Dye-Sensitized Titanium Dioxide

Dye-sensitized solar cells (DSSCs) based on nanocrystalline titanium dioxide (TiO₂) thin films have received much research attention in the past decade. These solar cells have reported solar-to-electric energy conversion efficiencies of greater than 10%. Given the low cost of their components and their demonstrated conversion efficiencies, DSSCs show true promise for being an inexpensive, renewable, and environmentally benign alternative to fossil fuels. DSSCs are made by covalently bonding a sensitizer dye to the surface of a thin film of nanocrystalline TiO₂, which does not absorb visible light because it is a wide-gap semiconductor. Sensitization with a dye, such as Ruthenium-535 [Ru(4,4'-dicarboxy-2,2'-bipyridine)₂(NCS)₂], greatly enhances the ability of these cells to absorb sunlight. Electrons are injected from the dye molecule's excited state to the conduction band of the TiO₂ on a sub-picosecond time scale, where they are conducted through the electrical circuit (shown below).

Optical studies have sought to understand the nature of electron transfer from the dye to the semiconductor, and to understand how the electrons are transported through the porous TiO₂ network. Unfortunately, these studies of electron transport are indirect. Recent mid-infrared probe studies have begun to measure the true dynamics of the photoexcited electrons. However, it is much better to probe the sample response in the far-infrared region of the spectrum because that is where the complex conductivity has greatest variation, and therefore greatest information content.

Time-resolved terahertz spectroscopy (TRTS) is a noncontact far-infrared probe of conduction electrons that is perfectly suited to the study of ultrafast carrier dynamics. We have already demonstrated that it is applicable to the study of semiconductor single crystals and DSSCs. We have extracted the frequency-dependent, time-resolved photoconductivity of both Ruthenium-535 and Coumarin-343 stained TiO₂ electrodes. The results of these experiments are currently being analyzed.

In our present configuration, the dye molecules are photoexcited at 400-nm because they do not absorb near the fundamental wavelength of our laser system (800-nm). The absorption spectra and molecular structures of Ruthenium-535 and Coumarin-343 dissovled in ethanol are shown to the right. Unsensitized TiO₂ also absorbs at 400-nm, so we would like to extend our studies to dyes and experimental configurations that are based on photoexcitation at wavelengths between 400 and 600 nm.

This investigation addresses fundamental questions about conduction in porous semiconductors, and should lead to new insight about the nature of photoconduction in dye-sensitized solar cells.