The design and the experimental results of some prototypes of SiC X-ray detectors are presented. The devices have been manufactured on top of 2 inch 4H-SiC wafer with 115 μm thick undoped high purity epitaxial layer, which constitutes the detection’s active volume. Pad and pixel detectors based on Ni-Schottky junctions have been tested. The residual doping of the epi-layer was found tobe extremely low, 3.7 x 1013 cm-3, allowing to achieve the highest detection efficiency and the lower specific capacitance of the detectors. At 22 °C and in operating bias condition, the reverse current densities of the detector’s Schottky junctions have been measured to be between J = 0.3 pA/cm2 and J = 4 pA/cm2; these values are more than two orders of magnitude lower than those of state of the art silicon detectors. With such low leakage currents, the equivalent electronic noise of SiC pixel detectors is as low as 0.5 electrons r.m.s at room temperature, which represents a new state of the art in the scenario of semiconductor radiation detectors.
Nickel-base alloys due to their high performances have been widely used in biomass and coal fired power plants. They can undertake plastic deformation with different strain rates such as those typically seen during creep and fatigue at elevated temperatures. In this study, the mechanical behaviours of Alloy 617 with strain rates from 10-2/s down to 10-6/s at temperatures of 650°C and 700°C have been studied using tensile tests. Furthermore, the microstructures have been investigated using electron backscatter detection and electron channeling contrast imaging. At relatively high strain rate, the alloy shows higher fracture strains at these temperatures. The microstructure investigation shows that it is caused by twinning induced plasticity due to DSA. The fracture strain reaches the highest value at a strain rate of 10-4/s and then it decreases dramatically. At strain rate of 10-6/s, the fracture strain at high temperature is now smaller than that at room temperature, and the strength also decreases with further decreasing strain rate. Dynamic recrystallization can also be observed usually combined with crack initiation and propagation. This is a new type of observation and the mechanisms involved are discussed.
Segregation bands which normally follow the outer contours of a casting are common in commercial magnesium alloy pressure die castings. Several models have been proposed in the literatures which attempt to explain mechanisms behind the formation of this type of segregation bands. However, it is difficult to explain some phenomena which occur in real die cast components. In this paper, a new theory concerning the formation of one common and detrimental segregation band defect (Type I) has been proposed, which is based on a coupled analysis of heat flow and volume changes during solidification. The formation of this type bands was related to a pressure drop in the liquid and resulting flow of segregated liquid from the surrounding two-phase regions. Mechanism on the formation of the other type segregation band (Type II or under surface band) is also proposed. The sudden increase of cooling rate at the moment of applying intensification pressure is believe to has main contribution to the formation of this type of bands.
A new tunnel Schottky diode based on SiC and a mixed conductor of BaSnO3 as the gate has been investigated. I-V curves at different operating temperatures and two different gas atmospheres have been measured. The device shows sensitivity to oxygen, with maximum at 400degreesC. A model that describes the behaviour of the device is proposed, which takes into account the different types of conduction of the BaSnO3 due to the temperature.
4H-SiC BJTs were fabricated using epitaxial regrowth instead of ion implantation to form a highly doped extrinsic base layer necessary for a good base ohmic contact. A remaining p(+) regrowth spacer at the edge of the base-emitter junction is proposed to explain a low current gain of 6 for the BJTs. A breakdown voltage of 1000 V was obtained for devices with Al implanted JTE.
Numerical simulations are one way to obtain a better and more detailed understanding of the chemical vapor deposition process of silicon carbide. Although several attempts have been made in this area during the past ten years, there is still no general model valid for any range of process parameters and choice of precursors, that can be used to control the growth process, and to optimize growth equipment design. In this paper a first step towards such a model is taken. Here, mainly the hydrocarbon chemistry is studied by a detailed gas-phase reaction model, and comparison is made between C3H8 and CH4 as carbon precursor. The results indicate that experimental differences, which previous models have been unable to predict, may be explained by the new model.
