A Deep Blue Strong Microcavity Organic Light-Emitting Diode Optimized by a Low Absorption Semitransparent Cathode and a Narrow Bandwidth Emitter

Herein, deep blue strong microcavity top-emitting organic light-emitting diodes (TEOLEDs) optimized by a low absorbed silver cathode and a narrow bandwidth emitter are reported for realizing high-ef ﬁ ciency and narrow spectral characteristics. Usually, the Mg:Ag cathode is widely used as a semitransparent cathode in TEOLED devices due to its favorable re ﬂ ectivity and proper electron injection ability. However, the high extinction coef ﬁ cient of Mg causes a decrease in the microcavity effect, causing a low optical ef ﬁ ciency of devices. A Ag cathode is applied as a semitransparent cathode to modify the high absorption property for a strong microcavity blue OLED. At optimal conditions, the optically calculated device ef ﬁ ciency of TEOLEDs using a BPPyA emitter is 70.2 and 91.1 cd (A * y) (cid:2) 1 for Mg:Ag and Ag TEOLEDs, respectively. To further improve the device ef ﬁ ciency, an alternative emitter (DABNA-NP-TB) is applied which has a narrow bandwidth of 25 nm. The optically calculated ef ﬁ ciency enhancement in the TEOLED is about 23% compared with a control emitter, which is well correlated with the result of fabricated TEOLEDs. 35nm in ﬁ lm state. DABNA-NP-TB is DABNA-based ﬂ uorescent dopant and it has spectral peak of 458nm and FWHM of 25nm. The N , N 0 -di[4-( N , N 0 -diphenylamino)phenyl]- N , N 0 -diphenylbenzidine (DNTPD) and N , N 0 -bis(naphthalen-1-yl)- N , N 0 -bis(phenyl)benzidine (NPB) were purchased from Jinlin OLED Materials Tech. Likewise, 1,4,5,8,9,11- hexaazatriphenylenehexacarbonitrile (HATCN) was purchased from EM Index, and 1,3-bis(9-phenyl-1,10-phenanthrolin-2-yl)benzene (BPPB) [47] was purchased from OSM. The current density – voltage – luminance ( J – V – L ) characteristics of TEOLED devices were measured by using Keithley SMU 2635A and Minolta CS-100A. EL spectra and CIE coordinate were obtained by using a Minolta CS-2000A.


Introduction
In the display applications, the strong microcavity top emission organic light-emitting diodes (TEOLEDs) are practically suitable due to the various optical advantages such as high aperture ratio, good color purity, and high efficiency in normal forward direction. [1,2] These TEOLEDs are composed of two highly reflective metal electrodes wherein these two electrodes generate strong microcavity effect. [3][4][5] Such strong microcavity effect exhibits a narrow full width at half maximum (FWHM) compared with the noncavity device due to the amplification of selective wavelength by the internal resonance condition. [6,7] As a result, the microcavity TEOLED has a good color purity and higher efficiency than the noncavity device at the normal forward direction. [8] In such microcavity device, the thick metal layer is typically used as a reflective anode owing to its high reflectivity. And the semitransparent cathode needs to have a high reflectance and high transmittance with almost no absorption property.
