Universität Stuttgart
Permanent URI for this communityhttps://elib.uni-stuttgart.de/handle/11682/1
Browse
Item Open Access Atomistic and continuum studies of deformation and failure in brittle solids and thin film systems(2004) Buehler, Markus J.; Gao, Huajian (Prof. Dr.)We describe joint atomistic and continuum studies of deformation and failure in brittle solids and thin film systems. The work is organized in four parts. In the first part, we present a review on atomistic modeling and analysis tools. The second part is dedicated to joint continuum-atomistic modeling of dynamic fracture of brittle materials, where we employ one-, two- and three-dimensional models. The main focus is a systematic comparison of continuum mechanics theory with atomistic viewpoints. An important point of interest is the role that material nonlinearities play in the dynamics of fracture. The elasticity of a solid clearly depends on its state of deformation. Metals will weaken or soften, and polymers may stiffen as the strain approaches the state of materials failure. It is only for infinitesimal deformation that the elastic moduli can be considered constant and the elasticity of the solid linear. However, many existing theories model fracture using linear elasticity. This can be considered questionable since material fails at the tip of a dynamic crack because of extreme deformation. We show by large-scale atomistic simulations that hyperelasticity, the elasticity of large strains, can play a governing role in the dynamics of fracture and that linear theory is incapable of capturing all phenomena. We introduce the concept of a characteristic length scale for the energy flux near the crack tip and demonstrate that the local hyperelastic wave speed governs the crack speed when the hyperelastic zone approaches this energy length scale. This length scale implies that in order to sustain crack motion, there is no need for long-range energy transport. Instead, only energy stored within a region defined by the characteristic energy length scale needs to be transported toward the crack tip in order to sustain its motion. This new concept helps to form a more complete picture of the dynamics of fracture. For instance, the characteristic energy length scale explains the observation of crack motion faster than all wave speeds in the solid, including recent experimental reports of mode I cracks faster than the shear wave speed. Further, we show that hyperelasticity also governs dynamic crack tip instabilities. Stiffening material behavior allows for straight crack motion up to super-Rayleigh speeds, and softening material behavior causes the crack tip instability to occur at speeds as low as one third of the theoretical limiting speed, in accordance with experimental results. The third part is devoted to the mechanical properties of ultra thin submicron copper films. We discuss a novel material defect referred to as a diffusion wedge, recently proposed theoretically and observed indirectly in experiment. The theory predicts that tractions along the grain boundary are relaxed by diffusional creep and a diffusion wedge is built up. Due to traction relaxation, the diffusion wedge behaves as a crack along the grain boundary in the long-time limit. As a consequence, large resolved shear stresses on glide planes parallel to the film surface develop that cause nucleation of dislocations on glide planes parallel to the film surface and close to the film-substrate interface, referred to as parallel glide dislocations. This new dislocation mechanism in thin films, though standing in contrast to the well known Mathews-Freund-Nix mechanism, has been observed recently in experiments of ultra thin submicron copper films subject to thermal stress. We propose a Rice-Thomson model for nucleation of parallel glide dislocations, and report a critical condition for initiation of grain boundary diffusion in thin films leading to a threshold stress for diffusion initiation. We model experimental thermal cycling curves, and find that the new model improves the stress-temperature curves at high temperatures. By large-scale atomistic modeling, we study the atomic details of buildup of the diffusion wedge and subsequent parallel glide dislocation nucleation. We calculate a critical stress intensity factor as a condition for nucleation of parallel glide dislocations. We observe that the structure of grain boundaries has impact on dislocation nucleation and on the motion of dislocations along grain boundaries. We also show by atomistic studies of diffusional creep in polycrystalline thin films that high-energy grain boundaries provide faster diffusion paths than low-energy grain boundaries. A deformation map summarizes the range of dominance of different strain relaxation mechanisms in ultra-thin films. In the fourth part of this thesis, we emphasize the potentials and limitations of molecular-dynamics simulations in studying small-scale materials phenomena, and include a critical assessment of the simulation methods employed in this work and the validity of the results.Item Open Access Atomistic simulation of interface controlled solid state phase transformations(2005) Bos, Cornelis; Mittemeijer, Eric J. (Prof. Dr. Ir.)A typical example of an interface controlled phase transformation is the massive transformation. The rate of transformation in a massive transformation is determined by processes at the interface. In many iron-based alloys the austenite to ferrite transformation is of massive nature. Most experimental data is collected after the transformation was completed. Data on the interface structure obtained during the transformation would greatly enhance the understanding of the transformation mechanism, but such data are scarce. Molecular dynamics simulations would allow the study of the structure of a moving interface in full atomic detail, as simulations of for example the martensitic transformation have shown. Unfortunately, the timescale required for massive transformations is generally too demanding computationally. To enhance the timescale that can be simulated, kinetic Monte Carlo (kMC) simulations can be made. However, then the possible atom positions are restricted to a set of discrete sites. To enable the use of kMC for the simulation of a phase transformation from austenite to ferrite a new multi-lattice kinetic Monte Carlo method has been developed here. The basis of the multi-lattice kMC method is the description of the multi phase system by multiple intertwining lattices. This means that at any given time only a part of the lattice sites are occupied. The transformation is simulated by atom jumps from occupied to unoccupied sites. To allow the atoms at and near the moving interface to take intermediate positions between the lattice sites, a collection of randomly placed sites is also included. These intermediate positions allow for irregularities in the atomic structure of transformation interfaces. Which jump occurs at a certain moment is determined by the relative jump rates of all possible jumps in the system. For the calculation of the jump rate, the jump activation energy must be known. In this work simulations have been performed where the jump activation energy was determined separately for every jump and where the activation energy was taken as the sum of the change in system energy caused by the jump and a constant energy barrier. Because the separate calculation of the jump activation energy for every jump is very slow it cannot be used in the simulations directly. However, it is possible to train a neural network with data from tens of thousands of jumps. The trained neural network can then be used to calculate the jump activation energy (for every jump separately) in the real simulations. For the atomic interactions two models were used. An embedded atom method (EAM) potential and a bond-counting model. The EAM potential gives more realistic interaction energies, but the bond counting model has the advantage that the driving force for the phase transformation and interface energy can be chosen almost arbitrarily. The massive austenite to ferrite transformation