Table of contents

The biennial conference of the Electron Microscopy & Analysis Group (EMAG) was this year held at Glasgow Caledonian University 5–7 September, whilst the EMAG Advanced School on 3–4 September was held at The University of Glasgow. The conference attracted 189 delegates from 21 countries. The conference focused on the dominant themes of Characterisation, Manipulation and Fabrication on the Nanoscale.

One of the motivations for running this conference series has been to encourage and develop the next generation of research scientists, to help maintain the UK's international profile in the areas of microscopy, analysis and innovation in micro- and nanotechnology. In this context, EMAG provided bursaries to cover the registration fees for 33 research students to help meet their costs of attending this event.

In addition to the 3 plenary lectures, there were 10 invited oral presentations and 53 contributed oral papers that ran in two parallel sessions. Furthermore, 95 posters were presented throughout the three days. These proceedings comprise 96 papers, beginning with a plenary paper, followed by the invited and contributed oral papers ordered chronologically by session as they appeared during the conference. The collated poster papers are then presented. The papers were submitted in advance of the conference, both electronically in Word and .pdf formats. Each paper was submitted to two referees for review. We are indebted to the efforts of the many delegates who kindly provided their valuable time to help in this process. Without their efforts it would not have been possible to produce these proceedings. We hope that readers of these proceedings will see this volume as a valuable snapshot of microscopy, microanalysis and nanoscale physics and technology in the UK at the time of writing.

In time honoured tradition, an Advanced School preceded the conference, with tutorial lectures on electron energy loss techniques to help research students gain a wider appreciation of the keynote scientific issues and to provide a background to the detailed conference themes. Our thanks to Maureen Makenzie and Pete Nellist for their sterling work running this event.

In addition, a Trade Exhibition was fully integrated into the conference site in the Glasgow Caledonian University Sports Hall, within close walking distance of the lecture theatres, giving delegates the opportunity to discuss recent developments in analytical instrumentation. In keeping with the previous EMAG-NANO 2005 conference at Leeds University, provision was made for commercial workshops for the promotion of products by the manufacturers and 'question and answer' sessions. The companies on show spanned the range of mainstream electron and scanning probe instrument makers, combined with a broad spectrum of smaller companies providing ancillary equipment, from services for sample preparation to vacuum system support. As ever, we are grateful to the exhibitors and sponsors for their valued contribution to this conference series.

Finally, we are extremely grateful for the many people who helped with the running of this conference. On behalf of the EMAG group we'd like to take the opportunity to thank the local organising committee of Ian MacLaren and Maureen Mackenzie. Our thanks also to Stephen Donnelly and Richard Baker for collating the scientific programme, to Stephen Donnelly for co-ordinating the award of student bursaries, and to Richard Baker and Guenter Moebus for their patient work guiding the editing of the proceedings. We'd also like to acknowledge the exceptional contribution of Jill Cowlard and Nicola Deedman of the CEM Group for co-ordinating the Trade Exhibition, and Claire Pantlin and her team at the Institute of Physics for their help keeping show on the road!

David McComb Imperial College London (EMAG Chair) Paul D Brown University of Nottingham (EMAG Proceedings Editor)

We describe some of the recent advances in the fluctuation microscopy technique for probing medium-range structural correlations in disordered materials. In particular we show that fluctuation microscopy is a surprisingly sensitive method for detecting trace quantities of C₆₀ in a disordered graphite matrix. This surprising sensitivity arises because C₆₀ does not have the same forbidden reflections as graphite. Modeling shows that the method should readily distinguish between C₆₀, C₇₀ and C₇₂. This result indicates that the technique can be used to discern dilute distributions of macromolecules in an otherwise disordered matrix.

We also describe preliminary interferometric fluctuation microscopy studies using cross-correlations in diffraction between coherent double probes. This is a form of holography where the diffraction patterns from two neighboring regions are allowed to overlap and interfere. Young's fringes appear wherever both regions scatter strongly. The cross-correlation can be examined as a function of probe separation to estimate a structure correlation length. At present, the method is being applied to x-ray and optical microscopies, but could also be applied to TEM. Since it isolates the essential four-body terms underpinning the fluctuation microscopy technique, this method holds much promise for studying medium-range order.

We present a TEM/STEM study of the structure and thermal stability of Co/AlO^x/Si, CoFe(B)/MgO/Si and FeNi/SiO^x/Si magnetic junctions deposited directly on patterned Si. In all three types of junctions the films are uniform and continuous. Annealing at 300 at 550 °C does not change their structure. CoFe and CoFeB layers remained amorphous after the annealing process.

As the SiO₂ or Si(O, N) layer in complementary metal oxide semiconductor (CMOS) devices reaches its atomic limits it is necessary to find suitable materials that will allow further device scaling. Leading candidate materials include hafnia and hafnium silicate; however, issues remain regarding their stability which must be resolved. This paper highlights some of these issues in hafnium-containing gate stacks. Hafnium silicate bulk powders were investigated to gain a better understanding of the chemistry and crystallisation processes involved. This was done using x-ray diffraction, thermal analysis, transmission electron microscopy and electron energy-loss spectroscopy.

The analysis of multiple EEL (electron energy-loss) spectra often requires normalisation of data to eliminate, for example, relative variations in thickness. In this case, data would only be reliable if it were normalised to a region in the spectrum, where the intensity varied with thickness alone. This raises the question of where an appropriate region is located, from which to extract the normalisation constants. The introduction of structure into the spectrum at certain thicknesses, such as energy lost via Cerenkov radiation emission, are clear regions to avoid. Identifying energy-loss contributions in the valence band region will provide further clarification for this process.

With the development of monochromators for transmission electron microscopes, valence electron energy-loss spectroscopy (VEELS) has become a powerful technique to study the band structure of materials with high spatial resolution. However, artefacts such as Cerenkov radiation and surface effects pose a limit for interpretation of the low-loss spectra; also the inelastic delocalisation restricts the spatial resolution on band gap mapping. For direct semiconductors, spectra acquired at thin regions can efficiently minimize the Cerenkov effects. Examples of h-GaN spectra acquired at different thickness showed that a correct band gap onset value can be obtained for sample thicknesses up to about 60 nm. For indirect semiconductors, the correct band gap onset can be obtained in the dark-field mode when the required momentum transfer for indirect transition is satisfied. At low energy-loss range the spatial resolution of this technique, which is mainly limited by the inelastic delocalisation, can be improved by dark-field VEELS at high collection angles.

In this paper we have collected multiple high angle annular dark field (HAADF) images of purified horse spleen ferritin mineral cores with an aberration corrected STEM (SuperSTEM). The images of individual cores have been separated into groups to determine characteristic views, and averaged to reduce the noise. We have found that these isolated cores do not have the same clear substructure as ferritin cores in plastic-embedded human liver sections (from a patient with haemochromatosis). There is a clear discrepancy between the in situ human liver and purified horse spleen ferritin, but also between Massover's observation of regular substructure in purified horse spleen ferritin by cryo-electron microscopy [1] and these results. A possible basis for this discrepancy is the difference in purification methods used between the two horse spleen ferritin preparations, with the one here purified using CdSO₄ precipitation, and the one used by Massover purified without the use of CdSO₄.

Single-particle analysis of short-term aerosol samplings can provide information on the rapid evolution of size distribution and chemical composition of pollution aerosols, notably in source areas where reactive compounds are present. The aim of this work is to evaluate the capability of automated particle analysis performed by SEM-EDX to describe such rapid evolutions. Two sampling campaigns were performed at a highly urbanised and industrialised coastal site. The first one corresponded to low atmospheric particle loads and low temporal variation of particle mass concentrations. In these conditions, the lower analysed particle number to yield representative results was 1 000 particles per impaction stage. During the second one, a pollution event with a significant increase in particle mass concentrations was recorded. The ability of automated SEM-EDX to describe short temporal variation in particle chemical composition was demonstrated here.

Studies of electron beam induced damage in synthetic HA powder produced by standard solution-precipitation are reported. Under intense irradiation, such as that produced by a focused field emission source used when analysing a grain boundary or heterophase interface, it was found that the material damages transforming to cubic calcium oxide. The damage process was quantitatively monitored as a function of total fluence, fluence rate, accelerating voltage and specimen temperature by measuring the Ca/P ratio determined by EDX or EELS. Conclusions as to the exact damage and also recovery mechanisms in operation are drawn and the conditions required for limited-damage measurement are derived.

