Software for scanning Auger microscopy

The procedures embodied in a program suite designed for use with a computer controlled digital scanning Auger electron microscope are elucidated. These include techniques for the collection and subsequent processing of spectra, linescans and images. Schemes found to be the most efficient for the normalization of data to correct for both incident beam current fluctuations and sample surface topography are explained. Averaging and economical smoothing algorithms are considered as a means of improving the signal-to-noise ratio when working with high spatial resolution (probe diameter less than 150 nm), which necessarily limits the available beam current to less than about 3 nA. Emphasis is laid upon the desirability of obtaining quantitative information and to this end software routines for use in ascertaining surface chemical composition from spectra are described. These include numerical differentiation of spectra and modelling and subtraction of the secondary electron background from beneath an Auger feature. The transfer of data to a faster computer and more lengthy processing operations, for improving the quality of Auger maps, are also discussed.     

On the calculation of magnetic fields generated by coils

The edge effect in electron microscopical images is proposed to be measured in length units by using a ratio of the edge or near-to-edge signal enhancement or depletion integrated over all this feature along a line-scan across an edge, to the mean signal away from the edge, giving the result in length units. This approach is applied to quantification of the edge effect in Auger electron images, which is examined by using specimens with sharp edge terminated overlayer terraces, both chemically homogeneous and heterogeneous. The following combinations of overlayer/substrate materials were used: Si/Si of three different overlayer thicknesses, W/W and W/Si, and line-scans across the edges were recorded in both the low- and the high-energy Auger electrons mode, i.e. Si LMM and Si KLL or W NOO and W MNN. The experimental results are presented for the 3-, 10- and 20-keV primary electron energies. Owing to a low signal-to-noise ratio in the measured data, basic relations between the effect appearance and the experimental conditions were revealed only: on both Si and W homogeneous specimens with a surface step, the edge enhancement is the dominating subphenomenon while at the W terrace edge on the Si substrate, the ‘penetration’ of the radiation characteristic to both areas separated by the step, to the neighbouring feature, is observed as the most significant effect. The quantification has shown that the effect is, first of all, proportional to the step height, amounting to one-third to up to the full height, while the material dependence was weak, equally to the dependence on the Auger electron energy. The primary electron energy dependence is increasing in accordance with expectation. The results indicate that the effect cannot be modelled simply by the interaction volume cut by the surface step but phenomena such as subsurface electron channelling along the sidewall have to be taken into account.     

An experimental method for calibration of the plasmon mean free path

Transmission electron microscopy specimens in the form of elongated, conical needles were made using a dual‐beam focused ion beam system, allowing the specimen thickness to be geometrically determined for a range of thickness values. From the same samples electron energy loss maps were acquired and the plasmon mean free path (λ) for inelastic scattering was determined experimentally from the measured values of specimen thickness. To test the method λ was determined for Ni (174 ± 17 nm), α‐Al₂O₃ (143 ± 14 nm), Si (199 ± 20 nm) and amorphous SiO₂ (238 ± 12 nm), and compared both to experimental values of λ taken from the literature and to calculated values. The calculated values of λ significantly underestimate the true sample thickness for high accelerating voltages (300 kV) and large collection angles. A linear dependence of λ on thickness was confirmed for t/λ < 0.5–0.6, but this method also provides an approach for calibrating λ at sample thicknesses for which multiple scattering occurs, thus expanding the thickness range over which electron energy loss spectroscopy can be used to determine the absolute sample thickness (t/λ > 0.6). The experimental method proposed in this contribution offers a means to calibrate λ for any type of material or phase that can be milled using a focused ion beam system.     

Validity of the dipole-selection rule for the Al-L2,3 edge of alpha-Al2O3 under channeling conditions.

