Realization of Red Iridium-Based Ionic Transition Metal Complex Light

Jun 4, 2019 - In both cases, the device shows a response time below 80 μs (limited ... 0.2%, 0.22 cd A–1, and 0.19 lm W–1 for a current density o...
13 downloads 0 Views 3MB Size
Download PDF
Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces 2019, 11, 22612−22620

Realization of Red Iridium-Based Ionic Transition Metal Complex Light-Emitting Electrochemical Cells (iTMC-LECs) by InterfaceInduced Color Shift Julia Frohleiks,†,‡,∥ Svenja Wepfer,†,‡,∥ Gerd Bacher,*,‡ and Ekaterina Nannen*,†,‡,§

Downloaded via WESTERN UNIV on August 4, 2019 at 04:30:16 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Research Group “Solid State Lighting”, NanoEnergieTechnikZentrum and ‡Werkstoffe der Elektrotechnik and CENIDE, University of Duisburg-Essen, 47057 Duisburg, Germany § Faculty of Electrical Engineering and Computer Science, Textile Innovatory, University of Applied Sciences Niederrhein, 47805 Krefeld, Germany S Supporting Information *

ABSTRACT: Red ionic iridium-based transition metal complex light-emitting electrochemical cells (iTMC-LECs) with emission centered at ca. 650 nm, maximum efficiency of 0.3%, maximum brightness above 650 cd m−2, and device lifetime well above 200 and 33 h at brightness levels of 10 and 210 cd m−2, respectively, are realized by the introduction of a p-type polymer interface to the standard design of [Ir(ppy)2(pbpy)]+[PF6]− (Hppy = 2-phenylpyridine, pbpy = 6-phenyl-2,2′-bipyridine) iTMC-LEC. The unexpected color shift from yellow to red is studied in detail with respect to operation conditions and material combination. The experimental data suggest that either exciplex formation or subordinate, usually suppressed optical transitions of the iTMC might become activated by the introduced interface, causing the pronounced red shift of the peak emission wavelength. KEYWORDS: iTMC, light-emitting electrochemical cell, iridium, red emission, interface effect orange12 light-emitting ionic transition metal complexes (λEL 580−610 nm), yielding devices with extrapolated lifetime above 650 and 2800 h at a maximum luminance above 1600 and 1000 cd m−2 and maximum efficiency of 16 and 3.5 cd A−1, respectively. At the same time, a lot of research efforts are undertaken to realize red light-emitting complexes.13−16 In particular, red and deep-red emitting complexes defined as those with a peak position of the electroluminescence spectrum in the range of 620−700 nm and position within the CIE 1931 color space of x from 0.52 to 0.72 and y from 0.28 to 0.34,17,18 which are highly desired in red signal color-related applications, suffer from a considerable trade-off in brightness/efficiency and stability/turn-on time.19 The devices with the best performance in brightness as well as external quantum efficiency (EQE) show luminance levels as high as 7500 cd m−2 and peak efficiencies >2% EQE,20−22 whereby the lifetime is limited to several hours20−22 and not necessarily reached the points of maximum luminance and efficiency. Only the recent development of a series of new cyclometalated iridium complexes yielded long-term stable red iTMC-LECs with extrapolated lifetimes >6000 h at 2% EQE and initial luminance of 200 cd m−2 (maximum luminance >1250 cd m−2).23

1. INTRODUCTION Light-emitting electrochemical cells (LECs) gain increasing interest due to their easy solution-based fabrication, mechanical flexibility, and large-area planar light-emission nature.1−3 The main difference to organic light-emitting diodes (OLEDs) is the incorporation of mobile ionic species that facilitate the charge carrier injection from the electrodes into the active materials due to the formation of electric double layers (EDLs) at the interface and subsequent electrochemical doping when a voltage is applied.4−6 Therefore, supporting layers for charge injection and transport as well as electrodes that are sensitive to atmosphere and moisture can be omitted, which allows processing under ambient air and much reduced encapsulation methods.7 In particular, LECs based on ionic transition metal complexes (iTMCs) are highly promising since they harvest both singlet and triplet excitons yielding intrinsically higher quantum efficiencies and higher brightness above 1000 cd m−2 during endurance tests compared to polymer-based LECs.2,7−9 Combining the most relevant benchmarks for lighting applications, such as high brightness and efficiency, with long-term stability for emitter molecules of different colors, is currently the main challenge for the market entry of iTMCbased LECs. In particular, the intramolecular π−π stacking approach is known to yield the most bright and long-term stable iTMC emitters with LEC lifetimes of more than 4000 h at initial brightness above 650 cd m−2 and maximum efficiency of 1.5 lm W−1, emitting yellow light.8 The multiple π-stacking approach could also be applied to green10,11 and reddish © 2019 American Chemical Society

