In this work the influence of solid electrolyte additives on the gas sensing properties of tin oxide layers was investigated systematically for the first time. NASICON (NAtrium, Super Ionic CONductor, Na(1+x)Zr2SixP(3-x)O12; 0 <= x <= 3) was used as a model for solid electrolyte additives. The structure of that material is ideally suitable for studies of the correlation between material parameters and the gas sensitivity of the layers. In the NASICON structure the content of mobile Na+-ions can be varied by a factor of four resulting in a simultaneous change of the ionic conductivity sigma(Na+) by approximately three orders of magnitude without considerable structural alterations. Powders of SnO2 and NASICON (x = 0; 2.2; 3) were prepared separately by means of sol-gel routes and mixed in a volume ratio of 80/20. Pastes were prepared from these powders with different compositions and screen printed on alumina substrates with a fourfold structure of thin film gold electrode combs. Four different compositions were characterised simultaneously at elevated temperatures in various gas atmospheres. The conductivity of the layers, when measured in air, decreases considerably with increasing Na+-content in the NASICON additive. This is correlated with enhanced activation energy of the electronic conductivity. The sensitivity of the layers to polar organic molecules like R-OH (alcohols), R-HO (aldehydes) and ROOH (carboxylic acids) is highly enhanced by the NASICON additive. This is observed especially on the admixtures with NASICON of high Na+-content (x = 2.2 and x = 3). On the other hand, the sensitivity to substances with mid-standing functional groups like 2-propanol or propanone can not be enhanced by NASICON additives. Furthermore the sensitivity of these composite layers to CO, H2, NH3, methane, propane, propene and toluene (all exposed as admixtures with air) is lower than the sensitivity of pure SnO2-layers. These observations are well correlated with the results of gas consumption measurements on SnO2/NASICON powders by means of FTIR spectroscopy. In spite of the lack of surface analytical data, a model of surface chemical gas reactions based on a triple phase boundary (SnO2/NASICON/gas atmosphere) was developed, which explains the experimental observations qualitatively. It is assumed that the decrease of the electronic conductivity as observed in the presence of NASICON additives with increasing Na+-content is due to an enhanced electron depletion layer. This is formed in the SnO2 grains by Na+/e- interactions across the SnO2/NASICON-interface. The enormous enhancement of the sensitivity to polar organic molecules may be due to specific nucleophilic interactions with the Na+-ions and coupled Na+/e--interactions at the triple phase reaction sites.

In this work the influence of solid electrolyte additives on the gas sensing properties of tin oxide layers was investigated systematically for the first time. NASICON (NAtrium, Super Ionic CONductor, Na(1+x)Zr2SixP(3-x)O12; 0 <= x <= 3) was used as a model for solid electrolyte additives. The structure of that material is ideally suitable for studies of the correlation between material parameters and the gas sensitivity of the layers. In the NASICON structure the content of mobile Na+-ions can be varied by a factor of four resulting in a simultaneous change of the ionic conductivity sigma(Na+) by approximately three orders of magnitude without considerable structural alterations. Powders of SnO2 and NASICON (x = 0; 2.2; 3) were prepared separately by means of sol-gel routes and mixed in a volume ratio of 80/20. Pastes were prepared from these powders with different compositions and screen printed on alumina substrates with a fourfold structure of thin film gold electrode combs. Four different compositions were characterised simultaneously at elevated temperatures in various gas atmospheres. The conductivity of the layers, when measured in air, decreases considerably with increasing Na+-content in the NASICON additive. This is correlated with enhanced activation energy of the electronic conductivity. The sensitivity of the layers to polar organic molecules like R-OH (alcohols), R-HO (aldehydes) and ROOH (carboxylic acids) is highly enhanced by the NASICON additive. This is observed especially on the admixtures with NASICON of high Na+-content (x = 2.2 and x = 3). On the other hand, the sensitivity to substances with mid-standing functional groups like 2-propanol or propanone can not be enhanced by NASICON additives. Furthermore the sensitivity of these composite layers to CO, H2, NH3, methane, propane, propene and toluene (all exposed as admixtures with air) is lower than the sensitivity of pure SnO2-layers. These observations are well correlated with the results of gas consumption measurements on SnO2/NASICON powders by means of FTIR spectroscopy. In spite of the lack of surface analytical data, a model of surface chemical gas reactions based on a triple phase boundary (SnO2/NASICON/gas atmosphere) was developed, which explains the experimental observations qualitatively. It is assumed that the decrease of the electronic conductivity as observed in the presence of NASICON additives with increasing Na+-content is due to an enhanced electron depletion layer. This is formed in the SnO2 grains by Na+/e- interactions across the SnO2/NASICON-interface. The enormous enhancement of the sensitivity to polar organic molecules may be due to specific nucleophilic interactions with the Na+-ions and coupled Na+/e--interactions at the triple phase reaction sites.

