The worldwide availability of safe and clean drinking water constitutes one of the main challenges of our era. Unfortunately, due to progress-related issues such as global warming, population growth and industrial development, the availability and quality of drinking water is expected to deteriorate in the coming years, posing new threats for humans and the environment. In view of these predictions, it becomes increasingly necessary to develop novel, economical and easily scalable methods for water treatment.

Within the context of this thesis, two novel devices for water treatment based on solar energy and Earth-abundant materials are proposed: a photoelectrochemical (PEC) cell using amorphous silicon carbide (a-SiC:H) photoelectrodes and a PV/electrochemical system based on graphite electrodes. In order to evaluate their performance, both systems are tested in the degradation of recalcitrant organic pollutants.

In the photoelectrochemical cell, a-SiC:H was used first as a photoanode (n-i-p structure), and afterwards as a photocathode material (p-i and p-i-n structures).  In both photoelectrode configurations, the degradation of phenol was monitored under different applied bias voltages. Additionally, photoresponse measurements were recorded in order to determine the behaviour of the photoelectrodes in the solution. Unfortunately, none of the photoelectrode configurations were able to develop noticeable photocurrents, resulting in a poor water treatment performance.

With regard to the PV/electrochemical system, the graphite electrodes were first tested under an external power supply. Here, the electrochemical oxidation of phenol was studied in detail under different applied voltages. Additionally, the chemical oxygen demand (COD) was measured in order to identify intermediate compounds in the reaction. Overall, it was found that the optimum voltage for the degradation reaction was 1.6V (80% phenol removal, 59% COD reduction). In addition to this study, the effect of an alternative copper counter-electrode was investigated, and the degradation of an organic dye (methylene blue) was examined. Finally, several PV structures were proposed for the integration with the graphite electrodes. The best overall performance was achieved by two micromorph solar cells connected in series. However, in terms of utilized area, an individual a-Si:H/a-Si:H tandem cell was the most appropriate choice

In the transition towards a renewable based energy supply, one of the main challenges is the intermittent nature of most renewable energy sources. In order to ensure a stable energy supply, different storage technologies attract a lot of research attention. For long term storage, using hydrogen as artificial fuel is a promising option. However, in order for hydrogen storage to be renewable, the hydrogen must be produced using renewable energy. A possible option is to use solar energy to drive the electrolysis of water. In this way, hydrogen is produced directly from sunlight by driving the redox reaction with a reversible potential of 1.23 V. However, due to overpotentials and possible losses a voltage of 1.6-2 V is needed form the used device to drive the reaction. Recently, a wafer based triple junction silicon solar cell was proposed to generate the voltage needed to drive the water splitting reaction. The cell uses an n-type crystalline silicon wafer as substrate, and has an a-Si/nc-Si/c-Si structure. Although this cell demonstrates the feasibility of this technology, the performance is not good enough yet to drive the water splitting reaction efficiently.
In this work, several ways to optimize the triple junction solar cell have been studied. First, optimising the tunnel recombination junctions improves the open circuit voltage of the device. For the tunnel recombination junction between the top a-Si and the middle nc-Si cell, this is done by using a combination of a thin n-SiOx layer and a thin n-a-Si layer. Apart from a small improvement in optical performance this configuration improves the conductivity of the junction and also slightly enhances charge separation and collection in both sub-cells. For the junction between the middle nc-Si and the bottom c-Si cell, the same result was achieved by increasing the doping in the n-layer. Subsequently, to improve current matching of the device, the nc-Si layer thickness can be increased to 3500 nm without sacrificing the electrical performance. In addition, optimizing the interface between the top p-nc-SiOx layer and the front transparent conductive oxide (TCO) also improves the electrical performance of the device. Moreover, it was found that using hydrogenated indium oxide as front TCO instead of indium doped tin oxide increases the current by decreasing the reflection in the wavelength region of the current limiting nc-Si middle cell by having reflection, and also increases the open circuit voltage of the device by 0.3 Volts. The final triple junction cell achieved an open circuit voltage of 1.93 V and a short circuit current density of 8.5 mA/cm2, with a fill factor of 0.65. With these improvements, it is estimated that a solar to hydrogen efficiency of 6.2 % can be reached by using an IrOx counter electrode.

