Hafnium oxide and silicate materials are now incorporated into working CMOS devices; however, the crystallization mechanism is still poorly understood. In particular, addition of SiO2 to HfO2 has been shown to increase the crystallization temperature of HfO2, hence, allowing it to remain amorphous under current processing conditions. Building on earlier study, we herein, investigate bulk HfxSi1−xO2 samples to determine the effect of SiO2 on the crystallization pathway. Techniques, such as XRD, HTXRD, thermal analysis techniques and TEM are used. It is found that the addition of SiO2 has very little affect on the crystallization path at temperatures below 900°C, but at higher temperatures, a second t‐HfO2 phase nucleates and is stabilized due to the strain of the surrounding amorphous SiO2 material. With an increase in SiO2 content, the temperature at which this nucleation and stabilization occurs is increased. The effect of strain has implications for inhibiting the crystallization of the high‐k layer, reduction of grain boundaries and hence diffusion, reduction of formation of interface layers and the possibility of stabilizing t‐HfO2 rather than m‐HfO2, hence, increasing the dielectric of the layer.
Hafnium‐containing compounds are of great importance to the semiconductor industry as a high‐κ gate dielectric to replace silicon oxynitrides. Here, the crystallization processes and chemistry of bulk hafnia powders are investigated, which will aid in interpretation of reactions and crystallization events occurring in thin films used as gate dielectrics. Amorphous hafnia powder was prepared via a sol–gel route using the precursor HfOCl2·H2O. The powders were subjected to various heat treatments and analyzed using X‐ray diffraction and thermal analysis techniques. A large change in the crystallization pathway was found to occur when the sample was heated in an inert environment compared with air. Instead of the expected monoclinic phase, tetragonal hafnia also formed under these conditions and was observed up to temperatures of ∼760°C. The tetragonal particles eventually transform into monoclinic hafnia on further heating. Possible mechanisms for the crystallization of tetragonal hafnia are discussed. It is proposed that, in an inert environment, tetragonal hafnia is stabilized due to the presence of oxygen vacancies, formed by the reduction of HfIV to HfIII. As the temperature increases the crystal grows until there are too few oxygen vacancies left in the structure to continue stabilizing the tetragonal phase, and hence transformation to monoclinic hafnia occurs.
The synthesis of semiconductor nanowires is more and more interested to the applications for building blocks of the innovative nano-sized devices and circuits, but the research and fabrication of these nanowires are also holding a number of difficulties and challenges. Among many different kinds of semiconductor nanowires, Ga2O3 is increasingly grown for many promising applications in nano-device production, namely nanowire LED and Laser. So far there are many synthesizing methods of semiconductor nanowires, among them the vapor–liquid–solid (VLS) method is simple, cheap and popular. However, when we use the VLS method for nanowire growth, various technological problems exist. This paper aims at investigating some influences of the growth technological conditions and Au metal catalyst on the morphology of Ga2O3 nanowire grown by VLS on GaAs substrate. The main considering factors include the different growing temperatures and times, the effects of Au diffusion, Au droplets formation, Au cluster islands formation, and gas volume of the growing tube/ampoule at the 10−1 torr low air pressure. The obtained experimental results regarding the structural properties of nanowires under these effects investigated by scanning electron microscopy, field emission scanning electron microscopy, high angle annular dark field and bright field, scanning transmission electron microscopy, energy-dispersive X-ray techniques, and focus ion beam are presented and discussed.
Metal (e.g. TiN, Ta2N) inserted hafnia-based high-k gate stacks on Si as well as gallium oxide/gallium gadolinium oxide gate stacks on III-V semiconductors (GaAs, InGaAs, AlInAs) are being investigated for current and future MOSFET applications. The combination of light, medium and heavy elements in multilayers with widths down to a nanometre is a major challenge for high spatial resolution EELS spectrum imaging, especially as the interfaces have significant roughness on this scale. However, it is possible to identify layers significantly thinner than the widths of the profiles that result from beam broadening and/or roughness. To deal with the range of edges present, it is essential to collect edges out to losses of ~3keV. Edges from some elements are difficult to extract but, in suitable case, the ELNES on other edges can be used to give information on such elements. Recording the low loss region of the spectrum as well as the core loss regions at each pixel allows much fuller processing of the data including determination of the absolute numbers of atoms per unit volume. In this way, the distributions of the different phases can be estimated and the overall accuracy is excellent. The results are also in good agreement with earlier work using ELNES.
