Periodic Reporting for period 3 - ULTRA-LUX (Ultra-Bright Thin-Film Light Emitting Devices and Lasers)
Reporting period: 2022-10-01 to 2024-03-31
It is the objective of this project to demonstrate brightness levels in thin film LEDs beyond state of the art and ultimately achieve an electrically pumped laser in the same.
Light sources are the basis of any photonic device and high brightness is a prerequisite for high accuracy application, e.g. of sensing, spectroscopy, or ranging in medicine, communication, automotive, or aerospace.
While light emitting devices composed of crystalline III-V semiconductors are used as high brightness light sources since decades, the maximum achievable current densities in thin film devices with organic, quantum dot, or perovskite based active media exhibit more than 100 times lower values. On the counter side, thin film materials are highly customizable active materials for light emission, where the wide range of composition variability allows for a tuning of photonic parameters. Thin film stacks can be produced on arbitrary surfaces as glass, plastic, or on top of electronic devices. This versatility opens numerous options for integration and custom designs and the utilization of massively parallel arrays.
Our target power density is P>30W/cm2 for a thin-film light source, by reaching an IQE>75% (EQE~15%) at J≤100A/cm2. This challenging target will be pursued by innovating simultaneously on new device architectures and perovskite materials as emission layers.
We are going to design and optimize photonic resonators based on our learnings from perovskite light emitting diodes. Ultralow optical losses in the resonators and controlled and clean quantum-confinement in the perovskite emission layer are targeted to reduce the lasing threshold to J≤100A/cm2.
Light sources are the basis of any photonic device and high brightness is a prerequisite for high accuracy application, e.g. of sensing, spectroscopy, or ranging in medicine, communication, automotive, or aerospace.
While light emitting devices composed of crystalline III-V semiconductors are used as high brightness light sources since decades, the maximum achievable current densities in thin film devices with organic, quantum dot, or perovskite based active media exhibit more than 100 times lower values. On the counter side, thin film materials are highly customizable active materials for light emission, where the wide range of composition variability allows for a tuning of photonic parameters. Thin film stacks can be produced on arbitrary surfaces as glass, plastic, or on top of electronic devices. This versatility opens numerous options for integration and custom designs and the utilization of massively parallel arrays.
Our target power density is P>30W/cm2 for a thin-film light source, by reaching an IQE>75% (EQE~15%) at J≤100A/cm2. This challenging target will be pursued by innovating simultaneously on new device architectures and perovskite materials as emission layers.
We are going to design and optimize photonic resonators based on our learnings from perovskite light emitting diodes. Ultralow optical losses in the resonators and controlled and clean quantum-confinement in the perovskite emission layer are targeted to reduce the lasing threshold to J≤100A/cm2.
The work is distributed over three work packages: (1) process development and device fabrication; (2) fundamental study of perovskite electro-optical performance and device physics; (3) optical design and characterization of perovskite films and photonic devices including stimulated emission. With a view on the development of an electrically pumped perovskite laser, the key development indicators were and are an increase of the external quantum efficiency, operation at extremely high current densities (> 1kA/cm2), and low laser and amplified spontaneous emission thresholds under optical pumping.
Initially, we have developed strategies to accurately extract device parameters from PeLEDs that are driven beyond the standard current densities. From perovskite solar cells it was known that the measurement protocol has an impact on the temporal response and the derived performance parameters. It was therefore a goal, to identify the underlying effects that influence the temporal response with a view on high current density driving in the future. The current density–voltage–radiance/luminance (J–V–R/L), and the EQE–J characteristics were chosen as the main figures of merit and the well-known perovskite methylammonium lead iodide (MAPI) was selected as the active layer in this study. The PeLED devices have been exposed to maximum bias voltages of 5.5V resulting in maximum current densities of ~ 500mA/cm2 while analysing the transient behaviour on a timescale from several 100µs up to 1000s. We have been able to determine ion migration, Joule heating, and irreversible device degradation as the main processes that influence the transient current density and electroluminescence as well as leading to hysteresis effects in the same.
As a conclusion of the peer-reviewed paper, we have suggested several guidelines to the community to make results between institutions comparable. A special emphasis was paid to outline the importance of the transient data in the interpretation of the results. Furthermore, we have demonstrated the advantage of pulsed voltage biasing to reduce the impact of hysteresis on the data extraction.