In the latter half of the 20th century the first active environmentalist movements such as Greenpeace and the International Energy Agency were born and initiated a gradual rethinking of environmental awareness. Against all expectations the sole agency under international law for climate protection policy, called the United Nations Framework Convention on Climate Change, was formed 20 years later. Today the awareness of sustained, regenerative and environmental policies permeates throughout all areas of life, science and industry. But energy provision is the most decisive topic, especially since the discussions concerning the phase out of nuclear power where the voices calling for alternative energy sources have become much more vociferous. In addition the depletion of fossil fuels is expected to occur in the not too distant future. All new energy generation methods are required to meet the present and future energy demands, need to be ecological and need to exhibit the same or significantly lower cost expenditure than current energy sources. Unfortunately mankind is confronted with the problem that current commercial alternative energies are more expensive and not yet remotely as efficient as the present energy sources. Although energy provision based on water, wind, sun and geothermal sources have a huge potential because of their continuous presence, unfortunately, they are plagued by inefficient energy conversion caused by the state of technology i.e. the conversion of sun light into electricity loses energy through heat emission, reflection of the sun light, the inability of the material to absorb the entire sun spectrum and the ohmic losses in the transmission of electric current. The sun power is the most exhaustless resource and moreover through photovoltaic action, one of the most direct and cleanest source for use in energy conversion. Presently incoming sun light is not transformed in its entirely, as much degradation occurs during photon absorption and electron transfer processes. A number of other innovative possibilities have also been researched. With respect to cost and efficiency one of the most promising devices is injection solar cells (ISC). By dint of the dye sensitised solar cell (DSSC) Grätzels findings provided the foundations for much research into this type of solar cell where the light absorbing molecule employed in is a dye.[1] The current is obtained through charge separation in the dye, which is initiated through the connection between the dye and a metal oxide on the one hand and a matched redox couple on the other. In a variant of the DSSC the charge separation processes can also occur between a nanoporous metal oxide and nanoparticles giving rise to a quantum dot sensitised solar cell (QDSSC).[2] The use of nanoparticle (NP) properties can be utilized for the harvesting of solar energy, as demonstrated by Kamat and coworkers[3] who were able to exploit these findings subsequently and prepare a number of nanoparticle based solar cells. Nanoparticle research has comprised a wide field of science and nanotechnology for a number of years. As the size of a material approaches dimensions on the nm scale the surface properties contribute proportionally more to the sum of the properties than the volume due to the increase in the surface to volume ratio. These dimensions also constitute a threshold in which quantum physical effects need to be taken into account. Hence the properties of devices or materials in this size regime are inevitably size dependent. The basic principles can be described by two different theories, one of which is based on molecular orbital theory in which the particle is treated as a molecule. For this reason n atomic orbitals with the same symmetry and energy can build up n molecular orbitals through their linear combination based on the LCAO method (Linear Combination of Atomic Orbitals).[4] In the case of solids the orbitals build up energy bands, where the unoccupied states form the quasi continous conduction band (CB) and the occuppied states form the quasi continous valence band (VB). The energy \"forbidden\" area in between these two bands is called the band gap. The band gap is a fixed material property for bulk solids but depends on size in the case of the nanoparticles. In contrast to the LCAO method, simplified solid state theory will be used throughout the present work, the theoretical background of which is provided by the effective mass approximation.[5] When an absorption of a photon occurs, an exciton (electron-hole pair) can be generated. By promoting an electron (e-) from the valence band into the conduction band a hole (h+) may be said to remain in the valence band. By comparison to bulk solids, in a small particle the free charges can sense the potential barrier i.e. the edges of the nanoparticle. Analogous to the particle in a box model this potential barrier interaction results in an increase in the band gap as the particle size decreases. In a solar cell NPs with a particle size which possess a band gap energy in the near infrared (NIR) may be utilised and therefore the NPs will be able to absorb in this spectral region. However NPs also have the ability to absorb higher energy photons due to the continuum present in their band structure, so that almost the entire sun spectral range from the NIR up to UV wavelengths may be absorbed just by using the appropriate NP material and size. Suitable NPs are metal chalcogenides e.g. MX (where M = cadmium, zinc or lead and X = sulfur, selenium or tellurium) because of their bandgap size[6–10] and their relative band positions compared to those of the semiconductor oxide states. Both the TiO2/CdSe[11–14] and TiO2/CdTe[15–18] systems have already been successfully fabricated and many of the anomalies reported.[3] Much interest in the lead chalcogenides has been generated by reports that they may feature the possibility to exhibit multiple exciton generation (MEG) where the absorption of one high energy photon can result in more than one electron-hole pairs.[19–25] Currently electrochemical impedance spectroscopy (EIS) is being used more and more to clarify processes at polarisable surfaces and materials such as nanoparticles. Likewise this method has been rediscovered in photovoltaic research and its use in the characterisation of DSSCs has been discussed in the literature.[26–31] In a number of publications the evaluation of nanoporous and porous structures has been quite extensively explored.