We find that a severely rolled FeCo alloy has anomalous enhancement of the rotated-cube {100}< 011 > texture component and a decrease of the {111} components after annealing, which is contrast to the recrystalliization behaviors reported in traditional BCC metals and alloys. The local texture measurements show that two kinds of grains with obviously different orientations, i.e. {100} and {111}, are heterogeneously distributed in the deformed specimen and the migration of high-angle grain boundaries is observed after annealing in the disordering temperature region.
We applied real-time spectroscopic ellipsometric (SE) measurements to assess the removal of overlayer material from 4H-SiC Si- and C-face surfaces in order to investigate the final step of an otherwise standard RCA cleaning regimen commonly used to prepare SiC surfaces for contact formation. The selected treatments (buffered hydrofluoric acid (HF), concentrated HF, dilute HF and 5% HF in Methanol) removed 4 to 40 Angstrom of effective SiO2 overlayer thickness from these surfaces. We also found that the concentrated HF treatment yielded the best surface, i.e. the most abrupt bulk-to-ambient transition region.
Residual stresses have been measured in green bodies after conventional compaction at 400 and 530 MPa and after high speed compaction with impact energies of 1400J and. 2200J. The former were rectangular bars and the latter cylinders. The stresses have been measured by X-ray and neutron diffraction. The full width half maximum peak widths were also recorded. :ft was found that all surfaces had compressive residual stresses in the range 5 to 90 MPa, the largest values found on side surfaces that had been deformed in shear during ejection from the die. The presence of lubricant reduces the residual stress values. The powders were basically water atomised iron powder: Hoganas ASC 100.29, Distalloy AE and Pasc.
Two-dimensional materials offer a unique platform for sensing where extremely high sensitivity is a priority, since even minimal chemical interaction causes noticeable changes inelectrical conductivity, which can be used for the sensor readout. However, the sensitivity has to becomplemented with selectivity, and, for many applications, improved response- and recovery times are needed. This has been addressed, for example, by combining graphene (for sensitivity) with metal/oxides (for selectivity) nanoparticles (NP). On the other hand, functionalization or modification of the graphene often results in poor reproducibility. In this study, we investigate thegas sensing performance of epitaxial graphene on SiC (EG/SiC) decorated with nanostructured metallic layers as well as metal-oxide nanoparticles deposited using scalable thin-film depositiontechniques, like hollow-cathode pulsed plasma sputtering. Under the right modification conditions the electronic properties of the surface remain those of graphene, while the surface chemistry can betuned to improve sensitivity, selectivity and speed of response to several gases relevant for airquality monitoring and control, such as nitrogen dioxide, benzene, and formaldehyde.
Large variations have been observed in the thickness uniformity and carrier concentration of epitaxial graphene grown on SiC by sublimation for samples grown under identical conditions and on nominally on-axis hexagonal SiC (0001) substrates. We have previously shown that these issues are both related to the morphology of the graphene-SiC surface after sublimation growth. Here we present a study on how the substrate polytype, substrate surface morphology and surface restructuring during sublimation growth affect the uniformity and carrier concentration in epitaxial graphene on SiC. These issues were investigated employing surface morphology mapping by atomic force microscopy coupled with local surface potential mapping using scanning Kelvin probe microscopy.
Clinching is a mechanical press joining method, which has become of growing interest in recent time since it has the potential to replace other conventional joining methods like e.g. spot welding. However, there still exists a lack of knowledge in terms of the mechanical behavior of clinched joints under quasistatic or cyclic loading. For that reason clinching is usually used for applications in structures which are not subjected to external loads. In particular the residual stress distribution in the vicinity of clinched joints and its influence on the mechanical behavior of the joints is unknown. Here diffraction methods are used for the determination of characteristic residual stress distributions in undismantled clinched samples. A combined residual stress determination by X-ray and neutron diffraction has been used to get a well-founded assessment of the residual stress distributions in the immediate vicinity of clinched joints. The residual stress analysis is supplemented by characterizations of the microstructures and the mechanical properties of single clinched joints. Two materials with different strain hardening behavior were used for clinching, a micro alloyed steel (ZStE340) and a non age hardenable aluminum base alloy (AlMg5). In addition two different common clinching techniques were applied - the TOX- [5] and the Eckold-technique [6]. Characteristic residual stress distributions were found for the combinations of clinching techniques and joined sheet materials investigated here. It has been determined that the clinching process induces predominantly compressive residual stresses inside the clinch and in the immediate vicinity of the clinch. The near surface residual stress distributions determined by X-ray diffraction measurements tend to reveal somewhat different residual stresses than measured by neutron diffraction, indicating a possible stress gradient through the sheet thickness. Further evaluation of the FWHM-values of the respective interference profiles shows that for both clinching techniques the largest amount of plastic deformation occurs in the clinch lock region.
