Search

Summary Electron microscopy is an essential tool for the evaluation of microstructure and properties of the catalyst layer (CL) of proton exchange membrane fuel cells (PEMFCs). However, electron microscopy has one unavoidable drawback, which is radiation damage. Samples suffer temporary or permanent change of the surface or bulk structure under radiation damage, which can cause ambiguity in the characterization of the sample. To better understand the mechanism of radiation damage of CL samples and to be able to separate the morphological features intrinsic to the material from the consequences of electron radiation damage, a series of experiments based on high-angle annular dark-field–scanning transmission scanning microscope (HAADF-STEM), energy filtering transmission scanning microscope (EFTEM), and electron energy loss spectrum (EELS) are conducted. It is observed that for thin samples (0.3–1 times λ), increasing the incident beam energy can mitigate the radiation damage. Platinum nanoparticles in the CL sample facilitate the radiation damage. The radiation damage of the catalyst sample starts from the interface of Pt/C or defective thin edge and primarily occurs in the form of mass loss accompanied by atomic displacement and edge curl. These results provide important insights on the mechanism of CL radiation damage. Possible strategies of mitigating the radiation damage are provided. SCANNING 36:338–346, 2014. © 2013 Wiley Periodicals, Inc.

Summary Generally, in scanning electron microscopy (SEM) imaging, it is desirable that a high-resolution image be composed mainly of those secondary electrons (SEs) generated by the primary electron beam, denoted SEI. However, in conventional SEM imaging, other, often unwanted, signal components consisting of backscattered electrons (BSEs), and their associated SEs, denoted SEII, are present; these signal components contribute a random background signal that degrades contrast, and therefore signal-to-noise ratio and resolution. Ideally, the highest resolution SEM image would consist only of the SEI component. In SEMs that use conventional pinhole lenses and their associated Everhart–Thornley detectors, the image is composed of several components, including SEI, SEII, and some BSE, depending on the geometry of the detector. Modern snorkel lens systems eliminate the BSEs, but not the SEIIs. We present a microfabricated diaphragm for minimizing the unwanted SEII signal components. We present evidence of improved imaging using a microlithographically generated pattern of Au, about 500 nm thick, that blocks most of the undesired signal components, leaving an image composed mostly of SEIs. We refer to this structure as a “spatial backscatter diaphragm.” SCANNING 35:1-6, 2013. © 2012 Wiley Periodicals, Inc.

A formula is presented for computing the polar angle of collisions from partial elastic Mott cross-sections. It will be useful in Monte Carlo simulations of electrons' trajectories in solids and is based on the Newbury and Myklebust (1981) expression. It also is valid for all energy ranging from 0.1 to 30 keV. Also given is the tabulation of the constants of this formula to compute polar angles of collisions for the first 94 elements of the periodic table. The constants of this formula have been obtained by the numerical integration of the values of partial Mott cross-sections computed by Czyzewski et al. (1990) using Thomas-Fermi-Dirac atomic potentials for atomic numbers >54 and Hartree-fock potential for atomic numbers

Electron beam-induced deposition (EBID) provides a simple way to fabricate submicron- or nanometer-scale structures from various elements in a scanning electron microscope (SEM). The growth rate and shape of the deposits are influenced by many factors. We have studied the growth rate and morphology of EBID-deposited nanostructures as a function of the tungsten hexafluoride (WF6) and tetraethylorthosilicate (TEOS) precursor gas pressure and growth time, and we have used Monte Carlo simulations to model the growth of tungsten and silicon oxide to elucidate the mechanisms involved in the EBID growth. The lateral radius of the deposit decreases with increasing pressure because of the enhanced vertical growth rate which limits competing lateral broadening produced by secondary and forward-scattered electrons. The morphology difference between the conical SiO(x) and the cylindrical W nanopillars is related to the difference in interaction volume between the two materials. A key parameter is the residence time of the precursor gas molecules. This is an exponential function of the surface temperature; it changes during nanopillar growth and is a function of the nanopillar material and the beam conditions.

A Monte Carlo simulation of the contrast variation across phase boundaries in backscattered electron images of multiple phase composition systems has been used to develop a model for prediction of the position of the interface based solely on a knowledge of the signal intensity levels on either side of the interface. A wide range of average atomic numbers of the phases on either side of the boundary have been investigated at incident beam energies of 5, 10, 15, and 25 kV and a least-squares minimisation procedure used to optimise the parameters of a generalised model.

