The EVACTRON is a new device (US Patent 6,105,589) designed to remove hydrocarbons from SEM specimens and SEM chambers to prevent contamination artifacts. The device uses a lowpowered RF plasma to make oxygen radicals from air. These radicals oxidize hydrocarbons to make CO, H2O, and CO2 that can be removed by the vacuum pump. The use of air as an oxygen source is convenient to the SEM operator but limits the cleaning effectiveness of the system to easily oxidized carbon species. The system does remove high percentages of the usual contamination problems of vacuum pump oil and skin oil hydrocarbons from surfaces in a relatively short time.

The EVACTRON SEM-CLEAN™ system provides a safe method for plasma cleaning inside the SEM. Air as a diluted oxygen gas mixture is chosen as the plasma reactant gas. Air is passed through a low-power RF glow-discharge to create oxygen radicals inside a generator mounted on a specimen chamber port. The generator chamber is subject to the same vacuum as the specimen chamber, and the vacuum pressure and gas flow are controlled by the EVACTRON SEM-CLEAN controller. The radicals are carried out of the plasma into the whole of the specimen chamber by convection. in the chamber they react with all exposed surfaces including the specimen if present. The plasma itself is confined to the generator chamber, which prevents ion and electron bombardment damage to the instrument or specimen.

Secondary Ion Mass Spectrometry (SIMS) has unique features that makes this technique a very attractive candidate for sub-micron analysis: high elemental sensitivity (ppm or better) if reactive primary ions are used ; isotopic analysis capability ; possible insulator analysis ; no sample preparation needed ; small probed depth (a few run) and small interaction volume (cascade collision dia. < 10 nm).

Similar to a SEM microscope, a beam of Cs+ or O− primary ions is focused and rastered on the surface of the sample. The ions sputtered by this primary beam are collected, and this secondary ion beam is shaped in order to be mass filtered by a magnetic sector mass analyzer. Five mass selected images can be simultaneously recorded, originating from the exact same sputtered volume, ensuring reliable isotopic ratio and perfect image registration.

Based on Pr. Slodzian and al. [1, 2] original design, the commercial version of the NANOSIMS 50 ion microprobe incorporates several new developments in order to improve performances:

With the advent of field emission gun equipped SEMs and TEMs with clean vacuum systems, it has been shown repeatedly that the surface cleanliness of samples exposed to the small, high current density electron beam available in these instruments is critical with respect to controlling hydrocarbon contamination under the electron beam. in previous studies of dedicated plasma cleaners for TEM samples by one of the authors (SDW), it was shown that surface analytical techniques are the best methods for determining the effectiveness of such cleaners in removing hydrocarbons from the samples’ surfaces. in the current study, some of the experiments that were used in these previous studies were repeated using the XEI Evactron® system in order to compare the relative effectiveness of this system for surface cleaning of EM samples.

Figure 1 shows a picture of the sample introduction chamber of a VG Scientific ESCALAB Mkll system (XPS) with the Evactron® in place. The Evactron® system consists of an RF head unit that attaches to the vacuum chamber with an adaptable flange, an RF impedance matching unit, and an RF generator/vacuum controller. Although a 100% O2 system can be used and is recommended, for safety reasons, a 20%O2/80%Ar by volume gas mixture was selected because of the hydrocarbon oil used in the mechanical pump. Cleaning was also performed using laboratory air. The major difference between the dedicated plasma cleaners and the Evactron® configuration is that the sample is not immersed in the plasma but is exposed to the activated oxygen species in the downstream flow.

Very often, the performance of modern electron microscopes is limited by the poor quality of the laboratory rather than by the specifications of the microscope. Therefore, with microscopes aiming at subangstrom resolution microscopy and highest performing holography methods, special measures have to be taken to reach the needed low level of AC-stray fields, vibrations, and acoustic noise. in an ”organically“ grown institute, this is very difficult to realize, if possible at all.

• The Triebenberg Lab is newly built well outside the city of Dresden, in an area essentially free from disturbances, i.e. free from power lines, away from roads, trains and airplanes; fortunately, the site is shielded from wind by means of a dense belt of trees. The lab houses six microscope units each consisting of a microscope room, a room for peripheral devices (power supply, cooling units, computers, etc), and an office for up to three scientists. in view of the nearly ideal environment, the main challenge is to avoid any self-made disturbances which might arise by the operation of the lab. Therefore, without any compromise using the best art of engineering, we designed a special lab virtually free from disturbances. in the following, the main guidelines for the construction are sketched.

