The mechanism of charge transport in molecularly doped polymers has been the subject of much discussion over the years. In this paper, data obtained from a new experimental variant of the time-of-flight (TOF) technique, called TOF1a, are compared to the predictions of a two-layer multiple trapping model (MTM) with an exponential distribution of traps. In the recently introduced TOF1a experimental variant, the charge generation depth is varied continuously, from surface generation to bulk generation, by varying the energy of the electron-beam excitation source. This produces systematic changes in the shape of the current transient that can be compared to predictions of the two-layer MTM. In the model, one additional assumption is added to the homogeneous MTM, namely: that there exists a surface region, on the order of a micrometer thick, in which the trap distribution is identical to the bulk, but has a higher trap concentration. We find that the characteristic experimental features of an initial spike, a flat plateau, and an anomalously broad tail, as well as the sometimes observed cusp or decreasing current occurring near the transit time, can all be described by such a two-layer model; that is, they can arise as a result of carriers delayed by a trap-rich surface layer. We find that we can semiquantitatively fit current transient data over the whole time range of the experiment, but only by using theoretical parameters that lie in a narrow range, the extent of which we quantify here.

Transient conductivity measurements have been carried out in a hydrazone-polycarbonate molecularly doped polymer over wide dynamic ranges of time and current as a function of the applied electric field. These data are plotted using both linear−linear and log−log axes. The plots on linear−linear axes give familiar results: an initial spike followed by a relatively flat current before the transit time; a mobility that equals published values and precisely follows the Poole−Frenkel law, being exponential in the square root of the electric field; and a current after the transit time that falls much more slowly than can be accounted for using Gaussian statistics. The same data plotted using log−log axes reveals the behavior over long times and low currents. Before the transit time, the decrease of the current can be characterized by two power laws, but it only decreases by about a factor of 2 from 10− 2 of the transit time to the transit time and is almost independent of electric field from 5 to 80 V/μm. After the transit time the current decreases algebraically as t −1.75 to a current value of about 10−2 of the value at the transit time and is also almost independent of electric field. Available theoretical models are not consistent with these data. We note a surprising coincidence: the current transients are virtually identical to the current transients obtained in molecular crystals, as if disorder does not play a role in determining the shape of the current transients in molecularly doped polymers.

We report results of specially planned experiments intended to verify the dispersive character of the charge carrier transport in polycarbonate molecularly doped with hydrazone at 30 wt% loading, using for this purpose samples specifically featuring a well-defined plateau on a linear-linear plot. For this purpose we propose a new variant of the time-of-flight technique which allows easy changing of the generation zone width from about 0.5 µm (surface excitation) through intermediate values to full sample thickness (bulk excitation). To achieve this, we use electron pulses of 3-50 keV energy rather than traditional light pulses provided by lasers. Experimental results corroborated by numerical calculations uniquely prove that carrier transport in this molecularly doped polymer is dispersive, with the dispersion parameter equal to 0.75. Nevertheless, the mobility field dependence follows the famous Poole-Frenkel law.

No abstract

No abstract