[28,29,32–34] Since the mid-20th century Jaffé’s[35] theoretical work concerning the steady- state ac response of solid and liquid systems lead to the formation of the basics of EIS. Further developments in the measurement technology have lead to a broader range of analysis becoming possible. Nevertheless the most challenging part still remains the interpretation of the results and especially to merge the measured data with the theoretical model. EIS quantifies the changes in a small ac current response at electrode electrolyte interfaces i.e. the rate at which the polarized domain will respond, when an ac potential is applied. In this way dielectric properties of materials or composites, such as charge transfers, polarization effects, charge recombination and limitations can be measured as a function of frequency and mechanistic information may be unveiled. Hence EIS allows one to draw a conclusion concerning chemical reactions, surface properties as well as interactions between the electrodes and the electrolyte. Other very useful tools that may be employed for quantifying electron transfer processes and their time domains are intensity modulated photocurrent spectroscopy (IMPS) and intensity modulated photovoltage spectroscopy (IMVS). IMPS permits the generation of time-resolved plots of particular photo-processes in the system, each of which may be specifically addressed through varying the excitation wavelength. For the IMPS technique a sinusoidal wave with a small amplitude is applied, analogous to that of electrochemical impedance spectroscopy, but in this case the modulation is applied to a light source and not to the electrochemical cell as in EIS.[35] The current response is associated with the photogenerated charge carriers which flow through the system and finally discharge into the circuit. The amount of generated and discharged charge carriers is often different due to the presence of recombination and capture processes in surface or trap states. Ultimately the phase shift and magnitude of these currents reveal the kinetics of such processes. The only processes that will be addressed will be those that occur in the same frequency domain or on the same time scale as that of the modulated frequency of the illuminated light. In the literature some explanation of the kinetics of simple systems can be found and basic theories and introductive disquisitions may be found elsewhere.[36–38] Furthermore in solar cell research a multiplicity of studies are available which give an account of IMPS measurements on TiO2 nanoporous structures. Such studies permitted proof for the electron trapping and detrapping mechanism in TiO2 surface states.[39,40] An analysis of TiO2 electrodes combined with a dye sensitization step was established in the work of Peter and Ponomarev.[41–43] Hickey et.al.[44,45] have previously published kinetic studies on CdS nanoparticle (NP) modified electrodes. A theory was presented which allows for the IMPS data to be the interpreted in the case of CdS NP based electrodes. The back transfer, recombination and surface states have been demonstrated to be important as was determined from their inclusion in the theory. Similar attempts to explain the kinetics of CdS quantum dots are described by Bakkers et.al.[46]. In the present work the most important questions concern the behaviour of the photovoltaic assembly. Such assemblies can be equated with an electrode in contact with an electrolyte. Preliminary remarks about such electrodes as components of an electrochemical cell will be introduced in the first part of chapter 2. Thereafter the properties of electrodes in contact with the electrolyte and under illuminated conditions are illustrated. This is followed by a description of the important electrochemical and opto-electrochemical methods which have been employed in these studies. In particular, two separate subsections are dedicated to the methods of EIS and IMPS and the experimental section which are then linked to the theoretical section. The synthesis of all substances used and the preparation of the solar cell substrates are also dealt with in this section as will the equipment used and the instrument settings employed. The optical response of the working photoactive electrode is not only dependent on the substances used but also on their arrangement and linkage. The substrate which was employed in chapter 3 consists of a nanoporous ZnO gel layer upon which an organic linker has been placed in order to connect the oxide layer with the light absorbing component, the PbS NPs. Chapter 3 deals with the linker dependence on the ZnO layer and reports the typical optical characteristics and assembly arrangements of six different linkers on the ZnO layer which is an important intermediate stage in the fabrication of an ISC. The questions concerning how the type of linking affects the photo response and other electrochemical interactions of the complete solar cell substrate will be outlined in chapter 4. Further an examination of the electrochemical and opto-electrochemical behaviours of the samples will be presented similar to that presented in chapter 3. The most interesting substrate resulting from the investigations as described in chapter 3 and 4 will be used for a more in-depth characterisation by EIS in chapter 5. A suitable model and the results of the calculation of the ISC and the intermediate stages will be presented. The potential dependence, the dependence on the illuminated wavelength and also the size dependence of the PbS nanoparticles will be discussed. It will be revealed that ZnO is chemically unstable in contact with some of the linkers. For that reason the same linker study has been repeated with the more stable TiO2 employed as the wide band metal oxide. Comparisons between the different semiconductor metal oxides are made in chapter 6. In addition a number of open questions which previously had remained unanswered due to the instability of the ZnO can now be answered. In chapter 7 another highly porous structure different from that of the ZnO gel structure has been studied to determine its suitability as an ISC substrate. The structure arises from the electrodeposition of a ZnO reactant in the presence of eosin Y dye molecules. In the end the desorption of the dye provides a substrate with a high degree of porosity. Compared to the ZnO gel which was prepared and used for measurements in chapter 3 and 4, the electrodeposited ZnO is of a higher crystallinity and possesses a more preferential orientation. This results in a lower amount of grain boundaries which in turn results in fewer trap processes and subsequently yields a higher effective diffusion of the electron through the layer.[47,48] Optical and (opto-)electrochemical methods have been used for the basic characterisation of the untreated ZnO/Eosin Y and all other materials used in the fabrication of the ISC and a comparison with the ZnO gel used in chapter 3 and 4 will be made. Finally in chapter 8 an alternative metal oxide structure will be discussed. The background to this last chapter is to examine the influence of the ISC where the oxidic layer is present as a highly periodic arrangement, known as a photonic crystal. The TiO2 metal oxide which was also used in chapter 6 has been structured to form an inverse opal. First preparative findings and the first illustration of the (opto-)electrochemical results are presented. Consequently suggestions for improvements will be made. It is envisaged that the information gathered and presented here will help to achieve a deeper understanding of solar cells and help to improve the device efficiency and the interplay of the materials. Elementary understanding paves the way for further developments which can also contribute to providing devices for more efficient energy conversion.

