Open Access
Add to BibSonomy
Abstract
Light extraction improvement is still an important issue for active-matrix organic light-emitting diode displays (AMOLEDs). In our previous work, a three-dimensional (3D) reflective pixel configuration embedding the OLED in the concave 3D reflector and patterned high-index filler had been proposed for significant enhancement of the pixel light extraction. In this work, influences of thin film encapsulation (TFE) on light extraction of such reflective 3D OLED pixels are considered as well by simulation studies. Unfortunately, the optical simulation reveals strong reduction of the light extraction efficiency induced by TFE layers. As such, an additional angle-selective optical film structure between the pixel and the encapsulation layers is introduced to control the angular distribution of the light coupled into the encapsulation layers and to solve TFE-induced optical losses. As a result, TFE-induced losses can be substantially reduced to retain much of light extraction efficiency. The results of this study are believed to provide useful insights and guides for developing even more efficient and power-saving AMOLEDs.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Due to their various merits, active-matrix organic light-emitting diode displays (AMOLED) have penetrated many applications, and even more efficient AMOLEDs for power saving are thus highly desired. Thanks to the progress of triplet-harvesting emitting materials, nearly 100% internal quantum efficiency (IQE) can in principle be achieved in OLEDs [1,2]. On the other hand, although there are many light extraction technologies reported, they in general are not so applicable to current AMOLED pixel structures due to considerations of fabrication compatibility or image quality [3–14]. Therefore, development of effective and compatible light extraction structures for AMOLEDs remains an important issue. In our recent reports [15–17], based on the conventional two-dimensional (2D) AMOLED top-emitting pixel structure as shown in Fig. S1(a) (Supplement 1), we proposed a promising three-dimensional reflective pixel (3D-RP) configuration as shown in Fig. 1(a) for enhancing optical out-coupling efficiency of AMOLEDs. It includes several key features: (1) selectively applying high-index filler material in the concave area to effectively extract OLED internal emission into the high-index filler region, (2) extending the reflective bottom electrode toward the surface of the pixel definition layer (PDL) bank slope to form the concave 3D reflector for redirecting and out-coupling the light otherwise confined and guided in the filler material as illustrated in Fig. 1(a), and (3) adopting low-loss, transparent conductive oxide (TCO) as the top electrode of top-emitting OLED to reduce optical loss due to absorption and surface plasmon modes (SP mode) of top metal electrodes. With such a configuration, multi-scale optical simulation indicated promise of ∼70-80% light extraction efficiency for AMOLED pixels [15–17].
Fig. 1. Schematic diagram of (a) the 3D-RP pixel structure with the reflective bank slope and patterned high-index filler, (b) the 3D-RP pixel structure with 5 TFE layers above the high-index filler, and (c) the 3D-RP pixel structure with ASOF between the filler and TFE layers.
Practically, to prevent organic materials/devices from water vapor or oxygen degradation, AMOLED pixel arrays are usually further overcoated/protected by thin film encapsulation (TFE) composed of alternating inorganic/organic multilayers, as shown in Fig. S1(b) in Supplement 1 [18–21]. When overcoated over the 3D-RP pixels (as shown in Fig. 1(b)), such TFE layers certainly would change the optical scenario and impact/degrade the light extraction performance of the proposed 3D-RP pixel structure, but their influences have not been analyzed before. In this work, through optical simulation, we first study the significant impacts and detrimental effects of the TFE layers on the light extraction efficiency of the 3D-RP pixel structure. Then, we further propose modified/improved pixel structures (Fig. 1(c)) that can significantly reduce optical losses caused by TFE layers and retain as much as possible light extraction efficiency of the 3D-RP pixel structure.