The build-up of damage in 4H SiC epitaxial layers implanted with 100 or 180 keV Al ions in the dose range of 10(13) to 10(15) cm(-2) has been studied by transmission electron microscopy (TEM) and Rutherford backscattering spectroscopy in the channeling mode (c-RBS). Implantations have been done at temperatures between room temperature and 800 degreesC and the samples have been analysed after implantation and after post implant anneals. In as implanted samples channeling results show that a major part of the damage can be avoided already at implantations at 200 degreesC, but complete removal of damage is not possible even at an implantation temperature of 800 degreesC. After post implant annealing at typically 1600 degreesC a distribution of planar faults are seen by TEM. The size is around 10 nm, but increases with increasing annealing temperature.
The deformation behavior of metastable austenitic stainless steel AISI 301, suffering different initial cold rolling reduction, has been investigated during uniaxial tensile loading. In situ high-energy x-ray diffraction was employed to characterize the residual strain evolution and the strain induced martensitic transformation. Moreover, the 3DXRD technique was employed to characterize the deformation behavior of individual austenite grains during elastic and early plastic deformation. The cold rolling reduction was found to induce compressive residual strains in the austenite along rolling direction and balancing tensile residual strains in the alpha-martensite. The opposite residual strain state was found in the transverse direction. The residual strain states of five individual austenite grains in the bulk of a sample suffering 2% cold rolling reduction was found to be divergent. The difference among the grains, considering both the residual strains and the evolution of these, could not be solely explained by elastic and plastic anisotropy. The strain states of the five austenite grains are also a consequence of the local neighborhood.
A first-principles calculation of the effective mass of electrons in quantum-well-like gap states induced by stacking faults and cubic inclusions in 4H- and 6H-SiC is performed, based on the density functional theory in the local density approximation. Our calculated effective electron masses for perfect crystals are in very good agreement with those previously determined both theoretically and experimentally. It has been found that electrons confined in the thin 3C-like regions have clearly heavier effective masses than that in perfect 3C-SiC.
First-principles band structure calculations of all the structurally different stacking faults that can be introduced by glide along the (0001) basal plane in 3C-, 4H-, and 6H-SiC are performed, based on the local-density approximation within the density-functional theory. Our calculations, using supercells containing 96 atoms, have revealed that both types of stacking faults in 4H-SiC and two of the three different SFs in 6H-SiC give rise to quasi-2D energy band states in the band gap at around 0.2 eV below the lowest conduction band, and are electrically active. The corresponding wave functions are strongly localized around the stacking fault plane. These results imply that stacking faults in these SiC polytypes are efficient planar traps for electron capture and responsible for subsequent electron-hole recombination. This can therefore have a profound influence on bipolar SiC technology.
A first-principles calculation of stacking faults in 15R-SiC is reported. All the geometrically distinguishable stacking faults which can be introduced by the glide of partial dislocations in (0001)-basal planes are investigated: there exist as many as five different stacking faults in 15R-SiC. Electronic properties and stacking fault energies of these extended defects are studied based on the density functional theory in the local density approximation. Stacking fault energies are also calculated using the axial next nearest neighbor Ising (ANNNI) model.
We report on a first-principles band structure calculation of twin boundaries in 3C-SiC, Si, and diamond, based on the density functional theory in the local density approximation. It is found that the electron wave functions belonging to the conduction and valence band edge states in 3C-SiC tend to be localized almost exclusively on different sides of the boundaries, while there is no such feature in Si and diamond. We have interpreted these localization and segregation phenomena as a consequence of the electrostatic field caused by the spontaneous polarization due to the hexagonal symmetry around twin boundaries. A mechanism for the creation of twin boundaries, i.e., propagation of partial dislocations in neighboring basal planes, has been investigated using total energy calculations, and it has been realized that the double-intrinsic-stacking-fault structure in 3C-SiC, coinciding with the extrinsic stacking faults, is much energetically favored.