In elektrochemischen Zellen vollziehen sich die Haupteaktionen in der Regel an Oberflächen von Elektroden und Katalysatoren und deren Elektrolytgrenzflächen, wodurch Änderungen dort die Effizienz der Zelle stark beeinflussen können. Diese Arbeit behandelt elementare Prozesse an solchen Ober- und Grenzflächen, die die Bandstruktur und damit möglicherweise auch die Reaktivität des Systems verändern. Mit Zwei-Photonen-Photoelektronenspektroskopie (2PPE) untersuche ich solche Prozesse in drei Modellsystemen für Elektrodenoberflächen beziehungsweise Elektrolyt/Elektroden-Grenzflächen: ZnO wird als Material für die photoelektrochemische Wasserspaltung diskutiert. In zeitaufgelösten 2PPE-Spektren beobachte ich Oszillationen des Dipols der (10-10)-Oberfläche, die bislang unbekannten kohärenten Oberflächenphononen zuzuordnen sind. Ich diskutiere ihre Erzeugung und entwickle eine Methode, um ultraschnelle Änderungen des Oberflächendipols anhand der Intensität des Sekundärelektronenschwanzes eines 2PPE Spektrums zu quantifizieren. An der D2O/ZnO(10-10)-Grenzfläche untersuche ich mehrere Effekte der Wasseradsorption, wie Veränderungen der Austrittsarbeit und der kohärenten Oberflächenphononen. Anders als in früheren Studien stelle ich keine Oberflächenmetallisierung durch Wasseradsorption fest. Auch gibt es keinen klaren Hinweis auf Elektronensolvatisierung, wie sie an Wasser/Metall-Grenzflächen zu beobachten ist. An der DMSO/Cu(111)-Grenzfläche, einem Modellsystem der Elektrolyt/Kathoden-Grenzfläche in Metall-Luft-Batterien, bestimme ich die elementaren Schritte der Sauerstoffreduktion. Im DMSO werden kleine Polaronen ultraschnell gebildet und zum Teil in Oberflächendefekten eingefangen. Die Lebensdauer dieser gefangenen Elektronen kann mehrere Sekunden betragen. Sie reagieren mit co-adsorbiertem O2, nachdem es in das DMSO diffundiert ist, zu O2-. Die Modellierung der Diffusion liefert eine Abschätzung des Elektroden-Reaktanten-Abstandes für Elektronentransfer in DMSO. / In electrochemical cells, the main reactions usually proceed at the surfaces of electrodes and catalysts and their interfaces with the electrolyte. Hence, changes there can have a huge impact on the efficiency of the cell. This thesis concerns elementary processes at such surfaces and interfaces, which affect the electronic band structure and, thus, potentially the reactivity of the surface. Using two-photon photoelectron spectroscopy (2PPE), I investigate such processes in three model systems for electrode surfaces and electrolyte/electrode interfaces: ZnO is discussed as material for photoelectrochemical water splitting. In time-resolved 2PPE spectra, I observe oscillations of the (10-10) surface dipole, which are assigned to previously unknown coherent surface phonons. I discuss their generation and develop a method to quantify ultrafast surface dipole changes from the intensity of the secondary electron tail of a 2PPE spectrum. At the D2O/ZnO(10-10) interface, I examine several effects of water adsorption, such as changes of the work function and the coherent surface phonons. Unlike in a previous study, I do not observe surface metallization upon water adsorption. Moreover, there is no clear indication of electron solvation as found at water/metal interfaces. At the DMSO/Cu(111) interface, a model system for the electrolyte/cathode interface in metal-air batteries, I determine the elementary steps of superoxide formation. In the DMSO, small polarons are formed and partly trapped in surface defects on an ultrafast time scale. These trapped electrons can persist for several seconds and react with co-adsorbed O2 to from O2-. Modelling the diffusion yields estimates for the electrode-reactant distance for electron transfer in DMSO.

Aussagen zur Reaktivität von Metalloxidclustern in der Gasphase wurden aus Messungen von totalen integralen Wechselwirkungsquerschnitten an Gastargets abgeleitet und diese mit Untersuchungen des Reaktionsverlaufes unter thermalisierten Bedingungen in einem Fließreaktor korreliert. Diese Untersuchungen konnten mit einer speziell entwickelten Molekularstrahlapparatur ausgeführt werden, die detailliert beschrieben wird. Mittels dieser beider Methoden wurden die Reaktionen von Bismutoxid-Clusterkationen mit Alkenen untersucht. Als Hauptreaktionskanal wird hier eine Assoziation des Alkens festgestellt. Einzelne Clusterkationen lagern zusätzlich zum Alken molekularen Sauerstoff an. Aus der massenspektrometrisch beobachteten Reaktionsfolge wird ein Mechanismus für die Aktivierung des molekularen Sauerstoffs und die Übertragung auf den Kohlenwasserstoff abgeleitet. Diese Resultate werden durch jüngste quantenchemische Rechnungen bestätigt. (M. Bienati, Dissertation, Humboldt-Universität zu Berlin, 2001) / Insights into the reactivity of metal oxide clusters in the gas phase have been gained from total integral cross sections measured with gas targets and investigations of the reaction sequences at thermalized conditions in a fast flow reactor. These experiments have been carried out with a newly designed molecular beam apparatus, which is described in detail. Applying these two techniques, the reactions of bismuth oxide cluster cations with alkenes have been probed. The main reaction channel is the association of alkenes, but particular clusters bind molecular oxygen additionally. A mechanism for this activation of molecular oxygen and its transfer towards the hydrocarbon has been derived from the mass spectrometrically measured reaction sequence, which is supported by recent theoretical calculations (M. Bienati, Doctoral Thesis, Humboldt-Universität zu Berlin, 2001)

