Electroluminescence Stability of Organic Light-Emitting Devices Utilizing a Nondoped Pt-Based Emission Layer.

We study the effects of using an emitting material (Pt(II) bis(3-(trifluoromethyl)-5-(2-pyridyl)pyrazolate-Pt(fppz)2) characterized by a preferred horizontal dipole alignment and a nearly unitary quantum yield regardless of concentration on the lifetime of organic light-emitting devices (OLEDs). Using such a material as a dopant in increasingly higher concentrations is found to lead to an increase in device stability, a trend that is different from that commonly observed with conventional OLED guests. The results are consistent with the newly discovered exciton-polaron-induced aggregation degradation mechanism of OLED materials. When this emitter is used as a neat emission layer, the material is already in a highly aggregated state, and the device is no longer affected by exciton-polaron interactions. The results demonstrate the potential stability benefits of using such materials in OLEDs.

Introduction

Organic light-emitting devices (OLEDs) have undergone rapid development since their invention almost 3 decades ago.[1] Owing to their advantages such as self-emission, large color gamut, excellent contrast, and large viewing angle, displays utilizing the efficient phosphorescent OLEDs, or PHOLEDs,[2,3] currently account for the second largest shipment in the mobile display industry, after liquid crystal displays.[4] However, the limited stability of PHOLEDs, especially the blue-emitting OLEDs, continues to be a challenge that prevents their wider adoption in consumer electronics.

A common reason behind the lower stability of blue PHOLEDs has been identified as the fast bond cleavage of the host materials used in these devices.[5,6] In comparison to those used in their green or red counterparts, host materials used in blue devices generally have a wider band gap, making the energy of excitons formed on them necessarily higher and hence increase the probability of bond cleavage. In addition to chemical deterioration of the host materials, we have recently discovered a new degradation mechanism[7] that appears to play a major role in the limited stability of PHOLEDs. When a device is under electrical bias, the interaction of positive polarons and excitons causes the host material to aggregate.[7−9] This aggregation results in a decrease in the emission quantum yield of the material. Moreover, this aggregation leads to phase segregation between the host and the guest, causing a decrease in energy transfer efficiency from the host to the guest and thus a gradual loss in device efficiency over time.[10] The aggregation rate of the host, hence also device lifetime, is found to correlate with the band gap of the host, where hosts with wider band gaps generally degrade faster. It has also been shown in our previous work that this degradation mechanism can be suppressed by reducing the density of excitons[11] and/or positive polarons[12] within the emission layer (EML). Developing materials that are immune to this aggregation mechanism would ultimately be a more effective approach. In this regard, it becomes interesting to see whether the use of emitter materials that are already aggregated may provide any advantage. In this work, we study the effects of using one such material, Pt(II) bis(3-(trifluoromethyl)-5-(2-pyridyl)pyrazolate) (Pt(fppz)2),[13] on device lifetime.

Pt(fppz)2 is a platinum-based phosphorescent orange/red emitter. The molecular structure is shown in Figure . Unlike most phosphorescent materials, Pt(fppz)2 does not exhibit concentration quenching, and its nearly unitary emission quantum yield is preserved even when used in a neat film.[14] This feature allows Pt(fppz)2 to be used as a neat EML in OLED. In 2014, Wang et al. demonstrated that an OLED with such an EML can exhibit an external quantum efficiency (EQE) as high as 31%.[14] The underlying reason for such a high efficiency is attributed to a 93% preferred horizontal dipole ratio of Pt(fppz)2 when used in a neat film.[15] Because of such high order in molecular packing, it has been shown that evaporated neat Pt(fppz)2 films form almost perfect single crystals based on glazing-incident wide-angle X-ray diffraction experiments.[15] Given this high level of crystallinity, it becomes interesting to see whether the material may be immune to further aggregation by exciton–polaron interactions. In this case, an OLED utilizing a neat Pt(fppz)2 EML may demonstrate longer lifetime than devices with Pt(fppz)2 used as a dopant.

Figure 1

Molecular structures of organic materials used.

