Influence of poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) in polymer LEDs

We investigate the influence of poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) on the optoelectronic properties of polymer light-emitting diodes containing poly(9,9-dioctylfluorene) (PFO). Electromodulation and $IV$ luminance measurements are reported for a series of devices with bare indium tin oxide (ITO) or PEDOT:PSS-coated ITO anodes and Ba or Al cathodes. The ITO/PFO/Al, ITO/PFO/Ba, and ITO/PEDOT:PSS/PFO/Al devices all exhibit conventional field-induced electromodulation behavior, in both forward and reverse bias, consistent with the Stark effect (SE). The ITO/PEDOT:PSS/PFO/Ba devices by contrast exhibit conventional behavior only for applied biases below the flat-band voltage; at higher biases, the field-induced SE features vanish and are replaced by anomalous charge-induced electromodulation features. This anomalous behavior is observed only when PEDOT:PSS is used in conjunction with a strongly electron-injecting cathode such as Ba, and is attributed to the presence of trapped electrons at the PEDOT:PSS-emitter interface, which screen the electric field from the bulk of the device. The enhanced field at the interface increases the rate of field-assisted hole injection into the highest occupied molecular orbital (HOMO) of the PFO, resulting in lower drive voltages and increased electroluminescence efficiencies.

Figure 1

A schematic energy-level diagram for the device materials used in this work.

Figure 2

$IV$ luminance measurements for the (a) ITO/PFO/Al, (b) ITO/PFO/Ba, (c) ITO/PEDOT:PSS/PFO/Al, and (d) ITO/PEDOT:PSS/PFO/Ba devices. Device (d) shows a sharp threshold for current injection at $3\mathrm{V}$ that is not observed in the other devices. This sharp threshold is observed only in the presence of a PEDOT:PSS anode and a low work-function cathode such as Ba or Ca.

Figure 3

The $6\text{−}\mathrm{kHz}$ first-harmonic electromodulation spectra of the (a) ITO/PFO/Al, (b) ITO/PFO/Ba, (c) ITO/PEDOT:PSS/PFO/Al, and (d) ITO/PEDOT:PSS/PFO/Ba devices at a variety of dc biases. The EM spectra of devices (a), (b), and (c) scale approximately linearly with the dc bias but otherwise do not change significantly in shape, consistent with electroabsorption. The reverse bias EM spectra of device (d) are similar to those of the other devices but, in forward bias, the spectra change completely and the EM features are replaced by new absorption features due to injected charge. These anomalous forward bias features are observed only in the presence of a PEDOT:PSS anode and a low work-function cathode such as Ba or Ca.

Figure 4

The dc bias dependence of the $6\text{−}\mathrm{kHz}$ $417\text{−}\mathrm{nm}$ first-harmonic electromodulation response of the (a) ITO/PFO/Al, (b) ITO/PFO/Ba, (c) ITO/PEDOT:PSS/PFO/Al, and (d) ITO/PEDOT:PSS/PFO/Ba devices. The $417\text{−}\mathrm{nm}$ signals vary linearly with dc bias for devices (a)–(c), consistent with electroabsorption. In the case of device (d), the signal varies linearly with dc bias in reverse bias but deviates sharply from linearity in forward bias due to charge-induced modulation. The contribution from charge-induced modulation is eliminated at high modulation frequencies ($20\mathrm{kHz}$, open circles), leaving an electromodulation signal that is due primarily to the Stark effect. This SE-only signal falls linearly to zero at $2.9\mathrm{V}$ (which is the approximate threshold for current injection), and remains at zero for higher biases.