A first-principles calculation of stacking fault energies in 3C-, 4H-, and 6H-SiC, based on the local-density approximation within the density-functional theory, is reported. All the structurally different stacking faults which can be introduced by glide along the (0001) basal plane are considered. The number of such stacking faults in these polytypes is one, two, and three, respectively. The stacking fault energies are also calculated using the simpler generalized axial next-nearest-neighbor Ising (ANNNI) model. Our calculations confirm that the stacking fault energy of 3C-SiC is negative, and we also find that one of the three types of stacking faults in 6H-SiC has a considerably higher stacking fault energy than the other two types.
First-principles density-functional calculations of the band structure and wave functions around narrow X-like inclusions in 4H-SiC have been performed. X-like inclusions of various thicknesses, corresponding to two, three, and four stacking faults in neighbouring basal planes, have been investigated. The results for the number of bound states in the inclusion, their energies, and wave functions are well described by a simple one-dimensional quantum-well square potential. The quantum-well property of these inclusions suggests that X-like regions in 4H-SiC are efficient planar traps for conduction band electrons.
A first-principles study of intrinsic stacking faults in GaN is reported. Our calculations are based on density functional theory in local density approximation. We have found that the electron wave functions belonging to the conduction and valence band edge states tend to be localized almost exclusively on different sides of the faulted layer. We suggest that the electrostatic field caused by the macroscopically polarized 2H-GaN parts below and above a thin 3C-like layer around the stacking fault is responsible for these possibly shallow localized states.
A three-dimensional computational model for chemical vapor deposition (CVD) of silicon carbide (SiC) in a hot wall reactor is developed, where the susceptor is tapered with a rectangular cross-section. The present work focuses on the advection-diffusion-reaction process in the susceptor. The precursors are propane and silane, and the carrier gas is hydrogen with mass fraction higher than 98%. Computed growth rates under different system pressures and precursor concentrations are compared with the experimental data measured on samples grown in the Linkoping CVD reactor. The gas composition distribution and the growth rate profile are shown. Dependence of the growth rate on precursor concentrations is investigated.
The evolution of micro- and macrostresses in a duplex stainless steel during uniaxial loading has been investigated in situ by X-ray diffraction. Due to differences in the coefficient of thermal expansion between the two phases, compressive residual microstresses were found in the ferritic phase and balancing tensile microstresses in the austenitic phase. The initial microstresses were almost two times higher in the transverse direction compared to the rolling direction. During loading the microstresses increase in the macroscopic elastic regime but starts to decrease slightly with increasing load in the macroscopic plastic regime. During unloading from the plastic regime the microstresses increases by approximately 35 MPa in the direction of applied load but remains constant in the other directions.
Results of a photoemission study of the (100) surface of ZrN and NbN are reported. For the N 1s level a negative surface core level shift is identified on both surfaces. For the metal 3d levels a negative surface shift is determined on ZrN while that on NbN is found to be positive. Negative surface core level shifts are predicted for the metal 3d level in both cases using the thermochemical model. Upon oxygen exposures chemically shifted components appear both in the metal 3d and N Is spectra for both samples. From the shape of the N Is spectra it appears as if different N Is sites are created in the surface region upon oxygen exposures.
Fluorescent silicon carbide was grown using the fast sublimation growth process on low off-axis 6H-SiC substrates. In this case, the morphology of the epilayer and the incorporation of dopants are influenced by the Si/C ratio. Differently converted tantalum foils were introduced into the growth cell in order to change vapor phase stochiometry during the growth. Fluorescent SiC grown using fresh and fully converted tantalum foils contained morphological instabilities leading to lower room temperature photoluminescence intensity while an improved morphology and optical stability was achieved with partly converted tantalum foil. This work reflects the importance of considering the use of Ta foil in sublimation epitaxy regarding the morphological and optical stability in fluorescent silicon carbide.
