Perylene Based Solution Processed Single Layer WOLED with Adjustable CCT and CRI

In solution processed single layer white organic light emitting diode (WOLED) applications, the choice of host matrix and optimization of dopant levels represent two crucial parameters to consider. In this work, poly(N-vinylcarbazole) (PVK): 2-(4-Biphenylyl)-5-phenyl-1,3,4-oxadiazole (PBD) and PVK:1,3-bis[(4-tert-butylphenyl)-1,3,4-oxadiazolyl] phenylene (OXD-7) matrices are used as hosts for perylene based devices. PVK:PBD presented better compatibility and lower turn-on voltages compared to PVK:OXD-7. Benefiting from the exciplex emission observed at 630 nm, a color rendering index (CRI) value of 90 is reached with the device containing PVK:PBD as the host and 0.1 wt.% of an orange emitting perylene derivative, i.e., PDI. Introduction of the perylene based green emitter, i.e., PTE, in this emitting layer not only caused a fading in the exciplex emission, but also resulted in disappearance of the electroplex peak at 535 nm, which is detected between PVK:PBD and PTE in bare PTE containing devices. Full visible range coverage is achieved by optimizing the PDI:PTE ratio. WOLED containing PVK:PBD:0.06 wt.% PDI:0.03 wt.% PTE presented high CRI (≥95) and adjustable correlated color temperatures (CCT, 3800 K-5100 K).


Introduction
Solution processed white organic light emitting diode (WOLED) applications are attracting much attention especially with their easy production process and consequently relatively cost-effective natures [1,2]. It is well known that the quality of the light obtained strongly depends on the balanced electron and hole injection into the emissive layer, exciton formation and its radiative recombination [3]. Utilization of materials with appropriate energy levels, electron and/or hole mobilities and device architecture enhances the electronhole balance and therefore the exciton formation efficiency. Although triplet sensitizers have the potential of reaching 100% internal quantum efficiency, they suffer from their long exciton diffusion lengths (<100 nm) compared to that of singlet excitons (<20 nm) which retard the radiative recombination step, increase the probability of losses and therefore need more complicated device architectures [4][5][6][7]. Recently, singlet emitters, free from those kinds of disadvantages, are getting more and more attention in WOLED applications [8,9].

Methods
UV-Vis absorption and photoluminescence (PL) measurements were performed by using Edinburgh Instruments FS5 spectrophotometer. ITO substrate cleaning was processed by ISOLAB ultrasonic bath and CUTE FC-10046 O 2 plasma system. All solutions were coated via Laurell WS-400B-6NPP LITE spin coater under atmospheric conditions. Film thickness values were determined by KlaTencor MicroXM-100 optical profilometer. Cathode evaporations were performed by the use of a vacuum thermal evaporator attached to a LC Technology Solution Inc. glove box system. The EL spectra and luminance-voltagecurrent density curves were obtained using a Keithley 2400 source measurement unit and calibrated Hamamatsu C9920-12 measurement system. Frequency varying capacitance values were determined up to 10 6 Hz by using an Agilent 4284A LCR meter.

Results and Discussion
This section may be divided by subheadings. It should provide a concise and precise description of the experimental results and their interpretation, as well as the experimental conclusions that can be drawn.

Characterization and Determination of Host Matrix
PVK is widely used as p-type host material due to its good film forming properties and compatible energy levels [26,27]. However, because of the low lying HOMO level of PDI (Figure 1), bare PVK cannot be considered an ideal host [20]. Therefore, in order to enhance electronic level alignment and also increase the electron density in the emissive layer, two dopants with different electron mobility values are also introduced in the PVK layer. These materials are PBD and OXD-7 with electron mobility values of 8 × 10 −8 cm 2 V −1 s −1 and 1-4 × 10 −5 cm 2 V −1 s −1 , respectively [33]. Energy levels of the materials are depicted in Figure 1 together with the open structures of PDI and PTE.
Electronics 2021, 10, x FOR PEER REVIEW 3 of 12 spin coated on top of PEDOT:PSS at 3000 rpm for 1 min and annealed at 125 °C for 30 min. Finally, LiF (1 nm) and Al (100 nm) cathodes were deposited through a shadow mask and by the use of the vacuum evaporator at 10 −6 mbar and deposition rates of LiF and Al were 0.2 Å/s and 2.5 Å/s, respectively. The active area of the devices was 7 mm 2 and six parallel devices were fabricated.