The ever increasing installed capacity of renewable, sustainable energy is essential in order to keep the earth
habitable. However, the intermittent nature of solar and wind energy do not strictly follow human energy
demands. Therefore an energy of sufficient magnitude buffer is essential to provide a constant supply of
energy that matches the demand as closely as possible. Chemical energy shows high potential in terms of
clean, long-term energy storage with efficient conversion to electricity.
The DEMO project intends on manufacturing a monolithic photovoltaic device for the production of hydrogen
gas and hydrocarbon products by the electrochemical splitting of water and carbon dioxide. an aSi:H/a-SiGe:H/nc-Si:H
triple junction device is proposed to generate a combination of voltage and current
to drive the electrochemical reactions efficiently, while simultaneously addressing scalability by using earthabundant
materials.
In this thesis, efforts were made to optimise the a-SiGe:H subcell as well as finding the optimal combination
of materials to use as tunnel recombination junctions as intermediate layers between subcells of the
multi-junction device.
The best performing a-SiGe:H p-i-n single-junction device achieved a VOC = 719 mV, JSC =17.2 mA/cm2
,
F F = 0.63 and η = 7.85% as processed on an Asahi VU superstrate. This performance was observed show a
strong decline with increasing intrinsic layer thickness and increasing deposition rate. Although a conclusive
optimal device structure was not obtained, it is believed that bandgap profiling by adjusting the germanium
content in the intrinsic layer as well as applying buffer layers can substantially improve the performance of
the a-SiGe:H single-junction solar cell.
To assess long-term stability, an a-SiGe:H single-junction device was subdued to 1000h light soaking.
Meta-stable defects induced by recombination of photo-generated charge carriers resulted is a strong degradation
in F F of 20% relative. VOC and JSC showed similar relative decreases in performance with 9.0% and
8.6%, respectively. The resulting relative drop in conversion efficiency for the degraded solar cell is observed
to be 34%.
The best tunnel recombination junction for an a-Si:H/a-SiGe:H double junction consists of a 5 nm highly
doped n+-type nc-SiOx :H layer with a 2 nm nc-Si:H p-layer, placed between 25 nm nc-SiOx :H n-layer and
a 16 nm nc-SiOx :H p-layer that resulted in good combination of F F, VOC and JSC by improved tunnelling,
charge separation and better light management.
The best tunnel recombination junction for an a-SiGe:H/nc-Si:H double-junction consists of a highly
doped n+ nc-SiOx :H layer with a 2 nm nc-Si:H p-layer. These layers are sandwiched between an 50 nm ncSiOx
:H n-layer, of which the first half has linearly increasing oxygen content, and a 16 nm nc-SiOx :H p-layer
of the a-SiGe:H subcell and nc-Si:H subcell, respectively.
These tunnel recombination junctions were used to further develop a p-i-n a-Si:H/a-SiGe:H/nc-Si:H triplejunction
solar cell on a wet-etched glass superstrate with ZnO:Al as both sacrificial layer for texturing and
as transparent conductive oxide. The best performing triple-junction device achieved a VOC = 1.96 V, JSC =
6.21 mA/cm2
, F F = 0.63 and η = 7.63% with respective intrinsic layer thicknesses of 175 nm/120 nm/3000 nm.
The conversion efficiency of this current mismatched device is throttled by the current limiting a-SiGe:H middle
cell. Enhancing the a-SiGe:H material quality allows for thicker absorber layer to increase current generation
without compromising the electrical performance, which can significantly improve the performance the
triple-junction device.
This device is able to achieve an estimated solar-to-fuels efficiency of 4.9% for producing hydrogen with
a high performance electrode and an an IrOx counter electrode. Assuming similar electrode performance,
the solar-to-fuels efficiency for producing both hydrogen and methane is expected to be 2.4%. Hydrocarbons
with higher electrochemical potential are not likely to be produced with such a device due to the higher
electrochemical potentials.
Using the measured degraded performance of the a-SiGe:H subcell, a relative decrease in water splitting
efficiency of 21% is expected for the triple-junction device.