Summary thesis Mark Huijben (English)

Perovskite oxides and structurally related compounds are receiving much attention in present-day materials science research. They comprise of a broad range of interesting electronic phases such as superconductors, metals, semiconductors, insulators, ferromagnets, antiferromagnets, ferroelectrics, multiferroics, dielectrics and piezoelectrics. This diversity in material properties has led in the last decades to an extensive amount of research into ‘oxide electronics’. In the last couple of years the focus of research has been directed into the fabrication of heteroepitaxial structures to explore new physical phenomena and new device concepts. This stacking of several different material compounds on top of each other is possible because of the close lattice-matching of the unit cells of the individual materials. Very small characteristic length scales of the order 0.1 - 1 nm determine the nature of the physical properties in oxide electronics. Therefore, a growth control with a precision on the atomic level is essential for novel epitaxial heterostructures.

For complex oxides, pulsed laser deposition (PLD) has proven to be a growth technique in which the deposited material can be controlled at the atomic scale. Stoichiometric transfer, high deposition rate and tunable energy of the arriving particles are properties, which make 2-dimensional layer-by-layer growth feasible for various complex oxides with a surface roughness of only one unit cell. This makes fabrication of thin films, heterostructures and superlattices possible with high quality surfaces and interfaces, which can be applied in electronic devices. However, the combination of pulsed laser deposition with reflection high-energy electron diffraction (RHEED), to monitor in-situ the thin film growth, is essential to obtain this high quality. Monitoring of the RHEED intensity provides information about the surface morphology as well as the diffusion and nucleation processes. This atomically controlled growth can only be fully exploited when the initial substrate crystals are prepared to have atomically flat surfaces and single unit cell steps. In this way atomically abrupt interfaces can be produced with a strong epitaxial relation between the two oxide layers.

The main aim of this thesis is to develop a controlled growth with atomic precision for the realization of artificial perovskite structures, to exploit the exceptional physical properties of complex oxide materials such as high-temperature superconductors and conducting interfaces between band insulators.

The superconducting La_2-xSr_xCuO₄₊_d compound was found to be an ideal material to study the influence of small variations in the stoichiometry on the electronic properties, because small alterations of the strontium and oxygen doping resulted in dramatic changes in the conductivity behavior. Variations in the strontium doping level (x=0, 0.05, 0.125, 0.15 and 0.25) and the background gas, oxygen (d=0) or ozone (d>0), provided the possibility to tune the electronic properties from insulating until superconducting, while maintaining the same tetragonal perovskite crystal structure.

The growth of La_2-xSr_xCuO₄₊_d thin films was dependent on the termination of the SrTiO₃ substrate surface. Experiments indicated that during the initial growth on the substrate surface the first atomic layers formed a rock-salt structure. In case of growth on a TiO₂-terminated surface, the film started with two (La,Sr)O layers, while for a SrO-terminated surface a single (La,Sr)O layer was formed first. Subsequent growth occurred in a layer-by-layer mode of half unit cell monolayers up to large thicknesses (~100 nm). The final surface roughness of the film remained comparable to the initial substrate surface with smooth terraces separated by half unit cell steps. For the case of a thin film on a SrO-terminated substrate however, the formation of very small precipitates on the surface was observed. The maximum superconducting transition temperature for optimum strontium doping, without extra oxygen doping, was found to be 26 K. This was lower than the bulk value of 38 K, due to the large tensile strain present in the thin films. However, when additional oxygen doping was used, the superconducting transition temperature was as high as the bulk value at optimum strontium doping and superconductivity was present in the underdoped region even down to the undoped case.

The ability to maintain the same crystal structure in the investigated doping region provided the possibility to fabricate multilayer structures of thin layers with variable doping levels. In this way a suppression of the superconducting transition temperature, due to the large epitaxial strain between the La_2‑xSr_xCuO₄₊_d thin films and the SrTiO₃ substrate, was found to exist over thicknesses up to 30 unit cells. On the other hand, the multilayer structures were used to investigate ultrathin superconducting films, where a superconducting transition temperature of 7.5 K was observed for only 5 unit cells.

Various difficulties, connected with the sensitivity of La_2‑xSr_xCuO₄₊_d to oxidation, complicated the fabrication of electronic devices. However, this sensitivity of the electronic properties on the carrier doping was used very successfully in the improvement of the fabrication process of electric field effect structures.