To identify the output spectral intensity of microcavity device, in 1993, Deppe et al. reported the variation of spectral intensity as the following equation. [9] jE cav ðλÞj 2 ¼ In Equation (1), where |E nc (λ)| 2 is an internal spectral intensity and |E cav (λ)| 2 is output spectral intensity. The interference of light generated inside the device is occurred, thereby affecting the internal spectral intensity. In this equation, the reflectivity of both electrodes and distance between these electrodes can be changed. Using the parameter variation of reflectivity and distance, we can control the output spectral intensity. In the TEOLED, the reflectance of reflective anode usually exhibits over 90% in the visible light region. [10] Therefore, the reflectance of anode can hardly be improved and hence we did not consider the optical property of reflective anode. In addition, term of (1 À R 2 ) represents the sum of transmittance and absorbance because of the relationship between transmittance, reflectance, and absorbance described by T þ R þ A ¼ 1. [11,12] When the absorbance decreases, it causes the enhancement of spectral intensity due to the increased transmittance. Consequently, the output spectral intensity of the microcavity device depends on the optical absorbance of the semitransparent cathode. Change in the type of cathode material and cathode thickness causes a variation of the microcavity effect in TEOLEDs. Thus, the control on the microcavity effect can enhance the efficiency of microcavity device due to the improvement of internal resonance condition. [13,14] DOI: 10.1002/adpr.202000122 Herein, deep blue strong microcavity top-emitting organic light-emitting diodes (TEOLEDs) optimized by a low absorbed silver cathode and a narrow bandwidth emitter are reported for realizing high-efficiency and narrow spectral characteristics. Usually, the Mg:Ag cathode is widely used as a semitransparent cathode in TEOLED devices due to its favorable reflectivity and proper electron injection ability. However, the high extinction coefficient of Mg causes a decrease in the microcavity effect, causing a low optical efficiency of devices. A Ag cathode is applied as a semitransparent cathode to modify the high absorption property for a strong microcavity blue OLED. At optimal conditions, the optically calculated device efficiency of TEOLEDs using a BPPyA emitter is 70.2 and 91.1 cd (A * y) À1 for Mg:Ag and Ag TEOLEDs, respectively. To further improve the device efficiency, an alternative emitter (DABNA-NP-TB) is applied which has a narrow bandwidth of 25 nm. The optically calculated efficiency enhancement in the TEOLED is about 23% compared with a control emitter, which is well correlated with the result of fabricated TEOLEDs.
To generate the strong microcavity effect, the thin metal film is generally used as a semitransparent cathode. Only Ag is known to have a low absorption; however, it is difficult to apply as a cathode due to its high work function for the cathode applications. [15] However, recently a pure Ag cathode with an improved electron injection property and stability has been reported. [16] This device showed much higher efficiency than the other cathode structure such as Mg:Ag (10:1) [17,18] and Al/Ag [19,20] device. The difference in efficiency is because of the high extinction coefficient of Al [21] and Mg, [22] although these are used in TEOLEDs for having a superior electron injection property. In addition, the microcavity effect can be enhanced by applying the capping layer (CPL) between the cathode and air. Generally, the CPL is used for enhancing the outcoupling efficiency at the normal direction due to the difference of refractive index between air and CPL. [23] Thus, the thickness of cathode and CPL mainly affect the microcavity effect in TEOLED. [24][25][26] In case of the emitter, we have to use the narrow emitter for enhancing the microcavity effect in specific wavelength region. With the implementation of microcavity effect in device, the current efficiency (cd A À1 ) enhancement is occurred when the device has a red or green color. On the contrary, the current efficiency is decreased at the blue color device because the region of deep blue color has very low sensitivity to human eye than other colors. [27] In this article, we have implemented a low absorption cathode to blue TEOLED for achieving the high efficiency and appropriate color purity. Before formulating the device, we performed the optical simulation to investigate the optical properties of cathode unit and optimum optical thickness of TEOLED. Based on the optical simulation, we fabricated devices of Mg:Ag and Ag cathode units to identify their properties. At the same thickness, Ag cathode shows high transmittance tendency than Mg:Ag cathode transmittance. The Ag TEOLED exhibited the efficiency enhancement of 1.26 times than that of the Mg:Ag TEOLED because Ag cathode has a low absorbance than Mg:Ag cathode. It was identified in transmittance calculation. In addition, we applied emitter which has different spectrum widths for comparing cavity effect as the emitter spectrum. This efficiency enhancement is believed to be due to the difference in applying the low absorption cathode and narrow emitter.

