Existing types of solar cells may be divided into two distinct classes: conventional solar cells, such as silicon p-n junctions, and excitonic solar cells, XSCs. Most organic-based solar cells, including dye-sensitized solar cells, DSSCs, fall into the category of XSCs. In these cells, excitons are generated upon light absorption, and if not created directly at the heterointerface as in DSSCs, they must diffuse to it in order to photogenerate charge carriers. The distinguishing characteristic of XSCs is that charge carriers are generated and simultaneously separated across a heterointerface. In contrast, photogeneration of free electron-hole pairs occurs throughout the bulk semiconductor in conventional cells, and carrier separation upon their arrival at the junction is a subsequent process. This apparently minor mechanistic distinction results in fundamental differences in photovoltaic behavior. For example, the open circuit photovoltage V oc in conventional cells is limited to less than the magnitude of the band bending Ø bi ; however, V oc in XSCs is commonly greater than Ø bi . Early work on solid-state excitonic solar cells is described as are excitonic processes in general and the use of carrier-selective (energy-selective) contacts to enhance V oc . Then studies of DSSCs, which provide a particularly simple example of XSCs, are described. A general theoretical description applicable to all solar cells is employed to quantify the differences between conventional and excitonic cells. The key difference is the dominant importance, in XSCs, of the photoinduced chemical potential energy gradient ∇µ hν , which was created by the interfacial exciton dissociation process. Numerical simulations demonstrate the difference in photoconversion mechanism caused solely by changing the spatial distribution of the photogenerated carriers. Finally, the similarities and differences are explored between the three major types of XSCs: organic semiconductor cells with planar interfaces, bulk heterojunction cells, and DSSCs.

No abstract

Conventional dye-sensitized solar cells function efficiently only with a single redox couple, I-/I2, because of the unusually slow kinetics for I2 reduction on SnO2 and TiO2 surfaces. When faster redox couples such as ferrocene/ferrocenium are employed, the rapid interfacial recombination of photoinjected electrons with the oxidized half of the redox couple eliminates the photovoltaic effect. To make use of other, perhaps more appropriate, redox couples in these cells, the interfacial recombination processes must be understood and controlled. Charge recombination at the SnO2/solution interface is clearly distinguishable from recombination at the nanoporous TiO2/solution interface. Dark current measurements probe mainly the former reaction, although the latter may be the dominant recombination mechanism under illumination. We introduce two methods for passivating the interfaces that decrease the recombination rates by orders of magnitude. One method involves electropolymerization of an insulating film of poly(phenylene oxide-co-2-allylphenylene oxide) on the solvent-exposed parts of the SnO2 surface. The other involves treating the dye-sensitized film with reactive methyltrichlorosilane vapor that forms an insulating film of poly(methylsiloxane) on both the exposed TiO2 and SnO2 surfaces. These methods make it possible for the first time to use kinetically fast redox couples in dye-sensitized solar cells, and they may facilitate the development of a viable solid-state version of these cells.

Charge carriers are photogenerated with very different spatial distributions in conventional inorganic photovoltaic (IPV) cells and in organic photovoltaic (OPV or excitonic) cells. This leads to a fundamental, and often overlooked, mechanistic difference between them. Carriers are generated primarily at the exciton-dissociating heterointerface in OPV cells, resulting in the production of electrons in one phase and holes in the other—the two carrier types are thus already separated across the interface upon photogeneration in OPV cells, giving rise to a powerful chemical potential energy gradient ∇μhv that promotes the photovoltaic effect. This occurs also in high-surface-area OPV cells, although their description is more complex. In contrast, both carrier types are photogenerated together throughout the bulk in IPV cells: ∇μhv then drives both electrons and holes in the same direction through the same phase; efficient carrier separation therefore requires a built-in equilibrium electrical potential energy difference ∅bi across the cell. The open-circuit photovoltage Voc is thus limited to ∅bi in IPV cells, but it is often greater than ∅bi in OPVs. The basic theory necessary to compare IPVs to OPVs is reviewed. Relevant experiments are described, and numerical simulations that compare semiconductor devices differing only in the spatial distribution of photogenerated carriers are presented to demonstrate this fundamental distinction between the photoconversion mechanisms of IPV and OPV devices.

A new sensitizing dye-semiconductor system comprised of perylene derivatives on SnO 2 has been characterized. The tetracarboxylic acid form ("PTCA") of the commercially available dye perylene-3,4,9,-10-tetracarboxylic dianhydride and the novel compound perylene-3,4-dicarboxylic acid-9,10-(5-phenanthroline)carboximide ("PPDCA") adsorb strongly to the surface of colloidal films of SnO 2 and inject electrons into the semiconductor film upon absorption of light. A film of PPDCA on SnO 2 yields a short circuit photocurrent density of 3.26 mA/cm 2 , a photovoltage of -0.45 V, and an overall cell efficiency of 0.89% at one sun light intensity. Estimates of the oxidation potential of adsorbed PPDCA indicate that it may also be useful in a water-splitting configuration. The results presented here indicate that the perylene-SnO 2 system is a promising dye-semiconductor combination and warrants further study.