Renewable energy sources, such as thin-film silicon solar cells, show a lot of potential as the technology promises low production temperatures, low cost and low material usage. Other benefits are the flexible and very lightweight modules that can be fabricated, which enable a wide variety of applications. Other benefits include the low-temperature coefficient and low toxicity of the materials used to create thin-film silicon solar cells. The downside of using thin-film silicon modules is the relatively low power conversion efficiency inherent to the amorphous and nanocrystalline silicon. Another downside of amorphous silicon solar cells is their significant initial degradation, known as the Staebler-Wronski effect. This initial degradation will stabilize after time but does limit the ultimate conversion efficiency, especially for single-junction devices.
Hydrogenated amorphous silicon (a-Si:H) and nano-crystalline silicon (nc-Si:H) solar cells can be added in series to form a tandem device, thus increasing the efficiency of thin-film silicon solar cells. Although the current will be limited for 2-terminal tandem devices, a significant increase in open circuit voltage can be achieved resulting in higher efficiencies. To achieve highly efficient thin-film silicon solar cells a combination of good optical, electrical and material properties need to be combined. Extensive research has been done to achieve such high-efficiency devices focussed on the silicon deposition conditions, glass texturing for effective light scattering and light incoupling through anti-reflection coatings. Further research is focused on Transparent Conductive Oxides (TCO) improvements and the creation of triple or quadruple junctions.
Firstly, this thesis will focus on a literature study on amorphous silicon with a high-energy bandgap, deposition conditions in PECVD chambers and an i-SiOx buffer layer. A top cell of amorphous silicon with a high-energy bandgap can improve the conversion efficiency of a-Si:H/nc-Si:H solar cells by increasing its overall V oc. The amorphous silicon is processed at high pressure (3-9mbar) and high power densities (150-400mW /cm2) in a VHF-PECVD chamber at 40.68M Hz. To improve the poor blue response of these solar cells, a buffer layer of i-SiOx is explored as such a buffer layer has proven itself under normal deposition power densities (28mW /cm2) and pressure (0.7mbar).The i-SiOx did not prove to be effective in improving the blue response. A narrow range of deposition conditions around 5mbar and 25W is found to give high-bandgap energy solar cells with a good blue response.
Secondly, literature findings on a bilayer TCO configuration of IOH (Hydrogenated Indium Oxide) and i-ZnO (intrinsic Zinc-Oxide) to improve nc-Si:H single junction solar cells are provided. Prior studies on the use of a back reflector made out of different ZnO(Zinc-Oxides) are presented. Findings in literature on micro and nano-textured glass are presented and a comparison is made in an experiment to justify the use of a specific texture. Further experiments are carried out to measure the effect of implementing the bilayer TCO and an i-ZnO back reflector and an experiment on the performance and reproducibility of a-Si on MST (modulated surface textured) glass substrates is performed. ITO textured glass is chosen as the best performing substrate and a thick 1μ i-ZnO layer as part of a bilayer TCO is shown to be most effective at light scattering. The i-ZnO as part of a bilayer TCO can also significantly improve the spectral response of a-Si:H solar cell on MST textured glass. Furthermore, the i-ZnO as a back reflector is shown to be more effective when its thickness is increased.
Thirdly, a study is performed on the functioning of the a-Si:H/nc-Si:H tandem cells with respect to the thickness of the p-layer in the tunnel recombination junction (TRJ). An experiment is carried out to improve the reproducibility of a-Si:H/nc-Si:H solar cells by increasing the hickness of p-layer in the TRJ.
The conclusion is made that the glass texturing is more influential than the thickness of the p-layer in this experiment. Further investigation of the p-layer at the front interface of the device is done by an experiment on both MST and Asahi substrates. It is shown that thicker front players do not improve the reproducibility of tandem solar cells, but a trend of more parasitic absorption is seen. Finally, a study on micro and nano-textured glass for a-Si:H/nc-Si:H solar cells is shown. The chosen ITO substrate is finally chosen to demonstrate the progress of ongoing research on the performance of a-Si:H/nc-Si:H solar cells. The bilayer TCO vs a single layer TCO is used in combination with and without a SiNx ARC.
The electrical performance of the solar cells deposited in the final experiment is likely limited by poor IOH depositions and a large contribution of parasitic absorption is shown for the samples with the SiNx ARC. Either interaction of SiNx with the IOH or a poor IOH deposition are the most likely causes. Good 1-R curves are obtained for the solar cells on ITO textured glass substrates, especially the sample with the bilayer TCO and SiNx ARC.

