1. Introduction

White organic light-emitting diodes (OLEDs) are attracting widespread attention as next-generation low-cost and high-efficiency thin-film electroluminescent devices for both flat panel displays and lighting applications [1–3] due to recent rapid technical evolution. The uniqueness of these light-emitting devices can for example be demonstrated in transparent or bi-directional OLEDs (TOLEDs or BiOLEDs), where light is emitted from both the bottom and the top side of the device; this intrinsic bi-directional emission capability of OLEDs clearly distinguishes them from other light-emitting devices [4–11].

To improve the light-emitting performance of TOLEDs/BiOLEDs, many approaches have been used, such as the modification of electrode components [5–7] and the introduction of a capping layer [8, 9]. More recently, we reported systematic studies on the influence of micro-cavity (MC) effects on BiOLEDs in which a thin metal layer is inserted in between organic and transparent conductive oxide (TCO) layers [10, 11]. Although these previous studies were successful in enhancing the efficiency of monochromatic OLEDs, limitations remain in the application of these methods to white-light BiOLEDs. This is mainly due to the spectral dependence of the cavity resonance condition that makes it very challenging to control the optical characteristics of MC-OLEDs such that the cavity resonance enhancement is achieved over a wide range of emission wavelengths. Fortunately, BiOLEDs rely inherently on top and bottom electrodes that are both relatively transparent. Hence it might be plausible to assume that the MC effect in BiOLEDs is “weak” at best and that the issue of spectral sensitivity is therefore of less concern for white BiOLEDs. However, questions still remain whether such a low degree of MC effect, if any, can still lead to efficiency enhancement in white BiOLEDs.

Here, we explore the characteristics of white BiOLEDs in which a 10-nm-thick Ag layer is added in between organic and ITO layers. We show that the added metal layer is thin enough not to cause a significant change in the spectral output of white BiOLEDs yet thick enough to enhance their efficiency and obtain balanced top/ bottom emissions. White BiOLEDs with a current efficiency (CE) enhancement of 26.2% are demonstrated, and the relative efficiency enhancement is shown to be relatively uniform across the whole visible spectrum.

2. Experiment

A series of white phosphorescent BiOLEDs based on p-i-n doped structures was fabricated with the following configuration: indium tin oxide (ITO) (90 nm)/inserted Ag layer (0 or 10 nm)/p-layer/1,1-bis[(di-4-tolylamino)phenyl]cyclohexane (TAPC) (10 nm)/emissive layer (EML, 20 nm)/ 2-(diphenylphosphoryl)spirofluorene (SPPO1) (10 nm)/n-layer/Ag (top cathode, 15 nm). The additional 10-nm-thick Ag layer deposited on top of the ITO anode was included to investigate the MC effect on the performance of the white BiOLEDs, as shown in Fig. 1. As the p-type hole injection and transport layer, we used 30-nm-thick films of N,N,N,N’-tetrakis(4-methoxyphenyl)-benzidine (MeO-TPD) doped with 4 wt % of the p-dopant 2,2'-(perfluoronaphthalene-2,6-diylidene)dimalononitrile (F6TCNNQ, Novaled AG, Dresden). The n-type electron injection and transport layer was 30 nm of 4,7-diphenyl-1,10-phenanthroline (BPhen) doped with cesium (Cs). A 10-nm-thick film of TAPC and a 10-nm-thick film of SPPO1 were used as electron and hole blocking layers, respectively, to confine charge carriers and excitons within the EML. Two blue-emitting EMLs, one containing 7 wt% blue light-emitting iridium(III)bis(4,6-difluorophenyl)-pyridinato-N,C^2’)picolinate (FIrpic) co-deposited with the hole-transport type host material 4,4’,4”-tri(N-carbazolyl)triphenylamine (TCTA, 5 nm) and another containing 10 wt% FIrpic co-deposited with the bipolar type host material 2,6-bis(3-(carbazol-9-yl)phenyl)pyridine (26DCzPPy, 5 nm), were successively deposited [12, 13]. Two types of white BiOLEDs were fabricated: Device A was based on a 2-color white approach, a 1-nm-thick layer of TCTA doped with 6 wt% of the red emitter iridium (III) bis[2-methyldibenzo-(f,h)quinoxaline](acetylacetonate) (Ir(MDQ)₂(acac)) was inserted between the two blue EMLs. Device B was a 3-color WOLED with a 1-nm-thick TCTA layer doped with 6 wt% Ir(MDQ)₂(acac) and another 1-nm-thick TCTA layer doped with 6 wt% of the orange emitter iridium (III) bis(2-(9,9-dihexylfluorenyl)-1-pyridine) (acetylacetonate) (Ir(dhfpy)₂(acac)). For all devices, a 50-nm-thick layer of TCTA was deposited as a dielectric capping layer on top of the cathode [9, 10].

