Simulation of organic light-emitting diodes


OLEDs (organic light-emitting diodes) have the potential to revolutionize general lighting and display technology.
Like their inorganic counterparts, they are way more efficient than classic light bulbs and, in contrast to gas-discharge lamps, can be dimmed and switched on instantaneously. Their unique property is the possibility of large-area deposition, which offers new opportunities for illumination concepts.
Displays based on OLEDs are extremely thin, show extraordinarily high levels of contrast and present colors independently of the viewing angle. Furthermore, they are flexible, which already shows in the existence of curved displays and will soon lead to the availability of foldable electronic devices.

schematic structure of a typical OLED Figure 1

Figure 1 shows the schematic structure of a typical OLED with reflecting cathode and transparent anode on a transparent substrate.
The high efficiency of state-of-the-art OLEDs can only be achieved through complex layer structures consisting of a large number of materials with specific functions. Different charge carrier mobilities and energy levels lead to confinement of the recombination zone in one or several emission layers, selected dopant materials in the emission layers enhance efficiency and enable the choice of the desired color.


Due to the wide range of available materials and the innumerable stack design possibilities resulting fom variations in thicknesses and compositions and sequences of layers, it is virtually impossible to determine the optimum device structure experimentally. This is, however, the way OLEDs are optimized right now, since reliable physical models are missing. So there is a strong demand for simulation tools which can support the development of devices as it has been the case in the classic semiconductor industry for many decades now.
The difficulties in achieving this goal arise from the fact that amorphous organic materials are fundamentally different from crystalline inorganic semiconductors and that their investigation is still in a relatively early state. Inorganic semiconductors consist of covalently bound atoms, which form a regular lattice. The carriers which contribute to current flow are delocalized in this material class. Organic semiconductors, on the other hand, consist of molecules which only interact via the relatively weak Van-der-Vaals force. Thus, a carrier is always localized on a specific molecule, and current flow is only possible through thermally activated tunnelling processes between molecules. Even the mathematical description of such a single transition process is still a matter of debate; two different expressions are currently used to describe it.
Many more questions arise from the amorphous structure of organic materials, i.e., from the fact that, as indicated in figure 2, the molecules do not form a spatially ordered lattice and that their energy levels are statistically distributed within a certain range. The theoretical knowledge about the influence of this spatial and energetic disorder on carrier transport is not even remotely complete yet.

  DOS, electron transport Figure 2

Figure 2: On the left, Gaussian densities of states for electrons (lowest unoccupied molecular orbitals – LUMOs) and for holes (highest occupied molecular orbitals – HOMOs) are shown. The rightern side illustrates electron transport along molecules statistically distributed with respect to position and energy level (indicated by color).

In order to account for the stochastic nature of carrier transport, Monte Carlo simulations or the more efficient master equation have to be used on a three-dimensional lattice. For device modelling, one of these methods can be used in combination with a solution of the Poisson equation to determine carrier and field distribution.
Alternatively, Monte Carlo simulations or a master equation approach can be used to determine the average carrier mobility as a function of electric field strength and carrier concentrations. Once a suitable empirical expression for this function has been found, it can be used in a very efficient one-dimensional device simulation based on self-consistent solution of the Poisson equation and the drift-diffusion equation.

Possible topics for bachelor or master theses consist in adjusting established numerical methods from semiconductor simulation to the drift-diffusion equation including mobility models for organic materials. For that purpose, simulation programs have to be developed in the language Fortran or C++. Programming experience is not required, but interest in physical and numerical problems are indispensable for this work.
Any thesis will be thoroughly supervised to enable a successful completion.




+49 241 80 23902