Quantifying the efficiency of electrodes for positive carrier injection into poly(9,9-dioctylfluorene) and representative copolymers

The perfect injecting contact for any semiconductor device is, by definition, an ohmic contact. When such a contact is made to an organic semiconductor the current density is limited by bulk space-charge effects. In the absence of charge carrier traps, J reaches the ultimate, trap-free, space-charge-limited value, J(TFSCLC)=(9/8)epsilon muV(2)/d(3). Knowledge of the mobility mu, permittivity epsilon, applied bias V, and film thickness d, thus allows the maximum possible current density to be calculated. The absolute injection efficiency of any specific contact can then be quantified via a figure of merit, chi =J/J(TFSCLC), namely the ratio of the actual current density to that expected for the ideal trap-free, space-charge-limited current. In this article we report on the injection efficiency of positive carriers into poly(9,9-dioctylfluorene) (PFO) and two representative copolymers, poly(9,9-diocytlfluorene-co-bis-N,N'-(4-methoxyphenyl)-bis-N, N-'-phenyl-1,4-phenylenediamine) (PFMO) and poly(9,9-dioctylfluorene-co-benzothiadiazole) (BT). Time-of-flight photocurrent, dark injection transient current, and current density-voltage (J-V) measurements were each performed on indium tin oxide (ITO)/polymer/Au or Al diode structures. The hole injection efficiency of various pretreated ITO electrodes and of the top Au and Al contacts was investigated. ITO coated glass substrates were cleaned by washing with solvents and then either not subjected to further treatment (untreated), exposed to an oxygen plasma (O-2 plasma), or coated in a poly(ethylenedioxythiophene)/polystyrenesulphonic acid (PEDOT/PSS) blend. The steady-state J-V characteristics for the different device structures were compared to the expected J(TFSCLC) and the figure of merit chi was calculated. At an applied field of 5x10(5) V/cm, the absolute injection efficiencies of holes into PFMO (ionization potential, I-p=4.98 eV) from untreated, O-2 plasma treated, and PEDOT/PSS treated ITO were found to be chi =10(-3), 1, and 1, respectively. For PFO (I-p=5.8 eV) the same contacts gave chi =10(-7), 10(-6), and 10(-3), respectively. For BT (I-p=5.9 eV) hole transport is highly dispersive, precluding determination of the hole mobility and hence an analysis of the injection efficiency. The injected current densities are, however, comparable to (within a factor of 5 of) those for PFO. The top Au and Al contacts exhibited injection limited behavior (10(-8)< chi < 10(-2)). Heat treatment of the Au top contacts did not improve injection, contrary to previous observations for a molecularly doped xerographic polymer system. (C) 2001 American Institute of Physics.