Results and Discussion
Before the fabrication of TEOLED, we performed optical simulation of cathode unit to identify their optical properties. Figure 1a shows the refractive index (n) and extinction coefficient (k) with wavelength of Mg and Ag. Mg shows higher n and k values than Ag at the visible wavelength region. For the optical simulations, the optical constants of Mg:Ag (10:1) were used from our previously reported article. [28] As a result, there is a difference in transmittance between Mg:Ag (10:1) and Ag cathode units, as shown in Figure 1b. Each cathode unit is designed as glass substrate (0.7 t)/organic layer (30 nm)/cathode unit (X nm)/CPL (60 nm). X is a variable value from 12 to 24 nm for checking tendency with variation of cathode thickness. At the wavelength of 460 nm, Mg:Ag (10:1) and Ag cathode units showed 26.3% and 69.4%, respectively, at same thickness of 18 nm. This tendency is intensified as the increase in cathode thickness. Especially, Ag 24 nm and Mg:Ag (10:1) 12 nm showed similar transmittance value. We can expect that the Mg:Ag (10:1) cathode has high absorption property due to the high n and k value of Mg. Mg, which occupies a large proportion in Mg:Ag (10:1) cathode, plays a role in lowering the transmittance of the Mg:Ag cathode. Therefore, we believe that a low absorption of Ag metal is playing a significant role in determining the transmittance of fabricated cathode units. As a consequence, the Mg:Ag [28] cathode with higher n and k values than that of Ag [21] seems to be responsible for the higher absorbance in Mg:Ag than Ag cathode according to the relationship between transmittance, reflectance, and absorbance values.
Considering previous cathode calculation, we performed the optical simulation of blue TEOLED for enhancing the efficiency and achieving the appropriate color purity. For the optical simulation, the optical constants of each organic layer, cathode metal, and glass are used like earlier in transmittance calculation of cathode unit. In addition, the optical constant of indium tin oxide (ITO) as refractive index of 2.0 and extinction coefficient of almost zero at wavelength of 460 nm from the previous report were used. [29] These optical constants of each layer were reported in previous article. [28] The thickness conditions of cathode and CPL are used as discussed in aforementioned explanation. Applying these optical conditions, we calculated the radiance distribution in accordance with the distance between anode to www.advancedsciencenews.com www.adpr-journal.com emissive layer (EML) distance (HTL thickness) and EML to cathode distance (ETL thickness), as shown in Figure 1c. From this radiance distribution, we optimized the location of EML in TEOLED device, which has a highest radiance and proper thickness condition. As mentioned previously in the optical interference in TEOLED, the interference condition is changed in accordance with the total thickness of deposited layers between both electrodes. At the thickness condition of proper interference, the internal light exhibits antinode condition (maximum amplitude), as well as there are various EML positions which have thick HTL or thick ETL condition. Also, there are interference thickness conditions according to the change in the antinode order with the increasing thickness. If the total thickness is thicker than the proper interference condition, the resonant wavelength will be longer, and the output spectral intensity will exhibit at a long wavelength region. [8] We decided a position of EML which has second-order cavity condition with the thick HTL layer. Due to the issues with the fabrication process of first-order microcavity condition with thin organic layer thickness, the second-order microcavity condition is applied. [30] The thick HTL layer condition is used as it is easy to control the charge balance and the cavity condition. Considering these conditions, we realized thickness condition as in the following structure: Mg:Ag TEOLED:Ag (150 nm)/ITO (10 nm)/organic layer (145 nm)/EML (20 nm)/organic layer (30 nm)/Mg:Ag (10:1) cathode/CPL. Ag TEOLED:Ag (150 nm)/ITO (10 nm)/organic layer (139 nm)/EML (20 nm)/organic layer (30 nm)/Ag cathode/CPL.
