Solid-state dye-sensitized solar cells (ssDSCs) offer the possibility of high power conversion efficiencies (PCEs) of over 20%. However, after more than a decade of research, devices still barely reach over 7% PCEs. In this thesis, limitations to device performance are studied in detail, and solutions for future advancement are put forward. In the first part of the thesis, factors limiting charge generation are explored by studying the crystallization environment of mesoporous TiO2 self-assembled through block copolymers. It was found that the density and distribution of sub band gap states are a function of the synthesis conditions and critically affect the performance characteristics of the self-assembled titania used in ssDSCs. As a result, the self-assembled mesoporous oxide system presented in this thesis outperforms for the first time the conventional nanoparticle based electrodes fabricated and tested under the same conditions, with demonstrated PCEs of over 5%. In chapters 6, 7, and 8, the factors limiting the diffusion length and hence, the thickness of the fabricated devices, are carefully examined. Previous literature points towards insufficient pore-filling of the hole transporting material (HTM) as the main limiting factor. In chapter 6, a pore-filling study is shown where a new technique to evaluate the pore-filling fraction of the HTM in the conventional mesoporous metal oxide electrode is also presented and conclude that sufficient pore-filling of thick films can easily be achieved. Another usual strategy to extend the electron lifetime in the devices and thus, the charge diffusion length, involving thin film coatings of insulating metal oxides is examined in chapter 7, with satisfactory results for SnO2-based ssDSCs. The diffusion length can also be extended if the factors limiting the diffusion of charges through the device are identified and removed, as presented in chapter 8. Finally, a study on the stability of the ssDSC is presented in chapter 9. The developments achieved enable long term stability to be effectively targeted, and represent a key milestone towards commercial realization of ssDSCs.

This thesis presents an experimental study of the incorporation and optimization of sol-gel processed aluminum doped zinc oxide (AZO) thin films in solar cell devices. I first optimized the optoelectronic properties of AZO thin films by manipulating the dopant incorporation, choice of precursor chemicals and post deposition anneal treatments. Results showed that improved performance could be attributed to several factors, including improved charge carrier concentration, mobility and conductivity. AZO thin films with transmittance of over 90% and resistivity values of the order of 10-2 Ω•cm have been achieved. I also demonstrated the successful application of these AZO thin films in organic photovoltaics (OPV), to serve as an alternative to ITO electrodes. I demonstrated greater than 2-fold improvement in device efficiency through the modification of the front contact/polymer interface using zinc oxide buffer layers. This improved the charge selectivity of the electrodes and energy level alignment at the interface while reducing the recombination of separated charges and the device's series resistance. Finally, I showed that the efficiency of inverted ZnO/PbS quantum dots solar cells can be enhanced by optimizing the p-type PbS thickness, UV treating the n-type ZnO layer and exposing the devices in the dark to nitrogen. Both ZnO and AZO systems were studied, and efficiency enhancement were demonstrated for a range of Al content from 0 to O.4at.%.

This research comprehensively reviews the properties of regular micro and nano structures fabricated by laser interference lithography and reports on their applications in the antireflection and self‐cleaning surface. The research systematically investigates the laser interference lithography technology taking into account its advantages and abilities to realize various potential applications. Multiple‐beam interference lithography systems are constructed. Laser interference interaction with silicon wafer is analysed and the optical and hydrophobic properties are obtained via measurements. In order to fabricate the extremely low reflection and very large contact angle for solar cells, fabrication methods of antireflection and self‐cleaning are surveyed and their advantages and disadvantages compared. The research investigates the effect of heat transfer and the radiation of laser interference plasma on silicon wafer surfaces and proposes equations of heat flow and radiation effects of laser plasma of interfering patterns in a four‐beam laser interference distribution. Following the irradiation, the silicon wafer surface is covered with a periodic array of micrometer and nanometer‐sized structures, which have the shape of grating, cone and hole. The research also investigates the effect of different laser parameters on the optical and hydrophobic properties of the structured silicon wafer surface. The results of periodic hexagonally‐distributed hole structures fabricated by three‐beam laser interference reveals excellent design guidelines for obtaining an extremely low solar‐weighted reflection, (SWR, 1.86%) and relatively large contact angle (140°) which can provide a strong self‐cleaning capability on the solar cell surface. In addition, the research creates a novel dual structure with antireflection and superhydrophobic properties fabricated by three‐beam laser interference lithography. The fabrication method is three‐beam laser interference combined with focused laser processing interacting on the silicon wafer surface. This kind of structure has a very low SWR (3.6 %) and extremely large contact angle which is more than 150° in the wavelength range from 380 nm to 780 nm. The research shows that the laser interference lithography technology can be employed and further developed to fabricate micro and nano structures of strong antireflection and self‐cleaning functions for applications in solar cells.

