Gertjan Koster

A suitable in situ monitoring technique for growth of thin films is reflection high energy electron diffraction (RHEED). Deposition techniques, like pulsed laser deposition (PLD) and sputter deposition, used for fabrication of complex oxide thin films use relatively high oxygen pressures (up to 100 Pa) and are, therefore, not compatible with ultrahigh vacuum RHEED equipment. We have developed a RHEED system which can be used for growth monitoring during the deposition of complex oxides at standard PLD conditions. We are able to increase the deposition pressure up to 50 Pa using a two-stage differential pumping system. Clear RHEED patterns are observable at these high pressures. The applicability of this system is demonstrated with the study of homoepitaxial growth of SrTiO3 as well as the heteroepitaxial growth of YBa2Cu3O7-δ on SrTiO3. Intensity oscillations of the RHEED reflections, indicating two-dimensional growth, are observed up to several tens of nanometers film thickness in both cases.

In thin films, new phases can be encountered near interfaces, whether it is the substrate–film interface or subsequent interfaces in the case of heterostructures. Both structural properties and surface morphology are a direct result of the thin film growth, controlled by deposition conditions and substrate properties, which in turn influence the electrical properties and determine their applicability in multilayer structures.

The fabrication of well-defined, atomically sharp substrate surfaces over a wide range of lattice parameters is reported, which is crucial for atomically regulated epitaxial growth of complex oxide heterostructures. By applying a framework for controlled selective wet etching of complex oxides on the stable rare-earth scandates (REScO3), apseudocubic = 0.394 – 0.404 nm, the large chemical sensitivity of REScO3 to basic solutions is exploited, which results in reproducible, single-terminated surfaces. Time-of-flight mass-spectroscopy measurements show that after wet etching the surfaces are predominantly ScO2 -terminated. Moreover, the morphology study of SrRuO3 thin-film growth gives no evidence for mixed termination. Therefore, it is concluded that the REScO3 surfaces are completely ScO2 -terminated.

As discovered by Ohtomo et al., a large sheet charge density with high mobility exists at the interface between SrTiO3 and LaAlO3. Based on transport, spectroscopic and oxygen-annealing experiments, we conclude that extrinsic defects in the form of oxygen vacancies introduced by the pulsed laser deposition process used by all researchers to date to make these samples is the source of the large carrier densities. Annealing experiments show a limiting carrier density. We also present a model that explains the high mobility based on carrier redistribution due to an increased dielectric constant.

We have studied the polar surface of singly terminated DyScOâ (110) crystals by reflective high-energy electron diffraction, surface x-ray diffraction and angle-resolved mass spectroscopy of recoiled ions. These techniques show that the surfaces are (1 x 1) reconstructed, which points to the absence of ordered cation vacancies at the surface. The best surfaces were obtained after a selective chemical wet

Inspired by the work of Ohtomo and Hwang in 2004, we shed new light on thin films of layered cuprate high-Tc superconductors (HTS). In principle all HTS materials consist of charged perovskite-like layers which in thin films can lead to polar discontinuities at the interfaces of different materials. The resulting charge redistribution has to occur but we expect it to be far more complex than in the LaAlO3/SrTiO3 system since copper can be multivalent. This makes it hard to predict what will happen in terms of transport or even magnetic properties compared to the 'simple' insulator LaAlO3. Nevertheless, we point out that the picture of systems of charged layers is important and necessary to fully understand heterostructures of these complex materials.

The perovskite SrTiO$_3$-LaAlO$_3$ structure has advanced to a model system to investigate the rich electronic phenomena arising at polar interfaces. Using first principles calculations and transport measurements we demonstrate that an additional SrTiO$_3$ capping layer prevents structural and chemical reconstruction at the LaAlO$_3$ surface and triggers the electronic reconstruction at a significantly lower LaAlO$_3$ film thickness than for the uncapped systems. Combined theoretical and experimental evidence (from magnetotransport and ultraviolet photoelectron spectroscopy) suggests two spatially separated sheets with electron and hole carriers, that are as close as 1 nm.

In order to understand the physical origin of the high charge carrier density at the conducting interface between SrTiO3 (STO) and LaAlO3 (LAO) the role of defects has to be clarified. In this study, LAO/STO-heterostructures modified in stacking sequence and growth conditions were investigated by means of high temperature conductance (HTC) measurements under changing oxygen ambience. Under measurement conditions the samples are in equilibrium with the surrounding oxygen atmosphere, which rules out the effect of mobile oxygen vacancies on the interface conductivity [Gunkel et al., APL 97(2010)]. The HTC characteristics show a significant dependency on the preparation procedure of STO and LAO close to the interface. Nevertheless, a common conduction and charge compensation mechanism can be identified. The results are discussed with respect to the defect chemistry model of perovskite oxides.

The manifestation of quantum behavior in two dimensional electron gases in semiconducting heterostructures and their progressive complexity towards fractional quantum Hall effect went hand-in-hand with the efforts to remove the effect of impurity scattering. For oxide materials, history is repeating itself and to date sample quality is reaching levels where quantum behavior starts to become accessible. To really understand the ground state of two dimensional electron gases in oxide systems, where electron-electron correlation effects seem more important, a step towards modulation doping is necessary, removing dopants away from a conduction channel. We will show that the impurity scattering of a 2DEG at the LaAlO3/SrTiO3 interfaces can be significantly suppressed by defect engineering, allowing the observation of quantum transport in a modulation doped oxide system.

