Search

The interchangeability of ion bombardment and deposition temperature during a-Si:H film growth at high deposition rates (10-42 Å/s) by means of the expanding thermal plasma has been studied. The ion bombardment is generated by applying an external rf bias voltage on the substrate. It is shown that the opto-electronic performance of the a-Si:H films improves considerably when a moderate rf substrate bias voltage (∼20-60 V) is applied, i.e. the photo response increases two orders of magnitude up to 106. Furthermore, it is also revealed that the additional energy supplied to the growth surface by the ion bombardment, makes a reduction of the deposition temperature by ∼100°C possible, while preserving good material properties. On the basis of the results obtained, three effects caused by the rf substrate bias can be distinguished: creation of an additional growth flux, a reduction of the void incorporation, and an increase in the vacancy density.

A set of 8 rf deposited a-Si:H thin films of various thickness (4-1031nm) have been used to explore the applicability of two optical techniques, thin film cavity ringdown spectroscopy (tf-CRDS) and second harmonic generation (SHG), for the measurement of small defect-related absorptions. In this paper we will give a first overview of the different aspects of these techniques, which are novel in the field of amorphous silicon materials. It is shown that tf-CRDS is capable of measuring defect-related absorptions (associated with dangling bonds) as small as 10-7 for a single measurement, without the need for elaborate calibration procedures. The results are compared with photothermal deflection spectroscopy (PDS) for a broad spectral range (0.7 -1.7 eV) and show good agreement. Furthermore the existence of a defect-rich surface layer with a defect density of 1.1 × 1012 cm-2 has been proven. The absorption spectrum of a 4 nm thin film has revealed a different spectral signature than a bulk dominated (1031 nm) film. The SHG experiments on a-Si:H films have shown that the second harmonic signal arises from the surface states and polarization dependent studies have revealed that the surface states probed have an ∞m-symmetry. From this it can be deduced that the absorbing surface states are isotropically distributed. A spectral scan suggests that the second harmonic signal, whose origin has not been unrevealed yet, has a resonance at an incident photon energy of 1.22 eV.

From investigations on the SiH3 and SiH radical density and the surface reaction probability in a remote Ar-H2-SiH4 plasma, it is unambiguously demonstrated that the a-Si:H film quality improves significantly with increasing contribution of SiH3 and decreasing contribution of very reactive (poly)silane radicals. Device quality a-Si:H is obtained at deposition rates up to 100 Å/s for conditions where film growth is governed by SiH3 (contribution ∼90%) and where SiH has only a minor contribution (∼2%). Furthermore, for SiH3 dominated film growth the effect of the deposition rate on the a-Si:H film properties with respect to the substrate temperature is discussed.