The validity of the dipole-selection rule for the Al-L2,3 electron energy-loss edge of alpha-Al2O3 is investigated. Dipole forbidden transitions can be observed in a transmission electron microscope operated with large collection apertures. In addition, it is shown that channeling along a highly symmetrical zone axis in alpha-Al2O3 can lead to the detection of dipole forbidden transitions even with small collection apertures. For an incident electron direction parallel to the <0001> orientation it is observed experimentally that the fine structure of the Al-L2,3 edge shows additional features compared to measurements with the electron beam parallel to <1100>. This effect is due to the occurrence of channeling conditions, indicated by the dependence of the additional dipole forbidden features on the sample thickness. These additional features disappear when tilting the crystal by 1.8 degrees (0.84 A(-1)) or even into the less-symmetrical <1100> zone axis. It is suggested that these observations are explained by differences in local symmetry at the excited center with respect to the incident beam directions.     

The imaging mechanism of single-walled carbon nanotubes on Si/SiO₂ wafer in scanning electron microscopy

Scanning electron microscopy imaging of both suspended single‐walled carbon nanotubes (SWNTs) and contacted SWNTs with Si/SiO₂ substrate has been studied in this paper. The voltage contrast has been investigated by supplying external electric field around the samples. The results show that the image contrast of SWNTs attributes to both voltage contrast from the area surrounding SWNTs (tens of nanometres in both sides of the SWNTs) and electron beam induced emission from SWNTs themselves under low primary beam energy. Under high primary beam energy, however, EBIE dominates the image contrast due to the fact that the voltage contrast caused by implanted charges of the SiO₂ layer is weakened. Imaging under the primary beam energy lower than 1 keV offers widened diameter of SWNTs, which promises that the SWNTs are observable at very low magnification (lower than 100×). At a larger magnification, however, imaging under the primary beam energy higher than 10 keV can display more realistic images of the SWNTs. In addition, an appropriate external electric field can improve the images.     

Bandgap measurement of thin dielectric films using monochromated STEM-EELS

High-resolution electron energy-loss spectroscopy (HR-EELS), achieved by attaching electron monochromators to transmission electron microscopes (TEM), has proved to be a powerful tool for measuring bandgaps. However, the method itself is still uncertain, due to Cerenkov loss and surface effects that can potentially influence the quality of EELS data. In the present study, we achieved an energy resolution of about 0.13 eV at 0.1 s, with a spatial resolution of a few nanometers, using a monochromated STEM-EELS technique. We also assessed various methods of bandgap measurement for a-SiNx and SiO₂ thin dielectric films. It was found that the linear fit method was more reliable than the onset reading method in avoiding the effects of Cerenkov loss and specimen thickness. The bandgap of the SiO₂ was estimated to be 8.95 eV, and those of a-SiNx with N/Si ratios of 1.46, 1.20 and 0.92 were measured as 5.3, 4.1 and 2.9 eV, respectively. These bandgap-measurement results using monochromated STEM-EELS were compared with those using Auger electron spectroscopy (AES)-reflective EELS (REELS).     

X-ray probe microanalyses and scanning x-ray microscopies

The analytical techniques based on excitation of inner-shell electrons by an incident X-ray photon beam suffer generally from their poor degree of lateral localization. Nevertheless it is possible to perform the microanalysis of a sample in the forms of a thin film by: (i) X-ray photoelectron spectroscopy (XPS or ESCA) and X-ray-induced Auger electron spectroscopy (XAES) for surface analysis; (ii) X-ray absorption spectroscopy (XAS) and X-ray microfluorescence spectroscopy (XMS) for bulk analysis. The corresponding images can also be obtained in the scanning mode: scanning X-ray photoelectron microscopy (SXPM), scanning X-ray-induced Auger electron microscopy (SXAEM) and scanning X-ray microradiography (SXM). The experimental arrangement and the results obtained are described here, together with further improvements and comparisons with other technical solutions.     