2. Device and pixel structures studied and optical simulation methods
For simplicity and consistency, the 3D-RP pixel structure for this study (Fig. 1(a)-(c)) basically adopts the 3D-RP pixel geometries and dimensions in our previous simulation study [15,16]. As shown in the Fig. S2 in Supplement 1, the 3D-RP pixel has a bank height H of 2 µm (and thus also ∼2 µm filler thickness), a taper angle θB of 30° for the bank slope, the bank opening of 13 µm×13 µm (i.e., the bottom surface width W1 of 13 µm), the actual active emission area of 11 µm ×11 µm as defined by the opening in the additional dielectric layer (assume 300 nm SiNx, used for insulating the reflective bottom electrode onto the bank slope and having a refractive index similar to the high-index filler). The patterned high-index filler layer filled into the pixel concave area assumes a refractive index similar to OLED organic layers.
Five layers of TFE composed of alternating inorganic and organic layers are further disposed over the 3D-RP structure (3D-RP + TFE structure, Fig. 1(b)) to examine their impacts on light extraction of 3D-RP pixels. As typical for AMOLEDs, the 1st, 3rd, and 5th layers in TFE adopt the higher-index inorganic materials (such as SiNx, with a refractive index nH,TFE of ∼1.85, similar to the filler index); the 2nd and 4th layers in TFE adopt lower-index organic materials (such as Acrylic, with a refractive index nL,TFE of 1.49). Inorganic layers generally serve as better barrier for water and oxygen permeation, while thicker organic layers contribute to surface planarization, mechanical flexibility, decoupling of defects in inorganic layers, and decrease of water/oxygen penetration depth [18–21]. The thicknesses of layers in the TFE are: 1st TFE (X nm)/ 2nd TFE (1 µm)/ 3rd TFE (600 nm)/ 4th TFE (1 µm)/ 5th TFE (300 nm). As the thickness of the higher-index 1st TFE layer has critical effects on light extraction of the 3D-RP + TFE structure, its thickness is varied in the simulation study. As simulation reveals significant TFE-induced optical losses and reduction of light extraction of the 3D-RP pixels, to reduce such optical and efficiency losses, additional angle-selective optical films (ASOF) composed of alternating low-index and high-index material layers are inserted between the 3D-RP/filler and the TFE, forming the 3D-RP + ASOF + TFE pixel configuration shown in Fig. 1(c). The low- and high-index materials are assumed properties of LiF and NPB (nLiF∼1.37 and nNPB∼1.84 at 520 nm), both being common OLED materials that can be facilely deposited by the compatible thermal evaporation.
Since the 3D OLED pixel configuration contains features of very different dimensional scales, i.e., nm-scale structures smaller than wavelengths (e.g., thicknesses of the OLED active layers or optical thin films) and µm-scale structures significantly larger than wavelengths (e.g., pixel size, bank height, filler and TFE thicknesses etc.), optical properties of the 3D pixel structures are analyzed with the multi-scale optical simulation, as detailed in the Supplement 1 (section 1, 2, and Fig. S3). It combines the rigorous electromagnetic wave-based model for dealing with detailed emission or optical properties of nm-scale layered structures, with the geometric optics simulation based on Monte Carlo ray tracing for dealing with larger-scale structures. The emission properties from the OLED device to the filler region and optical properties of optical thin films (e.g., optical reflection and transmission seen from the filler region for the bottom surface, the bank slope surface, and angle-selective optical films (ASOF) etc.) are calculated by the electromagnetic wave optics. Then they are used to set up the user-defined ray sources and surface/interface coating properties for further ray tracing simulation of the overall reflective 3D pixel structure. The optical properties (refractive indexes, emission spectrum) of all materials used in the structure and simulation are shown in Fig. S4 (Supplement 1).