The use of FIB milling as a TEM sample preparation tool for mineralised dentin was investigated in order to gain a better understating of the nanostructure of bone-like specimens. A clear advantage of FIB milling over ultramicrotomy, the traditional preparation route in biological systems, is that dehydration, embedding and section flotation can be obviated, thus reducing both physical and chemical damage to the specimen prior to examination. The characteristic periodic contrast of collagen fibrils is clearly visible in FIB sections without the need for any chemical staining. The nature of the organic/inorganic interface was studied using EELS and EFTEM mapping in a state-of-the-art monochromated FEG(S)TEM.

1D functional oxides at nm-scale are interesting for fundamental reasons and promising for future applications. Here, ferroelectric PbTiO₃ nanorods, produced through a hydrothermal process, have been studied in detail by transmission electron microscopy. The length (up to one μm) and the diameter (30–100 nm) as well as the growth direction ([001]) of the nanorods could easily be determined using conventional imaging and electron diffraction techniques. However, variations along the length of the rods were clearly visible in the bright field images. Steps on the outer surfaces of the rods could be identified using energy filtered transmission electron microscopy and spectrum imaging thickness maps. The thickness variation parallel to the electron beam affected the bright field contrast and energy dispersive spectroscopy of the nanorods. From cross-sectional specimens, it was determined that the outer surfaces of the rods were dominantly {110} type, leading to a rectangular cross-section. The cross section diameter of the rods was reduced by the introduction of {100} surfaces. In addition, the cross-sectioned specimen revealed the presence of internal channels in the growth direction, especially in the bottom part of the rods. Such a detailed structural description of the nanorods was necessary to study the possible ferroelectric domain structure and to reveal the growth mechanism of the rods.

Ferroelectric lead zirconate titanate (PZT) thin films have been analysed using electron backscatter diffraction (EBSD). Grain orientation mapping has been demonstrated, showing that features smaller than 100 nm may be successfully indexed. In conjunction with piezoresponse force microscopy (PFM), which was used to map and quantify the piezoelectric response from the same region of the films with a resolution of 10 nm, an analysis of the effects of grain orientation on the measured response at the nanoscale was possible. The microtexture of the film showed the presence of both mono- and multi-domains within grains exhibiting sizes of hundreds of nanometres.

Reliable automated orientation mapping of 90° domains in a tetragonal perovskite has been achieved for the first time using both EBSD and TEM-Kikuchi pattern analysis. This has been used to compare local measurements of c/a ratios in PZT with global measurements by X-ray diffraction. The local c/a rations are in broad agreement with the global measurements, but further work is needed to determine whether the small discrepancies are real local variations or are caused by experimental factors.

After previous work producing a successful 3D tomographic reconstruction of dislocations in GaN from conventional weak-beam dark-field (WBDF) images, we have reconstructed a cascade of dislocations in deformed and annealed silicon to a comparable standard using the more experimentally straightforward technique of STEM annular dark-field imaging (STEM ADF). In this mode, image contrast was much more consistent over the specimen tilt range than in conventional weak-beam dark-field imaging. Automatic acquisition software could thus restore the correct dislocation array to the field of view at each tilt angle, though manual focusing was still required. Reconstruction was carried out by sequential iterative reconstruction technique using FEI's Inspect3D software. Dislocations were distributed non-uniformly along cascades, with sparse areas between denser clumps in which individual dislocations of in-plane image width 24 nm could be distinguished in images and reconstruction. Denser areas showed more complicated stacking-fault contrast, hampering tomographic reconstruction. The general three-dimensional form of the denser areas was reproduced well, showing the dislocation array to be planar and not parallel to the foil surfaces.

Platinum 11 at.% vanadium undergoes an ordering transformation to form a Pt₈V superlattice. Deformation of the alloy results in the formation of single dislocations or superdislocations, depending on the degree of order. The superdislocations take the form of triple dislocations with a separation of 20 nm in the fully ordered alloy.

Transmission electron microscopy (TEM) has been used to characterize NaAlH₄ doped with Ti, after H cycling. NaAlH₄ was shown to be highly unstable under the electron beam, and 'knock on' damage lead to a decomposition of NaAlH₄ with Na and H evaporating from the sample. All Ti containing phases were stable under the electron beam. After H cycling, the Ti was present as a mixture of amorphous and crystalline Al_1-xTi_x.

The three-dimensional (3D) morphology of two related classes of CeO₂ nanostructures is reconstructed using electron tomography with three image acquisition modes for comparison. The samples include free standing nanoparticles of ∼ 40nm diameter with octahedral morphology and nanoscale precipitates of 100–200 nm diameter embedded in glass with dendritic morphology, both of fluorite-type CeO₂. Tomograms were successfully constructed from (i) bright field (BF) TEM, (ii) a novel single-window energy-filtered TEM (EFTEM), and (iii) annular dark field STEM (ADF) modes. Advantages and disadvantages of the imaging modes are summarized.

A lensless holographic in-line point source microscope was envisioned more than half a century ago, but its realization with electron waves has come short due to not only difficulties inherent in Fresnel-type reconstruction methods, but also to the lack of an adequate (spatially and temporally coherent) point source. With the recent creation of ultrasharp nanotips, which can field emit electrons from a single atom at their apex, an extremely coherent electron source is available that provides a great boost to the holographic method. The spatial coherence of such nanotips is a few Å, while their temporal coherence is characterized by a value of energy dispersion (FWHM) as low as 0.1 eV. In this work we ascertain the use of such a microscope in the imaging of nanoscale structures and interfaces. The method is suitable for two- and three- dimensional imaging of solid nanoparticles, thin crystals, and surfaces, but also for biological entities. We show how improvements in the reconstruction method can be made by applying the rigorous Fresnel-Kirchhoff diffraction theory adapted to Electron Optics. Sub-nanometer resolution is achievable for beam energy between 100–200 eV.

As a microanalytical technique, electron energy-loss (EEL) spectroscopy performed within a scanning transmission electron microscope (STEM) represents an outstanding method. The technique has led to remarkable success in spectroscopic chemical analysis. However, the question remains of whether the EEL signal of individual dopant atoms within a substrate can be detected, with simultaneous atomic-scale imaging of the microstructural surround. In the present contribution, results are presented of ultra-high spatially resolved EEL spectroscopy of ion implanted multi-walled carbon nanotubes (MWCNTs), performed within a C_s-corrected dedicated STEM. Quantifiable EEL signals with an absence of oversampling, suggest single implanted atoms were detected at certain locations.

The mechanical and microstructural properties of steels can be affected by dopant elements or the presence of impurities. Aluminium has historically been added to steels principally as a de-oxidant but it has also been shown to have a significant effect on the mechanical properties. It is well known that aluminium usually combines with nitrogen to form A1N precipitates at high temperature mostly in austenite (dependent on the relative concentrations of Al and N). Aluminium therefore acts as a grain refining element. A series of low carbon, low nitrogen steels have been prepared with varying additions of aluminium up to 1 wt%. These changes in properties of the steel cannot be attributed to the presence of aluminium nitride and hence it important to understand the specific role and mechanism which the aluminium plays. The location, concentration and chemistry of the aluminium have been investigated using SEM and TEM. The results show clear evidence for segregation of aluminium to the grain boundaries.

To investigate mechanical properties of different crystal orientations in titanium alpha and beta phases, nanoindentation was performed on two samples; commercially pure titanium (predominantly α phase) and a 15V-3Cr-3Sn-3Al (beta) alloy. This allowed the Young's Modulus and the hardness to be measured for individual grains, with a total of 100 measurements being made on each sample. The crystal orientations around the indents were mapped using Electron Backscattered diffraction (EBSD) in the scanning electron microscope. The mechanical property data was linked to the orientations of the grains at the points of the indents. This shows that for alpha titanium, hardness is strongly dependent on orientation, with the highest hardness being seen where indentation direction is normal to the basal plane. This is in agreement with work by other authors working on polycrystalline Ti-6A1-4V, although is in contrast to work carried out on pure titanium single crystals. It was also noted that some misorientation was seen in the region of the indent, with misorientations as high as 3° being seen over a distance of approximately 5 μm from the centre of the indent. Further work to examine dislocation substructure and misorientation beneath the indent is discussed.