was simulated in pure iron. The focus was on two different aspects of the transformation. First, simulations were performed to investigate how the growth mode depends on the driving force and interface energy. Here it was found that for a constant interface energy, the growth mode can change from continuous to plane-by-plane (comparable to ledge-wise growth) as the driving force is made smaller. For a constant driving force the same change can be induced by changing the interface energy (for example by changing the interface orientation), where a lower interface energy enables continuous growth. Second, the transformation kinetics were examined, by measuring the transformation rate for different temperatures. It was found that the activation energy for the interface mobility is determined by series of energetically unfavourable jumps by groups of atoms. This leads to an interface mobility activation energy that can be considerably larger than the activation energy for single atomic jumps. The series of jumps by groups of atoms are required because atoms can often not jump directly from the parent to the product lattice. The series of jumps by groups of atoms cause a redistribution of the free space available at the interface, enabling (locally) the atoms to jump from the parent to the product lattice. Furthermore, a comparison between the activation energies for the interface mobility and for the boundary self diffusion was made. The activation energies for single atomic jumps at the interface are equal for both processes. Because for interface movement series of jumps by groups of atoms are required, the migration part of the activation energy will generally be higher for interface mobility than for boundary self diffusion.Item Open Access Ausscheidungshärtung dünner Al-0,6Si-0,6Ge-Schichten: Studie zur Übertragbarkeit eines Massivmaterial-Legierungskonzeptes(2001) Kirchner, Steffen; Arzt, Eduard (Prof. Dr. phil.)Die Entwicklung neuer und leistungsfähiger Produkte setzt in vielen Technologiebereichen die fortschreitende Miniaturisierung von Systemkomponenten voraus. Viele der im makroskopischen Bereich bewährten Materialien versagen jedoch im Submikrometermaßstab. Am Beispiel der erstmalig hergestellten dünnen Al-0,6at-Si-0,6at-Ge-Schichten wurde untersucht, ob das an Massivlegierungen entwickelte Konzept der Ausscheidungshärtung auf dünne Legierungsschichten übertragbar ist. Zum Vergleich wurden auch Al-Schichten sowie das Al-Si-Ge- bzw. Al-Massivmaterial charakterisiert. Während das Al-Si-Ge-Massivmaterial sich wie eine ausscheidungsgehärtete Legierung verhält, weisen die Schichten mehrere Besonderheiten auf, die z. T. erstmalig bei Al-Legierungsschichten beobachtet wurden: 1. eine {110}-Textur 2. die vollständige Spannungsrelaxation bei hohen Temperaturen 3. eine stark beschleunigte Ausscheidungsreaktion. Die Ursache dafür scheint die beobachtete Ge-Segregation zur Grenz- bzw. Schichtoberfläche zu sein, die zu einer teilweisen Wiederausscheidung bei der Abkühlung nach der Homogenisierung führt. Wie die Ausscheidungskinetik zeigt, wirken die bereits vorhandenen Ausscheidungen als Keime für ein zweidimensionales, diffusionskontrolliertes Ausscheidungswachstum während der Alterung. Trotz der dadurch groben Ausscheidungsverteilung weisen Al-Si-Ge-Schichten eine ca. doppelt so grosse Nanohärte auf wie Al-Schichten mit gleicher Schichtdicke und Korngrösse. Die Ausscheidungshärtung ist prinzipiell auch in dünnen Schichten wirksam. Das Aushärtungspotential wird jedoch nur teilweise ausgeschöpft, da dünne Schichten eine starke Tendenz zur heterogenen Keimbildung aufweisen. Durch die Segregation von Elementen zur Grenz-/Schichtoberfläche können die Textur, das Ausscheidungsverhalten und die mechanischen Eigenschaften dünner Schichten gezielt beeinflusst werden, welches bei der Entwicklung ausscheidungsgehärteter Dünnschichtlegierungen besonders berücksichtigt werden muss.Item Open Access Behaviour of glasses and polymer derived amorphous ceramics under contact stress(2004) Burghard, Zaklina; Aldinger, Fritz (Prof. Dr. rer. nat.)A Vickers indentation study is presented focusing on the crack opening displacement (COD) method as one new approach for fracture toughness determination. COD measurements over the entire radial indentation crack lengths enable quantitative evaluations of residual stresses at the contact site. An alternative estimation of toughness, without knowledge of the calibration parameter which is required for the indentation crack length (ICL) method, is provided by COD measurements in the vicinity of the crack tip. In addition, this method allows to study slow crack growth which is another important phenomenon. The measurements generally require recourse to high magnification, high accuracy observation techniques like atomic force microscopy (AFM) that has been used in this work. Two different types of glasses, as reference materials, and fully dense, amorphous SiCN ceramics produced from precursor polymers through a casting route are investigated. Soda lime and borosilicate glass have been selected, which are well documented to behave as "normal" and "anomalous" glass under contact stress, respectively. A set of four different pyrolysis temperatures for polymer-derived ceramics differing in pyrolysis temperature (800°C, 900°C, 1000°C, 1100°C), were chosen. Indents for a given load in the investigated materials reveal substantially shorter radial cracks and smaller opening in polymer derived SiCN ceramics. This effect is attributed to the different levels of residual elastic-plastic contact stresses that drive the radial crack formation. In soda-lime glass, the plastic component of contact deformation is shear-driven, with conservation of material volume; in borosilicate, as well as polymer-derived amorphous SiCN ceramics, the plastic component is compression-driven, with resultant material densification. The latter deformation mode is less effective in expanding the surrounding elastic material outward upon removal of the indenter. Hence, the opening and crack lengths are consequently smaller in the amorphous SiCN ceramics which is in agreement with the calculated lowest residual stress level in these materials. A higher toughness in polymer derived amorphous SiCN ceramics relative to the glasses is obtained, and an increase with increasing pyrolysis temperature. The influence of subcritical crack growth on the crack opening profiles was investigated only in case of the glasses since the polymer-derived amorphous SiCN ceramics did not show subcritical crack length growth. The methodology presented in this study should prove useful as a means of characterizing the deformation response of glasses and other brittle materials under contact stress.Item Open Access Cementite in the Fe-N-C system(2008) Nikolussi, Marc; Mittemeijer, Eric Jan (Prof. Dr. Ir.)Gaseous nitriding and nitrocarburising are thermochemical heat treatments usually performed below the binary/ternary eutectoid temperature of the Fe-N/Fe-N-C solid solution. These processes result in the formation of a (i) diffusion zone where nitrogen and/or carbon are dissolved in the octahedral sites of the iron bcc-lattice and a (ii) compound layer. The compound layer is usually composed of epsilon and gamma’ iron-(carbo)-nitride. Under certain circumstances also cementite can form within the compound layer. The formation of massive cementite compound layers is usually accompanied by sooting and cementite disintegration, called metal dusting. The results obtained in the present work can be summarised as follows: (i) Sooting and metal dusting can be suppressed by the additional presence of ammonia in the carburising gas atmosphere. Moreover, a modified parabolic growth law was shown to hold for cementite-layer growth. The apparent activation energy for cementite-layer growth was determined. (ii) The Bagaryatsky orientation relationship holds for cementite grains of the compound layer and ferrite grains of the substrate. Favouritism of orientation variants according to the Bagaryatsky orientation relationship was observed. This was explained on the basis of misfit-strain