Electron energy-loss spectroscopy study of thin film hafnium aluminates for novel gate dielectrics

We have used conventional high‐resolution transmission electron microscopy and electron energy‐loss spectroscopy (EELS) in scanning transmission electron microscopy to investigate the microstructure and electronic structure of hafnia‐based thin films doped with small amounts (6.8 at.%) of Al grown on (001) Si. The as‐deposited film is amorphous with a very thin (～0.5 nm) interfacial SiO_x layer. The film partially crystallizes after annealing at 700 °C and the interfacial SiO₂‐like layer increases in thickness by oxygen diffusion through the Hf‐aluminate layer and oxidation of the silicon substrate. Oxygen K‐edge EELS fine‐structures are analysed for both films and interpreted in the context of the films’ microstructure. We also discuss valence electron energy‐loss spectra of these ultrathin films.     

Analysis of EEL spectrum of low-loss region using the C(s)-corrected STEM-EELS method and multivariate analysis

We analyzed a Si/SiO₂ interface using multivariate analysis and spherical aberration-corrected scanning transmission electron microscopy-electron energy loss spectroscopy which is characterized by using the electron energy loss spectrum of the low-loss region. We extracted the low-loss spectra of Si, SiO₂ and an interface state. Even if the interface is formed from materials with different dielectric functions, the present method will prove suitable for obtaining a more quantitative understanding of the dielectric characteristic.     

The spatial distribution and resolution of coaxial backscattered electrons in SEM, calculated by Monte Carlo simulations

We simulate, within a sample, the trajectories of the backscattered electrons detected in a scanning electron microscopy with a particular detection geometry. Thus we obtain the depth and lateral distributions, according to the adjustable parameter values, of the detected electrons. Finally, the scanline profile across a chemical edge is drawn. The conditions corresponding to the best lateral resolution are established; we obtain an ultimate resolution of the same order as the beam diameter.     

Polarization dependence in ELNES: influence of probe convergence, collector aperture and electron beam incidence angle

The differential scattering cross section in electron energy loss near edge spectroscopy (ELNES) generally depends on the orientation of the Q wave vector transferred from the incident electron to an atomic core electron. In the case where the excited atom belongs to a threefold, fourfold or sixfold main rotation axis, the dipole cross section depends on the angle of Q with respect to this axis. In this paper, we restrict to this situation called dichroism. Furthermore, if we take into account the relativistic effects due to the high incident electron velocity, this dipole cross section also depends on the angle of Q with respect to the electron beam axis. It is due to these dependences that the shape of measured electron energy loss spectra varies with the electron beam incidence, the collector aperture, the incident beam convergence and the incident electron energy. The existence of a particular beam incidence angle for which the scattering cross section becomes independent of collection and beam convergence semi-angles is clearly underscored. Conversely, it is shown that EELS spectra do not depend on the beam incidence angle for a set of particular values of collection and convergence semi-angles. Particularly, in the case of a parallel incident beam, there is a collection semi-angle (often called magic angle) for which the cross section becomes independent of the beam orientation. This magic angle depends on the incident beam kinetic energy. If the incident electron velocity V is small compared with the light velocity c, this magic angle is about 3.975theta(E) (theta(E) is the scattering angle). It decreases to 0 when V approaches c. These results are illustrated in the case of the K boron edge in the boron nitride.     

Analysis of bonding state in pure and gold-implanted amorphous SiC by a combination of UHVEM and AES

Single crystalline α-SiC films and Au(target metal)/α-SiC(substrate) bilayer films were irradiated with 2-MeV electrons in an ultrahigh-voltage electron microscope (UHVEM) to form, in situ, pure amorphous SiC (a-SiC) and gold-implanted a-SiC. respectively. In the latter case, sample films were set so that the electron beam was incident on the gold layer. Differences in the bonding state between pure a-SiC and gold-implanted a-SiC were studied by Auger valence electron spectroscopy (AES). In pure a-SiC. the carbon valence state was markedly different from that of α-SiC but the silicon valence state was similar to that of α-SiC. In gold-implanted α-SiC. both the carbon and the silicon valence states were different from those of α-SiC. It is suggested that a high density of molecularized carbon clusters are formed in the amorphous matrix in gold-enriched a-SiC.     