A field-emission scanning electron microscope (FESEM) equipped with the standard secondary electron (SE) detector was used to image thin (70-90 nm) and thick (1-3 microns) sections of biological materials that were chemically fixed, dehydrated, and embedded in resin. The preparation procedures, as well as subsequent staining of the sections, were identical to those commonly used to prepare thin sections of biological material for observation with the transmission electron microscope (TEM). The results suggested that the heavy metals, namely, osmium, uranium, and lead, that were used for postfixation and staining of the tissue provided an adequate SE signal that enabled imaging of the cells and organelles present in the sections. The FESEM was also used to image sections of tissues that were selectively stained using cytochemical and immunocytochemical techniques. Furthermore, thick sections could also be imaged in the SE mode. Stereo pairs of thick sections were easily recorded and provided images that approached those normally associated with high-voltage TEM.

Empirical forms for electron/atom scattering cross sections predict backscattering factors that compare well with those calculated using tabulated Mott data from 0.1 to 30 keV. The form of the empirical total cross section is similar to the screened Rutherford cross section. The fit to the tabulated differential Mott cross sections is decomposed into two parts, one part being of the same mathematical form as the screened Rutherford cross section (σR), and the second part being an isotropic distribution (σI). The ratio of the total cross sections (σR/σI) between the screened Rutherford part of the differential scattering cross section and the isotropic part of the distribution is fitted to give the same ratio of forward to backscattered currents as the tabulated Mott differential cross sections. The three equations, one for the total elastic cross section and two describing the differential cross section—one for the Rutherford screening parameter and one for the ratio σR/σI—give backscattering results covering all the major trends with energy and atomic number compared with the backscattering coefficients calculated using tabulated Mott cross sections. However, agreement with experiment is poor for some well-researched examples such as Au. Monte Carlo calculations using the empirical cross sections show that surface effects may be critical in interpreting experimental results.

Monte Carlo modeling of electron-solid interactions requires a detailed and accurate supply of experimental data on which to base its physics and against which to test its predictions. To meet this need, a collection of data—comprising measurements of secondary and backscattered electron yields, electron stopping powers, and x-ray ionization cross sections, as a function of energy—has been assembled from published sources. The quality and quantity of the compilation varies widely, with little or no data being available for the majority of elements in the periodic table, while results for complex materials of current technologic interest are also almost nonexistent. To meet the needs of Monte Carlo simulation in areas such as dimensional metrology or microanalysis, a program of systematic measurements is required.

Experimental nanotips have shown significant improvement in the resolution performance of a cold field emission scanning electron microscope (SEM). Nanotip electron sources are very sharp electron emitter tips used as a replacement for the conventional tungsten field emission (FE) electron sources. Nanotips offer higher brightness and smaller electron source size. An electron microscope equipped with a nanotip electron gun can provide images with higher spatial resolution and with better signal-to-noise ratio. This could present a considerable advantage over the current SEM electron gun technology if the tips are sufficiently long-lasting and stable for practical use. In this study, an older field-emission critical dimension (CD) SEM was used as an experimental test platform. Substitution of tungsten nanotips for the regular cathodes required modification of the electron gun circuitry and preparation of nanotips that properly fit the electron gun assembly. In addition, this work contains the results of the modeling and theoretical calculation of the electron gun performance for regular and nanotips, the preparation of the SEM including the design and assembly of a measuring system for essential instrument parameters, design and modification of the electron gun control electronics, development of a procedure for tip exchange, and tests of regular emitter, sharp emitter and nanotips. Nanotip fabrication and characterization procedures were also developed. Using a "sharp" tip as an intermediate to the nanotip clearly demonstrated an improvement in the performance of the test SEM. This and the results of the theoretical assessment gave support for the installation of the nanotips as the next step and pointed to potentially even better performance. Images taken with experimental nanotips showed a minimum two-fold improvement in resolution performance than the specification of the test SEM. The stability of the nanotip electron gun was excellent; the tip stayed useful for high-resolution imaging for several hours during many days of tests. The tip lifetime was found to be several months in light use. This paper summarizes the current state of the work and points to future possibilities that will open when electron guns can be designed to take full advantage of the nanotip electron emitters.