energy. (iii) Larger cementite-layer thicknesses were observed directly at the location of ferrite grain boundaries. This was explained on the basis of carbon grain boundary diffusion in cementite. (iv) Nitrogen diffusion through cementite was investigated using a calibrated microhardness-measurement technique. Values for the diffusivity of nitrogen through cementite, including the activation energy were determined. (v) The existence of the two-phase equilibrium alpha + epsilon in the Fe-N-C system - which was controversially discussed until now - was proven experimentally by following an invariant transition reaction in this system. (vi) First-principles calculations indicated extreme elastic anisotropy of cementite. This extreme elastic anisotropy was proven experimentally by X-ray diffraction stress measurements using synchrotron radiation.Item Open Access Characterization of the CO sensitivity of electrode materials by solid electrolyte galvanic cells(2004) Plashnitsa, Vladimir; Aldinger, Fritz (Prof. Dr.)The voltage of a galvanic cell using stabilized zirconia as a solid electrolyte can exhibit deviations from the equilibrium value given by the Nernst equation, if oxygen together with traces of an oxidizable gas like CO is exposed to one of the electrodes of the cell. This is called the non-Nernstian voltage behaviour. The basic principle of operation is not yet finally cleared since the experimental results are in accordance with two theoretical approaches. Aiming at a better understanding of the background of this phenomenon the mechanism of functioning of solid electrolyte galvanic cells with various sensing electrode materials (Pt1-xAux alloys) in oxygen containing atmospheres with different CO concentrations at moderate temperatures has been studied by means of electrochemical methods. The characterization of the CO sensitivity and understanding the behaviour of the Pt1-xAux sensing electrodes was done by means of the one-electrolyte galvanic cell and by new approaches based on the bi-electrolyte measuring principle. The CO sensitivity, defined as the difference between the experimental voltage at finite CO concentrations and that under zero CO content, was characterized by the cell voltage measurements within the temperature range of 400-700 °C as a function of the CO concentration in the measuring gas (0-40 000 ppm), the composition of the Pt1-xAux sensing electrodes as well as the reference electrode potential (under O2 and H2/H2O). The dependence of voltage of the one–electrolyte cell with Pt0.2Au0.8 and pure Au sensing electrodes repeats practically that for the Pt electrode, which is close to the theoretical curve. On the other hand, the voltage response of the cell with Pt0.8Au0.2 and Pt0.5Au0.5 sensing electrodes differs greatly from that expected theoretically. The CO sensitivity for the same sensing electrodes varies slightly using different reference electrode potentials. The Pt1-xAux (0Item Open Access Characterization of the conduction properties of alkali metal ion conducting solid electrolytes using thermoelectric measurements(2006) Gautam, Devendraprakash; Aldinger, Fritz (Prof. Dr. rer. nat.)Under certain circumstances the electronic conductivity of the solid electrolyte may play a pivotal role for the behaviour of a solid state galvanic cell. Quantitatively, the extent of the electronic conductivity is expressed by the electronic conduction parameters, a⊕ and a⊖, that denote the alkali metal activities at which the n and p-type electronic conductivities, respectively, of the electrolyte are equal to its ionic conductivity. Previous findings demonstrated the existence of a finite, non-negligible electronic conductivity in various alkali metal ion conductors like Na-beta-alumina (NBA), K-beta-alumina (KBA) and NASICON when these materials are employed as solid electrolytes in potentiometric CO2 sensors or for certain thermodynamic investigations. Particularly, a number of previous studies brought into the light the tendency of the p-type electronic conduction parameter a⊕ to adapt to the chemical potential of those species that represent the neutral counterpart of the mobile ion of the electrolyte. These findings are in contradiction with the basic assumptions of the conventional defect chemical considerations underlying the Schmalzried-Wagner theory of mixed ionic-electronic conduction. Because of the fundamental relevance, they were supposed to be substantiated further, at best by an experimental approach independent of the previously applied ones. Thermoelectric power measurements, under certain prerequisites, allow to determine the proportion of the electronic charge carriers transport in the total conductivity of a mixed ionic-electronically conducting material. Moreover, this technique has an inherent advantage over the isothermal potentiometric technique insofar as the two surfaces of the solid electrolyte need not be exposed to different electrode media as the voltage is primarily generated due to a temperature gradient. This makes the approach an essential and so far disregarded tool to clarify the above mentioned unconventional phenomenon regarding a⊕. The objective of the present work is the quantitative determination of the p-electronic conduction parameter a⊕ of several sodium and potassium ion-conducting solid electrolytes, namely NBA, KBA, NASICON, Na2CO3 and K2CO3, by thermoelectric power measurements under a possibly wide spectrum of the chemical potential of the potential determining species and temperature. The measurements aimed at finding further evidence for the dependence of the p-electronic conduction parameter on the surroundings as was first described by Näfe. In addition, impedance spectroscopy studies were designed to characterize the conduction properties of some exemplary materials by another measuring technique and to independently check the results of the thermoelectric power measurements, at least qualitatively. In view of the paucity of the knowledge regarding the stability of the constituent phases of NASICON and because of the relevance of this topic for the characterization of the electronic conduction parameter of this material, the thermodynamic stability of NASICON was determined as another task to be fulfilled within the present work.Item Open Access Characterization of the electronic conduction parameter of cation conducting solid electrolytes(2003) Shqau, Krenar; Aldinger, Fritz (Prof. Dr. rer. nat.)The thesis aims at the characterization of the p-electronic conduction parameter of cation conducting solid electrolytes. For this purpose potentiometric as well as thermoelectric measurements were carried out. The investigations have been done on commercially available potassium and sodium beta alumina as well as on laboratory-prepared samples of NASICON. The information about the p-electronic conduction parameter of K-beta-Al2O3 was obtained by evaluating the voltage of various galvanic cells using two different reference electrodes under variable measuring conditions, i.e. temperature (593 to 893 K) and potassium chemical potential of the electrodes. Under isothermal condition, a non-linear chemical potential dependence of the voltage response of galvanic cell is obtained. From the results obtained it is evident that the p-electronic conduction parameter of the solid electrolyte is not a constant but adapts to the potassium chemical potential in the surroundings which confirms previous findings on sodium beta alumina. Using the same technique the electronic conduction parameter of NASICON was characterized as a function of the sodium chemical potential at the measuring electrode as well as of the temperature. To prove the consistency of electronic conduction properties obtained from potentiometric methods on sodium beta alumina, another independent technique without employing a secondary electrode i.e. thermoelectric power measurement, was performed. The p-electronic conduction parameters obtained from thermoelectric power and potentiometric methods are in excellent agreement with each other. In addition, the impact of the electronic conduction on the behaviour of potentiometric