Atomic Layer Deposition (ALD) of Al2O3 on tall multiwalled carbon nanotube forests shows concentration variation with the depth in the form of discrete steps. While ALD is capable of extremely conformal deposition in high aspect ratio structures, decreasing penetration depth has been observed over multiple thermal ALD cycles on 1.3 mm tall multiwalled carbon nanotube forests. SEM imaging with Energy Dispersive X-ray Spectroscopy elemental analysis shows steps of decreasing intensity corresponding to decreasing concentrations of Al2O3. A study of these steps suggests that they are produced by a combination of diffusion limited delivery of precursors with increasing precursor adsorption site density as discrete nuclei grow during the ALD process. This conceptual model has been applied to modify literature models for ALD penetration on high aspect ratio structures, allowing several parameters to be extracted from the experimental data. The Knudsen diffusion constant for trimethylaluminum (TMA) in these carbon nanotube forests has been found to be 0.3 cm2s-1. From the profile of the Al2O3 concentration at the steps, the sticking coefficient of TMA on Al2O3 was found to be 0.003.

Au cours d’un traitement de surface de type dépôt assisté par plasma, les caractéristiques et propriétés de l’interface entre le dépôt et le substrat sont déterminées par la première couche atomique du dépôt, voire les premiers atomes qui commencent à recouvrir la surface du substrat. Aussi, la parfaite connaissance du comportement des particules incidentes et du réarrangement des atomes suite à l’impact d’une particule du plasma est-elle un élément essentiel à la description du comportement de la surface en cours de traitement et donc de ses propriétés ultérieures. Au cours de cette thèse, nous avons entrepris d’étudier, par une approche combinant expériences et simulation numérique par dynamique moléculaire, l’interaction d’espèces (C, Ti, W) avec une surface de silicium en fonction de paramètres tels que l’énergie, la fluence ou encore l’incidence des particules arrivant sur la surface. Une part importante de ce travail a consisté à adapter les codes de dynamique moléculaire (utilisation des champs de force réactifs) aux systèmes étudiés. La partie expérimentale a nécessité la mise en place de procédures spécifiques pour l’utilisation de l’équipement Storing Matter. Les résultats montrent que, quelles que soient l’espèce incidente, parmi celles étudiées, le coefficient de collage (SC) est dans la gamme [0.7 – 1] ; dans le cas de W, quasiment tous les atomes incidents restent sur la surface (SC~~1). Outre la détermination du coefficient de collage, pour différentes conditions initiales des espèces incidentes (énergie, incidence, fluence) les modifications apportées à la surface ont également été déterminées en termes d’implantation et de trajectoire dans le matériau des espèces incidentes, et de pulvérisation de la surface du substrat / During plasma assisted deposition, properties of the coating substrate interface depend on the first atomic layer of the deposit, or the atoms that first start to cover the surface. Therefore the good knowledge of the sticking coefficient and the reorganization of the surface following particle impact is an essential issue to achieve the description of the behavior of the processed surface and, therefore, its expected properties. Consequently, we investigated the interaction between incoming particles (C, Ti, W) and a silicon surface by using an approach combining molecular dynamic simulations and experiments. Various initial conditions were studied, energy, fluence and incidence angle of the incoming particles. An important part of this work has consisted in adapting the molecular dynamic codes (using reactive force fields) to the investigated systems. Meanwhile, experimental procedure specifically devoted to the use of the Storing Matter facility was also developed. Results show that the sticking coefficient (SC) value is in the range [0.7 – 1] irrespectively of the incoming species; in the case of W, almost all atoms stick on the surface (SC~~1). Besides the determination of sticking coefficient, the surface modification resulting from the particles impingement were determined for various initial conditions (energy, fluence, angle) in terms of implantation and displacement of the incoming species, and surface sputtering as well

This thesis supplies a contribution to the understanding of the interaction between carboncarbon surfaces, atoms with carbon-surfaces, and atoms with metallic surfaces. It is well established that the surface-surface and atom-surface interactions are interesting, important and challenging for reasons that vary from industrial interest up to the academic necessity of understanding it deeply. Currently, there are many measurements and simulations for the geometric and electronic properties of Graphite, these differ by more than 40%. This implies that our understanding of the nature of this material is quite poor. The interaction between small transition-metals clusters with a Graphene sheet is another example where our knowledge is very limited. There are many theoretical studies in the literature that describe the interaction between these clusters with a Graphene sheet, but they agree and disagree on many points, which calls for systematic study of this issue. In this thesis we will focus our efforts on studying the surface-surface and atom/clusters-surface interactions. This thesis is split into three projects. The first aims to contribute to the understanding of the interlayer interactions of the bulk Graphite. In the second, we intend to shed some light