The emergence of semiconducting nanowires as the new building blocks for photovoltaic (PV) devices has drawn considerable attention because of the great potential of achieving high efficiency and low cost. In special, nanowires with a coaxial structure, namely, core-shell structures have demonstrated significant advantages over other device configurations in terms of radial charge collection and cost reduction. In this dissertation, several core-shell nanowire structures, including ZnO/ZnSe, ZnO/ZnS, and CdSe/ZnTe, have been synthesized and the photovoltaic devices processed from a ZnO/ZnS core-shell nanowire array and a single CdSe/ZnTe core-shell nanowire have been demonstrated. By combining the chemical vapor deposition and pulsed laser deposition (PLD) techniques, type-II heterojunction ZnO/ZnSe and ZnO/ZnS core-shell nanowire array were synthesized on indium-tin-oxide substrates. Their structures and optical properties have been investigated in detail, which revealed that, despite highly mismatched interfaces between the core and shell, both systems exhibited an epitaxial growth relationship. The quenching in photoluminescence but enhancement in photocurrent with faster response upon coating the core with the shell provides the evidence that the charge separation and collection in the type II core-shell nanowire is greatly improved. This demonstration brings much greater flexibility in designing next generation PV devices in terms of material selection and device operation mechanisms for achieving their maximum energy conversion efficiencies at a low cost and in an environmentally friendly manner. In order to achieve a high quality interface in the core-shell nanowire, CdSe and ZnTe, which have close lattice parameters and thermal expansion coefficients, were chosen to fabricate nanowire solar cells. ZnTe and CdSe nanowires were first synthesized by thermal evaporation and the shells were subsequently deposited by PLD. ZnTe/CdSe nanowires represented an inhomogeneous coating while the CdSe/ZnTe core-shell exhibited a conformal coating with obvious ZnTe eptilayer. The final PV device based on an individual CdSe/ZnTe nanowire demonstrated an efficiency of ~1.7%. In addition, a controllable synthesis of CdSe nanowire array on muscovite mica substrate was presented, providing the possibility to harvest hybrid energies in an all-inorganic nanowire array.

In 2013, the worldwide production of renewable electricity accounted for 22.1% of the total energy production with 0.9% coming from solar photovoltaics (PVs). Recently, there has been a growing interest for Cu2ZnSnS4 (CZTS) quaternary semiconductor due to the abundance and low cost of its precursors. Moreover, this chalcopyrite material has an ideal direct band gap around 1.5 eV, high absorption coefficient (α >104 cm-1) and high conductivity, making it suitable for nanostructured and dye-sensitized solar cell (DSSC) applications. Here, CZTS nanoparticles have been synthesized by pulsed laser deposition (PLD) and simultaneously deposited in the interstitial space of ZnO nanowire arrays to form bulk heterojunction 3D nanostructured solar cells. Secondly, vertically oriented CZTS nanoplates have been synthesized by PLD and used as counter electrode in platinum-free dye-sensitized solar cells. These CZTS nanostructures proved to be suitable in achieving workable solar cells, which could significantly cut down the cell cost and provide environmentally friendly photovoltaic devices. Alternately, hybrid organic–inorganic perovskite solar cells have become one of the most attractive photovoltaic technologies with easy solution fabrication and high conversion efficiencies. However, the devices remain unstable under certain processing and environmental conditions. Herein, formamidinium lead tri-halide perovskite (FAPbI3) planar heterojunction solar cells have been fabricated under a controlled environment. The fabrication parameters (precursor concentration, annealing, etc) and the effect of humidity on the structural, optical, and electrical properties of FAPbI3 thin films and devices have been investigated and proved to be critical in the processing of efficient devices. Solar cells with conversion efficiency of 16.6% have been obtained. Furthermore, in-situ techniques such as in-situ (scanning) transmission electron microscopy and in-situ XRD were performed to understand the crystallization and degradation mechanisms of FAPbI3 thin films.The in-situ data were correlated with planar heterojunction FAPbI3 devices efficiency data in order to improve the device fabrication process.

