Diss. (sammanfattning) Uppsala : Univ., 2001. / Härtill 8 uppsatser.

Im Fokus der vorliegenden Arbeit liegt das Widerstandsschalten von HfO2- und TiO2-Dünnschichten, wobei insbesondere auf den Einfluss der Kristallstruktur, der Stöchiometrie und der Elektrodenmaterialien ((inert, Pt) und (reaktiv, Ti/Pt)) eingegangen wurde. Die HfO2-Dünnschichten wurden mittels Atomlagenabscheidung (ALD) durch die Reaktion von Tetrakis(dimethylamido)hafnium mit Wasser bei Temperaturen von 100°C bis 350°C abgeschieden. Das beste ALD-Wachstum mit der geringsten Konzentration an Wasserstoff wurde für eine Temperatur von 300°C erhalten. Sowohl eine geringere als auch eine höhere Abscheidetemperatur führte durch parasitäre CVD-artige Prozesse bzw. durch die ungewollte thermische Zersetzung des Präkursors zu einer größeren Wasserstoffkonzentration. Des Weiteren wurde eine Korrelation zwischen der Wachstumsrate pro Zyklus, der Schichtdickenhomogenität und der Wasserstoffkonzentration in den HfO2-Schichten festgestellt. Das Widerstandsschalten wurde an amorphen (150°C) und polykristallinen (300°C) HfO2-Schichten mit unstrukturierten Rückelektroden und kreisförmigen Vorderseitenelektroden untersucht. Die Verwendung von symmetrischen Pt-Pt-Elektroden führte bei beiden Schichttypen zum unipolaren Schaltmodus. Der unipolare Schaltmodus wurde ebenfalls bei Verwendung von asymmetrischen Pt-Ti/Pt-Elektroden im Zusammenspiel mit der amorphen HfO2-Phase beobachtet. Eine Ausnahme stellt die Kombination von asymmetrischen Elektroden mit der polykristallinen HfO2-Phase dar. Dabei wurde nach der Elektroformierung mit positiv angelegter Spannung an die Vorderseitenelektrode der bipolare Schaltmodus erhalten. Eine Erklärung dafür liefert die Betrachtung der Filament-Wachstumsrichtung während der Elektroformierung. Die TiO2-Dünnschichten wurden durch reaktives Magnetronsputtern gewachsen. Die Stöchiometrie dieser Schichten wurde durch die Bestrahlung mit Ar-Ionen gezielt verändert. Dabei führt das bevorzugte Herausschlagen von Sauerstoff im Vergleich zum Titan zu einer an Sauerstoff verarmten TiOx-Schicht, welche sich durch die verwendete Ionenenergie von 2keV in oberflächennahen Bereichen befindet. Die Ausdehnung der TiOx-Schicht wurde mit einer TRIDYN-Simulation auf ca. 4nm bestimmt. Während sich zwischen einer TiO2-Schicht und einer Pt-Elektrode ein Schottky-Kontakt ausbildet, führt das Einbringen einer TiOx-Schicht zu einem ohmschen Kontakt. Für die Charakterisierung des Widerstandsschaltens an den mit Ar-Ionen bestrahlten TiO2-Schichten wurde somit auf symmetrische Pt-Pt-Elektroden zurückgegriffen. Im Bereich der getesteten Fluenzen von 1 x 10^13 Ar+/cm2 bis 1 x 10^16 Ar+/cm2 wurde mit einer Fluenz von 1 x 10^14 Ar+/cm2 sowohl die beste Ausbeute als auch die größten Endurance-Werte erzielt. / The work focuses on the resistive switching of HfO2 and TiO2 thin films. Especially the influence of crystal structure, stoichiometry and electrode materials ((inert, Pt) and (reactive, Ti/Pt)) have been examined. HfO2 thin films have been grown at temperatures ranging from 100°C to 350°C through the process of atomic layer deposition (ALD) by the reaction of tetrakis(dimethylamido)-hafnium with water. The best ALD growth with the lowest concentration of hydrogen was achieved at a temperature of 300°C. Both lower and higher deposition temperatures caused higher hydrogen concentrations due to parasitic CVD-like processes or thermal decomposition of the precursor. Moreover a correlation between the growth rate per cycle, the layer thickness uniformity and the hydrogen concentration in the HfO2 films was observed. Resistive switching was examined in amorphous (150°C) and polycrystalline (300°C) HfO2 films with unstructured bottom electrodes and circular structured top electrodes. The use of symmetric Pt-Pt-electrodes caused the unipolar switching mode in both layer types. The unipolar switching mode was also observed when asymmetric Pt-Ti/Pt-electrodes were used with an amorphous phase of the HfO2 layer. An exception is the use of asymmetric electrodes with the polycrystalline phase of the HfO2 layer. In this case electroforming with the application of positive voltage to the top electrode resulted in the bipolar switching mode. This is explained when looking at the filament growth direction during electroforming. TiO2 thin films were grown by reactive magnetron sputtering. The stoichiometry of these layers was modified by irradiation with Ar-ions. The preferential sputtering of oxygen compared to titanium causes a surface-near oxygen deficient TiOx layer due to the used ion energy of 2 keV. The depth of the TiOx layer was estimated to be 4 nm by using a TRIDYN simulation. While a Schottky contact formed between a TiO2 layer and a Pt-electrode, the use of a TiOx layer led to an ohmic contact. Symmetric Pt-Pt-electrodes were used to characterize resistive switching of TiO2 layers which have been irradiated with Ar-ions. The tested fluences ranged from 1×10^13 Ar+/cm2 to 1 × 10^16 Ar+/cm2. A fluence of 1 × 10^14 Ar+/cm2 resulted in the best yield and highest endurance.