Growth of 3C or 6H-SiC epilayers on low off-axis 6H-SiC substrates can be mastered by changing the size of the on axis plane formed by long terraces in the epilayer using geometrical control. The desired polytype can be selected in thick (~200 µm) layers of both 6H-SiC and 3C-SiC polytypes on substrates with off-orientation as low as 1.4 and 2 degrees. The resultant crystal quality of the 3C and the 6H-SiC epilayers, grown under the same process parameters, deteriorates when lowering the off-orientation of the substrate.
We have designed and characterized preliminary versions of two wideband SiC-based RF power amplifiers using SiC MESFETs from Chalmers University and Lateral Epitaxy SiC MEESFETs fabricated at AMDS AB. When optimized transistors are available they will be used in the design of amplifiers for a 100-500 MHz multifunction EW system.
Over 150 μm thick epilayers of 4H-SiC with long carrier lifetime have been grown with a chlorinated growth process. The carrier lifetime have been determined by time resolved photoluminescence (TRPL), the lifetime varies a lot between different areas of the sample. This study investigates the origins of lifetime variations in different regions using deep level transient spectroscopy (DLTS), low temperature photoluminescence (LTPL) and a combination of KOH etching and optical microscopy. From optical microscope images it is shown that the area with the shortest carrier lifetime corresponds to an area with high density of structural defects.
The influence of chlorine has been investigated for high growth rates of 4H-SiC epilayers on 4o off-cut substrates. Samples were grown at a growth rate of approximately 50 and 100 μm/h and various Cl/Si ratios. The growth rate, net doping concentration and charge carrier lifetime have been studied as a function of Cl/Si ratio. This study shows some indications that a high Cl concentration in the growth cell leads to less availability of Si during the growth process.
The formation of Al/Si/p-4H SiC ohmic contacts at temperatures as low as 750 degreesC is reported in this paper. The dependence of electrical properties and contact morphology have been investigated as a function of the annealing regime in the interval 600-700 degreesC. The lowest contact resistivity of 3.8x10(-5) Omega .cm(2) was obtained at 700 degreesC annealing, however the most reproducible results were in the low 10(-4) Omega .cm(2) range. It has been established that the predominate current transport mechanism in the Al/Si/SiC contacts is thermionic-field emission. Atomic force microscopy showed that the addition of Si to the contact layer improves its morphology, and the pitting of annealed Al is not observed. The contacts developed are stable during ageing at 500 degreesC and at operating temperatures up to 450 degreesC. After the contacts testing with current density of 10(3) A/cm(2) at temperatures up to 450 degreesC, their contact resistivity decreases slightly.
We investigated three 3C-SiC samples grown on 6H SiC substrate by sublimation epitaxy under gas atmosphere. We focus on the low temperature photoluminescence and Raman measurements to show that compare to a growth process under vacuum atmosphere, the gas atmosphere favor the incorporation of impurities at already existing and/or newly created defect sites.
300 μm thick 3C-SiC epilayer was grown on off-axis 4H-SiC(0001) substrate with a high growth rate of 1 mm/hour. Dry oxidation, wet oxidation and N2O anneal were applied to fabricate lateral MOS capacitors on these 3C-SiC layers. MOS interface obtained by N2O anneal has the lowest interface trap density of 3~4x1011 eV-1cm-2. Although all MOS capacitors still have positive net charges at the MOS interface, the wet oxidised sample has the lowest effective charge density of ~9.17x1011 cm-2.