cells was evident by evaluating thermodynamic data on the system pyrochlore-NaSbO3. As a consequence, the thermodynamic stability obtained from these measurements proves to be much higher compared to that reported in the literature, thereby again confirming the non-conventional properties of the electronic conductivity of beta alumina.Item Open Access Correlation between the microstructure of porous materials and the adsorption properties of H2 and D2(2011) Krkljus, Ivana; Roduner, Emil (Prof. Dr.)One of the most challenging tasks toward the full implementation of the hydrogen based economy is the reversible storage of hydrogen for portable applications. Three main approaches have been investigated to store the hydrogen, storage as a compressed gas or a liquid, or through a direct chemical bond between the hydrogen atom and the material. The alternative approach, the most recently investigated, is the storage of hydrogen at cryogenic conditions. Storage by physisorption within porous adsorbents has particular advantages of complete reversibility, the fast refueling time, the low heat evolution, and above all increased safety. The nature of interaction of hydrogen, deuterium, and gas mixtures with porous adsorbents was exploited by performing thermal desorption spectroscopy (TDS) measurements. This sensitive experimental technique gives qualitative information about the different adsorption sites, which show different desorption temperatures depending on the interaction energy. After an appropriate calibration the amount of gas desorbed may be quantified. To gain a more fundamental insight into the available adsorption sites multiple TDS spectra were recorded, corresponding to different surface coverages (in the pressure range of 1 to 700 mbar), and different heating regimes. Different kind of porous adsorbents, conventional carbon–based materials and novel Metal Organic Framework Materials (MOFs), were used to investigate the hydrogen/deuterium physisorption mechanism. For carbon materials an increase in the hydrogen interaction potential was observed for adsorbents with narrow pore size. The confined geometry, where hydrogen simultaneously interacts with all the surrounding adsorbent walls, strengthens the interaction potential with the adsorbate molecule, thus, maximizing the total van der Waals force on the adsorbate. Crystalline MOFs are a new class of porous materials assembled from discrete metal centers, which act as framework nodes, and organic ligands, employed as linkers. The material properties can be optimized by changing these two main components. Owing to their high porosity, high storage capacity at low temperature, and excellent reversibility kinetics, MOFs have attracted a considerable attention as potential solid–state hydrogen storage materials. This novel class of porous adsorbents has been extensively investigated within this thesis. The greatest challenge for porous adsorbents is to increase the strength of the H2 binding interaction, and bring adsorption closer to RT conditions. Several strategies, aimed at improving hydrogen adsorption potential in MOFs are closely investigated. These strategies comprise the inclusion of open metal sites and the optimization of the pore size and, thus, the adsorption energy by ligand modification. The influence of the coordinatively unsaturated metal centers, liberated by the removal of metal–bound volatile species, has been particularly investigated. As for carbon materials, the H2–MOF interaction potential is especially enhanced in materials with the pore size comparable to the kinetic diameter of the hydrogen molecule. Such effects may result from the overlap of the potential field due to the proximity of the pore wall, which strengthen the interaction potential with the adsorbate molecule. However, smaller pores prevent hydrogen penetration and induce diffusion limitations. Furthermore, the molecular transport in confined pores at low temperatures may be significantly affected by quantum effects.Item Open Access Covalent and heterosupramolecular interaction of ceramic particles(2002) Stieger, Gregor; Aldinger, Fritz (Prof. Dr.)New concepts of particle interaction for the processing of ceramic powders are developed. They are based on chemical reactions either by heterosupramolecular or covalent interaction of proper reactants. For this the particles are functionalized so that they are able to undergo defined reactions with each other. A commercially available beta-cyclodextrin derivative, 3-chlor-5-sodium-hydroxyl-trianzinyl-beta-cyclodextrin, is covalently bound to the Si3N4 surface in an one step reaction. An effortful derivatisation of the cyclodextrin is not required. The triazinyl ring is not bound to a defined position at the cyclodextrin torus (in average 2.8 triazinyl groups per cyclodextrin). In the next covalent approach mono-6-amino-?-cyclodextrin is bound to epoxy-functionalized Si3N4. This route provides a regioselectivity and a spacer between the cyclodextrin torus and the surface ensuring flexibility. The ad- and desorption behavior is investigated by rotation angle measurements, the adsorbed amount is low: approximatively 25% of the total mass is adsorbed. This is in contrast to 9% for the non-modified Si3N4 revealing that mono-6-amino-beta-cyclodextrin binds to the epoxy group. Washing desorbs the loosely bound parts, 17% are still remaining. Because of the seven amino groups of heptakis-6-amino-?-cyclodextrin this molecule binds more easily to epoxy-functionalized Si3N4 than mono-6-amino-beta-cyclodextrin with one amino group. This is exhibited by the intensive DRIFT-signals. The cyclodextrinfunctionalized Si3N4 could heterosupramolecularly be crosslinked by a guest polymer bearing tert-butyl-anilide as an anchor for the cyclodextrin in the side chain. This is interesting for the consolidation step in plastic forming. Al2O3 powder could be cyclodextrin-functionalized in one step by the deposition of a polymeric beta-cyclodextrin derivative. A two layer system consisting of a cyclodextrin polymer and a guest polymer was built up on Al2O3 powder particles. Due to the enlarged steric barrier the dispersibility of the Al2O3 is improved which is interesting for ceramic powder processing. A covalent interaction between two ceramic particles is also achieved by functionalizing Si3N4 with two different commercially available silanes which can react with each other with their non silyl termini thus forming a molecular bridge between particles. 1-triethoxysilyl-3-isocyanato-propane is bonded to Si3N4 and 1-trimethoxysilyl-3-amino-propane reacts with Si3N4. The reaction is indicated in diluted suspension by the formation of agglomerates (50-500 micrometer) and sedimentation. Stable highly concentrated suspensions (20-37.6 vol%) could be prepared with a maximum solid loading of phimax=0.39, too, gelling at the critical volume fraction phicrit=0.35. The data of the relative viscosity etarel in dependence of the volume fraction phi can be fitted with the Mooney equation and the moduli G'(tau) and G''(tau) were measured. The viscosity of the 20-28 vol% gels is lower and G' and G'' have an additional cross over compared to suspensions of the iso-cyanatofunctionalized Si3N4 and the aminofunctionalized Si3N4. Concluding, Si3N4 powder could successfully be crosslinked by using low cost silanes. This technique is a promising possibility for the consolidation step in plastic forming.Item Open Access Deformation and fracture mechanical properties of precursor-derived Si-C-N ceramics(2007) Janakiraman, Narayanan; Aldinger, Fritz (Prof. Dr.)This thesis deals with the investigation of the deformation and fracture mechanical properties of precursor-derived (PDC) Si-C-N ceramics. The materials were synthesized from a liquid poly(ureamethylvinyl)silazane precursor. In order to access the intrinsic mechanical behavior of the materials, fully dense defect-free PDC specimens devoid of process induced extrinsic features were fabricated using a special casting and crosslinking process and controlled thermolysis procedures. The investigations were performed on a range of Si-C-N-(H) PDC