Nanostructured polymer solar cells (PSCs) have emerged as a promising low-cost alternative to conventional silicon-based photovoltaic devices. Since PSCs can be fabricated by processing polymers, eventually together with other organic materials, from solution and depositing them onto different types of substrates by roll-to-roll production processes, they offer important advantages over their inorganic counterparts due to easy and fast processing and low cost of fabrication. Despite the progress made in this field, PSCs are still in the early research and development stage. For PSCs to become practical efficient devices, several issues need to be addressed, including further understanding of their operation and stability and better control of the morphology formation, mainly morphology of the photoactive layer, which directly determines the performance of PSC devices. Photoactive layers of PSCs are ultra-thin films (200-250 nm at maximum) composed of a mixture of electron donor and electron acceptor materials. Both the processing conditions of donor/acceptor mixtures from solution and eventual postproduction treatments like thermal annealing have a large influence on performance of PSC, because they strongly affect the scale of phase separation (which should ideally be in the range of ~10 nm for efficient exciton dissociation), distribution of phases through the thickness of the layer (which largely determines free charges transport to the electrodes) and eventual crystalline order (beneficial for higher mobility of free charges). The research described in this thesis focuses on the relationship between 3D nanoscale morphology organization of photoactive layers and performance of corresponding PSC devices, with aim to identify the critical morphology parameters contributing to high power conversion efficiency of PSCs. For comprehensive 3D morphology characterization, the technique of electron tomography (or transmission electron microtomography) was intensively applied, among other techniques. In electron tomography, a series of 2D projections taken from a specimen tilted inside a transmission electron microscope (TEM) is used to reconstruct a 3D image of the specimen with a nanometer-scale resolution. This 3D image can then be used, voxel by voxel, to study in detail the specimen’s morphological organization. The general principles of electron tomography are introduced in Chapter 2, as well as the most important aspects of this technique specific for polymer-based films such as photoactive layers of PSCs. The first applications of electron tomography to photoactive layers of PSCs (described in Chapter 3) showed that the existence of 3D nanoscale networks of donor and acceptor components is indispensable for efficient PSC devices. Such networks ensure effective exciton dissociation at the donor/acceptor interface and effective charge carrier transport from any place in the photoactive layer to the electrodes. It has long been speculated that the volume of photoactive layers should consist of bi-continuous nanoscale interpenetrating networks of the donor and acceptor materials but it was for the first time that such networks were observed directly with nanometre resolution. The most efficient up to now PSC system, viz. the one based on combination of regioregular poly(3-hexylthiophene) (P3HT) as the electron donor and the fullerene derivative (PCBM) as the electron acceptor material, has been studied with electron tomography in more detail (see Chapters 4 and 5). Electron tomography has proved indispensable to quantify with high accuracy the amount of crystalline P3HT nanowires in (annealed) P3HT/PCBM layers through the whole layer thickness; information that was inaccessible before application of electron tomography. It was found that the most crucial parameters for performance of P3HT/PCBM solar cells are the high overall crystallinity of P3HT and enrichment of crystalline P3HT closer to the positive (hole collecting) electrode. Such enrichment should ensure that the percolation pathways within the P3HT and PCBM phases leading to the corresponding electrodes are more effective and that the possibilities of charge recombination are limited, while higher overall crystallinity of P3HT ensures higher mobility of holes (that have in this system intrinsically lower mobility than electrons). Different types of vertical distribution of crystalline P3HT in P3HT/PCBM layers prepared on the same (PEDOT:PSS) substrate were observed by electron tomography in this project, viz. more crystalline P3HT next to the positive electrode, more crystalline P3HT next to the negative electrode and homogeneous distribution of P3HT through the layer thickness (see e.g. Section 4.7). Previous studies suggested that PCBM was preferentially concentrated next to the positive electrode in the P3HT/PCBM films and P3HT next to the negative electrode and that the reason for this is the high surface energy of the PEDOT:PSS layer. Our observations do not support this hypothesis and dismiss the key role of the underlying substrate’s surface energy. Instead, the existence and type of composition gradients through the thickness of the active layer was found to be largely determined by the molecular weight distribution and regioregularity of P3HT (and possibly by impurities present in both P3HT and PCBM) and by kinetic aspects of film formation due to different solution viscosity, different time it takes for the solvent to evaporate and eventual differences in local solvent concentration. All these aspects have an impact on formation of nuclei and subsequent growth and distribution of (nano)crystals throughout the photoactive layer. What limits the application of electron tomography in photoactive layers of PSCs is the (often) rather poor contrast in TEM between the carbon-based donor and acceptor materials. The poor contrast complicates the segmentation of 3D datasets and thus limits the quantitative information that can be extracted from the electron tomography datasets. Sometimes the contrast in carbon-based (nano)crystalline specimens can be improved by applying HAADF-STEM (high-angle annular dark field scanning TEM) imaging mode instead of conventional bright-filed TEM mode, as we have observed in MDMO-PPV/PCBM system (see Chapter 3). To make it possible to segment and hence elaborately quantify the P3HT/PCBM and other datasets, one could also look into the possibilities that imaging filters offer, e.g. the median or smoothing filters for reducing noise, Laplacian (or differencing) operator for edge enhancing, Sobel or Kirsch transform for edge finding/detection, etc. The other approach would be to apply an alternative reconstruction method (instead of SIRT used in this project), viz. discrete tomography (DART) where the data are segmented in-situ during reconstruction. The 3D datasets that could be straightforwardly segmented and quantified in great detail were obtained from the photoactive layers of hybrid solar cells composed of P3HT and ZnO that give a high contrast in TEM. The nanoscale 3D morphology of P3HT/ZnO layers could be statistically analyzed for percolation pathways and for the scale of phase separation (by quantifying spherical contact distances). This quantitative analysis enabled differentiating between charge generation and charge transport as limiting factors to the device performance (see Chapter 7). It was also observed that the control over morphological organization of the photoactive layers is difficult to achieve by spontaneous morphology formation during film deposition from solution, especially in case of thicker films, as described in Chapter 4. A way to get a better control is to form desirable nanostructures, such as crystalline P3HT nanowires, in solution (dispersion) prior to the active layer deposition. Two approaches to cause P3HT crystallization have been tried in this project (see Chapter 6): one (a so-called solvatochromic approach) is based on using mixed solvents (i.e. adding an excess of a poor solvent to a solution of P3HT in a good solvent) and the other (thermochromic) approach is based on preparing a saturated solution of P3HT at elevated temperature and then cooling it down to room temperature. Both approaches have resulted in formation of crystalline P3HT nanowires, as confirmed by UV-Vis spectroscopy and microscopy techniques (TEM and AFM). However, the amount of the obtained P3HT nanowires was relatively low and formation of an undesirable precipitate was often observed. In case of thermochromic approach, the precipitate was most probably formed due to too rapid cooling of the saturated solutions. This research should be further continued to find out how to prevent the precipitate formation, increase the yield of crystalline P3HT nanowires and make their practical applications in solar cell devices possible. With a better understanding (facilitated by electron tomography) of how the 3D morphology of photoactive layers affects the device performance and with an improved control over the 3D morphology formation, bulk-heterojunction PSCs will eventually come closer to the stage of practical efficient devices.
|Kwalificatie||Doctor in de Filosofie|
|Datum van toekenning||14 dec 2009|
|Plaats van publicatie||Eindhoven|
|Status||Gepubliceerd - 2009|