Received: April 22, 2019 Accepted: June 4, 2019 Published: June 4, 2019 22612

DOI: 10.1021/acsami.9b07019 ACS Appl. Mater. Interfaces 2019, 11, 22612−22620

Research Article

ACS Applied Materials & Interfaces In this contribution, we present an alternative approach to shift the emission color of the Ir-based iTMC-LEC into the red spectral range. A traditional LEC architecture of LECs based on the yellow iTMC [Ir(ppy)2(pbpy)]+[PF6]−24−26 (Hppy = 2-phenylpyridine, pbpy = 6-phenyl-2,2′-bipyridine) emitter was accomplished with an additional organic hole injection and electron blocking layer (HIL/EBL) yielding hybrid LEC devices (HyLECs) with a pronounced color shift from yellow to red. The initial emitter is currently one of the most promising iTMCs for LECs in terms of stability at high brightness levels and efficiency.8 The novel red hybrid iTMCLECs emit in the range of 620−650 nm with a peak brightness above 650 cd m−2, sub-second turn-on times and a lifetime greater than 30 h at an initial brightness of 210 cd m−2, preserving the stability of the initial emitting complex. The electro- and photoluminescence experiments carried out on multiple devices with different HIL materials suggest an interfacial red emission zone stemming from exciplex formation and/or from activation of usually subordinate optical transitions of the iTMC as the possible reason for the observed color shift.

2. RESULTS The standard LEC design typically consists of an indium tin oxide (ITO)-covered glass substrate overcoated with poly(3,4ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) as a transparent cathode, followed by the active LEC layer and an evaporated aluminum anode. Figure 1a shows the schematic device concept for the hybrid LEC (HyLEC), which contains an additional layer of poly[N,N′bis(butylphenyl)-N,N-bis(phenyl)-benzidine] (poly-TPD, Figure 1c), complementing the standard LEC design (see the Experimental Section for information on device fabrication). Smooth layers were fabricated via spin-coating (Figure 1b), showing a homogeneous blue emission for the poly-TPD and a yellow emission for LECs under UV light excitation. Photoluminescence (PL) measurements (Figure 1e) reveal typical main intensity peaks of the poly-TPD at 425 and 450 nm in addition to subordinate transitions in the blue-green spectral range.27 The iTMC PL spectrum is centered at 585 nm with a full-width half-maximum of 106 nm, which is typical for emitters based on charge transfer transitions.28 In the case of the HyLEC, we observe a superposition of blue poly-TPD and yellow iTMC emission in the PL measurements (Figure 1e). At first sight, similar emission color should be expected for the HyLEC under electrical excitation, since a poly-TPD layer is also known to actively contribute to the overall electroluminescence spectrum of various light-emitting devices.29−34 At the same time, also other thin films of organic35,36 and inorganic37 materials could be used in stacked hybrid LEC architectures to tune the emission color36,37 and to achieve even white emission35,36 by the superposition of luminescence spectra. In our case, surprisingly, the additional poly-TPD layer leads to a considerable red shift of the emission spectrum and the devices show large-area deep-red emission, as can be seen from the images of LEC and HyLEC devices in the inset of Figure 1f. The electroluminescence (EL) spectrum of the reference LEC, located at 585 nm, as well as the EL spectrum of the HyLEC at the equivalent voltage of 5 V are shown in Figure 1f. The emission spectrum of the HyLEC is now located at λEL = 645 nm and is thus red-shifted approximately