The sequence of precipitation in five different alkali-borosilicate glasses is studied by HREM, HAADF-STEM and EELS. The glasses are doped or supersaturated with 4 mol% of CeO₂, and have varying concentrations of other dopant oxides, such as Fe₂O₃, Nd₂O₃, Cr₂O₃, and Ag₂O. The shape and distribution of precipitates range from amorphous droplet shape to dendritic crystals. Oxidation states of Ce and other cations in glass and precipitates are measured by ELNES. It is found that Ce in glass preferentially adopts +IV valence in Ce-crystal-free glasses while it adopts +III valence in glass nanocomposites that contain Ce(+IV)O₂ nanocrystals.

There is a major drive to introduce γ-TiAl into gas turbine engines in order to reduce weight. However, this will require the development of coatings that protect against oxidation at high temperature, but do not adversely affect the mechanical properties. This work reports the high temperature degradation mechanisms of a nanoscale CrAlYN/CrN multilayer coating deposited on γ-TiAl(8Nb) by a combined high power impulse magnetron sputtering / unbalanced magnetron sputtering. Detailed TEM/STEM of FIB prepared specimens from isothermal static oxidation tests at 850°C for up to 1030 hours is presented. The evolution of the complex oxide structure and the implications for future coating development is discussed.

The structure of amorphous and crystalline SiGe nanoparticles, embedded in a dielectric medium, SiO₂, and its stability under 'in situ' electron beam irradiation is reported. High-resolution transmission electron microscopy and electron-diffraction pattern simulation by fast Fourier transform was used to analyze the crystal structure of the SiGe nanoparticles. Electron beam irradiation induces structural alternate order-disorder transitions in the nanoparticles for irradiation effects are mainly associated to the density of current. For irradiation with current densities < 7 A.cm-^-2 no effects are observed in the as-deposited amorphous samples, whereas in the crystallized samples, SiGe nanocrystals show higher stability and no effects are observed for irradiation densities of current < 50 A-cm^-2. Irradiation with densities of current greater than these thresholds cause consecutive amorphous-crystalline or crystalline-amorphous structure transitions respectively for both amorphous and crystallized nanoparticles. A hexagonal structure is proposed for those nanocrystals obtained after irradiation in the as deposited amorphous samples.

Two examples of electron beam induced deposition (EBID) are described in terms of the fabrication of metal nanostructures with transmission and scanning electron microscopy (TEM and SEM). Both need the reaction of gas molecules, W(CO)₆ and Fe(CO)₅, near substrate surfaces, which is introduced by a nozzle from outside. For the first example, electric conductive substrates were used to deposit metal nanostructures right on the beam irradiated area, and self-standing 2D and 3D structures were grown in the range of diameters of 3 to 100 nm. In case of iron deposit, magnetic properties were analyzed with electron holography. For the second example, insulator substrates were used to cause the electron charge-up on the surface. Due to the fractal distribution of electrons, metal deposits grew dendrite-like nano-trees with many branches of 2 to 3 nm width to 10 to 100 nm in length.

With electron beam induced deposition structures as small as 1 nm can be deposited. The nucleation stage of the growth of such deposits is studied in dependence of the substrate material and substrate temperature. On amorphous carbon foils larger deposit sizes are obtained at higher substrate temperatures during growth. On SiN membranes unusually high growth rates are observed which seem to relate to charging effects.

GaN nanowires (NWs) have been grown on Si(111) substrates byt plasma-assisted moleucular bearn epitaxy. The Morphology and optical properties of the NWs are influenced by te growth parameters as investigated by the scanning electron microscope. The nucleation process of GaN-NWs is explained in terms of nucleation density and wire evolution with time. The wire length in the nucleation stage shows a linear time dependence. The wire density increases rapidly with time and then it saturates. We explain GaN-NWs growth by making use of the diffusion-induced (D-I) mechanism that explains the dependence of the length on wire diameter for a deposition time longer than the nucleation stage.

Film formation of ultrathin polymers on microheterogeneous surfaces is strongly influenced by molecular surface patterns which cause local wettability differences for liquid phases in contact with the surface. Surface coverage with polymers transferred by dip-coating from polymer solutions is controlled by surface heterogeneities prepared by softlithography or by electron beam lithography of self-assembled monolayers. Using crystallisable polymers for ultrathin film formation (polyethyleneoxide (PEO) we could carefully investigate the competition in pattern formation resulting from the dewetting process and patterns which result from lamella crystallization of PEO in ultrathin films. For the first time we could observe that ultrathin polymer films of crystallisable polymers which are prepared in absence of any substrate surface defects form a metastable state in which they can exist over days and weeks without crystallization. Heterogenous nucleation in these metastable films can be done by external stresses such as contact with an AFM tip. The nucleation on demand allowed us to study the diffusion controlled pattern formation that is observed during lamella crystallization and the growth process resulting in different morphological features could be studied at elevated temperatures and in lateral confined areas which were realized by the film preparation on the microheterogenized surfaces.

Transmission electron microscope (TEM) specimens are today routinely prepared using focussed ion beam (FIB) instruments. Specifically, the lift-out method has become an increasingly popular technique and involves removing thin cross-sections from site-specific locations and transferring them to a TEM grid. This lift-out process can either be performed ex situ or in situ. The latter is mainly carried out on combined dual-beam FIB and scanning electron microscope (SEM) systems whereas conventional single-beam instruments often are limited to the traditional ex situ method. It is nevertheless desirable to enhance the capabilities of existing single-beam instruments to allow for in situ lift-out preparation to be performed since this technique offers a number of advantages over the older ex situ method. A single-beam FIB instrument was therefore modified to incorporate an in situ micromanipulator fitted with a tungsten needle, which can be attached to a cut-out FIB section using ion beam induced platinum deposition. This article addresses the issues of using an ion beam to monitor the in situ manipulation process as well as approaches that can be used to create stronger platinum welds between two objects, and finally, views on how to limit the extent of ion beam damage to the specimen surface.

Materials Scientists need information on the kinetics of the microstructural evolution processes that determine the finished microstructure, and hence the properties, of any material. E.g. recrystallisation, grain growth and phase changes. Such kinetic information requires reliable discrimination of differently oriented crystallites and/or different crystal phases coupled with useful spatial resolution and temporal resolution (i.e. high frame rates). These imaging results must be realised from a hot and changing specimen, in an instrument that is compatible with that hot specimen and with a practical specimen heater. Focused Ion Beams (FIB) offer strong contrast between crystallites and phases, and hence offer the ability to discriminate between these features even while imaging at fast frame rates, however their compatibility with hot specimens was unproven. Here we report results from a novel combination of FIB with an in-situ heating stage, to produce in-situ, real-time microstructural imaging from a variety of metallic specimens at temperatures up to about 700°C. FIB has additional capabilities, such as site-specific milling, which we show remains practical with hot specimens and which can be very useful with certain materials.

The Internet has enabled researchers to communicate over vast geographical distances, sharing ideas and documents. e-Science, underpinned by Grid [1] and Web Services, has enabled electronic communications to the next level where, in addition to document sharing, researchers can increasingly control high precision scientific instruments over the network. The Oxford CyberSEM project developed a simple Java applet via which samples placed in a JEOL 5510LV Scanning Electron Microscope (SEM) can be manipulated and examined collaboratively over the Internet. Designed with schoolchildren in mind, CyberSEM does not require any additional hardware or software other than a generic Java-enabled web browser. This paper reflects on both the technical and social challenges in designing real-time systems for controlling scientific equipments in collaborative environments. Furthermore, it proposes potential deployment beyond the classroom setting.

Gold nanowires were successfully fabricated by a DC electrodeposition technique into Anodic Aluminium Oxide (AAO) templates. The microstructure of 55nm gold nanowires released from AAO templates was observed by SEM and TEM to be polycrystalline, with a bamboo-type structure and grain sizes 20nm to several micrometers. Individual gold nanowires were picked up from bundles of gold nanowires using a super-sharp W tip attached to an in-situ Kleindiek nanomanipulator fitted in a SEM-FIB. The picked-up gold nanowires were then deposited onto a silicon wafer, or connected between two nanomanipulator tips, to fabricate single nanowire nano-circuits for electrical testing. The electrical properties of single manipulated nanowires are compared to that of bundles of gold nanowires for the two circuit types. The lowest resistance is achieved by connecting the gold nanowires between two FIB-milled tungsten tips.

Room temperature electrical transport measurements have been made on suspended multi-walled carbon nanotubes (MWNTs) using a remote controlled manipulation system within a scanning electron microscope. It is shown that the current-voltage characteristics of the MWNTs are symmetric with respect to voltage and that the conductance improves with multiple cycling of the voltage. Estimations of the semiconducting sub-bands and the contact transmission coefficients of the MWNTs have also been made.

