This special issue of Microscopy and Microanalysis contains 17 articles from a topical conference on microbeam characterization of nonconductive materials. The conference was held at the Department of Mining, Metals, and Materials Engineering of McGill University in Montreal on August 2–3, 2002, as the pre-meeting congress prior to the Microscopy & Microanalysis 2002 meeting in Quebec City, Canada.

The physical mechanisms involved in electron irradiation of insulating specimens are investigated by combining some simple considerations of solid-state physics (trapping mechanisms of electrons and secondary electron emission) with basic equations of electrostatics. To facilitate the understanding of the involved mechanisms only widely irradiated samples having a uniform distribution of trapping sites are considered. This starting hypothesis allows development of simple models for the trapped charge distributions in ground-coated specimens as investigated in electron probe microanalysis (EPMA) as well as for the bare specimens investigated in scanning electron microscopy (SEM) and environmental SEM (ESEM). Governed by self-regulation processes, the evolution of the electric parameters during the irradiation are also considered for the first time and practical consequences in EPMA, SEM, and ESEM are deduced. In particular, the widespread idea that the noncharging condition of SEM is obtained at a critical energy E2 (where δ + η = 1 with δ and η yields obtained in noncharging experiments) is critically discussed.

In the context of electron microscopists' changing attitudes to charging effects, some basic aspects of these phenomenona are surveyed. Methods of mapping internal charge distributions such as doping levels in semiconductors, trap distributions, or internal electric fields in insulators are discussed.

When a dielectric is irradiated by electrons with energy E of several kiloelectron volts, a large number of processes take place: backscattering of incident electrons, excitation and ionization of the electrons in the dielectric with binding energies lower than E, creation of excitons, radiative and nonradiative decays of the excited and ionized states, slowing down of the primary and secondary electrons, and thermalization in the conduction band. The thermalized electrons can move freely in the unoccupied conduction states of the material. If electric connection exists between the dielectric and the apparatus, then the charges normally flow out. Thermalized electrons can also be trapped in excited levels localized in the band gap of the dielectric and nonradiative and radiative recombinations from these levels can be observed. The number of the trapped electrons varies with the structural characteristics of the dielectric. In a monocrystal, this number is weak because the number of the defect states in the band gap is small, making the localization of the charges restricted. In contrast, in a polycrystal or amorphous material, the number of the trapped electrons can be large and increases with the disorder. Information on the charge effects suffered by the sample during its irradiation can be deduced by studying the trapping of electrons in localized states and, consequently, by analyzing radiations emitted from these states in the visible and X-ray ranges. In the case of oxides, F+ centers (oxygen–ion vacancy having trapped one electron) and F centers (F+ center having trapped a second electron) are generally present. We will show that the F+ [harr ] F conversion can be used to study the dynamic of the trapping in the oxides. Application to various samples of crystallized and amorphous alumina will be presented.

The goal of this article is first to review the charging effects occurring when an insulating material is subjected to electron irradiation in a scanning electron microscope (SEM) and next their consequences from both scanning electron microscopy and electron probe microanalysis (EPMA) points of view. When bare insulators are observed, the so-called pseudo mirror effect leads to an anomalous contrast and also to an erroneous surface potential, VS, measurement when a Duane–Hunt limit (DHL) method is used. An alternative possibility is to use an electron toroidal spectrometer (ETS), specially adapted to a SEM, which directly gives the VS value. In the case of a bulk specimen coated with a grounded layer, although the layer prevents external effects of the trapped charge, the electric field beneath the coating is reinforced and leads to loss of ionizations that reduces the number of generated X-ray photons. To take into account both effects mentioned above, whether the studied insulator is coated or not, a method is proposed to deduce the trapped charge inside the insulator and the corresponding internal or external electric field.

A framework is presented for understanding charging processes in low vacuum scanning electron microscopy. We consider the effects of electric fields generated above and below the specimen surface and their effects on various processes taking place in the system. These processes include the formation of an ionic space charge, field-enhanced electron emission, charge trapping and dissipation, and electron–ion recombination. The physical mechanisms behind each of these processes are discussed, as are the microscope operating conditions under which each process is most effective. Readily observable effects on gas gain curves, secondary electron images, and X-ray spectra are discussed.

The variable pressure scanning electron microscope (VP-SEM) allows imaging of insulators without the need for a conductive coating, due to charge neutralization at the surface from recombination of positive ions and surface electrons. Varying certain parameters such as pressure, bias, and working distance creates incomplete neutralization, and localized charging develops called charge contrast. Although the exact mechanism creating charge contrast imaging (CCI) is unknown, it is agreed that it is related to an optimum charge compensation. The behavior of the CCI is still vague, which presents a problem for determining the mechanisms. This article provides user-friendly methods of finding the optimum levels of charge contrast in the VP-SEM. We show that the CCI is obtained at optimum operating conditions where the specimen current is between 2.5 nA and 3.5 nA. The specimen current is a function of secondary electrons (SE) emission and ionization potential, producing an ion flux. Therefore an optimum specimen current represents the balanced conditions of SE emission and ion flux. Controlling the pressure, working distance, bias, scan rate, and beam current allows the microscopist to set the specimen current at this optimum level for charge contrast imaging. All the work was performed on gibbsite using the S3000N VP-SEM from Hitachi.