Figure 1. (a) Schematic device design for HyLEC devices containing an additional poly-TPD layer in addition to the standard LEC design. (b) Spin-coated poly-TPD layer as well as fabricated LEC and HyLEC devices under UV light excitation. (c) Chemical structure of poly-TPD. (d) Chemical structure of the light-emitting complex [Ir(ppy)2(pbpy)]+. (e) Normalized PL spectra of a poly-TPD layer, an iTMC layer, and HyLEC layers. (f) Normalized LEC and HyLEC EL spectra at 5 V. The inset shows images of emitting LEC and HyLEC devices at an applied voltage of 5 V.

by 60 nm (ΔE ∼ 200 meV) compared to the original LEC emission at the same applied voltage. The unexpected red-shift of the emission color in HyLECs was further studied as a function of the applied external voltage to gain more insight into its origin. The results are shown in detail in Figure 2. Figure 2a displays the normalized EL spectra for the HyLEC between 4 and 6 V compared to the EL spectrum of the reference LEC at 5 V. A continuous blue shift of the emission peak with increasing operation voltage of in total 83 meV was observed (λEL = 660−626 nm), never reaching the yellow emission peak of the LEC (λEL = 585 nm) though. When the voltage was further increased (Figure 2c, 7− 9 V), the spectrum shifts back to the red emission region. Moreover, the same trend in the dependence of the emission color on time was observed. At a constant applied voltage of 6.5 V (Figure 2c), the emission peak shifts back to deep red 22613

DOI: 10.1021/acsami.9b07019 ACS Appl. Mater. Interfaces 2019, 11, 22612−22620

Research Article

ACS Applied Materials & Interfaces

Figure 3. (a) Normalized luminance versus time for the HyLEC biased under pulsed current operation (2 and 20 mA cm−2, 1 kHz, 50% duty cycle). (b) Corresponding normalized efficiency and RMS voltages versus time. The inset in (a) summarizes the maximum values for the luminance (Lv,max), the EQE (EQEmax), the current efficacy (CEmax), and luminous efficacy (ηmax) for the respective pulsed current density.

efficiency are normalized to the respective maximum. The corresponding maximum values are included in the inset of Figure 3a. In both cases, the device shows a response time below 80 μs (limited by the measurement setup). At 20 mA cm−2, the HyLEC reaches the maximum luminance of 210 cd m−2 after 1 h of operation. The HyLEC exhibits a lifetime (defined as luminance decay time down to 50% of the maximum value, t1/2) greater than 33 h preserving the stability of the initial iTMC complex (see Supporting Information Figure S1). The increase in brightness with time strongly depends on the formation of n- and p-doped regions and thus on the applied current density. The gentle operation at lower current densities as expected increases the device lifetime, resulting in a lifetime far above 50 h at 2 mA cm−2 pulsed current operation. However, due to the delayed formation of the doped regions at lower current densities, the HyLEC exhibits a slower increase in brightness, reaching the maximum value after 45 h of operation. The efficiency curves show equivalent dependence on the applied pulsed current density, reaching peak efficiencies of 0.2%, 0.22 cd A−1, and 0.19 lm W−1 for a current density of 20 mA cm−2. The corresponding root mean square operation voltage (RMS voltage, Figure 3b) drops during the first minutes of operation in both cases and remains constant (3.4 V for 20 mA cm−2, 3.1 V for 2 mA cm−2). In particular, for deep-red iTMC-LECs, this is a promising combination of lifetime, luminance, and efficiency. Except for the pronounced shift of the emission color, HyLECs showed regular LEC behavior in terms of device dynamics with time. In the next step, it is important to know if HyLECs also preserve the typical current−voltage−luminance (IVL) behavior of the LECs38−40 and how the additional polyTPD layer possibly impacts the device performance. Due to the incorporated ions, LECs are known to show the dynamic behavior of the current and luminance or radiance as a

Figure 2. (a) Normalized EL spectra for HyLEC at 4, 5, and 6 V applied voltages compared to the EL of the reference LEC at 5 V. (b) Normalized EL spectra for HyLEC at 7, 8, and 9 V applied voltages compared to the LEC EL at 5 V. (c) Normalized EL spectra for HyLEC at 6.5 V for two different operation times (t2 = t1 + 1 min, t1 approx. 3 min) compared to the LEC EL at 5 V.