In the heteroepitaxial growth of the superconducting YBa₂Cu₃O_7-_d compound the atomic stacking sequence at the substrate–film interface played an essential role. During initial growth, the atomic interface configuration influenced the surface morphology and structural properties of the film, due to the formation of anti-phase boundaries by coalescence of islands with different stacking sequences. The interface configuration was accurately controlled by both the terminating atomic layer of the SrTiO₃ substrate and the stoichiometry of the first unit cell layer. Using this capability the network of anti-phase boundaries and, therefore, the in-plane ordering was tuned, allowing the study of its influence on the structural and electrical properties of the YBa₂Cu₃O_7-_d film. The superconducting transition temperature was found to be reduced by improvement of the in-plane ordering, which indicated that the absence of anti-phase boundaries hampered the oxygen in-diffusion.

Due to the small lattice mismatch between YBa₂Cu₃O_7-_d and SrTiO₃ the first deposited atomic layers were strained to match the substrate. However, above a critical thickness for this strained pseudomorphic layer the strain relaxed by the introduction of defects. The initial defect density, which was already present at the substrate-film interface, determined the value for this critical thickness during subsequent growth, because the process of strain relaxation was improved by the number of antiphase boundaries. This engineering of the interface between YBa₂Cu₃O_7-_d and SrTiO₃ was analyzed in-situ during growth, besides reflection high-energy electron diffraction, by surface X-ray diffraction.

Sub-unit cell layer epitaxy was investigated as a novel growth technique for the artificial growth of perovskite structures. Using this technique new crystals were formed by sequentially depositing its sub-unit cell layers. Initial results demonstrated the fabrication of crystalline films with low roughness surfaces, while further investigations are necessary to improve the superconducting transition temperature.

The electronic properties of the heteroepitaxial interfaces between the band-insulators LaAlO₃ and SrTiO₃ were found to be very sensitive to the atomic stacking sequence. An accurate control of the atomic configuration at the interface tuned the conductivity of the 2-dimensional system between metallic and insulating. Whereas SrTiO₃ and LaAlO₃ are seemingly similar, the Sr²⁺O^2- and Ti⁴⁺O^2-₂ layers were charge-neutral, while in the ionic limit the charge states in the LaAlO₃ were positive for La³⁺O^2- and negative for Al³⁺O^2-₂. The polarity discontinuity at the interfaces led to a metallic conductivity for the ‘n-type’ LaO-TiO₂ interface, where by electronic reconstruction through mixed valence Ti states extra electrons were placed in the SrTiO₃ conduction band. In the other case, the AlO₂-SrO interface was found to be insulating. Still ‘p-charging’ is conceivable, which will result most likely in an atomic reconstruction by the introduction of oxygen vacancies.

To grow both types of interfaces with an atomic control, the surfaces of the SrTiO₃ substrates had to be single terminated by either TiO₂ or SrO. The first was achieved by a chemical and thermal treatment, while the latter was obtained by the deposition of an extra monolayer of SrO. In both cases the subsequent growth of LaAlO₃ was perfectly similar and resulted in a highly crystalline thin film with a surface of smooth terraces separated by unit cell steps. The atomic ordering at the heteroepitaxial interface was investigated also in-situ directly after growth by surface X-ray diffraction to obtain structural information at high deposition temperatures without the influences of possible strain relaxation and impurity incorporation. A single unit cell of LaAlO₃ on a TiO₂-terminated SrTiO₃ substrate was found to adapt to the cubic bulk structure of the substrate at high deposition temperatures, while for lower temperatures the atoms in the LaAlO₃ unit cell are displaced from their bulk positions.

The strong epitaxial relation was used to grow superlattices, which were composed of alternating ‘n-type’ and ‘p-type’ interfaces. The observed oscillations in the RHEED monitoring during growth were used to accurately control the thicknesses of the individual layers. Structural analysis indicated a high quality ordering, in both the crystallinity of the total structure as well as the imposed periodicity along the c-axis. The atoms in the SrTiO₃ layers were found to be in their bulk positions, while the c-axis parameter of the LaAlO₃ unit cells was shortened due to strain. The expected atomic ordering at both the LaO-TiO₂ and the AlO₂-SrO interfaces was proven by high-quality scanning transmission electron microscopy. In order to investigate the influence of the layer thicknesses on the abruptness of the interfaces, a multilayer was fabricated with variable thicknesses of the individual LaAlO₃ and SrTiO₃ layers. In this way, the thickness was reduced down to a single unit cell, while maintaining a high-quality atomic ordering.

