Electron Beam Diameter Research Articles

Progress and research at the Shanghai EBIT

In this report, a brief description of the current progress at the Shanghai EBIT project is presented. This is followed by a short discussion of the measurement of various parameters (electron beam diameter and ion density) under a number of operational conditions. A brief introduction to di-electronic recombination measurements for highly ionized xenon is given. Next, we present a preliminary measurement of the time dependence of xenon X-ray emission lines. Finally, a comparison of calculated and experimental charge-state distributions is given. This shows the influence of multi-electron capture and different distributions of the ion cloud on the charge state distribution.PACS Nos.: 41.85.–p; 34.80.Kw; 34.80Lx

Use of focused ion beam sample preparation for Auger analysis

Abstract FIB cross‐sections are often used to get access to sample features of interest. Due to the inherent surface sensitivity an AES measurement is favourable for small feature compositional analysis on such samples. Even though the primary electron beam diameter of a modern field emission AES instrument is about 30 nm and thereby nearly ten times that of a SEM, the analysis volume of an AES measurement is much smaller than that of SEM/EDX. As AES measurements on small sample features can last for many hours it is mandatory to compensate a likely mechanical drift. In this procedure at regular intervals SE images are compared with an SE image recorded at the beginning of the measurement. Lateral drifts are compensated by an additional electrostatic deflection of the primary beam. For this purpose on low contrast samples or samples with periodically repetitive structures a FIB crater cut into the sample surface can be successfully applied as a marker feature to determine sample drift. (© 2007 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Modeling of the integrated magnetic focusing and gated field-emission device with single carbon nanotube

A gated single carbon nanotube field emitter with magnetic focusing is proposed and simulated using a parallelized Poisson’s equation solver, coupled with the ray tracing of electrons, on an unstructured tetrahedral adaptive mesh. The magnetic focusing for the electrons can be achieved by a vertically downward magnetic focusing field (−Bz) through the use of either external solenoids or permanent magnets around the field-emission array. The simulation results, assuming uniform magnetic field inside a field-emission unit, are compared with those conventional tetrode-type field emitters using an electrostatic focusing structure. The results reveal that the magnetic focusing design can promise much higher emission current, while a much smaller spot size results at the anode. In addition, the magnitude of the applied gate voltage in the range of 60–120V shows little influence on the electron-beam diameter at the anode. The proposed magnetic focusing method can also possibly reduce the complexity of the fabrication without the electrostatic focusing structure. Noticeably, a distribution, similar to the Airy function, is obtained that shows the dependence of the spot size at the anode on the magnetic flux intensity. Thus, under suitable magnetic focusing conditions, it is possible to produce well-defined microelectron sources for many field-emission applications, such as novel parallel electron-beam lithography or field-emission displays.

Custom design of optical-grade thin films of silicon oxide by direct-write electron-beam-induced deposition

This work describes a rapid fabrication approach of thin silicon oxide films on confined areas by electron-beam-induced deposition. This maskless direct-write process utilizes a localized chemical vapor deposition (CVD) on specific areas utilizing a focused electron beam. The deposition from siloxane vapor in the presence of oxygen is initiated by the energy of an electron beam of 1nm diameter. By scanning the beam, thin films with arbitrary geometries and three-dimensional structures were deposited. In contrast to blanket deposition with conventional methods such as thermal CVD or plasma-enhanced CVD, the lateral confined layers can be fabricated at room temperature. With a maskless process, the final structure is fabricated within a single process step. The process was optimized towards a high deposition rate and high material purity. The influence of process parameters on the deposition efficiency is discussed. A characterization of the chemical composition and of the surface roughness was performed with auger electron spectroscopy, energy-dispersive x-ray, and atomic force microscopy, respectively. The optical properties were investigated by transmission measurement at 248nm. The correlation to processing conditions and the growth mechanism induced by the electron beam is discussed. This work illustrates the flexibility of this maskless method and the potential to control material properties via the process parameters. The fabrication of exemplary structures such as three-dimensional silicon oxide pillars and transparent films illustrates the application potential of this versatile direct-write method.

Technique to automatically measure electron-beam diameter and astigmatism:BEAMETR

Accuracy and resolution of scanning electron microscopy and an electron-beam lithography system directly depend on beam diameter; it should be monitored and tuned frequently. A technique is described to determine beam size using an automatic procedure. In the developed method, a specially designed and fabricated test pattern is scanned using an e-beam. A spectrum of the signal is analyzed; beam diameter is automatically determined using a software program. Results of design, fabrication, and analysis of the beam calibration test pattern are presented.

