Dye-Sensitized Approaches to Photovoltaics

Sensitization of wide band-gap semiconductors to photons of energy less than the band-gap is a key step in two technically important processes - panchromatic photography and photoelectrochemical solar cells. In both cases the photosensitive species is not the semiconductor - silver halide or metal oxide - but rather an electrochemically active dye. The gap between the highest occupied molecular level (HOMO) and the lowest unoccupied molecular level (LUMO) is less than the band-gap of the semiconductor with which it is associated. It can therefore absorb light of a wavelength longer than that to which the semiconductor itself is sensitive. The electrochemical process is initiated when the dye molecule relaxes from its photoexcited level by electron injection into the semiconductor, which therefore acts as a photoanode. If the dye is in contact with a redox electrolyte, the negative charge represented by the lost electron can be recovered from the reduced state of the redox system, which in return is regenerated by charge transfer from a cathode. An external load completes the electrical circuit. The system therefore represents a conversion of the energy of absorbed photons into an electrical current by a regenerative device in every functional respect analogous to a solid-state photovoltaic cell. As in any engineering system, choice of materials, their optimization and their synergy are essential to efficient operation. While a semiconductor-electrolyte contact is analogous to a Schottky contact, in that a barrier is established between two materials of different conduction mechanism, with the possibility of optical absorption, charge carrier pair generation and separation, it should be remembered that the photogenerated valence band hole in the semiconductor represents a powerful oxidizing agent. Given that the band-gap is related to the strength and therefore the stability of chemical bonding within the semiconductor, for narrow-gap materials the most likely reaction of such a hole is the photocorrosion of the semiconductor itself. However, only relatively narrow band-gap materials have an effective optical absorption through the visible spectrum, towards and into the infra-red. Materials with an optimal band-gap match to the solar spectrum, of the order of 1.5eV, are therefore electrochemically unstable. A stable photoelectrochemical cell, without some process of optical sensitization, and necessarily using a wide-gap semiconductor is sensitive only to the ultra-violet limit of the visible spectrum. Over recent years a suitable combination of semiconductor and sensitizer has been identified and optimized, so that now a solar spectrum conversion efficiency of over 11{\%} has been verified in a sensitized photoelectrochemical device. One key to such an efficient system is the suppression of recombination losses. When the excited dye relaxes by electron loss, the separated charge carriers find themselves on opposite sides of a phase barrier -- the electron within the solid-state semiconductor, the positive charge externally, in association with the dye molecule. There is no valence---band involvement in the process, so the system represents a majority-carrier device, avoiding one of the major loss mechanisms in conventional photovoltaics. In consequence also a highly-disordered, even porous, semiconductor structure is acceptable, enabling surface adsorption of a sufficient concentration of the dye to permit total optical absorption of incident light of photon energy greater than the HOMO-LUMO gap of the dye molecule. The accepted wide-band semiconductor for photoelectrochemical applications is titanium dioxide in the anatase crystal structure. The size of the nanocrystals making up the semiconductor photoanode can be determined by hydrothermal processing of a precursor sol, and the film can be deposited on a transparent conducting oxide (TCO) substrate by any convenient thin-film process such as screen printing or tape casting. The preferred dye system is inspired by the natural processes involving chlorophyll, the coloring material in plants on which all earthly life depends. Chlorophyll is an organometallic dye, with a metal ion, Mg, within a porphyrin cage of nitrogen atoms. The synthetic chemist of course can select any convenient metal within the periodic table, and experience shows that ruthenium has the optimal properties expected. A ruthenium-pyridyl complex provides the chromophore of the dye, with the HOMO-LUMO gap, and thence the absorption spectrum bring modified by substitution with thiocyanide groups. Chemisorptive attachment of the dye to the metal oxide surface is obtained by carboxyl groups attached to the pyridyl components. The energetics of the dye is such that the LUMO level is just above the conduction band edge of the semiconductor, enabling relaxation by electron injection as required. A satisfactory electroactive dye structure, with good attachment properties and a wide optical absorption spectrum is therefore a sophisticated molecular engineering product. The electrolyte is also an optimized electrochemical system. The basic redox behavior is provided by the iodine/iodide system, with the advantage that the ions, both oxidized and reduced are relatively small, and therefore mobile in the supporting electrolyte. Energy losses due to slow diffusion are minimized. Early experiments used aqueous electrolytes, though with limited cell lifetime due to hydrolysis of the chemisorptive dye---semiconductor bond. A wide range of organic systems were therefore investigated, with the present favored formulation being based on imidazole salts. These have the additional advantage of low vapor pressure, very necessary as the photoactive sites under mid---day sun illumination may reach 80\r{ }C or higher. Low losses at the cathode counterelectrode are also a requirement for cell efficiency. The cathode is not necessarily transparent, and prototype cells on thin metal foils have been produced. However a TCO on glass or polymer counterelectrode is widely used. In either case suitable electrocatalytic behavior is required and frequently a nanodispersed Pt precipitated from haxachloride solution is employed. It is by now evident that the achievement of an industrially-competitive sensitized photoelectrochemical solar cell is the result of the optimization of several components, associated obviously with their effective synergy. Each change of a single component has repercussions on the choice and performance of others. However as already mentioned an efficiency of over 11{\%} has now been certified, and a stability of over 14,000 hours under accelerated testing with continuous simulated AM1.5 illumination was recently reported. In consequence there is increasing confidence on the part of industry. Several licensees of EPFL patents on dye---sensitized photovoltaic systems are now preparing for large-scale production. G24 Innovations PLC in Wales is commissioning a manufacturing plant, and Dyesol PLC in Australia is making available the required materials on an industrial scale. In conclusion, then, it can be stated that the DSC system is much more than a fascinating scientific artifact illustrating charge-transfer mechanisms at electrochemical interfaces; an efficiency and reliability with industrial credibility have been demonstrated and verified, and a significant role in competition with other photosystems can be foreseen.