Charge accumulation in insulating or semiconducting samples due to electron beam irradiation is one of the key problems in electron microscopy. One of the most promising techniques for reducing the severity of such charging is to surround the sample with a low-pressure atmosphere of a gas. The charging behavior of a number of materials, surrounded by a variety of gases, has been determined to identify the important factors which control charging under these conditions. The magnitude of the surface potential was deduced from an analysis of x-ray spectra from the surface. The relationship between surface charge, gas pressure, and gas type are measured, and the charging reduction efficiency (CRE) is compared.

Charging is one of the most important problems encountered in scanning electron microscopy and as a result this phenomenon it has received a lot of both theoretical and experimental attention. Despite this, many questions remain about the nature and behavior of charging because of the limitations of the experimental techniques available to study it. For example, although it is now straightforward to determine in situ the surface potential of a sample that is charging during irradiation, it is difficult to measure the lateral extent of the charging, or its persistence once the incident beam is switched off. We describe here a simple technique which provides a rapid way of visualizing the temporal and spatial behavior of charging phenomena.

The relationship between convolution/correlation operation and the data acquisition process of the scanning microscope and spectrometer families is analyzed. It is shown that a coordinate or event sensitive detector, and the intrinsic or extrinsic property of the specimen response, are two important factors in distinguishing the data acquisition mode of such systems. Four types of convolution- and correlation-based modes are extracted to illuminate the physical characteristics of scanning imaging and analysis systems by focusing on the probe, specimen, detector, and their relationships. Criteria for identifying these modes are explored. In addition, the physical meanings of general existing coefficients between the independent variables of convolution and correlation are investigated.

Experimental measurements are presented of the total scattering cross-section of four gases: He, Air, CH 4 , and Argon as a function of electron beam energy. The method used was that of Gauvin. A comparison of theoretical estimates of the total scattering cross-section with the experimental data indicates that the model used gives generally reliable results. The data confirm that the gases tend to be molecular rather than atomic in nature in this pressure range (1-300 Pascal). The factors affecting the accuracy of these results are examined.

The absolute efficiency of generation of a selection of K, L, and M- x-ray lines has been measured as a function of the incident electron beam energy. At an overvoltage U=2 this efficiency falls within the range 1E-4 to 1E-7, with K-lines being highest and L-Lines usually being the lowest. It is shown that for all three families of lines the efficiency has a functional variation which has the form A. (U- 1)n, as first suggested by Compton and Allison, where A and n are constants. Values of A and n for the K, L, and M shells are tabulated. The smoothly varying behavior of the efficiency makes it well suited for analytical use and spectrum simulation purposes.

By applying the basic principles of metrology we discuss how to define the standards that any experimental method to measure resolution has to obey. Our results clearly indicate the need to apply a calibration procedure when designing algorithms to estimate resolution to satisfy accuracy requirements. Similarly, the precision of an algorithm has to be clearly specified. We compare here the performances of a variety of commonly used implementations of published methods, with that of an algorithm based on an approach known to be reliable. Our results confirm that when an algorithm is designed with the clear intent of satisfying metrology requirements, it demonstrates excellent accuracy, precision, and lack of sensitivity to the noise level, as is desirable. As a consequence, the algorithm will have the ability to measure accurately the point spread function convoluted in the image, thus paving the way for quantitative deconvolution techniques.

A simple method is described to determine the effective gas path length when incident electrons scatter in the gas above the specimen. This method is based on the measurement of a characteristic x-ray line emitted from a region close to the incident beam. From various experimental measurements performed on various microscopes, it is shown that the effective gas path length may increase with the chamber pressure and that it is also often dependent of the type of x-ray bullet.

Secondary electron imaging is not possible in the variable pressure scanning electron microscope because the mean free path of the secondaries in the gas is too short to permit them to reach the detector. This paper therefore investigates an alternative strategy for producing an image containing significant amounts of secondary electron contrast. This involves modifying the microscope by the addition of a biased electrode above the sample and then collecting a specimen current signal. This system, originally described by Farley and Shah (1988), is found to produce true secondary electron detail over a wide range of conditions.