Results and Discussion
This section may be divided by subheadings. It should provide a concise and precise description of the experimental results and their interpretation, as well as the experimental conclusions that can be drawn.

Characterization and Determination of Host Matrix
PVK is widely used as p-type host material due to its good film forming properties and compatible energy levels [26,27]. However, because of the low lying HOMO level of PDI (Figure 1), bare PVK cannot be considered an ideal host [20]. Therefore, in order to enhance electronic level alignment and also increase the electron density in the emissive layer, two dopants with different electron mobility values are also introduced in the PVK layer. These materials are PBD and OXD-7 with electron mobility values of 8 × 10 −8 cm 2 V −1 s −1 and 1-4 × 10 −5 cm 2 V −1 s −1 , respectively [33]. Energy levels of the materials are depicted in Figure 1 together with the open structures of PDI and PTE. PVK:OXD-7 host presented higher turn-on voltages (Vturn-on) and, lower current density and luminance values (Figure 2a,b) than those of PVK:PBD host in bare PDI containing devices. These results may be attributed to the deeper HOMO and LUMO energy levels of OXD-7, which may act as trap center for the injected holes and may cause unnecessarily excess electron injection in to the emission layer, respectively. Hole mobility values of PVK:PBD and PVK:OXD-7 hosts are calculated to be 2 × 10 −6 and 8 × 10 −7 cm 2 V −1 s −1 , whereas the electron mobility values are 8.7 × 10 −7 and 1.1 × 10 −5 cm 2 V −1 s −1 , respectively PVK:OXD-7 host presented higher turn-on voltages (V turn-on ) and, lower current density and luminance values (Figure 2a,b) than those of PVK:PBD host in bare PDI containing devices. These results may be attributed to the deeper HOMO and LUMO energy levels of OXD-7, which may act as trap center for the injected holes and may cause unnecessarily excess electron injection in to the emission layer, respectively. Hole mobility values of PVK:PBD and PVK:OXD-7 hosts are calculated to be 2 × 10 −6 and 8 × 10 −7 cm 2 V −1 s −1 , whereas the electron mobility values are 8.7 × 10 −7 and 1.1 × 10 −5 cm 2 V −1 s −1 , respectively ( Figure S3). The dramatic difference obtained between the electron mobility values confirm the above made suggestion on the lower luminance values, but the difference in hole mobility values is incapable of addressing the deeper HOMO level of OXD-7 as the main reason for higher V turn-on . The AFM micrographs of PVK:PBD (RMS of 0.31) and PVK:OXD-7 (RMS of 0.44) films presented homogenous and smooth surfaces ( Figure S4). However, dramatic thickness difference between the host media was also detected; the thicknesses of PVK:OXD-7 and PVK:PBD films are 126 ± 4.3 nm and 84 ± 2.0 nm, respectively (Table S1). Therefore, we suggest that in addition to the deeper HOMO energy level, a higher thickness of PVK:OXD-7 layer may have caused charge accumulation on both sides of the active layer and resulted in an increased V turn-on for the PVK:OXD-7 based OLED [36].
coordinates of (0.36, 0.29), CCT of 4170 K and CRI of 90 could be reached with 0.1 wt.% PDI doping. Further increment of PDI wt.% caused a reduction in white light properties as a result of decreasing of both the blue and the orange-red emission intensities ( Figure  S5a).  The V turn-on values are increased with the increase in PDI doping wt.% for both of the host media and this increment is more notable with PVK:OXD-7 host (Figure 2b). The explanation given above is also pertinent for the devices those of which PDI doping wt.% is increased. Presence of 0.1 wt.% PDI in PVK:OXD-7 caused approximately 10 nm of thickness increment whereas for PVK:PBD:PDI (0.1 wt.%) this increment is only 5 nm compared to their un-doped films. The thicknesses of PVK:OXD-7: PDI (0.1 wt.%) and PVK:PBD:PDI (0.1 wt.%) films are 135 ± 4.9 nm and 89 ± 3.4 nm, respectively (Table S1). This huge thickness difference is attributed to the different molecular volumes of OXD-7 (378 Å 3 ) and PBD (270 Å 3 ) (molecules are drawn in the chemsketch software and their volumes are calculated by using an online tool [37]). The V turn-on value increments obtained can also be explained by the high dipole moments and dipole moment differences of the hosts and PDI molecule, which may create energy disorder and consequently reduce the carrier mobility and conductivity of the emitting layer [38].
