chapter  2
5 Pages

5.1 Block Copolymers in the Dye-Sensitized Solar Cell

Dye-sensitized solar cells (DSCs) are a paricularly successful example of a bulk heterojunction cell architecture. A wide bandgap inorganic semiconductor (typically a metal oxide) is sensitized to the solar spectrum by attaching a surface-adsorbed monolayer of an organic or organic-metal complex dye [96, 97]. If the redox potential of the photoexcited dye lies above the conduction band edge of the inorganic semiconductor then an electron may be injected into the layer. The oxidized dye is regenerated by electron transfer from a surrounding donor species. The first, and still most successful version of this concept (though there are now many variations) uses a donor species dissolved in a liquid electrolyte, usually an organic or ionic liquid solvent containing the I−/I3− redox couple [98]. The oxidized donor diffuses away to a counterelectrode where it is subsequently reduced to complete the cell circuit.The problem of exciton diffusion is avoided since all excited states are generated directly at a charge-separating interface. However, the confinement of the cell’s absorbing component to a single interfacial monolayer makes it very difficult to achieve significant light absorption in a flat layer. The bulk heterojunction concept again offers a solution in which a highly structured (or porous) inorganic layer provides the massively increased surface required to load with dye. Using a typical transition metal complex dye, a surface area enhancement of around 1000 over the flat surface equivalent is needed to achieve significant (ca. 90%) light harvesting. Electron extraction occurs via the diffusion of electrons through the nanocrystalline inorganic layer [99], again requiring a fully continuous network structure. This well-studied and modeled system provides an ideal testing ground for probing both the device performance of BCP morphologies and the dependence of important optoelectronic characteristics such as charge transport and recombination rates on the semiconductor topology.Fig. 2.26a-f summarizes the templated BCP morphologies accesible using electrochemical replication of TiO2 detailed in Section In addition to standing nanowires and the gyroid network, a disordered mesoporous layer of thermally sintered,

20 nm TiO2 nanoparticles is taken as comparison, representing the current state of the art inorganic nanostructure in DSCs. The predicted geometric surface areas (Table 2.3) expressed as the surface enhancement (or roughness factor (RF)) per micron in array thickness, are comparable for the three arrays; 100, 125, and 120 for wires, gyroid and nanoparticles, respectively. We expect the primary difference between the morphologies to lie rather in the dimensionality of charge transport out of the array. Standing wires offer direct and, in principle, independent extraction pathways while gyroid and nanoparticle layers privide an ordered or disordered 3D network of routes, respectively. The TiO2 crystal phase (indexed in all cases to anatase) and characteristic grain size in each of the three morphologies is shown by high resolution transmission electron microscopy (Fig. 2.26). In both electrochemically replicated morphologies the grain size is of the same order as the structural confinement (∼10 nm). Crystallite dimensions in nanowire templated structures may exceed this confinement length along the wire axis in some places (Fig. 2.26g, inset). The nanoparticle network (Fig. 2.26c) is made up of ∼20 nm randomly contacted single-crystal nanoparticles.