Monte carlo simulation of charging effects in linewidth metrology (II): On insulator substrate

Charging effects have been investigated quantitatively using Monte Carlo (MC) simulation when the linewidth of polymethylmethacrylate (PMMA) insulator patterns on SiO₂ insulator substrate are measured by scanning electron microscope (SEM). We established reference operating and shape conditions for array patterns and we have calculated the offset on linewidth metrology according to the change in each condition. We have used a 50% threshold algorithm for the edge determination, calculated the offsets in those conditions, and compared them with the results in the case of Si substrate. Finally, the question of which factor is the most sensitive in linewidth metrology is discussed.     

Influence of UV irradiation in low frictional performance of CNx coatings

The influence of ultraviolet (UV) irradiation on low frictional performance of CN_x coatings with 100 nm thickness having nitrogen contents of 9%, 14% and 19% deposited on Si(100) substrate by ion beam mixing was investigated in N₂ atmosphere environment. Three UV lights of 254, 312 and 365 nm were used to irradiate the surface of CN_x / Si(100) for 60 min. The changes of N / C ratio and atomic binding energy in the coating were analysed using Auger electron spectroscopy and X‐ray photoelectron spectroscopy, respectively. The friction coefficient of Si₃N₄ ball sliding against CN_x was measured by a pin‐on‐disc tribometer, and wear tracks were analysed by the transmission electron microscope image. The results showed that UV irradiation on CN_x coating can decrease the critical frictional cycles for low friction coefficient and that the mechanism is due to the formation of graphite‐like structure in the topmost CN_x coating. Copyright © 2012 John Wiley & Sons, Ltd.     

Electronic core level microanalyses and microscopies in multipurpose apparatus

Using an instrument equipped with two electron guns, an electron analyzer, and a Si(Li) diode detector, we developed microanalytical techniques based on inner-shell electron excitations by incident electrons and X-rays, that is, electron energy-loss spectroscopy (EELS) in the reflection mode; electron probe microanalysis (EPMA) and X-ray appearance potential spectroscopy (XAPS); electron-induced Auger electron spectroscopy (e-AES); X-ray photoelectron spectroscopy (XPS), X-ray absorption spectroscopy (XAS); X-ray induced AES (XAES), X-ray fluorescence analysis (XRF), and scanning X-ray radiography (SXR). The corresponding characteristic images (including X-ray microradiography and X-ray photoelectron microscopy) were obtained in the scanning mode. The principle of the apparatus is described. Each spectroscopy and microscopy is illustrated by an example. Their performance and limits are discussed.     

Analysis of Kikuchi band contrast reversal in electron backscatter diffraction patterns of silicon

We analyze the contrast reversal of Kikuchi bands that can be seen in electron backscatter diffraction (EBSD) patterns under specific experimental conditions. The observed effect can be reproduced using dynamical electron diffraction calculations. Two crucial contributions are identified to be at work: First, the incident beam creates a depth distribution of incoherently backscattered electrons which depends on the incidence angle of the beam. Second, the localized inelastic scattering in the outgoing path leads to pronounced anomalous absorption effects for electrons at grazing emission angles, as these electrons have to go through the largest amount of material. We use simple model depth distributions to account for the incident beam effect, and we assume an exit angle dependent effective crystal thickness in the dynamical electron diffraction calculations. Very good agreement is obtained with experimental observations for silicon at 20 keV primary beam energy.     