Due to the comparatively better luminance and lower V turn-on values obtained, PVK:PBD was chosen as the host matrix. The white light properties of the device structure of ITO/PEDOT:PSS/PVK:PBD:PDI (x wt.%)/LiF/Al are summarized in Table 1. The CIE coordinates of (0.36, 0.29), CCT of 4170 K and CRI of 90 could be reached with 0.1 wt.% PDI doping. Further increment of PDI wt.% caused a reduction in white light properties as a result of decreasing of both the blue and the orange-red emission intensities ( Figure S5a). The EL maximum of PVK:PBD host presented 15 nm of red shift compared to its PL as a result of exciplex emission [39]. The orange-red region of the EL was mainly dominated Electronics 2021, 10, 725 5 of 12 by a peak at 610 nm ( Figure 3a). Generally, this peak has been observed in PL spectrum of bare PDI films and attributed to aggregation of the PDI core due to π-π stacking [23]. However, 0.1 wt.% of PDI content corresponds approximately 5 × 10 −5 M of PDI concentration and aggregation induced PL behavior might not be expected. Therefore, PL measurements are carried on in order to further understand the mechanism and origin of this peak. PL spectra of PVK:PBD:(0.1 wt.%) PDI film presented the same emission wavelength maxima (λ 1 ems = 551 nm, λ 2 ems = 598 nm and λ 3 ems = 650 nm) with the excitation wavelengths of 510 nm and 310 nm which are used for the excitation of the PDI and host, respectively (Figure 3a). Although they preserved characteristic three peak decreasing intensity structure of PDI, λems maxima and PL offset are red shifted approximately 20 nm and 90 nm, respectively, compared to the solution phase ([10 −6 ] in chloroform) and polystyrene (PS): 0.1 wt.% PDI film ( Figure S5b). These shifts are evaluated as possible formation of another excited state. Interestingly, in the EL spectrum, λ1ems of PDI was the same while λ 2 ems of it also presented approximately 15 nm of red shift and gave a peak at 610 nm. It is deduced that the EL obtained at 610 nm was a result of an excited/ground state energy or charge transfer interaction between the host and PDI. In applied voltage dependent EL measurements [40][41][42] arising of a new peak at 630 nm is observed and intensities of both the blue and the orange-red regions are affected equally from the voltage increments ( Figure 3b). Therefore, it is suggested that the EL at 610 nm was actually the combination of λ 2 ems of PDI and exciplex emission with a peak point at 630 nm. This suggestion is supported by the EL intensity increments observed at 630 nm with the increase in PDI doping wt.% in PVK:PBD host system ( Figure S5a, inset). The EL maximum of PVK:PBD host presented 15 nm of red shift compared to its PL as a result of exciplex emission [39]. The orange-red region of the EL was mainly dominated by a peak at 610 nm ( Figure 3a). Generally, this peak has been observed in PL spectrum of bare PDI films and attributed to aggregation of the PDI core due to π-π stacking [23]. However, 0.1 wt.% of PDI content corresponds approximately 5 × 10 −5 M of PDI concentration and aggregation induced PL behavior might not be expected. Therefore, PL measurements are carried on in order to further understand the mechanism and origin of this peak. PL spectra of PVK:PBD:(0.1 wt.%) PDI film presented the same emission wavelength

Optimization of White Color Properties
In order to increase the visible light region coverage and consequently, the CRI value of bare orange-red emitting PDI containing devices, green light emitting PTE is thought to be introduced in the emission layer. Absorption and emission spectra of PVK:PBD, PDI