The influence of a lack of sufficient electrical conductivity on the results of quantitative electron probe microanalysis has been investigated on a number of oxides. The effect of surface charging and the way it alters the emitted X-ray signals has been studied. It is shown that the presence of conducting coatings, such as carbon or copper, will affect the interelement X-ray intensity ratios, whatever the thickness of the coating may be. Although the effects for heavier elements may be acceptable, they cannot be ignored for a light element such as oxygen, where strong variations with coating thickness were observed. Quantitative analyses of oxygen, on uncoated well-conducting oxide specimens, using uncoated well-conducting hematite (Fe2O3) as a standard yielded excellent results in the range between 4 and 40 kV with the φ(ρz) software used. As soon as coated nonconducting specimens were examined, using the same hematite standard, coated under exactly the same conditions, widely scattering and noncoherent results were obtained. These discrepancies can only be attributed to a lack of conductivity.

Energy dispersive X-ray spectrometry of uncoated insulators performed at low beam energy (incident energy ≤ 5 keV) and in the variable pressure scanning electron microscope and the environmental scanning electron microscope is subject to spectral artifacts. Charging decelerates the incident beam electrons and reduces the impact energy, lowering the available overvoltage to excite characteristic X-ray peaks. The Duane–Hunt limit of the X-ray bremsstrahlung continuum is commonly used as a diagnostic of charging. Dynamic charging effects can hide the true impact of charging on the X-ray spectrum. Careful examination of the behavior of the X-ray spectrum with time and other variables is needed to avoid spectral artifacts, particularly on relative X-ray intensities.

A signal loss is generally reported in electron probe microanalysis (EPMA) of porous, highly divided materials like heterogeneous catalysts. The hypothesis generally proposed to explain this signal loss refers to porosity, roughness, energy losses at interfaces, or charging effects. In this work we investigate by Monte Carlo simulation all these physical effects and compare the simulated results with measurements obtained on a mesoporous alumina. A program using the PENELOPE package and taking into account these four physical phenomena has been written. Simulation results show clearly that neither porosity nor roughness, nor specific energy losses at interfaces, nor charging effects are responsible for the observed signal loss. Measurements performed with analysis of carbon and oxygen lead to a correct total of concentration. The signal loss is thus explained by a composition effect due to a carbon contamination brought by the sample preparation and to a lesser extent by a stoichiometry of the porous alumina different from a massive alumina. For this kind of high specific surface porous sample, a little surface contamination layer becomes an important volume contamination that can produce large quantification errors if the contaminant is not analyzed.

Quantitative chemical analysis by energy-dispersive X-ray spectrometry (EDS) in the environmental scanning electron microscope (ESEM) is difficult. This analysis is complicated by the spread of the electron beam by chamber gas molecules and the necessity for surface charge neutralization. Without charge neutralization, errors in quantitative analysis can range up to 15–20% relative. It is possible to achieve the error expected of traditional EDS, ±5% relative error, using a newly developed surface charge neutralization scheme for the ESEM. Estimates of accuracy and precision are based on studies of the National Bureau of Standards (now National Institutes for Science and Technology) Standard Reference Material 482, a series of certified copper–gold alloys. The scheme for charge neutralization requires an independent path to ground at or near the surface of the specimen. The current through the ground path must be maintained at zero by adjusting the voltage on the Gaseous Secondary Electron DetectorTM when the sample chamber is at a gas pressure of 1–2 torr. This procedure forms the exact number of chamber gas positive ions to neutralize negative electrical charge on the specimen surface from electron bombardment.

Very low energy electrons in a solid should behave like Bloch electrons and will interact with perturbations of the atomic lattice, that is, with phonons. So we use the acoustic phonon scattering for replacing the elastic binary encounter approximation of the Mott scattering for electrons with low energies E < 100 eV. For ballistic electrons (1 eV < E < Eg) and higher energies up to 1 keV we determined the acoustic phonon scattering and the impact ionization rate by means of the “backscattering-versus-range” proof and respective η(E0) − R(E0) diagrams. Electron trajectories demonstrate the relatively short range of primary electrons (PE) with energies E > 50 eV due to strong impact ionization losses (cascading) and the much greater range of secondary electrons (SE) with E < 50 eV, finally as a consequence of less effective phonon losses. The field-dependent transport parameters allow us to model the self-consistent charge transport and charging-up of insulating SiO2 layers during electron bombardment maintained by the current components of primary electrons jPE, secondary electrons jSE, and associated ballistic holes jBH, as well as by Fowler–Nordheim field injection jFN from the substrate. The resulting distributions of currents j(x,t), charges ρ(x,t), electric fields F(x,t), and the potential V(x,t) across the dielectric layer explain the phenomena of field-enhanced and field-blocked secondary electron emission with rates δ [gel ] 1.

