Search

International audience

Ceria (CeO2) is known as a refractory oxide for dissolution in nitric acid, since the leaching reaction is thermodynamically unfavorable, except when it is complexed by nitrates but with very slow kinetics. To enhance dissolution, surface activation was achieved using high-energy milling. With the mechanically-activated cerium oxide, leaching in nitric acid reached 36%. The mechanical activation of the solid caused structural and microstructural changes (particle size, specific surface area, crystallite size, lattice strain, defects…). After one hour, the cleavage induced by energetic milling generated two populations: nanoparticles and grains containing defects like dislocations. Beside crystallite size and micro-strain evaluation using X-ray diffraction, cerium oxidation state was measured by Electron Energy-Loss Spectroscopy (EELS) analyses while linear defects were pictured by Transmission Electron Microscopy (TEM) observations. On one hand, it was found that the nanoparticles formed during milling process greatly enhance the dissolution reaction by the creation of Ce3+ thin layers of a few nanometer depth on their surfaces. On the other hand, it is shown that dislocations represent another way to increase the kinetics by activation energy. In conclusion, dissolution rate's growth can be due to different parameters like the leaching of the smallest particles, the presence of reduced oxidation state on nanoparticles and some highly reactive sites concentrating structural defects such as dislocation nodes. Finally, as ceria is also well known to be a safe analogue of PuO2, especially for dissolution studies, a solution for improving the dissolution of ceria would probably also be useful for dissolving the oxides rich in Pu.

Convergent beam electron diffraction (CBED), nano-beam electron diffraction (NBED or NBD), high resolution imaging (HRTEM and HRSTEM) and dark field electron holography (DFEH or HoloDark) are five TEM based techniques able to quantitatively measure strain at the nanometer scale. In order to demonstrate the advantages and disadvantages of each technique, two samples composed of epitaxial silicon-germanium layers embedded in a silicon matrix have been investigated. The five techniques are then compared in terms of strain precision and accuracy, spatial resolution, field of view, mapping abilities and ease of performance and analysis. (C) 2013 Elsevier By. All rights reserved.

Dark field electron holography is a new TEM-based technique for measuring strain with nanometer scale resolution. Here we present the procedure to align a transmission electron microscope and obtain dark field holograms as well as the theoretical background necessary to reconstruct strain maps from holograms. A series of experimental parameters such as biprism voltage, sample thickness, exposure time, tilt angle and choice of diffracted beam are then investigated on a silicon-germanium layer epitaxially embedded in a silicon matrix in order to obtain optimal dark field holograms over a large field of view with good spatial resolution and strain sensitivity.

Convergent Beam Electron Diffraction (CBED) experiments and simulations associated with Finite Element calculations were performed in order to measure strain and stress in a complex device such as periodic MOS transistors with a spatial resolution of about 2 nm and a sensitivity that could reach 50 MPa. A lamella of a thickness of about 475 nm was extracted from the wafer with the transistors by Focus Ion Beam (FIB) and was observed in cross-section in a Transmission Electron Microscope (TEM). When approaching the transistors, the HOLZ lines of the CBED patterns acquired in the silicon substrate, become broader and broader. This HOLZ line broadening, which is due to the stress relaxation in the thin foil, was used to determine quantitatively the strain and stress in the lamella and then in the bulk device. We showed that this procedure could be applied to a complex device. Two parameters, the intrinsic material strains – or equivalently the intrinsic material stresses – in the nickel silicide (NiSi) and nitride (Si 3 N 4 ) layers on the top of the transistors gate, were successfully fitted by trial and error, in the procedure.

GaAs specimens containing p-n junctions have been prepared using focused ion beam (FIB) milling for examination using off-axis electron holography. By lowering the FIB operating voltage from 30 to 8 kV, we have shown a systematic reduction of the electrically 'inactive' thickness from 220 to 100 nm, resulting in a significant increase in the step in phase measured across the junctions as well as an improvement in the signal-to-noise ratio. We also show that the step in phase measured across the junctions can be influenced by the intensity of the electron beam.