Silicon carbide (SiC) and silicon oxide (SiOx) have garnered increased attention to reduce the parasitic absorption at the p layer when compared to p-aSi.An outstanding question is how the integration of p-SiC and p-SiOx impacts the metastability of the solr cells. Previous work in the PVMD group at TU Delft has studied the stability of the p-nc-SiOx in detail revealing from experimental results that stable efficiency remains elusive.
This thesis investigates the impact of p-a-SiC on the light-induced degradation (LID) of p-i-n solar cells, aiming to compare its degradation performance with that of p-nc-SiOx. Prior to comparison, p-a-SiC fabrication via plasma-enhanced chemical vapor deposition (PECVD) is extensively explored, as it serves as a widely used light absorber in thin-film silicon solar cells. Various parameters such as power pressure and gas flow rate of the source gas for silicon carbide formation are meticulously studied to achieve optimal P-a-SiC performance. Once the optimal fabrication conditions are identified, light absorption experiments are conducted and characterized using JV and EQE measurements to assess LID performance.

The characteristic feature of ”thin” film solar cells is that they are thin, ranging over thicknesses of a few micrometers. This results in a drop in the light trapping capability of the solar cell, owing to the Beer-Lambert law. Thus, there is a need to optimize light management to enhance the amount of light being trapped in the photoactive region. This can be done by imparting texture in the glass superstrate to scatter the light and thereby allowing the light to travel over a longer distance in the absorber.

In this thesis, the focus is mainly on developing a high-quality reproducable procedure for achieving periodic texturing in glass and understanding the enhancement in the optical impact of the surface. the first objective was to understand the influence of the different processes involved in the fabrication procedure of periodic textured glass superstrates. This was done by varying the parameters associated with the individual processes and observing the resultant impact on the deterministic characteristics of the texturing process. The influence of the mask design, employed in the photolithography process, on the dimensionality of the texture was also analyzed. Based on the aforementioned results, an optimized recipe was developed and successfully implemented to enhance the depth of the periodic hexagonal texturing imparted onto the glass. Additionally, an indicative parameter Re/P was developed to imply the effect the processes had on the quality of the texturing.

This project also ventured into studying the growth of nanocrystalline silicon on the deep textures developed using the optimized recipe. Following this, an extensive optical analysis of the honeycomb textured superstrates was carried out to realize the scattering effect induced by this texturing on interacting with light. This was done by studying the transmittance components and the angular intensity distribution of the periodic textured glass. The interaction mechanism of the textures was found to be dependent on the wavelength and
the size of the textures. In general, there was a linear increase in the diffuse transmittance with texture height, for all wavelengths of light. However, it was noticed that the slope of this relationship varied for the different wavelengths, since it was mainly dominated by different scattering mechanisms for different wavelengths. Finally, the influence of the optical impact of these hexagonal textures on solar cell performance was also studied by studying the Jsc of the micromorph tandem solar cells, developed on deep hexagonal textures of different periodicities.

Low-cost multijunction photovoltaic devices are the next step in the solar energy revolution. Adding a bottom junction with a low bandgap energy material through plasma-enhanced chemical vapour deposition (PECVD) processing could provide a low-cost boost in conversion efficiency. A logical candidate for this low-bandgap material is germanium. A showerhead configuration instead of a direct PECVD reactor is investigated to better control reaction mechanics. In this work, hydrogenated amorphous germanium (a-Ge:H) films processed with a wide range of deposition pressures, powers, temperatures and GeH4 dilution in hydrogen are characterized using elemental analysis, vibrational analysis and analysis of the optoelectrical properties. We have identified a small processing window in which intrinsic a-Ge:H films are processed reproducibly. Two types of degradation were also studied: light-induced and oxidation in air. The lowest E04 achieved for intrinsic films was 0.8 eV with an activation energy of 0.39 eV which lies at the centre of the bandgap. The degradation experiments showed no sign of oxygen signature from EDX measurements performed after two months and FTIR measurements performed after two weeks. Light-induced degradation (LID) was performed by exposing the samples to over 80 hours; photoconductivity did not degrade below the initial value for the duration of the exposure. We believe that the showerhead configuration has increased the surface mobility of Ge-radicals during deposition, allowing them to reach sites in the layer that would otherwise be void.