Results and Discussion

To study the effects of using Pt(fppz)2 as a dopant versus as the sole material in a neat EML on device lifetime, we fabricate devices with the following architecture: indium tin oxide (ITO)/MoO3 (5 nm)/4,4′-bis(carbazol-9-yl)biphenyl (CBP) (48 nm)/CBP:Pt(fppz)2 (x%) (30 nm)/2,2′,2″-(1,3,5-benzinetriyl)-tris(1-phenyl-1-H-benzimidazole) (TPBi) (35 nm)/LiF (1 nm)/Al (80 nm), where x = 5, 20, 60, and 100 (shown in Figure a). Figure b presents the current efficiency versus luminance characteristics of these devices. It should be noted that the device with a neat Pt(fppz)2 EML is less efficient than the 60% doped one. Because the photoluminescence (PL) quantum yields of the 60% doped film and the neat film are both around 95%,[14] the lower efficiency of the device with a neat Pt(fppz)2 EML is likely due to a lower carrier mobility of the EML compared to that of the 60% doped EML, as can be seen in Figure S1a, and thus a less optimal charge balance factor. The external quantum efficiency (EQE) of these devices is presented in Figure S1b. We would like to point out that our devices have slightly lower EQE values than those of the previously reported ones.[15] The lower efficiency is likely a result of a lower charge balance factor due to the use of different transport layers. Nonetheless, the use of CBP and TPBi in our devices significantly improves their stability (the lifetime comparison between our devices and that reported by Wang et al. is shown in Figure S2). Because the main focus of this work is to investigate device stability, the choice of using more electrically stable charge transport materials such as CBP and TPBi is more appropriate. Figure presents the changes in device electroluminescence (EL) with time in these devices. The devices are subjected to a constant current of 20 mA/cm2. The change in EL is presented in the form of normalized luminance, that is, luminance/initial luminance. As can be seen clearly, the device lifetime increases monotonically as the concentration of Pt(fppz)2 increases. It is important to point out that this phenomenon is very different from that of the devices using doped iridium-based emitters, where the device lifetime would peak at around 5% emitter concentration.[16] In iridium-based devices, the main degradation mechanism shifts from exciton–polaron-induced aggregation of the host (at low emitter concentrations) to a degradation in the emission quantum yield of the emitter (at high emitter concentrations). A trade-off between these two phenomena, and hence maximum device stability, therefore occurs at intermediate concentrations (∼5%).

Figure 2

(a) Device structure. (b) Current efficiency versus luminance characteristics of devices with Pt(fppz)2 as a dopant or as the sole material in a neat EML.

Figure 3

Changes in EL intensity with time for devices with 5, 20, 60, and 100% Pt(fppz)2, under 20 mA/cm2 current density.

To understand why the device lifetime increases with emitter concentration and to investigate the effect of using already aggregated emitters (i.e., 100% Pt(fppz)2) on device EL stability, we study device EL spectra and their evolution with time as the device ages. Figure a presents the normalized EL spectra for the device with 5% Pt(fppz)2 at various time intervals. The change in absolute EL is shown in the inset of Figure a. It can be seen that the peak of the EL spectrum for the device with 5% Pt(fppz)2 shifts to longer wavelengths over time. The red shift suggests that the intermolecular Pt···Pt distance is decreasing, pointing to an increase in molecular aggregation.[14] One can see that this device also has the lowest stability. Similar results are also observed in devices with 20% Pt(fppz)2 and are shown in Figure b. Here again, a red shift in the EL spectra can be observed. The magnitude of the red shift is however less than in the case of with 5% Pt(fppz)2. Although the device stability is generally limited, it is higher in comparison to the previous case. In contrast, when the Pt(fppz)2 concentration is further increased to 60 or 100%, as shown in Figure c,d, the normalized EL spectra of these devices do not show any spectral shift during device degradation. The changes in CIE coordinates of these devices before and after 16 h of electrical driving are included in Table . Knowing that the intermolecular stacking interaction is much higher in these devices, especially in the neat film where the Pt(fppz)2 molecules are packed in a way similar to a single crystal, one can attribute this effect to the fact that the material is already well aggregated and therefore does not undergo any further aggregation. As a result, it is expected that the EML would no longer be susceptible to exciton–polaron-induced aggregation, which may explain the higher stability of these devices. To better illustrate the changes in molecular packing during device operation, shifts in device EL peak over time for all four devices are presented in Figure .

Figure 4

Normalized EL spectra for the device with (a) 5% Pt(fppz)2, (b) 20% Pt(fppz)2, (c) 60% Pt(fppz)2, and (d) 100% Pt(fppz)2 at various time intervals. (Insets) changes in absolute EL over time.

Table 1

CIE Coordinates of 5, 20, 60, and 100% Pt(fppz)2-Doped Devices before and after 16 h of Electrical Driving

Pt(fppz)2 doping concentration (%)	CIE color coordinates (x,y) of pristine devices	CIE color coordinates (x,y) after 16 h of electrical driving
5	0.410, 0.506	0.416, 0.477
20	0.467, 0.515	0.468, 0.487
60	0.560, 0.438	0.559, 0.438
100	0.600, 0.400	0.599, 0.400

Figure 5

Changes in EL peak wavelength of devices with various Pt(fppz)2 concentrations over time.

Although the device with the neat EML layer is the most stable one, it should be noted that the lifetime of such a device (T50) is still only about 16 h. To examine whether the stability of the hole transport material is the limiting factor, we fabricate a device with the structure ITO/MoO3 (5 nm)/Spiro-CBP (48 nm)/Pt(fppz)2 (30 nm)/TPBi (35 nm)/LiF (1 nm)/Al (80 nm) for comparison. Spiro-CBP is used here because it has been shown to be more stable than CBP.[8]Figure presents the relative changes of device EL (Figure a) and driving voltage (Figure b) over time for these two devices. It can be seen clearly that both devices have very similar EL degradation trends despite the fact that the CBP device has a faster rise in driving voltage. The almost identical stability trends despite the use of different hole transport layers suggests that the stability of the hole transport material is not the limiting factor for the operational stability of our PHOLEDs.