The change of the spontaneous polarization due to a change of temperature is known as the pyroelectric effect and is restricted to crystalline, non-centrosymmetric and polar matter. Its main application is the utilization in infrared radiation sensors, but usage for waste heat energy harvesting or chemical catalysis is also possible. A precise quantification, i.e. the measurement of the pyroelectric coefficient p, is inevitable to assess the performance of a material. Hence, a comprehensive overview is provided in this work, which summarizes and evaluates the available techniques to characterize p. A setup allowing the fully automated measurement of p by utilizing the Sharp-Garn method and the measurement of ferroelectric hysteresis loops is described. It was used to characterize and discuss the behavior of p with respect to the temperature of the doped bulk III-V compound semiconductors gallium nitride and aluminum nitride and thin films of doped hafnium oxide, as reliable data for these materials is still missing in the literature. Here, the nitride-based semiconductors show a comparable small p and temperature dependency, which is only slightly affected by the incorporated dopant, compared to traditional ferroelectric oxides. In contrast, p of HfO2 thin films is about an order of magnitude larger and seems to be affected by the present dopant and its concentrations, as it is considered to be responsible for the formation of the polar orthorhombic phase.:1. Motivation and Introduction 2. Fundamentals 2.1. Dielectrics and their Classification 2.2. Polarization 2.3. Pyroelectricity 2.4. Ferroelectricty 2.5. Phase Transitions 2.6. Applications and Figures of Merit 3. Measurement Methods for the Pyroelectric Coefficient 3.1. General Considerations 3.1.1. Heating Concepts 3.1.2. Thermal Equilibrium 3.1.3. Electric Contact 3.1.4. Separation of Contributions 3.1.5. Thermally Stimulated Currents 3.2. Static Methods 3.2.1. Charge Compensation Method 3.2.2. Hysteresis Measurement Method 3.2.3. Direct Electrocaloric Measurement 3.2.4. Flatband Voltage Shift 3.2.5. X-ray Photoelectron Spectroscopy Method 3.2.6. X-ray Diffraction and Density Functional Theory 3.3. Dynamic Methods 3.3.1. Temperature Ramping Methods 3.3.2. Optical Methods 3.3.3. Periodic Pulse Technique 3.3.4. Laser Intensity Modulation Methods 3.3.5. Harmonic Waveform Techniques 4. Pyroelectric and Ferroelectric Characterization Setup 4.1. Pyroelectric Measurement Setup 4.1.1. Setup and Instrumentation 4.1.2. Automated Sharp-Garn Evaluation of Pyroelectric Coefficients 4.1.3. Further Examples 4.2. Hysteresis Loop Measurements 4.2.1. Instrumentation 4.2.2. Measurement and Evaluation 4.2.3. Examples 5. Investigated Material Systems 5.1. III-Nitride Bulk Semiconductors GaN and AlN 5.1.1. General Structure and Spontaneous Polarization 5.1.2. Applications 5.1.3. Crystal Growth and Doping 5.1.4. Pyroelectricity 5.2. Hafnium Oxide Thin Films 5.2.1. General Structure and Applications 5.2.2. Polar Properties in Thin Films 5.2.3. Doping Effects 5.2.4. Pyro- and Piezoelectricity 6. Results 6.1. The Pyroelectric Coefficient of Free-standing GaN and AlN 6.1.1. Sample Preparation 6.1.2. Pyroelectric Measurements 6.1.3. Lattice Influence 6.1.4. Slope Differences 6.2. Pyroelectricity of Doped Hafnium Oxide 6.2.1. Sharp-Garn Measurement on Thin Films 6.2.2. Effects of Silicon Doping 6.2.3. Dopant Comparison 7. Summary and Outlook A. Pyroelectric Current and Phase under Periodic Thermal Excitation B. Loss Current Correction for Shunt Method C. Conductivity Correction D. Comparison of Pyroelectric Figures of Merit Bibliography Publication List Acknowledgments / Die Änderung der spontanen Polarisation