In this paper the electron-beam lithography conditions and the nanofabrication process are described for current-perpendicular-to-plane (CPP) pillar devices with 30 nm critical dimensions. This work combines a RAITH-150 tool with a negative e-beam resist (AR-7520) so that dense nanopillar arrays are patterned fast into large area samples. The resist dilution and coating conditions are optimized, aiming at its thickness reduction down to 80 nm. The exposure parameters are tuned for different geometries and dimensions, so that features down to 30 nm are exposed with good accuracy (+/- 1.9 nm) and reproducibility. The complete integration of these nanoelements into CPP devices involved electron beam lithography, ion milling for pattern transfer and chemical-mechanical polishing (CMP). Results on devices incorporating very low resistance-area (R x A) MTJ films deposited by Ion beam assisted deposition are shown, for MTJ stacks with R x A down to 0.8 omega x microm2. Device characterization includes electrical measurement of the pillar resistance and the transfer curves under dc magnetic fields (TMR up to 40%).
The surface morphology, crystal structure, and in-plane magnetic anisotropy of sputter deposited Co70Fe30 films on GaAs(110) epilayer substrates have been investigated. The surface morphology of thin Co70Fe30 films appears to closely follow that of the underlying GaAs(110) substrate. The study of the crystal structure by x-ray diffraction shows that no crystallographic orientation other than Co70Fe30(110) is present in our films, although we cannot unambiguously deconvolute characteristic reflections of Co70Fe30(110) from those of GaAs(110) due to the very small lattice mismatch. High-resolution transmission electron microscopy shows that the crystal structure of the Co70Fe30 film is coherently matched to the GaAs(110) substrate. The magnetic anisotropies in both 35 and 1100 Å thick films are consistent with those of similar epitaxial films, and demonstrate that a strong bcc (110) texture is propagated throughout the entire film thickness. The cubic and uniaxial anisotropy constants extracted by fitting with the Stoner–Wohlfarth model are consistent with those of molecular beam epitaxy grown bcc CoxFe(100−x)/GaAs(110). This work shows that sputter deposition may be used to produce structurally coherent bcc-Co70Fe30/GaAs(110) contacts suitable for spin-injection applications.
Thin HfO2 films grown on the lightly oxidized surface of Si(100) wafers have been examined using dark-field transmission electron microscopy and selected area electron diffraction in plan view. The polycrystalline film has a grain size of the order of 100 nm, and many of the grains show evidence of twinning on (110) and (001) planes. Diffraction studies showed that the film had a strong [110] out-of-plane texture and that a tiny volume fraction of a metastable (possibly tetragonal) phase was retained. The reasons for the texture, the twinning, and the retention of the metastable phase are discussed.
Determining the bonding environment at a rough interface, using for example the near-edge fine structure in electron energy loss spectroscopy (EELS), is problematic since the measurement contains information from the interface and surrounding matrix phase. Here we present a novel analytical method for determining the interfacial EELS difference spectrum (with respect to the matrix phase) from a rough interface of unknown geometry, which, unlike multiple linear least squares (MLLS) fitting, does not require the use of reference spectra from suitable standards. The method is based on analysing a series of EELS spectra with variable interface to matrix volume fraction and, as an example, is applied to a TiN/poly-Si interface containing oxygen in a HfO(2)-based, high-k dielectric gate stack. A silicon oxynitride layer was detected at the interface consistent with previous results based on MLLS fitting.
We present results of the magnetic, structural, and chemical characterizations of Mn+-implanted Si displaying n-type semiconducting behavior and ferromagnetic ordering with Curie temperature, TC, well above room temperature. The temperature-dependent magnetization measured by superconducting quantum interference device from 5 to 800 K was characterized by three different critical temperatures (TC*∼45 K, TC1∼630–650 K, and TC2∼805–825 K). Their origins were investigated using dynamic secondary ion mass spectroscopy and transmission electron microscopy (TEM) techniques, including electron energy loss spectroscopy, Z-contrast scanning TEM imaging, and electron diffraction. We provided direct evidences of the presence of a small amount of Fe and Cr impurities which were unintentionally doped into the samples together with the Mn+ ions as well as the formation of Mn-rich precipitates embedded in a Mn-poor matrix. The observed TC* is attributed to the Mn4Si7 precipitates identified by electron diffraction. Possible origins of TC1 and TC2 are also discussed. Our findings raise questions regarding the origin of the high-TC ferromagnetism reported in many material systems without a careful chemical analysis.