Optimization of White Color Properties
In order to increase the visible light region coverage and consequently, the CRI value of bare orange-red emitting PDI containing devices, green light emitting PTE is thought to be introduced in the emission layer. Absorption and emission spectra of PVK:PBD, PDI and PTE are provided in Figure S2b. Beforehand, the behavior of ITO/PEDOT:PSS/PVK:PBD:PTE (y wt.%)/LiF/Al device is monitored. In a similar manner with the bare PDI devices, the V turn-on values increased with the increasing doping ratio of PTE (Figure 4a) but the maximum luminance value produced by the bare PTE containing device was slightly higher than that of the bare PDI containing one (Figure 4a vs. Figure 2b). This is attributed to more efficient interaction between the host and PTE due to bigger integral area (overlap) between the absorption of PTE and emission of PVK:PBD compared to that of PDI ( Figure S2b). It is well known that in host:guest systems the emission is usually controlled by energy or electron transfer between them [43]. Therefore, in order to determine the measure of transfer, steady state quenching experiments are performed. PDI and PTE are used as quenchers for the emission of PVK:PBD host and quenching rate constants of 9 × 10 10 M −1 s −1 and 3.6 × 10 11 M −1 s −1 , respectively, are extracted from the corresponding Stern-Volmer plots (Figure 4b). These rate constants are higher than the diffusion control limit (k q ≈ 10 10 M −1 s −1 ) and suggests the probability of charge transfer process between the carbazole based host and perylene derivatives [44].
FOR PEER REVIEW 6 of 12 area (overlap) between the absorption of PTE and emission of PVK:PBD compared to that of PDI ( Figure S2b). It is well known that in host:guest systems the emission is usually controlled by energy or electron transfer between them [43]. Therefore, in order to determine the measure of transfer, steady state quenching experiments are performed. PDI and PTE are used as quenchers for the emission of PVK:PBD host and quenching rate constants of 9 × 10 10 M −1 s −1 and 3.6 × 10 11 M −1 s −1 , respectively, are extracted from the corresponding Stern-Volmer plots (Figure 4b). These rate constants are higher than the diffusion control limit (kq ≈ 10 10 M −1 s −1 ) and suggests the probability of charge transfer process between the carbazole based host and perylene derivatives [44]. Resulting from a better donor-acceptor relationship between the host and PTE, the ratio needed to obtain white light was much lower than that of the PDI based devices. The devices that contain 0.03, 0.06 and 0.1 wt.% PTE presented greenish white light and the CIE values remained above the Planckian locus, while for the bare PDI devices these values are below it ( Figure 5). Further increment of PTE wt.% did not cause a significant change in the EL intensities of blue and green regions (Figure 6a). The green region of the EL was dominated by a peak at 535 nm which corresponds to a wavelength of 15 nm red shifted second PL peak of PTE in solution phase (λ 2 ems = 520 nm, Figure S2b). The intensity of 535 nm peak is increased with the increase in PTE doping wt.%. Although both the red shift and increasing EL intensity behaviors are similar with the situation in PVK:PBD:PDI devices, it is unlikely that the applied voltage dependent EL spectra presented significant intensity differences between the blue and green regions; the intensity of 535 nm peak increased more than two folds of the 445 nm peak (Figures 6b and S7). Therefore, it is suggested that the EL peak of 535 nm is generated as a result of electroplex formation between the host and PTE [40][41][42]. Resulting from a better donor-acceptor relationship between the host and PTE, the ratio needed to obtain white light was much lower than that of the PDI based devices. The devices that contain 0.03, 0.06 and 0.1 wt.% PTE presented greenish white light and the CIE values remained above the Planckian locus, while for the bare PDI devices these values are below it ( Figure 5). Further increment of PTE wt.% did not cause a significant change in the EL intensities of blue and green regions (Figure 6a). The green region of the EL was dominated by a peak at 535 nm which corresponds to a wavelength of 15 nm red shifted second PL peak of PTE in solution phase (λ 2 ems = 520 nm, Figure S2b). The intensity of 535 nm peak is increased with the increase in PTE doping wt.%. Although both the red shift and increasing EL intensity behaviors are similar with the situation in PVK:PBD:PDI devices, it is unlikely that the applied voltage dependent EL spectra presented significant intensity differences between the blue and green regions; the intensity of 535 nm peak increased more than two folds of the 445 nm peak (Figure 6b and Figure S7). Therefore, it is suggested that the EL peak of 535 nm is generated as a result of electroplex formation between the host and PTE [40][41][42].  The CRI values of PTE containing devices are only around 75 because of the lack of orange-red region and CIE and CCT values of PTE wt.% < 0.1 devices presented the closest values to white light (Table 1). Therefore, these doping ratios of PTE are considered to be  The CRI values of PTE containing devices are only around 75 because of the lack of orange-red region and CIE and CCT values of PTE wt.