Application of FTIR spectra for evaluating interfacial reactions in metal matrix composites.

Manufacturing of Saffil/MgLi metal matrix composites by the melt infiltration process is accompanied by extensive interfacial redox reaction between delta-Al(2)O(3) fibers (Saffil) and lithium. The present paper deals with the Fourier transform infrared spectroscopy examination of Saffil fibers isolated from Mg-8 wt% Li alloy by the bromine/methylacetate agent focusing on the insertion of Li(+) ions into delta-Al(2)O(3) and their influence on water adsorption. Insertion of Li(+) into delta-Al(2)O(3) is monitored by gradual change of Al-O stretching bands (400-900 cm(-1)) towards more simple patterns of a spinel-like product assigned as delta(Li) which transforms to LiAl(5)O(8) during subsequent annealing. Rapid increase in the water adsorption with increase in Li content, indicated by the changes in H-O-H bending (about 1,650 cm(-1)) and O-H stretching (about 3,500 cm(-1)), is connected with the ionicity of the delta(Li) phase, which attracts polar water molecules.

Determination of manganese valency in La1-xSrxMnO3 using ELNES in the (S)TEM.

In this article several experimentally identified Mn valence-sensitive ELNES quantities for the La1-xSrxMnO3 compound class are presented, namely the energy separations between Mn-L3 and O-Ka, between O-Kb and O-Ka edges, the Mn-L2,3 white line intensity ratio, and the Mn-L3 line width. Valence sensitivities of these quantities are evaluated, and possible additional influences on them are considered. At high signal-to-noise ratio the two energy separations display most sensitively changes of the Mn valency. An experiment-based estimation of the total uncertainties of the quantities indicates that at low signal-to-noise ratio, which is the case when studying interface effects at high spatial resolution, again both energy separations allow to resolve the smallest valency changes.

Structure and giant magnetoresistance behaviour of Co-Cu/Cu multilayers electrodeposited under various deposition conditions.

Electrodeposited Co-Cu/Cu multilayers were prepared under a variety of deposition conditions on either a polycrystalline Ti foil or on a silicon wafer covered by a Ta buffer and a Cu seed layer. X-ray diffraction (XRD) revealed a strong (111) texture for all multilayers with clear satellite peaks for the multilayers on Si/Ta/Cu substrates, in some cases for up to three reflections. Cross-sectional transmission electron microscopy investigations indicated a much more uniform multilayer structure on the Si/Ta/Cu substrates. The bilayer periods from XRD satellite reflections were in reasonable agreement with nominal values. An analysis of the overall chemical composition of the multilayers gave estimates of the sublayer thickness changes due to the Co-dissolution process during the Cu deposition pulse. The XRD lattice spacing data indicated a behaviour close to a simple "multilayer" Vegard's law which was, however, further refined by taking into account elastic strains as well. In agreement with the structural studies, magnetoresistance data also indicated the formation of more perfect multilayers on the smooth Si/Ta/Cu substrates. An analysis of the magnetoresistance behaviour revealed the presence of superparamagnetic (SPM) regions in the magnetic layers. The contribution of these SPM regions to the total observed giant magnetoresistance was found to be dominating under certain deposition conditions, e.g., for magnetic layer thicknesses less than 1 nm (about 5 monolayers).

Extraction of EELS white-line intensities of manganese compounds: methods, accuracy, and valence sensitivity.

This article evaluates the valence sensitivity and accuracy of selected EELS white-line extraction methods based on the white-line intensity ratio in case of manganese oxides, predominantly in the valency range [Mn3+ ; Mn4+]. For this purpose Mn-L2,3 ionization edges of several Mn oxides were measured and four different methods were applied to extract the intensities. The obtained ratios have been analyzed in terms of their errors, their scatter, and their sensitivity on manganese valency. We found that the maximum-intensity as well as the Pearson and Walsh-Dray methods show larger relative changes of the ratio between Mn3+ and Mn4+ than curve-fitting. Taking the scatter into account, Walsh-Dray distinguishes more clearly between those two valencies than curve-fitting. Thus the Walsh-Dray method promises the most accurate determination of small valency shifts.

Radio frequency glow discharge source with integrated voltage and current probes used for sputtering rate and emission yield measurements at insulating samples.