with progressively varying material structures from hydrogenated amorphous to phase-separated nanocrystalline microstructures, from which the influence of material structure on the mechanical behavior was analyzed. The crack-tip fracture toughness (KI0) of the materials were evaluated using the novel crack opening displacement (COD) method. The estimated KI0 values ranged from 0.6 to 1.2 MPa m1/2. The variation in KI0 was well correlated with the structural evolution in the materials, effected by the progressive stripping of the one-fold coordinated hydrogen in the amorphous materials leading to increased network connectivity, and the nucleation and growth of turbostratic graphite (TG) and nanocrystalline SiC in the phase-separated materials. The net change in the resistance to fracture in these materials was effected by the change in the average fracture surface energy and crack deflection toughening. Crack deflection observed even in the amorphous materials revealed the presence of structural and compositional inhomogeneities. To further understand the cause and effect of crack deflection in terms of crack tip damage mechanisms, roughness analysis of fracture surfaces was carried out using the fractal approach. The analysis revealed self-affine scaling up to a correlation length of around 50 nm and a self-affine roughness exponent (ζ) of 0.8 ± 0.1 in all the materials, the latter in agreement with the universal roughness exponent conjectured in literature. However, no correlation was observed between the observed roughness exponents and the fracture toughness of the corresponding materials. Examination of the crack opening near the crack tip by high resolution AFM imaging revealed no persistent damage cavities along the crack, concluding that the fracture in the investigated Si-C-N ceramics proceeded in a brittle manner in the resolvable length scales, at crack velocities employed in the present experiments. The deformation behavior of the present Si-C-N ceramics under contact loading conditions was investigated using spherical as well as sharp (Vickers and Berkovich) indentation experiments. The elastic moduli and nanoindentation hardness evaluated from the analysis of the nanoindentation load-displacement curves correlated well with the evolution of material resistance to elastic and plastic deformation, commensurate with the structural and microstructural evolution in the materials. Analysis of the elastic and plastic deformation work quantities derived from the load-unload cycle in the Berkovich nanoindentation enabled the discrimination of the different plastic deformation characteristics of the amorphous and phase segregated materials. The equi-proportional variation of elastic and plastic deformation in the amorphous Si-C-N materials identical to vitreous silica indicated the anomalous character of plastic deformation in these materials that induced appreciable strain hardening under progressive densification. This was manifested in the load-dependant increase in hardness. The contrasting variation of plastic deformation work in the phase-separated materials indicated the emergence of an additional plastic deformation mechanism in these materials, that proceeded by a shear deformation promoted by the TG-phase. The anomalous densification behavior in amorphous Si-C-N materials also led to a load dependence of the strain rate sensitivity (m), also observed in vitreous silica, and controlled the evolution of indentation size effect (ISE). The magnitude and direction of ISE was determined by the relative dominance of the two concurrent effects, namely strain hardening and indentation creep deformation. The evolution of strain rate sensitivity in the range of investigated materials showed good agreement with the cluster model, which relates the increase in the number density of isolated regions in the microstructure to a corresponding increase in m. The non-densifying shear mode of plastic deformation in phase-separated materials led to a decrease in the strain hardening capability, increase in m and increased vulnerability to ISE.Item Open Access Deposition of metal oxide thin films from solutions containing organic additives(2007) Lipowsky, Peter; Aldinger, Fritz (Prof. Dr.)In bio-inspired materials synthesis the principles of biomineralization are employed for the fabrication of materials with favourable functional properties at near-ambient temperature and with little expenditure: Organic templates direct the formation of inorganic matter. In aqueous solution, zinc compounds with manifold morphologies are produced by thermal hydrolysis of zinc nitrate in the presence of biomolecules like amino acids and dipeptides. In methanol, ZnO films are deposited by hydrolysis of zinc acetate in the presence of polymers like polyvinylpyrrolidone (PVP) and polyethylene glycol. With PVP, particularly smooth, uniform and stable films are fabricated. Their thickness is determined by the deposition time and the polymer concentration. Various microscopic and spectroscopic measurements prove that the films consist of textured nanocrystalline zinc oxide. Selected properties of the films, such as their photoluminescence, are investigated. Film deposition is possible on substrates with organic coatings bearing certain functional groups. Patterned films can be deposited after local decomposition of the organic coating by UV light. The mechanism of film formation is treated in detail. Like in biomineralization, an amorphous transient state of matter occurs before crystallization. This state succumbs to ZnO nanocrystals, which either aggregate in solution or adsorb to the substrate. It is demonstrated in what way the additive controls the reaction. Sulfonate-modified polystyrene beads are coated with zinc oxide and used as sacrificial templates for the fabrication of zinc oxide hollow spheres. Laminates of alternating layers of zinc oxide and poly(amino acids) are deposited and exhibit an improved mechanical performance compared to the monolithic zinc oxide.Item Open Access Diffraction analysis of materials in a state of stress: elastic loading and phase transformations(2013) Koker, Margaret Kolbe Annellen; Mittemeijer, Eric J. (Prof. Dr. Ir.)Externally or internally applied stresses and strains can pronouncedly influence material properties. Hence, the role of stress on material behavior is an important and developing field of research. Variations in stress throughout a material can lead to either strengthening or weakening of a specimen or engineering component. A clear understanding of stress and strain, and the ability to predict the magnitude of its variation, in a material (as a function of material processing), in the absence and presence of external loading, is of utmost importance to optimize material properties. Lattice-strain variation in massive, polycrystalline aggregates provides a wealth of information about the grain interaction in an externally loaded specimen. Each grain within the body is confined by its neighbors, and the compliance of these neighboring bodies provides the extent to which a grain in a massive polycrystalline body may deform under loading. As single crystals (with tungsten being an exception) are intrinsically elastically anisotropic, the direction of the applied loading with respect to the grain's crystallographic orientation must also be considered. Various elastic grain-interaction models can be used to approximate the average lattice strain within a crystallite based on its orientation with respect to the aggregate and the external loading. Each of these grain-interaction models is based on its own set of assumptions for the grain interaction. (See Table 2.1 in Ch. 2 for details on each of the discussed elastic grain-interaction models.) Two main categories of grain interaction can be defined: (i) isotropic grain interaction, where the interactions of the grains in all directions adhere to the same assumptions, and (ii) anisotropic grain interaction, where, conversely, the interactions of the grains do not adhere to the same assumptions in all directions. Three categories of strain variation may be present in an