% < 0.1 devices presented the closest values to white light (Table 1). Therefore, these doping ratios of PTE are considered to be The CRI values of PTE containing devices are only around 75 because of the lack of orange-red region and CIE and CCT values of PTE wt.% < 0.1 devices presented the closest values to white light (Table 1). Therefore, these doping ratios of PTE are considered to be introduced in the 0.1wt.%PDI containing device. Device configuration of is prepared and labeled as Device 1, 2 and 3 in accordance with the increasing y' wt.%. As expected, an increment at the green and a decrement at the blue regions of the EL spectra are detected as the wt.% of PTE is increased (Figure 7a). The fading of 630 nm emission of a bare PDI device and disappearance of the electroplex peak at 535 nm in bare PTE devices are also monitored. The V turn-on values of Device 1-3 are in the range of bare PDI and bare PTE containing devices. Although the current density at the same applied voltage is in the order of Device 1 > Device 2 ∼ = Device 3 (Figure 7b), the EL intensity decrement at 630 nm is continued in Device 2 and Device 3. This situation is evaluated as more evidence of electric field independent formation of 630 nm peak, and the fading is ascribed to the reduction of host concentration ready to form exciplex with PDI as the PTE wt.% is increased. Absence of 535 nm peak is attributed to the wide overlap between the emission of PTE and absorption of PDI ( Figure S2). is prepared and labeled as Device 1, 2 and 3 in accordance with the increasing y' wt.%. As expected, an increment at the green and a decrement at the blue regions of the EL spectra are detected as the wt.% of PTE is increased (Figure 7a). The fading of 630 nm emission of a bare PDI device and disappearance of the electroplex peak at 535 nm in bare PTE devices are also monitored. The Vturn-on values of Device 1-3 are in the range of bare PDI and bare PTE containing devices. Although the current density at the same applied voltage is in the order of Device 1 > Device 2 ≅ Device 3 (Figure 7b), the EL intensity decrement at 630 nm is continued in Device 2 and Device 3. This situation is evaluated as more evidence of electric field independent formation of 630 nm peak, and the fading is ascribed to the reduction of host concentration ready to form exciplex with PDI as the PTE wt.% is increased. Absence of 535 nm peak is attributed to the wide overlap between the emission of PTE and absorption of PDI ( Figure S2).  Figure S8).
The white light properties of Device 1-3 are given in Table 2. Combination of blue, green and orange-red emitter resulted in a CRI value higher than 90. However, white light characteristic is reduced from Device 1 to Device 3. By considering the discussions provided above, in order to regain the green photons absorbed by PDI and increase the host concentration in emitting layer, reducing the PDI wt.% is employed. Device 4, with an emitting layer of PVK:PBD: 0.06 wt.% PDI: 0.03 wt.% PTE, is prepared. Its EL spectrum presented a full visible range coverage (Figure 8) with the combination of dominated peaks at 445, 535, 558 and 620 nm. These peak points are attributed to PVK:PBD, The white light properties of Device 1-3 are given in Table 2. Combination of blue, green and orange-red emitter resulted in a CRI value higher than 90. However, white light characteristic is reduced from Device 1 to Device 3. By considering the discussions provided above, in order to regain the green photons absorbed by PDI and increase the host concentration in emitting layer, reducing the PDI wt.% is employed. Device 4, with an emitting layer of PVK:PBD: 0.06 wt.% PDI: 0.03 wt.% PTE, is prepared. Its EL spectrum presented a full visible range coverage (Figure 8) with the combination of dominated peaks at 445, 535, 558 and 620 nm. These peak points are attributed to PVK:PBD, PVK:PBD:PTE electroplex, PDI and addition of PVK:PBD:PDI exciplex and PDI emission Electronics 2021, 10, 725 9 of 12 peaks, respectively. Electron (7.3 × 10 −7 cm 2 V −1 s −1 ) and hole (8.5 × 10 −7 cm 2 V −1 s −1 ) mobility values of this device addressed a balanced charge injection in the emissive layer ( Figure S9). White light properties of 96 CRI, 4916 K CCT and (0.34, 0.36) CIE coordinates are achieved (Device #4 in Table 2), which can be manipulated with the applied voltage between 86 and 96, 3800 K and 5100 K, and (0.38, 0.34) and (0.33, 0.40), respectively ( Figure S10). emitting layer of PVK:PBD: 0.06 wt.% PDI: 0.03 wt.% PTE, is prepared. Its EL spectrum presented a full visible range coverage (Figure 8) with the combination of dominated peaks at 445, 535, 558 and 620 nm. These peak points are attributed to PVK:PBD, PVK:PBD:PTE electroplex, PDI and addition of PVK:PBD:PDI exciplex and PDI emission peaks, respectively. Electron (7.3 × 10 −7 cm 2 V −1 s −1 ) and hole (8.5 × 10 −7 cm 2 V −1 s −1 ) mobility values of this device addressed a balanced charge injection in the emissive layer ( Figure  S9). White light properties of 96 CRI, 4916 K CCT and (0.34, 0.36) CIE coordinates are achieved (Device #4 in Table 2), which can be manipulated with the applied voltage between 86 and 96, 3800 K and 5100 K, and (0.38, 0.34) and (0.33, 0.40), respectively ( Figure  S10).