Pulsed laser deposition (PLD) has become thin film deposition technique with increasing prominence. One of the advantages above other techniques is the possibility to growth at relative high background pressures, with a large freedom in choosing the kind of gas. An example is oxygen in the case of high Tc superconductors and giant magnetic resistors. However, the advantage of relative high pressures hinders the use of a number of diagnostics and monitor techniques, like reflective high energy electron diffraction (RHEED). With the introduction of the possibility to use RHEED at standard PLD pressures, it became possible to study the growth of oxide materials under different oxygen and temperature conditions. In this paper we employed this technique on SrTiO3, which can be grown in different growth modes depending on growth temperature and oxygen pressure during deposition. Applying a modified etch treatment on SrTiO3 single crystals, a real 2D growth mode could be observed by the homo-epitaxial growth of SrTiO3, as indicated by RHEED oscillations. In addition to the RHEED oscillations another phenomenon is observed, typical for PLD. The pulsed way of deposition leads to discontinuities in the intensity of the diffracted pattern. This is caused by the mobility of the deposited material from a disordered distribution till an ordered one and leads to a characteristic exponential slope with characteristic relaxation time constants. These time constants give extra information about relaxation, crystallization, and nucleation of the deposited material during growth. Finally, a new approach to deposit these complex oxide materials will be introduced. This, so-called interval deposition, is based upon the results obtained from the intensity oscillations as well as relaxations. The basic idea is to deposit an equivalent of one unit cell of material in such a short time that no coalescence in larger islands can occur, followed by a relaxation time before the next unit cell layer is deposited. This interval deposition leads to an imposed layer by layer growth.

We have grown and studied high quality SrRuO3 films grown by MBE as well as PLD. By changing the oxygen activity during deposition we were able to make SrRuO3 samples that were stoichiometric (low oxygen activity) or with ruthenium vacancies (high oxygen activity). Samples with strontium vacancies were found impossible to produce since the ruthenium would precipitate out as RuO2. The volume of the unit cell of SrRuO3 becomes larger as more ruthenium vacancies are introduced. The residual resistivity ratio (RRR) and room temperature resistivity were found to systematically depend on the volume of the unit cell and therefore on the amount of ruthenium vacancies. The RRR varied from ~30 for stoichiometric samples to less than two for samples that were very ruthenium poor. The room temperature resistivity varied from 190 microOhm cm for stoichoimetric samples to over 300 microOhm cm for very ruthenium poor samples. UPS spectra show a shift of weight from the coherent peak to the incoherent peak around the Fermi level when samples have more ruthenium vacancies. Core level XPS spectra of the ruthenium 3d lines show a strong screened part in the case of stoichiometric samples. This screened part disappears when ruthenium vacancies are introduced. Both the UPS and the XPS results are consistent with the view that correlation increases as the amount of ruthenium vacancies increase.

We present a study of the lattice response to the compressive and tensile biaxial stress in La0.67Sr0.33MnO3 (LSMO) and SrRuO3 (SRO) thin films grown on a variety of single crystal substrates: SrTiO3, DyScO3, NdGaO3 and (La,Sr)(Al,Ta)O3. The results show, that in thin films under misfit strain, both SRO and LSMO lattices, which in bulk form have orthorhombic (SRO) and rhombohedral (LSMO) structures, assume unit cells that are monoclinic under compressive stress and tetragonal under tensile stress. The applied stress effectively modifies the BO6 octahedra rotations, which degree and direction can be controlled by magnitude and sign of the misfit strain. Such lattice distortions change the B-O-B bond angles and therefore are expected to affect magnetic and electronic properties of the ABO3 perovskites.

In recent years, well-defined and nearly perfect single crystal surfaces of oxide perovskites have become increasingly important. A single terminated surface is a prerequisite for reproducible thin film growth and fundamental growth studies. In this work, atomic and lateral force microscopy have been used to display different terminations of SrTiO3. We observe hydroxylation of the topmost SrO layer after immersion of SrTiO3 in water, which is used to enhance the etch-selectivity of SrO relative to TiO2 in a buffered HF solution. We reproducibly obtain perfect and single terminated surfaces, irrespective of the initial state of polished surfaces and the pH value of the HF solution. This approach to the problem might be used for a variety of multi-component oxide single crystals. True two-dimensional reflection high-energy electron diffraction intensity oscillations are observed during homo epitaxial growth using pulsed laser deposition on these surfaces.

Monoclinic CuO is anomalous both structurally as well as electronically in the 3$d$ transition metal oxide series. All the others have the cubic rock salt structure. Here we report the synthesis and electronic property determination of a tetragonal (elongated rock salt) form of CuO created using an epitaxial thin film deposition approach. In situ photoelectron spectroscopy suggests an enhanced charge transfer gap $\Delta$ with the overall bonding more ionic. As an end member of the 3d transition monoxides, its magnetic properties should be that of a high $T_N$ antiferromagnet.

We have studied the polar surface of singly terminated DyScOâ (110) crystals by reflective high-energy electron diffraction, surface x-ray diffraction and angle-resolved mass spectroscopy of recoiled ions. These techniques show that the surfaces are (1 x 1) reconstructed, which points to the absence of ordered cation vacancies at the surface. The best surfaces were obtained after a selective chemical wet