elastically loaded polycrystalline aggregate: (i) macro-, (ii) meso-, and (iii) microvariation in strain. The applicability of each set of grain-interaction assumptions (e.g. the individual grain-interaction model) is highly dependent on the specimen and the loading conditions. One shortcoming of the elastic grain-interaction models is that all grains of the same crystallographic orientation are considered to experience identical (average) lattice strains. Also, the lattice-strain variation within an individual grain cannot be calculated according to the models. Hence, these models only calculate approximate solutions for the macrovariation of strain and a portion of the mesovariation of strain, not taking into account any strain variation induced by local heterogeneities in the neighborhood around the individual crystallites. Therefore, the elastic grain-interaction models provide only an underestimate of the total strain variation in a loaded, polycrystalline body. Near equiatomic compositions of NiTi are shape memory alloys, which are characterized by two unique behaviors: pseudoelasticity (also called super elasticity) and shape memory effect. These material properties make NiTi thin films a prime candidate for application in microelectromechanical systems (MEMS). Due to the abrupt structural transition associated with the phase transformation, the material lends itself well to investigation via x-ray diffraction (XRD) techniques. Diffraction line-profile analysis of XRD patterns during such in situ experiments is a powerful tool. The study in Ch. 4 focuses on synchrotron XRD experiments of substrate-bound NiTi thin films (49.2(5) at.%Ni) during in situ heating. Through in situ high temperature XRD measurements, experiments have been performed to track the phase fraction, stress, and crystallite size of the phases during the transformation. A combination of XRD techniques were applied for the investigation to measure the phase fraction (Rietveld analysis) and stresses (curvature and sin2psi methods) as a function of temperature. Measurements of similar samples demonstrated good reproducibility. The macroscopic film stress was observed to increase with temperature. The stresses pertaining to the individual phases were separated, demonstrating that the magnitude of stress is highly dependent on the phase fraction of austenite in the substrate-bound NiTi thin film. Despite increasing in magnitude, the stress in the austenite phase remains biaxially, rotationally symmetric throughout the entire transformation.Item Open Access Diffraction analysis of residual stress; modelling elastic grain interaction(2002) Welzel, Udo Siegfried; Mittemeijer, Eric J. (Prof. Dr. Ir.)X-ray (and neutron) diffraction can be used for the non-destructive analysis of residual and load stresses. Suitable parameters as (X-ray) elastic constants are required for relating the measured lattice strains to the components of the mechanical stress tensor. For the first time a unifying, rigorous treatment for the diffraction stress analysis of both macroscopically elastically isotropic and anisotropic polycrystals is given. The notion 'surface anisotropy' of bulk specimens is revisited as a special case of direction-dependent grain interaction. Evidence for direction-dependent grain interaction, i.e. the direction-dependence of the elastic coupling of grains in a polycrystal, was obtained only very recently for the first time in the diffraction stress analysis of an untextured, polycrystalline nickel thin film by van Leeuwen et al. (van Leeuwen, M., Kamminga, J.-D. & Mittemeijer, E. J. (1999), J. Appl. Phys. 86 [4], 1904). The direction-dependent grain-interaction model employed by van Leeuwen et al. is elaborated for the textured case in this work. The concept of direction-dependent grain interaction for diffraction analysis of stress is generalised, overcoming some unrealistic (extreme) grain interaction assumptions involved in the model presented by van Leeuwen et al.. Experimental verification has been achieved by X-ray diffraction strain measurements performed on a vapour deposited copper film. For quantitative diffraction stress analysis the crystallographic texture of the specimen has to be taken into account in terms of an orientation distribution function. In this work, texture analysis was performed employing a polycapillary X-ray lens. The corrections for instrumental aberrations in texture measurements using an X-ray lens were for the first time rigorously investigated and suitable correction procedures are proposed.Item Open Access Diffraction stress analysis of thin films - investigating elastic grain interaction(2005) Kumar, Atul; Mittemeijer, Eric J. (Prof. Dr. Ir.)The components of the macroscopic, mechanical stress tensor of a polycrystal can be determined by X-ray diffraction stress analysis. Traditional diffraction stress analyses generally presuppose that (i) the specimen is macroscopically elastically isotropic (ii) no depth gradients (of, for example, the stress state) occur. These conditions are often not fulfilled for thin films or surface regions of bulk polycrystals. As a consequence, erroneous results can be obtained. This work is dedicated to the investigation of specimens exhibiting anisotropic microstructures (and thus macroscopic elastic anisotropy) and/or inhomogeneous microstructures, as met near surfaces and in textured materials. The following aspects are covered: (i) Analysis of specimens with direction-dependent (anisotropic) elastic grain-interaction. Elastic grain-interaction determines the distribution of stresses and strains over the (crystallographically) differently oriented grains of a mechanically stressed polycrystal and the mechanical and diffraction (X-ray) elastic constants (relating (diffraction) lattice strains to mechanical stresses). Grain interaction models that allow for anisotropic, direction-dependent grain interaction have been developed very recently. The notion 'direction-dependent' grain-interaction signifies that different grain-interaction constraints prevail along different directions in a specimen. Practical examples of direction-dependent grain interaction are the occurrence of surface anisotropy in thin films and the surface regions of bulk polycrystals and the occurrence of grain-shape (morphological) texture. In this work, for the first time, stress analyses of thin films have been performed on the basis of these newly developed grain-interaction models and it could be shown that curvature observed in so-called sin2ψ plots (i.e. plots of the lattice strain versus sin2ψ, where ψ is the specimen tilt angle), which is incompatible with traditional grain-interaction models, can be fully explained on this basis. It has also been demonstrated that the identification of the (dominant) source of direction-dependent grain interaction is possible. The results for the grain interaction have been discussed in the light of microstructural investigations of the specimens by microscopic techniques. (ii) Analysis of specimens with depth gradients: Diffraction stress analysis can be hindered if gradients of the stress state, the composition or the microstructure occur in the specimen under investigation, as the so-called information depth varies in the course of a traditional stress measurement: Ambiguous results are thus generally obtained. In this work, a strategy for stress measurements at fixed information depths has been developed and from such stress measurements at fixed information depths employing a laboratory diffractometer and a diffractometer at a synchrotron source, the stress gradients and gradients in the elastic grain-interaction constraints of Nickel layers (layer thicknesses 2 micron and 4 micron) have been successfully deduced. Thereby the first evidence ever for the depth-dependence of the so-called surface anisotropy has been obtained.Item Open Access Diffusion in stressed thin films(2005) Chakraborty, Jay; Mittemeijer, Eric Jan (Prof. Dr. Ir.)Interdiffusion in layered thin film structures may occur at temperatures as low as room temperature due to the presence of a high density of short-circuit