Renewable forms of energy are becoming increasingly important as the world quickly depletes its current energy reserves, and rapidly increases the concentration of pollutants in our environment. Solar technology based on organic semiconductors provides a promising candidate to fulfill a portion of our future energy needs in an environmentally sustainable manner. Organic semiconductors are a collection of pi-conjugated small molecules and polymers which can be implemented in photovoltaic cells that are potentially quite low cost. Currently, however, their commercial applications are limited due to a relatively low efficiency in converting sunlight into usable power. The fundamental physics of such devices must be clarified if these materials are to compete with traditional inorganic solar cells. In this dissertation, two emerging experimental tools are implemented in investigations of the internal electric fields present within operating organic photovoltaic cells. The first set of investigations utilizes the vibrational Stark effect to quantify the electric fields which often form at the interfaces between two organic semiconducting materials. Such interfaces are at the heart of the photocurrent generation process in these devices and any electric fields formed crucially alter device performance. We quantitatively determine the interfacial field present in blends of poly(3-hexylthiophene) (P3HT) and phenyl-C61-butyric acid methyl ester (PCBM) and show that this field depends strongly on annealing conditions. Finally we discuss a correlation between this interfacial electric field, crystalinity and device performance. The second set of investigations take advantage of electric field induced second harmonic generation microscopy to examine the electric potential across active organic solar cells. We again investigate blends of PCBM and P3HT as well as poly(4,4-dioctyldithieno(3,2-b:2',3'-d)silole)-2,6-diyl-alt-(2,1,3-benzothiadiazole)-4,7-diyl) (PSBTBT) and PCBM. In the former we find that the potential drop across the device shifts dramatically over time under illumination, while in the latter we find a nearly linear drop which remains constant through device operation. We then extend our examinations of PSBTBT:PCBM with EFISH by quantifying the extent of space charge accumulation throughout such devices. / text

In this Ph.D. work, we investigate the optoelectronic properties of low-bandgap semiconducting polymers and project the potential for employing these materials in electronic and photonics devices, with a particular emphasis on use in organic solar cells. The field of organic solar cells is well developed and many of the fundamental aspects of device operation and material requirements have been established. However, there is still more work to be done in order for these devices to ultimately reach their full potential and achieve commercialization. Of immediate concern is the low power conversion efficiency demonstrated in these devices so far. In order to improve upon this efficiency, several routes are being explored. Because the optical bandgaps of semiconducting polymers are larger than in inorganic semiconductors, one of the most promising routes currently under exploration is the development of low-bandgap materials. Using polymers with lower band gaps will allow more of the solar irradiance spectrum to be absorbed and converted into electricity and thus possibly boost the overall efficiency. The bandgap of these semiconducting polymers is determined by the chemical structure, and therefore can be tailored through synthesis if the relevant structure-property relationships are well-understood. The materials studied in this work, a new series of Poly(thienylenevinylene) (PTV) derivatives, posses lower band gaps than conventional polymers through a design that incorporates aromatic-quinoid structural disturbances. This type of chemical structure delocalizes the electronic structure along the polymer backbone and reduces the energy of the lowest excited-state leading to a smaller band-gap. We investigate these materials through a variety of techniques including linear spectroscopy such as absorption and photoluminescence, pump-probe techniques like cw-photoinduced absorption and transient photo-induced absorption, and the non-linear electroasborption technique in order to interrogate the consequences of the delocalized electronic structure and its response to optical stimuli. We additionally consider the effects of environmental factors such as temperature, solvents and chemical doping agents. During the course of these investigations, we consider both of the two primary categorical descriptions of structure-property relationships for polymers within the molecular exciton model, namely the role of inter-molecular interactions on the electronic properties through the variation of supermolecular order and the fundamental determination of electronic structure due to specific intra-molecular interaction along the backbone of the polymer chain. We show that the dilution of aromaticity in semiconducting polymers, while being a viable means of reducing the optical band gap, results in a significant increase in the role of electron-electron interactions in determining the electronic properties. This is observed to be detrimental for device performance as the highly polarizable excited state common to polymers gives way to highly correlated state that extinguishes both the emissive properties and more importantly for solar cells, the charge-generating characteristics. This situation is shown to be predominant regardless of the nature of interchain interactions. We therefore show that the method of obtaining low-bandgap polymers here comes along with costly side-effects that inhibit their efficient application in solar cells. Further, we