Thin-film silicon photovoltaic (PV) technology has an enormous potential in the solar power market due to its light weight, flexibility and ease of integration. This enables it to have a wide range of applications like roof top and building integrated installations making it more sustainable. Some of the other key benefits of thin-film silicon PV technology include energy efficient production, less material usage, low temperature coefficient and low manufacturing cost. The relatively low conversion efficiency, however, is the major drawback of thin film silicon solar cells. Hydrogenated amorphous silicon (a-Si:H) grown by plasma-enhanced chemical vapor deposition (PECVD) is one of the extensively employed light absorbers in thin-film silicon solar cells. However, light-induced degradation (LID) of a-Si:H is one of the biggest limiting factors of this technology. LID is a metastable defect creation phenomenon that decreases the conversion efficiency of the a-Si:H solar cell after prolonged exposure to light. This is known as the Staebler-Wronski effect (SWE). SWE has been a subject of research over the past 4 decades. Despite tremendous efforts, the complete suppression of the LID has not been demonstrated yet. Nevertheless, there has been a great progress in characterisation of the defect creation and also the optimisation of solar cells to minimise the effect of LID.
In this thesis, LID of three single junction a-Si:H thin film solar cells with different qualities of a-Si:H absorber layer namely high bandgap a-Si:H, low bandgap a-Si:H and high/low bandgap a-Si:H are studied. Light soaking experiments were carried out for several hundred hours and the performance of the solar cells was analysed through external quantum efficiency (EQE) and illuminated JV characterisations at regular time intervals. Some of the key findings of the experiments include severe LID of high bandgap a-Si:H absorber layer due to increased defect creation, enhanced recombination at the p/i interface in all of the devices and also a significant relationship between the performance and light soaking temperature. Furthermore, the thermal annealing study revealed that, maximum recovery of the performance parameters can be observed at an annealing temperature of 180 ^◦C.
Based on the findings of the above mentioned study, additional experiments were carried out to investigate the stability of the devices with different materials as a buffer layer at the p/i interface. The results of the experiments revealed the significance of the buffer layer in the performance of the devices and also its influence on the LID. In addition, the analysis suggests that intrinsic silicon oxide (i-SiO_X), intrinsic hydrogenated amorphous silicon (i-a-Si:H), and p-doped hydrogenated amorphous silicon (p-aSi:H) are some of the potential candidates for a buffer layer in an a-Si:H solar cell that could give an improved stabilised performance. Furthermore, the speculation of light induced boron diffusion into
the a-Si:H bulk through the buffer layer was tested by another light soaking study to investigate effects of window layer (p) doping on LID. The study found no correlation between the window layer doping and the p/i interface degradation indicating the possibility of a different defect creation mechanism responsible for the enhanced recombination at the p/i interface.

Thin-film silicon solar cells are an innovative approach to utilizing solar energy. Thanks to their flexibility, lower material usage, and production cost, they have the potential to be used in a range of applications. The conversion efficiencies of thin-film silicon solar cells need to be improved to make them commercially viable. This involves further optimizing the different layers of the solar cells. During this project, different strategies for enhancing thin film silicon solar cells were studied, including the deposition conditions for nc-Si:H layer, the sacrificial texturing used, the use of an additional back reflector layer, and the type of TCO layer used. A more consistent quality of the nc-Si:H samples were obtained by varying the hydrogen flow rate compared to the silane flow rate in the processing chamber. Using a silane flow rate of 2.3 sccm and a hydrogen flow rate of 120 sccm, a silane concentration 2.6 was achieved, resulting in nc-Si:H layer with a crystallinity of 60%, close to the desired amorphous-nanocrystalline silicon transition region. The texturing used in the glass sample can play a significant role in scattering the incident light into the solar cell. Craters of different sizes are formed depending on the material used for the sacrificial layer. Making smaller craters on top of larger craters using modulated surface textures (MST) is also possible. Using intrinsic Zinc Oxide (i-ZnO) sacrificial texturing resulted in the highest spectral utilization in single junction nc-Si: H with a Jsc of 25.2 mA/cm2. Indium-doped tin oxide (ITO) sacrificial texturing created micro-sized textures, resulting in the highest spectral utilization in micromorph samples (Jsc of 24.6 mA/cm2). Using an MST of ITO and i-ZnO also resulted in high spectral utilization with a Jsc of 24.3 mA/cm2 in micromorph cells Having an additional back reflector on top of the metal back contact can further improve photon absorption in solar cells. Using a material with a low refractive index, such as i-ZnO, increased the spectral utilization and improved the short-circuit current (Jsc) by approximately 10%. Depositing the i-ZnO back reflector layer using a higher heater temperature (300°C), improved the spectral utilization of the sample further by 9% The transparent conductive oxide (TCO) layer must be highly transparent and conductive. Usually, ITO is used as a TCO layer, although it has limited absorption in the infrared region and lower conductivity at higher temperatures. A TCO with higher optoelectrical properties was obtained using a bilayer of hydrogenated indium oxide (IOH) and i-ZnO. nc-Si:H samples with the TCO bilayer showed a higher spectral utilization and an improvement in their Jsc by 9-12%. The performance of micromorph samples with TCO bilayer was also higher with Jsc improvement of 6%. The influence of the bilayer thickness on the performance of micromorph samples was also inspected, and it was found that using a thicker bilayer with a thickness of 1100 nm instated of 600nm boosted the performance of the sample further and improved its Jsc by 3%.