In a next step, we have focused on reducing the driving voltage of perovskite LEDs, so that devices under the same current densities are exposed to a reduced power loss and thus a lower internal heat generation. With this reduce Joule heating, degradation processes as observed before and linked to high internal temperatures are suppressed and long-term stability is considerably improved. The key to the voltage reduction was the selection of PCBM/ZnMgO over an organic stack as the electron transport layer system, which led to device stability improvement from 3min to over 4h at 100mA/cm2 current densities for MAPI. The concept has been applied to high performance perovskite active material systems of formamidinium based (FAPb3)0.95(MAPbBr3)0.05 which exhibited a maximum EQE of 11.4% at 330mA/cm2. With this system, a T50 stability of >500h at 50mA/cm2 has been demonstrated in the same article [DOI: 10.1002/adom.202100586].
With the learnings from the previous work, that heat generation in combination with low thermal conductivity of perovskite are limiting factors on the route to electrically pumped laser diodes, scaling of the active area was the selected device development track. Shrinking the device area would allow to achieve a small pumping volume, such that the lasing-level current density injection could be realized at a low absolute current passing through the device. Device diameters of 1000µm down to 50µm have been systematically studied under different pulsing regimes to quantify the thermal impact on device performance for glass and sapphire substrates. The voltage bias was varied from quasi-DC pulsing down to 250ns long pulses. For the smallest devices (50µm), current densities beyond 5kA/cm2 have been reached. In line with expectations, we could prove that scaling devices and ultras-short pulsing is reducing but not eliminating the detrimental impact of Joule heating. Using a PeLED with a diameter of 50µm on sapphire substrate, we achieve a high quasi-DC radiance of 2059W/m2sr, half-life time above 10 h under continuous operation at 500 mA/cm2, as well as pulsed EQE-J product value of 6.7 A/cm2 at a current density of 4.2 kA/cm2.
To further reduce the impact of Joule heating, the operation of PeLEDs under intense electrical pulsing in cryogenic conditions was investigated. At 77K we could demonstrate high EQEs of up to 2.9% at 1kA/cm2 current density for 250ns pulses. Furthermore, we were able to identify pre-biasing schemes during for optimal device performance by controlling the ion distribution in cryogenic conditions [DOI: 10.1002/adom.202200024]
In further works, we have studied the charge transfer and energy migration in quasi-2D perovskites and provided evidence that at low T (15K) charge carrier transfer is dominating over energy migration [DOI: 10.1002/adfm.202010076].
Alternating cations in the interlayer space have been investigated in PeLEDs as an alternative 2D perovskite fabrication method, with a stability and EQE improvement with increasing interlayer order number [DOI: 10.1515/nanoph-2021-0037].
As an alternative to solution processing of the active layers in PeLEDs and lasers, we have started to use vacuum deposition to improve parameter control during the film preparation. In a first demonstration, the concept was applied to perovskite-based photodetectors and solar cells where the promising device performance indicate the potential of this technique also for light emitting devices.
All the work on electrically driven devices is accompanied by optically pumped analyses of thin film devices and device stacks to investigate the properties of amplified spontaneous emission (ASE). In parallel resonators are developed for the generation of lasers. The results of this topic are currently in preparation for peer-reviewed publication.
We have published to date 10 peer-reviewed articles and presented 6 talks in virtual conferences (MRS, SPIE, SID).
Initially, we have developed strategies to accurately extract device parameters from PeLEDs that are driven beyond the standard current densities. From perovskite solar cells it was known that the measurement protocol has an impact on the temporal response and the derived performance parameters. It was therefore a goal, to identify the underlying effects that influence the temporal response with a view on high current density driving in the future. The current density–voltage–radiance/luminance (J–V–R/L), and the EQE–J characteristics were chosen as the main figures of merit and the well-known perovskite methylammonium lead iodide (MAPI) was selected as the active layer in this study. The PeLED devices have been exposed to maximum bias voltages of 5.5V resulting in maximum current densities of ~ 500mA/cm2 while analysing the transient behaviour on a timescale from several 100µs up to 1000s. We have been able to determine ion migration, Joule heating, and irreversible device degradation as the main processes that influence the transient current density and electroluminescence as well as leading to hysteresis effects in the same.