Two high-k gate stacks with the structure Si/SiO2/HfO2/TiN/poly-Si are characterised using nanoanalytical electron microscopy. The effect of two key changes to the processing steps during the fabrication of the stacks is investigated. Electron energy-loss spectroscopy is used to show that the TiN layer has a very similar composition whether it is deposited by PVD or ALD. Spectrum imaging in the electron microscope was used to profile the distribution of elements across the layers in the stack. It was found that when the anneal after HfO2 deposition is carried out in a NH3 atmosphere instead of an O2 atmosphere, there is diffusion of N into the SiO2 and HfO2 layers. There is also significant intermixing of the layers at the interfaces for both wafers.
Hf-based systems are going into production this year as the high-k materials replacing amorphous SiO2 and Si(O,N) as the gate dielectric in Si MOSFETs. At the same time, metal inserted poly-Si gate electrodes are being used in the gate stacks to remove problems associated with poly-Si gate electrodes. Significant interface interactions can occur in such systems as a result of the thermal budget received during device processing.
The mechanisms by which coatings develop on weathered grain surfaces, and their potential impact on rates of fluid-mineral interaction, have been investigated by examining feldspars from a 1.1 ky old soil in the Glen Feshie chronosequence, Scottish highlands. Using the focused ion beam technique, electron-transparent, foils for characterization by transmission electron microscopy were cut from selected parts of grain surfaces. Some parts were bare whereas others had accumulations, a few micrometres thick, of Weathering products, often mixed with mineral and microbial debris. Feldspar exposed at bare grain surfaces is crystalline throughout and so there is no evidence for the presence of the amorphous 'leached layers' that typically form in acid-dissolution experiments and have been described from some natural Weathering contexts. The weathering products comprise sub-mu m thick crystallites of an Fe-K aluminosilicate, probably smectite, that have grown within an amorphous and probably organic-rich matrix. There is also evidence for crystallization of clays having been mediated by fungal hyphae. Coatings formed within Glen Feshie soils after similar to 1.1 ky are insufficiently continuous or impermeable to slow rates Of fluid-feldspar reactions, but provide valuable insights into the complex Weathering microenvironments oil debris and microbe-covered mineral surfaces.
Electron energy-loss spectroscopy (EELS) analyses have been performed on a sol–gel deposited lead zirconate titanate film, showing that EELS can be used for heavy as well as light element analysis. The elemental distributions within the sol–gel layers are profiled using the Pb N6,7-edges, Zr M-edges, Ti L-edges and O K-edge. A multiple linear least squares fitting procedure was used to extract the Zr signal which overlaps with the Pb signal. Excellent qualitative information has been obtained on the distribution of the four elements. The non-uniform and complementary distributions of Ti and Zr within each sol–gel deposited layer are observed. The metal:oxygen elemental ratios are quantified using experimental standards of PbTiO3, PbZrO3, ZrO2 and TiO2 to provide relevant cross-section ratios. The quantitative results obtained for Ti/O and Pb/O are very good but the Zr/O results are less accurate. Methods of further improving the results are discussed.
Replacement high-k dielectrics for Si(O,N) in MOSFETs undergo many physical and chemical changes during deposition and processing. The situation is even more complicated when a metal electrode is inserted into the gate stack. Investigation of such systems using advanced nanoanalytical techniques in the transmission electron microscope is discussed.