durch eine Änderung der Temperatur ist bekannt als der pyroelektrische Effekt, welcher auf kristalline, nicht-zentrosymmetrische und polare Materie beschränkt ist. Er findet vor allem Anwendung in Infrarot-Strahlungsdetektoren, bietet aber weitere Anwendungsfelder wie die Niedertemperatur-Abwärmenutzung oder die chemische Katalyse. Eine präzise Quantifizierung, d. h. die Messung des pyroelektrischen Koeffizienten p, ist unabdingbar, um die Leistungsfähigkeit eines Materials zu bewerten. Daher bietet diese Arbeit u.a. einen umfassenden Überblick und eine Bewertung der verfügbaren Messmethoden zur Charakterisierung von p. Weiterhin wird ein Messaufbau beschrieben, welcher die voll automatisierte Messung von p mit Hilfe der Sharp-Garn Methode und auch die Charakterisierung der ferroelektrischen Hystereseschleife ermöglicht. Aufgrund fehlerender Literaturdaten wurde dieser Aufbau anschließend genutzt, um den temperaturabhängigen pyroelektrischen Koeffizienten der dotierten III-V-Verbindungshalbleiter Gallium- und Aluminiumnitrid sowie dünner Schichten bestehend aus dotiertem Hafniumoxid zu messen und zu diskutieren. Im Vergleich zu klassichen ferroelektrischen Oxiden zeigen dabei die nitridbasierten Halbleiter einen geringen pyroelektrischen Koeffizienten und eine kleine Temperaturabhängigkeit, welche auch nur leicht durch den vorhandenen Dotanden beeinflusst werden kann. Dagegen zeigen dünne Hafniumoxidschichten einen um eine Größenordnung größeren pyroelektrischen Koeffizienten, welcher durch den anwesenden Dotanden und seine Konzentration beeinflusst wird, da dieser verantwortlich für die Ausbildung der polaren, orthorhombischen Phase gemacht wird.:1. Motivation and Introduction 2. Fundamentals 2.1. Dielectrics and their Classification 2.2. Polarization 2.3. Pyroelectricity 2.4. Ferroelectricty 2.5. Phase Transitions 2.6. Applications and Figures of Merit 3. Measurement Methods for the Pyroelectric Coefficient 3.1. General Considerations 3.1.1. Heating Concepts 3.1.2. Thermal Equilibrium 3.1.3. Electric Contact 3.1.4. Separation of Contributions 3.1.5. Thermally Stimulated Currents 3.2. Static Methods 3.2.1. Charge Compensation Method 3.2.2. Hysteresis Measurement Method 3.2.3. Direct Electrocaloric Measurement 3.2.4. Flatband Voltage Shift 3.2.5. X-ray Photoelectron Spectroscopy Method 3.2.6. X-ray Diffraction and Density Functional Theory 3.3. Dynamic Methods 3.3.1. Temperature Ramping Methods 3.3.2. Optical Methods 3.3.3. Periodic Pulse Technique 3.3.4. Laser Intensity Modulation Methods 3.3.5. Harmonic Waveform Techniques 4. Pyroelectric and Ferroelectric Characterization Setup 4.1. Pyroelectric Measurement Setup 4.1.1. Setup and Instrumentation 4.1.2. Automated Sharp-Garn Evaluation of Pyroelectric Coefficients 4.1.3. Further Examples 4.2. Hysteresis Loop Measurements 4.2.1. Instrumentation 4.2.2. Measurement and Evaluation 4.2.3. Examples 5. Investigated Material Systems 5.1. III-Nitride Bulk Semiconductors GaN and AlN 5.1.1. General Structure and Spontaneous Polarization 5.1.2. Applications 5.1.3. Crystal Growth and Doping 5.1.4. Pyroelectricity 5.2. Hafnium Oxide Thin Films 5.2.1. General Structure and Applications 5.2.2. Polar Properties in Thin Films 5.2.3. Doping Effects 5.2.4. Pyro- and Piezoelectricity 6. Results 6.1. The Pyroelectric Coefficient of Free-standing GaN and AlN 6.1.1. Sample Preparation 6.1.2. Pyroelectric Measurements 6.1.3. Lattice Influence 6.1.4. Slope Differences 6.2. Pyroelectricity of Doped Hafnium Oxide 6.2.1. Sharp-Garn Measurement on Thin Films 6.2.2. Effects of Silicon Doping 6.2.3. Dopant Comparison 7. Summary and Outlook A. Pyroelectric Current and Phase under Periodic Thermal Excitation B. Loss Current Correction for Shunt Method C. Conductivity Correction D. Comparison of Pyroelectric Figures of Merit Bibliography Publication List Acknowledgments