Scale dependence of Weibel instability of relativistic electron beam in high-dense plasma

Nonlinear dynamics of relativistic electron current in dense plasmas has been investigated by using two-dimensional Fluid-Particle Hybrid code (FPH code). FPH simulations show that relativistic electron filaments are generated by the Welbel instability and merge each other to be self-organized. When the relativistic electron beam diameter is large and the total current is over 100 times Alfven limit current, many filaments are confined in the initial beam area. It is found that the number of relativistic electron filament decreases in proportion to t -0.9 in the fixed ion case. This asymptotic behavior is explained by the random motion of filaments driven by fluctuating magnetic fields.

Simulation of the nonlinear evolution of large scale relativistic electron flow in dense plasmas

A relativistic electron beam of about 100MA is transported through overdense plasmas in the fast ignition. The nonlinear dynamics of the relativistic electron beam in dense plasmas has been investigated using a two-dimensional fluid-particle hybrid (FPH) code [T. Taguchi et al., Phys. Rev. Lett. 86, 5055 (2001)] that combines a particle-in cell code with an Rational Cubic Interpolated Pseudo-Particle fluid code [F. Xiao et al., Comput. Phys. Commun. 93, 1 (1996)]. These simulations show that the relativistic electron beam breaks up into filaments from the Weibel instability and the filaments merge successively to larger filaments. When the relativistic electron beam diameter is large and the total current is over 100 times Alfvén limit current, many large scale filaments remain after the merging process and are confined in the initial beam diameter. It is found that the number of relativistic electron filaments decreases in proportion to t−0.9 in the fixed ion case and t−0.2 in the mobile ion case until ωpe0τ≈1000, respectively. This asymptotic behavior is caused by the random motion of filaments driven by fluctuating magnetic fields.

A Novel Method of Analytical Transmission Electron Microscopy for Measuring Highly Accurately Segregation to Special Grain Boundaries or Planar Interfaces

A new technique of analytical transmission electron microscopy has been developed for determining accurately the amount of solute atoms incorporated into well-defined planar defects in solids, such as stacking faults, special grain boundaries or interfaces. The method is based on recording series of analytical spectra taken with different electron beam diameters on the same position centred above a defect that is oriented nearly edge-on. The matrix/solute ratio is then plotted as a function of the electron beam radius, linear regression is performed and the segregation level is determined from the slope of the best fit. This method can be applied to energy-dispersive X-ray (EDX) or electron energy-loss spectroscopy (EELS) in a transmission electron microscope. It necessitates a nano-probe mode but no scan unit. Reliability and accuracy have been tested numerically. Simulations suggest an accuracy in the determination of the Gibbsian solute excess at a special grain boundary down to ±0.1 atoms nm−2 under typical experimental conditions, with a maximum error about twice as large.

Multi-stacked MEMS nano-membranes for coherent extreme ultraviolet emission

Fabricating precise freestanding metal membranes tens of nanometers thick in a continuous stack with a constant air gap of several microns is a key challenge in a free-electron extreme ultraviolet (EUV) source design program. Such a stack, when exposed to a high energy (∼9 MeV) electron beam, will operate as an EUV light emitting device. A novel MEMS-based microfabrication technique enables manufacturing of a precise-sized large-area membrane infinite stack for exposure by 1 mm diameter electron beam. Thermal self-tensioning is used to reach the required planarity of the membranes.

Static properties of a femtosecond electron diffraction system

Ultrafast electron diffraction is an important technique to study the ultrafast phenomenon in physical, chemical and biological processes. This paper introduces a femtosecond electron diffractometer. The diameter of electron beam and deflection sensitivity of X-Y deflection plates are reported. We also demonstrate the static diffraction pattern of a 300 nm thick gold film taken by the femtosecond electron diffractometer.

