Characterization of Polymer/TiO2 Photovoltaic Cells by Intensity Modulated Photocurrent Spectroscopy

A dynamic model for the intensity modulated photocurrent spectroscopy (IMPS) of bilayer polymer/TiO2 photovoltaic cells is developed in consideration of the exciton generation in the polymer layer, exciton diffusion to the polymer/TiO2 interface, electron injection into the TiO2 layer, and electron diffusion through the TiO2 layer; particularly, the phase shift 0,,(a)) due to the time delay between exciton generation and dissociation is included in the continuity equations for electron transport. Bilayer polymer/TiO2 cells consisting of poly(2-methoxy-5-(2-ethyihexyloxy)-1,4-phenylene vinylene) (MEH-PPV) and nanostructured TiO2 were prepared for experimental purposes, with a TiO2 layer thickness d of 120 and 65 nm. Experimental data obtained confirm all the main expectations of the model, providing important information on incident photon-to-current conversion efficiency (IPCE), exciton dissociation, and electron transport. The frequency-dependent phi(n)(omega) affects the location (P-high) where IMPS response crosses with the positive real axis at high frequency, by imposing a phi(n)(omega) effect on electron transport in the TiO2 layer to the collection electrode. A more remarkable phi(n)(omega) effect results from either a larger electron diffusion coefficient D-e in TiO2 or a smaller d. An increased d value makes the Phigh point tend toward the origin and the electron transit time TD through Ti02 layer increase, but both S and IPCE values decrease, and the IPCE value depends greatly on the exciton dissociation rate S at the polymer/TiO2 interface. The measured IMPS responses of the cells are satisfactorily fitted to the model. Dynamic results show that both cells exhibit a phi(n)(omega) effect with wo of 6 x 10(4) rad/s and a D-e of 2 x 10(-5) cm(2)/s, but the different d values lead to varied tau(D), S, and IPCE values, which are, respectively, 1.90 ms, 141 cm(2)/s and 0.380% for d = 120 nm, but 0.48 ms, 153 cm(2)/s, and 0.410% for d = 65 tim. Our studies demonstrate that the exciton dissociation efficiency at the polymer/TiO2 interface is crucially important for highly efficient device performance.