Polymer light-emitting diodes (PLEDs) are attractive for use in large-area displays and lighting panels, but their limited stability under current stress impedes commercialization. In spite of large efforts over the last two decades a fundamental understanding of the degradation mechanisms has not been accomplished. Here we demonstrate that the voltage drift of a PLED driven at constant current is caused by the formation of hole traps, which leads to additional non-radiative recombination between free electrons and trapped holes. The observed trap formation rate is consistent with exciton-free hole interactions as the main mechanism behind PLED degradation, enabling us to unify the degradation behaviour of various poly(p-phenylene) derivatives. The knowledge that hole trap formation is the cause of PLED degradation means that we can suppress the negative effect of hole traps on voltage and efficiency by blending the light-emitting polymer with a large-bandgap semiconductor. Owing to trap-dilution these blended PLEDs show unprecedented stability.

Electron and hole transport properties of the organic small molecule N,N 0-Di(1-naphthyl)-N,N 0diphenyl-(1,1 0-biphenyl)-4,4 0-diamine are investigated by space-charge-limited current measurements. The hole transport shows trap-free behavior with a mobility of 2.3 Â 10 À8 m 2 /Vs at vanishing carrier density and electric field. The electron transport, on the other hand, shows heavily trap-limited behavior, which leads to highly unbalanced transport. A trap concentration of 1.3 Â 10 24 m À3 was found by modeling the electron currents, similar to the universal trap concentration found in conjugated polymers. This indicates that electron trapping is a generic property of organic semiconductors, ranging from vacuum-deposited small-molecules to solution-processed conjugated polymers.

In amorphous organic semiconductor devices, electrons and holes are transported through layers of small organic molecules or polymers. The overall performance of the device depends both on the materials and the device configuration. Measuring a single device configuration requires a large effort of synthesizing the molecules and fabricating the device, rendering the search for promising materials in the vast molecular space both non-trivial and time-consuming. This effort could be greatly reduced by computing the device characteristics from first principles. Here we compute transport characteristics of unipolar single-layer devices of prototypical hole and electron transport materials respectively α-NPD and TPBi using a first principles multiscale approach that requires only the molecular constituents and the device geometry. This approach of generating a digital twin of the entire device can be extended to multi-layer stacks and enables computer design of materials and devices to facilitate systematic improvement of organic light-emitting diode (OLED) devices.

Aizawa et al. combined thermally crosslinkable materials and orthogonal solvents to achieve a high-efficiency solutionprocessed white-light emitting small molecule OLED. [10] A different approach using cross-linking was shown by Png et al. by adding fluorophenyl azides that cross-link the alkyl side chains of commercially available polymers, resulting in an insoluble layer. [19] Another proposed option is the use of a liquid buffer layer. Here, a buffer layer is deposited on top of the first layer. The buffer layer does not dissolve in the solvents of the subsequent layer. The second layer then floats on top of the buffer layer. The buffer layer evaporates either during the deposition or during a subsequent baking step. [20,21] Furthermore, it has been shown that for polymers also differences in molecular weight can be used to realize a bilayer structure. Polymers with a high molecular weight take a long time to dissolve in solvents such that a second layer can be processed on top. [22] Recently, He et al. reported that crosslinking the surface using a mixed acetylene and argon plasma makes it possible to resist redissolution in organic solvents. [23] An excellent review of the recent progress of solution-processed multilayer OLEDs is given by So and co-workers. [24] These approaches all have certain disadvantages, for example, multilayer structures based on solvent polarity or cross-linkable units typically rely on elaborate and often cumbersome synthetic strategies. The approaches involving liquid buffer layers or molecular weight differences only allow for a limited number of layers in the final stack. An interesting alternative approach has been published by Zhou et al. [25] A polyfluorene-based hole transport material was blended with two insulating, commercially available cross-linkable materials, ethoxylated (4) bisphenol a dimethacrylate (SR540, Sartomer), and NOA83H (Norland Products), to tune the solubility of the resulting blend layer. Upon cross-linking with UV-light these materials form an insoluble host matrix around the semiconductor that makes the whole blend layer insoluble.In contrast to the approaches mentioned before, in this approach standard organic semiconductors can be used without the requirement for chemical modifications. However, in order to make the layer insoluble in either toluene and chloroform these blends contained nearly 70% of cross-linkable host matrix, which might have a severe effect on the charge transport properties. Nevertheless, an improved performance of a polymer light-emitting diode (PLED) containing such an insoluble hole transport layer was observed. To further validate this approach knowledge on the effect of the amount of matrix on the solubility and the charge transport properties of the Poly[2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylenevinylene] (MEH-PPV) is blended with two different inert UV-cross-linkable matrices to tune the solubility of the solution-processed films. It is found that only 10 wt% of theses matrices is required to make the blend layer insoluble...