Please enable JavaScript to view this site.

Application Gallery

Navigation: OLEDs > Patterned OLED

Simple 2D OLED

Scroll Prev Top Next More

cover_picture_oled_2d_small_zoom15This example has been updated. Find the latest version at OLED (2D).






This section shows how multi-layer OLED with PC patterning can be simulated with FDTD using the methodology described in OLED Simulation methodology



Associated files



See also

OLED Simulation methodology

Related publications

J. Hauss et al., Periodic Nanostructures Fabricated by Laser Interference Lithography for Guided Mode Extraction in OLEDs, SOLED (2010)


We gratefully acknowledge the collaboration of Horst Greiner of Philips Research in the development of this application example.



Simulation Setup

The screenshot above shows the device drawn up in the Layout editor as in simulation file OLED_2D.fsp. The thin blue layer is the aluminum cathode layer, and the light gray top layer is the glass substrate. It is through this glass layer where the normal emitted power is measured. In a real experiment, the light would eventually pass through the glass into air, but that is not considered in this example. Here, we use the model analysis group to adjust the simulation span, the PC period (a), as well as the orientation and location of the dipoles.


A screenshot of the Model analysis group

A screenshot of the Model analysis group


For this simple 2D OLED example, we consider two dipole positions: aligned with a hole (x=0) and between two holes (x=a/2).




For the patterned case, this leads to six simulations, since each dipole position requires three dipole orientations (x/y/z). For the unpatterned case (ie. setting the "add_PC" variable in the model analysis group to "0"), only three simulations are required since dipole position in the plane has no effect. The process of sweeping through the different dipole locations and orientations (and summing the result from each incoherently) can be automated using parameter sweep projects:




For the patterned case, the outer sweep changes the dipole location, whereas the inner sweep changes the dipole orientation. For the unpatterned case, we only have one sweep that changes the dipole orientation. The script OLED_sweepresults.lsf can be used to run the sweeps and obtain the results shown below.


Results and Discussion

Extraction efficiency analysis

The far_field_change_index analysis group uses FDTD' Far field projection functions to calculate the far field angular distribution. This analysis object also accounts for reflection and refraction that would occur at a far field Glass-Air interface. Once the parameter sweeps are complete, we can analyze the farfield results averaged over all the dipole orientations and locations. For example, we can calculate the ratio between the patterned and unpatterned OLED for the power transmitted into a 5 degree cone normal to the surface. The associated analysis script file uses the farfield2dintegrate command to integrate the far field intensity over the 5 degree cone.



oled_2d_extraction_farfield_nopc_zoom63 oled_2d_extraction_farfield_pc_zoom63

The above plots show that extraction efficiency is highly wavelength dependent, and one can achieve significant enhancement at particular wavelengths with the use of PC patterning. Therefore, by setting the operating wavelengths of the OLED at these wavelengths, one can potentially achieve extraction efficiencies much higher than that of the conventional OLED. To analyze the OLED for different operating wavelengths, one can simply integrate the results over the desired emission spectrum (see 3D OLED with square symmetry for an example). These results are fairly similar to the ones shown in Figure 2 of the Hauss reference provided above, although for a different OLED structure.


Radiative decay rate enhancement

The dipole_power Analysis group is constructed with 4 frequency domain power monitors (6 for 3D) enclosing the dipole source. The dipole_power result returns the power radiated by the dipole, normalized to the amount of power the dipole would radiate in a homogeneous material. This quantity is calculated in two ways: from the dipolepower script command and from a box of 4 monitors surrounding the dipole.



                                                 = T(from trans_box) = dipolepower / sourcepower



Since parameter sweep projects can access the results of analysis groups, one can compute the average power radiated from all the dipole locations and orientations by setting the "Results" section of the parameter sweep project as shown:




Below is the dipole power for the patterned and unpatterned OLED, as calculated by the dipolepower script function and the box of monitors method:


The figure shows two things:

a) The two techniques give very similar results

b) The dipole power has a much more complicated shape when the PC structure is present.


Tip: Dipole sources in dispersive media

The dipolepower script function should not be used when the dipole is located in a dispersive media. For example, if the active layer alq3 object is changed from being a simple dielectric with index 1.68 to using the dispersive alq3 material mode in the material database, the dipolepower script function and box of monitor technique will give very different results, as shown below. In such cases, the box of monitor technique is more reliable.



Fraction of emitted power

Below is a fraction of the power emitted by OLED, cumulative in different layers. In this example, about 60% of the emitted light is absorbed in the OLED layers, about 10-15% is trapped in glass, and only 20-25% is able to reach air.



Tip: Slow far field projections

You may notice that the time required for the far field projections is often on the same order as the time to run the actual FDTD simulations. The time required for far field projections can be substantially reduced by recording fewer frequency points, using spatial downsampling, and by lowering the projection resolution.

In many cases, using the spatial downsampling option (Geometry tab of the monitor) will give a significant speed up with very little loss of accuracy, as long as the nyquist limit is observed (at least two spatial points per the shortest wavelength). If the mesh size is 15 nm and the wavelength is 400nm, then it might be possible to set the downsampling as large as 10 or 13 (400/15/2 = 13.333).

Copyright Lumerical Inc. | Privacy | Site Map