The heteroepitaxial interfaces between the band insulators LaAlO₃ and SrTiO₃ were either metallic or insulating depending on the atomic stacking sequences. The charge carriers in these two-dimensional systems were determined to be electrons at the conducting LaO-TiO₂ interface with high carrier mobilities up to ~1000 cm² V^‑1 s^‑1 at low temperatures. The complementary AlO₂-SrO interface was found to be insulating. The conductivity properties were found to be independent of the orientation with respect to the terrace step direction of the substrate. The negligible influence of step heights of one unit cell suggested a thickness for the interface conductivity of a couple of unit cells. However, large variations in the electronic properties with orientation were observed when the atomic ordering at the interfaces was not perfect.

All heteroepitaxial LaAlO₃/SrTiO₃ interfaces showed photoconductivity due to photo carrier injection. Measurements of the transmittance and reflectance of the two types of interfaces at variable wavelengths resulted in a clear observation of the absorption edges for LaAlO₃ and SrTiO₃. However, no differences were observed between the metallic and the insulating interfaces. An abrupt increase in the conductance was observed for wavelengths below ~380 nm. This change at precisely the bandgap of SrTiO₃ suggested an intrinsic optical absorption, and as a result the rising of an electron from the valence band to the conduction band in the SrTiO₃ substrate without any influence of the LaAlO₃-SrTiO₃ interfaces. High-intensity ultraviolet light illumination on single LaAlO₃ layers on SrTiO₃ substrates increased the conductivity by factors of 4 and 12000 for the LaO-TiO₂ and AlO₂-SrO interfaces, respectively. The response times to a higher conductance, when the interfaces were illuminated, were very short. In case of a AlO₂-SrO interface even below 0.1 second. Still, the response time to a lower conductance, when the ultraviolet light was removed, was very long. It took hours before this process reached equilibrium.

To study the electronic coupling of these complementary interfaces, high-quality heterostructures were fabricated in which a variable number of LaAlO₃ unit cell layers were stacked between SrTiO₃, and vice versa. The sheet resistances and sheet carrier densities at room temperature, for both types of heterostructures, were found to change below a critical separation distance of 6 unit cells, corresponding to 23 Å. Both types of heterostructures showed a similar dependence on the interface separation distance, although there was a small difference in the absolute values.

The temperature dependence of the electronic properties provided further insight into coupled complementary interfaces. For the SrTiO₃/LaAlO₃/SrTiO₃ heterostructures, the energy-scale over which charge carriers seemed to be frozen out was 6.0 meV, which was comparable to observations in SrTiO₃ at low La-doping. At temperatures above 100 K, the carrier density was approximately constant for a separation distance of 2 unit cells, while the thermally activated increase continued for a separation distance of 5 unit cells. The T^-2 power law dependence in the carrier mobilities, observed above 50 K, suggested electron-electron scattering, which was known to be relevant in transition metals with partially filled d-shells. For all heterostructures, the carrier mobilities at room temperature were found to be constant at 6.0 cm² V^-1 s^-1, without any dependence on the separation distances. Impurities or crystalline defects in the nearby interface were not dominating effects, because of the mean free path of 100 nm – 1 mm, which was estimated from the very large mobilities in the zero-temperature limit. Interestingly, these high carrier mobilities at low temperatures, characterizing the separate electron doped interfaces, were found to be maintained in coupled structures down to sub-nanometer interface spacing.

These different research topics demonstrate the importance of a controlled growth with an atomic level precision. To develop novel oxide electronic devices for future applications, the fundamental growth processes have to be studied down to the atomic level. This will be essential to reach the high-quality levels, which are needed for the implementation of ‘oxide electronics’ in daily life. An important advantage of such a highly controllable growth is the fact that it opens the door to a world of new fascinating phenomena.