In the latter half of the 20th century the first active environmentalist movements such as Greenpeace and the International Energy Agency were born and initiated a gradual rethinking of environmental awareness. Against all expectations the sole agency under international law for climate protection policy, called the United Nations Framework Convention on Climate Change, was formed 20 years later. Today the awareness of sustained, regenerative and environmental policies permeates throughout all areas of life, science and industry. But energy provision is the most decisive topic, especially since the discussions concerning the phase out of nuclear power where the voices calling for alternative energy sources have become much more vociferous. In addition the depletion of fossil fuels is expected to occur in the not too distant future. All new energy generation methods are required to meet the present and future energy demands, need to be ecological and need to exhibit the same or significantly lower cost expenditure than current energy sources. Unfortunately mankind is confronted with the problem that current commercial alternative energies are more expensive and not yet remotely as efficient as the present energy sources. Although energy provision based on water, wind, sun and geothermal sources have a huge potential because of their continuous presence, unfortunately, they are plagued by inefficient energy conversion caused by the state of technology i.e. the conversion of sun light into electricity loses energy through heat emission, reflection of the sun light, the inability of the material to absorb the entire sun spectrum and the ohmic losses in the transmission of electric current. The sun power is the most exhaustless resource and moreover through photovoltaic action, one of the most direct and cleanest source for use in energy conversion. Presently incoming sun light is not transformed in its entirely, as much degradation occurs during photon absorption and electron transfer processes. A number of other innovative possibilities have also been researched. With respect to cost and efficiency one of the most promising devices is injection solar cells (ISC). By dint of the dye sensitised solar cell (DSSC) Grätzels findings provided the foundations for much research into this type of solar cell where the light absorbing molecule employed in is a dye.[1] The current is obtained through charge separation in the dye, which is initiated through the connection between the dye and a metal oxide on the one hand and a matched redox couple on the other. In a variant of the DSSC the charge separation processes can also occur between a nanoporous metal oxide and nanoparticles giving rise to a quantum dot sensitised solar cell (QDSSC).[2] The use of nanoparticle (NP) properties can be utilized for the harvesting of solar energy, as demonstrated by Kamat and coworkers[3] who were able to exploit these findings subsequently and prepare a number of nanoparticle based solar cells. Nanoparticle research has comprised a wide field of science and nanotechnology for a number of years. As the size of a material approaches dimensions on the nm scale the surface properties contribute proportionally more to the sum of the properties than the volume due to the increase in the surface to volume ratio. These dimensions also constitute a threshold in which quantum physical effects need to be taken into account. Hence the properties of devices or materials in this size regime are inevitably size dependent. The basic principles can be described by two different theories, one of which is based on molecular orbital theory in which the particle is treated as a molecule. For this reason n atomic orbitals with the same symmetry and energy can build up n molecular orbitals through their linear combination based on the LCAO method (Linear Combination of Atomic Orbitals).[4] In the case of solids the orbitals build up energy bands, where the unoccupied states form the quasi continous conduction band (CB) and the occuppied states form the quasi continous valence band (VB). The energy \"forbidden\" area in between these two bands is called the band gap. The band gap is a fixed material property for bulk solids but depends on size in the case of the nanoparticles. In contrast to the LCAO method, simplified solid state theory will be used throughout the present work, the theoretical background of which is provided by the effective mass approximation.[5] When an absorption of a photon occurs, an exciton (electron-hole pair) can be generated. By promoting an electron (e-) from the valence band into the conduction band a hole (h+) may be said to remain in the valence band. By comparison to bulk solids, in a small particle the free charges can sense the potential barrier i.e. the edges of the nanoparticle. Analogous to the particle in a box model this potential barrier interaction results in an increase in the band gap as the particle size decreases. In a solar cell NPs with a particle size which possess a band gap energy in the near infrared (NIR) may be utilised and therefore the NPs will be able to absorb in this spectral region. However NPs also have the ability to absorb higher energy photons due to the continuum present in their band structure, so that almost the entire sun spectral range from the NIR up to UV wavelengths may be absorbed just by using the appropriate NP material and size. Suitable NPs are metal chalcogenides e.g. MX (where M = cadmium, zinc or lead and X = sulfur, selenium or tellurium) because of their bandgap size[6–10] and their relative band positions compared to those of the semiconductor oxide states. Both the TiO2/CdSe[11–14] and TiO2/CdTe[15–18] systems have already been successfully fabricated and many of the anomalies reported.[3] Much interest in the lead chalcogenides has been generated by reports that they may feature the possibility to exhibit multiple exciton generation (MEG) where the absorption of one high energy photon can result in more than one electron-hole pairs.[19–25] Currently electrochemical impedance spectroscopy (EIS) is being used more and more to clarify processes at polarisable surfaces and materials such as nanoparticles. Likewise this method has been rediscovered in photovoltaic research and its use in the characterisation of DSSCs has been discussed in the literature.[26–31] In a number of publications the evaluation of nanoporous and porous structures has been quite extensively explored.