M.Sc. / Techniques for the fabrication of polycrystalline silicon solar cells have advanced in recent years with efficiencies exceeding 17%. The major advantage of polycrystalline silicon is its low cost relative to single-crystalline silicon. The disadvantage is the significantly smaller minoritycarrier bulk diffusion length and inhomogeneous nature of the material. These two drawbacks are due to the presence of grain boundaries as well as high concentrations of dislocations and other physical and chemical defects. In this study the experimental conditions were determined to fabricate solar cells on polycrystalline silicon substrates. The controlled diffusion of phosphorous into silicon and subsequent evaluation of the doped layers (by spreading resistance profiling and chemical staining) were important aspects of this study. From these results the diffusion parameters (i.e. temperature and reaction times) could be optimized in order to improve the solar cell output parameters. Additional material improvement (increase in surface- and bulk minority carrier lifetimes) was demonstrated by the hydrogen passivation of electrically active defects in polycrystalline silicon. However. measurements on hydrogenated silicon samples also indicated that excess passivation can result in surface damage and subsequent reduction in the minority carrier lifetimes. Preliminary solar cells were fabricated on polycrystalline silicon with efficiencies ranging between 0.5 and 6% (total area = 16 cm2).

I takt med en förstärkt global uppvärmning har åtgärdsplaner skrivits om att begränsa temperaturökningen, där många länder är överens om att mängden koldioxidutsläpp måste minska. Lösningen till det anses dels som att gå från en fossilberoende energianvändning till en förnybar. Installering av solceller växer och i Sverige kan 5 – 10 % av elproduktionen förväntas komma från solkraft år 2040. Det kvarstår dock en hel del problematik kring effektförlust till följd av partiell skuggning på solceller. Idag tillverkas solpaneler med inbyggda bypass-dioder som förbättrar elproduktionen vid skugga. Forskning påstår även att skuggningsproblematik kan minskas genom att förändra konfigurationen mellan solcellspaneler. Idag sitter panelerna vanligen ihop i serie- och parallellkopplingar, men det kan finnas energivinster i att byta ut mot tvärkopplingar. Syftet med denna studie har varit att ge beslutsunderlag för investerare av solceller, som planerar att placera dem där det finns risk för att skugga kan uppstå. Målet delades in i två delar. Det fanns en experimentell del som jämförde elproduktion utan och med skugga för två olika typer av systemkonfigurationer där solpanelerna var serie-, respektive tvärkopplade. Det fanns även en simuleringsdel som beräknade producerad elenergi på årsbasis för en solpanel som var placerad intill en skorsten. Utifrån det skapades riktlinjer för hur solpaneler bör placeras för att minimera skuggproblematik av närliggande föremål som riskerar att skugga. Experimenten utfördes utomhus och till det användes 16 monokristallina solmoduler, som först sammankopplades i serie och därefter i tvärkopplingar. Systemen anslöts till en växelriktare som omvandlade från likström till växelström. Först uppmättes producerad eleffekt för systemen, varav en modul utsattes för olika skuggförsök. Därefter separerades en solmodul från anläggningen och samma skuggförsök genomfördes på modulen. Resultaten påvisade att vid solcellssystemen hade en verkningsgrad på 16 % för solinstrålning mellan 400 – 700 W/m2. För intensitet lägre än 400 W/m2 och högre än 700 W/m2 avtog verkningsgraden med 1 – 2 %. Hög solinstrålning inträffar vanligen mitt på dagen och solcellerna har troligen då en högre temperatur, vilket sänker effektiviteten. Experimenten påvisade att ju fler av en solmoduls slingor som skuggan faller på, desto lägre blir den producerade eleffekten. Även mängden skugga som drabbar panelen spelar roll. Det beror på att om skugga enbart faller på en slinga kan de övriga arbeta ostört och får därför en högre maxeffektpunkt (MPP) på IU-kurvan. Vid full skugga på en modul minskade effekten med 7,72 % för det seriekopplade systemet (S), respektive 13,87 % vid tvärkoppling (TCT). Att det tvärkopplade systemet producerar lägre effekt vid skugga beror dels på att kretsen innehåller en högre ström, troligen för hög för vad bypass-dioderna klarar av. Om strömmen i en skuggad modul leds genom dess bypass-dioder utvecklas ingen spänning i modulen. Om systemet innehåller fler parallellkopplade strängar kommer även de oskuggade strängarna att minska sin spänning, då den är identisk över hela systemet. Den tvärkopplade kretsen innehöll två parallellkopplade strängar. Känslighetsanalysen där en solmodul avlägsnades från systemet och skuggades enskilt, redovisade snarlika resultat som för systemen. Att placera skuggobjektet längre ifrån visade sig i några försök ge en högre elproduktion, då det är enbart den direkta strålningen som tas bort. Resultaten från simuleringen uttrycktes i enheten ”skuggprestanda”, som anger hur mycket en solmodul ger skuggad jämfört med vad den hade gett oskuggad. Störst betydelse för resultaten hade skorstenens höjd och hur långt ifrån skorstenen som solpanelen placeras, samt val av väderstreck. Lägst skuggprestanda blev 41 % för en