on comprehending the interaction of an adatom with a Graphene sheet. Both of these projects, are studied within DFT framework with the inclusion of the van der Waals (vdW) corrections. In the Graphite project, we found that the electronic and geometric properties depend on the vdW correction employed in the calculation. In the adatom supported on Graphene project, we combined a modified Anderson-Newns model to describe the coupling between the adatom with Graphene. In addition, we found the existence of competition between quantum and classical forces, which determine the type of site in which the adatom prefers to adsorb. The last project is a dynamical study of an atom that impinges upon a metallic surface. In this project, we focus on the calculation of the sticking coefficient, a measure of the amount of nuclear density attached to the metal surface after collision. At this time the project is not one 100% ready, but our preliminary results indicates that, a small part of the nuclear density stays stuck on the metal. / Esta tese ajuda a entender as interações entre duas superfícies de carbono. A natureza da interação de átomos ou aglomerados atômicos adsorvidos sobre uma superfície de carbono. Além disso, visa esclarecer a dinâmica de um átomo sendo adsorvido sobre uma superfície metálica. As interações superfície-superfície e átomos-superfícies são importantes por razões que variam desde o interesse industrial até a necessidade acadêmica para compreendê-la profundamente. Entendê-las ainda é um desafio. Diversos trabalhos apresentam medidas experimentais e simulações para as propriedades geométricas e eletrônicas do grafite. Tais medidas diferem em mais ide 40% umas da outra. Isso mostra que nossa compreensão sobre a natureza desse material ainda é bastante pobre. A interação entre pequenos grupos de metais de transição com uma folha de grapheno é outro exemplo em que nosso conhecimento é limitado. Existem muitos estudos teóricos na literatura que descrevem a interação desse tipo de aglomerado com uma folha de grafeno, porém há numerosas discordâncias. Tais controvérsias parecem suplicar por um estudo sistemático. Nesta tese focamos nossos estudos nas interações superfície-superfície e de átomos ou aglomerados atômicos com superfícies de carbono e de um metal. A tese foi dividida em três projetos. O primeiro visa compreender melhor a interação entre as camadas do grafite. No segundo, pretendemos lançar alguma luz no entendimento da interação de átomos e aglomerados atômicos com uma folha de grafeno. Esses dois projetos, são estudados à luz da Teoria do Funcional da Densidade com a inclusão das correções van der Waals (vdW). No Projecto sobre o grafite, mostramos que as propriedades eletrônicas e geométricas dependem do tipo de correção de vdW empregada no cálculo. No projeto sobre átomos e aglomerados atômicos adsorvidos no grapheno, combinamos um modelo modificado de Anderson-Newns para descrever o acoplamento entre um átomo adsorvido e o grafeno. Além disso, encontramos uma competição entre forças quânticas e clássicas, a qual determina o tipo de sítio no qual o átomo prefere ser adsorvido. O último projeto é um estudo dinâmico de um átomo colidindo contra uma superfície metálica. Nesse projeto o foco é posto no cálculo do coeficiente de aderência, o qual mede a taxa de densidade nuclear presa na superfície metálica após a colisão. Resultados preliminares indicam que, uma pequena parte da densidade nuclear permanece aderida ao metal depois da colisão.

Desenvolve-se um novo método de cálculo da probabilidade de adsorção química de átomos incidentes em superfícies metálicas. Introduz-se um modelo teórico de adsorção cujo Hamiltoniano descreve um átomo incidindo normalmente e interagindo com os elétrons da banda de condução de uma superfície metálica. Como interações, são levadas em consideração (1) a possibilidade de transferência de energia cinética e de carga do átomo para o metal e (2) o potencial de carga imagem do átomo ionizado. A solução do modelo consiste em se tratar a parte eletrônica e a nuclear do Hamiltoniano separadamente. A parte eletrônica é tratada com a técnica de grupo de renormalização introduzida por Wilson e a parte nuclear, através da solução numérica da equação de Schrödinger para o movimento nuclear. O acoplamento entre as duas componentes do hamiltoniano é tratado como perturbação à aproximação adiabática. A probabilidade de adsorção é calculada em função da energia cinética do átomo incidente através da regra de ouro de Fermi. Os resultados, mostrando que a probabilidade de adsorção decai rapidamente acima de uma energia cinética característica, são interpretados fisicamente. / A new procedure that calculates sticking coefficients for atomic beams incident upon metallic surfaces is discussed. A model Hamiltonian describing the normal incidence of an ad-atom and its interaction with the conduction electrons of the adsorbate is introduced. The Hamiltonian accounts for two couplings: (1) the overlap between the atomic orbital and the metallic conduction states, allowing charge transfer between incident particle and adsorbate, and (2) the image potential associated with the ionized ad-atom. The electronic and nuclear parts of the model Hamiltonian are diagonalized separately, the former by renormalization group techniques and the second by numerical integration of the Schrödinger equation for the nuclear motion. Through the perturbative treatment, the first order corrections to the adiabatic approximation are presented. The results, showing that the sticking coefficient diminishes rapidly above a characteristic kinetic energy o£ the incident atom, are interpreted.