Soon after the discovery of the problem with electrical degradation of bipolar SiC devices, we started to perform ab initio calculations in order to evaluate the hypothesis that the degradation is caused by the expansion of stacking faults (SF) created by the propagation of partial dislocations in the (0001) basal plane. These investigations have created a wealth of important information, and constitutes a major part of our present understanding of the degradation phenomenon. Salient features are: (1) In 3C-, 4H-, 6H-, and 15R-SiC there are one, two, three, and five structurally different SFs, respectively, with different properties. (2) In 4H-, 6H- and 15R-SiC two of the different types of SFs give rise to states with energies around 0.2 eV (0.1-0.15 eV in 15R) below the conduction band. These states extend along the SF plane but are strongly localized to within around 10 A in the direction perpendicular to the SF plane. (3) These states and their one-dimensional confinement can be interpreted in terms of a quantum-well whose depth is determined by the conduction band offset between the relevant polytype and 3C-SiC. (4) Very shallow, localized (gap) states appear in some cases and can be related to the change in electronic polarization induced by the SF. (5) Calculated SF energies (SFE) are very close to both measured values and to the predictions of the simpler ANNNI (axial next nearest neighbour Ising) model. (6) The SFE in 3C-SiC is negative. (7) In 6H-SiC, the SFE for one of the SFs is considerably larger than for the other two. (8) In 15R-SiC, the SFEs for two of the SFs are almost zero. (9) The localized states described in item 2 are, beyond reasonable doubt, responsible for the electrical degradation. We have also investigated the electronic properties of two (2SF), three (3SF), and four SFs (4SF) in neighbouring planes in 4H-SiC, leading to thin 3C-like inclusions. Especially double SFs (2SF) have been observed, and may also be present in degraded devices. For these systems, some salient features are: (1) Like in the case of an isolated SF, localized gap states in the upper part of the band gap appear. The number of bound states, their energies and wave function localizations are well described by a quantum-well model. (2) The electronic polarization of the host crystal gives rise to a clear displacement of the wave functions for the localized gap states. (3) The SFE for a second SF in the presence of an already existing one (i.e., the change in total energy in going from ISF to 2SF) is around a factor four less than the SFE for the first SF. This is compatible with recent experimental observations.
We present a theoretical investigation of how n- and p-type doping affect the band structure around the band gap of 3C-, 2H-, 4H-, and 6H-SiC. For comparison we also consider Si. We have calculated for various values of the dopant concentration (i) the shift in energy of the bottom (top) of the conduction (valence) band, (ii) the band gap narrowing, (iii) the shift of the optical band gap, and (iv) the doping-induced changes in conduction band curvature, i.e., changes in effective electron masses in n-type materials. In addition we have also (v) estimated the critical concentration for Mott transitions and (vi) calculated the shifts in conduction- and valence bands caused, not by doping, but by injection of an electron-hole plasma of various concentrations. To study the effects of doping we have considered a system consisting of impurity ions immersed in a (high-density) gas of majority carriers and a low-density gas of minority carriers. The changes in the bands relative to the idealised crystal are then regarded as being due to interparticle Coulomb interactions and associated particle correlation in and between the gases, as well as to electron and hole interactions with the randomly distributed ions. We have considered two models. The simplest model for band edge displacements is analytical and based on relatively simple assumptions like parabolic energy bands and simple modelling of electron correlation effects. The second model is numerical and includes full band non-parabolicity, and the electron and hole gas interactions are treated in the random-phase approximation.
A variable angle of incidence spectroscopic ellipsometer equipped with a compensator has been used to determine the dielectric functions in the 0.74 - 6 eV photon energy range of n-type bulk 4H-SiC with doping concentrations between 10(17) and 10(19) cm(-3). The resulting dielectric function for different SiC wafers depends on the doping concentration, especially around the absorption onset and higher photon energies. Measurements on different wafers with the same doping show good reproducibility. Simulations and preliminary measurements show that ellipsometry might be useful for thickness determination of thin (<1 m) homoepitaxial films.
Deuterium was introduced in p-type SiC from a gas ambient. The samples were partially coated with 200 Angstrom thick metal layer of titanium, nickel, platinum or gold. Heat treatments were performed in the temperature range 500-800 degreesC during 4 h. Secondary ion mass spectrometry (SIMS) was used to measure the deuterium content after deuterium exposure. The catalytic metal coating is shown to play an important role for introducing deuterium into SiC. Nickel and platinum facilitate hydrogen incorporation in p-type SiC, which may be due to an increased hydrogen concentration at the metal/SiC interface and/or an increase the H+ ions to H ratio. No in-diffusion of deuterium is observed using titanium although large quantities of deuterium are stored in the titanium film. Furthermore, gold reveals an inert character and does not promote in-diffusion of deuterium.