Surprising as it may seem, aberration correction for the scanning transmission electron microscope (STEM) is now a practical proposition. The first-ever commercial spherical aberration corrector for a STEM was delivered by Nion to IBM Research Center in June 2000, and other deliveries have taken place since or are imminent. At the same time, the development of corrector hardware and software is still proceeding at full speed, and our understanding of what are the most important factors for the successful operation of a corrector is deepening continuously.

Fig. 1 shows two high-angle dark field (HADF) images of [110] Si obtained with the IBM VG HB501 STEM operating at 120 kV, about 2 weeks after we fitted a quadrupole-octupole corrector into it. Fig. 1(a) shows the best HADF image that could be obtained with the corrector's quadrupoles on but its octupoles off. Sample structures were captured down to about 2.5 Å detail, easily possible in a STEM with a high resolution objective lens with a spherical aberration coefficient (Cs) of 1.3 mm. Fig. 1(b) shows a HADF image obtained after the Cs-correcting octupoles were turned on and the corrector tuned up. The resolution has now improved to 1.36 Å. This is sufficient to resolve the correct separation of the closely-spaced Si columns.

Sub-Ångstrom TEM to a resolution of 0.78Å has been demonstrated by the one-Ångstrom microscope (OÅM) project at the National Center for Electron Microscopy. The OÅM combines a modified CM300FEG-UT with computer software able to generate sub-Angstrom images from experimental image series.

Sub-Ångstrom HREM is gaining in importance as researchers design and build artificially-structured nanomaterials such as semiconductor devices, ceramic coatings, and nanomachines. Commonly, such nanostructures include atoms with bond lengths shorter in projection than the point resolution of a mid-voltage HREM. in addition, image simulations have shown that structure determinations of defects such as dislocation cores require sub-Angstrom resolution, as will hold true for grain boundaries and other interfaces.

Sub-Ångstrom microscopy with a transmission electron microscope requires meticulous attention to detail. As resolution is improved, resolution-limiting parameters need to be reduced. in particular, aberrations must be minimized, power supplies must be stabilized, and the microscope environment optimized to reduce acoustic and electromagnetic noise in addition to vibration.

One of the most striking problems in electron optics, the correction of resolution limiting aberrations by means of a corrector incorporated into the electron microscope column, has been solved during the last six years by demonstrating the improvement of resolution beyond the theoretical limit of the uncorrected Electron Microscope (EM). At first, in 1995 [1] with the correction of spherical and chromatic aberration of a dedicated Low Voltage Scanning Electron Microscope (LVSEM) and later, in 1997, with the correction of only spherical aberration of a commercially available 200 kV TEM [2]. The correction of spherical aberration of a dedicated Scanning Transmission Electron Microscope (STEM) at 100 keV primary energy has been demonstrated [3] and further improvements can be anticipated within the near future.

These achievements could only be obtained due to the emergence of new computer technology and especially CCD-cameras in the case of TEM correctors. These two developments made it possible first to calculate the electron optical components more precisely and hence, to achieve a better understanding of the requirements on the hardware and second, to have a better computer control of the electron microscope and the corrector itself. The combination of these two new technologies made it possible to go towards an automatisation of the alignment. This simplification of the alignment of an even more complex system is achieved by means of a proper combination of image acquisition and dedicated software in order to analyze and measure the aberrations of an electron optical system on one side and on the other to have appropriate tools to compensate these aberrations by computer controlled power supplies [4,5].

One of the limiting parameters in high-resolution transmission electron microscopy (HRTEM) are the high values for the spherical aberration Cs of the objective lens, which are the reason, that for TEM's with field-emitter the point resolution at Scherzer defocus is about two times lower than the information limit. Another effect of Cs-values of about one mm is the rather large disc of least confusion, which contributes to a high amount of contrast derealization. Furthermore the images are very sensitive towards beam tilt. These disadvantages contribute to a difficult interpretation of HRTEM-images especially at interfaces and defects. In-situ applications as well as diffraction contrast experiments require a gap of at least ten mm between the pole pieces. For the present TEM's this degrades the resolution at an acceleration voltage of 200 kV to approximately 0.27 nm. Cs-correction offers the ability to combine high resolution with a large space for the sample, which can be used for in-situ experiments.