Silicon solar cells are the technology with the highest market share in the photovoltaic industry. In particular, Silicon Heterojunction (SHJ) solar cells have been strongly developed in the last years due to the high power conversion efficiency that can be achieved with them. In order to improve the production of SHJ solar cells, many factors such as efficiency, materials cost, ecological impact of the materials extraction should be researched. This thesis contains experimental work related to the development of indium-free materials for the contact layers of SHJ solar cells. Transparent Conductive Oxide (TCO) layers of Aluminium doped Zinc Oxide (ZnO:Al) were investigated as a potential replacement of Indium Tin Oxide (ITO) for the front contact in SHJ solar cells in the rear emitter configuration. It was found that ZnO:Al can be tuned to yield cell performance almost at the same level as ITO with a power conversion efficiency of 22.6% and 22.8%, for ZnO:Al and ITO, respectively. The main reason for the slight underperformance of ZnO:Al compared to ITO was found to be a higher contact resistivity between this material and the silver grid on the front side. An entirely indium-free SHJ solar cell, replacing the ITO on the rear side by ZnO:Al as well, reached a power conversion efficiency of 22.5%. Nickel oxide (NiOx) was also investigated as an ITO replacement material. NiOx layers deposited by e-beam evaporation in a reactive process with oxygen were investigated as a possible charge selective contact for holes in SHJ solar cells. As a first step, the viability of depositing NiOx using this process was confirmed. A variation of the oxygen flux during the deposition was performed in order to obtain a variation of the material properties, ranging from metallic nickel layers to transparent nickel oxide layers, when the oxygen deposition pressure varies from 2 × 10−7 mbar to 1 × 10−4 mbar. These samples were characterized mainly by X-ray Photoelectron Spectroscopy (XPS) and Energy Dispersive X-ray difraction (EDX) measurements. For future reference and other possible applications of the NiOx, a material study was performed, in which the optical and electronical properties of the layers were characterized. The characterization techniques include ultraviolet and visible (UV-VIS) spectroscopy, X-ray Diffraction (XRD) and Hall mobility measurements. It was found, that the deposition of NiOx on silicon wafers passivated with a-Si:H leads to a strong decrease in the minority charge carrier lifetime in the samples, making them unsuitable for fabricating photovoltaic devices. The NiOx e-beam deposition on silicon was analyzed in detail in order to find the reason for the decrease in charge carrier lifetime. Although it was not possible to find the root cause for the degradation with certainty, a number of factors could be ruled out as possible causes for the lifetime decrease. A detrimental effect of X-ray radiation (present in every e-beam process due to the deflection of the electrons) on the wafer, as well as an oxidation of the passivation layers due to the oxygen atmosphere during the deposition, were ruled out using photoluminescence measurements. Diffusion of nickel atoms through the passivation layers and up to the interface between a-Si and c-Si which would lead to increased recombination at these metal atoms was considered as a possible reason, and also discarded after designing an experiment to measure the diffusion coefficient of nickel atoms in silicon, which showed that no significant diffusion is taking place. In the final part of this work, the XPS spectra of nickel oxide layers with different thicknesses were compared. One of them is thin enough to allow the measurement of the a-Si/NiOx interface, in which an XPS signal related to nickel silicide was detected. This nickel silicide interlayer is identified as the probable reason for the detrimental effect of the e-beam deposition of nickel oxide on passivated silicon wafers, due to the fact that it acts as a recombination site for the charge carriers, and also that the formation of NiSi consumes a-Si which leads to a thinner passivation layer. Similar results were found in the literature for nickel depositions on a-Si, however not for depositions under an oxygen atmosphere., Silizium-basierte Solarzellen haben den größten Marktanteil im Bereich der Photovoltaik. Insbesondere Silizium Heteroübergang-Solarzellen (SHJ, engl. silicon heterojunction) wurden aufgrund ihrer potentiell hohen Wirkungsgrade stark weiterentwickelt. Um die Produktion von SHJ Solarzellen zu verbessern, müssen zahlreiche Faktoren wie Effizienz, Materialkosten oder ökologische Auswirkungen der genutzten Materialien untersucht werden. Diese Doktorarbeit befasst sich mit der Entwicklung von Indium-freien Kontaktschichten für SHJ Solarzellen. Schichten des transparenten leitfähigen Oxids (TCO, engl. transparent conductive oxide) Aluminium dotiertes Zinkoxid (ZnO:Al) wurden als potentieller Ersatz für Indium-Zinnoxid (ITO) als Vorderseitenkontakt in SHJ Solarzellen in Rückseiten-Emitter- Konfiguration untersucht. Mit der Nutzung von ZnO:Al am Vorderseitenkontakt konnte eine Effizienz der SHJ Solarzelle von 22,6% erreicht werden, welche nur geringfügig unter der, der ITO Referenzzelle von 22,8% lag. Als Hauptgrund für diese etwas niedrigere Effizienz wurde ein höherer Kontaktwiderstand zwischen TCO und dem Silbergitter an der Vorderseite gefunden. Eine Indium-freie Solarzelle, bei der ITO auch auf der Rückseite durch ZnO:Al ersetzt wurde, erreichte eine Effizienz von 22.5%. In einer zweiten Materialstudie wurden per Elekronenstrahl-Verdampfung aufgebrachte Nickeloxid (NiOx) Schichten als mögliche TCO und selektive Kontakte für die photogenerierte Löcher in SHJ Solarzellen untersucht. Nachdem die Funktionsfähigkeit des Prozesses untersucht und bestätigt wurde, wurde der Sauerstoffdruck während der Deposition zwischen 2×10−7 mbar und 1 × 10−4 mbar variiert, um die Materialeigenschaften in einer Bandbreite von metallischem Nickel zu transparentem Nickeloxid zu variieren. Die resultierenden Probe wurden anschließend hauptsächlich mit Röntgenphotoelektronenspektroskopie (XPS, engl. X-ray photolectron spectroscopy) und Energiedispersiver Röntgenspektroskopie (EDX, engl. Energy dispersive X-ray spectroscopy) untersucht. Als Referenz für zukünftige Experimente wurde eine Materialstudie an den NiOx Schichten durchgeführt. Hierzu wurden UV-VIS Spektroskopie, Röntgendiffraktometrie (XRD, engl. X-ray diffraction) und Hall Mobilitätsmessungen angewandt.Es stellte sich heraus, dass die Deposition von NiOx auf passivierte Siliziumwafer zu einer starken Reduktion der Minoritätsladungsträgerlebensdauer in der resultierenden Probe führt, weshalb diese nicht für die Anwendung in einer Solarzelle geeignet ist. Um den Grund für die reduzierte Minoritätladungsträgerlebensdauer zu finden, wurde die NiOx Elektronenstrahl Deposition näher untersucht. Hierbei konnten sowohl ein schädlicher Effekt der während des Prozesses entstehenden Röntgenstrahlung, als auch eine Oxidation der Passivierungsschichten durch die Sauerstoffatmosphäre während der Aufbringung anhand von Photolumineszenz Messungen als Ursachen für die Degradation ausgeschlossen werden. Weiterhin wurde die Diffusion von Ni-Atomen zur Grenzfläche zwischen amorphem und kristallinem Silizium als Ursache ausgeschlossen. Im abschließenden Teil dieser Arbeit wurden die XPS Spektren von NiO Schichten verschiedener Dicken analysiert. Diese Spektren deuten auf die Präsenz eines Nickelsilizides hin, welches als Rekombinationszentrum für Ladungsträger dienen kann und daher die wahrscheinliche Ursache für die negativen Auswirkungen der Elektronenstrahl Deposition von NiOx auf passivierte Siliziumwafer ist. Einerseits wird das Si von der a-Si Schicht konsumiert um das NiSi zu bilden, wodurch die passivierungseigenschaften vom a-Si verloren gehen und andererseits ist das NiSi ein rekombinationsaktives Material. Ähnliche Ergebnisse wurden in der Literatur gefunden für Nickel Abscheidungen auf a-Si, jedoch nicht in einem reaktiven Prozess unter einer Sauerstoff Atmosphäre.