A chemical gas sensor based on a silicon carbide field effect transistor with a catalytic gate metal has been under development for a number of years. The buried gate design allows the sensor to operate at high temperatures, routinely up to 600degreesC and for at least three days at 700degreesC. The chemical inertness of silicon carbide makes it a suitable sensor technology for applications in corrosive environments such as exhaust gases and flue gases from boilers. The selectivity of the sensor devices is established through the choice of type and structure of the gate metal as well as the operation temperature. In this way NH3 sensors with low cross sensitivity to NOx have been demonstrated as potential sensors for control of selective catalytic reduction (SCR) of NOx by urea injection into diesel exhausts. The hardness of the silicon carbide makes it for example more resistant to water splash at cold start of a petrol engine than existing technologies, and a sensor which can control the air to fuel ratio, before the exhaust gases are heated, has been demonstrated. Silicon carbide sensors are also tested in flue gases from boilers. Efficient regulation of the combustion in a boiler will decrease fuel consumption and reduce emissions.
A three dimensional computational model for temperature and flow predictions in hot wall chemical vapor deposition (CVD) reactors, heated by induction, is presented. It includes heating by a Radio Frequency (RF) coil, flow and heat transfer. Thermal radiation is modeled by a modified Monte Carlo method. Model predictions are compared to full scale experiments at Linkoping CVD reactor for epitaxial growth of silicon carbide (SIC). Both streamwise and spanwise temperature gradients are well predicted, with the temperature maximum location shifted slightly upstream compared to the measured. Additionally, the model succeeds in predicting a recirculation zone just downstream of the susceptor. It is demonstrated how thermal gradients can be greatly reduced by simple geometrical changes.
Sublimation-grown 3C-SiC crystals were implanted with B ions at elevated temperature (400 °C) using multiple energies (100 to 575 keV) with a total dose of 1.3×1017 atoms/cm2 in order to form intermediate band (IB) in 3C-SiC. The samples were then annealed at 1400 °C for 60 min. An anomalous area in the center was observed in the PL emission pattern. The SIMS analysis indicated that the B concentration was the same both within and outside the anomalous area. The buried boron box-like concentration profile can reach ~3×1021 cm-3 in the plateau region. In the anomalous area a broad emission band (possible IB) emerges at around ~1.7-1.8 eV, which may be associated with B-precipitates having a sufficiently high density.
Single crystals of SiC were etched in hydrofluoric acid to remove the native oxide. Ozone exposure at room temperature is shown to give an oxide of about 0.7nm. The differences of interface and bulk oxides regarding their elemental composition and their oxidation states are reported as determined by photoelectron spectroscopy utilizing synchrotron radiation.
The drain current-voltage (I-d-V-D) characteristics of a chemical gas sensor based on a catalytic metal insulator silicon carbide field effect transistor (SiC-FET) were measured in H-2 or O-2 ambient while applying negative substrate bias, V-sub, at temperatures up to 600degreesC. An increase in the negative V-sub gives rise to an increase of the drain voltage at a given drain current level, which can be used to adjust the device baseline. In addition, we found that the difference in drain voltage between H-2 and O-2 ambient at a given drain current level (the gas response to H-2) increases for an increased negative substrate bias. By modifying an equation for the drain current in a SIT (static induction transistor), the influence of substrate bias on the amplification factors, mu and eta, was estimated using the temperature dependence of the I-d-V-D characteristics. From this, the effect of substrate bias on the gas response to hydrogen was calculated. It was clarified that the increase in the gas response caused by the negative substrate bias is due to a substrate bias dependence of the amplification factor of the short channel device.
We present thermally stimulated current (TSC) measurements made on metal-oxide-semiconductor (MOS) structures fabricated on off-axis (0001) or on-axis (1120) face n-type 4H-SiC with wet or dry oxides. The TSC measurements show the interface trap spectra of traps with activation energies in the range from 0.1 to 0.6 eV. Varying the charging and discharging conditions, we are able to distinguish between two types of traps which are both present on (0001) and (1120) face samples. One type is sensitive to the electric field during discharging but is insensitive to the charging temperature, while the other type is insensitive to the electric field during discharging but can not capture electrons at low temperatures. We find that, compared to the (0001) face, the traps at the (1120) face are shifted in energy about 0.1 eV towards higher activation energies. In all cases, For wet or dry oxides made on the (0001) or the (1120) face, the number density of traps is above 7x10(12) cm(-2).