seriekopplad modul. Det för en 2 m hög skorsten som placerades intill solpanelen. Resultaten påvisade att skuggprestandan avtog med en ökad höjd på skorstenen, men för en tillräckligt hög skorsten förändrades inte skuggprestandan. Genom att välja ett avstånd som var tillräckligt långt ifrån skorstenen bidrog skuggan inte lika mycket till minskning av elproduktionen. Ju mer sydligt skorstenen var placerad om panelen krävdes ett längre avstånd. Följande riktlinjer för placering av solceller intill skuggande föremål rekommenderas för att garantera att solcellerna ger minst 95 % av vad de skulle ha gjort oskuggade. För ett skuggande föremål med höjden 0,5 m bör solcellerna placeras 0,3 m ifrån föremålet. För ett skuggobjekts höjd på 2 m bör avståndet vara 2,1 m. / The effects of global warming has lead to a temperature rise in the atmosphere. Many countries have agreed to reduce the greenhouse gas emissions in order to mitigate climate change. The energy production today is highly dependent on fossil fuels and thus one solution to reduce emissions is to use more renewable resources. Solar cells are today growing on the market and in Sweden one can expect that solar energy will represent 5 – 10 % of the electricity production in year 2040. Partial shading is still a problem causing losses in the electricity production for solar cells. To reduce the losses due to shading bypass-diodes are built-in on modules. Further successful method can also be to change the interconnections among moduls in a system. At the time of writing this paper, configurations with series-parallel connections between modules are most commonly used, but during partial shading it has been shown that increasing interconnections can increase the outgoing effect. The purpose of this study was to show guidelines for the investor of solar cells, who plans an installation near objects that will creates shadows on the cells. The target consisted of two parts. First, there was an experimental objective which compared two different system configurations where the solar modules first were connected in series and thereafter in “Total Cross Tied”. The other part of the study considered a theoretical simulation to answer questions as how the energy production during a year was changed for a solar panel by placing it near a chimney. Thereafter it should also state guidelines on how to place the panel around the chimney so that the shading effects are no noticeables. The experiments were done outdoors and the system consisted of sixteen monocrystalline solar panels. First the panels were connected in series (S) and thereafter in “Total Cross Tied” (TCT). The system included also an inverter that changed direct current from the solar panels to alternating current. In the first experiment the electricity produced from the systems was measured, whereof one module was exposed to different types of shading. As a final experiment, one module was separated from the system where is was exposed to the same shading pattern as before. The results showed that the PV systems had an efficiency of 16 % for an irradiance between 400 – 700 W/m2. For lower irradiances than 400 W/m2 and higher than 700 /m2 the systems experienced a decreasing efficiency of about 1 – 2 %. A high irradiance occurs normally during the middle of the day when the solar cells have a higher temperature, which decreases the efficiency. The experiments proved that shading more cells so that current will flow through more bypass diodes resulted in a lower electricity production. For example if only one bypass diode is activated it means that the remaining cells that are not connected to that bypass diode can work freely and gets a higher maximum power point (MPP) on the IU characteristics. When a module got completely shaded it reduced the electricity output power with 7,72 % for the system with series connected modules. For the TCT-configuration the system experienced a reduction of 13,87 %. These results may be a contradiction according to previously research. A conceivable explanation is that the TCT-configuration produces higher current, which can exceed the allowed value for a functionally bypass diode. If the current in a shaded module flows through its bypass diodes it results in a short-circuit of the module so that it cannot contribute to the voltage. The more parallel PV arrays a system includes, a higher effect loss will occur since the unshaded strings may experience an identical voltage drop as the shaded one. The TCT-configuration included two strings connected in parallel. The sensitivity analysis where one module was separated from the whole system showed nearly the same results as for the whole system. Placing the shading object far away from the module resulted in a higher electricity production for some tests, as it is only the direct radiation that disappears. The simulations results were expressed in terms of “shading performance” and it showed how much a shaded solar module produced compared to an unshaded. It was shown that the parameters that had the biggest influence on the results were the height of the chimney and how far the solar module was placed with respect to the chimney. As a worst-case scenario, the shading performance was 41 % for a series-connected module. This was for a chimney with 2 m of height that placed next to the solar module. The results indicated that having a higher chimney had a decreasing effect of the shading performance for the solar module. To get a higher shading performance one can choose to place the solar cells at a larger distance from the chimney. Conclusively, placing the solar module so that the chimney is located to the south of the module the shadow will affect the panel more, and a longer distance is therefore needed. If the solar cells are located on a distance that is at least 0,3 m from a shading object with a height of 0,5 m, it means that the solar cells will produce 95 % of what that would have given unshaded. For a shading object with a height of 2 m it means that the solar cells must be placed at a distance of 2,1 m from the object.