A slight difference in organic layer thickness between Mg:Ag and Ag TEOLEDs was attributed to the different optical condition of each cathode unit. At the device structure with optimized EML position, we simulated the radiance distribution in accordance with the thickness of cathode and CPL for achieving the high performance. To design a highly efficient blue TEOLED with appropriate color characteristic, the color coordinate has to be considered due to the low eye sensitivity at deep blue color coordinates. Figure 2a shows a normalized radiance distribution of Mg:Ag TEOLED in accordance with the thickness of cathode and CPL. A variation in thickness of cathode unit affects the optical properties in TEOLED, and hence it causes the change in radiance distribution by microcavity effect. A thickness condition of highest radiation is located at 10-15 nm in Mg:Ag cathode and CPL thickness of 50-70 nm in the radiance distribution of Mg:Ag TEOLED. To identify the suitability about the thickness condition with highest radiance, we fitted the variation of radiance with consideration of Commission Internationale de l'éclairage  (CIE) y coordinate in accordance with the thickness of cathode and CPL. Figure 2b shows the variation of radiance divided by CIE y color coordinate, and the thickness condition with highest radiance exhibited high radiance/CIE y values. Before dividing radiance by CIE y, high radiance condition is broadly distributed. On the contrary, in radiance/CIE y distribution, their distribution is narrow. Dividing CIE y means it only considers blue emission intensity with considering blue color coordinate, and high value of radiance/CIE y means it corresponds to high efficiency and high color purity condition. Using these various conditions, we decided the thickness condition with high radiance and appropriate color purity, which produced a thickness value of Mg:Ag cathode and CPL of 14 and 60 nm, respectively. As described previously, the optimized thickness of cathode and CPL of Ag TEOLED was simulated. Figure 2c,d shows a normalized radiance and radiance/CIE y distribution of Ag TEOLED. The Ag cathode and CPL thickness of 24 and 60 nm conditions were obtained as an optimal condition, respectively. Considering these results of simulation with radiance distribution and consideration of CIE y color coordinate, we determined the thickness of metal cathode and CPL. Using these simulated thickness conditions, the efficiency of TEOLED was calculated.
The calculation based on the simulated efficiency of bottomemitting OLED (BEOLED) is performed by optical simulation by using measured BEOLED characteristics, as shown in Figure 3a.
We fabricated BEOLED structure of ITO (50 nm)/HATCN (7 nm)/NPB (65 nm)/PhPC:5%BPPyA (20 nm)/BPPB (25 nm)/ Liq (1.5 nm)/Al (100 nm). It exhibited measured current efficiency of 8.3 cd A À1 and CIE color coordinates of (0.14, 0.12) at 1 mA cm À2 condition. Based on the measured BEOLED efficiency, we calculated TEOLED efficiency by using relative value. Especially, applying different cathode between Mg:Ag and Ag cathode causes difference of device efficiency. However, there is different tendency of efficiency with color coordinate. It is originated by low eye sensitivity of blue emission (strong emission intensity in blue region reduces device efficiency). To consider the influence of low eye sensitivity in current efficiency, we used cd (A * y) À1 unit which is current efficiency divided by CIE y coordinate. Calculated MgAg and Ag TEOLED device characteristics are shown in Table 1. At the optimal condition, MgAg TEOLED and Ag TEOLED showed 70.2 and 91.1 cd (A * y) À1 , respectively. There is efficiency enhancement of almost 30% between MgAg and Ag TEOLED in optical calculation. These results indicate that the Ag cathode with low absorption is more efficient than Mg:Ag cathode to use as a semitransparent cathode in blue TEOLEDs.