The OLED structure in the simulation adopted the realistic and general top-emitting green phosphorescent OLED structure and the transparent ITO top electrode: ITO (20 nm)/ Ag (150 nm)/ ITO (20 nm)/ HATCN (10 nm)/ NPB (130 nm)/ TCTA (10 nm)/ CBP:Ir(ppy)2(acac) (20 nm) / TPBi (50 nm)/ HATCN (30 nm)/ITO (80 nm). The ITO/Ag/ITO tri-layer serves as a common reflective bottom electrode. HATCN [Dipyrazino[2,3-f:2’,3'-h]quinoxaline2,3,6,7,10,11-hexacarbonitrile] is the hole injection layer (HIL). NPB [N,N’-Di(1-naphthyl)-N,N’-diphenyl-(1,1’-biphenyl)-4,4’–diamine] and TCTA [Tris(4-carbazoyl- 9-ylphenyl)amine] are the hole transport layers (HTLs) [11,22–25]. CBP [4,4’-bis(carbazol-9-yl)biphenyl] doped with 8 wt.% Ir(ppy)2(acac) [bis(2-phenylpyridine) (acetylacetonato) iridium(III)], with a horizontal emitting dipole ratio of 76%, is the typical green phosphorescent emitting layer (EML) [24]. TPBi (2,2’,2’’-(1,3,5-benzinetriyl)-tris(1-phenyl-1-H- benzimidazole)) is the electron transport layer (ETL) [11,23]. HATCN serves as the electron injection layer (EIL) and the buffer layer for the top ITO electrode deposition [23,25–28]. The filler material on top of the OLED in the 3D pixel configuration assumes optical properties (n, k) of NPB due to its suitably high refractive index (nNPB, 1.84 at 520 nm) and its use as the effective filler material in previous experiments of the 3D-RP OLED pixel (cost-effective and patternable by thermal evaporation through the fine metal mask) [17]. The various layer thicknesses in the OLED are chosen to give roughly optimized optical coupling efficiency of OLED internal radiation into the high-index filler region (ηfiller, ∼81.3%, device 1 in Table 1). The calculated emission properties from such OLED (device 1) into the filler (e.g., emission intensity as the function of the wavelength and the initial internal angle θint into the filler, and the spectrally integrated emission intensity as the function of the initial internal angle θint in the filler; see Fig. 1 for definition of θint) are shown in Fig. 2. They are then used to set up ray sources for further multi-scale optical (ray-tracing) simulation for overall light extraction properties of the 3D-RP pixel. In this study, we also find that by inserting a low-index material layer (e.g., LiF in this study) between the OLED (device 1) and the high-index filler can further increase ηfiller, and benefit overall light extraction efficiency from the OLED device to air, ηext. Such an OLED device (device 1 + low-index LiF capping) is named as device 2, with their emission properties into the filler also shown in Fig. 2. With these OLED device configurations, calculated reflection properties (reflection as function of polarization, wavelength and angle of incidence θAOI seen from the filler) for the bottom surface and the bank slope surface of the 3D-RP pixel are shown in Fig. S5 and Fig. S6 (Supplement 1). These properties are used to set up the user-defined coating for the Monte Carlo ray tracing simulation of overall pixel. As a reference for efficiency comparison, ηext’s (optical out-coupling efficiency from OLED to air) for device 1 and device 2 in the conventional AMOLED pixel configuration having no TFE (i.e., Fig. S1(a)) are calculated to be 25.8% and 24.6%, respectively. ηext’s for device 1 and device 2 in the conventional AMOLED pixel configuration having TFE (i.e., Fig. S1(b)) are 20.1% and 20.5%, respectively.
Fig. 2. (a)-(b) Calculated OLED emission intensity Iint(λ, θint)=Is,int(λ, θint)+Ip,int(λ, θint) as the function of the wavelength and the initial internal angle θint into the filler for Device 1 and Device 2. (c) Spectrally integrated emission intensity as the function of the initial internal angle θint in the filler for Device 1 and Device 2 (normalized to 0° intensity).
Table 1. Optical out-coupling efficiencies under various structures and conditions.