Evaluation and Optimization of Electron Capture Dissociation Efficiency in Fourier Transform Ion Cyclotron Resonance Mass Spectrometry

Electron capture dissociation (ECD) efficiency has typically been lower than for other dissociation techniques. Here we characterize experimental factors that limit ECD and seek to improve its efficiency. Efficiency of precursor to product ion conversion was measured for a range of peptide (approximately 15% efficiency) and protein (approximately 33% efficiency) ions of differing sizes and charge states. Conversion of precursor ions to products depends on electron irradiation period and maximizes at approximately 5-30 ms. The optimal irradiation period scales inversely with charge state. We demonstrate that reflection of electrons through the ICR cell is more efficient and robust than a single pass, because electrons can cool to the optimal energy for capture, which allows for a wide range of initial electron energy. Further, efficient ECD with reflected electrons requires only a short (approximately 500 micros) irradiation period followed by an appropriate delay for cooling and interaction. Reflection of the electron beam results in electrons trapped in or near the ICR cell and thus requires a brief (approximately 50 micros) purge for successful mass spectral acquisition. Further electron irradiation of refractory precursor ions did not result in further dissociation. Possibly the ion cloud and electron beam are misaligned radially, or the electron beam diameter may be smaller than that of the ion cloud such that remaining precursor ions do not overlap with the electron beam. Several ion manipulation techniques and use of a large, movable dispenser cathode reduce the possibility that misalignment of the ion and electron beams limits ECD efficiency.

Surface Micro Modification of Machined Surfaces by Wide-Area Electron Beam (EB) Irradiation(Electrical machining)

Productive finishing of very smooth surface for dies and molds are still in huge demand in industries. The conventional method to obtain such high quality finished surface has mainly relied on the manual polishing in most cases. In some cases, concentrated energy flows such as intense pulsed laser, electron and ion beams and pulsed plasma flows have been tried or applied to treat workpiece surfaces in industries and academies. In this study, a large diameter electron beam (LDEB) with a maximum diameter of nearly 60mm was applied to treat very quickly machined surfaces of hardened alloys. The preliminary experimental studies performed on two difficult-to-machine materials, WC/Co and NAK80, have shown that surface finish could be significantly improved under optimized beam application conditions while over irradiation could worsen surface quality. The effect of EB irradiation on residual stress was explored. The defects induced by LDEB irradiation process were also observed and discussed. The mechanism of the modification mechanism, especially the interaction between the electrons and the material has also been discussed.

A 20 keV electron gun system for the electron irradiation experiments

An electron gun consisting of cathode, focusing electrode, control electrode and anode has been designed and fabricated for the electron irradiation experiments. This electron gun can provide electrons of any energy over the range 1–20 keV, with current upto 50 μA. This electron gun and a Faraday cup are mounted in the cylindrical chamber. The samples are fixed on the Faraday cup and irradiated with electrons at a pressure ∼10 −7 mbar. The special features of this electron gun system are that, at any electron energy above 1 keV, the electron beam diameter can be varied from 5 to 120 mm on the Faraday cup mounted at a distance of 200 mm from the anode in the chamber. The variation in the electron current over the beam spot of 120 mm diameter is less than 15% and the beam current stability is better than 5%. This system is being used for studying the irradiation effects of 1–20 keV energy electrons on the space quality materials in which the irradiation time may vary from a few tens of seconds to hours.

Development of a new analytical electron microscopy technique to quantify the chemistry of planar defects and to measure accurately solute segregation to grain boundaries

A new technique of analytical transmission electron microscopy called ConceptEM has been developed for determining highly accurately small amounts of solute or dopant atoms incorporated into well-defined planar defects such as stacking faults, grain boundaries or interfaces. The method is based on recording series of analytical spectra taken with different electron beam diameters on the same position centred above a defect that is orientated either edge-on or slightly inclined with respect to the electron beam. It can be applied to energy-dispersive X-ray spectroscopy or electron energy-loss spectroscopy and necessitates only a nano-probe modus but no scanning unit. Reliability and accuracy have been tested numerically under various conditions using simulations for a specific geometry, as a function of specimen thickness, material, acceleration voltage, collection angle, random beam displacements and solid solubility. The accuracy has been found to be substantially better (by factors of 5-10) than that of any other current standard technique based on single measurements. Our calculations suggest an accuracy in the determination of the Gibbsian solute excess at a special grain boundary down to +/- 1% of a monolayer, i.e. around +/- 0.1 atoms nm(-2) under typical experimental conditions, with a maximum error about twice as large. The parameter limiting a straightforward analysis is found to be the solid solubility, which itself, however, can be measured accurately by the technique so that it can be taken into account quantitatively and the above-stated precision is retained.