	Home > ... > summary's > Summary thesis Mark Huijben (English)
Home News Events Strategic Orientations research groups Publications PhD students Education MESA+ NanoLab Starting entrepreneurs Tech Park Industrial Partners Vacancies Links Sitemap Search	Summary thesis Mark Huijben (English) Perovskite oxides and structurally related compounds are receiving much attention in present-day materials science research. They comprise of a broad range of interesting electronic phases such as superconductors, metals, semiconductors, insulators, ferromagnets, antiferromagnets, ferroelectrics, multiferroics, dielectrics and piezoelectrics. This diversity in material properties has led in the last decades to an extensive amount of research into ‘oxide electronics’. In the last couple of years the focus of research has been directed into the fabrication of heteroepitaxial structures to explore new physical phenomena and new device concepts. This stacking of several different material compounds on top of each other is possible because of the close lattice-matching of the unit cells of the individual materials. Very small characteristic length scales of the order 0.1 - 1 nm determine the nature of the physical properties in oxide electronics. Therefore, a growth control with a precision on the atomic level is essential for novel epitaxial heterostructures. For complex oxides, pulsed laser deposition (PLD) has proven to be a growth technique in which the deposited material can be controlled at the atomic scale. Stoichiometric transfer, high deposition rate and tunable energy of the arriving particles are properties, which make 2-dimensional layer-by-layer growth feasible for various complex oxides with a surface roughness of only one unit cell. This makes fabrication of thin films, heterostructures and superlattices possible with high quality surfaces and interfaces, which can be applied in electronic devices. However, the combination of pulsed laser deposition with reflection high-energy electron diffraction (RHEED), to monitor in-situ the thin film growth, is essential to obtain this high quality. Monitoring of the RHEED intensity provides information about the surface morphology as well as the diffusion and nucleation processes. This atomically controlled growth can only be fully exploited when the initial substrate crystals are prepared to have atomically flat surfaces and single unit cell steps. In this way atomically abrupt interfaces can be produced with a strong epitaxial relation between the two oxide layers. The main aim of this thesis is to develop a controlled growth with atomic precision for the realization of artificial perovskite structures, to exploit the exceptional physical properties of complex oxide materials such as high-temperature superconductors and conducting interfaces between band insulators. The superconducting La_2-xSr_xCuO₄₊_d compound was found to be an ideal material to study the influence of small variations in the stoichiometry on the electronic properties, because small alterations of the strontium and oxygen doping resulted in dramatic changes in the conductivity behavior. Variations in the strontium doping level (x=0, 0.05, 0.125, 0.15 and 0.25) and the background gas, oxygen (d=0) or ozone (d>0), provided the possibility to tune the electronic properties from insulating until superconducting, while maintaining the same tetragonal perovskite crystal structure. The growth of La_2-xSr_xCuO₄₊_d thin films was dependent on the termination of the SrTiO₃ substrate surface. Experiments indicated that during the initial growth on the substrate surface the first atomic layers formed a rock-salt structure. In case of growth on a TiO₂-terminated surface, the film started with two (La,Sr)O layers, while for a SrO-terminated surface a single (La,Sr)O layer was formed first. Subsequent growth occurred in a layer-by-layer mode of half unit cell monolayers up to large thicknesses (~100 nm). The final surface roughness of the film remained comparable to the initial substrate surface with smooth terraces separated by half unit cell steps. For the case of a thin film on a SrO-terminated substrate however, the formation of very small precipitates on the surface was observed. The maximum superconducting transition temperature for optimum strontium doping, without extra oxygen doping, was found to be 26 K. This was lower than the bulk value of 38 K, due to the large tensile strain present in the thin films. However, when additional oxygen doping was used, the superconducting transition temperature was as high as the bulk value at optimum strontium doping and superconductivity was present in the underdoped region even down to the undoped case. The ability to maintain the same crystal structure in the investigated doping region provided the possibility to fabricate multilayer structures of thin layers with variable doping levels. In this way a suppression of the superconducting transition temperature, due to the large epitaxial strain between the La_2‑xSr_xCuO₄₊_d thin films and the SrTiO₃ substrate, was found