To study the mechanisms of dopant contrast in secondary electron (SE) imaging in the SEM, we have measured the image widths of a series of thin p-doped layers in Si, from 1 nm upwards. We have used computer modelling to simulate the effects of surface band-bending due to a realistic density of surface states on the Si, and we have also calculated the magnitude of the external patch fields. We have found a good correlation between the intensity widths and slopes of experimentally measured SE images of thin p-doped layers and the calculated widths and slopes of the energy distributions across these layers at a depth of 5–10 nm below the surface. This is consistent with the mean escape depth of SEs in Si being about 7 nm. We conclude that doping contrast in the SEM is mainly a function of bulk built-in voltages modified by surface band-bending effects within about 5–10 nm of the surface.

Aberration correction leads to a substantial improvement in the interpretable resolution of Transmission Electron Microscopes. Electron optical correctors based on two strong hexapole elements linked through a round lens transfer doublet enables direct correction of all axial aberrations to third order. Subsequent, indirect computational analysis of a focal or tilt series of images offers the possibility of further compensation of the axial aberrations to fifth order. This paper describes 1st and 2nd generation aberration corrected instrumentation installed in Oxford and also the use of combinations of direct and indirect correction / compensation in a variety of different geometries to achieve specimen exit plane wavefunctions containing directly interpretable structural information significantly below 0.1 nm.

The principle of the ptychographical method is to collect a number of high-angle Fraunhofer diffraction patterns from an object as it is moved relative to a substantially localised illuminating beam. The many diffraction patterns are then used to solve the phase problem in order to synthesise a computational lens than can achieve much higher resolution than any of the lens or aperture components used to generate the illumination function. Although ptychography has now been demonstrated using visible and hard X-ray photons, there remain many unanswered questions relating to the scattering approximations involved in ptychography. We examine the breakdown of the kinematical and 2D approximations.

A reduction in the focal depth of field as a result of the installation of aberration correctors in scanning transmission electron microscopy, allows three-dimensional information to be retrieved by optical depth sectioning. A three-dimensional representation of the specimen is achieved by recording a series of images over a range of focal values. Optical depth sectioning in zone-axis crystals is explored computationally using a Bloch wave analysis to explain the form of the electron intensity in the crystal as a function of depth. We find that the intensity maximum deviates from that of the expected defocus value due to pre-focusing by the atomic column and also due to channelling pendellosung. The possibility of performing bright-field imaging in a double corrected two lens system in a confocal arrangement is also investigated computationally. The method offers some advantages over depth sectioning using conventional transmission electron microscopy.

The continued development of electron probe aberration correctors for scanning transmission electron microscopy has enabled finer electron probes, allowing atomic resolution column-by-column electron energy loss spectroscopy. Finer electron probes have also led to a decrease in the probe depth of focus, facilitating optical slicing or depth sectioning of samples. The inclusion of post specimen aberration corrected image forming lenses allows for scanning confocal electron microscopy with further improved depth resolution and selectivity. We show that in both scanning transmission electron microscopy and scanning confocal electron microscopy geometries, by performing a three dimensional raster scan through a specimen and detecting electrons scattered with a characteristic energy loss, it will be possible to determine the location of isolated impurities embedded within the bulk.

The density functional theory (DFT) code WJEN2k is one of the principle codes used for the simulation of electron energy loss spectra. As DFT codes scale poorly with increasing complexity of calculation, it is important to choose input parameters in such a way as to minimise computational time and produce spectra reliable enough to interpret experimental data. Using graphite as an example, we discuss the effects of these input parameters on the timing of the calculation and the spectrum produced.

Magnetic ordering has been shown to have a significant effect on the shape of the oxygen K-edge energy-loss near edge structure (ELNES) in a series of chromite spinels. However, the ELNES of these materials has only been simulated using a rough approximation of antiferromagnetism - the true nature of the magnetic interactions responsible for the detailed oxygen K-edge shape is still unknown. Chromite spinels typically undergo transitions to long range ordered antiferromagnetic (AFM) structures at temperatures below ∼15K. Dynamic short range magnetic order (SRO) has been observed at temperatures up to 150K using neutron powder diffraction (NPD). It is not clear whether long range magnetic order, short range magnetic order or paramagnetic effects are responsible for the characteristic oxygen K-edge ELNES observed at room temperature. Here we discuss the possibility of carrying out paramagnetic simulations using the real space multiple scattering program FEFF8.2, and show preliminary results of paramagnetic simulations of the oxygen K-edge ELNES of magnesium chromite.

In this paper, the benefits of processing Energy-Filtered TEM (EFTEM) datasets with Multivariate Statistical Analysis will be described. These include: Isolation of the main sources of information, identification of features masked by noise and/or background and, more importantly, almost 'noise-free' data reconstruction.

In this paper we present considerations for experimental recovery of electron phase using the TIE algorithm as applied to TEM images. Realistic simulations have considerably aided interpretation of results. Experimental reconstructions show that low spatial frequency artefacts in the phase reconstruction, resulting from very small variations in the detected signal, can be reduced by using images with smaller numbers of pixels.

Aberration correction can be achieved using either direct electron optical correction or indirect image restoration. In the past focal series restoration has been applied to aberration corrected images in order to improve the quality of aberration correction and retrieve the complex specimen exit wavefunction. Image restoration can also be performed from a number of images with differing illumination tilts (a tilt series) instead of the more common focal series datasets where only the defocus value is varied. Here we apply tilt series image restoration to aberration corrected images and discuss the advantages of this approach. Preliminary results demonstrate the potential of this technique to provide interpretable structural information at resolutions beyond the axial information limit of the microscope.

In Sudan, X-rays are routinely used at least once for measurements of pelvis during the gestation period, though this is highly prohibited worldwide, except for a few life threatening cases. To demonstrate the effect of diagnostic ionizing radiation on uterus, fetus and neighboring tissues to the ovaries, two independent experiments on pregnant rabbits were conducted. The first experiment was a proof of concept that diagnostic ionizing radiation is hazardous throughout the gestation period. The second experiment was done through Transmission Electron Microscopy (TEM) to elucidate the morphological changes in the ultra structure of samples taken from irradiated pregnant rabbits. This study uses TEM to test the effect of diagnostic radiation of less than 0.6 Gray on the cellular level. Morphological changes have been captured and the images were analyzed to quantify these effects.

Interest in alumina-on-alumina total hip replacements (THR) continues to grow for the young and active patient due to their superior wear performance and biocompatibility compared to the alternative traditional polymer/metal prostheses. While alumina on alumina bearings offer an excellent solution, a region of high wear, known as stripe wear, is commonly observed on retrieved alumina hip components that poses concern. These in-vivo stripe wear mechanisms can be replicated in vitro by the introduction of micro-separation during the simulated walking cycle in hip joint simulation. However, the understanding of the mechanisms behind the stripe wear processes is relatively poor. 3D topographic reconstructions of titled SEM stereo pairs from different zones have been obtained to determine the local worn surface topography. Focused ion beam (FIB) microscopy was applied to examine the subsurface damage across the stripe wear. The paper presents novel images of sub-surface microcracks in alumina along with 3D reconstructions of the worn ceramic surfaces and a classification of four distinct wear zones following microseparation in hip prostheses.

Preliminary EELS analyses of range of rare Mn-bearing ore minerals from the N'chwaning II mine in the Kalahari Manganese field have been undertaken. The EEL data reveal a range of Mn oxidation states which reflect the oxidation changes due to post-sedimentary and early metamorphic changes. Further detailed investigation and quantification of the various Mn-valences, will be used to track oxidation changes within these geochemical significant deposits.

Possibilities of 'in-situ' observation of non-conductive biological samples free of charging artefacts in dynamically changed surrounding conditions are the topic of this work. The observed biological sample, the tongue of a rat, was placed on a cooled Peltier stage. We studied the visibility of topographical structure depending on transition between liquid and gas state of water in the specimen chamber of VP SEM.

This paper discusses the use of a focused ion beam for the preparation of cross-sectional transmission electron microscopy specimens from small samples, typically 150 micron in diameter, that have been recovered from a laser-heated diamond anvil cell. We present preliminary observations of iron melted with helium as the pressure-transmitting medium to pressures near 4 GPa. The project is designed to investigate entrapment mechanisms of the inert gases in metals and silicates at high pressure and temperatures.