paths for diffusion. Whereas the effect of the presence of crystalline defects and grain boundaries on diffusion in thin films has received considerable attention in the past, the effect of mechanical stresses on diffusion has attracted only little attention and few quantitative assessments exist, though possible effects of mechanical stresses on diffusion are frequently invoked in a qualitative discussion of diffusion results obtained in thin film systems. With respect to the theoretical basis for the effect of mechanical stresses on diffusion, a rigorous re-thinking of standard textbook knowledge is required. Limits of Fick's first law in a state of non-hydrostatic stress have not been generally recognised: In cases of non-hydrostatic stress, chemical potentials for defining equilibrium are not applicable. Following an approach proposed by Larche and Cahn [Larche F. & Cahn J.W. (1973), Acta Metall 21, 1051], so-called diffusion potentials have to be utilised for defining equilibrium instead of the chemical potentials. Thus, the diffusion flux of a species under non-hydrostatic states of stress is taken proportional to the gradient of a diffusion potential (instead of a chemical potential). The resulting equation for the diffusion flux demonstrates that the flux is not only proportional to the concentration gradient, but also proportional to the product of the gradient of the trace of the mechanical stress tensor and the difference in the partial molar volumes of the diffusing species. Not only the effect of stresses on diffusion has to be considered: Concentration changes arising from diffusion generally may lead to the build-up of stresses. Their magnitude and distribution depends on the differences of the partial molar volumes of the diffusing species, the elastic constants of the alloy considered and the boundary conditions (e.g. if the thin film diffusion couple is freestanding or attached to a rigid substrate). The situation can be (and generally is) complicated by the presence of residual stresses and/or externally applied stresses. In this work, complete flux equations for non-hydrostatic states of stress have been derived both in the lattice-fixed frame of reference and the laboratory frame of reference for the case of interdiffusion in a binary, substitutional diffusion couple. Fick’s first and second laws have been solved numerically by an explicit finite difference method for Pd-Cu thin film diffusion couples (as model examples) subjected to a planar, rotationally symmetric state of stress. The effects of both applied stress gradients as well as diffusion-induced stress gradients have been studied. A general conclusion is that the ‘larger’ atoms are driven towards regions of tensile stress whereas the ‘smaller’ atoms are driven towards regions of compressive stress by the stress gradient acting as a driving force for diffusion. Experimental investigations of interdiffusion in Pd-Cu thin (both 50nm thick) polycrystalline bilayers in the temperature range 175°C to 250°C are reported. To this end, (mainly) Auger electron spectroscopy (AES) in combination with sputter-depth profiling, X-ray diffraction phase, texture and (macro-) stress analysis and transmission electron microscopy (TEM) have been employed. Upon annealing at relatively low temperatures (175°C to 250°C) for durations up to 10 hours, considerable diffusional intermixing occurs. Interdiffusion coefficients have been determined using the so-called ‘centre gradient’ and ‘plateau rise’ methods. Grain boundary diffusion coefficients of Pd through Cu have been determined employing the ‘Whipple-Le Claire’ method. Both interdiffusion and grain boundary diffusion coefficients have been found to decrease (roughly exponentially) with increasing annealing time. Also, a corresponding increase in activation energies for both volume and grain boundary diffusion has been observed. X-ray diffraction phase analysis indicates that interdiffusion is accompanied by the sequential formation of ordered phases. Defect annihilation during annealing and ordering in the Pd-Cu solid solution have been proposed as potential causes for the observed decrease in diffusion coefficients (and corresponding increase in activation energies). The evolution of the stress state upon annealing has been investigated employing ex-situ stress measurements by X-ray diffraction and in-situ wafer curvature stress measurements. The results reveal that tensile stresses are generated during annealing in both sublayers. The stress evolution is discussed in the light of possible mechanisms of stress generation.Item Open Access Dünnschichtplastizität und Wechselwirkung von Gitterversetzungen mit der Film/Substrat Grenzfläche(2004) Edongué, Hervais; Arzt, Eduard (Prof. Dr. phil)Die Gefüge und das thermomechanische Spannungsverhalten von 100-2000 nm dicken epitaktischen und polykristallinen Cu Schichten auf (0001)-alpha-Al2O3 Substraten, die in Ultrahochvakuum hergestellt und bei 600 °C ausgelagert wurden, wurden untersucht. Epitaktische Schichten zeigen zwei {111} zwillingsorientierte Domänen mit einer Domänengröße, die mehr als das 30-fache der Schichtdicke beträgt, während die lognormal verteilte Korngröße der ebenfalls {111} orientierten polykristallinen Schichten nur das 1-2-fache der Schichtdicke aufweist. Dieser Unterschied in den Gefügen macht sich bei dem Verformungsverhalten der beiden Schichtgruppen bemerkbar. Bei hohen Temperaturen werden die Spannungen in polykristallinen Schichten durch Diffusionsplastizität an den Korngrenzen abgebaut. Dagegen ist das plastische Verformen der epitaktischen Schichten unabhängig von der Temperatur immer auf Versetzungsplastizität zurückzuführen. Fließspannungen bei Raumtemperatur nehmen in epitaktischen und polykristallinen Schichten mit abnehmender Schichtdicke zu. Aufgrund ihrer großen Domänen sind epitaktische Schichten jedoch weicher als polykristalline Schichten gleicher Schichtdicke. Allerdings sind die Fließspannungswerte der epitaktischen Schichten größer als Vorhersagen durch ein Modell, das eine Einengung der Versetzungsbewegung durch die Schichtdicke annimmt. Diese Diskrepanz liegt in Zwillingsgrenzen begründet, die parallel zur Schicht-Substratgrenzfläche verlaufen und die Versetzungsbewegung in der Schicht zusätzlich einengen, ähnlich wie die Schicht-Substratgrenzfläche selbst. Diese Einengung der Versetzungsbewegung in dünnen Schichten durch Grenzflächen führt zu den erhöhten Fließspannungswerten.Item Open Access The effect of capillary forces on adhesion of biological and artificial attachment devices(2007) Souza , Emerson Jose de; Arzt, Eduard (Prof. Dr. phil.)The presence of a liquid meniscus can cause far greater adhesion between a particle and a surface than occurs under dry conditions. Recent studies on biological attachment systems have highlighted the unique and important effect of liquid capillarity at the micro- and nanometer scale. The results demonstrate that macroscopic considerations of the classic meniscus theory must be modified to take into account new scaling laws and geometric relationships. A general description of wetting and capillary condensation as it applies to interfaces of small scales and to arbitrary substrates is clearly desirable but remains an unsolved challenge. In this work, I have performed numerical simulations of wet adhesion under less restrictive conditions then has been done before. In particular, I calculated the capillary force as a function of the distance between two substrates for the general case of different properties and different geometries of the substrates. The results are in excellent agreement with analytical results and with measurements of the capillary force. They allow us to propose a novel, effective method to evaluate the contact angle hysteresis of a liquid bridge between arbitrary substrates. The numerical calculations also include the effect of contact splitting which has proven to be a powerful mechanism in many biological attachment systems that are based on dry adhesion. Our results show that this mechanism does in principle also apply to wet adhesion and that splitting of one large bridge