Germanium dots have been grown on high twist angle (twist angle as high as 20°) molecular bonded silicon (0 0 1) substrates. We show that, depending on the thickness of the silicon film, the strain field generated by an ordered array of mixed edge interfacial tilt (miss-cut) dislocations may induce an ordered growth of germanium dots. We also show that in order to observe an influence of the mixed edge interfacial dislocations on the growth of germanium dots, the thickness of the film has to be much lower that the period of the mixed edge dislocations array. Germanium dots grown by molecular beam epitaxy on 10–15 nm thick silicon films with the period of tilt dislocation array of 43 nm show a high degree of self-ordering.

Highly ordered germanium nanostructures are grown by molecular beam epitaxy (MBE) on molecularly bonded silicon (0 0 1) substrates. For high twist-angle-bonded substrates (twist angle as high as 20°), a one-dimensional organization of growth is induced by an interfacial network of mixed tilt dislocations. Depending on growth conditions, we were able to achieve an organized growth of germanium dots, hut islands and wires.

The conditions to grow GaN quantum dots (QDs) by plasma-assisted molecular beam epitaxy will be examined. It will be shown that, depending on the Ga/N ratio value, the growth mode of GaN deposited on AlN can be either of the Stranski–Krastanow (SK) or of the Frank–Van der Merwe type. Accordingly, quantum wells or QDs can be grown, depending on the desired application. In the particular case of modified SK growth mode, it will be shown that both plastic and elastic strain relaxation can coexist. Growth of GaN QDs with N-polarity will also be discussed and compared to their counterpart with Ga polarity.

The growth of Ge dots has been achieved by molecular beam epitaxy on a Si (001) surface covered by an ultra thin (20 A) silicon nitride Si 3 N 4 layer. The Ge dots exhibit an epitaxial relationship with the underlying silicon substrate, despite the presence of the dielectric layer. The atomic force microscopy results show that the Ge dots have a hemispherical shape with 10 nm average size and the density of the dots is as high as 4 × 10 11 cm -2 . Moreover, we show that the silicon, grown in order to embed the Ge dots deposited on silicon nitride Si 3 N 4 layer, exhibits a quasi-two-dimensional growth of single crystal.

Self-assembled CdTe quantum dots made on ZnTe buffer have been obtained in special growth conditions: the dot-like islands form themselves in a metastable zone which occurs just before the onset of the plastic relaxation transition. This was evidenced by comparing the in situ in-plane relaxation and structural data for two growth processes, the conventional molecular beam epitaxy (MBE), and the atomic layer epitaxy (ALE). These results demonstrate the importance of the kinetic effect on the formation of CdTe nanostructures. The kinetics during growth, studied here by comparing MBE and ALE processes, not only delays the onset of the plastic relaxation by revealing an intermediate zone where the islands can be formed, but also allows some interdiffusion between the cations Cd and Te to occur, as evidenced by studying the optical and structural properties of these dot-like islands.

In this paper, we report the molecular beam epitaxy of germanium dots on Si (001) surface covered by an ultra-thin (12–20 A thick) dielectric layer. Despite the presence of a dielectric layer, the germanium dots exhibit an epitaxial relationship with the underlying silicon substrate. The structural properties of germanium dots grown on Si (001) surface covered by a thin layer of SiO 2 and on Si (001) surface covered by a thin layer of Si 3 N 4 are discussed. The characterisation of dielectric layers and germanium dots was performed using reflection of high-energy electron diffraction, atomic force microscopy (AFM) and high-resolution transmission electron microscopy (HRTEM). The atomic force microscopy results show that the Ge dots have a hemispherical shape with 10 nm average size and that the density of dots is as high as 2–5×10 11 cm −2 .

Ultra-thin (0 0 1) silicon films (thickness less than 25 nm) directly bonded onto (0 0 1) silicon wafers have been investigated by transmission electron microscopy. Twist interfacial dislocations accommodate the twist between the two crystals, whereas tilt interfacial dislocations accommodate the tilt resulting from the residual vicinality of the initial surfaces. In low-twist angle grain boundaries, twist interfacial dislocations are dissociated and no precipitates are detected. In high-twist angle grain boundaries, there is no dissociation and a high density of silicon oxide precipitates is observed at the interface. Tilt interfacial dislocations are pinned by these precipitates, they are more mobile than precipitates. Without precipitates, their lines are straighter than those with precipitates, and this is especially when the bonded wafers are annealed at a high temperature. When no precipitates are present, tilt interfacial dislocations are associated by pairs, and we demonstrate that each tilt interfacial dislocations introduce a diatomic interfacial step at the interface.