This thesis supplies a contribution to the understanding of the interaction between carboncarbon surfaces, atoms with carbon-surfaces, and atoms with metallic surfaces. It is well established that the surface-surface and atom-surface interactions are interesting, important and challenging for reasons that vary from industrial interest up to the academic necessity of understanding it deeply. Currently, there are many measurements and simulations for the geometric and electronic properties of Graphite, these differ by more than 40%. This implies that our understanding of the nature of this material is quite poor. The interaction between small transition-metals clusters with a Graphene sheet is another example where our knowledge is very limited. There are many theoretical studies in the literature that describe the interaction between these clusters with a Graphene sheet, but they agree and disagree on many points, which calls for systematic study of this issue. In this thesis we will focus our efforts on studying the surface-surface and atom/clusters-surface interactions. This thesis is split into three projects. The first aims to contribute to the understanding of the interlayer interactions of the bulk Graphite. In the second, we intend to shed some light on comprehending the interaction of an adatom with a Graphene sheet. Both of these projects, are studied within DFT framework with the inclusion of the van der Waals (vdW) corrections. In the Graphite project, we found that the electronic and geometric properties depend on the vdW correction employed in the calculation. In the adatom supported on Graphene project, we combined a modified Anderson-Newns model to describe the coupling between the adatom with Graphene. In addition, we found the existence of competition between quantum and classical forces, which determine the type of site in which the adatom prefers to adsorb. The last project is a dynamical study of an atom that impinges upon a metallic surface. In this project, we focus on the calculation of the sticking coefficient, a measure of the amount of nuclear density attached to the metal surface after collision. At this time the project is not one 100% ready, but our preliminary results indicates that, a small part of the nuclear density stays stuck on the metal. / Esta tese ajuda a entender as interações entre duas superfícies de carbono. A natureza da interação de átomos ou aglomerados atômicos adsorvidos sobre uma superfície de carbono. Além disso, visa esclarecer a dinâmica de um átomo sendo adsorvido sobre uma superfície metálica. As interações superfície-superfície e átomos-superfícies são importantes por razões que variam desde o interesse industrial até a necessidade acadêmica para compreendê-la profundamente. Entendê-las ainda é um desafio. Diversos trabalhos apresentam medidas experimentais e simulações para as propriedades geométricas e eletrônicas do grafite. Tais medidas diferem em mais ide 40% umas da outra. Isso mostra que nossa compreensão sobre a natureza desse material ainda é bastante pobre. A interação entre pequenos grupos de metais de transição com uma folha de grapheno é outro exemplo em que nosso conhecimento é limitado. Existem muitos estudos teóricos na literatura que descrevem a interação desse tipo de aglomerado com uma folha de grafeno, porém há numerosas discordâncias. Tais controvérsias parecem suplicar por um estudo sistemático. Nesta tese focamos nossos estudos nas interações superfície-superfície e de átomos ou aglomerados atômicos com superfícies de carbono e de um metal. A tese foi dividida em três projetos. O primeiro visa compreender melhor a interação entre as camadas do grafite. No segundo, pretendemos lançar alguma luz no entendimento da interação de átomos e aglomerados atômicos com uma folha de grafeno. Esses dois projetos, são estudados à luz da Teoria do Funcional da Densidade com a inclusão das correções van der Waals (vdW). No Projecto sobre o grafite, mostramos que as propriedades eletrônicas e geométricas dependem do tipo de correção de vdW empregada no cálculo. No projeto sobre átomos e aglomerados atômicos adsorvidos no grapheno, combinamos um modelo modificado de Anderson-Newns para descrever o acoplamento entre um átomo adsorvido e o grafeno. Além disso, encontramos uma competição entre forças quânticas e clássicas, a qual determina o tipo de sítio no qual o átomo prefere ser adsorvido. O último projeto é um estudo dinâmico de um átomo colidindo contra uma superfície metálica. Nesse projeto o foco é posto no cálculo do coeficiente de aderência, o qual mede a taxa de densidade nuclear presa na superfície metálica após a colisão. Resultados preliminares indicam que, uma pequena parte da densidade nuclear permanece aderida ao metal depois da colisão.