[28,29,32–34] Since the mid-20th century Jaffé’s[35] theoretical work concerning the steady- state ac response of solid and liquid systems lead to the formation of the basics of EIS. Further developments in the measurement technology have lead to a broader range of analysis becoming possible. Nevertheless the most challenging part still remains the interpretation of the results and especially to merge the measured data with the theoretical model. EIS quantifies the changes in a small ac current response at electrode electrolyte interfaces i.e. the rate at which the polarized domain will respond, when an ac potential is applied. In this way dielectric properties of materials or composites, such as charge transfers, polarization effects, charge recombination and limitations can be measured as a function of frequency and mechanistic information may be unveiled. Hence EIS allows one to draw a conclusion concerning chemical reactions, surface properties as well as interactions between the electrodes and the electrolyte. Other very useful tools that may be employed for quantifying electron transfer processes and their time domains are intensity modulated photocurrent spectroscopy (IMPS) and intensity modulated photovoltage spectroscopy (IMVS). IMPS permits the generation of time-resolved plots of particular photo-processes in the system, each of which may be specifically addressed through varying the excitation wavelength. For the IMPS technique a sinusoidal wave with a small amplitude is applied, analogous to that of electrochemical impedance spectroscopy, but in this case the modulation is applied to a light source and not to the electrochemical cell as in EIS.[35] The current response is associated with the photogenerated charge carriers which flow through the system and finally discharge into the circuit. The amount of generated and discharged charge carriers is often different due to the presence of recombination and capture processes in surface or trap states. Ultimately the phase shift and magnitude of these currents reveal the kinetics of such processes. The only processes that will be addressed will be those that occur in the same frequency domain or on the same time scale as that of the modulated frequency of the illuminated light. In the literature some explanation of the kinetics of simple systems can be found and basic theories and introductive disquisitions may be found elsewhere.[36–38] Furthermore in solar cell research a multiplicity of studies are available which give an account of IMPS measurements on TiO2 nanoporous structures. Such studies permitted proof for the electron trapping and detrapping mechanism in TiO2 surface states.[39,40] An analysis of TiO2 electrodes combined with a dye sensitization step was established in the work of Peter and Ponomarev.[41–43] Hickey et.al.[44,45] have previously published kinetic studies on CdS nanoparticle (NP) modified electrodes. A theory was presented which allows for the IMPS data to be the interpreted in the case of CdS NP based electrodes. The back transfer, recombination and surface states have been demonstrated to be important as was determined from their inclusion in the theory. Similar attempts to explain the kinetics of CdS quantum dots are described by Bakkers et.al.[46]. In the present work the most important questions concern the behaviour of the photovoltaic assembly. Such assemblies can be equated with an electrode in contact with an electrolyte. Preliminary remarks about such electrodes as components of an electrochemical cell will be introduced in the first part of chapter 2. Thereafter the properties of electrodes in contact with the electrolyte and under illuminated conditions are illustrated. This is followed by a description of the important electrochemical and opto-electrochemical methods which have been employed in these studies. In particular, two separate subsections are dedicated to the methods of EIS and IMPS and the experimental section which are then linked to the theoretical section. The synthesis of all substances used and the preparation of the solar cell substrates are also dealt with in this section as will the equipment used and the instrument settings employed. The optical response of the working photoactive electrode is not only dependent on the substances used but also on their arrangement and linkage. The substrate which was employed in chapter 3 consists of a nanoporous ZnO gel layer upon which an organic linker has been placed in order to connect the oxide layer with the light absorbing component, the PbS NPs. Chapter 3 deals with the linker dependence on the ZnO layer and reports the typical optical characteristics and assembly arrangements of six different linkers on the ZnO layer which is an important intermediate stage in the fabrication of an ISC. The questions concerning how the type of linking affects the photo response and other electrochemical interactions of the complete solar cell substrate will be outlined in chapter 4. Further an examination of the electrochemical and opto-electrochemical behaviours of the samples will be presented similar to that presented in chapter 3. The most interesting substrate resulting from the investigations as described in chapter 3 and 4 will be used for a more in-depth characterisation by EIS in chapter 5. A suitable model and the results of the calculation of the ISC and the intermediate stages will be presented. The potential dependence, the dependence on the illuminated wavelength and also the size dependence of the PbS nanoparticles will be discussed. It will be revealed that ZnO is chemically unstable in contact with some of the linkers. For that reason the same linker study has been repeated with the more stable TiO2 employed as the wide band metal oxide. Comparisons between the different semiconductor metal oxides are made in chapter 6. In addition a number of open questions which previously had remained unanswered due to the instability of the ZnO can now be answered. In chapter 7 another highly porous structure different from that of the ZnO gel structure has been studied to determine its suitability as an ISC substrate. The structure arises from the electrodeposition of a ZnO reactant in the presence of eosin Y dye molecules. In the end the desorption of the dye provides a substrate with a high degree of porosity. Compared to the ZnO gel which was prepared and used for measurements in chapter 3 and 4, the electrodeposited ZnO is of a higher crystallinity and possesses a more preferential orientation. This results in a lower amount of grain boundaries which in turn results in fewer trap processes and subsequently yields a higher effective diffusion of the electron through the layer.