Radio frequency glow discharge optical emission spectroscopy (RF-GD-OES) is routinely used for the chemical analysis of solid samples. Two independent electrical signals from the discharge are required for quantification. When sputtering insulating samples, the voltage over the discharge is not directly measurable. The coupling capacity of the sample is required in order to calculate the discharge voltage. A procedure is outlined where the coupling capacity is determined using an electrical measurement without discharge. The calculated time-dependent discharge voltage and current are evaluated using a plasma equivalent circuit. An insulating sample is sputtered at constant cathode voltage and current. The emission yield for an aluminium line is comparable to that of conducting reference material.

Nanometer-size cluster formation in alkali-metal-doped fullerene layers.

Kinetic Monte Carlo methods have been used to simulate structural transformations in fullerene layers during electrochemical intercalation with alkali-metal ions (A). Special attention is paid to the thermodynamic stability of the A(x)C(60) phases. The calculations point out a phase separation in the doped fullerene layer into alkali-metal-rich and alkali-metal-depleted areas at room temperature. The final state is represented by two phases which coexist as a stable fine mixture of nanoscale particles. The instability of homogeneous layers has potentially critical impact on their electrical properties and can explain the formation of nanostructures (20-50 nm) at the fullerene-electrolyte interface. Rb(3)C(60) clusters are predicted to be larger than K(3)C(60) ones for equal mean alkali-metal concentrations. Experimental data on electrochemical metal deposition on alkali-metal-doped fullerene substrates-in particular, atomic force microscopy measurements-are also consistent with the model proposed.

Microstructure and composition of annealed Al/Ti-metallization layers.

Al/Ti multilayers with columnar grains were deposited by electron-beam evaporation on piezoelectric LiNbO(3) substrates. After annealing in air and under vacuum conditions dissolution of the Ti interlayer was observed for all samples. The original Ti interlayer dissolved completely and globular Al(3)Ti grains were formed within an Al matrix. All samples had an oxidized adhesive Ti bottom layer and a 10 nm thin Al layer below this adhesive Ti bottom layer, which remains intact after the applied heat treatment. This resistance against dissolution by interdiffusion could be caused by the oxidation. These changes in the microstructure and in the chemical composition were investigated by conventional and analytical TEM.

Comparison of the annealing behavior of thin Ta films deposited onto Si and SiO2 substrates.

Structural changes at annealing temperatures (T(an)) of 500-1,100 degrees C were investigated for thin Ta films which were sputter-deposited onto pure Si substrates and onto thermally oxidized Si. In the as-deposited state, the Ta layers predominantly consist of metastable tetragonal beta-Ta, whereby the [001] texture is independent of the substrate material. At lower annealing temperatures, the microstructural evolution is essentially the same for both Ta films. Incorporation of O atoms causes an increase of the intrinsic compressive stress, and diffusion of C atoms into the Ta layer leads to the formation of Ta(2)C. Additionally, a partial transformation of the original beta-Ta phase into a second phase with tetragonal unit cell (denoted as beta'-Ta) occurs. For the Ta/Si system, the formation of a Ta-Si intermixing layer is initiated at T(an)=550 degrees C, and nucleation of crystalline TaSi(2) occurs at T(an)=620 degrees C. The formation of a second Ta silicide was not detected up to T(an)=900 degrees C. In the case of the Ta film deposited onto the SiO(2) substrate, the metastable beta-Ta and the beta'-Ta transform completely into the thermodynamically stable cubic alpha-Ta at T(an)=750 degrees C. A marked reaction with the substrate indicated by the formation of Ta(2)O(5) and Ta(5)Si(3) occurs at T(an)=1,000 degrees C.

Specific properties of fine SnO2 powders connected with surface segregation.

The effect of surface segregation in Sb- and In-doped SnO2 fine-grained powders has been analyzed in comparison with single-crystalline samples. The kinetics and thermodynamics of the Sb and In segregation processes were studied as a function of annealing temperature by X-ray photoelectron spectroscopy (XPS) after annealing in an oxygen-containing atmosphere. Significant differences between diffusion and segregation were revealed for doped powders and single crystals, obviously because of simultaneous diffusion and particle-growth processes proceeding during annealing of powders. For doped single crystals the thermodynamic equilibrium is approached after 24 h annealing above 850 degrees C and at 1000 degrees C for Sb and In, respectively. Higher effective activation energies of diffusion are observed for doped powders and the thermodynamic equilibrium is not achieved under technologically relevant annealing conditions. On the basis of dopant profile measurements anomalies in the electrical resistivity at 300 degrees C of Sb-doped SnO2 powders annealed at 700 and 900 degrees C were attributed to an Sb-depleted zone formed beneath the segregated surface during the kinetic regime. To achieve optimum resistivity behavior for commercial application, inhomogeneous doping of powders must be avoided by appropriate preparation steps.

