(103f) Modeling of Solid and Particulate Passivating Layers in Dye Sensitized Solar Cells

Dye-sensitized solar cells (DSCs) utilize a porous layer of wide-bandgap semiconductor nanoparticles which have been deposited on a transparent conductive oxide (TCO) surface such as tin-doped indium oxide or fluorine-doped tin oxide. This porous layer, typically about 10-20 microns thick, is coated with a dye which is highly absorbent over the visible spectrum. This dye functions as the means of charge separation in these cells, injecting excited electrons into the porous TiO2 layer and interacting with a redox couple suspended in a liquid electrolyte.

A common source of current loss in DSCs is the transfer of electrons from the TCO substrate to the liquid redox couple, occurring at the undesirable interface of the liquid and the TCO. This problem is typically addressed by depositing a passivating layer of TiO2 on the bare TCO before the deposition of the porous layer. While this effective method for stopping TCO-electrolyte recombination is standard practice when building DSCs, there is little consensus on the specific impact of this passivating layer on DSC performance.

Our lab group has fabricated DSCs using passivating layers consisting of both solid-phase TiO2 and particulate TiO2, and we have compared these to results from other published works. We note that, when using passivating layers consisting of particulate TiO2 which have been deposited via electrostatics, we observe a 9% improvement in open-circuit voltage and up to a 50% improvement in short-circuit currents as opposed to cells which are not passivated. Additionally, cells which are passivated using particulate layers show higher short-circuit currents than cells passivated using solid TiO2, but have lower fill factors.  We have used these data as an aid to the building of a theoretical and computational model for the steady-state and time-dependent operation of a DSC, with an emphasis on the effects of interactions occurring at the TCO-electrolyte interface. Our calculations have shown that by simply allowing for TCO-electrolyte interaction via a Butler-Volmer mechanism, we are able to account for commonly-cited, non-ideal current-voltage behavior of the anode in DSCs. Additionally, taking this interaction into account when modeling steady-state cell behavior allows for higher accuracy when calculating overpotential losses in the liquid components of the DSC. Time-dependent studies, solved numerically, have proven to be qualitatively similar to experimental impedance measurements, indicating that our additions to the modeling of DSCs carry physical significance. We believe that the differences in resistance, as well as the effective passivated area, between solid and particulate blocking layers results in the differences in their effect on DSC performance.

This work is the first suggestion for an analytical mechanism by which TCO-electrolyte interaction affects DSC performance, as well as an explanation for non-ideal anode behavior. While experimental optimization of DSC blocking layers is common practice, we hope that this work will begin more fundamentals-driven optimization.