Recently, we have constructed a lens system for Cs-correction based on hexapole lenses for a commercial 200 kV instrument equipped with a field emission gun. This system allows choosing the Cs-value between +2.0 mm and -0.05 mm. The objective lens of the microscope has a Cs value of 1.2 mm resulting in a point resolution of 0.24 nm.

In electron microscopy high-resolution imaging of finest object structures is generally hampered by the influence of aberrations of the lens system, especially the high spherical aberration of the objective lens. The delocalization of contrast details induced by aberrations is especially strong for microscopes equipped with a field emission gun providing a high spatial coherence. in recent years a prototype of an aberration correction system has been constructed by Haider et al., following a suggestion by Rose, consisting of two hexapole elements and four additional round lenses. The correction system was adapted to a Philips CM 200 FEG ST microscope with an information limit of 0.13 nm. The alignment is carried out using aberration measurements deduced from Zemlin tableaus. By appropriately exciting the hexapole elements it is possible to reduce or even fully compensate the spherical aberration of the objective lens.

With the freedom of a variable spherical aberration Cs new operation modes can be accessed that are not available in standard microscopes. with Cs = 0 and defocus Z = 0 pure amplitude contrast occurs, together with a vanishing contrast delocalization; phase contrast with a single, narrow pass-band up to the information limit can still be achieved by Z = ±7 nm, which introduces a delocalization of R = 0.13 nm. with Cs = 97 μm and Z = −18 nm the broad Scherzer pass-band for phase contrast can be extended to the information limit, with R = 0.35 nm. For the CM 200 Cs = 43 fim and Z = −12 nm still produces a high level of phase contrast, comparable with the extended Scherzer pass-band, but with R = 0.08 nm only. in the latter mode Scherzer’s defocus equals Lichte's defocus of least confusion.

Field emission electron microscopes operating at 200kV or 300kV and incorporating aberration correctors for either the incident electron probe or for the primary aberrations of the objective lens (OL) are currently under development for several laboratories in the world. OL-corrected instruments require monochromators for the electron beam, built into the electron gun prior to the accelerating stages, in order to optimize the contrast transfer characteristics of the objective lens to push the instrumental resolution limit to well beyond 0.1nm. This will allow the point resolution limit as controlled by the correction of spherical aberration Cs to potentially extend to the instrumental limit of better than 0.1nm. Figure 1 shows the contrast transfer characteristics of a Cs-corrected 200kV TEM, both without and with a beam monochromator.

Dedicated STEM instruments such as the 300kV VG-603 and lOOkV VG-501 at Oak Ridge National Laboratory, and other VG instruments at Cornell University and IBM Co. are also being adapted (by Nion Co., Kirkland, WA) to incorporate aberration correctors for the incident probe. The aim is to improve the resolution of the VG-603 instrument in dark-field imaging mode, for example, from 0.13nm to 0.05nm. in another ORNL project, the High Temperature Materials Laboratory has contracted JEOL Ltd. to construct a STEM-TEM instrument with a probe corrector designed and built by CEOS GmbH (Heidelberg, Germany).

A new type of electron energy loss spectrometer for use with monochromated microscopes is presented. The energy resolution of the spectrometer is better than 0.100 eV. A completely new electron optical design with a number of extra optical elements and advanced tuning software makes it possible to correct spectrum aberrations to 4th order, which increases sensitivity and collection angles. New high-stability electronics make it possible to maintain energy resolution over a period of several minutes in a practical laboratory environment.

The energy resolution of Transmission Electron Microscopes (TEMs) equipped with electron energy loss spectrometers is determined by a combination of the energy spread of the electron source, the stability of the microscope’s high voltage power supply, and the energy resolution of the spectrometer. Commercial microscopes usually employ electron sources with an energy distributions of around 0.5 eV or more (FWHM), limiting the energy ultimate energy resolution that can be achieved. Recently FEI constructed a special 200 kV TEM with a built-in monochromator which makes it possible to monochromize the electron source to better than 0.100 eV. A prototype of the presented spectrometer has been installed on this microscope.