Diffraction effects and inelastic electron transport in angle‐resolved microscopic imaging applications

We analyse the signal formation process for scanning electron microscopic imaging applications on crystalline specimens. In accordance with previous investigations, we find nontrivial effects of incident beam diffraction on the backscattered electron distribution in energy and momentum. Specifically, incident beam diffraction causes angular changes of the backscattered electron distribution which we identify as the dominant mechanism underlying pseudocolour orientation imaging using multiple, angle‐resolving detectors. Consequently, diffraction effects of the incident beam and their impact on the subsequent coherent and incoherent electron transport need to be taken into account for an in‐depth theoretical modelling of the energy‐ and momentum distribution of electrons backscattered from crystalline sample regions. Our findings have implications for the level of theoretical detail that can be necessary for the interpretation of complex imaging modalities such as electron channelling contrast imaging (ECCI) of defects in crystals. If the solid angle of detection is limited to specific regions of the backscattered electron momentum distribution, the image contrast that is observed in ECCI and similar applications can be strongly affected by incident beam diffraction and topographic effects from the sample surface. As an application, we demonstrate characteristic changes in the resulting images if different properties of the backscattered electron distribution are used for the analysis of a GaN thin film sample containing dislocations.     

Modeling the point-spread function in helium-ion lithography

We present here a hybrid approach to modeling helium-ion lithography that combines the power and ease-of-use of the Stopping and Range of Ions in Matter (SRIM) software with the results of recent work simulating secondary electron (SE) yield in helium-ion microscopy. This approach traces along SRIM-produced helium-ion trajectories, generating and simulating trajectories for SEs using a Monte Carlo method. We found, both through simulation and experiment, that the spatial distribution of energy deposition in a resist as a function of radial distance from beam incidence, i.e. the point spread function, is not simply a sum of Gauss functions.     

Solid Lubrication of Silicon Nitride with Cesium-Based Compounds: Part II — Surface Analysis

Energy dispersive X-ray spectroscopy (EDS), X-ray photoelectron spectroscopy (XPS), and Auger electron spectroscopy (AES) were used to characterize the wear surfaces of selected samples from Part 1 of the authors study. Results are presented for films generated on silicon nitride (Si₃N₄) originally coated with cesium oxytrithiotungstate (Cs₂WOS₃), cesium sulfate (Cs₂SO₄), and a hydrated cesium silicate (Cs₂O·3SiO₂·nH₂O), all applied in a sodium silicate binder (Na₂SiO₃). Results show the presence of mostly Si, O, and Cs within the wear tracks of post-tested specimens. In some cases, W and S were not detected on samples that originally contained these elements, suggesting that decomposition had taken place. To simulate the reactions that might occur in a tribo-contact, mixtures of Si₃N⁴ and Cs₂WOS₃ powders were heated in air to 700°C and analyzed using XPS and Bremsstrahlung-excited AES. It was found that Cs₂WOS₃ accelerates the formation ofSiO₂ on Si₃N₄ under static conditions. These results support our hypothesis that high temperature chemical reactions between the cesium-containing compounds and the Si₃N₄ surface form a lubricious cesium silicate film. A mechanism is proposed based on the glass-modifying tendency of alkali metals and the hot-corrosion of Si₃N₄     

Large-scale process optimization for focused ion beam 3-D nanofabrication

With the ever-decreasing size of manufactured objects, fabrication processes driven by charged particle beams, such as focused ion beam (FIB), become important for a wide spectrum of interdisciplinary applications. A designed three-dimensional (3-D) pattern to fabricate may contain millions of pixels, which will require solving an unprecedented large-scale problem for planning. This paper proposes a general framework of planning FIB milling for fabricating 3-D nanostructures, including model formulations to enable FIB for scalable and automated applications and a corresponding optimization model to support the process planning. The implementation of proposed work does not affect the fabrication quality and yet tremendously reduces the required computational time and data storage during planning. The proposed framework of process planning is further illustrated and verified by simulation and milling experiments of submicron features on Si and Si₃N₄. This research offers an accurate and economical solution to the realization from designs to actual micro/nanoscale models and builds a scientific foundation for immediate development of complex, yet more accurate and cost-effective, beam scanning techniques.