We demonstrate that the growth mode of GaN, AlN and InN in molecular beam epitaxy is two or three dimensional, depending on the competing kinetics of the metal species (Ga, Al or In) and of N. In this view, we show that the presence of foreign species acting as surfactants profoundly modifies the kinetics of the adatoms, eventually leading to an improvement in both structural and optical properties of the material. Next, we discuss the interplay between the growth mode and the strain relaxation in nitride heterostructures. In particular, we show that GaN and InGaN can experience a Stranski–Krastanov growth mode leading to the formation of quantum dots. A mechanism of quantum dot nucleation is proposed in the case of GaN on AlN.

In this communication two defect-selective etching methods for GaN are evaluated and critically compared: (i) orthodox etching in molten bases (KOH}NaOH eutectic denoted E) and in hot H 2 SO 4 /H 3 PO 4 acids (denoted HH etching) and (ii) photoelectrochemical (PEC) etching in aqueous KOH solutions. The parameters of etching are given for di!erent type of materials, i.e. bulk crystals and epitaxial layers, for both Ga- and N-polar M 0001 Nsurfaces. It is shown that molten bases are e!ective in revealing nano-pipes, inversion domains (IDs) and some dislocations, but the optimal etching parameters depend on the type of material, type and density of defects and polarity. Both orthodox etchants that result in the formation of pits on dislocations are also suitable for revealing micro-defects in heavily Mg-doped GaN single crystals but instead of pits, protruding etch features are formed. The reliability and limitations of these etching methods in revealing defects are demonstrated by means of transmission electron microscopy (TEM) calibration and by the selective formation of etch pits on dislocations introduced by indentation. Comparison with PEC etching, newly developed for GaN, is brie#y discussed. ( 2000 Published by Elsevier Science B.V. All rights reserved. PACS: 61.72.Ff

This paper reviews our studies of wurtzite GaN/AlN quantum dots grown by MBE via the Stranski–Krastanow mode. High resolution transmission electron microscopy shows that dots are dislocation free, without any evidence of interdiffusion effects. They form arrays of truncated hexagonal pyramids (height of 4 nm and base diameter of 15 nm), nucleating next to threading dislocations on top of a wetting layer, and exhibiting vertical correlation between different quantum dot layers. The existence of polarization and piezoelectric fields of several MV/cm is demonstrated by photoluminescence and decay time measurements. Nevertheless quantum dots are found to be the most efficient emitters in nitrides at room temperature.

Taking adavantage of the Stranski–Krastanov growth mode of GaN deposited on AlN, the formation of self-organized GaN islands has been achieved in both hexagonal and cubic phases. It has been shown that the dot size variation as a function of time obeys an Ostwald’s ripening mechanism. Vertical correlation effects between stacked layers of dots have been shown to result in a better size homogeneity and a decrease in density. The optical properties of the GaN dots have been studied by photoluminescence. A blue shift is oberved which is definitely assigned to quantum confinement effects.

We observe in strained GaN self-assembled quantum dots grown on an AlN layer, a dramatic modification of the optical emission spectra as the dot size varies. In “large” quantum dots with an average height of 4.1 nm, the photoluminescence (PL) peak is centered at 2.95 eV, nearly 0.5 eV below the bulk GaN bandgap. We attribute this enormous redshift to a giant 5.5 MV cm−1 piezoelectric field present in our dots. Temperature-dependent PL studies reveal the strongly zero-dimensional character of this QD system and are consistent with an intrinsic PL mechanism.

With a view to understand the misfit relaxation mechanisms for binary III-nitride semiconductors, a combined RHEED and TEM study of strain accommodation for AlN/GaN, GaN/AlN and InN/GaN MBE heteroepitaxial combinations is presented. It is shown that AlN grows on GaN (2.7% tensile strain) in a two-dimensional way in accordance with a Frank-van der Merwe mode with plastic relaxation occurring through the formation of non-interacting edge-type dislocations. On the other hand, GaN epitaxy on AlN (2.7% compressive strain) proceeds in different modes according to the growth temperature: in particular, a strained Stranski–Krastanow mode can be evidenced at intermediate temperatures. For the case of InN on GaN (10% compressive strain), strain relaxation occurs through the formation of dislocated islands on the surface, i.e. through a relaxed Stranski–Krastanow mode. The implications of these different growth processes to the development of nitride-based quantum dots heterostructures are presented.