In dieser Arbeit werden verschiedene Prozesse zur Atomlagenabscheidung (ALD) von TiO2 und HfO2 experimentell untersucht. Die Untersuchungen schließen eine experimentelle Charakterisierung des Schichtwachstums sowie eine massenspektrometrische Analyse der Reaktionsprodukte ein. Im Detail wurden der ALD-Prozess mit Cp*Ti(OMe)3 und Ozon zur Abscheidung von TiO2 sowie der ALD-Prozess mit TEMAHf und Ozon zur Abscheidung von HfO2 untersucht. Der theoretische Teil der Arbeit beginnt mit einer Methode zur Bestimmung des absoluten Haftkoeffizienten. Anschließend werden numerische Modelle entwickelt, welche die Adsorption von Präkursormolekülen durch strukturierte Substrate beschreiben. Diese Modelle enthalten die Substratstruktur und den absoluten Haftkoeffizienten. Es wird eine statistische numerische Methode entwickelt, mit der der Gastransport in dem ALD-Reaktor statistisch beschrieben wird. Die statistischen Größen, welche die Gasdynamik im Reaktor beschreiben, werden mit der Discrete Simulation Monte Carlo (DSMC) Methode bestimmt. Mit dieser Methode und den Modellen der Adsorption kann der komplette ALD-Prozess simuliert werden. Die neu entwickelte Methode wird verwendet um die Effizienz verschiedener ALD-Reaktoren in Abhängigkeit des absoluten Haftkoeffizienten, der Substratstruktur sowie der Prozessbedingungen zu untersuchen. Die Geometrie des Reaktors wird variiert und mit der Referenzgeometrie verglichen. / This dissertation is divided into an experimental part and a theoretical part. The experimental part describes the atomic layer deposition (ALD) of TiO2 and HfO2. TDMAT and Cp*Ti(OMe)3 were used as titanium precursors, while TEMAHf was used as the hafnium precursor. Ozone was used as the oxygen source. The self limiting film growth and the temperature window of these ALD processes were investigated. The reaction by-products of the Cp*Ti(OMe)3/O3 process were identified by quadrupol mass spectrometry (QMS). The QMS analysis of the TEMAHf/O3 process revealed that water is formed during the metal precursor pulse. The theoretical part of this thesis describes the development of models and numerical methods to simulate the ALD as a whole. First of all, a model for the adsorption of precursor molecules by planar substrates was developed. This model was extended to describe the adsorption of precursor molecules inside a cylindrical hole with an aspect ratio of 20, 40 and 80. The adsorption of precursor molecules is dominated by the absolute sticking coefficient (SC), i.e., the reactivity of the precursor molecules. From the numerical model the saturation profiles along the wall of a cylindrical hole can be determined. From the comparison of the simulated profile with an experimentally determined thickness profile the SC can be determined. This method was used to determine the SC of the precursors examined in the experimental part. The SC of TEMAHf increases exponentially with the substrate temperature. A discrete particle method (DSMC) was used to derive a statistical description of the gas kinetics inside an ALD reactor. Combining the statistical description of the gas transport and the numerical models of the adsorption, it is possible to simulate the ALD for any combination of reactor, substrate and SC. It is possible to distinguish the contribution of the reactor geometry, the process parameters and the process chemistry (SC) to the process efficiency. Therefore, the ALD reactor geometry can be optimized independently of the process chemistry. This method was used to study a shower head ALD reactor. The reactor geometry, the composition of the gas at the inlet and the position of the inlet nozzles was varied in order to find more efficient ALD reactors. The efficiency of the reference geometry is limited by the inlet nozzles close to the exhaust and the decrease of the pressure on the substrate near the exhaust. The efficiency of ALD processes with different SCs was simulated for planar and structured substrates with a diameter of 300 mm and 450 mm.