directly probe the efficacy of these materials in the common bulk-heterojunction architecture with both spectroscopy and device characterization in order to determine the limiting and beneficial factors. We show that, while from the point of view of absorption of solar radiation these low-bandgap polymers are more suited for solar cells, the ability to convert the absorbed photons into electron-hole pairs and generate electricity is lacking, due to the internal conversion into the highly correlated state and thus, the absorbed photon energy is lost. For completeness, we fabricate devices and verify that both the charge-transport properties and alignment of charge extraction levels with those of the contacts can not be responsible for the dramatic decrease in efficiency found from these devices as compared to other higher band gap polymers. We thus conclusively determine that the lack of power converison efficiency is governed by the inefficiency of charge-generation resulting from the intrinsic defective molecular structures rendering a low-lying optically forbidden state below the lowest optical allowed state that consumes the majority of the photogenerated excitons. It is emphasized that our means of investigation allow us to truly access the potential of these materials. In contrast, the direct application of these systems in devices and interpretation of the performance is exceedingly complex and may obscure their true potential. In other words, poor performance from a device may be extrinsic in nature and the optimization process may be very costly with respect to both time and materials. The methods used here however, allow us to determine the intrinsic potential. Not only is this beneficial in terms of preserving the resources that would be used on the trial-and-error method for devices, but it also allows us to learn more on a fundamental level about the structure-property relationships and their implications for device performance. The benefits of this increased understanding are two-fold. First, by learning about the fundamental response of a material, a new application may be realized. For example, the rapidly efficient internal conversion process that renders the materials in this study as poor candidates for solar cells may make them useful for photonics applications, as optical switches, for instance. Secondly, this type of investigation has implications for the whole organic electronics community instead of just being limited to the particular material system and the primary application attempted. In this case, we are essentially able to determine a threshold for aromaticty necessary in a structure that will preserve the stability of the ionic excited state that is useful for charge generation in solar cells.

Les récents développements dans le domaine du photovoltaïque ont permis l'apparition de cellules utilisant des technologies nouvelles. Parmi elles on trouve les cellules photovoltaïques hybrides, cependant les méthodes de fabrication utilisées actuellement présentent des défauts. Cette thèse a pour but de proposer deux nouvelles approches pour la préparation de cellules photovoltaïques hybrides sous forme d'hétérojonctions en volume d'un oxopolymère de titane et d'un polythiophène. Dans un premier temps la formation de mésostructures vermiculaires dans des couches minces de TiO2 par autoassemblage par évaporation a été étudiée, l'oxopolymère amorphe obtenu a ensuite été cristallisé en conditions douces. Ces résultats ont ensuite été utilisés afin de préparer des matériaux hybrides à partir d'homopolymères de thiophène portant des substituants hexyl et acide, ainsi qu'à partir de copolymères. Des matériaux hybrides ne présentant pas de macroségrégation ont été obtenus pour un polymère portant des fonctions acides et pour les copolymères. Dans un second temps l'électrochimie du ferrocène et du cuivre ont été étudiés dans des films de TiO2 mésostructuré et mésoporeux, puis deux dérivés thiophène : le mot et l'edot ont été électropolymérisés dans ces structures. Des cellules photovoltaïques ont été préparées en utilisant ces matériaux hybrides et caractérisées par des mesures de la courbe I/E ainsi que par l'étude du rendement quantique externe, des facteurs de forme et des rendements ont été calculés. En conclusion, deux nouvelles approches de synthèse de matériaux hybrides ont été proposées et menées à bien les propriétés photovoltaïques et ces matériaux ont été mesurées. / Recent advances in the field of photovoltaics have led to the emergence of new solar cell technologies. Among them can be found the hybrid solar cells, unfortunately the way such cells are built is still a source of problems. The aim of this phd is to develop two new approaches in the synthesis of hybrid materials as bulk heterojunctions. In first place the titanium dioxide component vas prepared by sol-gel process and its mesostructure was studied, low temperature crystallization was also investigated. Those results were used in order to prepare hybrid materials from preformed polymers. The behavior of polythiophènes with hexyl and carboxylic acid functions were used as well as copolymers bearing both functions. Hybrids without macrosegregations phenomena were obtained using acid bearing homopolymers as well as copolymers. The second approach was about investigating the electrochemical behavior of ferrocene and copper ions inside the mesoporosity, this first study was followed by a study of the electropolymerization of mot and edot inside the porosity in order to prepare hybrid materials. The obtained hybrids were studied in solar cells by measuring the I/V curve as well as the external quantum efficiency, fill factors and efficiencies were also obtained. To conclude, both approaches leaded to hybrid materials with measurable photovoltaic properties.