TEM image acquisition is subject to limitations in resolution imposed by the physics of the position-sensitive detection process. in existing on-line digital cameras detection consists of scintillation, image transfer and photon detection by a CCD. Scintillator resolution is governed primarily by the incident beam energy. Material parameters affect the resolution only weakly. Some improvement can be made with a thin transmission scintillator. Still, backscatter of electrons from below can form a noisy background. Optical transfer resolution also has an important effect on overall camera system resolution. Leakage of light within scintillator, between fibers, at interfaces, and the inherent resolution loss due to binning into a finite pixel all contribute to optical resolution loss. This point spread function is narrower than the electron point spread function above 100kV but has an affect at all voltages. Long-range light leakage (psf tail) affects the background noise in diffraction work where there is often a 10e6 ratio between central peak and diffraction spot of interest. Thus, the number of resolved pixels in a TEM image detector is limited by the area of the detector, the effective thickness of the scintillator and the quality of the image transfer.

Gatan has developed several improvements which will address both electron and lightoptical resolution loss.

A unique in situ nanoindentation stage has been built and developed at the National Center for Electron Microscopy in Berkeley, CA. By using piezoceramic actuators to finely position a 3-sided, boron-doped diamond indenter, we are able to image in real time the nanoindentation induced deformation of thin films. Recent work has included the force-calibration of the indenter, using silicon cantilevers to establish a relationship between the voltage applied to the piezoactuators, the displacement of the diamond tip, and the force generated.

In this work, we present real time, in situ TEM observations of the plastic deformation of Al thin films grown on top of lithographically-prepared silicon substrates. The in situ nanoindentations require a unique sample geometry (see Figure 1) in which the indenter approaches the specimen normal to the electron beam. in order to meet this requirement, special wedge-shaped silicon samples were designed and microfabricated so that the tip of the wedge is sharp enough to be electron transparent.

Recently several different methods have been proposed to reconstruct the projected crystal potential from electron diffraction patterns. These methods envolve either diffraction patterns at many different orientations (as many orientations as beams in the pattern) and/or images, which makes their experimental realization difficult. We propose an entirely new method for reconstructing the projected crystal potential from fully dynamical [including absorption and multiple scattering effects to all orders] diffraction patterns from only a single crystal orientation and no image at all. Knowledge of the specimen thickness is not necessary. However, it requires diffraction patterns at many different accelarating voltages, which is a parameter that can easily be varied (within a certain range) in most modern electron microscopes. Since the intensities in the electron diffraction pattern are not affected by lens abberations this method is capable of reconstructing the projected potential with a resolution far better than that of any method using HRTEM images.

Two optimum defocus conditions are well known to users of high-resolution transmission electron microscopes. Scherzer defocus is useful in high-resolution electron microscopy (HREM) because it produces an image of the specimen “projected potential” to the resolution of the microscope. Lichte defocus is useful in electron holography because it optimizes sampling in frequency-space by minimizing the slope of the microscope objective lens phase change out to the highest spatial frequency in the hologram, consequently minimizing dispersion. For focal-series reconstruction, the requirement to maximize transfer into the image of high-frequency diffracted beam amplitudes leads to a third optimum defocus condition.

Image reconstruction methods allow the achievement of super-resolution - resolution beyond the native (Scherzer) resolution of the microscope - by correction of the phase changes introduced by the microscope objective lens. One such method is focal-series reconstruction, in which diffracted-beam information obtained at several different focus values is combined. to produce a valid super-resolution result, it is necessary to ensure that every spatial frequency is represented appropriately. Suitable choice of an optimum defocus produces optimum transfer of diffracted-beam amplitudes at any chosen spatial frequency.

We developed a 1-MV field-emission transmission electron microscope to help in further improving electron holography, Lorentz microscopy, and high-resolution electron microscopy. This microscope is characterized by an electron beam having the highest brightness ever, 2×1010 A/cm2, and by the highest lattice-resolution below 0.5 Å. These two features were attained by minimizing the mechanical vibration of the whole column and by improving the stability of both the electron beam and the high voltage. If the tiny electron source located at the top of the 7-m-high microscope moves by as little as a fraction of the source size, 50 Å in diameter relative to the column, due to mechanical vibration or beam deflection by the AC magnetic fields, the beam brightness will be greatly degraded. If the ripples ΔE of the high-voltage E exceed ΔE/E = 5 × 10−7 /min, then the inherent monochromatic feature of the beam is deteriorated by the increase in energy spread.