Desenvolve-se um novo método de cálculo da probabilidade de adsorção química de átomos incidentes em superfícies metálicas. Introduz-se um modelo teórico de adsorção cujo Hamiltoniano descreve um átomo incidindo normalmente e interagindo com os elétrons da banda de condução de uma superfície metálica. Como interações, são levadas em consideração (1) a possibilidade de transferência de energia cinética e de carga do átomo para o metal e (2) o potencial de carga imagem do átomo ionizado. A solução do modelo consiste em se tratar a parte eletrônica e a nuclear do Hamiltoniano separadamente. A parte eletrônica é tratada com a técnica de grupo de renormalização introduzida por Wilson e a parte nuclear, através da solução numérica da equação de Schrödinger para o movimento nuclear. O acoplamento entre as duas componentes do hamiltoniano é tratado como perturbação à aproximação adiabática. A probabilidade de adsorção é calculada em função da energia cinética do átomo incidente através da regra de ouro de Fermi. Os resultados, mostrando que a probabilidade de adsorção decai rapidamente acima de uma energia cinética característica, são interpretados fisicamente. / A new procedure that calculates sticking coefficients for atomic beams incident upon metallic surfaces is discussed. A model Hamiltonian describing the normal incidence of an ad-atom and its interaction with the conduction electrons of the adsorbate is introduced. The Hamiltonian accounts for two couplings: (1) the overlap between the atomic orbital and the metallic conduction states, allowing charge transfer between incident particle and adsorbate, and (2) the image potential associated with the ionized ad-atom. The electronic and nuclear parts of the model Hamiltonian are diagonalized separately, the former by renormalization group techniques and the second by numerical integration of the Schrödinger equation for the nuclear motion. Through the perturbative treatment, the first order corrections to the adiabatic approximation are presented. The results, showing that the sticking coefficient diminishes rapidly above a characteristic kinetic energy o£ the incident atom, are interpreted.

Ce travail de thèse a été réalisé, dans le cadre de la convention CIFRE 1538/2010, au sein d'adixen Vacuum Products (aVP) à Annecy (France). Il a été en partie financé par le projet S.P.A.M. (Surface Physics for Advanced Manufacturing). Il s'agit d'un projet ITN financé par le programme Pierre et Marie Curie de la Communauté Européenne rassemblant des partenaires universitaires et industriels dont aVP. L'objectif de ce programme était de contribuer à l'étude et au développement de la lithographie et en particulier la lithographie à ultraviolet extrême (EUVL). Ce travail porte sur la problématique de la contamination moléculaire dans l'industrie des semi-conducteurs ainsi que les besoins de maitrise de contamination pour la photolithographie EUVL. Pour ce faire, des modèles mathématiques de sorption ont été recherchés, testés et validés à l'aide d'une microbalance à quartz (QCM). Cette technique, possédant une très haute sensibilité (au niveau du ng), permet d'étudier les phénomènes de sorption relatifs à tout matériau déposable sur un cristal de quartz mis au contact de différents gaz dont la pression partielle est maitrisée. Par conséquent, le protocole détaillé dans cette thèse peut être utilisé pour d'autres types d'expériences dans toute discipline nécessitant une telle précision. Le déroulement de notre plan d'expérience comprend deux types de matériaux naturellement différents : un polymère (PCBA) d'une part et deux substrats métalliques (SS AISI 304 et CuC1) d'autre part pour lesquels le transfert de masse n'intervient pas de la même manière. Les gaz d'étude ont été sélectionnés pour leur intérêt dans l'industrie des semi-conducteurs (vapeur d'eau, HF). Le résultat de l'interaction des gaz d'étude avec les substrats ciblés est suivi en direct par la QCM, ce qui permet non seulement de valider et/ou améliorer les modèles mathématiques déjà disponibles dans la bibliographie mais aussi de les ajuster aux données obtenues expérimentalement. Nous pouvons ainsi non seulement prévoir le comportement des contaminants à l'équilibre (isothermes) et à l'état transitoire mais aussi réaliser des estimations de sorption à des températures autres que celles retenues pour notre plan d'expérience / This thesis was carried out within the framework of the CIFRE 1538/2010 convention at adixen Vacuum Products (aVP) in Annecy (France). It is has been partly funded by the ITN project SPAM (Surface Physics for Advanced Manufacturing). SPAM is an ITN project funded by the Pierre and Marie Curie program of the European Community bringing together academic institutions and industrial partners including aVP. The objective of this program was to contribute to the study and development of lithography and extreme ultraviolet lithography (EUVL). This work deals with the issues caused by the airborne molecular contamination (AMC) in the semiconductor industry and their control needs in EUVL and the current photolithography. In order to tackle the problem, sorption mathematical models have been investigated and validated using