Perovskite solar cells (PSCs) are an emerging and promising photovoltaic technology that have demonstrated impressive power conversion efficiencies exceeding 25% after only fourteen years of development. This rapid progress can mostly be attributed to the development and optimisation of PSCs fabricated by solution-based methods, because of their easy processing and inexpensive equipment. However, to commercialise PSCs, alternative methods are required. Among these methods, the vacuum-based technique thermal evaporation is a suitable candidate. In thermal evaporation, solid precursors are evaporated and deposited onto a substrate in a high vacuum chamber. Thermal evaporation is a mature technique in the semiconductor industry to fabricate pinhole-free and highly uniform thin films at a large scale on different types of substrates. To date, the efficiency of PSCs by thermal evaporation is lagging behind their solution-based counterpart, partly due to limited research on vacuum deposited PVK films and several challenges unique to thermal evaporation.
In this MSc thesis project, organic-inorganic metal halide perovskite absorbers based on thermally evaporated FAxCs1-xPb(IyBr1-y)3 and spin-coated MAPbI3 have been developed together with the additional supporting layers for device fabrication. Several PSCs with different p-i-n architectures have been fabricated and characterised. The device comprises of: ITO as front electrode, spiro-TTB, spiro-OMeTAD and/or MoOx as hole transport layer, PVK absorber layer, C60 as electron transport layer, BCP as buffer layer, and silver and aluminium as metallic back electrode. The goal of this work is to develop and optimise the structural and opto-electronic properties of different device layers and demonstrate a working perovskite solar cell.
PVK engineering is carried out on both thermally evaporated and spin-coated PVK thin films with the aim at obtaining perovskite that possesses desirable opto-electronic properties. For FAxCs1-xPb(IyBr1-y )3 fabricated by sequential layer thermal evaporation, the uniformity and tooling factors were optimised to grow layers with a correct precursor ratio showing a high crystallinity and absorptance. No clear effect of post annealing could be detected for the temperatures and times tested in this work. Achieving the target thickness and obtaining a satisfactory reproducibility were not fully reached. For spin-coated MAPbI3, it was found that increasing the precursor solution concentration helped to achieve a desirable thickness for PSCs. Moreover, an adequate bandgap, diffractogram and absorptance were measured. However, based on the small crystal size determined with SEM, and poor device performance, further improvement is needed.
Transport, contact, and buffer layers were successfully fabricated by developing new deposition recipes for silver, BCP, and spiro-TTB. Furthermore, spectrophotometry measurements show that thermally evaporated HTLs (MoOx and spiro-TTB) exhibit much lower parasitical absorption losses compared to spin-coated spiro-OMeTAD. An optical model to fit spectroscopic ellipsometry data measured on spiro-TTB thin films is created to determine its thickness and optical properties. Moreover, time resolved microwave conductivity results indicate that C60 effectively extracts electrons from MAPbI3. In contrast, spiro-TTB films did not appear to possess any hole extracting abilities. Strong quenching of the steady-state PL signal is observed for bi-layers of MAPbI3 with either C60, spiro-OMeTAD and MoOx . This quenching is possibly caused by extraction of either electrons or holes and/or enhanced non-radiative recombination at the interface.
The optimised films were combined in complete PSCs with most layers fabricated by thermal evaporation through metal masks structuring cells with active areas of 0.16 and 0.36 cm2. Solar cells were characterised in the dark and under illumination using a solar simulator integrated with a glove box to prevent degradation. A wide range of J-V characteristics are identified and classified: (1) ohmic responses, possibly caused by poor quality PVK absorber layers with pinholes filled with thermally evaporated metal, (2) S-shaped curves under illumination, possibly caused by a mismatch in energy band alignment or increased non-radiative recombination at the MoOx/MAPbI3 interface, (3) diode behaviour in the dark, and (4) a high series resistance and low shunt resistance, possibly caused by low mobility, non-optimal thicknesses, internal currents, and a mismatches in band alignment. The most promising PSC featuring the ITO (200 nm)/Spiro-OMeTAD/MoOx (5 nm)/MAPbI3/C60 (20 nm)/BCP (4 nm)/Ag (150 nm) architecture, had a short circuit current of 10.0 mA/cm2 and an open circuit voltage of 0.67 V.
The experiments carried out in this MSc thesis project supported the development of several steps of PSC fabrication process. This work contributes towards understanding thermal evaporation processes of perovskite films and fabricating fully evaporated devices with high efficiency and a large area.