This thesis discusses vacuum deposited organic solar cells. It focuses on the investigation of new donor molecules blended with the standard electron acceptor C60. These donor-acceptor heterojunctions form the photoactive system of organic solar cells. In addition, the influence of the processing conditions on the morphology of the blend layers is investigated, as the morphology is crucial for an efficient generation of free charge carriers upon photon absorption. Bulk heterojunction solar cells with the donor DTDCTB are deposited at different substrate temperatures. We identify three substrate temperature regimes, discriminated by the behavior of the fill factor (FF ) as a function of the blend layer thickness. Devices deposited at RT have a maximum FF between 50 and 70 nm blend thickness, while devices deposited at 110 °C have a monotonically decreasing FF. At Tsub=85 °C, the devices have an S-kinked current-voltage curve. Grazing incidence wide angle X-ray scattering measurements show that this peculiar behavior of the FF is not correlated with a change in the crystallinity of the DTDCTB, which stays amorphous. Absorption measurements show that the average alignment of the molecules inside the blend also remains unchanged. Charge extraction measurements (OTRACE) reveal a mobility for the 110 °C device that is an order of magnitude higher than for the RT device. The difference in mobility can be explained by a higher trap density for the RT samples as measured by impedance spectroscopy. Despite slightly higher carrier lifetimes for the RT device obtained by transient photovoltage measurements, its mobility-lifetime product is still lower than for the 110 °C devices. Based on DTDCTB, three new donor materials are designed to have a higher thermal stability in order to achieve higher yields upon material purification using gradient sublimation. For PRTF, the thermal stability is increased demonstrated by a higher yield upon sublimation. However, all new materials have a reduced absorption as compared to DTDCTB, which limits the short current density, and the FF is more sensitive to an increase of the blend layer thickness. The highest power conversion efficiency is achieved for a PRTF:C60 solar cell with 3.8%. Interestingly, PRTF:C60 solar cells show exceptionally low nonradiative voltage losses of only 0.26 V. Another absorber molecule is the push-pull chromophore QM1. Scanning electron microscope (SEM) measurements show a growth of the molecule in nanowires on several substrates. The nanowires have lengths up to several micrometers and are several tens of nanometers wide. The formation of the nanowires is accompanied by a strong blue shift (650 meV) of the thin film absorption spectrum in comparison to the absorption in solution, which is attributed to H-aggregation of the molecules. Furthermore, the thin film absorption onset reaches up to 1100 nm, making the material a suitable candidate for a near infrared absorber in organic solar cells. For a solar cell in combination with C60, a power conversion efficiency of 1.9% was achieved with an external quantum efficiency of over 19% for the spectral range between 600 and 1000 nm. The method of “co-evaporant induced crystallization” as a means to increase the crystallinity of blend layers without increasing the substrate temperature during the deposition is investigated. Mass spectrometry (LDI-ToF-MS) measurements show that polydimethylsiloxane (PDMS), which is used as a co-evaporant, decomposes during the evaporation and only lighter oligomers evaporate. Quartz crystal microbalance (QCM) measurements prove that the detection of PDMS saturates at higher amounts of evaporated material. LDI-ToF-MS measurements show further that the determination of the volatilization temperature by QCM measurements is highly error prone. The method was applied to zinc phthalocyanine (ZnPc) :C60 solar cells, accepting the insertion of PDMS into the blend layer. Diffraction (GIXRD) measurements show a large increase in crystallinity. ZnPc:C60 solar cells produced by applying the method reveal a similar behavior as solar cells processed at a higher substrate temperature.

