Contribution of Top and Bottom Devices to Efficiency in a Tandem QLED

ContributorZhengqi ZHANG

Regular Tandem Quantum Dot Light-Emitting Diodes with over 51% External Quantum Efficiency for Next-Generation Displays Dawei Yang; Yiduo Wang; Jing Xie; Daocheng Pan; Bingsuo Zou; Heng Zhang*

Highlights

In this study, novel regular tandem QLEDs are demonstrated with an ultrathin (≈4 nm) indium tin oxide (ITO) charge generation layer (CGL). Through systematic optimization of charge injection balance and light out-coupling efficiency, the resulting tandem QLEDs can exhibit a record-breaking EQE of 51.2%, which represent the highest performance metrics reported to date for regular tandem QLEDs.

Figure 1. Device structure and working mechanism of regular tandem QLEDs. (a) Schematic device structure of regular tandem QLEDs. The inset on the right shows chemical structures of 2PACz and 6PA. (b) Cross-sectional TEM images and corresponding color-mixed EDS elemental mapping image of regular tandem QLEDs. (c) Working mechanism of regular tandem QLEDs.

Figure 2. Calculated power fraction of (i) bottom device, (j) top device, and (k) tandem device.

Preparation

Install and activate Leda according to the Installation and Licensing guide.

The authors provided the refractive index data of the materials in the Supporting Information.

The refractive indices of each layer in the tandem QLED structure at the target wavelength of 632 nm are as follows:

• ITO: 1.73+0.01i
• PEDOT:PSS: 1.48
• TFB: 1.72
• QDs: 1.95
• ZnMgO: 1.59
• Al₂O₃: 1.37+7.52i
• 2PACz:6PA: 1.72
• Al: 1.37+7.52i

The device structure in the paper is: glass/ITO anode (150 nm)/PEDOT:PSS (40 nm)/TFB-up (30 nm)/QDs-up (26 nm)/ZnMgO-up (130 nm)/Al:Al₂O₃ (2 nm)/ITO-CGL (4 nm)/2PACz:6PA (5 nm)/TFB-down (40 nm)/QDs-down (35 nm)/ZnMgO-down (50 nm)/Al cathode (110 nm)

The data is for this case study only. If you use this data in a real project, please cite: Yang et al., 2025

Regular Tandem Device

First, build the regular tandem QLED structure according to Figure 1(a), importing the thickness and refractive index data provided by the authors for each layer. Note that this structure emits light from the ITO anode side. When building the model, the emitting side should face upwards, which is the reverse of the order shown in Figure 1(a). The ITO anode is at the top, and the Al cathode is at the bottom.

Activate the Emis. property for both QDs-up and QDs-down. Since the substrate is 1 mm thick glass, activate the Inco. property for the glass layer.

Since we are only investigating the device performance at the target wavelength of 632 nm, set the spectral type of both emitters to Unit White in the Emitter configuration page.

Enable the Mode Detector, set Wavelength Type to Single, and the wavelength to 632 nm. Increase the Error Tolerance to 20 to allow for some deviation.

In Sweep, add a thickness scan for the ZnMgO-up layer, with a range from 22 nm to 200 nm and a step size of 2 nm.

Reason for setting Error Tolerance to 20 and starting the From value not at 0 nm

When the From value of the scan is too small, the distance between the dipole and the metal layer (i.e., the thickness of the transport layer) is too small. This causes the intensity of the Evanescent (SPP) Mode to be excessively high, leading to inaccurate calculation results, and Leda will stop the calculation and report an error.

This is manifested as a Mode detector error: The result fails to converge in a finite in-plane wave vector

Solution:

• Increase the From value of the scan (increase the transport layer thickness).
• Increase the Error Tolerance of the Mode detector (i.e., tell Leda to allow results with a larger error margin).

Click Run Sweep to obtain the scan results of the mode distribution of the regular tandem device as a function of the ZnMgO-up thickness. The mode distribution result is similar to the fraction of power in Figure 2(k).

The difference between TOC in Leda and OCE in the paper

In the original paper, the theoretically predicted peak OCE (54.6%) of the regular tandem device is obtained by summing the OCE of the top device (32.2%) and the bottom device (22.4%). Figure 2(k) shows that the total value for the tandem device is 2, which is 200%.

Original paper’s description:

As shown in Figure 2i–k, the power fraction of the top device, bottom device, and tandem device was calculated, and the results demonstrate that the optimal thickness for ZnMgO is 120 nm, where the tandem device achieves a peak light out-coupling efficiency of 54.6%. It can also be seen that this excellent light out-coupling efficiency comprises individual contributions of 22.4% from the bottom device and 32.2% from the top device.

However, the TOC obtained from the Leda simulation is the average OCE of the device, which adheres to the definition of OCE (the number of photons entering free space from the emitting surface / the number of photons emitted by the emitting layer). Therefore, the value is about half of that in the original paper.

Top Device

According to Figure 1(a), the top device consists of the ITO-CGL, 2PACz:6PA, TFB-down, QDs-down, ZnMgO-down, and Al cathode layers.

Based on the tandem device structure built above, disable the Emis. property of QDs-up and activate the Emis. property of QDs-down.

The difference of structure orientation

The device structure shown in Figure 1(a) is inverted compared to the structure in Leda. This is because Leda requires the light-emitting surface to face upwards for modeling. Since the original structure is bottom-emitting, the layer order is reversed in Leda.

Therefore, to simulate the top device from the paper, we activate the Emis. property of QDs-down and disable the Emis. property of QDs-up.

Click Run Sweep to obtain the scan results of the mode distribution for the top device as a function of the ZnMgO-up thickness. The mode distribution result is similar to the fraction of power in Figure 2(j).

Bottom Device

According to Figure 1(a), the bottom device consists of the ITO-anode, PEDOT:PSS, TFB-up, QDs-up, ZnMgO-up, and Al:Al₂O₃ layers.

Based on the constructed device, disable the Emis. property of QDs-down and activate the Emis. property of QDs-up.

Click Run Sweep to obtain the sweep results of the mode distribution for the bottom device as a function of the ZnMgO-up thickness. The mode distribution result is similar to the fraction of power in Figure 2(i).

Summary

This case study demonstrates the optimization of a regular tandem QLED structure to achieve balanced charge injection and a high EQE. Optical simulations were used to investigate the contributions of the top and bottom devices to the overall OCE of the tandem structure.

The simulation results from Leda are generally consistent with the original paper. Some minor discrepancies in the specific values may exist due to potential differences in the data and algorithms used.