Figure 6

Changes in (a) EL intensity and (b) driving voltage for devices using CBP and Spiro-CBP as the hole transport material.

Seeing that the stability of hole transport material is not the main bottleneck behind the limited stability of the PHOLEDs with the neat Pt(fppz)2 EMLs, we investigate the stability of Pt(fppz)2 under excitation by measuring changes in its relative emission quantum yield as a result of the electrical stress and under UV irradiation. The same neat EML device structure—ITO/MoO3 (5 nm)/CBP (48 nm)/Pt(fppz)2 (30 nm)/TPBi (35 nm)/LiF (1 nm)/Al (80 nm)—is used in this set of experiment. Figure presents the relative photoluminescence (PL) intensity from Pt(fppz)2 before and after the device has been electrically driven @ 20 mA/cm2 for 12 h (I_12 h) and after the device has been illuminated under UV (@ ∼380 nm with a power density of 2.3 mW/cm2) for 26 h (UV_26 h). It is clear that, in both cases, the emission quantum yield decreases after aging, suggesting that the material has limited stability under excitons (i.e., limited photostability).

Figure 7

Relative PL intensity from Pt(fppz)2 before and after the device has been aged under various conditions.

To examine the effect of varying the chemical structure of the emitter, devices with the architecture ITO/MoO3 (5 nm)/CBP (48 nm)/CBP:Pt-Bu or Pt-Me (x%) (30 nm)/TPBi (35 nm)/LiF (1 nm)/Al (80 nm), where x = 5, 20, 60, and 100, are fabricated. (The structural drawings of Pt-Bu and Pt-Me are illustrated in Figure .) The J–V characteristics as well as current efficiency data are presented in Figure S3. The changes in device EL over time, as well as shifts in EL peak for Pt-Bu and Pt-Me devices are presented in Figure (the actual EL data are shown in Figures S4 and S5). Similar to the Pt(fppz)2 devices, it can be seen that both Pt-Bu and Pt-Me devices show no spectral shift when the emitter concentration reaches 60%, suggesting that the devices are no longer susceptible to further aggregation as a result of exciton–polaron interactions. However, the device with Pt-Bu emitter shows the longest device lifetime and, hence, the greatest stability among the three. Again, this possibility is verified by measuring changes in the emission quantum yield of Pt-Bu and Pt-Me emitters by application of electrical stress as well as under UV irradiation (shown in Figure ). The results show that Pt-Bu has the highest photostability, suggesting that the higher chemical stability of this molecule under excitation might be the reason for its higher EL stability in devices. The results demonstrate that better molecular design of these emitters may improve their stability dramatically.

Figure 8

Changes in EL intensity over time for devices with various (a) Pt-Bu and (c) Pt-Me concentrations. Changes in EL peak wavelength of devices with various (b) Pt-Bu and (d) Pt-Me concentrations over time.

Figure 9

Relative PL intensity from (a) Pt-Bu and (b) Pt-Me before and after the device has been aged under various conditions.

Conclusions

In conclusion, we have investigated the EL stability performance of Pt(fppz)2 as a model compound of an emitter that has a preferred horizontal dipole alignment and a nearly unitary emission quantum yield regardless of material concentration. Using such a material as a dopant in increasingly higher concentrations is found to lead to an increase in device stability, a trend that is different from that commonly observed with conventional (disordered) OLED guest materials. The results are consistent with the newly discovered exciton–polaron-induced aggregation degradation mechanism of OLED materials. When this emitter is used as a neat emissive layer, the material is already in a highly aggregated state and thus the device is no longer affected by exciton–polaron interactions. The results demonstrate the potential benefits of these materials for achieving higher stability and more stable color coordinates. Although the photostability of this class of Pt(II) emitters is found to be relatively limited, the results show that altering the molecular structure has the potential to effectively improve the exciton stability of these materials.

Experimental Section

The organic materials used in this study are 4,4′-bis(carbazol-9-yl)biphenyl (CBP), 2,2′,2″-(1,3,5-benzinetriyl)-tris(1-phenyl-1-H-benzimidazole) (TPBi), and Pt(fppz)2. CBP and TPBi were obtained from Luminescence Technology Corp. and used as received without further sublimation. Pt(fppz)2 was synthesized in house. Prior to device fabrication, the ITO-coated glass substrates were cleaned by sonication in acetone and 2-propanol for 5 min each, in respective order. Devices with an active area of 2 × 2 mm2 were then fabricated in an EvoVac system from Angstrom Engineering Inc. All materials were thermally evaporated at a rate of 0.1–2 Å/s at a base pressure of 5 × 10–6 Torr. The devices were kept and measured in a nitrogen atmosphere at all times.