reconstructions is extremely challenging, yet critical to many applications. Currently, the most prevalent approach to detect clusters is the maximum separation method, a heuristic that relies heavily upon parameters manually chosen by the user. In this work, a new clustering algorithm, Gaussian mixture model Expectation Maximization Algorithm (GEMA), was developed. GEMA utilizes a Gaussian mixture model to probabilistically distinguish clusters from random fluctuations in the matrix. This machine learning approach maximizes the data likelihood via expectation maximization: given atomic positions, the algorithm learns the position, size, and width of each cluster. A key advantage of GEMA is that atoms are probabilistically assigned to clusters, thus reflecting scientifically meaningful uncertainty regarding atoms located near precipitate/matrix interfaces. GEMA outperforms the maximum separation method in cluster detection accuracy when applied to several realistically simulated data sets. Lastly, GEMA was successfully applied to real APT data.
Correlative microscopy approaches offer synergistic solutions to many research problems. One such combination, that has been studied in limited detail, is the use of atom probe tomography (APT) and transmission Kikuchi diffraction (TKD) on the same tip specimen. By combining these two powerful microscopy techniques, the microstructure of important engineering alloys can be studied in greater detail. For the first time, the accuracy of crystallographic measurements made using APT will be independently verified using TKD. Experimental data from two atom probe tips, one a nanocrystalline Al–0.5Ag alloy specimen collected on a straight flight-path atom probe and the other a high purity Mo specimen collected on a reflectron-fitted instrument, will be compared. We find that the average minimum misorientation angle, calculated from calibrated atom probe reconstructions with two different pole combinations, deviate 0.7° and 1.4°, respectively, from the TKD results. The type of atom probe and experimental conditions appear to have some impact on this accuracy and the reconstruction and measurement procedures are likely to contribute further to degradation in angular resolution. The challenges and implications of this correlative approach will also be discussed.
In the course of a thorough investigation of the performance-structure-chemistry interdependency at silicon grain boundaries, we successfully developed a method to systematically correlate aberration-corrected scanning transmission electron microscopy and atom probe tomography. The correlative approach is conducted on individual APT and TEM specimens, with the option to perform both investigations on the same specimen in the future. In the present case of a Σ9 grain boundary, joint mapping of the atomistic details of the grain boundary topology, in conjunction with chemical decoration, enables a deeper understanding of the segregation of impurities observed at such grain boundaries.
Due to the intrinsic evaporation properties of the material studied, insufficient mass-resolving power and lack of knowledge of the kinetic energy of incident ions, peaks in the atom probe mass-to-charge spectrum can overlap and result in incorrect composition measurements. Contributions to these peak overlaps can be deconvoluted globally, by simply examining adjacent peaks combined with knowledge of natural isotopic abundances. However, this strategy does not account for the fact that the relative contributions to this convoluted signal can often vary significantly in different regions of the analysis volume; e.g., across interfaces and within clusters. Some progress has been made with spatially localized deconvolution in cases where the discrete microstructural regions can be easily identified within the reconstruction, but this means no further point cloud analyses are possible. Hence, we present an ion-by-ion methodology where the identity of each ion, normally obscured by peak overlap, is resolved by examining the isotopic abundance of their immediate surroundings. The resulting peak-deconvoluted data are a point cloud and can be analyzed with any existing tools. We present two detailed case studies and discussion of the limitations of this new technique.
In this work, we report on the atom probe tomography analysis of two metallic hydrides formed by pressurized charging using an ex situ hydrogen charging cell, in the pressure range of 200–500 kPa (2–5 bar). Specifically we report on the deuterium charging of Pd/Rh and V systems. Using this ex situ system, we demonstrate the successful loading and subsequent atom probe analysis of deuterium within a Pd/Rh alloy, and demonstrate that deuterium is likely present within the oxide–metal interface of a native oxide formed on vanadium. Through these experiments, we demonstrate the feasibility of ex situ hydrogen analysis for hydrides via atom probe tomography, and thus a practical route to three-dimensional imaging of hydrogen in hydrides at the atomic scale.
Atom probe tomography is routinely used for the characterization of materials microstructures, usually assuming that the microstructure is unaltered by the analysis. When analyzing ionic conductors, however, gradients in the chemical potential and the electric field penetrating dielectric atom probe specimens can cause significant ionic mobility. Although ionic mobility is undesirable when aiming for materials characterization, it offers a strategy to manipulate materials directly in situ in the atom probe. Here, we present experimental results on the analysis of the ionic conductor lithium-manganese-oxide with different atom probe techniques. We demonstrate that, at a temperature of 30 K, characterization of the materials microstructure is possible without measurable Li mobility. Also, we show that at 298 K the material can be deintercalated, in situ in the atom probe, without changing the manganese-oxide host structure. Combining in situ atom probe deintercalation and subsequent conventional characterization, we demonstrate a new methodological approach to study ionic conductors even in early stages of deintercalation.
