Using a two-step deposition technique to prepare perovskite (CH3NH3PbI3) for thin film solar cells based on ZrO2 and TiO2 mesostructures

Undoubtedly, perovskite solar cells have become a key player in 3rd generation photovoltaics over the last few years. Although it is only in 2012 that the first solid state perovskite solar cell was reported, power conversion efficiencies have increased so rapidly that PSCs are now serious contenders to the well-established and marketed thin-film and wafer technologies. Over the timespan on this work, over 3€™000 published articles in peer-review journals have been published. The work that I report in this thesis is merely a small contribution to the gigantic amount of data, models and theories that have been made over the last 4 years. I hope that my modest contribution will be of value to the scientific community. Over the course of this work, I attempted not to focus my attention on a single issue that relates to perovskite photovoltaics, but to approach it from different perspectives. As a result, a wide array of subjects have been treated. Starting with the fabrication of perovskite solar cells, an innovative way to deposit thin film of perovskite from solution is reported, using a two-step deposition technique. Using this approach, the power conversion efficiency of perovskite solar cells has been increased from 12.3% to over 17% in less than a year. Subsequently, I looked into the possible compositional variations of the perovskite layer. This study showed that the cation methylammonium and formamidinium can be used inter- changeably using the deposition technique that we have reported on earlier. We showed that a mixture of the two cations leads to a stabilization of the tetragonal perovskite phase and shrinks the optical band gap of the photoabsorber, allowing more photons to be absorbed and converted into electrons. Following this work, I studied some of the reactivity properties of the phase pure organic inorganic CH3NH3PbX3(X=Cl,Br,I). Notably, it could be shown that the halide component can be easily exchanged by another one from a simple solution based anion exchange reaction. Later, the power conversion characteristics of perovskite solar cells were investigated. By developing a electronic circuit allowing versatile power point tracking of the solar cells, a new algorithm that deals with the issues specifically related to perovskite solar cells is reported. i A major topic of my work was the study of the frequency resolved dynamics that results for the generation carriers by light. Using a novel way to measure the intensity modulated spectral response of perovskite solar cells, I was able to identify some of the key limiting factors that cap the power conversion efficiency of the devices. A numerical model that answers to the laws of carrier dynamics was assembled and used as a tool to better understand perovskite solar cells. Finally, I reported on a way to deal with the end of life of perovskite solar cells. Some of the relevant regulatory framework in the European Union that deals with the recycling of solar panels are highlighted. Using the concept of upcycling, a way to recycle the valuable components of the solar cell in an efficient way was outlined.

Photovoltaics literature survey (no. 149)

