Abstract
We report on electroabsorption (EA) and built-in voltage (V(BI)) measurements of polymer light-emitting diodes with the general structure ITO/PEDOT:PSS/emitting polymer/LiF/Ca/Al where ITO is indium tin oxide, PEDOT:PSS is poly(3,4-ethylene dioxythiophene) doped with poly(styrene sulfonate), and the emitting polymer is either poly(9,9-dioctylfluorene), poly(9, 9-dioctyl-fluorene-alt-bis-N,N(')-(4-butyl-phenyl)-bis-N, N(')-phenyl-1, 4-phenylenediamine), or poly(9,9-dioctylfluorene-alt-N-(4-butylphenyl)-diphenylamine). We find that the EA nulling voltage, i.e., the dc bias at which the EA signal vanishes, depends on the frequency of the ac voltage and on the incident photon wavelength. Such dependence poses a problem for accurate measurement of the built-in voltage (V(BI)), which is the voltage generated between the electrodes upon equilibration of the Fermi levels through the heterostructure. We find that the EA signal is mixed with a smaller intensity signal which can be ascribed to excited state absorption (ESA). We propose a method for separating the excited state absorption signal and producing accurate V(BI) measurements. We also demonstrate that in our devices the ESA contribution to the electromodulated (Delta T/T) signal is negligible with respect to the accuracy with which V(BI) can be determined from the nulling voltage of Delta T/T and can thus be safely ignored.