3D imprint technology using substrate voltage change

Spin-on-glass (SOG) is used as an electron beam (EB) resist whose depth is controlled by changing the EB acceleration voltage. Exposed SOG area and depth were developed with only 1 EB exposure using buffered hydrofluoric acid (BHF), yielding a three-dimensional (3D) SOG mold. Two acceleration voltage changes were used, i.e., changing the EB gun bias and changing the substrate voltage. When the EB gun bias was changed, patterned depth increased linearly by increasing with acceleration voltages and depth deviations of each acceleration voltages were within ±3%. The width resolution was 125nm on SOG using a 100nm EB diameter and the depth resolution was 10nm per 100V of acceleration voltage change. When the substrate voltage was changed, the relationship between the apparent acceleration voltage and pattern depth almost coincided with the change in EB gun bias, and the deviation in acceleration voltage was within ±4.4%. The depth gradation resolution limit was less than 10nm changing the substrate voltage. Imprinted patterns were transferred by pressing the fabricated 3D SOG mold onto photocurable resin at 0.5MPa and curing with a 1J/cm2 ultraviolet dose. Transferred patterns of photocurable resin were faithful and multigradational, corresponding to the mold pattern and attaining both 10nm mold depth resolution and 10nm transfer resolution.

A New Method to Measure Small Amounts of Solute Atoms on Planar Defects and Application to Inversion Domain Boundaries in Doped Zinc Oxide

We demonstrate the application of a new method of analytical transmission electron microscopy for measuring very accurately small amounts of solute atoms within a well-defined planar defect such as a stacking fault, grain boundary or an interface. The method is based on acquiring several spectra with different electron beam diameters from the same position centred on the defect. It can be applied to energy-dispersive X-ray microanalysis (EDXS) or electron energy-loss spectroscopy (EELS) and does not necessitate a scanning unit. The accuracy has been tested numerically under different conditions using simulations for a specific geometry and has been found to be substantially better than that of any other current standard technique. Our calculations suggest an extremely high accuracy theoretically achievable in the determination of e.g. the Gibbsian solute excess or the doping level of a grain boundary down to about ±1% of an effective monolayer, i.e. ±0.1 atoms/nm2 under typical experimental conditions. The method has been applied to zinc oxide, which forms inversion domain boundaries (IDBs) when doped with different transition metal oxides such as SnO2 or Sb2O3. We obtained an experimental precision of ±0.4 atoms/nm2, which has allowed us to solve the atomic structure of the IDBs.

Dresden EBIT: Trap properties and ion production studied by X-ray spectroscopy of helium-like argon

The Dresden EBIT – a compact room-temperature electron beam ion trap (EBIT) – is able to produce argon ions in all stages of ionization. From measured argon X-ray spectra we determine electron beam current densities of 176–514 Acm−2 at measured electron beam diameters of 145 and 102 μm, respectively. Direct excitation processes in Ar16+ are studied in dependence of additional helium and neon admixtures acting here as coolant. It is shown that the addition of different light gases leads to a specific increase of some ten percent of the argon X-ray output from the trap.

Grazing-exit electron probe X-ray microanalysis of ultra-thin films and single particles with high-angle resolution

Ultra-thin metallic films, artificial particles and aerosols were analyzed by grazing-exit electron probe X-ray microanalysis (GE-EPMA). It is beneficial to measure the characteristic intensity emitted from these specimens at very small exit angles. Such an approach reduces the high background originated from the substrate. The reduction of background X-rays is very important for thin film analysis. Therefore, it is shown that GE-EPMA is a powerful tool for localized analysis of thin-films and the surface of the sample. It is also found that GE-EPMA solves the problem of the overlapping of X-ray lines emitted from the substrate and the particle or the aerosol. Performing the measurement of the characteristic radiation from the particle under grazing exit conditions allows particles smaller than the electron beam diameter to be analyzed. Thus any limitations of the electron beam size can be removed.