Through the optical simulation, we designed highly efficient TEOLED with Ag cathode and it has better properties than conventional MgAg TEOLED. This device has optimal EML position and cathode unit, and it is optically optimal condition at same material structure. To enhance the TEOLED efficiency, we applied different fluorescent dopant 2,12-di-tert-butyl-N,N,5,9-tetrakis (4-(tert-butyl)phenyl)-5,9-dihydro-5,9-diaza-13b-boranaphtho[3,2,1de]anthracen-7-amine (DABNA-NP-TB) which has narrow spectrum. Figure 3b shows film photoluminescence of N,N 0bis-dibenzofuran-4-yl-N,N 0 -bis-(2,5-dimethyl-phenyl)-pyrene-1,6diamine (BPPyA) and DABNA-NP-TB, and they have FWHM of 34 and 25 nm, respectively. There is difference in FWHM of  www.advancedsciencenews.com www.adpr-journal.com almost 10 nm; it can cause difference in device calculation. We performed optical simulation to identify device efficiency of DABNA-NP-TB emitter. It has almost similar peak wavelength with BPPyA; we applied same structure to calculate TEOLED. In Table 1, Ag TEOLED_M with DABNA-NP-TB emitter showed 112.5 cd (A * y) À1 value. In the same structure, only a change in the emitter produced an efficiency enhancement of 23%. The cause of this enhancement in efficiency is a change in the FWHM of the emitter. BPPyA is commonly used fluorescent material design based on pyrene core. On the contrary, DABNA-NP-TB is boron acceptor-based structure which causes narrow FWHM due to its multiresonance effect. [31,32] We can represent the spectrally integrated enhancement of emission as where G cav and G noncav indicate resonant emission enhancement in cavity and noncavity medium. [33] Δλ em and Δλ cav indicate emission spectrum widths and cavity mode width, respectively. At this equation, low Δλ em means narrow FWHM, and it causes increase in spectrally integrated enhancement of emission. As a result, we can expect the efficiency of TEOLED with optimized cathode unit can be improved by controlling emitter FWHM at the same structure. In these structures, the ITO deposited on Ag electrode was used as a reflective anode. As a consequence, the Ag/ITO has a high reflectivity due to the thick Ag layer. Furthermore, it has a high hole injection property due to the high work function of ITO. We applied the DNTPD as a hole injection layer (HIL), NPB as a hole transport layer, and HATCN as a n-type HIL, which has a role of improved hole injection property. [33] The BPPB is used for ETL. Well-known lithium fluoride (LiF) is used for Mg:Ag cathode device because of the good electron injection property of LiF/Mg:Ag cathode. [34,35] The lithium could be liberated during the deposition due to the reaction between LiF and Mg. [36] On the contrary, we applied Li-doped BPPB in the Ag TEOLED structure to improve the electron injection property because alkali metal was previously reported as an electron injection layer in the pure Ag cathode structure. [37,38] PhPC was used for blue host. As a blue dopant, BPPyA and DABNA-NP-TB were used. These fabricated devices were characterized and results are presented as follows. Figure 4a shows the current density and luminance versus voltage characteristics of fabricated TEOLEDs. At 1 mA cm À2 , the driving voltage values of MgAg and Ag TEOLEDs are almost similar of 2.7 V. These results show that the MgAg and Ag TEOLEDs have similar electrical characteristics. However, current density of MgAg TEOLED shows relatively low values at high voltage, because EIL (LiF) has slightly low injection and transport characteristics than Li-doped BPPB. Also, Mg:Ag (10:1) cathode has low sheet resistance than thick Ag cathode due to its different conductivity. Accordingly, luminance curves are also varied. Hence, we focused on the optical characteristics regardless of electrical performances. The current efficiency values of MgAg, Ag TEOLED, and Ag TEOLED_M showed 3.6, 3.9, and 3.9 cd A À1 at the current density condition of 1 mA cm À2 , www.advancedsciencenews.com www.adpr-journal.com respectively. Similar current efficiency of fabricated devices does not distinguish which device has better in blue emission. As previously discussed, we calculated cd (A * y) À1 to distinguish the device performance. Figure 4b) shows calculated cd (A * y) À1 versus luminance values of fabricated devices. Unlike the current efficiency result, there is large difference in efficiency between each device due to the consideration of CIE y coordinate. Change in the cathode from Mg:Ag (10:1) to Ag showed efficiency enhancement of 26% (71.9 cd (A * y) À1 ! 90.7 cd (A * y) À1 ). The cause for the higher efficiency of Ag TEOLED than that of MgAg TEOLED is expected to be due to the difference of absorption properties. They have almost similar spectral peak and FWHM, and it indicates TEOLEDs with different cathode have similar cavity condition at optimal thickness condition.
