The objective of this work is to develop and optimize the new dye-sensitized solar cell technology. For the first time, crack-free nanocrystalline rutile TiO 2 films with thicknesses of up to 12 µm were prepared and characterized.The photoelectrochemical properties of the rutile-based solar cell were compared with those of the conventional anatase-based cell. Intensity-modulated photocurrent spectroscopy and scanning electron microscopy studies indicate that electron transport is slower in the rutile layer than in the anatase layer due to differences in the extent of inter-particle connectivity associated with the particle packing density. In view of the infancy of rutile material development for solar cells, the PV response of the dyesensitized rutile-based solar cell is remarkably close to that of the anatase-based cell.

We report transient photocurrent measurements on solar cell structures based on dye-sensitized, porous TiO2 films filled with a liquid electrolyte. The measurements are interpreted as ambipolar diffusion; under most measurement conditions, the ambipolar diffusion coefficient is dominated by electrons diffusing in the TiO2 matrix. We report a strong dependence of the ambipolar diffusion coefficient upon the photoexcitation density, as has been proposed previously. The coefficients vary from 10-8 cm2 s-1 at low density to 10-4 cm2 s-1 for densities of 1018 cm-3. At a specified photoexcitation density, ambipolar diffusion coefficients measured using weak laser pulses and optical bias are about 10 times larger than coefficients measured using large-intensity laser pulses. We describe trapping models for these effects based on an exponential distribution (T 0 = 650 K) of electron trap levels in TiO2. We infer an electron recombination cross section less than 2 × 10-27 cm2; this value is nearly 10 orders of magnitude smaller than typical values in compact semiconductors and indicates the extraordinarily effective separation of electrons in the TiO2 matrix from electrolyte ions only nanometers distant.

The effect of lithium intercalation on the transport dynamics and recombination kinetics in dye-sensitized nanoparticle TiO 2 solar cells at lithium levels below 5 atom % was investigated by photocurrent and photovoltage transient and spectroelectrochemical techniques. Titanium dioxide films were doped electrochemically in the dark and under illumination. It was discovered that when Li + is present in the electrolyte, lithium intercalates irreversibly into dye-sensitized TiO 2 films at open circuit (ca. -0.7 V) under normal solar light intensities. Photocurrent transients of doped nonsensitized TiO 2 films indicate that lithium doping decreases the diffusion coefficient of electrons through the nanoparticle network. Photocurrent and photovoltage transients of sensitized TiO 2 films provide the first evidence that electron transport limits recombination with the redox electrolyte in working cells. As the Li density in the films increases, the diffusion and recombination times of photoelectrons increase proportionately, indicating a causal link between electron transport and recombination. The electron diffusion coefficient in dye-sensitized solar cells exhibits a power-law dependence on photocharge at all concentrations of inserted lithium in the TiO 2 film. With increasing doping, the dependence of the electron diffusion coefficient on the photocharge becomes stronger, a phenomenon attributed to widening of the exponential conduction band tail resulting from disorder induced by randomly placed lithium defects in TiO 2 . The photovoltaic characteristics of dye-sensitized solar cells are largely unaffected by lithium intercalation, implying that intercalation has only a small effect on the charge collection efficiency and the rate of recombination. A simple model is presented that explains the observed transport-limited recombination. The results suggest that increasing the electron transport rate will not significantly improve the solar cell performance.

Percolation theory is applied to understand the influence of network geometry on the electron transport dynamics in dye-sensitized nanocrystalline TiO 2 solar cells, and the predicted results are compared with those measured by transient photocurrent. The porosity of the films was varied experimentally from 52 to 71%. Electron transport was modeled using simulated mesoporous TiO 2 films, consisting of a random nanoparticle network, and the random-walk approach. The electron transport pathway through the network was correlated with the film porosity and the coordination numbers of the particles in the film. The experimental measurements and random-walk simulations were in quantitative agreement with percolation theory, which predicts a powerlaw dependence of the electron diffusion coefficient D on the film porosity as described by the relation: D ∝ |P -P c | µ . The critical porosity P c (percolation threshold) and the conductivity exponent µ were found to be 0.76 ( 0.01 and 0.82 ( 0.05, respectively. The fractal dimension of the nanoparticle films was estimated from the simulations to be 2.28, which is in quantitative agreement with gas-sorption measurements. It is shown that as the porosity increases, the distribution of the coordination numbers of the particles shifts from an emphasis on high coordination numbers to low ones, causing the electron transport pathway to become more tortuous and electron transport to slow. Another consequence of increasing the porosity is that the fraction of terminating particles (dead ends) in the TiO 2 film increases markedly, from less than 1% for a 50% porous film to 31% for a 75% porous film. It is estimated that during their respective transit through 50 and 75% porous 10-µm thick films, the average number of particles visited by electrons increases by 10-fold, from 10 6 to 10 7 . This study provides the first clear evidence that network topology has a strong influence on the electron transport dynamics in mesoporous TiO 2 films.

The role of electrical potential, charge transport, and recombination in determining the photopotential and photocurrent conversion efficiency (IPCE) of dye-sensitized nanocrystalline solar cells was studied. Electrostatic arguments and electrical impedance spectroscopy (EIS) are used to obtain information on the electrical and electrochemical potential distribution in the cell. It is shown that on the macroscopic level, no significant electrical potential drop exists within the porous TiO2 when it contacts the electrolyte and that the electrical potential drop at the transparent conducting oxide substrate (TCO)/TiO2 interface occurs over a narrow region, one or two layers of TiO2. Analyses of EIS and other data indicate that both the photopotential of the cell and the decrease of the electrical potential drop across the TCO/TiO2 interface are caused by the buildup of photoinjected electrons in the TiO2 film. The time constants for the recombination and collection of the photoinjected electrons are measured by EIS and intensity-modulated photocurrent spectroscopy (IMPS). As the applied bias is varied from short-circuit to open-circuit conditions at 1 sun light intensity, recombination becomes faster, the collection of electrons becomes slower, and the IPCE decreases. The decrease of IPCE correlates directly with the decline of the charge-collection efficiency ηcc, which is obtained from the time constants for the recombination and collection of the photoinjected electrons. Significantly, at open circuit, η cc is only 45% of its short-circuit value, indicating that the dye-sensitized nanocrystalline TiO2 solar cell behaves as a nonideal photodiode.