Hybrid inorganic-organic thin-film solar cells, such as dye-sensitized solar cells (DSCs), can be assembled with low-cost materials and manufactured with cost-effective methods, and are considered promising photovoltaic technologies. A typical DSC is composed of a layer of mesostructured wide-band gap metal oxide semiconductor on which is adsorbed a light-absorbing sensitizer. Upon illumination, the adsorbed sensitizer is photoexcited and injects an excited electron into the conduction band of the semiconducting oxide. The oxidized dye is then regenerated by electron donation from a redox couple present in the electrolyte. The redox couple is regenerated at the counter electrode with electrons that have migrated through an external circuit. Despite over 20 years of development, the performance and long-term stability of these solar cells are still lagging behind major photovoltaic technologies, such as silicon-based solar cells. This is partly because some of the factors limiting device performance are not fully understood. Moreover, novel photoactive materials that have both high light absorption capability and fast carrier mobility are highly desired. The aim of the research presented in this thesis was to contribute to the development of high-performance hybrid thin-film solar cells by designing novel device architectures and exploiting hybrid perovskite materials as novel light absorbers along with providing a deeper understanding of device operation mechanisms. In order to enhance the charge transport in DSCs, a nanostructured collector–shell electrode was developed. A collector–shell electrode consists of a porous scaffold material having high electronic conductivity and a thin metal oxide shell. The shell provides sites for dye adsorption and photoelectron injection while the conductive backbone allows fast electron transport. DSCs fabricated using the collector–shell electrodes along with a cobalt redox couple and an organic sensitizer, MK-2, exhibited a promising power conversion efficiency (PCE) of 3.3% and a charge transport rate that was 2.6 times faster than observed for devices utilizing P25-based TiO2 electrodes. In the quest to explore novel efficient light absorber materials, an organic-inorganic hybrid perovskite material, viz., CH₃NH₃PbI₃, was utilized in planar structured thin film solar cells. The CH₃NH₃PbI₃ perovskites have been shown to exhibit excellent light harvesting, high carrier mobility and facile solution processability. A one-step, solvent-induced fast crystallization method was developed which produced high quality CH₃NH₃bI₃ perovskite thin films. These thin films exhibit full surface coverage and are composed of micron-sized grains. The application of these films in solar cell construction led to highly efficient devices with an average PCE of 13.9±0.7% and a steady state efficiency of 13%. The champion device fabricated using this deposition method achieved a PCE of 16.2%. In an attempt to further understand the perovskite-based device working mechanism, inverted structure perovskite-based solar cells with single or double selective contacts were fabricated which employed CH₃NH₃PbI₃ as light absorber and different inorganic metal oxides as interlayers. Solar cells fabricated on metal oxide layer coated substrates exhibit promising PCEs of over 10%. In addition, solar cells fabricated on ITO substrate without any metal oxide layer also exhibit a high PCE of 11.5%. This result indicates that perovskite solar cells can also perform well with single selective contact. The inverted structure perovskite solar cells, with or without the metal oxide interlayers, all exhibit very weak hysteresis in J–V measurements. This finding suggests that the hysteresis effect is not an intrinsic characteristic of the CH₃NH₃PbI₃ perovskite material and can be alleviated by optimization of the cell structure and judicious selection of the contact material. To understand the perovskite crystallization process and hence obtain a better control of film morphology for device fabrication, a series of perovskite materials including FAPbI₃, MAPbBr₃, MASnI₃ and mixed perovskites were investigated. Tuning of the relative rates of nucleation and crystal growth is crucial to achieve control over the final film morphology. For the FAPbI₃ system, smooth and uniform perovskite films were obtained over a large area by simultaneously applying a gas-assisted deposition method and adding HI solution in the perovskite precursor solution. Optimization of fabrication process resulted in a solar cell with a best performance of 12.0%. For MAPbBr₃ system, planar structure photovoltaic devices using these MAPbBr3 films achieved a PCE of 0.5%. The device performance is further increased to 2.2% by deposition of a mesoporous TiO₂ layer on top of the dense TiO2 blocking layer. For MASnI₃ system, application of a precursor solution combining MASnI₃ with MAPbI3 in a molar ratio of 1:1, and using the gas-assisted deposition method, smooth MASn₀.₅Pb₀.₅I₃ films over large area were obtained. In summary, high performance hybrid thin film solar cells have been developed by engineering the device architecture and employing an alkylammonium metal halide perovskite material as light absorber. The studies presented herein highlight the potential of hybrid mesoscopic thin film solar cells to become a promising photovoltaic technology.

In TiO2-based perovskite solar cells (PSC), the preparation of the CH3NH3PbI3/TiO2 composite layer is very important, since the morphology of the perovskite adsorbed onto the surface of the TiO2 nanoparticles has decisive significance for the absorption of the incident sunlight and separation of the generated carrier. The traditional two-step spin-coating method for the deposition of CH3NH3PbI3 into the porous mesoporous TiO2 layer usually suffers from filling block problems. In this study, the PbO-TiO2 mesoporous layer was prepared with various ratios of Pb:Ti in the raw materials. Morphological, optical, and element analysis of the prepared thin films indicated that Pb was gradually mixed into the TiO2 mesoporous layer with the increased Pb:Ti ratios. The element distribution characteristics of the optimal thin films showed that the distribution of Pb was uniform throughout the whole TiO2 thin film, which indicates the successful mixing of Pb into the TiO2 electrode layer. Combined with dip coating, the PbO-TiO2 mesoporous layer was prepared into a CH3NH3PbI3/TiO2 composite layer and subsequently to a solar cell device. The prepared solar cell shows a short-circuit photocurrent density of 16.4 mA/cm2, an open-circuit voltage of 900 mV, a fill factor of 61%, and a power conversion efficiency (PCE) of 9.00%. The PCE of the PSC is promoted by nearly 25% when compared with that prepared with the traditional method. The proposed preparation method that combines TiO2 nanoparticle electrode with a mixing and dip coating provides a new effective way to improve the deposition of perovskite into the mesoporous TiO2 layer, which is very helpful for the fabrication of high-efficiency and low-cost PSC.