We have grown CdTe/MnTe tilted superlattices by atomic layer epitaxy. These heterostructures are formed by depositing, in the step-flow growth mode, fractional monolayer superlattices (CdTe) m (MnTe) n , with p = m + n ≈ 1, onto 2° A and 2° B Cd 0.95 Zn 0.05 Te vicinal substrates. Transmission electron microscopy images reveal that we obtain a uniform short-range superlattice period and an in-plane modulation which is approximately 0.1 of that expected for perfect lateral atomic ordering. This reduced lateral modulation is evidenced by polarization anisotropy in optical experiments.

AlGaN is an important material in the design of nitride devices. However, little is known concerning its growth with high Al contents. We have studied the growth of AlxGa1-xN epilayers on c-face sapphire by low pressure MOVPE (76 Torr), using triethylgallium, trimethylaluminum and ammonia as precursors. The solid versus gas phase composition relationship was determined experimentally and was fitted using a kinetic model. Then the structural properties of the layers (x=0–1) were studied, using X-ray diffraction, scanning electron microscopy and transmission electron microscopy. We demonstrate that at high Al content, the buffer layer defects are replicated into the AlGaN layer.

GaN layers grown by MOCVD or by MBE on (0001) sapphire have been characterised by transmission electron microscopy (TEM). We make a review of the different crystallographic structures found in theses GaN layers. We comment shortly on the nitridation of the sapphire and the structure of the buffer layer (BL). We point out that the roughness of the BL can be an important parameter for releasing the residual strain of the GaN layer. We compute the Keating energies of the main inversion domain boundaries (IDBs) and translation domains boundaries (TDBs) observed in some GaN layers. The observed structures correspond to the lowest energy models. Perfect dislocations have Burgers vectors equal to a, a+c and c. The dislocation lines are generally parallel to the c-axis. a-Edge dislocations are generally not dissociated and we propose an atomic model for them. Screw dislocations with a Burgers vector equal to c, can `open and close' during growth leaving holes (the so-called nanopipes) in the structure.

The first stages of the growth of strained GaN on AlN were studied using reflection high energy electron diffraction, atomic force microscopy and high resolution electron microscopy. It was shown that GaN grows in the Stranski–Krastanov mode, with three-dimensional islanding occuring after deposition of two monolayers. This 2D/3D transition was found to depend on the growth temperature. At low growth temperature, coalescence of 3D islands rapidly leads to a smooth surface. At high temperature, no smoothing process is observed. It is shown that the size of the 3D islands is controlable and that it is small enough to expect quantum effects.

The epitaxial growth of both cubic and hexagonal GaN epilayers is considered here with the aim of comparing their physical properties. In particular, the growth mechanisms at the first stages of growth will be dealt with together with the quality of the growth front. The optical characteristics of the epilayers will be compared by reference to the structure of the defects present within the different types of layers.

High levels of energy deposition by electronic excitations can induce damage creation in some bulk metallic targets as soon as the rate of energy deposition in electronic excitation is higher than a few 10 keV/nm. The present study is aimed at determining whether such excitations can also induce interdiffusion at the interface of metallic bilayers. Ni Ti bilayers were irradiated at 80 K with GeV Pb, Ta and U ions up to a few 10 13 ions/cm 2 . Damage creation and mixing were followed using various methods: electrical resistance, grazing X-ray and neutron reflectometry, X-ray diffraction, electron microscopy and electron energy loss on target cross-sections. A very strong mixing is indeed observed at the Ni Ti interface as a result of electronic energy deposition: the observed mixing efficiency is at least two orders of magnitude higher than that induced by elastic collisions with target nuclei. We finally study the influence of various parameters (irradiation temperature, symetry of the interface) on the mixing efficiency and follow the kinetics of the mixing process.