L’une des tendances naturelle des technologies photovoltaïque est la réduction systématique de l’épaisseur des cellules solaires, que cela soit pour des raisons de coûts, d’économie d’éléments rares ou toxiques ou encore pour limiter les recombinaisons. Jusqu’à présent, les technologies couche minces cristallines (GaAs) et poly-crystallines (CIGS) trouvent des optimum d’efficacité de conversion pour des épaisseurs aux alentours de 1 ou 2 microns. Typiquement, cette gamme d’épaisseur ne nécessite pas de nouvelles solutions de piégeages optiques comme cela est le cas pour la filière silicium amorphe. Cependant, si l’on veut réduire ces épaisseurs d’un facteur 10 voire même 100 afin de s’orienter vers les nouveaux concepts de collections et conversions (GaAs ou GaSb) ou encore de réduire l’utilisation d’indium (CIGS), de nouveaux besoin en matière d’absorption efficace de la lumière sont nécessaires pour ces technologies. Ce manuscrit de thèse se concentre sur la conception, la simulation et la réalisation de solutions nanophotoniques nouvelles pour de futures cellules solaires cristallines ultrafines.Dans un premier temps, nous nous sommes engagé dans une approche en rupture avec la conception habituelle des cellules solaires pour piéger la lumière dans une cellule ultrafine (≤100 nm de matériaux couche-mince (GaAs, GaSb et CIGS). Nous proposons un réseau métallique nanostructuré placé en face avant de la cellule reportée sur un miroir métallique afin d'obtenir une absorption très élevée et multi-résonante, indépendante de l’angle d’incidence et de la polarisation. Une analyse numérique approfondie des mécanismes résonants en jeu a été menée ainsi que la fabrication et la caractérisation optique de démonstrateurs. Les résultats de cette étude sont motivants pour des travaux futurs sur les dispositifs ultrafins, mettant en jeu de nouveaux concepts de collection (transport balistique) ou de conversion (cellules solaires à porteurs chauds).Dans un deuxième temps, nous avons étudié la possibilité d’intégrer à court terme un contact arrière nanostructuré en or à des cellules solaires fines (200-400 nm) en CIGS afin d’augmenter potentiellement le courant de court-circuit et la tension de circuit ouvert. Nous avons proposé un procédé innovant pour réaliser cette structure et ce piégeage optique, jusqu’à lors inédits pour les cellules en CIGS. Etude numérique, fabrications de démonstrateurs et premières caractérisations de cellules solaires ultrafines sont présentés. / One of the natural tendencies of photovoltaic technologies is the systematic reduction of the thickness of the solar cells in order to reduce the cost, to save rare or toxic elements or to limit recombination. So far, crystalline thin-film (GaAs) and poly-crystalline (CIGS) technology are reaching optimum conversion efficiency for thicknesses around 1 or 2 microns. Typically, this thickness range does not require new solutions of optical trappings as it is the case for amorphous silicon. However, if we want to reduce these thicknesses by a factor of 10 or even 100 to study new concepts of collections and conversions (GaAs or GaSb) or reduce the use of indium (CIGS), new needs for efficient light absorption are necessary for these technologies. This manuscript is focused on the design, simulation and realization of innovative nanophotonic solutions for future ultra-thin crystalline solar cells.As a first step, we were engaged in an approach at odds with the usual design of solar cells to trap light in a ultra-thin (≤100 nm) layer of material (GaAs, GaSb and CIGS). We propose an array of metal nanostructure placed in front of the cell, transferred on a metal mirror in order to obtain a high, multi-resonant absorption independent of the angle of incidence and polarization. Numerical analysis of the resonant mechanisms involved was conducted as well as the fabrication and optical characterization of demonstrators. The results of this study are motivating for future work on the ultra-thin devices, involving new concepts of collection (ballistic transport) or conversion (hot carrier solar cells).On the other hand, we studied the possibility of integrating a rear gold nanostructured back contact (200-400 nm) in thin CIGS solar cells to potentially increase the current of short circuit and open circuit voltage. We have proposed an innovative process to achieve this structure and the optical trapping for CIGS solar cells. Numerical study, manufacture of demonstrators and first measurements are presented.