Perovskite solar cells attract great attention due to its rapidly increasing efficiency. Conventional structure of perovskite solar cell contains FTO glass substrate, compact TiO2 layer, mesoporous TiO2/CH3NH3PbI3 layer, hole transport- ing material layer and Au counter electrode. In this work, we fabricated perovskite solar cells with the above conventional structure. The mesoporous TiO2 layer thickness are 500, 350, 150 and 100 nm. Thickness of CH3NH3PbI3 capping layer is about 300 nm. The perovskite films and solar cells were characterized by SEM, XRD, UV-Vis absorption spectrum, photo- current-photovoltage characteristics and electrochemical impedance spectra. XRD patterns of the perovskite films are similar, indicating the complete transfer from PbI2 to CH3NH3PbI3. Statistical results of short-circuit current, open-circuit voltage, fill factor and power conversion efficiency are compared, revealing that as mesoporous TiO2 layer thickness increasing, both photovoltage and fill factor decrease whereas short-circuit current slightly increases. Solar cells with thinner mesoporous TiO2 layer can give higher efficiency. Besides, the devices with 100 and 150 nm mesoporous TiO2 layers can present the average efficiency of 15%. The decrement of efficiency is supposed to be caused by stronger carrier recombination. Electro- chemical impedance spectra and current-voltage characteristics under dark condition were applied to characterize the carrier recombination process. Nyquist plots demonstrated an increment of the recombination as the mesoporous TiO2 layer thick- ness increases. Charge transfer resistances were obtained by fitting Nyquist plots. The charge transfer resistances of solar cells with 100 and 350 nm mesoporous TiO2 layer decrease with bias voltage exponentially in similar slope, indicating that this change of recombination do not influence the diode quality factor. Reverse saturated current density was obtained by fitting dark current-voltage curves. The reverse saturated current densities have positive correction with mesoporous TiO2 layer thickness. As a conclusion, the change of the recombination is caused by reverse saturated current density rather than diode quality factor. Further investigation revealed that the devices with thinner mesoporous TiO2 layers exhibit relatively stronger hysteresis behavior. 15.56% of certified efficiency has been obtained for the perovskite solar cell with 150 nm-thickness mesoporous TiO2 layer. Keywords perovskite; solar cell; carrier recombination; TiO2; film thickness

This work presents a utilization of spray deposition technique in the fabrication of thin film perovskite solar cell devices with an assistance of a homemade CNC machine. An 80 nm-thick compact TiO2 was sprayed at a substrate temperature of 400 °C for electron blocking layer. Both one and two-step spray deposition technique are used to compare the quality of the perovskite absorber layer. A large grain size and voids free CH3NH3PbI3 film achieved by using the two-step spray deposition process. When such films are incorporated into a solar cell device with conductive carbon counter electrode, the maximum power conversion efficiencies of up to 9.58% are realized. This approach could pave the way to develop low-cost perovskite solar cell.

Mesoporous TiO2 (m-TiO2) layer has been widely used as a photoelectrode of solar cells. Compared with conventional planar TiO2 layer, appropriate pore size dramatically improves infiltration of perovskite (PVK) into the mesoporous layer because of the larger voids formed within the TiO2 mesoporous layer and further enhances the light absorption efficiency of PVK for the better light-scattering ability of 20 nm TiO2 nanoparticles. In our study, 75–95 nm pore size of TiO2 scaffold layer was successfully prepared by adding different contents of ethyl cellulose (EC) as a template and terpineol as a leveling agent into conventional TiO2 paste composed of 20 nm-sized TiO2 nanoparticles and then applied to PVK solar cells (PSCs).The novel mesoporous layer was conducive to increase light-harvesting capacity and photovoltaic performance of PSCs. The influence of different contents of EC was discussed systematically for TiO2 scaffold layer, the charge carrier dynamic, and the hysteresis behavior for our devices. Finally, by optimizing the parameters of the electron transport layer, the power conversion efficiency of the champion device with a 16.5% mass fraction of EC reached as high as 18%.

Improved crystallization of perovskite films using PbTiO3-decorated mesoporous scaffold layers for high stable carbon-counter-electrode solar cells

Rapid and low temperature processing of mesoporous TiO2 for perovskite solar cells on flexible and rigid substrates

Sputtered transparent electrodes for optoelectronic devices: Induced damage and mitigation strategies

AlN thin films were deposited on flexible Hastelloy tapes and Si (100) substrate by middle-frequency magnetron sputtering. A layer of Y2O3 films was used as a buffer layer for the Hastelloy tapes. A two-step deposition technique was used to prepare the AlN films. The effects of deposition parameters such as sputtering power, N2/Ar flow rate and sputtering pressure on the microstructure of the AlN thin films were systematically investigated. The results show that the dependency of the full width at half maximum (FWHM) of AlN/Y2O3/Hastelloy on the sputtering parameters is similar to that of AlN/Si (100). The FWHM of the AlN (002) peak of the prepared AlN films decreases with increasing sputtering power. The FWHM decreases with the increase of the N2/Ar flow rate or sputtering pressure, and increases with the further increase of the N2/Ar flow rate or sputtering pressure. The FWHM of the AlN/Y2O3/Hastelloy prepared under optimized parameters is only 3.7° and its root mean square (RMS) roughness is 5.46 nm. Based on the experimental results, the growth mechanism of AlN thin films prepared by the two-step deposition process was explored. This work would assist us in understanding the AlN film’s growth mechanism of the two-step deposition process, preparing highly c-axis–oriented AlN films on flexible metal tapes and developing flexible surface acoustic wave (SAW) sensors from an application perspective.

