Studies of the degradation mechanism of organic light-emitting diodes based on tris (8-quinolinolate)aluminum Alq and 2-tert-butyl-9,10-di(2-naphthyl)anthracene TBADN

Abstract

Previously, radical cation of tris(8-quinolinolate)aluminum (Alq•+) has been associated with the instability of Alq films subjected to holes-only electrical current. Yet, the questions remain (i) whether Alq•+ is the primary source of the intrinsic degradation of bipolar organic light-emitting diodes (OLEDs) based on Alq, (ii) whether Alq•+ reactions result in deep charge traps in holes-only devices as found in bipolar counterparts, and (iii) whether radical cations can be a common source of degradation of OLEDs irrespective of materials. With regards to generality of hole-current-related degradation, it is interesting to examine the behavior of 9,10-diarylanthracenes (DAAs)—the practically important class of blue-fluorescing light-emitting-layer hosts. These questions prompted our comparative study of the effects of unipolar currents in Alq and 2-t-butyl-9,10-di(2-naphthyl)anthracene (TBADN), which was chosen as a representative material of the DAA class. First, we identified device structures allowing for rigorous and stable unipolar conduction. Interestingly, even in pristine holes-only devices, our voltammetric measurements indicated that Alq contains a substantial density of deep hole traps (far deeper than what can be explained by energetic disorder), which can be charged by passing holes-only current and seemingly discharged by exposure to white light. As for aged holes-only Alq devices, they exhibited symptoms qualitatively matching those of aged bipolar Alq devices, viz., photoluminescence (PL) loss, transition voltage (V0) rise, and drive voltage (Vd) rise. Notably, PL and V0 are linearly correlated in both holes-only and bipolar devices, which reinforces the supposed link between Alq•+ and the degradation in both types of devices. Yet, there are indications the Alq•+ instability may not be the only degradation pathway in bipolar devices. Even though our observations for holes-only Alq devices agree qualitatively with previously reported ones, we observe far slower degradation rates [Alq PL fades up to ∼500 times slower in holes-only devices, while Alq electroluminescence (EL) fades ∼50 times slower in bipolar control devices]. It is possible that impurities play a significant, perhaps crucial role in the degradation mechanism of both bipolar and holes-only devices, especially the relatively shorter-lived ones. In sharp contrast to Alq, all three observables (PL, V0, and Vd) indicate that holes-only current in TBADN (neat or doped with a perylene-based blue dopant) does not result in degradation in the time that is sufficient for the corresponding bipolar control devices to lose 60%–80% of EL and 20%–30% of PL. We find that the electrons-only current in Alq or TBADN does not result in degradation either. Thus, the degradation of Alq and DAA bipolar devices may be caused by fundamentally dissimilar mechanisms: while hole current may damage the former, it does not appear to affect the latter, suggesting that the initiation step is different.