a quartz crystal microbalance (QCM). This technique, which confers a high sensitivity (ng level), allows the study of the sorption phenomena related to any deposable material onto a quartz crystal in contact with different gases whose concentrations are accurately controlled. Consequently, the protocol detailed in this thesis may be used for other types of experiments in any discipline requiring such precision. The conduct of our experimental plan includes two types of naturally different materials: a polymer (PCBA) on the one hand and two metallic substrates (stainless steel AISI 304 and CuC1) on the other hand, for which the matter transfer does not occur in the same manner. Studied gases were selected for their interest in the semiconductor industry (water vapor, HF). The resulting interaction between the studied gases and the targeted substrates is continuously followed by the QCM, which allows not only to validate the mathematical models already proposed by the literature but also to fit the experimentally obtained data. This enables us not only to predict the behavior of the AMC at equilibrium (isotherms) and the transient state but also to provide sorption estimations at temperatures other than those specified in our experimental plan

In dieser Arbeit werden verschiedene Prozesse zur Atomlagenabscheidung (ALD) von TiO2 und HfO2 experimentell untersucht. Die Untersuchungen schließen eine experimentelle Charakterisierung des Schichtwachstums sowie eine massenspektrometrische Analyse der Reaktionsprodukte ein. Im Detail wurden der ALD-Prozess mit Cp*Ti(OMe)3 und Ozon zur Abscheidung von TiO2 sowie der ALD-Prozess mit TEMAHf und Ozon zur Abscheidung von HfO2 untersucht. Der theoretische Teil der Arbeit beginnt mit einer Methode zur Bestimmung des absoluten Haftkoeffizienten. Anschließend werden numerische Modelle entwickelt, welche die Adsorption von Präkursormolekülen durch strukturierte Substrate beschreiben. Diese Modelle enthalten die Substratstruktur und den absoluten Haftkoeffizienten. Es wird eine statistische numerische Methode entwickelt, mit der der Gastransport in dem ALD-Reaktor statistisch beschrieben wird. Die statistischen Größen, welche die Gasdynamik im Reaktor beschreiben, werden mit der Discrete Simulation Monte Carlo (DSMC) Methode bestimmt. Mit dieser Methode und den Modellen der Adsorption kann der komplette ALD-Prozess simuliert werden. Die neu entwickelte Methode wird verwendet um die Effizienz verschiedener ALD-Reaktoren in Abhängigkeit des absoluten Haftkoeffizienten, der Substratstruktur sowie der Prozessbedingungen zu untersuchen. Die Geometrie des Reaktors wird variiert und mit der Referenzgeometrie verglichen.:Inhaltsverzeichnis................................................................................ i Tabellenverzeichnis.............................................................................. iii Abbildungsverzeichnis ......................................................................... v Abkürzungsverzeichnis ........................................................................ ix Formelverzeichnis ................................................................................ xi 1. Einführung ....................................................................................... 1 1.1. Motivation und Zielstellung ........................................................... 1 1.2. Grundlagen der Atomlagenabscheidung ....................................... 3 1.3. Materialien und Anwendungen ..................................................... 6 2. Experimentelle Grundlagen .............................................................. 9 2.1. ALD-Anlage ................................................................................... 9 2.2. Physikalische Probencharakterisierung ........................................ 11 2.2.1. Röntgenmethoden ..................................................................... 11 2.2.2. Elektronenstrahl-Methoden ....................................................... 12 2.2.3. Spektrometrische Methoden ...................................................... 13 2.3. Experimentelle in-situ Prozesscharakterisierung .......................... 14 3. Atomlagenabscheidung von TiO2 und HfO2 ..................................... 21 3.1. Abscheidung von Titandioxid ........................................................ 21 3.1.1. TDMAT als Titanpräkursor .......................................................... 21 3.1.2. Cp*Ti(OMe)3 als Titanpräkursor ................................................ 25 3.2. Abscheidung von Hafniumdioxid mit TEMAHf und Ozon ................. 30 3.3. Massenspektrometrie an ALD-Prozessen mit Ozon ...................... 32 3.3.1. Cp*Ti(OMe)3 mit Ozon .............................................................. 32 3.3.2. TMA mit Ozon ............................................................................ 36 3.3.3. TEMAHf mit Ozon ....................................................................... 37 3.3.4. Prozessüberwachung mit Massenspektrometrie ....................... 39 3.4. Zusammenfassung zur ALD von TiO2 und HfO2 ........................... 41 4. Modellierung der Adsorption ........................................................... 