Defects in the active layers of MOCVD-grown InGaN/GaN blue-emitting multiple quantum well LED heterostructures on (111) silicon substrates are examined by conventional and high resolution transmission electron microscopy (HRTEM). The quantum wells contain small gaps, interfacial steps, sphalerite stacking and well thickness variations. No characteristic phase separation of the InGaN into In-rich clusters is observed within the wells. Despite the defective active layers and the absence of In clusters, the internal quantum efficiency of the device quantum wells is 23%, hence mechanisms by which carriers are localised in the quantum wells via the defects are considered. The roles of well gaps, thickness variations, stacking faults and sphalerite grains in exciton confinement are discussed.

In this paper we report initial results from structural and optical investigations on bilayers of InAs/GaAs quantum dots (QDs) grown by molecular beam epitaxy with different GaAs spacer layer thicknesses. We have used bright field imaging in transmission electron microscopy and high angle annular dark-field imaging in scanning transmission electron microscopy to study strain and compositional correlations in these QD structures. Our goal is to utilise valence electron energy-loss spectroscopy (VEELS) to probe the bandgap and optical properties of the quantum dots. We discuss the influence of Cerenkov losses on VEELS and approaches to mediating their effect on bandgap measurements.

Defects in Ruddlesden-Popper phase CaO·[(CaMnO₃)]₄ epitaxial films grown on SrTiO₃ (001) by pulsed laser deposition have been investigated using high angle annular dark field imaging in an aberration-corrected STEM. The stacking faults perpendicular and parallel to the substrate formed during the growth are discussed in detail. The desired n = 4 RP phase is imaged and chemically analyzed at the atomic scale using electron energy loss spectroscopy.

La_{0. 7}Ca_{0. 3}MnO₃ (LCMO) films grown on (110) surfaces of various (pseudo-) cubic substrates, SrTiO₃ (STO), LaAlO₃ (LAO) and La_0.3Sr_0.7Al_0.5Ta_0.35O₃ (LSAT), were studied by means of transmission electron microscopy (TEM). (110) LSAT substrate and LCMO film have the smallest lattice mismatch. (110) LAO substrate induces compressive strain while (110) STO induces tensile strain in the LCMO film. In all cases, the (010)_° plane (long axis) of LCMO is parallel to the substrate surface. In the case of LSAT, the [001]_° is parallel to [011] LSAT while the films on the other 2 substrates have [100]_°// to the [011] substrates. Films grown on STO and LAO show an easy magnetisation when the applied magnetic field is in plane along the long-axis of the LCMO. In the case of LSAT, there is no in-plane anisotropy. The films on STO and LAO have twins as the dominant defects. The film on LSAT has less twining. High resolution TEM shows that LCMO on (110) LSAT (the least lattice mismatch) possesses a perfect interface with no misfit dislocations in the area of examination. The LCMO films on (110) LAO and (110) STO have nearly perfect interfaces with atomic sharpness but with misfit dislocations. The different in-plane crystal directions and strain distributions in the films may account for the different magnetisation behaviours of the LCMO films on the different (011) substrates.

Work elsewhere has shown that the hardnesses of coherent In_xGa_1-xAs multilayer structures in which the misfit stresses are controlled by varying the indium content in each layer are influenced by the thicknesses and the coherency strains in the layers. These results have been interpreted in terms of the length-scale for flow being greater than the layer thickness and associated with deformation size effects. However, nanoindentation of In_xGa_1-x As suggests an influence of the flow stress of the individual layers on the overall multilayer flow stress. Consequently, initiation of flow in the weaker layer is expected. A deformed multilayer has been characterized by measurement of the lattice rotations measured from energy-filtered convergent beam electron diffraction (CBED) patterns recorded in scanning transmission electron microscopy (STEM) mode. Based on the movement of Kikuchi lines as a result of rotations of the lattice the local orientation of the crystal can be extracted, allowing the local orientations to be estimated. The small probe-size in CBED of ∼1 nm used here ensures the region sampled is smaller than the thickness of the individual layers. These measurements have been used to construct a map of the axes of rotation in the lattice which demonstrates the ability to distinguish the individual layers.

High quality oxides layers are now available for MOSFETs on GaAs. For successful devices, suitable process schemes are required. In this paper we show an investigation of an etching process on a GaAs/Ga₂O₃/GGO dielectric gate stack. This investigation has been carried out using EFTEM and EELS SI. EFTEM provides a quick analysis on the structure while EELS SI offers much better resolution and the possibility to quantitatively characterize the material.

Grain contrast imaging can be performed with several techniques. In order to be able to choose the most suitable one it is important to know which techniques are available and also to be aware of the strengths and weaknesses of each technique. In this work, the grain contrast imaging is performed with secondary electrons, backscattered electrons, forward scattered electrons, transmitted electrons in the scanning electron microscope, and with secondary electrons in the focused ion beam instrument. The advantages and disadvantages of each method are discussed in order to make it easier to choose the most appropriate technique for grain contrast imaging.

Lorentz Transmission Electron Microscopy (LTEM) allows the observation of magnetization evolution in thin magnetic layers. Electron Microscopy which enables observation at a nanometric scale is sensitive to magnetic induction perpendicular to the beam. Here we show that this technique allows the magnetization process in thin layers having perpendicular magnetic anisotropy to be studied. LTEM has been carried out on a JEOL 3010 and a FEI Titan 300kV (equipped with a Lorentz Lens). We determine how the magnetization is distributed along the domain walls (where the magnetization is in plane). Hence we could observe the evolution of a domain configuration during the magnetization process versus the orientation of its boundaries. This study is based on phase reconstruction methods and is completed by Lorentz image simulations.

A variety of nanostructures of carbon in nuclear graphite were revealed by transmission electron microscopy (TEM) and high resolution TEM. Theses nanostructures include nanosized graphite particles, quinoline insoluble particles, a chaotic structure and a non-graphitizable structure of carbon. The basic structure of these nanostructures was observed to be nanosized packets of graphitic sheets or nanosized graphene.

In this work we have investigated the influence of nanoscale and microscale structure on the tribo-mechanical performance and failure mechanisms of two biocompatible dental polymer composites, with different reinforcing particulates, using advanced microscopy techniques. Nano- and micro structural analysis reveals the shape, size and distribution of the particles in the composites. In the microparticle filled polymer composite (microcomposite), the particles are of irregular shape with sharp edges with non-uniform distribution in the matrix. However, in the nanoparticle filled composites (nanocomposite), filler particles are spherical in shape with uniform distribution in the matrix. From nanoindentation measurements, hardness and reduced modulus of the microcomposite were found to be heterogeneous. However, the hardness and reduced modulus of the nanocomposite were found to be homogeneous. The nanocomposite shows better tribo-mechanical performance compared to that of the microcomposite.

TEM characterization of a CoFeB/MgO/CoFeB magnetic tunnel junction (MTJ) has been performed. A TEM cross section was prepared using focused ion beam (FIB) milling techniques. However, the high energy Ga⁺ beam causes sample damage and induces the redeposition of the sputtered materials on the section surface. Complementary investigation of the crystal structure of the active trilayer in the MTJ was performed by depositing films directly onto a TEM holey carbon film. The TEM imaging, selected area electron diffraction (SAED) and energy dispersive X-ray (EDX) analysis were employed to study the nanostructure. The MgO layer is found to be incompletely crystalline with randomly oriented MgO crystallites. The CoFeB layer is amorphous and is homogenously deposited.

Focused ion beam (FIB) milling techniques are presented aiming at the manipulation of both tin dioxide (SnO₂) inverted opals and polystyrene (PS) direct opals. Different SnO₂ opals are considered in order to estimate the regularity of their bulk after the production. A SnO₂ mesoporous monolith is FIB micromachined to make it suitable for optical applications. PS direct opals are structured by FIB milling at different scales. Ordered arrays of PS opals are modified by selectively removing a single sphere. In performing this task, we discuss the effects on the FIB milling due to the gas-assisted enhanced etching and to the binding of the nearest neighbours. Techniques to achieve imaging of PS opals in absence of a conductive coating are also brought up. Furthermore, isolated PS spheres are drilled with or without enhanced etching in order to produce controlled defects on them. The FIB-assisted manipulations we show may find potential applications in the field of photonic crystals, (bio)sensors and lithography assisted by colloidal masks.