Through the preliminary experiments testing the vibration and magnetic shielding of the acceleration tube as well as the high stability of the high voltage, and through the numerical simulations on the vibration modes of the whole column, we were led to the conclusion that the microscope must be separated into three parts that are connected by cables.

Ultrahigh-vacuum transmission electron microscopy (UHVTEM) has become increasingly popular for the direct observation of nanostructures having clean surfaces, since industrial requirements to make and research nano-scale materials have been rapidly growing for quantum or nanoscale electronic devices. Since we have first developed high resolution UHVTEM in 1986, the UHVTEMs have been evolved with steady advances such as UHV compatible goniometer, field emission gun or etc. Furthermore, the UHVTEM started to combine analytical capabilities such as energy dispersive X-ray spectrometer, in-column type energy filter and etc., and to combine STM (scanning tunneling microscope). The UHV technology is essential for the analysis, because the portion of contaminant in a nano-scale specimen increases as the size of the specimen goes down. This paper reports the results of gold nanostructures by recently the developed UHVTEM.

Figure 1 shows recently developed UHVTEM with Schottky type field emission gun.

The attainable information in transmission electron microscopes is limited today primarily by chromatic aberrations, because the new generation of spherical aberration correctors has practically eliminated resolution limits caused by spherical aberration. If it would be possible to develop an electron source with 0.2 eV energy width, the limitations of chromatic aberrations could be minimized, and techniques like Electron Energy Loss Spectroscopy (EELS) could provide much more detail on the type of bonding between atoms. At present, the main approach to reduce the energy width of present day field emission electron sources is to incorporate a beam monochromator in the electron gun. The challenge for such a monochromator is not only to provide a small energy width of 0.2 eV or less, but also to preserve the brightness of the gun as much as possible at a level of current sufficient for high resolution imaging on a routine basis.

Over the last few years, several different designs for monochromators have been proposed. All designs are based on energy dispersive components which separate electrons of different energies spatially and eliminate the ones with high energy deviation by means of a selection slit.

The strong technological trend towards higher and higher integrated systems is resulting in considerably increasing demands on nanotechnique and -electronics. One of the most important sets of information is the characterization of the inner interfaces within such materials not only concerning the atomic structure but also concerning chemical composition and bonding and concerning local strain. New generation energy filtering TEMs using corrected in-column spectrometers and providing considerably increased power of resolution concerning imaging and analysis are thus essential in order to be able to understand, define, built and characterize such devices.

Within the SESAMe (Sub-eV-sub-Angstrom Microscope) [1] project, LEO and CEOS realize such a microscope for the MPI of Metal Research, Stuttgart in 3 steps. For step 1 and 2 an in second order completely corrected and in third order minimized 90° OMEGA energy filter has been developed. This filter is operating since one year in a 200 kV LEO EFTEM with LaB6 emitter (Step 1) [4]. Results will be presented by W. Sigle during this meeting. This filter will also be incorporated in the Sub-Angstrom-TEM (SATEM) with Cs corrector for the Research Center Jülich.

The potential utility of FIB for routine (and novel) applications has come to forefront recently due to advances in ion optics which now allow formation of focused ion probe of better than ∼10-20 nm containing current density exceeding several A/cm2, with a liquid metal source (typically Gallium). The small ion probe size, coupled with shallow sputtering depth - yet high sputtering yield of ions, has opened several opportunities in machining, lithography and ion-assisted deposition.[1-3] These developments, including automation, multi-specimen stages, cross-compatible specimen holders for FIB/TEM/SEM, use of in-situ electron beam (so-called dual beam), coupled with innovations such as the “lift-off” process[4], have provided an invaluable set of tools for microelectronic defect characterization. However, re-deposition (contamination), ion implantation/damage especially for desirable thinner sections (<∼50 nm) remain major concerns for further applications.

While much of the excitement in TEM community for FIB is due to thin foil specimen preparation (especially in microelectronics), we have been exploiting the site-specific micromachining aspect of FIB beyond specimen preparation for TEM, which is the focus of this contribution. Two broad themes will be presented: One exploits the site-specificity of FIB in making thin sections (including lift-off) at and across localized deformation as in indentation response of micro/nanocomposites. The other involves FIB as a fabrication tool for sputtering/drilling arbitrary shapes and sizes down to 20-50 nm, to enhance functional aspects.