This paper is a description of the stopping power routine utilized in the CASINO program that is based on the experimental measurement of the energy loss function (ELF). In addition, we present an ANSI C standard program that can be used to generate the data needed for the stopping power routine. Both optical and energy loss spectrum (ELS) measurements of the ELF can be used as input to compute the stopping power. For ELS, only the single scattering spectrum is needed. Hence, measurement of the stopping power for a given element or compound of interest can easily be performed and used in the CASINO program. The resulting effect of using these stopping powers in Monte Carlo simulations is generally to increase the backscattering coefficient. Except for carbon, the change of stopping power for pure elements so far compiled is relatively small. In some compounds (i.e., Al2O3 and ZnSe), the discrepancy with the Joy and Luo (1989) expression is significant.

Electron microscopy is an essential tool for the evaluation of microstructure and properties of the catalyst layer (CL) of proton exchange membrane fuel cells (PEMFCs). However, electron microscopy has one unavoidable drawback, which is radiation damage. Samples suffer temporary or permanent change of the surface or bulk structure under radiation damage, which can cause ambiguity in the characterization of the sample. To better understand the mechanism of radiation damage of CL samples and to be able to separate the morphological features intrinsic to the material from the consequences of electron radiation damage, a series of experiments based on high-angle annular dark-field-scanning transmission scanning microscope (HAADF-STEM), energy filtering transmission scanning microscope (EFTEM), and electron energy loss spectrum (EELS) are conducted. It is observed that for thin samples (0.3-1 times λ), increasing the incident beam energy can mitigate the radiation damage. Platinum nanoparticles in the CL sample facilitate the radiation damage. The radiation damage of the catalyst sample starts from the interface of Pt/C or defective thin edge and primarily occurs in the form of mass loss accompanied by atomic displacement and edge curl. These results provide important insights on the mechanism of CL radiation damage. Possible strategies of mitigating the radiation damage are provided.

Generally, in scanning electron microscopy (SEM) imaging, it is desirable that a high-resolution image be composed mainly of those secondary electrons (SEs) generated by the primary electron beam, denoted SE(I) . However, in conventional SEM imaging, other, often unwanted, signal components consisting of backscattered electrons (BSEs), and their associated SEs, denoted SE(II) , are present; these signal components contribute a random background signal that degrades contrast, and therefore signal-to-noise ratio and resolution. Ideally, the highest resolution SEM image would consist only of the SE(I) component. In SEMs that use conventional pinhole lenses and their associated Everhart-Thornley detectors, the image is composed of several components, including SE(I) , SE(II) , and some BSE, depending on the geometry of the detector. Modern snorkel lens systems eliminate the BSEs, but not the SE(II) s. We present a microfabricated diaphragm for minimizing the unwanted SE(II) signal components. We present evidence of improved imaging using a microlithographically generated pattern of Au, about 500 nm thick, that blocks most of the undesired signal components, leaving an image composed mostly of SE(I) s. We refer to this structure as a "spatial backscatter diaphragm."

The diffusion matrix model proposed by Desai and Reimer (1990) has been investigated. After some modifications it appeared to be a powerful tool for rapid determination of the SE and BSE signals from two-dimensional surface topographies. Results for low atomic number specimens show good agreement between experimental and theoretical data.

Electron-beam-induced deposition (EBID) and etching (EBIE) provides a simple way to fabricate or etch submicron or nanoscale structures of various materials in a direct-write (i.e.nonlithographic) fashion. The growth rate or the etch rate are influenced by many factors such as beam energy, beam current, temperature of the substrate material, pressure of the chamber, and geometry of the gas injector etc. The mechanism of EBID and EBIE involves the interaction of the incident electron beam or emitted electron from the target material. The role of these electrons is still not completely understood although the contribution of low energy secondary electrons (SE) has been assumed to be the dominant contributor of EBID and EBIE based on its overlap with the dissociation cross section. We have studied the growth and etching phenomenon under various biasing conditions to investigate how low voltage biasing of the substrate affects secondary electron trajectories and subsequently modifies electron-beam-induced deposition and etching.SCANNING 29: 000–000, 2007. © 2007 Wiley Periodicals, Inc.

Although electron beams with energies of a few keV can excite fluorescent X-ray production from solids, ion beams of comparable energy cannot do so. The reason for this situation is that it is the velocity of the incident particle, rather than its energy, which determines whether an ionization event can be generated. SCANNING 29: 1–4, 2007. Copyright © 2007 Wiley Periodicals, Inc.