into many smaller ones enhances the capillary forces for all possible contact angles. This results in new scaling laws of the capillary force as a function of the number of liquid bridges. They predict, for example, an unexpected maximal force for moderately hydrophilic surfaces (i.e. contact angles around 70 degrees) and a maximal force per contact area for cylindrical bridges. These novel scaling laws lead to a deeper basic understanding of wet adhesion and can also serve as an important guideline as to how artificial attachment devices can be engineered to have specific properties.Item Open Access Effect of H2S on the thermodynamic stability and electrochemical performance of Ni cermet-type of anodes for solid oxide fuel cells(2006) Manga, Venkateswara Rao; Aldinger, Fritz (Prof.Dr.rer.nat)For SOFCs to be main means of power generation, they should be able to exploit wide variety of fuels. Among Ni-cermets, Ni-YSZ is the state-of-the-art materials for SOFC-anode which is the fuel electrode. But sulphur impurity present in different gaseous fuels (e.g Biogas), depending on its concentration, is highly poisonous to the stability and electrochemical performance of the Ni catalyst in the cermet anodes. Thus in this study the microstructural stability of Ni-YSZ, Ni-CGO and Ni-LSGM cermets in H2S-containing hydrogen gas is studied in the intermediate temperature range of SOFC operation. Thermodynamic modelling of Ni-S-O-H quaternary system was performed for the calculation of thermodynamic stability and sulphur-tolerance limit of Ni in the gaseous atmosphere made up of H, O and S. The effect of presence H2S in fuel gas, in the concentrations well below the thermodynamic tolerance limit, on the electrochemical performance of the anodes is studied by using model Ni-patterned electrodes on YSZ and LSGM. Thermodynamic modelling of the Ni-S-O-H quaternary was performed by employing CALPHAD methodology. The modelling of Ni-S binary phase diagram was performed by using sublattice models for the non-stoichiometric phases. The optimised binaries of Ni-O, and Ni-H were taken from the literature. The Ni-O-S and Ni-O-H ternaries were extrapolated from the lower order binaries. In Ni-O-S ternary, NiSO4 is the only ternary compound present. The ternary compounds, Ni(OH)2 and NiOOH in the Ni-O-H ternary were considered as stoichiometric line compounds. The model parameters of the ternary compounds were optimised using the experimental data. The Ni-S-O-H quaternary was calculated by extrapolation method as employed in the CALPHAD methodology. Inorder to understand the H2-oxidation mechanism and the role played by the electrolyte in the reaction mechanism, symmetrical cells of Ni-patterned YSZ single crystals with different crystallographic orientations were studied in H2+H2O gas in the intermediate temperature range. EIS analysis of the symmetrical cells with Ni-patterned electrodes was carried out to understand the electrode processes. Effect of 5 ppm of H2S on the H2-oxidation kinetics at Ni-patterned electrodes was studied. It was found that even 5 ppm of H2S, in H2+H2O gas, is poisonous to the Ni-electrodes on YSZ. The addition of 5ppm of H2S doubled the total polarisation resistance of symmetrical cell. But there was no change in the shape of the impedance arc and a single arc was recorded also with the 5 ppm of H2S impurity. The activation energy obtained from the impedance arc measured with and without 5 ppm of H2S was also same. Hence it can be said that the addition of 5 ppm of H2S does not lead to change of the rate limiting step, but leads to a decrease in the rate of the reaction.Item Open Access The effect of substrate orientation on the kinetics and thermodynamics of initial oxide-film growth on metals(2007) Reichel, Friederike; Mittemeijer, Eric J. (Prof. Dr. Ir.)This thesis addresses the effect of the parent metal-substrate orientation on the thermodynamics and kinetics of ultra-thin oxide-film growth on bare metals upon their exposure to oxygen gas at low temperatures (up to 650 K). For such thin oxide overgrowths on their metals, the resulting oxide-film microstructures often differ from those predicted by bulk thermodynamics, because of the relatively large contributions of interface and surface energies to the total energetics of the various metal-substrate/oxide-film systems. To this end, a model description has been developed to predict the thermodynamically stable microstructure of a thin oxide film grown on its bare metal substrate as function of the oxidation conditions and the substrate orientation. An amorphous state for ultra-thin oxide films grown on e.g. Al, Ti, Zr or Si can be thermodynamically, instead of kinetically, preferred up to a certain critical oxide-film thickness, because of the lower sum of surface and interface energies as compared to the corresponding crystalline modification. Beyond this critical oxide-film thickness, bulk thermodynamics will strive to stabilize the competing crystalline oxide phase. For Mg and Ni, the critical oxide-film thickness is less than 1 oxide monolayer and therefore the initial development of an amorphous oxide phase on these metal substrates is unlikely. Finally, for Cu and densely packed Cr and Fe metal surfaces, oxide overgrowth is predicted to proceed by the direct formation and growth of a crystalline oxide phase. Further, polished Al single-crystals with {111}, {100} and {110} surface orientations were introduced in an ultra-high vacuum system for specimen processing and analysis. After surface cleaning and annealing, the bare Al substrates have been oxidized by exposure to pure oxygen gas. During the oxidation, the oxide-film growth kinetics has been established by real-time in-situ spectroscopic ellipsometry. After the oxidation, the oxide-film microstructures were investigated by angle-resolved X-ray photoelectron spectroscopy and low energy electron diffraction. Finally, high-resolution transmission electron microscopic analysis was applied to study the microstructure and morphology of the grown oxide films on an atomic scale. The stoichiometric (i.e. Al2O3) oxide films grown on Al{111} are amorphous up to 450 K, whereas at higher temperatures epitaxial crystalline oxide films with a coherent metal/oxide interface develop. The oxide films grown on Al{100} are also overall stoichiometric, have uniform thicknesses and atomically flat metal/oxide interfaces. They are amorphous up to 400 K, but are transformed into crystalline gamma-Al2O3 upon annealing beyond a critical thickness. At more elevated temperatures (> 400 K), a crystalline Al2O3 film with a semi-coherent metal/oxide interface develops. For the crystalline gamma-Al2O3 overgrowth on Al{100}, an unexpected high lattice mismatch (> 15%) between the Al{100} substrate and the gamma-Al2O3 overgrowth is found with a semi-coherent metal/oxide interface. The oxide films grown on Al{110} for temperatures smaller than 550 K are also overall stoichiometric and amorphous. At more elevated temperatures (> 550 K), the original bare Al{110} surface becomes reconstructed at the onset of oxidation and {111}-facets develop. The kinetics of the oxide-film growth on the bare Al{100} and Al{110} substrates can be subdivided into a initial, very fast and a subsequent, very slow oxidation stage. For the oxidation of the bare Al{111} substrate up to 450 K, a distinction between an initial, very fast and a subsequent, very slow oxidation stage cannot be made. Instead, the initial oxide-film growth rate on Al{111} decreases only gradually with increasing oxidation time. The experimental growth curves for the thermal oxidation of Al single-crystals in the temperature regime of 350 – 600 K can be accurately described by considering the coupled currents of Al3+ cations and electrons in an uniform surface-charge field. As such, a gradual transformation of the initial amorphous oxide film on Al{100} into gamma-Al2O3, was observed with increasing oxidation temperature in the range of 350 – 600 K for Al{100}, as well as up to 450 K for Al{110}. Whereas, on Al{111}, the corresponding amorphous-to-crystalline transition was found to be more abrupt.