Charging phenomena of a mask material during electron beam exposure are analyzed in an electron beam projection lithography system. First, the three-dimensional charge deposition distribution by the electron beam irradiation is obtained. Next, in every time step, the distributions of the accumulated charge and the potential are obtained considering the current flow due to the diffusion and the drift. As a narrow bridge pattern defined in a 5 μm × 5 μm area is assumed to lay out all over the field of 1 mm × 1 mm and the potential is grounded at the circumference of the field (1 mm × 1 mm × 1 mm), the saturated potential distribution is obtained at the central 5 μm × 5 μm area in the field. The maximum potential attained is around 4.23 μV at the center of the bridge, if the accelerating voltage of the electron beam is 100 kV, the current density is 10 A/cm2, and the material of the mask is the intrinsic Si. The contribution of the charging may be negligible to the electron beam with such a high accelerating voltage, which is going through the opening beside the bridge pattern in the projection lithography.

The microanalysis of nonconductive specimen in a scanning electron microscope is limited by charging effects. Using a charge density model for the electric field buildup in a nonconductive specimen irradiated by electrons, a Monte Carlo simulation method has been applied to alumina (Al2O3). The results show a change in the depth distribution for characteristic and bremsstrahlung X-ray, φ(ρz) curves, and ψ(ρz) curves (with absorption) for both elements' Kα lines. The influence of the electric field on the measured X-ray intensity is shown. The dependency of this influence by the three parameters, electron energy, X-ray energy, and charge density, is clarified.

Specimen charging may be one of the most significant factors that contribute to the high variability and generally low quality of images in cryo-electron microscopy. Understanding the nature of specimen charging can help in devising methods to reduce or even avoid its effects and thus improve the rate of data collection as well as the quality of the data. We describe a series of experiments that help to characterize the charging phenomenon, which has been termed the Berriman effect. The pattern of buildup and disappearance of the charge pattern has led to several suggestions for how to alleviate the effect. Experiments are described that demonstrate the feasibility of such charge mitigation.

Although the most familiar consequences of specimen charging in transmission electron microscopy can be eliminated by evaporating a thin conducting film (such as a carbon film) onto an insulating specimen or by preparing samples directly on such a conducting film to begin with, a more subtle charging effect still remains. We argue here that specimen charging is in this case likely to produce a dipole sheet rather than a layer of positive charge at the surface of the specimen. A simple model of the factors that control the kinetics of specimen charging, and its neutralization, is discussed as a guide for experiments that attempt to minimize the amount of specimen charging. Believable estimates of the electrostatic forces and the electron optical disturbances that are likely to occur suggest that specimen bending and warping may have the biggest impact on degrading the image quality at high resolution. Electron optical effects are likely to be negligible except in the case of a specimen that is tilted to high angle. A model is proposed to explain how both the mechanical and electron-optical effects of forming a dipole layer would have much greater impact on the image resolution in a direction perpendicular to the tilt axis, a well-known effect in electron microscopy of two-dimensional crystals.

Kelvin probe microscopy (KPM) is a specialized atomic force microscopy technique in which long-range Coulomb forces between a conductive atomic force probe and a specimen enable the electrical potential at the surface of a specimen to be characterized with high spatial resolution. KPM has been used to characterize nonconductive materials following their exposure to stationary electron beam irradiation in a scanning electron microscope (SEM). Charged beam irradiation of poorly conducting materials results in the trapping of charge at either preexisting or irradiation-induced defects. The reproducible characteristic surface potentials associated with the trapped charge have been mapped using KPM. Potential profiles are calculated and compared with observed potential profiles giving insight into the charging processes and residual trapped charge distributions.

A major issue in low voltage lithography is surface charging, which results in beam deflection presented as uneven exposure between adjacent structures. In this study, charge-induced pattern distortions in low-voltage energy beam lithography (LVEBL) were investigated using a silicide direct-write electron beam lithography process. Two methodologies have been proposed to avert charging effects in LVEBL, namely, pattern randomizing and lithography using the crossover voltage. Experimental results demonstrated that these methods are effective in significantly reducing the problems associated with charging. They indicate that charging on a sample is a function of time interval and proximity between line structures. In addition, the optimum time and distance between exposures for no charge-induced pattern distortion were determined. By using the crossover voltage of the material for lithography, charging effect can be significantly minimized.

Following is a list of microscopy-related meetings and courses. The editors would greatly appreciate input to this list via the electronic submission form found in the MSA World-Wide Web page at http://www.msa.microscopy.com. We will gladly add hypertext links to the notice on the web and insert a listing of the meeting in the next issue of the Journal. Send comments and questions to JoAn Hudson, [email protected] or Nestor Zaluzec, [email protected].