On constate une forte demande mondiale d' énergie propre et renouvelable en raison de la consommation rapide des combustibles fossiles non renouvelables et l'effet de serre qui en résulte. Une solution prometteuse pour produire une énergie propre et renouvelable est d'utiliser des cellules solaires pour convertir l' énergie solaire directement en électricité. Comparativement à leurs homologues inorganiques, les cellules solaires organiques (OSCs) sont maintenant intensivement étudiées en raison des avantages tels que le poids léger, la flexibilité, la compatibilité avec les procédés de fabrication à faibles coûts. Malgré ces avantages, l'efficacité de conversion (PCE) des OSCs doit encore être améliorée pour la commercialisation à grande échelle. Les cellules solaires organiques sont réalisées en pile de couches minces comprenant des électrodes, la couche de transport d' électrons, la couche de polymère actif et la couche de transport de trous. Dans cette étude, nous sommes concernés par la couche de PEDOT:PSS qui est couramment utilisée comme une couche tampon entre l'électrode anodique et la couche de polymère actif de cellules solaires organiques. Cette étude vise à intégrer différentes concentrations de nanoprismes (NPSMs) d'argent de taille sub-longueur d'onde dans du PEDOT: PSS afin de profiter de leurs propriétés optiques uniques nées de résonances de plasmons de surface localisées (LSPR) pour améliorer la collecte lumineuse et l'efficacité de génération de charge en optimisant l' absorption et la diffusion de la lumière. Nous avons constaté que les facteurs clés qui contrôlent les performances des cellules solaires plasmoniques comprennent non seulement les propriétés optiques, mais également les propriétés structurelles et électriques des couches hybrides de PEDOT:PSS comprenant des NPSMs d' Ag. D'une part, l'ajout de NPSMs d' Ag conduit ¨¤ (1) une augmentation de l'absorption optique; (2) de la diffusion de la lumière ¨¤ de grands angles ce qui pourrait conduire ¨¤ un meilleur piégeage de la lumière dans les OSCs. D'autre part, (1) la rugosité de surface est augment¨¦e en raison de la formation d'agglomérats de NPSMs d' Ag, ce qui conduit ¨¤ une meilleure efficacité de collecte de charge; (2) la résistance globale des films hybrides est également augment¨¦e en raison de l'excès de PSS introduit par les NPSMs d' Ag incomplètement purifiées, inférieur courant de court-circuit (Jsc) qui en résulte; (3) les Ag NPSMs et leurs agglomérats ¨¤ l'interface PEDOT:PSS/couche photo-active pourraient agir comme des centres de recombinaison, conduisant ¨¤ une réduction de la résistance de shunt, du Jsc et de la tension en circuit ouvert (Voc). Afin de résoudre partiellement l'inconvénient (2) et (3), en intégrant des NPSMs d¡¯Ag davantage purifiés et une petite quantité de glycérol dans le PEDOT:PSS, la résistance des couches hybrides de PEDOT:PSS-Ag-NPSMs peut ¨être réduite à une valeur comparable ou inférieure ¨¤ celles couches vierges. Les futurs progrès en chimie de surface colloïdale et l'optimisation sur le processus d'incorporation des nanoparticules seront nécessaires pour produire des cellules solaires organiques plasmoniques de meilleures performances. / Nowadays there has been a strong global demand for renewable and clean energy due to the rapid consumption of non-renewable fossil fuels and the resulting greenhouse effect. One promising solution to harvest clean and renewable energy is to utilize solar cells to convert the energy of sunlight directly into electricity. Compared to their inorganic counterparts, organic solar cells (OSCs) are now of intensive research interest due to advantages such as light weight, flexibility, the compatibility to low-cost manufacturing processes. Despite these advantages, the power conversion efficiency (PCE) of OSCs still has to be improved for large-scale commercialization. OSCs are made of thin film stacks comprising electrodes, electron transporting layer, active polymer layer and hole transporting layer. In this study, we are concerned with PEDOT:PSS layer which is commonly used as a buffer layer between the