to exist over thicknesses up to 30 unit cells. On the other hand, the multilayer structures were used to investigate ultrathin superconducting films, where a superconducting transition temperature of 7.5 K was observed for only 5 unit cells. Various difficulties, connected with the sensitivity of La_2‑xSr_xCuO₄₊_d to oxidation, complicated the fabrication of electronic devices. However, this sensitivity of the electronic properties on the carrier doping was used very successfully in the improvement of the fabrication process of electric field effect structures. In the heteroepitaxial growth of the superconducting YBa₂Cu₃O_7-_d compound the atomic stacking sequence at the substrate–film interface played an essential role. During initial growth, the atomic interface configuration influenced the surface morphology and structural properties of the film, due to the formation of anti-phase boundaries by coalescence of islands with different stacking sequences. The interface configuration was accurately controlled by both the terminating atomic layer of the SrTiO₃ substrate and the stoichiometry of the first unit cell layer. Using this capability the network of anti-phase boundaries and, therefore, the in-plane ordering was tuned, allowing the study of its influence on the structural and electrical properties of the YBa₂Cu₃O_7-_d film. The superconducting transition temperature was found to be reduced by improvement of the in-plane ordering, which indicated that the absence of anti-phase boundaries hampered the oxygen in-diffusion. Due to the small lattice mismatch between YBa₂Cu₃O_7-_d and SrTiO₃ the first deposited atomic layers were strained to match the substrate. However, above a critical thickness for this strained pseudomorphic layer the strain relaxed by the introduction of defects. The initial defect density, which was already present at the substrate-film interface, determined the value for this critical thickness during subsequent growth, because the process of strain relaxation was improved by the number of antiphase boundaries. This engineering of the interface between YBa₂Cu₃O_7-_d and SrTiO₃ was analyzed in-situ during growth, besides reflection high-energy electron diffraction, by surface X-ray diffraction. Sub-unit cell layer epitaxy was investigated as a novel growth technique for the artificial growth of perovskite structures. Using this technique new crystals were formed by sequentially depositing its sub-unit cell layers. Initial results demonstrated the fabrication of crystalline films with low roughness surfaces, while further investigations are necessary to improve the superconducting transition temperature. The electronic properties of the heteroepitaxial interfaces between the band-insulators LaAlO₃ and SrTiO₃ were found to be very sensitive to the atomic stacking sequence. An accurate control of the atomic configuration at the interface tuned the conductivity of the 2-dimensional system between metallic and insulating. Whereas SrTiO₃ and LaAlO₃ are seemingly similar, the Sr²⁺O^2- and Ti⁴⁺O^2-₂ layers were charge-neutral, while in the ionic limit the charge states in the LaAlO₃ were positive for La³⁺O^2- and negative for Al³⁺O^2-₂. The polarity discontinuity at the interfaces led to a metallic conductivity for the ‘n-type’ LaO-TiO₂ interface, where by electronic reconstruction through mixed valence Ti states extra electrons were placed in the SrTiO₃ conduction band. In the other case, the AlO₂-SrO interface was found to be insulating. Still ‘p-charging’ is conceivable, which will result most likely in an atomic reconstruction by the introduction of oxygen vacancies. To grow both types of interfaces with an atomic control, the surfaces of the SrTiO₃ substrates had to be single terminated by either TiO₂ or SrO. The first was achieved by a chemical and thermal treatment, while the latter was obtained by the deposition of an extra monolayer of SrO. In both cases the subsequent growth of LaAlO₃ was perfectly similar and resulted in a highly crystalline thin film with a surface of smooth terraces separated by unit cell steps. The atomic ordering at the heteroepitaxial interface was investigated also in-situ directly after growth by surface X-ray diffraction to obtain structural information at high deposition temperatures without the influences of possible strain relaxation and impurity incorporation. A single unit cell of LaAlO₃ on a TiO₂-terminated SrTiO₃ substrate was found to adapt to the cubic bulk structure of the substrate at high deposition temperatures, while for lower temperatures the atoms in the LaAlO₃ unit cell are displaced from their bulk positions. The strong epitaxial relation was used to grow superlattices, which were composed of alternating ‘n-type’ and ‘p-type’ interfaces. The observed oscillations in the RHEED monitoring during growth were used to accurately control the thicknesses of the individual layers. Structural analysis indicated a high quality ordering, in both the crystallinity of the total structure as well as the imposed periodicity along the c-axis. The atoms in