This thesis discusses vacuum deposited organic solar cells. It focuses on the investigation of new donor molecules blended with the standard electron acceptor C60. These donor-acceptor heterojunctions form the photoactive system of organic solar cells. In addition, the influence of the processing conditions on the morphology of the blend layers is investigated, as the morphology is crucial for an efficient generation of free charge carriers upon photon absorption. Bulk heterojunction solar cells with the donor DTDCTB are deposited at different substrate temperatures. We identify three substrate temperature regimes, discriminated by the behavior of the fill factor (FF ) as a function of the blend layer thickness. Devices deposited at RT have a maximum FF between 50 and 70 nm blend thickness, while devices deposited at 110 °C have a monotonically decreasing FF. At Tsub=85 °C, the devices have an S-kinked current-voltage curve. Grazing incidence wide angle X-ray scattering measurements show that this peculiar behavior of the FF is not correlated with a change in the crystallinity of the DTDCTB, which stays amorphous. Absorption measurements show that the average alignment of the molecules inside the blend also remains unchanged. Charge extraction measurements (OTRACE) reveal a mobility for the 110 °C device that is an order of magnitude higher than for the RT device. The difference in mobility can be explained by a higher trap density for the RT samples as measured by impedance spectroscopy. Despite slightly higher carrier lifetimes for the RT device obtained by transient photovoltage measurements, its mobility-lifetime product is still lower than for the 110 °C devices. Based on DTDCTB, three new donor materials are designed to have a higher thermal stability in order to achieve higher yields upon material purification using gradient sublimation. For PRTF, the thermal stability is increased demonstrated by a higher yield upon sublimation. However, all new materials have a reduced absorption as compared to DTDCTB, which limits the short current density, and the FF is more sensitive to an increase of the blend layer thickness. The highest power conversion efficiency is achieved for a PRTF:C60 solar cell with 3.8%. Interestingly, PRTF:C60 solar cells show exceptionally low nonradiative voltage losses of only 0.26 V. Another absorber molecule is the push-pull chromophore QM1. Scanning electron microscope (SEM) measurements show a growth of the molecule in nanowires on several substrates. The nanowires have lengths up to several micrometers and are several tens of nanometers wide. The formation of the nanowires is accompanied by a strong blue shift (650 meV) of the thin film absorption spectrum in comparison to the absorption in solution, which is attributed to H-aggregation of the molecules. Furthermore, the thin film absorption onset reaches up to 1100 nm, making the material a suitable candidate for a near infrared absorber in organic solar cells. For a solar cell in combination with C60, a power conversion efficiency of 1.9% was achieved with an external quantum efficiency of over 19% for the spectral range between 600 and 1000 nm. The method of “co-evaporant induced crystallization” as a means to increase the crystallinity of blend layers without increasing the substrate temperature during the deposition is investigated. Mass spectrometry (LDI-ToF-MS) measurements show that polydimethylsiloxane (PDMS), which is used as a co-evaporant, decomposes during the evaporation and only lighter oligomers evaporate. Quartz crystal microbalance (QCM) measurements prove that the detection of PDMS saturates at higher amounts of evaporated material. LDI-ToF-MS measurements show further that the determination of the volatilization temperature by QCM measurements is highly error prone. The method was applied to zinc phthalocyanine (ZnPc) :C60 solar cells, accepting the insertion of PDMS into the blend layer. Diffraction (GIXRD) measurements show a large increase in crystallinity. ZnPc:C60 solar cells produced by applying the method reveal a similar behavior as solar cells processed at a higher substrate temperature.