The effect of lens aberrations on the spatial resolution of an energy-filtered TEM image

High-Energy Electron Scattering in Thick Samples Evaluated by Bright-Field Transmission Electron Microscopy, Energy-Filtering Transmission Electron Microscopy, and Electron Tomography

Energy-filtering transmission electron microscopy (TEM) and bright-field TEM can be used to extract local sample thickness $t$ and to generate two-dimensional sample thickness maps. Electron tomography can be used to accurately verify the local $t$. The relations of log-ratio of zero-loss filtered energy-filtering TEM beam intensity ($I_{{\rm ZLP}}$) and unfiltered beam intensity ($I_{\rm u}$) versus sample thickness $t$ were measured for five values of collection angle in a microscope equipped with an energy filter. Furthermore, log-ratio of the incident (primary) beam intensity ($I_{\rm p}$) and the transmitted beam $I_{{\rm tr}}$ versus $t$ in bright-field TEM was measured utilizing a camera before the energy filter. The measurements were performed on a multilayer sample containing eight materials and thickness $t$ up to 800 nm. Local thickness $t$ was verified by electron tomography. The following results are reported:• The maximum thickness $t_{{\rm max}}$ yielding a linear relation of log-ratio, $\ln ( {I_{\rm u}}/{I_{{\rm ZLP}}})$ and $\ln ( {I_{\rm p}}/{I_{{\rm tr}}} )$, versus $t$.• Inelastic mean free path ($\lambda _{{\rm in}}$) for five values of collection angle.• Total mean free path ($\lambda _{{\rm total}}$) of electrons excluded by an angle-limiting aperture.• $\lambda _{{\rm in}}$ and $\lambda _{{\rm total}}$ are evaluated for the eight materials with atomic number from $\approx$10 to 79.The results can be utilized as a guide for upper limit of $t$ evaluation in energy-filtering TEM and bright-field TEM and for optimizing electron tomography experiments.

OUP accepted manuscript

We first review the significance of resolution and contrast in electron microscopy and the effect of the electron optics on these two quantities. We then outline the physics of the generation of secondary electrons (SEs) and their transport and emission from the surface of a specimen. Contrast and resolution are discussed for different kinds of SE imaging in scanning electron microscope (SEM) and scanning-transmission microscope instruments, with some emphasis on the observation of individual atoms and atomic columns in a thin specimen. The possibility of achieving atomic resolution from a bulk specimen at SEM energies is also considered.

Limits to the spatial, energy and momentum resolution of electron energy-loss spectroscopy

We discuss various factors that determine the performance of electron energy-loss spectroscopy (EELS) and energy-filtered (EFTEM) imaging in a transmission electron microscope. Some of these factors are instrumental and have undergone substantial improvement in recent years, including the development of electron monochromators and aberration correctors. Others, such as radiation damage, delocalization of inelastic scattering and beam broadening in the specimen, derive from basic physics and are likely to remain as limitations. To aid the experimentalist, analytical expressions are given for beam broadening, delocalization length, energy broadening due to core-hole and excited-electron lifetimes, and for the momentum resolution in angle-resolved EELS.

Valence state map of iron oxide thin film obtained from electron spectroscopy imaging series

This paper demonstrates the applicability of electron-spectroscopic imaging (ESI) for valence-state mapping of the iron oxide system. We have previously developed a set of signal-processing methods for an ESI series, to allow mapping of sp(2)/sp(3) ratio, dielectric function and energy bandgap. In this study, these methods are applied to generate a valence-state map of an iron oxide thin film (Fe/alpha-Fe(2)O(3)). Two problems, data undersampling and a convolution effect associated with extraction of the image-spectrum from the core loss image series, were overcome by using cubic-polynomial interpolation and maximum-entropy deconvolution. As a result, the reconstructed image-spectrum obtained from the ESI series images has a quality as good as that of conventional electron energy-loss spectra. The L(3)/L(2) ratio of the reconstructed ESI spectrum is determined to be 3.30+/-0.30 and 5.0+/-0.30 for Fe and alpha-Fe(2)O(3), respectively. Our L(3)/L(2) ratio mapping shows an accurate correspondence across the Cu/Fe/alpha-Fe(2)O(3) region. The effect of delocalization and chromatic aberration on the ESI resolution is discussed and estimated to be about 2 nm for the case of L(3)/L(2) ratio mapping.

EFTEM assistant: A tool to understand the limitations of EFTEM

The first version of a free tool for Gatan's Digital Micrographtrade mark is presented which aims to aid the energy-filtered TEM (EFTEM) community by predicting and correcting the most common sources of degradation. The software allows selection of either Krivanek's or Egerton's approach to account for the spatial resolution degradation caused by the electron optical aberrations. The effects of aberrations and signal 'delocalization' are combined to simulate the blurring caused in EFTEM elemental maps. Two microstructural features with ideal geometry are used to illustrate use of the software: spherical particles and parallel sided interfaces. The software also allows the simulation of the effects of the noise and drift in the final elemental map, independently or in combination. It can be easily demonstrated that when the dimensions of the feature of interest are comparable in scale to the image degradation factors, the effects of the latter should not be neglected. More importantly, the software can deconvolute the effects of the degradation factors, revealing the true dimensions and signal intensity of the feature of interest.

Beam-induced damage to thin specimens in an intense electron probe

We have investigated the changes produced in single-element and two-layer transmission electron microscope (TEM) specimens irradiated by an intense nanometer-sized electron probe, such as that produced in a field-emission or aberration-corrected TEM. These changes include hole formation and the accumulation of material within the irradiated area. The results are discussed in terms of mechanisms, including electron-beam sputtering and surface diffusion. Strategies for minimizing the effect of the beam are considered.