(660 nm) by 96 meV within 1 min of operation time. The pronounced redshift is thus dependent on the applied voltage and also on time, but still observable under all operating conditions. Please note that the reference LEC based on the standard [Ir(ppy)2(pbpy)]+[PF6]− emitter is spectrally stable within the wide range of operation voltages and with time.37,38 In particular, for red iTMC-LECs, the combination of high luminance and good efficiency with long-term stability remains challenging. Therefore, the stability of the HyLEC device at high as well as low luminance levels was examined. Figure 3 shows the luminance and efficiency behavior as a function of operation time for the HyLEC device at pulsed current operation (1 kHz, 50% duty cycle) for two applied current densities (20 and 2 mA cm−2). For better comparison of the stability for high luminance (at 20 mA cm−2) as well as lower luminance (at 2 mA cm−2) levels, the intensity as well as the 22614

DOI: 10.1021/acsami.9b07019 ACS Appl. Mater. Interfaces 2019, 11, 22612−22620

Research Article

ACS Applied Materials & Interfaces

increases up to 3.8 W sr−1 m−2 (1483 cd m−2) and 0.24% EQE (0.62 cd A−1) at peak voltages of 7 and 6.4 V, respectively. The HyLEC shows a typical LEC behavior as well, with a pronounced shift of the turn-on voltage (from 3 V down to 2 V) and increase in radiance and efficiency over multiple orders of magnitude upon conditioning, when comparing the performance of the device at the same operation voltage. The maximum values increased as well from 0.37 W sr−1 m−2 and 65 cd m−2 for the pristine device to 3.9 W sr−1 m−2 and 690 cd m−2 for the conditioned device and from 0.15% EQE (0.16 cd A−1) to 0.31% EQE (0.34 cd A−1). The comparison between the IVR-behavior of the HyLECs and the reference LEC delivers, however, some unexpected features. In all cases, the HyLEC shows higher emission intensity and in particular higher efficiency than the reference LEC in the low-voltage range, indicating improved electron− hole-balance within the active light-emitting region. At the same time, the performance of both devices in the high voltage range is more or less comparable. Both devices reach a maximum radiance of around 4 W sr−1 m−2, indicating no major loss in radiative excitons in the case of the HyLEC. Another striking feature is the reduced turn-on voltage of the HyLEC compared to the reference LEC. Although the shift of the turn-on voltage by ca. 0.2 eV in the conditioned state can be rationalized by the shift of the peak emission energy from 2.12 eV to ca. 1.9 eV, the significant shift of the turn-on voltage in the pristine state is at first sight unexpected. This finding is also in agreement with the pulsed current operation mode, revealing as well significantly lower operation voltage of the HyLEC device than the standard reference [Ir(ppy)2(pbpy)]+[PF6]− LEC (3.4 vs 4.6 V RMS voltage) under identical operation conditions.38 Based on the transient as well as voltage-dependent results, it can thus be concluded that the fabricated HyLEC devices can be attributed to the class of LEC devices with both formation of EDLs and doping fronts by mobile ions being crucial for efficient device operation and resulting in a typical dynamic behavior of the device with operation time.

function of voltage and time, which makes measurements of the current−voltage−luminance/radiance (IVL/IVR) behavior challenging.39,40 Therefore, ultra-fast IVR-scans (1.7 V s−1 scanning rate) on both the pristine and conditioned states of the devices were performed,38 so that the impact of the ionic movement can be subdued as much as possible.39 First, pristine devices were measured without any previous biasing or pretreatments to retrieve “pristine” IVR-curves (i.e., negligible electric double layer (EDL) formation and doping). Afterward, the devices were preconditioned with a defined and slowly increasing current density (reaching 0.25 A cm−2 within 2 min) to reach the steady state of the LEC where the EDLs and the n- and p-doped regions are already formed. Finally, IVR-scans were performed again at the same measurement conditions as in the case of the pristine devices to retrieve the “conditioned” curves of the device in its efficient operation state. For better comparison of the emission intensity behavior of both devices without the impact of the emission color, radiance [W sr−1 m−2] as a radiometric unit was used to describe emission intensity of both devices instead of the more common photometric unit luminance [cd m−2]. Figure 4 compares the radiance as well as the resulting EQE for the reference LEC and the HyLEC as a function of the applied voltage.

