Modelling Dye-Sensitised Nanocrystalline Solar Cells

Abstract

Dye-sensitized Solar Cells (DSCs) and Solid State Dye-sensitized Solar Cells (ss-DSCs) are photovoltaic devices with short energy payback times compared toexisting technologies. DSCs and ssDSCs have the potential to become cheapexible devices with new building integrated and portable applications.Despite over 20 years of research fundamental steps in the operation, constructionand measurement of DSCs and ssDSCs are still not completely understood. Forexample, the competition between the rate of electron transport through theanode and the rate of electron transfer to I {3 ions in the electrolyte determines theeciency of the cell. The diusion length is proportional to the average distancean electron will travel though the anode before reacting with I {3 . Conventionallyelectron transfer to I {3 has been described by the rate equation r = krnc wherenc is the concentration of photoelectrons and kr is a constant. Recently there hasbeen an increasing body of evidence suggesting so called non-linear recombinationwhere the reaction order is in fact less than one where r = krncand < 1. Manyof the standard methods for cell characterization have been reinterpreted in lightof this result, however an obvious omission is that there is no general modelfor electrochemical impedance spectroscopy (EIS) experiments when < 1. Wepresent a full EIS model for DSCs anodes, valid for the case where < 1, andthe resulting EIS spectra at dierent points along the J-V curve and for dierentreaction orders . The results are consistent with published experimental dataand show that the diusion length calculated from EIS experiments ought todecrease as the cell is moved from open to short circuit conditions.In light of non-linear reaction kinetics it has become increasingly important tomeasure the diusion length as simply and unambiguously as possible. As such, anew method of measuring the diusion length has been developed using patternedelectrodes. The electron extracting electrode is etched such that the requireddiusion length for 100% electron collection is changed. A simple 1D analyticalmodel for patterned electrode experiments is presented so that the diusion lengthcan easily be extracted from experimental data. The 1D model is veried with a2D nite element model.Understanding the processes involved in DSC construction is equally important asthose involved in DSC operation if the devices are to be manufactured on a large scale. The application of light sensitizing dye to the anode is one such example.Since the anode is porous on the nanoscale the dying process is slow, commonly 12hours are required to apply dye to the anode. To best understand dye applicationoptical waveguide spectroscopy (OWS) has been used to monitor the amountof dye entering the anode. This thesis presents a model of dye diusion andadsoption within the anode and an optical model of the OWS experiment. Thetwo models are used to interpret the OWS data and show that dye is enteringthe anode well after 12 hours. The model is then used to demonstrate that thedistribution of dye within the lm can be observed if a high refractive index rutileprism is used to couple light into the lm.In ssDSCs the conventional liquid electrolyte is replaced with a chemically dopedorganic hole conductor. Kinetic Monte Carlo Simulations are used explain therelationship between the conductivity of the organic matrix and the doping level.The results, which are consistent with experimental observations, suggest that atlow doping levels carriers fail to disassociate from their host dopants causing adecrease in mobility. This explains the only modest increase in conductivity seenat low doping levels.

Date of Award	9 Oct 2013
Original language	English
Awarding Institution	University of Bath
Supervisor	Alison Walker (Supervisor)