Cloaking Contacts on Large‐Area Organic Light‐Emitting Diodes

Abstract

In this paper, we suggest a feasible way to homogenize the OLED light emission, i.e., to cloak the contacts or to make them "invisible" in emission. This task is schematically illustrated in Figure 1 . At fi rst sight, this task may appear the same as that of cloaking contact grids on the sunfacing side of solar cells. [ 8 ] These devices for solar cells operate in the regime of ballistic light propagation. [ 8 ] In sharp contrast, the cloaking devices for OLEDs described in the present paper operate in the distinct regime of multiple light scattering.We start in Section 2 by discussing two different blueprints based on the light-diffusion equation. The fi rst blueprint is based on coordinate transformations, leading to piecewise homogeneous and anisotropic diffusivities, the second blueprint uses piecewise homogeneous and isotropic diffusivities. The latter is postoptimized in Section 3 for large light throughput by using Monte-Carlo simulations, which can cover the regime between diffusive and ballistic light transport. Finally, the outcome is validated by model experiments in Section 4. Numerical Calculations Based on Light-Diffusion TheoryFigure 2 a illustrates the geometry of the problem considered in this paper. A homogeneous layer (yellow) emits light from its surface according to a Lambertian angular distribution. We assume that this layer is thin compared to all other dimensions.Metallic contact grids on top of large-area organic light-emitting diodes (OLEDs) are important for laterally homogenized current spreading leading to a lower series resistance and reduced large-scale brightness gradients. However, the contacts are also visible by unwanted shadows, i.e., by spatially inhomogeneous Lambertian light emission. In this paper, inspired by recent work on diffuse-light core-shell invisibility cloaking, a practical tailored distribution of light scattering centers is designed on top of an OLED to homogenize its light emission. Numerical solutions of the light-diffusion equation as well as Monte Carlo ray-tracing simulations are presented. The blueprint is verifi ed by experiments on a scaled-up model structure. The contacts can be made "invisible" in emission while maintaining large light throughput.

Invisibility cloaking in light‐scattering media

“…This untransformed structure is illustrated in Fig. (a) . While such a layer “smears out” the shadow cast by the contact wire, only a very thick layer would lead to truly homogeneous emission.…”

“…Alternatively, it is also possible to obtain a perfectly homogeneous intensity distribution by replacing the anisotropic tensorial diffusivity by a constant isotropic scalar diffusivity D 1 according to This approach is in close analogy to the core–shell cloaking as presented above.…”

“…Therefore, we start with a simple homogeneous diffusive layer (diffusivity D 0 ) of height h on top of the OLED's light-emitting area (with an assumed reflectivity of 100%), which is interrupted by contacts of width d and spacing L. This untransformed structure is illustrated in Fig. 15(a) [97]. While such a layer "smears out" the shadow cast by the contact wire, only a very thick layer would lead to truly homogeneous emission.…”

“…Alternatively, it is also possible to obtain a perfectly homogeneous intensity distribution by replacing the anisotropic tensorial diffusivity ↔ D by a constant isotropic scalar diffusivity D 1 according to [97]…”

Photon management in solution-processed organic light-emitting diodes: a review of light outcoupling micro- and nanostructures

“…However, in case of a large variation in the local brightness, as the one introduced by shadowing metal grids used for charge carrier transport, more advanced concepts are required to obtain a spatially homogeneous emission. 30 In order to test their potential for external outcoupling, those MLAs were integrated in white OLEDs (WOLEDs) by using a single layer of the white-emitting co-polymer SPW087 by Merck KGaA, and the measured efficiencies (defined in terms of luminous flux normalized to the electrical power consumption) were compared with the ones obtained without the MLA on the same devices. 29 The results shown in Fig.…”