As a conclusion of the peer-reviewed paper, we have suggested several guidelines to the community to make results between institutions comparable. A special emphasis was paid to outline the importance of the transient data in the interpretation of the results. Furthermore, we have demonstrated the advantage of pulsed voltage biasing to reduce the impact of hysteresis on the data extraction.
In a next step, we have focused on reducing the driving voltage of perovskite LEDs, so that devices under the same current densities are exposed to a reduced power loss and thus a lower internal heat generation. With this reduce Joule heating, degradation processes as observed before and linked to high internal temperatures are suppressed and long-term stability is considerably improved. The key to the voltage reduction was the selection of PCBM/ZnMgO over an organic stack as the electron transport layer system, which led to device stability improvement from 3min to over 4h at 100mA/cm2 current densities for MAPI. The concept has been applied to high performance perovskite active material systems of formamidinium based (FAPb3)0.95(MAPbBr3)0.05 which exhibited a maximum EQE of 11.4% at 330mA/cm2. With this system, a T50 stability of >500h at 50mA/cm2 has been demonstrated in the same article [DOI: 10.1002/adom.202100586].
With the learnings from the previous work, that heat generation in combination with low thermal conductivity of perovskite are limiting factors on the route to electrically pumped laser diodes, scaling of the active area was the selected device development track. Shrinking the device area would allow to achieve a small pumping volume, such that the lasing-level current density injection could be realized at a low absolute current passing through the device. Device diameters of 1000µm down to 50µm have been systematically studied under different pulsing regimes to quantify the thermal impact on device performance for glass and sapphire substrates. The voltage bias was varied from quasi-DC pulsing down to 250ns long pulses. For the smallest devices (50µm), current densities beyond 5kA/cm2 have been reached. In line with expectations, we could prove that scaling devices and ultras-short pulsing is reducing but not eliminating the detrimental impact of Joule heating. Using a PeLED with a diameter of 50µm on sapphire substrate, we achieve a high quasi-DC radiance of 2059W/m2sr, half-life time above 10 h under continuous operation at 500 mA/cm2, as well as pulsed EQE-J product value of 6.7 A/cm2 at a current density of 4.2 kA/cm2.
To further reduce the impact of Joule heating, the operation of PeLEDs under intense electrical pulsing in cryogenic conditions was investigated. At 77K we could demonstrate high EQEs of up to 2.9% at 1kA/cm2 current density for 250ns pulses. Furthermore, we were able to identify pre-biasing schemes during for optimal device performance by controlling the ion distribution in cryogenic conditions [DOI: 10.1002/adom.202200024]
In further works, we have studied the charge transfer and energy migration in quasi-2D perovskites and provided evidence that at low T (15K) charge carrier transfer is dominating over energy migration [DOI: 10.1002/adfm.202010076].
Alternating cations in the interlayer space have been investigated in PeLEDs as an alternative 2D perovskite fabrication method, with a stability and EQE improvement with increasing interlayer order number [DOI: 10.1515/nanoph-2021-0037].
As an alternative to solution processing of the active layers in PeLEDs and lasers, we have started to use vacuum deposition to improve parameter control during the film preparation. In a first demonstration, the concept was applied to perovskite-based photodetectors and solar cells where the promising device performance indicate the potential of this technique also for light emitting devices.
All the work on electrically driven devices is accompanied by optically pumped analyses of thin film devices and device stacks to investigate the properties of amplified spontaneous emission (ASE). In parallel resonators are developed for the generation of lasers. The results of this topic are currently in preparation for peer-reviewed publication.
We have published to date 10 peer-reviewed articles and presented 6 talks in virtual conferences (MRS, SPIE, SID).
As outlined in the previous section, we have pushed the limits in device operation beyond the state of the art concerning high current density pulsing – a prerequisite for the goal of an electrically pumped laser. In subsequent development steps, we have investigated the transient behaviour of perovskite LEDs, first at low current densities and ambient conditions, later in scaled devices for very high current densities under cryogenic conditions. At all development stages we have set new references concerning EQE and stability. With our recent work on sub-µs pulsing at cryogenic temperatures, we have demonstrated the operation of perovskite LEDs at current densities close to the inversion threshold.
We therefore expect to gain even more insights into the device behavior at ultrahigh current densities bringing us even closer to thin film injection lasers.
We therefore expect to gain even more insights into the device behavior at ultrahigh current densities bringing us even closer to thin film injection lasers.