43 4.1. Adsorptionsverhalten planarer Substrate .................................... 43 4.2. Adsorptionsverhalten strukturierter Substrate ............................ 49 4.2.1. Numerische Simulationsmethode .............................................. 52 4.2.2. Gaskinetik in einem zylindrischen Graben ................................. 54 4.2.3. Effektive Haftkoeffizienten und Sättigungsdosen ..................... 55 4.2.4. Sättigungsprofile entlang der Grabenwand .............................. 59 4.3. Methode zur Bestimmung des absoluten Haftkoeffizienten von ALD-Präkursoren ........................................................................................ 61 4.3.1. Methode am Beispiel von TDMAT mit Ozon ................................ 66 4.3.2. Absoluter Haftkoeffizient von TEMAHf mit Ozon ......................... 74 4.3.3. Absoluter Haftkoeffizient von Cp*Ti(OMe)3 mit Ozon ................ 78 4.3.4. Temperaturabhängigkeit absoluter Haftkoeffizienten ............... 79 4.4. Zusammenfassung zur Modellierung der Adsorption .................... 81 5. Gekoppelte Prozesssimulation ........................................................ 83 5.1. Statistische Methode zur Simulation der ALD ............................... 83 5.1.1. Statistische Größen der Gasdynamik ......................................... 85 5.1.2. Algorithmus der gekoppelten ALD-Simulation ............................ 90 5.2. Anwendung der Methode zur Optimierung einer Gasdusche ........ 93 5.2.1. Geometrie und Randbedingungen ............................................. 93 5.2.2. Ergebnis der Reaktorsimulation ................................................. 96 5.2.3. Gekoppelte ALD-Simulation für planare Substrate ................... 102 5.2.4. Gekoppelte ALD-Simulation für strukturierte Substrate ........... 110 5.3. Einfluss der Randbedingungen auf die geometrische Effizienz ... 113 5.4. Vergleich zwischen Simulation und Experiment .......................... 114 6. Zusammenfassung und Ausblick .................................................... 117 Literaturverzeichnis ........................................................................... 121 Anhang .............................................................................................. 129 Parameter der modellierten effektiven Haftkoeffizienten ................... 129 Hafnium-Dotierung von Titandioxidschichten ..................................... 131 Eigene Veröffentlichungen ................................................................. 133 Lebenslauf ......................................................................................... 135 / This dissertation is divided into an experimental part and a theoretical part. The experimental part describes the atomic layer deposition (ALD) of TiO2 and HfO2. TDMAT and Cp*Ti(OMe)3 were used as titanium precursors, while TEMAHf was used as the hafnium precursor. Ozone was used as the oxygen source. The self limiting film growth and the temperature window of these ALD processes were investigated. The reaction by-products of the Cp*Ti(OMe)3/O3 process were identified by quadrupol mass spectrometry (QMS). The QMS analysis of the TEMAHf/O3 process revealed that water is formed during the metal precursor pulse. The theoretical part of this thesis describes the development of models and numerical methods to simulate the ALD as a whole. First of all, a model for the adsorption of precursor molecules by planar substrates was developed. This model was extended to describe the adsorption of precursor molecules inside a cylindrical hole with an aspect ratio of 20, 40 and 80. The adsorption of precursor molecules is dominated by the absolute sticking coefficient (SC), i.e., the reactivity of the precursor molecules. From the numerical model the saturation profiles along the wall of a cylindrical hole can be determined. From the comparison of the simulated profile with an experimentally determined thickness profile the SC can be determined. This method was used to determine the SC of the precursors examined in the experimental part. The SC of TEMAHf increases exponentially with the substrate temperature. A discrete particle method (DSMC) was used to derive a statistical description of the gas kinetics inside an ALD reactor. Combining the statistical description of the gas transport and the numerical models of the adsorption, it is possible to simulate the ALD for any combination of reactor, substrate and SC. It is possible to distinguish the contribution of the reactor geometry, the process parameters and the process chemistry (SC) to the process efficiency. Therefore, the ALD reactor geometry can be optimized independently of the process chemistry. This method was used to study a shower head ALD reactor. The reactor geometry, the composition of the gas at the inlet and the position of the inlet nozzles was varied in order to find more efficient ALD reactors. The efficiency of the reference geometry is limited by the inlet nozzles close to the exhaust and the decrease of the pressure on the substrate near the exhaust. The efficiency of ALD processes with different SCs was simulated for planar and structured substrates with a diameter of 300 mm and 450 mm.:Inhaltsverzeichnis................................................................................ i Tabellenverzeichnis.............................................................................. iii Abbildungsverzeichnis ......................................................................... v Abkürzungsverzeichnis ........................................................................ ix Formelverzeichnis ................................................................................ xi 1. Einführung ....................................................................................... 1 1.1. Motivation und Zielstellung ........................................................... 1 1.2. Grundlagen der Atomlagenabscheidung ....................................... 3 1.3. Materialien und Anwendungen ..................................................... 6 2. Experimentelle Grundlagen .............................................................. 9 2.1. ALD-Anlage ................................................................................... 9 2.2. Physikalische Probencharakterisierung ........................................ 11 2.2.1. Röntgenmethoden ..................................................................... 11 2.2.2. Elektronenstrahl-Methoden ....................................................... 12 2.2.3. Spektrometrische Methoden ...................................................... 13 2.3. Experimentelle in-situ Prozesscharakterisierung .......................... 14 3. Atomlagenabscheidung von TiO2 und HfO2 ..................................... 21 3.1. Abscheidung von Titandioxid ........................................................ 21 3.1.1. TDMAT als Titanpräkursor .......................................................... 21 3.1.2. Cp*Ti(OMe)3 als Titanpräkursor ................................................ 25 3.2. Abscheidung von Hafniumdioxid mit TEMAHf und Ozon ................. 30 3.3. Massenspektrometrie an ALD-Prozessen mit Ozon ...................... 32 3.3.1. Cp*Ti(OMe)3 mit Ozon .............................................................. 32 3.3.2. TMA mit Ozon ............................................................................ 36 3.3.3. TEMAHf mit Ozon ....................................................................... 37 3.3.4. Prozessüberwachung mit Massenspektrometrie ....................... 39 3.4. Zusammenfassung zur ALD von TiO2 und HfO2 ........................... 41 4. Modellierung der Adsorption ........................................................... 43 4.1. Adsorptionsverhalten planarer Substrate .................................... 43 4.2. Adsorptionsverhalten strukturierter Substrate ............................ 49 4.2.1. Numerische Simulationsmethode .............................................. 52 4.2.2. Gaskinetik in einem zylindrischen Graben ................................. 54 4.2.3. Effektive Haftkoeffizienten und Sättigungsdosen ..................... 55 4.2.4. Sättigungsprofile entlang der Grabenwand .............................. 59 4.3. Methode zur Bestimmung des absoluten Haftkoeffizienten von ALD-Präkursoren ........................................................................................ 61 4.3.1. Methode am Beispiel von TDMAT mit Ozon ................................ 66 4.3.2. Absoluter Haftkoeffizient von TEMAHf mit Ozon ......................... 74 4.3.3. Absoluter Haftkoeffizient von Cp*Ti(OMe)3 mit Ozon ................ 78 4.3.4. Temperaturabhängigkeit absoluter Haftkoeffizienten ............... 79 4.4. Zusammenfassung zur Modellierung der Adsorption .................... 81 5. Gekoppelte Prozesssimulation ........................................................ 83 5.1. Statistische Methode zur Simulation der ALD ............................... 83 5.1.1. Statistische Größen der Gasdynamik ......................................... 85 5.1.2. Algorithmus der gekoppelten ALD-Simulation ............................ 90 5.2. Anwendung der Methode zur Optimierung einer Gasdusche ........ 93 5.2.1. Geometrie und Randbedingungen ............................................. 93 5.2.2. Ergebnis der Reaktorsimulation ................................................. 96 5.2.3. Gekoppelte ALD-Simulation für planare Substrate ................... 102 5.2.4. Gekoppelte ALD-Simulation für strukturierte Substrate ........... 110 5.3. Einfluss der Randbedingungen auf die geometrische Effizienz ... 113 5.4. Vergleich zwischen Simulation und Experiment .......................... 114 6. Zusammenfassung und Ausblick .................................................... 117 Literaturverzeichnis ........................................................................... 121 Anhang .............................................................................................. 129 Parameter der modellierten effektiven Haftkoeffizienten ................... 129 Hafnium-Dotierung von Titandioxidschichten ..................................... 131 Eigene Veröffentlichungen ................................................................. 133 Lebenslauf ......................................................................................... 135