Ferroelectric field effect transistor (FeFET) memories based on a new type of ferroelectric material (silicon doped hafnium oxide) were studied within the scope of the present work. Utilisation of silicon doped hafnium oxide (Si:HfO2) thin films instead of conventional perovskite ferroelectrics as a functional layer in FeFETs provides compatibility to the CMOS process as well as improved device scalability. The influence of different process parameters on the properties of Si:HfO2 thin films was analysed in order to gain better insight into the occurrence of ferroelectricity in this system. A subsequent examination of the potential of this material as well as its possible limitations with the respect to the application in non-volatile memories followed. The Si:HfO2-based ferroelectric transistors that were fully integrated into the state-of-the-art high-k metal gate CMOS technology were studied in this work for the first time. The memory performance of these devices scaled down to 28 nm gate length was investigated. Special attention was paid to the charge trapping phenomenon shown to significantly affect the device behaviour.

Ferroelectric field effect transistor (FeFET) memories based on a new type of ferroelectric material (silicon doped hafnium oxide) were studied within the scope of the present work. Utilisation of silicon doped hafnium oxide (Si:HfO2) thin films instead of conventional perovskite ferroelectrics as a functional layer in FeFETs provides compatibility to the CMOS process as well as improved device scalability. The influence of different process parameters on the properties of Si:HfO2 thin films was analysed in order to gain better insight into the occurrence of ferroelectricity in this system. A subsequent examination of the potential of this material as well as its possible limitations with the respect to the application in non-volatile memories followed. The Si:HfO2-based ferroelectric transistors that were fully integrated into the state-of-the-art high-k metal gate CMOS technology were studied in this work for the first time. The memory performance of these devices scaled down to 28 nm gate length was investigated. Special attention was paid to the charge trapping phenomenon shown to significantly affect the device behaviour.:1 Introduction 2 Fundamentals 2.1 Non-volatile semiconductor memories 2.2 Emerging memory concepts 2.3 Ferroelectric memories 3 Characterisation methods 3.1 Memory characterisation tests 3.2 Ferroelectric memory specific characterisation tests 3.3 Trapping characterisation methods 3.4 Microstructural analyses 4 Sample description 4.1 Metal-insulator-metal capacitors 4.2 Ferroelectric field effect transistors 5 Stabilisation of the ferroelectric properties in Si:HfO2 thin films 5.1 Impact of the silicon doping 5.2 Impact of the post-metallisation anneal 5.3 Impact of the film thickness 5.4 Summary 6 Electrical properties of the ferroelectric Si:HfO2 thin films 6.1 Field cycling effect 6.2 Switching kinetics 6.3 Fatigue behaviour 6.4 Summary 7 Ferroelectric field effect transistors based on Si:HfO2 films 7.1 Effect of the silicon doping 7.2 Program and erase operation 7.3 Retention behaviour 7.4 Endurance properties 7.5 Impact of scaling on the device performance 7.6 Summary 8 Trapping effects in Si:HfO2-based FeFETs 8.1 Trapping kinetics of the bulk Si:HfO2 traps 8.2 Detrapping kinetics of the bulk Si:HfO2 traps 8.3 Impact of trapping on the FeFET performance 8.4 Modified approach for erase operation 8.5 Summary 9 Summary and Outlook