Copper layers were deposited from acidic electrolytes containing different amounts of organic additives, designed for the formation of Cu-interconnect structures. Amorphous Ni-P substrates allow to study the unbiased growth of the electrodeposits. The crystallographic texture was investigated by the determination of X-ray diffraction (XRD) pole figures and the calculation of the orientation distribution functions. XRD results are discussed in relation to the morphologies of the electrodeposits as investigated with light optical microscopy and correlated with the process parameters during electrodeposition.
The deformation behaviour of four super duplex stainless steels of the grade SAF 2507 (UNS S32750) were studied by X-ray diffraction experiment with in-situ uniaxial tensile load. The steels had different nitrogen contents, between 0.2 and 0.33%, and/or different volume fractions of the ferrite, between 37% and 49%, in balance with austenite. The development of phase-specific stresses under external loading up to over 10% tensile strain was followed. The X-ray diffraction measurements revealed that load partitioning between the phases changed with increasing applied load, as the ferrite and austenite exhibited different deformation hardening behaviours. At the onset of macroscopic yielding and low plastic strains, a load transfer from γ to α occurred due to higher yield strength and strain hardening rate of the ferrite but vice versa at larger plastic strains when the austenite hardened more rapidly than the ferrite. It was also concluded that both the yield and tensile strengthen of the steels increased with increasing nitrogen content due to increased strengthen of the austenite by additional solid solution hardening, whereas a higher volume fraction of austenite contributed to higher tensile strength.
The deformation behaviour of the super duplex stainless steel SAF2507 (UNS S32750) under successive uniaxial tensile loading-unloading was investigated with respect to load sharing and inter-phase interactions. The steel consists of 58% austenite and 42% ferrite in volume. By insitu X-ray diffraction experiment the evolution of phase-specific stresses with applied load was monitored for three successive loading-unloading cycles with the maximum total strains being 0.34%, 0.75% and 1.63%, respectively. It was found that yielding occurred earlier in the austenitic phase than in the ferritic phase during the first loading cycle. In the followed loading cycles, both phases yielded under larger but similar applied stresses. Due to a similar behavior of the phases in the elasto-plastic regime inter-phase interactions were relatively weak. Low microstresses induced by the plastic straining resulted in somewhat larger stresses in the ferritic phase.
In-situ neutron diffraction experiments under tensile loading were carried out to study the micromechanical behaviour of two iron-manganese based steels, a TWIP (twinning induced plasticity) steel with 30 wt% Mn and a TRIP steel (transformation induced plasticity) with 20 wt% Mn. The former was loaded to 31.3% strain and the latter to 20% strain. The 30 wt.% Mn steel had a fully austenitic microstructure which remained stable over the loading range studied, while stress induced austenite to α´- and ε-martensite transformations occur in the 20 wt.% Mn steel which initially contained an α´-martensite in addition to the austenite. The evolution of lattice strains under tensile loading differs between the two steels, reflected their different plastic deformation mechanisms. A stronger grain-orientation dependent behaviour is observed during deformation for the 20 wt.% Mn in contrast to the 30wt.% Mn steel.
Microstresses due to intergranular and inter-phase interactions in an austenitic-ferritic super duplex steel (SAF 2507) under uniaxial compressive deformation have been studied by in-situ neutron diffraction experiments. Lattice strains of several hkl planes of austenite respective ferrite were mapped as a function of sample direction at a number of load levels during loading into the plastic regime and unloading. The analysis of the experimental results has shown that during loading both grain-orientation-dependent and inter-phase stresses were generated under plastic deformation that was inhomogeneous at the microstructural level. Residual stresses depending on the grain-orientation and phase have been found after unloading. The results also indicate stronger intergranular interactions among the studied hkl planes of austenite than those of ferrite.