Determination of the Fe content of embedded Cu-rich particles in ferritic alloys using energy-filtered TEM

A novel technique for the quantification of the iron content of copper precipitates in ferritic steels is presented. Energy-filtered (EF) imaging has been used to extract elemental maps with high spatial resolution. These maps contain enough information to attempt the quantification of the signal produced by the precipitates when either a line profile is measured across them or the whole image signal is integrated. Assumptions such as sphericity of the precipitates and composition variations are discussed. Special attention to the assessment of drift on the information extracted from EF images has been taken. Minimum detectability and optimum acquisition conditions are discussed.

Structural characterization of TiN/NbN multilayers: X-ray diffraction, energy-filtered TEM and Fresnel contrast techniques compared

Two TiN/NbN multilayers with wavelength 13.6 and 6.15 nm have been characterized by X-ray diffraction (XRD), Fresnel contrast analysis (FCA) and energy-filtered transmission electron microscopy (EFTEM). Good agreement between the composition profile obtained by FCA and EFTEM is obtained if the lower resolution of the EFTEM images is taken into account. The relative advantages and disadvantages of the techniques are discussed. Used together the two TEM techniques provide a quantitative characterization that is consistent with, and for some parameters provides more precise values than, that from XRD. The analysis shows that the multilayers have narrow interfaces (< 1 nm) and a composition amplitude close to 95% for the longer wavelength.

Electron spectroscopy imaging to study ELNES at a nanoscale

Series of energy-filtered TEM images have been acquired with very narrow energy slit using a post-column energy filter. This allowed us to reconstruct spectra with an energy resolution estimated to 2 eV, and a spatial resolution in the order of 0.5 nm. In that way, fine structures of the N-K edge in AlN/GaN heterostructures have been investigated and compared to EELS spectra. The fine structure in the two nitrides is very sensitive to the local environment. Very good agreement between ESI and EELS spectra was found. Moreover, this technique allowed analysis of the AlN/GaN interface at a nanoscale. The second example is an application of the technique to construct bonding maps. In this case, maps differentiating AlN nanoprecipitates with either the cubic or the hexagonal phase were created.

A modeling and convolution method to measure compositional variations in strained alloy quantum dots

We have developed a method to quantitatively measure the absolute composition of nanometer sized capped quantum dots in semiconductor alloys. The method uses spatially resolved electron energy-loss spectroscopy in a scanning transmission electron microscope to measure compositional profiles across the center of the quantum dot and the adjacent nanometer wide wetting layer. The measurements from the wetting layer are used to derive a spatial broadening function which includes the effects of probe size, instabilities and beam spreading in the sample. This broadening function is employed to simulate compositional profiles from the quantum dots. Information on the dimensions of dots is extracted from annular dark-field images. The method is applied to In(y)Ga(1-y)As (y = 0.5) quantum dots grown on a GaAs substrate. In this system, a simple truncated cone model is found to give an adequate description of the compositional variations across the dot. We find a substantial enrichment in In at the center of the dots, in agreement with theoretical predictions.

Molecular immunolabeling with recombinant single-chain variable fragment (scFv) antibodies designed with metal-binding domains

To study the molecular structure and function of gene products in situ, we developed a molecular immunolabeling technology. Starting with cDNA from hybridomas producing monoclonal antibodies against biotin, catalase, and superoxide dismutase, we bioengineered recombinant single-chain variable fragment antibodies (scFv) and their derivatives containing metal-binding domains (scFv:MBD). As tested with surface plasmon resonance and enzyme-linked immunosorbent assay, affinity binding constants of the scFv (5.21 x 10(6) M(-1)) and scFv:MBD (4.17 x 10(6) M(-1)) were close to those of Fab proteolytic fragments (9.78 x 10(6) M(-1)) derived from the parental IgG antibodies. After saturation of MBD with nickel or cobalt, scFv:MBD was imaged with electron spectroscopic imaging at each element's specific energy loss, thus generating the element's map. Immunolabeling with scFv:MBD resulted in a significant improvement of the labeling fidelity over that obtained with Fab or IgG derivatives, as it produced a much heavier specific labeling and label-free background. As determined with radioimmunoassay, labeling effectiveness with scFv:MBD was nearly the same as with scFv, but much higher than with scFv conjugated to colloidal gold, Nanogold, or horseradish peroxidase. This technology opens possibilities for simultaneous imaging of multiple molecules labeled with scFv:MBD at the molecular resolution within the same sample with electron spectroscopic imaging. Moreover, the same scFv:MBD can also be imaged with fluorescence resonance energy transfer and lifetime imaging as well as positron emission tomography and magnetic resonance imaging. Therefore, this technology may serve as an integrative factor in life science endeavors.

Energy-filtered transmission electron microscopy of multilayers in semiconductors

Quantitative analysis of novel semiconductors with wide or ultrathin multilayers of atomic dimensions is very important in order to control electronic and optical properties, but rather difficult due to the limited resolution in most techniques. In this paper we attempt to assess how effectively the total As dopant concentration in ultrathin As doped layers in InP and the Ti atomic fraction in a TixAl1-xN multilayer can be analysed quantitatively using energy-filtered imaging. These two materials have characteristic edges located at widely different energy losses, with the L edge of As being above 1000 eV, while that of Ti is around 450 eV. We have quantified the As concentration using the three-window technique and theoretical cross-sections and we find that the resolution limit is dominated by the signal-to-noise ratio in this delta-doped specimen. However, the accuracy of the Ti atomic fraction in TixAl1-xN can be as good as 10 at% for specimens of uniform thickness made by focused ion beam milling. We will compare our results with measurements of the composition made using Fresnel contrast, high resolution imaging and high angle annular dark field techniques.