[47,48] Optical and (opto-)electrochemical methods have been used for the basic characterisation of the untreated ZnO/Eosin Y and all other materials used in the fabrication of the ISC and a comparison with the ZnO gel used in chapter 3 and 4 will be made. Finally in chapter 8 an alternative metal oxide structure will be discussed. The background to this last chapter is to examine the influence of the ISC where the oxidic layer is present as a highly periodic arrangement, known as a photonic crystal. The TiO2 metal oxide which was also used in chapter 6 has been structured to form an inverse opal. First preparative findings and the first illustration of the (opto-)electrochemical results are presented. Consequently suggestions for improvements will be made. It is envisaged that the information gathered and presented here will help to achieve a deeper understanding of solar cells and help to improve the device efficiency and the interplay of the materials. Elementary understanding paves the way for further developments which can also contribute to providing devices for more efficient energy conversion.:Contents List of Abbreviations vii Legend of Symbols ix 1 Introduction and Motivation 1 2 Theoretical and Experimental Introduction 7 2.1 Basics of the (Opto-)Electrochemistry . . . . . . . . . . . . . . . . 7 2.1.1 Electrode-Electrolyte Interface Non-Illuminated . . . . . . 8 2.1.2 Electrode-Electrolyte Interface Under Illumination . . . . . 10 2.1.3 The Processes in the Injection Solar Cell (ISC) . . . . . . . 12 2.1.4 Cyclic Voltammetry (CV) . . . . . . . . . . . . . . . . . . 15 2.1.5 Chronoamperometry (CA) . . . . . . . . . . . . . . . . . . 16 2.1.6 Incident Photon to Current Conversion Efficiency (IPCE) . 16 2.1.7 Electrochemical Impedance Spectroscopy (EIS) . . . . . . 17 2.1.8 Intensity Modulated Photocurrent Spectroscopy (IMPS) . 21 2.2 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.2.1 Synthesis of ZnO Sol-Gel . . . . . . . . . . . . . . . . . . . 23 2.2.2 Synthesis of TiO2 Sol-Gel . . . . . . . . . . . . . . . . . . 24 2.2.3 Preparation of the ZnO/Eosin Y Substrate . . . . . . . . . 24 2.2.4 Syntheses and Preparation of the Inverse Opal . . . . . . . 25 2.2.5 The Syntheses for PbS Nanoparticle . . . . . . . . . . . . . 26 2.2.6 Preparation of the PbS Coated Substrates . . . . . . . . . 30 2.2.7 Preparation of the ISC . . . . . . . . . . . . . . . . . . . . 31 2.2.8 Material Characterisations and Instrument Settings . . . . 33 3 The Linker Attachment on a ITO/ZnO Substrate 37 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.2 The ITO/ZnO Film . . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.2.1 The ZnO Layer and the ITO/ZnO Substrate Preparation . 40 3.2.2 The ZnO Structure as a Function of the Sintering Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.3 The Linker on the ITO/ZnO Film . . . . . . . . . . . . . . . . . . 48 3.3.1 The Linker Orientation on the ZnO layer . . . . . . . . . . 48 3.3.2 The Linker Interaction with the ZnO Gel . . . . . . . . . . 52 3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 4 The PbS Sensitized ITO/ZnO/linker Substrate 59 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 4.2 The ITO/ZnO/Linker/PbS Substrate . . . . . . . . . . . . . . . . 61 4.2.1 Spectroscopic Evidence for PbS on the ITO/ZnO/Linker Substrate . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 4.2.2 The Cyclic Voltammetry Study on the Substrates . . . . . 63 4.2.3 The Opto-Electrochemistry on the Substrates . . . . . . . 70 4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 5 The EIS Study of the ITO/ZnO/MPA/PbS Substrate 75 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5.2 The Substrate Assembly . . . . . . . . . . . . . . . . . . . . . . . 77 5.3 The Substrate Characteristics . . . . . . . . . . . . . . . . . . . . 78 5.4 The Model for the EIS Analysis . . . . . . . . . . . . . . . . . . . 83 5.5 The Results of EIS Data Fitting . . . . . . . . . . . . . . . . . . . 86 5.5.1 The EIS Results of the FTO/ZnO Substrate . . . . . . . . 86 5.5.2 The EIS Results of the FTO/ZnO/MPA Substrate . . . . 89 5.5.3 The EIS Results of the FTO/ZnO/MPA/PbS Substrate . . 92 5.5.4 The EIS Results for Shorter Illumination Wavelength . . . 96 5.5.5 The Resistance of the Linker . . . . . . . . . . . . . . . . . 111 5.6 General Remarks on the Modelling . . . . . . . . . . . . . . . . . 112 5.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 6 TiO2 based Injection solar Cell 119 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 6.2 The ITO/TiO2 Film . . . . . . . . . . . . . . . . . . . . . . . . . 121 6.3 The Linker and PbS Attachment on the ITO/TiO2 Substrate . . . 123 6.4 The Cyclic Voltammetry Study on the Substrates . . . . . . . . . 125 6.4.1 The Linker Sensitized ITO/TiO2 Film . . . . . . . . . . . 125 6.4.2 The ITO/TiO2/Linker/PbS Substrate . . . . . . . . . . . 126 6.5 The Opto-Electrochemistry on the Substrates . . . . . . . . . . . 127 6.6 Comparison Between ZnO and TiO2 Based ISCs . . . . . . . . . . 129 6.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 7 ZnO-Eosin Y based Injection Solar Cell 135 7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 7.2 The FTO/ZnO-Ey Film . . . . . . . . . . . . . . . . . . . . . . . 137 7.3 The PbS Attachment to the FTO/ZnO-Ey Film . . . . . . . . . . 137 7.4 The Cyclic Voltammetry Study on the Substrates . . . . . . . . . 140 7.5 The Opto-Electrochemistry on the Substrates . . . . . . . . . . . 142 7.5.1 The Linear Sweep Voltammetry (LSV) Study on the Substrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 7.5.2 The IPCE Measurements on the Substrates . . . . . . . . 144 7.5.3 The Photo Transient Measurements on the Substrates . . . 145 7.6 Comparison between ZnO and ZnO-Ey based ISC . . . . . . . . . 146 7.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 8 Injection Solar Cell meets Photonic Crystal 151 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 8.2 The Opal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 8.3 The Inverse Opal . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 8.4 The Inverse Opal based ISC . . . . . . . . . . . . . . . . . . . . . 159 8.4.1 The Substrate Characteristics . . . . . . . . . . . . . . . . 159 8.4.2 The Cyclic Voltammetry . . . . . . . . . . . . . . . . . . . 160 8.4.3 The Opto-Electrochemistry . . . . . . . . . . . . . . . . . 161 8.4.4 The EIS Measurements . . . . . . . . . . . . . . . . . . . . 163 8.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 9 Overall Conclusion 167 10 Outlook 173 Bibliography I A Acknowledgement XXV B Erklärung XXVII