 

Fig. 1 Device structures of the bi-directional OLEDs tested in this study.

Download Full Size | PDF

The current density-voltage-luminance (J-V-L) characteristics of the devices were measured with a source measure unit (Keithley 2400), and the spectral radiant intensity was determined by a calibrated spectroradiometer (Instrument Systems GmbH CAS140).

3. Results and discussions

Figure 2 shows the J-V-L characteristics of our 3-color white BiOLEDs (Device B) with and without a 10-nm-thick Ag layer (B1: without Ag, B2: with Ag). The J-V curves as well as the EL onset voltages of the OLEDs are nearly identical because doped MeO-TPD layer deposited onto Ag or ITO can realize ohmic contacts. By using a moderate doped transporting layer, interface effects as well as the work function of electrode only play a minor role for the energy alignment [14]. As a result, it is demonstrated that the additional Ag layer does not alter the electrical characteristics of our BiOLEDs. However, noticeable differences in the L-V curves are observed for both the top and bottom emission, depending on the presence of the Ag layer. As has been discussed in several previous reports, the bottom emission (black symbols) is much stronger than the top emission (red symbols) in Device B1 because the reflectance of the thin metal cathode is higher than that of the ITO anode [4–11]. On the other hand, in Device B2 the bottom emission (green symbols) and the top emission (blue symbol) are nearly identical due to a significant enhancement in the top emission. This can be understood by the reflectance of the bottom electrode that increases when one introduces the Ag layer on top of the ITO electrode.

 

Fig. 2 Current density versus voltage (J-V) and voltage versus bottom- and top-luminance (V-L) for Devices B1 and B2.

Download Full Size | PDF

It is noteworthy that the top/bottom balancing effect due to the insertion of the Ag layer in Device B2 is achieved without a significant compromise in the bottom emission, as shown in Fig. 3 and Table 1. In fact, the current efficiency (CE) for bottom emission decreases only slightly from 17.8 cd/A to 15.7 cd/A at J of 15 mA/cm², when the Ag layer is added. The CE for top emission, on the other hand, increases significantly by a factor of 2.65 from 4.8 cd/A to 12.8 cd/A, resulting in an overall enhancement in the CE for total emission by more than 25%.

 

Fig. 3 Current efficiency of bottom- and top-emission versus current density of (a) Devices A1 and A2 and (b) Devices B1 and B2.

Download Full Size | PDF

Table 1. Performance parameters of the white BiOLEDs under study*

View Table

To better understand the experimental results, a micro-cavity (MC) OLED model [10, 15–22] has been adopted to analyze the optical properties of the proposed white BiOLEDs. The precise treatment of the MC effect should consider dipole emitters embedded vertically or horizontally within OLEDs, which are optically a multilayer stack consisting of dielectric and/or metallic layers, and the environment-induced change in their radiative lifetime as well as coupling to various modes such as waveguide modes and surface plasmon polariton (SPP) modes [20, 21]. For the emission in forward-direction (θ = 0), however, the Fabry-Perot (FP) formulation which uses a plane-wave approximation is also popular for simplicity’s sake and can still be useful in illustrating various phenomena in MC-OLEDs [18, 19]. Within the framework of the FP formulation, one can describe the emission intensity of an OLED in a simplified structure of front electrode/ organic layer/ rear electrode with a sheet-like emission zone by [18, 19]:

From the result of optical analysis shown in Fig. 4, we can attribute the observed enhancement to the multiple-beam cavity resonance enhancement as well as to constructive two-beam interference that can occur efficiently in both emission directions as the reflectance of the bottom Ag layer (anode side) is comparable to that of the top Ag layer (cathode side) [11]. For top emission, the ratio of g_res(λ) obtained for devices with the inserted Ag [ = g_res^(Ag)(λ) ] to that for devices without it [ = g_res^{(no Ag)}(λ) ] indicates that there is indeed a resonance-induced enhancement centered around λ of 550 nm with the peak enhancement ratio of 1.56. However, the two-beam interference effect appears to be more important by comparison; as can be seen in Fig. 4, the enhancement ratio for f_TB(λ) ranges from 1.75 to 2.20 throughout the visible spectral range.

 

Fig. 4 Calculated enhancement in multiple-beam resonance [g_res(λ)] and two-beam interference [f_TB(λ)] terms due to the additional Ag layer in the bottom electrode of the white BiOLEDs under study. Calculation was based on Fabry-Perot formalism in a simplified structure of front electrode/ organic layer/ rear electrode using Eqs. (1)-(3).