This work focuses on the preparation of zeolite and alumina hydrocracking catalysts for investigation by electron energy-loss spectroscopy (EELS). EELS can potentially give new insights into the location and structure of coke which can result in catalyst deactivation. Three sample preparation techniques have been used - microtoming, focussed ion beam milling (LIB) and conventional ion beam milling. Crushing and grinding the catalyst pellets has been discounted as a preparation technique as the spatial relationship between the coke and the catalyst is lost using this method. Microtomed sections show some mechanical damage while sections milled in a single beam LIB microscope show gallium decoration in pores and were too thick for EELS. Conventional ion beam milling has proved to be most successful as it results in extensive thin regions and maintains the spatial distribution of the zeolite and alumina phases.

The combination of electron energy loss (EEL) spectroscopy and density functional theory can lead to an understanding of how changes within a material lead to changes in electronic structure. In the past, C₇₀ EEL spectra have been interpreted using molecular and crystalline calculations. Although the shape of EEL spectra was explained well in both cases, the molecular calculation attributed the second peak in the spectrum to the ten equatorial atoms, which was not found to be the case in the crystalline calculations. In this paper we have studied the effect of variations in bond length on the density of states, and use this to explain the differences between the two previous calculations.

Low-voltage transmission electron microscopy (LVEM) using very low accelerating voltage ≤ 5keV for observation of low density materials (unstained biological samples and organic thin films) has the advantage of high contrast due to the high scattering efficiency of atoms at low electron energy. The disadvantage of LVEM is the very low transmittable sample thickness. To quantify the transmittance of the samples and the contrast formation, we simulated the interaction of an electron beam with a thin carbon film by our Monte-Carlo code. As the result of our calculations, we obtained the spatial distribution of the transmitted electrons in conditions like those in LVEM, and we were able to find the dependence of the image contrast on the thickness and density of the samples and on the size of the microscope objective aperture. The critical thicknesses for elastic, inelastic and both types of scattering at a density of 2.0 g/cm³ and energy 5 keV were 17.7, 10.3 and 6.3 nm, respectively. The calculated dependences of the electron microscopical parameters show the limits of sample thickness in LVEM

Electron tomography allows the three-dimensional (3D) quantitative characterization of nanostructures, provided a monotonic relationship is fulfilled between the projected signal and the atomic number and thickness of the specimen. This requirement is not satisfied if the micrographs are affected by (i) diffraction contrast, (ii) absorption-thickness limit or (iii) detector saturation. The effects of non-monotonic tomography acquisition have been examined using computer simulations and experimental tilt series of conical tungsten tips. It is shown that the reconstruction artefacts arising from non-linearity can be best predicted by considering geometric tomography (which reconstructs the object external shape) and quantitative tomography (which reconstructs the 3D density function) as limiting cases.

The microstructure and phase stability of a CuAl-on-sapphire thin film material is examined using in-situ and ex-situ heating experiments and TEM, SAD, HREM, and EELS as analysis technique. An in-situ heating sequence has revealed formation and decomposition of the metastable tetragonal θ-alloy phase CuAl₂ at two consecutive reaction temperatures. Pockets of unreacted Al remain at the interface to the substrate. The epitaxial orientation relationship of the θ <=> α-Al₂O₃ interface is examined by HREM, while composition mapping across the metal-alloy-oxide regions is demonstrated using plasmon (low-loss) spectrum imaging and EDX mapping. A 2nd derivative technique is shown to be powerful for enhancing phase changes which exhibit similar plasmon resonance energies.

Many methods have been extended for the preparation of nanoparticles. One of the most important methods is the chemical wet process, e.g. the co-precipitation method that has been used for the preparation of Fe-Ni nanoparticles by the authors. XRD patterns show that the nanoparticles are amorphous before calcination and crystallized after calcination. SEM images show that the grain size of the Fe-Ni particles is in 50–300 nanometre range and that the texture of the nanoparticles after calcination was smoother than before calcination. Hysteresis loops show that the Fe-Ni nanoparticles are superparamagnetic before calcination, because the carbonate phase still exists in the sample, and that they are ferromagnetic materials after calcination. For 40Fe-60Ni nanoparticles after calcination, H_c = 0.12 and B_s = 4800 Oe, being very different in respect to the bulk 40Fe-60Ni alloy.

Silicon nano crystals (NCs) have attracted considerable interest for possible uses in optoelectronics¹ As the particle size decreases the properties of NCs become increasingly sensitive to the surface termination.^{2, 3} Monolayer chemistries^4-10 have been exploited to control the physicochemical properties. NCs are often prepared by vapour-phase deposition techniques; using these they can be conveniently analysed via gas phase analysis techniques, such as mass spectrometry. This cannot be employed, however, if NCs are not synthesized in the gas phase. Here we present a STEM study of undecyl-capped SiNCs, evaporated intact upon heating in ultrahigh vacuum at 200 °C and collected on a variety of solid substrates, including carbon-coated TEM grids. The BF- and HAADF lattice images confirm that the particles have a crystalline core with Si-lattice spacings. The presence of Si in the core is also confirmed by Si-L edge EELS, which reveals furthermore the presence of a surface oxide.

Manganese nanoparticles have been produced using the inert gas condensation technique. XRD has been used to analyse the particles, which contain a mixture of Mn-Mn oxide phases. Too few isolated reflections from any one phase have meant that broadening measurements are limited, and strain measurements are unreliable. Another way to measure strain in a sample is through analysis of lattice images taken using TEM. Observation of strains over long atomic distances can prove difficult, but a method of magnifying strain is by the optical moiré technique. A crystal lattice image with a grating superimposed over the top creates dark bands due to interference between the two. These dark bands are called moiré fringes and they can occur as an artefact in TEM images due to interference between two real crystal lattices. Here though they are implemented using a synthetic grating where the strain in crystalline solids can be studied. The manganese nanoparticle samples have been found to contain fully oxidised particles below 20nm diameter, and core shell particles above 20nm diameter. The technique has been used to study the core-shell particles, which consist of a β-Mn core and an oxide shell. The core is a single crystal and the moiré fringes show that there is no strain in that region. The oxide crystals in the shell are too fine for the technique to work, and so any strain in them must be measured using a different method.

Titanium dioxide based powders are regularly used in the development of latent fingerprints on dark surfaces. For analysis of prints on adhesive tapes, the titanium dioxide is suspended in a surfactant and used in the form of a small particle reagent (SPR). Analysis of commercially available products shows varying levels of effectiveness of print development, with some powders adhering to the background as well as the print. Scanning electron microscopy (SEM) images of prints developed with different powders show a range of levels of aggregation of particles.

Analytical transmission electron microscopy (TEM) of the fingerprint powder shows TiO₂ particles with a surrounding coating, tens of nanometres thick, consisting of Al and Si rich material. X ray photoelectron spectroscopy (XPS) is used to determine the composition and chemical state of the surface of the powders; with a penetration depth of approximately 10nm, this technique demonstrates differing Ti: Al: Si ratios and oxidation states between the surfaces of different powders. Levels of titanium detected with this technique demonstrate variation in the integrity of the surface coating. The thickness, integrity and composition of the Al/Si-based coating is related to the level of aggregation of TiO₂ particles and efficacy of print development.

The fabrication of novel atomic force microscopy (AFM) probes for nanoindentation and nanoimprint lithography (NIL) is presented. Nanomachining induced by focused ion beam (FIB) were employed in order to modify the original tip shape of commercial silicon AFM probes. The FIB-modified probes are used both to perform experiments as to image the corresponding tip-induced surface modifications. With this approach, a relationship between the hardness of a material and the shape of the indenter has been found in the nanoindentation application, and we have obtained information related to the force acting on the mold during its detaching from the polymer film in the AFM-NIL application.

Etching techniques are used to fabricate nickel and tungsten tips suitable as probes for in-situ SEM and TEM experiments. The smallest tip radius achieved is 10nm for Ni and 1.5 nm for W. Depending on the electrolyte concentration and speed of tip movement during etching, the macroscopic morphology can be changed between conical and concave shapes and the tip end can be modified. HREM and EELS characterisation is used to determine oxide layer thickness. For Ni the oxide layer is found to be extremely thin, while for W various thicknesses can be achieved. The particular amorphous WO₃ layer is imaged by plasmon resonance mapping in EELS, which turns out to be a sensitive and discriminative technique exploiting the much larger plasmon peak width of the oxide compared to the pure metal.