AES and SIMS investigation of diffusion barriers for copper metallization in power-SAW devices.

Barrier layers for Cu-metallization in surface acoustic wave (SAW) devices were investigated by AES and SIMS depth profiles. Two layered systems on LiNbO(3) substrate have been analyzed after annealing in air up to 400 degrees C. The investigated systems were (A) Ta(20 nm)/Cu(150 nm)/Ti(30 nm), deposited by electron beam evaporation, and (B) Ta(30)Si(18)N(52)(50 nm)/Cu(150 nm)/Ta(30)Si(18)N(52)(50 nm) deposited by magnetron sputtering. In system A the Ta layer shows oxidation in air for T>or=300 degrees C. Ti from the buffer layer diffuses into the Cu at about 100 degrees C, and segregates at the Ta/Cu interface for T>or=200 degrees C. Oxidation of the Ti layer starts at 300 degrees C. But no remarkable amounts of oxygen could be found in the Cu film. The depth profiles show that the TaSiN layer in system B operates as a more effective barrier for the Cu-SAW technology up to more than 300 degrees C.

XPS investigations of thin tantalum films on a silicon surface.

Ultra thin tantalum-based diffusion barriers are of great interest in copper metallisation technology. Even the smallest amounts of copper that diffuse into the active silicon regions on a microprocessor will alter their semiconducting properties thus leading to failure of the device. In the present work Ta films were deposited on silicon by electron beam evaporation and magnetron sputtering. The background of this study is investigation of interface formation, which is expected to have substantial influence on the properties of thin Ta films. All experiments were carried out under UHV conditions. This was necessary because Ta is a very reactive metal and is readily oxidized even at low oxygen partial pressure. The Ta4f peak, as a sensitive indicator of the chemical state, was analysed and compared to that for standard samples. Silicide formation is assumed to occur at the Ta/Si interface.

Calibration of XPS - energy scale for determination of the oxidation states of doping elements in SnO 2 powders.

X-ray photoelectron spectroscopy (XPS) has been used to investigate the oxidation states of doping elements in doped SnO(2) powders. Because of low conductivity, however, charging and resulting peak shift is observed. To obtain the real peak position a suitable reference peak must be found in the XPS spectrum. In this study both internal (Sn3d(5/2) peak) and external references (Au4f(7/2) and C1 s) were examined. When external references were used a shift of all the peaks studied was observed; the extent of this depended on the doping element and the doping concentration. By doping with an element of valence >4 (Nb and Sb) we obtained peaks at binding energy (BE); doping with a trivalent element (In) led to peaks at values of the BE. This peak shift is connected with changes of Fermi level. In contrast, by using Sn3d(5/2) as reference we obtained results which enabled, for example, observation of the dependence of changes of the oxidation state of Sb on doping concentration.

Characterization of laser-irradiated YNi 2 B 2 C surfaces by Auger electron spectroscopy.

The suitability of laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) for precise analysis of YNi(2)B(2)C has been investigated. The intensity ratios B/Y and Ni/Y were found to vary during ablation as a function of the ablation conditions. This could be because of fractionation, owing to incongruent ablation or transport and plasma effects. The bottoms and surroundings of the craters were investigated by scanning Auger electron spectrometry. The bottoms of the craters produced by ablation are covered with a thin oxide comparable with that on the polished crystal surface. The craters are surrounded by an oxide layer the dimensions and thickness of which depend on the laser conditions. The formation of this oxide can be assumed to be a result of partial oxidation of sample material during the ablation process; the oxide is then redeposited around the laser crater.

Nanostructure and thermoelectric properties of ReSi 2+/-x thin films.

Anomalies in the nanostructure evolution of ReSi(2+/-x) thin films have proved to be of large interest in connection with their thermoelectric properties. By means of electron microscopic methods the correlation between structural properties and transport behaviour has been studied. The short-range order of the amorphous state was characterised by reduced density functions calculated from diffuse electron diffraction diagram and is found to correlate with the temperature dependence of the electrical resistance. The crystallisation process observed in situ in the transmission electron microscope starts with the formation of relatively large ReSi(1.75) grains (up to 100 nm). In later stages, only smaller grains are growing. This leads to a decrease in the mean grain size and to the increase of the nanocrystalline volume fraction during the heat treatment. This behaviour allows the investigation of the thermopower as function of the nanocrystalline volume fraction. Thus, at a nanocrystalline content of about 35% the thermopower exhibits a maximum in accordance with calculations.