Download Full Size | PDF

For bottom emission, on the other hand, the two-beam interference effect remains unchanged because there is no change in the reflectance from the top electrode. The multiple-beam cavity resonance enhancement given by the denominator in Eq. (2) is same as that in the top-emission case, but it is over shadowed by the reduced transmittance due to the Ag layer, which leads to a decrease in the ‘g_res^(Ag)/ g_res^{(no Ag)}’-ratio by a factor of T_bot^(Ag)/T_bot^{(no Ag)} (shown as a dashed gray line in Fig. 4) at a given wavelength. This reduced transmittance leads to a small net decrease in light output for the bottom direction. In addition, we observe that the bottom-to-top CE ratio (γ_CE = CE for bottom-emission / CE for top-emission) is strongly influenced by the inserted Ag layer. The γ_CE values of the reference Devices A1 and B1 without Ag are 3.89 and 3.67, respectively, implying imbalanced emissions in these devices, with the bottom emission being much stronger than the top emission. In contrast, the γ_CE values of Devices A2 and B2 with the Ag layer are 1.20 and 1.23, respectively. The introduction of the Ag layer on top of the ITO can thus achieve a good balance between the bottom emission and the top emission. As mentioned previously, spectrally balanced broadband white emission is of great importance for practical applications of white MC-OLEDs but achieving such characteristics has turned out to be very challenging, especially for unidirectional bottom- or top-emitting white MC-OLEDs.

For the BiOLEDs investigated here, the EL spectra shown in Fig. 5 indicate that balanced broadband white emission is emitted in the top and bottom direction. Although introducing the additional Ag layer tends to increase MC effects, the EL spectra of Devices A2 and B2 do not exhibit significant distortions compared to the respective reference (Devices A1 and B1). As a result, the color rendering index (CRI), the Commission Internationale del'Eclairage (CIE) color coordinates, and the correlated color temperature (CCT), of the devices with additional Ag layer are comparable to the values for the reference devices. CRI, CIE and CCT are summarized in Table 1. For instance, for bottom emission, Device B2 exhibits a CRI of 74.8, CIE coordinates of (0.43, 0.44), and a CCT of 3404 K (at a current density of 15 mA/cm²). These values are near the values necessary to satisfy the Energy Star requirements for solid-state lighting applications [22].

 

Fig. 5 Electroluminescence spectra emitted in bottom and top-direction for (a) Devices A1 and A2 and (b) Devices B1 and B2 at a driving current density of 15 mA/cm².

Download Full Size | PDF

As can be seen in Fig. 6(a), the simulation results obtained under a full classical electromagnetic formalism describing dipole emitters in a cavity structure [20, 21, 23] show a good match to the experimental data obtained for Devices A, and it indicates that the intensity enhancement ratio between those with and without the bottom Ag layers follows a broad spectral envelope with the full-width half-maximum (FWHM) even comparable to the full visible spectral range, being consistent with the spectrally balanced enhancement in the proposed white BiOLEDs. (See Fig. 6(a)) The simulation results presented in Fig. 6(b) and 6(c) for top-emission direction further show that increasing the MC effect by making both of the Ag layers too thick indeed reduce the spectral width of the cavity-induced intensity enhancement, which may then cause undesirable side effects such as spectral distortion or reduced CRI. Nevertheless, the proposed white BiOLEDs are expected to exhibit spectrally balanced enhancement over the entire visible spectral range provided that the thickness of top and bottom Ag layers is maintained within the range of approximately 10 nm - 20 nm.

 

Fig. 6 (a) The overall enhancement ratio of the forward intensity I(λ) due to the additional Ag layer in the bottom electrode of white BiOLEDs: comparison between experimental data (Device A) and simulation results. (b)-(c) Simulated intensity for various thickness values of the bottom Ag layer ( = d_Ag(bott.)) for the thickness of the top Ag layer of 15 nm (b) and 30 nm (c). Shown in (b) and (c) are the values obtained for top-emission direction. Simulation in Fig. 6 was done using the full classical electromagnetic model considering dipole emitters embedded in a microcavity structure as described in Ref. 20 with the unity radiative quantum efficiency assumed.

Download Full Size | PDF

4. Conclusions

In summary, the performance of white-emitting BiOLEDs has been enhanced and the balance between bottom- and top-emission has been improved by controlled introduction of MC effects. We find that the overall current efficacy of BiOLEDs can be increased without significant distortion of the white EL spectra when the reflectance of the electrodes in the BiOLED is carefully managed.