A direct method for the determination of the polarity of (0001) wurtzite GaN films is presented, using a combination of ion channeling and convergent beam electron diffraction experiments (CBDE). The samples were grown by metal organic chemical vapor deposition (MOCVD) on (0001) Al2O3. Depending on the nitridation conditions of the buffer, the samples are flat or rough. It is found that flat samples are Ga-terminated. The rough samples exhibit numerous pyramidal defects. In that case, a mixture of N- or Ga-terminated domains is put in evidence. These results unambiguously demonstrate that the nitride layer quality is closely related to the control of the polarity from the very first stages of the growth.

We present a two-stage procedure for the analysis of high resolution transmission electron microscopy (HRTEM) images. First the original image, or more precisely the underlying unit cells, is decomposed according to correspondence analysis (CA), a branch of multivariate statistics (MS), into a certain number of “eigencells” with respective coordinates. CA orders these eigencells in a hierarchical way according to the information content via the associated eigenvalues. The initial image can then be partially reconstructed by considering only those eigencells carrying the relevant information present in the original image and discarding by this way “useless” information (“noise”). Second a reduced set of coordinates equivalent to this partially reconstructed image is then submitted to several classification algorithms known in remote sensing (RS). The specificities of RS and its relationship to other methods will be highlighted. The combination of these two techniques will be applied in order to localize the (γ, γ′)-interface in a Ni base superalloy.

It has been recently established that high levels of energy deposition in electronic excitations can induce damage creation in a few metallic targets as soon as the linear rate of energy deposition in electronic excitation is of the order of a few 10 keV/nm. The present study is aimed at determining whether high electronic excitations can induce interdiffusion at the interface of metallic bilayers. Ni Ti bilayers were irradiated at 80 K with GeV Ta ions up to a few 1013 ions/cm2. Damage creation and mixing were followed using various methods: X-ray and neutron reflectometry, X-ray diffraction, electron microscopy and electron energy loss on transverse cuts. A very strong mixing is observed at the Ni Ti interface as a result of high electronic excitations.

Dy/Zr bilayers and multilayers were grown on oxydized Si(100) substrates by evaporation under ultrahigh vacuum conditions. X-ray and electron diffraction shows that both Zr and Dy layers are hexagonal, strongly textured with the c-axis along the growth direction. The grains can extend over several layers and have an in-plane diameter of about 200 A. The Dy lattice displays an in-plane compression of about 1.5%. Interface profiles determined from high-resolution electron microscopy studies are sharp, extending over less than 3 lattice spacings. Magnetization measurements and polarized neutron reflectometry reveal a strong enhancement of TC, accompanied with a suppression of the helical antiferromagnetic phase in the multilayer samples. This effect can be understood as the result of the change in the balance between exchange and magnetoelastic energies caused by in-plane epitaxial compression of the Dy lattice by Zr. A large out-of-plane magnetic anisotropy energy is observed when the Dy layers are made thinner than 15 A. Volume and interface contributions to this anisotropy compensate each other for a Dy layer thickness of about 8 A. Comparison is made between the results of crystal-field calculations based on the point-charge approximation and the measured anisotropy energy.

Electron cyclotron resonance plasma-assisted molecular-beam epitaxy has been used to grow hexagonal and cubic GaN crystal layers on Si(100). By a combined application of in-situ reflection high-energy electron diffraction, X-ray photoelectron spectroscopy and cross-sectional transmission electron microscopy, the state of the Si(001) surface, the structure of GaN grown on this surface and the orientation relationship between substrate and layer were determined. Substrate cleaning, surface reconstruction and carbon surface contamination were found to have a strong effect on GaN growth in the early stage of epitaxy (up to 2000 Angstrom). While polycrystalline GaN in its hexagonal phase is obtained on a clean Si(001) surface, cubic GaN grows epitaxially on a Si(001)c(4 x 4) surface covered by small three-dimensional cubic beta-SiC crystallites.

We present recent results on the structure and on the magnetic properties of Fe/Ru superlattices. Detailed knowledge of the local structure is obtained from X-ray diffraction, including in the anomalous mode, and by EXAFS, thus bringing evidence that Fe layers correspond to an expansion exceeding 10%. The magnetic state is mostly characterized by Mossbauer spectroscopy and also by magnetization measurements, thus demonstrating that the internal Fe layer carry a moment of about 2μ B , whereas the two interfacial layers bear no moment. Except for the extension of the non-magnetic interface, these results are in agreement with theoretical predictions of Kubler, which indicate an high-spin metastable state for such an expanded hcp Fe phase.