Our ability to identify thin non-stoichiometric and amorphous layers beneath mineral surfaces has been tested by undertaking X-ray photoelectron spectroscopy (XPS) and transmission electron microscopy (TEM) work on alkali feldspars from pH 1 dissolution experiments. The outcomes of this work were used to help interpret XPS and TEM results from alkali feldspars weathered for <10,000 years in soils overlying the Shap Granite (north-west England). The chemistry of effluent solutions indicates that silica-rich layers a few nanometers in thickness formed during the pH 1 experiments. These layers can be successfully identified by XPS and have lower Al/Si, Na/Si, K/Si and Ca/Si values than the outermost ∼9nm of unweathered controls. Development of Al–Si non-stoichiometry is coupled with loss of crystal structure to produce amorphous layers that are identifiable by TEM where >∼2.5nm thick, whereas the crystallinity of albite is retained despite leaching of Na to depths of tens to hundreds on nanometers. Integration of XPS data over the outermost 6–9nm of naturally weathered Shap feldspars shows that they have stoichiometric Al/Si and K/Si ratios, which is consistent with findings of previous TEM work on the same material that they lack amorphous layers. There is some XPS evidence for loss of K from the outermost couple of nanometers of Shap orthoclase, and the possibility of leaching of Na from albite to greater depths cannot be excluded using the XPS or TEM results. This study demonstrates that the leached layer model, as formulated from laboratory experiments, is inapplicable to the weathering of alkali feldspars within acidic soils, which is an essentially stoichiometric reaction.
A Si(100)/SiO2/HfSiO/TaN/poly-Si gate stack is investigated using analytical electron microscopy and, in particular, electron energy-loss spectroscopy. Oxygen is shown to be present in the TaN layer and at the TaN/poly-Si interface in the form of oxidised TaN and SiO2, respectively. Phase separation of the HfSiO into crystalline HfO2 and amorphous SiO2 is also observed with a resulting widening of the SiO2 layer.
The strength of a microalloyed high strength low alloy steel is controlled by a number of factors including the grain size and the fine dispersion hardening precipitates. For sheet steel, the latter form after the final rolling pass, during and following the austenite to ferrite phase transformation. The majority of these particles are sub-10nm. Both 3 dimensional atom probe (3-DAP) and spectrum imaging (SI) using electron energy loss spectroscopy (EELS) have shown the peak in the size distribution in a V-based steel is around 2nm [1]. As well as the size distribution, the composition of the particles is also of great interest. Earlier work using extraction replicas had shown that particles above 4nm in size were essentially nitrogen rich carbonitrides but there was some uncertainty as to whether a plasma process, used to thin the support film, was modifying the composition [2]. The more recent work, using both 3-DAP and EELS SI to analyse the particles within the ferritic matrix, confirmed that these results also applied to particles below 4nm in size [1]. 3-DAP had difficulty in quantifying the N content while confirming the low C content. In fact, the C seemed to form an “atmosphere” around the particles rather than being contained within them. SI was able to quantify the N but had difficulty determining the C content because of a perturbation of the background shape in front of the C K-edge. This perturbation is a major problem since its features are much bigger than the C K-edge intensity from a small particle.
A system that allows the collection of the low loss spectrum and the core loss spectrum, covering different energy regions, at each pixel in a spectrum image is described. It makes use of a fast electrostatic shutter with control signals provided by the spectrum imaging software and synchronisation provided by the CCD camera controller. The system also allows simultaneous collection of the X-ray spectrum and the signals from the imaging detectors while allowing the use of the existing features of the spectrum imaging software including drift correction and sub-pixel scanning. The system allows acquisition of high-quality spectra from both the core and the low loss regions, allowing full processing of the EELS data. Examples are given to show the benefits, including deconvolution, absolute thickness mapping and determination of numbers of atoms per unit area and per unit volume. Possible further developments are considered.
We demonstrate an isolated magnetic interface anisotropy in amorphous CoFeB films on (Al)GaAs(001), similar to that in epitaxial films but without a magnetocrystalline anisotropy term. The direction of the easy axis corresponds to that due to the interfacial interaction proposed for epitaxial films. We show that the anisotropy is determined by the relative orbital component of the atomic magnetic moments. Charge transfer is ruled out as the origin of the interface anisotropy, and it is postulated that the spin-orbit interaction in the semiconductor is crucial in determining the magnetic anisotropy.
The thickness of HfO2 and hafnium silicate HfxSi1−xOy thin films with a range of compositions are investigated using three complementary analytical techniques. We compare results obtained from Medium Energy Ion Scattering spectroscopy, spectroellipsometry and high-resolution Transmission Electron Microscopy. Our results demonstrate that the thickness of the silicate layers decreases with the Hf content.