Blanket chemical vapor deposited tungsten (CVD-W) offers the potential to fabricate reliable contacts for submicron multilevel metallization. In via filling application, various aluminum fluorides were formed by the reduction of WF6 with aluminum underlayer. These compounds will stay at CVD-W/Al interface and act as insulating layers which cause electrical degradation. In addition, impurities like fluorine or oxygen induce the formation of β-W lattices as well as high film resistivity. In this work, a two-step chemical vapor deposition of tungsten was developed to suppress the fluorine impurities in tungsten films and at CVD-W/Al interface for blanket CVD-W application. The first step involves a gas phase nucleation with high SiH4/WF6 flow ratio (i.e., 2.5) to deposit a thin tungsten film as the glue layer. It was found that the probability of the WF6 reduction with underlying aluminum was suppressed because the WF6 was completely consumed by SiH4 before arriving onto aluminum surface. Meanwhile, this gas phase nucleated tungsten exhibits blanket deposition capability and an amorphous structure. The second step includes typical CVD-W process (i.e., SiH4/WF6 flow ratio <1) to grow thick tungsten film. Secondary Ion Mass Spectroscopy (SIMS) measurements indicate that the fluorine impurities in tungsten film and at CVD-W/Al interface are drastically reduced. Also, the gas phase nucleated tungsten can be reproducibly deposited without attacking the aluminum underlayer. Moreover, a lower tungsten resistivity, lower via resistance and longer electromigration lifetime are achieved in the Al/W/Al Kelvin structures produced by two-step deposition technique than those of the typical CVD-W films deposited directly on aluminum.

Tuning the band alignment is proved to be an effective way to facilitate carrier transportation and thus enhance the power conversion efficiency (PCE) of solar cells. Doping the compact layer with metal ions or modifying the interfaces among functional layers in perovskite solar cells (PSCs) can appreciably improve the PCE of PSCs. Inspired by the rare earth elemental doping of TiO2, which has witnessed the success in photocatalysis and dye-sensitized solar cells, we firstly demonstrated here that La3+ doping in the mesoporous TiO2 layer of a mesostructured PSC can tune its Fermi level and thus significantly enhance the device PCE. Systematic analysis reveals that doping La3+ into TiO2 raises the Fermi level of TiO2 through scavenging oxygen and inducing vacancies, which subsequently increases the open circuit voltage and the fill factor while reducing the series resistance of the PSC using La3+-doped TiO2 as a mesoporous layer. As a result, a PCE of 15.42% is achieved, which is appreciably higher than the PCE of a device with undoped TiO2 (12.11%).

This work explores a mechanism behind hysteresis in CH3NH3PbI3 perovskite solar cells. The solar cells in this work employed either compact TiO2, mesoporous TiO2, or a combination of compact and mesoporous TiO2 as an electron transport layer. The solar cells using compact TiO2 layer displayed the most pronounced hysteresis compared to those which made use of mesoporous TiO2. Different hysteretic behavior is attributed to difference in the built-in electric fields present in the architecture of perovskite solar cell. The solar cells with a compact TiO2 layer have a built-in field which allows for iodide ions to migrate and accumulate near to the interface of indium-tin-oxide electrode, ultimately causing a reduction in the measured power conversion efficiency for forward bias scans. In case of the cells with a mesoporous TiO2 layer, they have the built-in fields configured in such a way that iodide ions are blocked from migrating on a large scale to the vicinity of the ITO electrode. This results in the reduced hysteresis in perovskite solar cells when a mesoporous TiO2 electron transport layer is employed.

Meso-structured perovskite solar cells (PSC), utilizing a mesoporous absorber layer consisting of mesoporous metal oxide and the perovskite material inside, are still delivering the highest solar cell efficiency for perovskite-based solar cells up to date. Their outstanding performance critically depends on the nanoscopic morphology formed inside the mesoporous absorber layer. This, however, is not accounted for in most of the perovskite device models, as they are based on an effective-medium formulation for the mesoporous absorber layer, and the details of its underlying morphology are ignored. The mesoporous absorber layer is treated as a two-phase model that describes intrinsic solar cell physics such as free charge carrier generation, carrier transport, and recombination within the two phases, as well as at the interface between the two phases. We derive a spatially smoothed device model for meso-structured PSCs based on volume-averaging of electric potential and electron and hole concentrations of the two-phase model, and this spatially smoothed formulation captures two essential morphological descriptors that are not found in existing effective-medium formulations for meso-structured PSCs, namely, surface-to-volume ratio and porosity inside the mesoporous layer. Furthermore, we determine the explicit functional forms of the effective parameters in the spatially smoothed model for the case of an ideal “Spaghetti” blend morphology.