In der vorliegenden Arbeit wurde mit Hilfe eines Photon-STM die Korrelation zwischen optischen Eigenschaften und der lokalen Morphologie an zwei unterschiedlichen Systemen untersucht. Hierfür wurden zum einem oxidgetragene Ensemble von Silber-Partikeln präpariert, wobei sowohl die Partikelform (Kuppel- und Scheibenform) als auch die deponierte Partikeldichte variiert werden konnte. Neben der Präparation solcher Partikel auf Al10O13/NiAl, konnten sphärische Silber-Kolloide geordnet, als auch ungeordnet auf HOPG aufgebracht und untersucht werden. Dabei zeigte sich, dass das Verhältnis von Höhen zu Breiten nicht nur einen signifikanten Einfluss auf die Mie-Resonanz des einzelnen Partikels hat, sondern auch die elektromagnetische Kopplung der Partikel in einem Ensemble stark kontrolliert. Die energetische Lage der Mie-Resonanz zeigt im Fall der kuppelförmigen Ag-Partikel eine starke Abhängigkeit vom Intepartikel-Abstand, was sich in einer Verschiebung zu höheren Energien für eine steigende Partikeldichte äußert. Eine solche Abhängigkeit konnte bei den Ensembles der scheibenförmigen Partikel nicht beobachtet werden. Des weiteren zeigte sich, dass, verglichen mit den ungeordneten Ensembles, die selbstorganisierte langreichweitige Ordnung der Silber-Kolloide auf HOPG nur einen schwachen Einfluss auf die energetische Position der Mie Resonanz hat.Das zweite hier untersuchte System sind dünne MgO Filme unterschiedlicher Dicken auf einem Mo(001) Substrat. Diese zeigen ein reichhaltiges Wachstumsverhalten, welches durch eine Differenz in den Gitterkonstanten von 5.3% begründet ist und erst ab etwa 25 ML zu einem flachen und defektarmen Film führt. Die so induzierte Spannung relaxiert bis zu einer Dicke von etwa 7 ML in einer periodischen Überstruktur die aus abwechselnd flachen und verkippten Ebenen an der MgO-Mo Grenzschicht hervorgeht. Für MgO Filme mit einer Dicke von etwa 12 ML werden dann Schraubenversetzungen, ausgedehnte verkippte Ebenen und Stufenkanten mit einer Orientierung entlang der Richtung beobachtet. Die optische Charakterisierung durch Feldemission von Elektronen aus der STM-Spitze in den MgO-Film wird dominiert von zwei Emissionsmaxima bei Energien von 3.1 eV und 4.4 eV. Die kontrollierte Nukleation von Gold Partikeln und die Erzeugung von Farbzentren im MgO Film erlaubten eine Zuordnung dieser Emissionen zu strahlenden Zerfällen von Exitonen an Ecken, Kinken bzw. Stufen des Magnesiumoxids. Solche Emissionsprozesse konnten allerdings nur unter Einstellungen beobachtet werden, bei denen ein gleichzeitiges Rastern der Oberfläche unmöglich ist. Bei moderaten Einstellungen war auch eine ortsaufgelösten Spektroskopie möglich, wobei dann neue Emissionsmechanismen beobachtet wurden. Dabei sind zwei Prozesse wesentlich; zum einen die Ausbildung von sog. Spitzen-induzierten Plasmonen im Bereich zwischen Spitze und dem Mo-Substrat, zum anderen strahlende Elektronenübergänge zwischen sog. Feldemissionsresonanzen, die sich im Spitze/MgO-Film System ausbilden. / In this thesis, the correlation between the optical properties and the local morphology of supported silver nanoparticle ensembles and MgO thin films deposited on Mo(001) systems is explored by means of Photon-STM. In the first section, dome and disk shaped Ag nanoparticle ensembles with increasing density on an alumina film on NiAl(110) were analyzed as well as ordered and disordered ensembles of Ag nanocolloids on HOPG. The aspect ratio of the Ag nanoparticles was found to have a significant influence not only on the Mie plasmon resonance of a single particle, but also on the electromagnetic coupling within the nanoparticle ensembles. The Mie resonance in the ensemble of dome shaped Ag nanoparticles shows a strong dependence on the interparticle distance, where it shifts to higher energies with increasing particle density, due to destructive interference effects. In the disk-like Ag ensembles, however, the plasmon energy is independent of particle-particle separation. The long-range lateral ordering of size-selected Ag nanocolloids is found to induce a high dipole-dipole coupling within the ensemble. This is mainly reflected by the enhancement of the spectral intensity of the in-plane Mie mode, due to constructive coupling. However, ensembles with either well-ordered or disordered arrangements reveal no important difference in their optical properties, reflecting the weak influence of the long-range order in the particle ensemble. Thin MgO films with different thicknesses were grown on a Mo(001) surface. The stress resulting from the 5.3% lattice mismatch between the MgO(001) and the Mo(001) lattice parameters is found to control the surface morphology of the MgO film until thicknesses of around 25ML at which flat and defect-poor films are obtained. The relaxation of the stress induces a periodic network in the first 7ML of the MgO film, consisting of alternated flat and tilted mosaics. The presence of screw dislocations, steps oriented along the MgO directions, and tilted planes is observed when the MgO films are approximately 12ML thick. In addition, an increase of the MgO work function around these new surface features is revealed from STM spectroscopy. The photon emission induced by field-emitted electron injection from the STM tip into the MgO films is dominated by two emission bands located at 3.1eV and 4.4eV. To check the origin of these bands, further experiments, namely, nucleation of Au particles and creation of F-centers on the MgO surface, have been performed. The nucleation of Au particles at the low coordinated sites is found to quench the MgO optical signal, while the creation or annihilation of F-centers does not alter the MgO emission bands. The 3.1eV and the 4.4eV bands are therefore assigned to the radiative decay of MgO excitons at corner and kink sites, and step sites, respectively. Besides, spatially resolved optical measurements in the tunneling mode of the STM revealed different light emission mechanisms. These radiative processes are mainly related to tip-induced plasmons that form between the tip and the Mo support and to electron transitions between field-emission-resonance states in the STM tip-MgO film junction. The signal from exciton decays at corners and kinks of the MgO surface is however only observed at excitation conditions where the spatial resolution is already strongly reduced.