3. DISCUSSION 3.1. Generality of the Approach. The origin of the unexpected red shift in the HyLEC device architecture might have several reasons. The time-dependence as well as the IVRbehavior of the HyLEC devices already suggest that the red emission can not only be reasoned by the poly-TPD layer but has to be related to the iTMC and the polymer/iTMC interface. To exclude that the red emission is mainly a result of poly-TPD-specific internal optical transitions and/or a chemical reaction caused by the poly-TPD layer or its solvent residues that somehow alters the iTMC, we performed additional experiments with a poly(9-vinylcarbazole) (PVK) HIL layer instead of the poly-TPD. The new HyLEC device also shows typical LEC behavior, as can be seen from the pristine as well as conditioned radiance curves depicted in Figure 5a. Although the HyLEC with a PVK HIL layer does not reach the same performance as the HyLEC with a polyTPD HIL layer in the conditioned state (0.8 W sr−1 m−2 instead of 2.4 W sr−1 m−2 at 7 V), it shows comparable features in the pristine state as for example increased brightness, especially in the low-voltage regime and an even more reduced turn-on voltage compared to the reference LEC in the pristine state.

Figure 4. (a) Radiance as well as (b) EQE for pristine and conditioned LEC and HyLEC as a function of the applied voltage. The insets summarize the maximum values for both devices (Rmax and EQEmax) supplemented by the corresponding maximum luminance (Lv,max) and maximum current efficacy (CEmax).

The pristine reference LEC shows a turn-on voltage (defined here as the one to reach 10−5 W sr−1 m−2, corresponding to 0.005 cd m−2) of above 3.5 V, followed by an increase in radiance and external quantum efficiency (EQE), reaching peak values of 2.1 W sr−1 m−2 (840 cd m−2) and 0.2% EQE (0.5 cd A−1) at a peak voltage of 8 V. After the conditioning procedure, the performance of the reference LEC significantly improves due to the establishment of efficient charge carrier injection conditions and the formation of doping regions. The turn-on voltage reduces down to 2.2 V, the maximum radiance 22615

DOI: 10.1021/acsami.9b07019 ACS Appl. Mater. Interfaces 2019, 11, 22612−22620

Research Article

ACS Applied Materials & Interfaces

intensity in the low-voltage regime can therefore be attributed to the increased amount of electrons leading to an improved electron−hole-balance within the device especially in the pristine state under the assumption of the emitter quantum yield remaining the same. Please note that the injection conditions for the electrons at the cathode and likewise the emitter quantum yield remain initially the same in both reference and HyLEC. This is in agreement with Figure 4 demonstrating the improved EQE and higher emission intensity only in the low-voltage operation regime. At high operation voltages and in the conditioned state, HyLECs show more or less the same or even a slightly reduced amount of emitted photons (indicative of the amount of injected and transported electrons) as the reference LEC. The assumption of the additional efficient recombination zone at the interface emitting light at lower (red-shifted) energies therefore explains the overall reduction of the turn-on/operation voltages. 3.3. Origin of the Voltage- and Time-Dependent Color Shift. It is known that [Ir(ppy)2(pbpy)]+[PF6]− iTMC is not significantly changing its emission profile with time or at different operation voltages.38 On the other hand, its intensity is known to change both with time and operation voltage first increasing due to EDL and doping front formation and afterward decreasing at elevated voltages and proceeding operation time due to exciton quenching and degradation.7,8,38 Also, as known from the literature7,24,38 and also evident from our experiments, the [Ir(ppy)2(pbpy)]+[PF6]− iTMC layer is neither bright nor efficient in the low-voltage range