anodic electrode and the organic photoactive layer of the OSC thin film stack. We incorporated different concentrations of silver nanoprisms (NPSMs) of sub-wavelength dimension into PEDOT:PSS. The purpose is to take advantage of the unique optical properties of Ag MPSMs arisen from localized surface plasmon resonance (LSPR) to enhance the light harvest and the charge generation efficiency by optimizing absorption and scattering of light in OSCs. We found that the key factors controlling the device performance of plasmonic solar cells include not only the optical properties but also the structural and electrical properties of the resulting hybrid PEDOT:PSS-Ag-NPSM-films. On one hand, the addition of Ag NPSMs led to (1) an increased optical absorption; (2) light scattering at high angles which could possibly lead to more efficient light harvest in OSCs. On the other hand, the following results have been found in the hybrid films: (1) the surface roughness was found to be increased due to the formation of Ag agglomerates, leading to increased charge collection efficiency; (2) the global sheet resistance of the hybrid films also increases due to the excess poly(sodium styrenesulphonate) introduced by incompletely purified Ag NPSMs, resulting in lower short circuit current (Jsc); (3) the Ag nanoprisms and their agglomerates at the PEDOT:PSS/photoactive layer interface could act as recombination centers, leading to reductions in shunt resistance, Jsc and open circuit voltage (Voc). In order to partially counteract the disadvantage (2) and (3), by incorporating further purified Ag NPSMs and/or a small amount of glycerol into PEDOT:PSS, the sheet resistance of hybrid PEDOT:PSS-Ag-NPSM-films was reduced to a resistance value comparable to or lower than that of pristine film.

With increasing global energy demand and diminishing fossil fuel supplies, the development of clean and affordable renewable energy technology is more important than ever. Photovoltaic devices harvest the sun’s energy to produce electricity and produce very little pollution compared to nonrenewable sources. In order to make these devices affordable, however, technological advances are required. In this dissertation a novel photovoltaic device architecture that is designed to enhance sunlight-to-electricity conversion efficiency of photovoltaics is proposed and demonstrated. The increase in efficiency arises due to enhancement of the internal quantum efficiency of photoexcitation in the semiconductor absorber. In other words, the probability that the absorption of a single photon will produce two or more electron-hole pairs, instead of just one, is increased. This occurs through the process of impact ionization, by which a highly excited charge carrier (via absorption of a high energy photon) relaxes by excitation of a second electron-hole pair. The result is an increased photocurrent, and efficiency, of the photovoltaic device. Using thin films of ZnS on Si substrates, we demonstrate that the probability of impact ionization is enhanced at the (unbiased) heterojunction between these layers. The magnitude of enhancement depends on material properties, including crystallinity of the ZnS film as well as concentration of oxygen (impurity) at the interface. Thin films of ZnSe on Si substrates do not exhibit heterojunction-assisted impact ionization, but they do display promising characteristics that make them an intriguing system for future work. The same is true for ZnS/Si materials fabricated by O2-free chemical bath deposition. For the analysis of plain Si as well as ZnS/Si and ZnSe/Si heterostructures, we employ a novel pump-probe transient transmission and reflection spectroscopy technique. A method is demonstrated for using this technique to quantify internal quantum efficiency as well as interface recombination velocity in each of these materials. In bulk silicon, a free carrier absorption cross section that depends on free carrier concentration (above 1018 cm-3) is observed and the relationship is quantified. This dissertation includes unpublished and previously published co-authored material.