Conclusions
One of the many problems in WOLED generation is the energy transfer between R G B emitters and local color losses. In this study, we produced an optimized host: perylene single layer WOLED with white light properties of 96 CRI, 4916 K CCT and 0.34, 0.36 CIE from selected emitters that prevent aggregation with low doping rates in host [15]. To the best of our knowledge, this device presents the best white light properties of current solution processed single layer perylene based WOLED literature. Comparison of PL and EL spectra of the films and devices, respectively, addressed that the orange-red emission was dominated by the exciplex formation, whereas green emission was mainly generated from the electroplex emission. Energy transfer from the host to the green emitting PTE is more efficient than that of to the orange-red emitting PDI. The lack of the red region resulted in CRI values of only around 75 in bare PTE devices. The highest CRI value obtained with bare PDI devices was 90. Even though the devices do not contain additional layers, optimization of the dopant ratio and detailed investigation of the electronic mechanism during the device operation allowed for reporting the best white light properties of solution processed WOLED literature [28].

Conclusions
One of the many problems in WOLED generation is the energy transfer between R G B emitters and local color losses. In this study, we produced an optimized host: perylene single layer WOLED with white light properties of 96 CRI, 4916 K CCT and 0.34, 0.36 CIE from selected emitters that prevent aggregation with low doping rates in host [15]. To the best of our knowledge, this device presents the best white light properties of current solution processed single layer perylene based WOLED literature. Comparison of PL and EL spectra of the films and devices, respectively, addressed that the orange-red emission was dominated by the exciplex formation, whereas green emission was mainly generated from the electroplex emission. Energy transfer from the host to the green emitting PTE is more efficient than that of to the orange-red emitting PDI. The lack of the red region resulted in CRI values of only around 75 in bare PTE devices. The highest CRI value obtained with bare PDI devices was 90. Even though the devices do not contain additional layers, optimization of the dopant ratio and detailed investigation of the electronic mechanism during the device operation allowed for reporting the best white light properties of solution processed WOLED literature [28].