Investigation of thermoelectric silicide thin films by means of analytical transmission electron microscopy

The microstructure of rhenium silicide thin films and its progress by annealing were investigated by means of analytical transmission electron microscopy. Sputtered amorphous films were characterised by analysis of the radial distribution function (RDF). The position of the first maximum of RDF represents the most probable distance between neighbouring atoms and decreases from 2.75 to 2.62 A in films with an increasing Si-content from 60 to 75 at%. This decrease correlates with the change of the temperature coefficient (TC) of the electrical resistivity. During in situ annealing, the formation of nanocrystals in films with different Si-contents was observed. In thin films with 64 at% the quantity of nanocrystals increases after 1 h at 900 K whereas their sizes remain unchanged. The crystallisation in Re-rich thin films proceeds lower and produces larger crystals than in films near to the ReSi1.75 stoichiometry. Sputtered epitaxial ReSi1.75 films on Si (1 0 0) consist of crystals with nanometer size and an azimuthal torsion of 45 degrees.

Depth profile analysis of various titanium based coatings on steel and tungsten carbide using laser ablation inductively coupled plasma--"time of flight" mass spectrometry.

A homogenized 193 nm ArF* laser ablation system coupled to an inductively coupled plasma-"Time of Flight"-mass spectrometer (LA-ICP-TOFMS) was tested for depth profiling analysis on different single-layer Ti based coatings on steel and W carbides. Laser parameters, such as repetition rate, pulse energy and spatial resolution were tested to allow optimum depth related calibration curves. The ablation process using a laser repetition rate of 3 Hz, 120 microm crater diameter, and 100 mJ output energy, leads to linear calibration curves independent of the drill time or peak area used for calibrating the thickness of the layer. The best depth resolution obtained (without beam splitter) was 0.20 microm per laser shot. The time resolution of the ICP-TOFMS of 102 ms integration time per isotope was sufficient for the determination of the drill time of the laser through the coatings into the matrix with better than 2.6% RSD (about 7 microm coating thickness, n = 7). Variation of the volume of the ablation cell was not influencing the depth resolution, which suggests that the depth resolution is governed by the ablation process. However, the application on the Ti(N,C) based single layer shows the potential of LA-ICP-TOFMS as a complementary technique for fast depth determinations on various coatings in the low to medium microm region.

New hardware for radio frequency powered glow discharge spectroscopies and its capabilities for analytical applications.

A new radio frequency (rf) hardware is developed for glow discharge spectroscopic methods. The resulting features and its capabilities for analytical applications are discussed. The electrical equipment developed allows to work as quickly, stably, reliably and easily as known from the direct current (dc) mode. Moreover, the rf power measurement has been improved. The hardware has been developed, optimised and tested for glow discharge optical emission spectroscopy (GDOES), but nevertheless it is possible to use it for all procedures applying glow discharge sources.

Quantitative depth profiling in glow discharge spectroscopies - A new deconvolution technique to separate effects of an uneven erosion crater shape.

An algorithm is presented as a concept for the quantification in direct current and radiofrequency glow discharge (GD) modes for GD optical emission spectroscopy. The algorithm is divided into excitation and sputtering part and thus it is possible to distinguish between the different excitation processes and to consider equivalent sputtering crater formations in both modes. Intensity-time profiles are affected corresponding to the method by several effects. One important effect is that sputtering occurs at a single time in different depths because of curved crater bottoms, this is usually called crater effect. The main purpose is to introduce an iterative deconvolution technique which for the quantification numerically takes into account the curved sputtering crater bottom. Input data for the deconvolution technique are the calibrated mass-time profile, the partial densities of the sample constituents and the measured final shape of the sputtering crater. Using a relatively simple model for ion sputtering the deconvolution technique improves iteratively the calculated layer structure by means of information on crater formation. The mathematical handling is illustrated for the quantification of a depth profile of a multilayer sample of ten 100 nm layers. The resulting concentration-depth profile reflects excellently the real elemental distribution of the multilayer system.

Thin film analysis in the nanometer scale.

A survey is presented on the present state of the art in analytical transmission electron microscopy (ATEM). An essential advantage of this method is the simultaneous use of imaging, analytical and microdiffraction techniques with a lateral resolution in the 1.5 nm range. Two different analytical techniques are frequently used as ATEM attachments, energy dispersive X-ray spectrometry (EDXS) and electron energy loss spectrometry (EELS). Microscopic images with nanometer resolution may be also produced by energy selected imaging (ESI) with characteristic energy loss electrons. Advantages and limitations of all these methods will be discussed using actual material problems in the field of thin film research.