Surprisingly large strain plasticity has been demonstrated for ceramic SiC nanowires through in-situ deformation experiments near room temperature. This article reports a detailed electron energy-loss spectroscopy (EELS) study of deformation-induced localized plastic zones in a bent SiC nanowire. Both the 'red shift' of the plasmon peak and the characteristic fine structure at Si L-edge absorption are consistent with local amorphisation of SiC. The recorded C K-edge fine structure is processed to remove the contribution from the surface amorphous carbon and the extracted C K-edge fine structure has no characteristic sp2-related pre-edge peak and hence is also consistent with amorphous SiC. These results suggest that the large strain plasticity in SiC nanowires is enabled by crystalline-to-amorphous transition.

Electron beam deposition of tungsten from W(CO)₆ has been studied for a range of different height deposits. The resistance of the deposited tracks was found to decrease with increasing height implying that thicker deposits have a higher metallic cross section. It was also found that when the current limits were increased to values in excess of 100 μA, the resistance decreased with successive voltage cycles and increasing current limit. This improvement continued until a point where the structure of the deposit appeared to start breaking down and its resistance increased slightly. Subsequently, the structure was found to break near to the contact. It was also found that as the resistance of the deposit decreased, the structure of the deposited tracks changed implying that ohmic heating induced high enough temperatures within the deposit to be able to cause the structure of the material to change.

Based on individual piezodriven nanomotoric units a nanomanipulation platform is integrated into a low voltage field emission SEM (Zeiss Gemnini). The instrumentation allows the controlled movement of individual particles (ca. 50 nm), the cutting of ultrathin polymer membranes or the deformation of hollow polymer spheres with an ultrathin elastic membrane (ca. 20 nm thickness). Interaction with the objects can be done under direct microscopic observation. Imaging up to 20000 x magnification without any detectable influences of external vibrations guarantees the precise manipulation. Under appropriate imaging conditions (e.g. using low energy electrons) specimen charging can be avoided and even uncoated glass tips can be used to manipulate samples.

This paper explores the possibilities for imaging and chemical analysis of thin foil specimens in the SEM. Bright field and dark field imaging provide high resolution imaging with crystallographic information within the grains. In multiphase materials with varying electron transmission the dark field images generally provide a more even contrast in all phases. It is possible to obtain high-quality quantitative EDX data with high spatial resolution.

Bioactive silica gels/polymer systems have been produced using a sol-gel route and their bio-compatibility has been investigated by immersing them in simulated body fluid (SBF). The porous monoliths have been characterised by SEM and EDX analysis where images obtained show pores on the surface of 10–200 μm. The silica gels are not homogeneous and distinct regions of silicon and calcium are observed. The growth of an apatite layer on the surface of the gels was evident after steeping in SBF.

We have employed Environmental Scanning Electron Microscopy (ESEM) and Energy Dispersive X-ray (EDX) analysis to study the microstructural evolution of acrylic latex stabilised with a novel polysaccharide derived from agricultural waste. The analysis revealed that the micro structure of the latex in the 'wet' state consists of individual particles and clusters. Upon evaporation a discontinuous film is formed, with voids present within its structure, which is inconsistent with the conventional descriptions of film formation. Further ESEM examination of 'dry' specimens revealed that aging resulted in the formation of dendritic structures on the surfaces of the latex films, which EDX analysis confirmed to have been formed via crystallisation of salt. The experimental evidence suggests that the clusters, which are part of the structure of the latex, may act as nucleation centres that would allow the dendrites to form.

A multi-scale approach to stress corrosion cracking is presented in this paper. It will be shown that the same crack and, more importantly, the crack tip region can be characterized with surface techniques such as NanoSIMS and EBSD and later prepared for TEM or Atom Probe. This will offer a unique insight into the chemistry and microstructure, proving that the right combination of techniques can provide most of the information needed to correctly understand the mechanisms of crack propagation.

The use of a combined Focussed Ion Beam/Environmental Scanning Electron Microscope (FIB/ESEM) offers new possibilities for imaging complex heterogeneous polymeric structures. The use of the focussed ion beam, using positively charged gallium ions in conjunction with a measured 'defocused' low energy primary electron beam has permitted milling through the heterostructure to be achieved in a controlled way, exposing the inner structure without introducing significant charge damage into the sample. The use of the Environmental Scanning Electron Microscope for imaging the revealed milled sections has then enabled insulating polymer structures to be imaged without charging problems, despite the absence of a conductive coating. Cross sections of a 900nm thick spun cast film of phase separated polystyrene-polybutadiene blends have been successfully milled and imaged, the results agreeing with previous experiments produced using ultramicrotomy and TEM. In addition, elliptical shaped titanium dioxide particles approximately 220nm in diameter have been dispersed in commercial film forming latices at concentrations between 0 and 100 percent volume. The films have been cast, milled 2μm deep and imaged. Results on the arrangement of titanium dioxide particles in the polymer matrices are presented.

3D Focused Ion Beam tomography is increasingly being used for 3D characterisation of microstructures in the 50 nm −20 μm range. Here FIB tomography has been used to study crack morphologies under Vickers microindentations in R-cut alumina and in soda-lime-silicate glass samples. 3D tomographic reconstruction of crack distribution around 100g microindentation sites shows that the crack density in alumina is very much greater than glass. In addition the cracks observed are influenced by the location of the FIB milled surface trenches due to localised stress changes.

The application of an automated procedure for the rapid assessment of selected area electron diffraction patterns is described. Comparison with complementary EDX spectra has enabled the thermal decomposition reactions within high velocity oxy-fuel thermally sprayed WC-Co coatings to be investigated.

A novel method is presented, combining site-specific TEM sample preparation and in-situ STEM analysis in a dual-beam microscope (FIB/SEM) fitted with a chamber mounted nano-manipulator. TEM samples are prepared using a modified in-situ, lift-out method, whereby the samples are thinned and oriented for immediate in-situ STEM analysis using the tilt, translation, and rotation capabilities of a FIB/SEM sample stage, a nano-manipulator, and a novel cassette. This cassette can provide a second tilt axis, orthogonal to the stage tilt axis, so that the STEM image contrast can be optimized to reveal the structural features of the sample (true STEM imaging in the FIB/SEM). The angles necessary for stage rotation and probe shaft rotation are calculated based on the position of the nano-manipulator relative to the stage and door and the stage tilt angle. A FIB/SEM instrument, equipped with a high resolution scanning electron column, can provide sufficiently high image resolution to enable many failure analysis and process control applications to be successfully carried out without requiring the use of a separate dedicated TEM/STEM instrument. The benefits of this novel approach are increased throughput and reduced cost per sample. Comparative analysis of different sample preparation methods is provided, and the STEM images obtained are shown.

The use of spherical aberration correctors in the scanning transmission electron microscope (STEM) has the effect of reducing the depth of field of the microscope, making three-dimensional imaging of a specimen possible by optical sectioning. Depth resolution can be improved further by placing aberration correctors and lenses pre and post specimen to achieve an imaging mode known as scanning confocal electron microscopy (SCEM). We present the calculated incoherent point spread functions (PSF) and optical transfer functions (OTF) of a STEM and SCEM. The OTF for a STEM is shown to have a missing cone region which results in severe blurring along the optic axis, which can be especially severe for extended objects. We also present strategies for reconstruction of experimental data, such as three-dimensional deconvolution of the point spread function.

Transmission electron microscopy (TEM) was used to characterise the microstructure of non-polar α-plane (11-20) GaN grown on (1-102) r-plane sapphire. Conventional TEM found that the microstructure of the non-polar GaN layers is dominated by basal plane and prismatic stacking faults (BSFs and PSFs, respectively). Partial dislocations (PDs) are found at the intersection of, and bounding, these stacking faults. The density of BSFs and PDs in the non-polar films were determined as 5x10⁵ cm^-1 and 4x10¹⁰cm^-2, respectively. In cross section, the PD distribution was anisotropic. When viewed along <0001> the density of PDs appeared to decrease through the film thickness, which may be related to the initial growth conditions of the epilayer.