Six precursors were evaluated for use as in situ electron beam-induced deposition capping layers in the preparation of atom probe tomography specimens with a focus on near-surface features where some of the deposition is retained at the specimen apex. Specimens were prepared by deposition of each precursor onto silicon posts and shaped into sub-70-nm radii needles using a focused ion beam. The utility of the depositions was assessed using several criteria including composition and uniformity, evaporation behavior and evaporation fields, and depth of Ga+ ion penetration. Atom probe analyses through depositions of methyl cyclopentadienyl platinum trimethyl, palladium hexafluoroacetylacetonate, and dimethyl-gold-acetylacetonate [Me2Au(acac)] were all found to result in tip fracture at voltages exceeding 3 kV. Examination of the deposition using Me2Au(acac) plus flowing O2 was inconclusive due to evaporation of surface silicon from below the deposition under all analysis conditions. Dicobalt octacarbonyl [Co2(CO)8] and diiron nonacarbonyl [Fe2(CO)9] depositions were found to be effective as in situ capping materials for the silicon specimens. Their very different evaporation fields [36 V/nm for Co2(CO)8 and 21 V/nm for Fe2(CO)9] provide options for achieving reasonably close matching of the evaporation field between the capping material and many materials of interest.
Two challenges exist in laser-assisted atom probe tomography (APT). First, a drastic decline in mass-resolving power is caused, not only by laser-induced thermal effects on the APT tips of bulk oxide materials, but also the associated asymmetric evaporation behavior; second, the field evaporation mechanisms of bulk oxide tips under laser illumination are still unclear due to the complex relations between laser pulse and oxide materials. In this study, both phenomena were investigated by depositing Ni- and Co-capping layers onto the bulk LaAlO3 tips, and using stepwise APT analysis with transmission electron microscopy (TEM) observation of the tip shapes. By employing the metallic capping, the heating at the surface of the oxide tips during APT analysis became more symmetrical, thereby enabling a high mass-resolving power in the mass spectrum. In addition, the stepwise microscopy technique visualized tip shape evolution during APT analysis, thereby accounting for evaporation sequences at the tip surface. The combination of “capping” and “stepwise APT with TEM,” is applicable to any nonconductors; it provides a direct observation of tip shape evolution, allows determination of the field evaporation strength of oxides, and facilitates understanding of the effects of ultrafast laser illumination on an oxide tip.
In 1994, a new type of atom probe instrument, named the scanning atom probe (SAP), was proposed. The unique feature of the SAP is the introduction of a small extraction electrode, which scans over a specimen surface and confines the high field, required for field evaporation of surface atoms in a small space, between the specimen and the electrode. Thus, the SAP does not require a sharp specimen tip. This indicates that the SAP can mass analyze the specimens which are difficult to form in a sharp tip, such as organic materials and biomolecules. Clean single wall carbon nanotubes (CNT), made by high-pressure carbon monoxide process are found to be the best substrates for biomolecules. Various amino acids and dipeptide biomolecules were successfully mass analyzed, revealing characteristic clusters formed by strongly bound atoms in the specimens. The mass analysis indicates that SAP analysis of biomolecules is not only qualitative, but also quantitative.
The formation of copper-rich precipitates of 17-4 precipitate hardened stainless steel has been investigated, after tempering at 350–570°C for 4 h, by atom probe tomography (APT). The results reveal that the clusters, enriched only with Cu, were observed after tempering at 420°C. Segregation of Ni, Mn to the Cu-rich clusters took place at 450°C, contributing to the increased hardening. After tempering at 510°C, Ni and Mn were rejected from Cu-rich precipitates and accumulated at the precipitate/matrix interfaces. Al and Si were present and uniformly distributed in the precipitates that were <1.5 nm in radius, but Ni, Mn, Al, and Si were enriched at the interfaces of larger precipitates/matrix. The proxigram profiles of the Cu-rich precipitates formed at 570°C indicated that Ni, Mn, Al, and Si segregated to the precipitate/matrix interfaces to form a Ni(Fe, Mn, Si, Al) shell, which significantly reduced the interfacial energy as the precipitates grew into an elongated shape. In addition, the number density of Cu-rich precipitates was increased with the temperature elevated from 350 up to 450°C and subsequently decreased at higher temperatures. Also, the composition of the matrix and the precipitates were measured and found to vary with temperature.
The microstructural evolution of precipitation in two model alloys, Fe–NiAl and Fe–NiAl–Cu, was investigated during aging at 500°C for different times using atom probe tomography (APT). The APT results reveal that the addition of Cu effectively increases the number density of NiAl precipitates. This is attributed to Cu promoting the nucleation of NiAl particles by increasing the chemical driving force and decreasing the interfacial energy. The NiAl precipitates of the Fe–NiAl–Cu alloy grow and coarsen at a slower rate than that of the Fe–NiAl alloy, mainly due to the slower diffusion rate of the Cu atoms. The mechanical properties of the two alloys were characterized by Vickers hardness and tension tests. It was found that the addition of Cu results in the formation of core–shell precipitates with a Cu-rich core and a NiAl shell, leading to a dramatic improvement of peak hardness and strength. The effect of Cu on precipitation strengthening is discussed in terms of chemical strength and coherency strength.
In creep resistant Z-phase strengthened 12% Cr steels, MX (M=Nb, Ta, or V, and X=C and/or N) to Z-phase (CrMN, M=Ta, Nb, or V) transformation plays an important role in achieving a fine distribution of Z-phase precipitates for creep strengthening. Atom probe tomography was employed to investigate the phase transformation in a Nb-based Z-phase strengthened trial steel. Using iso-concentration surfaces with different concentration values, and subtracting the matrix contribution enabled us to reveal the core-shell structure of the transient precipitates between MX and Z-phase. It was shown that Z-phase forms by diffusion of Cr into NbN upon ageing, and Z-phase has a composition corresponding to Cr1+xNb1−xN with x=0.08.