In the variable pressure scanning electron microscope (VP-SEM) the incident electrons pass through a gaseous environment and the beam is scattered by these interactions. We show here that the experimental intensity profile of the scattered beam can be described as Gaussian in form to a high level of accuracy. This provides a rapid means of accounting for the effects of beam scatter in imaging and microanalysis because the standard deviation of the Gaussian is a simple function of parameters such as working distance, beam energy, gas type and pressure.

A simple method is described to determine the effective gas path length when incident electrons scatter in the gas above the specimen. This method is based on the measurement of a characteristic x-ray line emitted from a region close to the incident beam. From various experimental measurements performed on various microscopes, it is shown that the effective gas path length may increase with the chamber pressure and that it is also often dependent of the type of x-ray bullet.

A recent study of secondary electron (SE) spectra in an Auger spectrometer demonstrated unique features indicative of the chemical nature of the tested material. The scanning electron microscope (SEM) naturally generates SEs; therefore, in this paper, we combine the concept of using differential voltage contrast (DVC) with SE spectroscopy to identify the chemical nature of a material. It is demonstrated that this method reveals the uniqueness of electron energy distribution in the conduction band of a solid or, what is the same, the uniqueness of a build-up of the outer electron shell system, and avoids errors due to the changes in the angular distribution or yield of the SE in the SEM. A theory of this new microanalytical method is developed. The experimental limitations of the SEM for this type of study are examined as well.

The charging of polymeric resist materials during electron beam irradiation leads to significant problems during imaging and lithography processes. Charging occurs because of charge deposition in the polymer and charge generation/trapping due to formation of electron-hole pairs in the dielectric. The presence of such charge also results in the phenomena of electron beam-induced conductivity (EBIC). Electron beam-induced conductivity data have been obtained for three commercial e-beam resists under a variety of dose rate and temperature conditions. From the observed values of induced conductivity under varying conditions significant information about the generation of electron-hole pair and the transport of charge in the resist can be obtained. Three electron beam resists, EBR900, ZEP7000, and PBS are examined by an external bias method. The difference in resist chemistry is considered to play the role in the initial state EBIC behaviors among three resists even though the way that it affects the behaviors is not clear. A comparison of the power consumption comparison is proposed as a measure to give a preliminary estimate of the carrier concentration and carrier drift velocity differences among the resists. A simple single trap model with constant activation energy is proposed and provides good agreement with experiment.

A simple digital method of measuring the performance of detectors in the scanning electron microscope (SEM) is described. The value derived is absolute and can therefore be used to compare both different detectors on the same instrument as well as different detectors on different instruments. The technique can be applied to secondary electron, backscattered electron, and energy-dispersive x-ray detectors. Examples are given of measurements made on a variety of commercial detectors installed on a number of current SEMs.

This note generalizes the Archard model for electron backscattering. It is shown that the energy-dependent variation of electron backscattering can be predicted.

Using a new method for improving the signal-to-noise of electron channeling pattern (ECP) contrast, preliminary quantitative measurements of the linewidths of ECP lines in silicon have been performed. The signal-to-noise was improved by utilizing a boxcar integrating circuit whose gating logic was referenced to the SEM horizontal scan signal. The measurements were done on the (220) series of reflections from 18 and 25 keV incident electrons by monitoring the specimen current as the SEM was scanning in the line mode. The measured linewidths are consistent with those predicted by the two beam theory of ECP intensity when the affects of the incident beam divergence, α, are accounted for. The appearance of ECP lines from very high order reflections is an indication that the limit of the reciprocal lattice vector, g, of a line that can be resolved in an ECP is limited more by the sample than by α. The finite α appears only to broaden the linewidths of higher order lines. These studies are being pursued to explore the potential of this technique for the study of crystalline perfection.

Charging of the specimen under electron beam irradiation is a common problem in scanning electron microscopy (SEM). It results in unstable imaging conditions and a loss in resolution due to defocus of the beam. In addition, it can cause permanent changes in some specimens from translocation of mobile ions under the influence of the induced electrostatic field. To minimize charging and its associated problems, the incident beam energy must be carefully chosen to be the value E2 at which a dynamic charge balance is obtained. This article presents data on E2 values for a variety of materials and demonstrates how E2 is affected by the choice of angle of beam incidence.