Under a gaseous atmosphere at high temperatures, almost all the materials (metal, catalysts, etc.) change their structures and properties. For the research and development of materials, it is of vital importance to clarify mechanisms of solid-gas and liquid-gas reactions. Recently an in situ TEM system combined with an environmental holder, which has a gas injection nozzle close to a specimen-heating element, has been developed. The gas injection nozzle permits gas to flow around the specimens sitting on the heating element made of a fine W filament. The newly developed in situ TEM has a differential pumping system; therefore, the pressure in the specimen chamber is maintained in the range of higher than 1 Pa, while the pressure in the electron gun chamber can be kept in the range of 10^-5 Pa. This system was applied to in situ observation of chemical reactions of metals with gases: Observation of oxidation and reduction under a gas pressure ranging from 10^-5 Pa to 1 Pa at high temperatures (room temperature to ∼1473 K) were successfully carried out on pure metal and rare metal catalysts at near-atomic resolution. This in situ environmental TEM system is promising for clarifying mechanisms of many solid-gas and liquid-gas reactions that take place at high temperatures under a gas atmosphere.

Strain arising in epitaxial thin films can be beneficial in some cases but devastating in others. By altering the lattice parameters, strain may give a thin film properties hitherto unseen in the bulk material. On the other hand, heavily strained systems are prone to develop lattice defects in order to relieve the strain, which can cause device failure or, at least, a decrease in functionality. Using convergent beam electron diffraction (CBED) and high-resolution transmission electron microscopy (HRTEM), it is possible to determine local strains within a material. By comparing the results from CBED and HRTEM experiments, it is possible to gain a complete view of a material, including the strain and any lattice defects present. As well as looking at how the two experimental techniques differ from each other, I will also look at how results from different image analysis algorithms compare. Strain in Si/SiGe samples and BST/SRO/MgO capacitor structures will be discussed.

Using a dedicated FEG STEM, we present highly spatially-resolved electron energy-loss (EEL) studies of individual multi-walled carbon nanotubes (MWCNTs), each with the inner cavity possessing regions completely filled with silver. The transmission and attenuation of graphite π-collective mode E-fields through the MWCNT walls are established. Noticeable changes in the graphite π-surface mode are witnessed, concomitant with coupling of the silver Mie mode and the graphite π-surface mode. The resulting collective mode is significantly red-shifted to below 5 eV, with considerable intensity in the visible frequency regime. It appears that silver retains its ability to enhance E-fields when surrounded by a MWCNT. Present observations lead to the possibility of collective modes propagating on graphene monolayers being tuned in frequency by the presence of a metal.

Specimen heating holder consisting of a gas injector and two sets of heating element has been developed for use with conventional high resolution analytical TEMs. One of the heating elements is used for specimen heating and the other for metal deposition. Both heating elements are made of fine wire of tungsten. The specimen holder allows high resolution in-situ TEM observation of a process of gas-solid reaction followed by a metal deposition on the reactant product. The microscope used in the study was a standard TEM equipped with EDX detector and TV camera system. Thanks to the high speed pumping system and well designed column, the pressure of the electron gun area was kept at 5x10^-5Pa or lower for more than 5 hours while the specimen chamber was gas injected to the pressure of 10^-1Pa. Oxidation of Al to synthesize A1₂O₃ carrier, deposition of AuPd catalyst on the Al₂O₃ carrier and behaviour of the deposited AuPd nano-particles at high temperatures in the air injected environment were successfully observed at atomic resolution.

A new generation of imaging detectors is being considered for application in TEM, but which device architectures can provide the best images? Monte Carlo simulations of the electron-sensor interaction are used here to calculate the expected modulation transfer of monolithic active pixel sensors (MAPS), hybrid active pixel sensors (HAPS) and double sided Silicon strip detectors (DSSD), showing that ideal and nearly ideal transfer can be obtained using DSSD and MAPS sensors. These results highly recommend the replacement of current phosphor screen and charge coupled device imaging systems with such new directly exposed position sensitive electron detectors.

Energy-dispersive X-ray spectroscopy in a transmission electron microscope is a standard tool for chemical microanalysis and routinely provides qualitative information on the presence of all major elements above Z=5 (boron) in a sample. Spectrum quantification relies on suitable corrections for absorption and fluorescence, in particular for thick samples and soft X-rays. A brief presentation is given of an easy way to improve quantification accuracy by evaluating the intensity ratio of two measurements acquired at different detector take-off angles. As the take-off angle determines the effective sample thickness seen by the detector this method corresponds to taking two measurements from the same position at two different thicknesses, which allows to correct absorption and fluorescence more reliably. An analytical solution for determining the depth of a feature embedded in the specimen foil is also provided.

Electron microscopic imaging methods usually provide only indirect chemical information on thin layers. Energy-dispersive X-ray spectroscopy in a transmission electron microscope is a more direct analytical method of microanalysis but lacks atomic spatial resolution. Hence, information on ultra-thin layers must be interpreted carefully. Two approaches are compared here. They are based on either recording, in scan mode, line profiles across thin layers or, in transmission imaging mode, series of spectra with different electron beam diameters centred on the same position. The accuracy of quantification is judged using Monte Carlo simulations of electron beam broadening within a GaAs specimen. It is shown for a 0.28 nm thin InAs layer that the indirect approach based on linear regression analysis of series of spectra reliably yields sub-monolayer accuracy in the effective chemical layer width.

A possible configuration for undertaking diffractive imaging microscopy in the conventional scanning transmission electron microscope (STEM) is to defocus the probe by a large distance in order to minimise the number of diffraction patterns required to reconstruct the ptychographical image. Although the characteristics of STEM probes have been well explored and measured near the beam crossover, they are rarely observed (or calculated) at very large defoci. In this paper we compare probes calculated under a variety of approximations with measured data from a JEOL 2010F in STEM mode. When the illumination aperture is narrow and the defocus is large (near parallel illumination useful for ptychographical imaging) the Ronchigram cannot be used easily to characterize aberration, We develop an alternative method of estimating the effective aperture size and condenser lens spherical aberration.

We demonstrate experimentally an application of ptychographical iterative phase retrieval using a focussed-probe geometry, analogous to the scanning transmission electron microscope (STEM). Unlike conventional iterative phase retrieval methods, the technique is inherently able to handle complex, softly varying illumination, due to its use of an update function rather than a hard support mask to localise the modulation of the object plane. This feature is crucial to application in electron microscopy, where hard-edged localisation of the illumination on the specimen is difficult to achieve experimentally.

An energy stabilizing system for electron energy loss spectrometers in a transmission electron microscope (TEM) is described. A wire-based fast response detector is used to position the zero-loss peak. A digital-based real-time processor is used to stabilize the energy-dispersed electron spectrum and simultaneously to acquire a drift corrected electron energy loss spectrum. The performance of the stabilizing system is evaluated.

A miniaturised nanomanipulation and nanoindentation system has been designed and manufactured to perform sub-micron localised in-situ deformation studies in a high resolution transmission electron microscope (HRTEM). The nanomanipulation drive comprises two independent mechanisms for both coarse and fine positioning of sharp indenter tips. Small slip-stick inertial sliders are used to coarsely position a tip which fits inside a bespoke hollowed specimen holder for a JEOL (Japan) 2010/3010 series microscope. The coarse drive comprises three fully independent sliders which are set mutually perpendicular to one another (x, y and z) with a range > 1 mm and resolution ∼ 100 nm. Fine positioning is achieved with a quartered piezoelectric tube with range ∼ 2 μm and resolution < 1 Å. Optical displacement sensors have been used to characterise the nanomanipulation drive performance including total displacement rate and step size in ambient conditions. These are compared to the operation of the drive within a TEM under vacuum conditions. TEM observations at high magnification enable optimisation of the fine and coarse motion and overall drive stiffness.

Slick-stick inertial sliders have been used in the manufacture of a novel miniaturised nanomanipulation and nanoindentation system. This has been designed to perform sub-micron localised in-situ deformation studies in a high resolution transmission electron microscope (HRTEM). Coarse position is realised by three independent sliders which are set mutually perpendicular to one another (x, y and z) with a range > lmm and resolution <100nm In this paper we discuss the effect that gravity has on the performance of the nanomanipulation slip-stick inertial drives by monitoring displacement rate, resolution and the effect of the normal force across the slip-stick slider. We report on the successful actuation of slip-stick sliders operated in a vertical orientation.

Focused ion beam / scanning electron microscopy (FIBSEM) procedures are examined for the sectioning of biological cell / biomedical implant materials. Conventional site-specific lift-out techniques enable the structural characterisation of hydroxyapatite/Ti coatings. Cryo-FIBSEM procedures, as used to investigate human osteoblasts/ hydroxyapatite/Ti structures, emphasise the need for the repeated deposition of coating layers to protect such delicate structures and to minimise beam curtaining effects.

The PDF is the author index.

The PDF is the subject index.