the SrTiO₃ layers were found to be in their bulk positions, while the c-axis parameter of the LaAlO₃ unit cells was shortened due to strain. The expected atomic ordering at both the LaO-TiO₂ and the AlO₂-SrO interfaces was proven by high-quality scanning transmission electron microscopy. In order to investigate the influence of the layer thicknesses on the abruptness of the interfaces, a multilayer was fabricated with variable thicknesses of the individual LaAlO₃ and SrTiO₃ layers. In this way, the thickness was reduced down to a single unit cell, while maintaining a high-quality atomic ordering. The heteroepitaxial interfaces between the band insulators LaAlO₃ and SrTiO₃ were either metallic or insulating depending on the atomic stacking sequences. The charge carriers in these two-dimensional systems were determined to be electrons at the conducting LaO-TiO₂ interface with high carrier mobilities up to ~1000 cm² V^‑1 s^‑1 at low temperatures. The complementary AlO₂-SrO interface was found to be insulating. The conductivity properties were found to be independent of the orientation with respect to the terrace step direction of the substrate. The negligible influence of step heights of one unit cell suggested a thickness for the interface conductivity of a couple of unit cells. However, large variations in the electronic properties with orientation were observed when the atomic ordering at the interfaces was not perfect. All heteroepitaxial LaAlO₃/SrTiO₃ interfaces showed photoconductivity due to photo carrier injection. Measurements of the transmittance and reflectance of the two types of interfaces at variable wavelengths resulted in a clear observation of the absorption edges for LaAlO₃ and SrTiO₃. However, no differences were observed between the metallic and the insulating interfaces. An abrupt increase in the conductance was observed for wavelengths below ~380 nm. This change at precisely the bandgap of SrTiO₃ suggested an intrinsic optical absorption, and as a result the rising of an electron from the valence band to the conduction band in the SrTiO₃ substrate without any influence of the LaAlO₃-SrTiO₃ interfaces. High-intensity ultraviolet light illumination on single LaAlO₃ layers on SrTiO₃ substrates increased the conductivity by factors of 4 and 12000 for the LaO-TiO₂ and AlO₂-SrO interfaces, respectively. The response times to a higher conductance, when the interfaces were illuminated, were very short. In case of a AlO₂-SrO interface even below 0.1 second. Still, the response time to a lower conductance, when the ultraviolet light was removed, was very long. It took hours before this process reached equilibrium. To study the electronic coupling of these complementary interfaces, high-quality heterostructures were fabricated in which a variable number of LaAlO₃ unit cell layers were stacked between SrTiO₃, and vice versa. The sheet resistances and sheet carrier densities at room temperature, for both types of heterostructures, were found to change below a critical separation distance of 6 unit cells, corresponding to 23 Å. Both types of heterostructures showed a similar dependence on the interface separation distance, although there was a small difference in the absolute values. The temperature dependence of the electronic properties provided further insight into coupled complementary interfaces. For the SrTiO₃/LaAlO₃/SrTiO₃ heterostructures, the energy-scale over which charge carriers seemed to be frozen out was 6.0 meV, which was comparable to observations in SrTiO₃ at low La-doping. At temperatures above 100 K, the carrier density was approximately constant for a separation distance of 2 unit cells, while the thermally activated increase continued for a separation distance of 5 unit cells. The T^-2 power law dependence in the carrier mobilities, observed above 50 K, suggested electron-electron scattering, which was known to be relevant in transition metals with partially filled d-shells. For all heterostructures, the carrier mobilities at room temperature were found to be constant at 6.0 cm² V^-1 s^-1, without any dependence on the separation distances. Impurities or crystalline defects in the nearby interface were not dominating effects, because of the mean free path of 100 nm – 1 mm, which was estimated from the very large mobilities in the zero-temperature limit. Interestingly, these high carrier mobilities at low temperatures, characterizing the separate electron doped interfaces, were found to be maintained in coupled structures down to sub-nanometer interface spacing. These different research topics demonstrate the importance of a controlled growth with an atomic level precision. To develop novel oxide electronic devices for future applications, the fundamental growth processes have to be studied down to the atomic level. This will be essential to reach the high-quality levels, which are needed for the implementation of ‘oxide electronics’ in daily life. An important advantage of such a highly controllable growth is the fact that it opens the door to a world of new fascinating phenomena.