CONTROL OF A SELF-SUSTAINED DISCHARGE BY RUNAWAY ELECTRON BEAMS

The method of generating runaway electron beams (EB) in a pulsed discharge in a middle-pressure gas suggested in [1] turns out to be very efficient for pumping of lasers on atomic and ionic transitions in rare gases and metal vapors [2]. Of interest is its practical use for control of the electric discharge parameters. This problem has not yet been considered, and this paper is aimed at filling this gap. In the present paper, we investigated the EB control of a self-sustained discharge to provide the voluminous and homogeneous mode of the discharge. Figure 1a shows the design of a discharge cell and of its power supply circuit. An electron gun contains an 11-stage cathode Kn and a common grid electrode E. A ceramic plate with 11 holes that determine the accelerating distance and the electron beam diameters (~8 mm) is inserted between them. Since the energy of electrons in the beam was several keV and the electron free path in a middle-pressure gas was several centimeters [3], the interelectrode gap of the controlled discharge was chosen equal to 1.8 cm. A grid anode of the gun E represents a steel plate 4 mm thick with 11 holes each 12 mm in diameter. A W-wire 50 μm in diameter was wound on the plate with a pitch of 220 μm. The grid facing the cathode serves as the anode of the gun, whereas the grid on the outside serves as the cathode of the main (controlled) discharge. The aluminum plate A serves as its anode. For this construction of the electrode E, the discharge in the accelerating gap of the gun and the main discharge were separated by a 4 mm space to reduce the negative effect of the main discharge on the EB generation [4]. Eleven capacitors Cn and also the capacitor C1 are charged from a high-voltage rectifier to the voltage U0 = 4– 12 kV through resistors Rn and R1. Capacitors Cn are used to supply power to the electron gun; the energy stored in C1 determines the energy deposited in the gas from the main discharge. The electron gun is actuated through the circuit comprising ground, thyratron T, capacitors Cn, accelerating gap, C2, and ground. Simultaneously, the capacitor C1 is discharged on C2 which then discharges through the cell. A delay between the electron beam current pulse and the current pulse of the main discharge is determined by the characteristic times of the charge exchange circuit and also by the discharge times of Kn–E and E–A gaps. The advantage of this circuit is that it operates with a single commutator T; its disadvantage is the impossibility of independent switching on of the electron gun and main discharge. Figure 1b shows the photograph of the main discharge in the cell. It can be seen that the geometry of the discharge is determined by electron beams, and glow of discharge columns is fairly homogeneous both in the longitudinal direction and in the column cross sections. The column diameter depended on the gas pressure and discharge current. On the electrode E, it was 8–12 mm. We failed to realize the volumetric main discharge mode under any conditions when the electron gun was switched off. Figure 2 shows the peak current of the volumetric main discharge versus the charge voltage U0 in the cell filled with helium. The discharge current pulse duration at half-maximum was 18–22 ns. The EB pulse parameters were measured when the main discharge was switched off. It was found that each element of the electron gun created a peak EB pulse current as high as 10 A for the initial gun voltage 8–9 kV at helium pressures indicated in Fig. 2. In this case, the EB current pulse duration was in the range 15–20 ns. However, when the main discharge was switched on, the voltage on the electron gun, equal to the potential difference between the cathodes Kn and electrode E, did not exceed half the initial applied voltage, and the EB current through each element did not exceed 5 A. The EB was generated simultaneously with the main discharge, and the values of the beam pulse and main discharge duration were almost identical, that is, the main discharge was completely controlled by the electron beam. Quantitatively, the efficiency θ of the

EELS analysis of PMMA at high spatial resolution

We performed the electron energy-loss spectroscopic (EELS) analysis of electron-sensitive polymers in the analytical transmission electron microscope in order to evaluate the possibility to obtain chemical information on polymers at a nanometre scale (i.e. at 2.4nm diameter probe). In the acquired spectra, we propose an identification of the ELNES fine structure to the different chemical bonding in agreement with molecular orbital calculations (EHT) and with previous XANES experiments.The main results confirm that poly(methyl mettacrylate) (PMMA) is very sensitive to electrons when a large probe size is used, with a critical dose of about 102Cm−2. However a high dose rate in a nanometre diameter electron beam is less destructive and the EELS spectra of far less degraded PMMA could be obtained even at 107Cm−2. Irradiation damage was however thought to be the main limitation of the field-emission gun microscope, since high electron doses are required to acquire an EELS spectrum. This surprising behaviour was already observed in the case of poly(ethylene terephthalate), which is however more resistant to the electron beam (Varlot et al., 1997. Ultramicroscopy 68 (2), 123–133). Several possible explanations were studied, such as the influence of the accelerating voltage, a wrong calculation of the electron dose, the excitation delocalisation and the electron dose rate.