In case of change the emitter material at same structure, Ag TEOLED_M, showed 107.2 cd (A * y) À1 and this is the value increased by 18% over the Ag TEOLED. As discussed previously, it is due to the difference in FWHM of emitter BPPyA (34 nm) and TDBA-TBN (25 nm). Narrow FWHM emitter causes more amplified enhanced intensity by cavity effect; it makes TEOLED with increased efficiency. Slightly difference between calculated efficiency and measured efficiency is due to the slight difference in thickness and optical condition. It is shown in Figure 4c-e which is comparison of calculated and measured spectra of device. We obtained almost similar spectrum between calculated and measured results; however, there is slight discrepancy such as maximum spectral peak and FWHM about 1-2 nm. This result indicates that the tendencies of calculations and measurements are consistent, but there may be slight difference due to optical characteristic of cathode unit or different thickness of organic layer. In addition, we evaluated the stability of TEOLEDs with Mg:Ag and pristine Ag cathodes with BPPyA emitter. The Ag TEOLED showed relatively short device lifetime (LT 90 ¼ 15 h) than the MgAg TEOLED (LT 90 ¼ 26 h). Such short device lifetime seems to be related to easy aggregation of pure Ag under operating condition through the high surface energy of Ag. We also added summarized table which have reported blue TEOLED characteristics and it can show the importance of our devices ( Table 2).
To examine the efficiency enhancement of Ag TEOLED and Ag TEOLED_M due to the influence of a low absorption property of Ag cathode and narrow FWHM emitter, the optical simulation was performed using the analysis of dissipated power at 460 nm which is maximum spectral peak of blue emitters. Figure 5 shows the dissipated power versus normalized in-plane wavevector of optimized TEOLEDs. When the cathode material is varied from Mg:Ag (10:1) to Ag, there is an overall increase such as out-coupled, wave-guided, and evanescent-coupled. It indicates decreased absorption in Ag TEOLED makes enhanced power dissipation in whole layers. Different power dissipation at normalized in-plane wavevector is due to the different n and k value of cathode, and different thickness of device structure. In case of emitter change condition, Ag TEOLED_M exhibited almost similar dissipated power curve because they have same structure. However, there is difference in dissipated power intensity between both devices. It indicates change in the emitter material causes enhancement of dissipated power at 460 nm which is maximum spectral peak. Through these results, optimization of cathode unit and applying narrow FWHM emitter have to be considered to design structure of blue TEOLED.

Conclusion
In summary, we optimized the blue TEOLED devices having a high efficiency and good color purity. To enhance the device efficiency, we applied Ag cathode, which has a low absorption property compared with Mg:Ag cathode. The optical simulation results of device are almost matched with the measured device characteristics. Furthermore, we applied narrow bandwidth emitter to enhance cavity effect in TEOLED. Measured Mg:Ag and Ag TEOLED showed the efficiency values of 70.2 and 91.1 cd (A * y) À1 , respectively. When we applied different emitter (narrow FWH), it was increased to 112.5 cd (A * y) À1 . The significant efficiency enhancement of about 1.6 times was noted for Ag device compared with Mg:Ag device. In conclusion, we have demonstrated that the optical properties of cathode unit and emitter spectrum widths are mainly determining the optical characteristics of TEOLEDs.

Experimental Section
Fabrication of Cathode Unit: We have fabricated cathode unit which is composed of the following structure: bare glass/ETL/semitransparent cathode (Ag, Mg:Ag)/CPL. BPPB and NPB layers are used ETL and CPL, respectively. All the layers were deposited under a pressure of 10 À7 -10 À6 Torr. The deposition rate of BPPB and NPB was 0.5 Å s À1 . The deposition rates of Mg and Ag were 2.0 and 0.2 Å s À1 , respectively.  www.advancedsciencenews.com www.adpr-journal.com Optical Simulation: To optimize the optical condition of device, we performed the optical simulation, which is SETFOS, Fluxim. In this program, the refractive index (n), extinction coefficient (k), dipole orientation, photoluminescence spectrum of emitting material, and thickness of each layer were used. In case of the cathode simulation, we utilized the simple transfer matrix method for calculating the transmittance of cathode units. It uses only n, k value of metal and organic layer and their thickness. Based on the measured BEOLED efficiency, calculated TEOLED efficiency changed to relative value.
The current density-voltage-luminance ( J-V-L) characteristics of TEOLED devices were measured by using Keithley SMU 2635A and Minolta CS-100A. EL spectra and CIE coordinate were obtained by using a Minolta CS-2000A.