3. Impacts of thin film encapsulation on light extraction of the reflective 3D pixel
In the 3D-RP structure without TFE shown in Fig. 1(a), the light incident at the filler-air interface with an angle θF→A larger than its TIR (total internal reflection) critical angle θc,0, (i.e. θF→A>θc,0) would be reflected back to and guided by the filler and then be redirected by the reflective bank surface until θF→A<θc,0 and out-coupled. Thus, for the 3D-RP pixel having OLED-to-filler coupling efficiency ηfiller of 81.3% (device 1), the calculated ηext (overall optical out-coupling efficiency from OLED to air, by multi-scale optical simulation) can be as high as 68% (as listed in Table 1). On the other hand, in the 3D-RP + TFE structure shown in Fig. 1(b), since TFE layers have larger refractive indexes than air, some of the light with the filler-to-TFE incident angle θF→TFE larger than θc,0 (i.e., θF→TFE>θc,0≈θc,1, θc,1 is the TIR critical angle at the interface of the higher-index 5th TFE layer and air), such as rays (2), (4) and (5) in Fig. 1(b), may enter the TFE layer(s) but not be reflected back into the 3D-RP pixel for redirection and out-coupling. Instead, they may be guided by the TFE layer/stack, leading to optical leakage/losses and reduced light extraction. Indeed, for the 3D-RP + TFE having typical TFE stack of: 1st TFE (900 nm)/ 2nd TFE (1 µm)/ 3rd TFE (600 nm)/ 4th TFE (1 µm)/ 5th TFE (300 nm), calculated ηext drops significantly to 43% only (Table 1).
Since the TIR at the interface of higher-index 1st TFE/lower-index 2nd TFE layers and the corresponding light leakage [e.g., ray (4) in Fig. 1(b), not coming back to the 3D-RP region for re-direction and out-coupling] is one main optical loss mechanism, it is intuitive to reduce such loss by reducing the higher-index 1st TFE layer thickness to have more TIR light return to the 3D-RP region for re-direction/out-coupling, as illustrated in Fig. 3. Indeed, as shown in Fig. 4, calculated ηext monotonically increases with reducing the higher-index 1st TFE layer thickness X (while keeping other TFE layers same), reaching 50.4% at X = 0 nm (vs. 43% at X = 900 nm, Table 1). Practically, adopting the minimal X necessary would surely benefit ηext. Although ηext of the 3D-RP + TFE structure can be recovered to some degree by reducing the 1st high-index TFE layer thickness, it is still not satisfactory. To gain more insights for the efficiency loss and possible improvement, we further examine the optical coupling efficiency into each TFE layer, ηi for the ith TFE layer, and the optical out-coupling efficiency from the filler to air as a function of the initial internal angle θint of emission into the filler, φ(θint).
Fig. 3. Schematic illustration of the influence of the 1st TFE layer thickness on light leakage/loss and optical out-coupling.
Fig. 4. ηext of the 3D-RP + TFE pixel structure, incorporating either the OLED Device 1 or OLED Device 2, as a function of the higher-index 1st TFE layer thickness.
Figure 5(a) shows the optical coupling efficiency of device 1 emission into different regions (filler, 1st to 5th TFE layer, air) for the 3D-RP + TFE pixel structure having either 900 nm 1st TFE layer or no 1st TFE layer (i.e., X = 900 or 0 nm). In both cases, there are only slight decreases from η2 to η5, due to Fresnel loss between layers (no TIR). Thus, as long as the light can be coupled into the 2nd TFE layer, it can mostly propagate throughout all remaining TFE layers. In the case having thick (X = 900 nm) 1st TFE layer, ηfiller is roughly the same as η1 due to their similar indexes (nNPB≈nH,TFE). Meanwhile, a significant drop occurs from η1 to η2, due to TIR and corresponding light leakage in the higher-index 1st TFE layer (as illustrated in Fig. 1(b) and Fig. 3). In contrast, in the case having no higher-index 1st TFE layer (X = 0 nm), the efficiency drop from ηfiller to η2 is much minor, clearly indicating effective suppression of light leakage/loss by the 1st TFE layer. However, both cases exhibit another significant efficiency drop from η5 and ηext, mainly caused by the TIR at this interface and corresponding light leakage/loss through the whole TFE stack. Overall, the results of Fig. 5(a) indicate that TIR at the 1st TFE/2nd TFE interface and 5th TFE/air interface and corresponding light leakage/losses in the 1st TFE layer and the whole TFE stack are major mechanisms degrading light extraction efficiency of the 3D-RP + TFE structure.
Fig. 5. The optical coupling efficiencies of OLED emission into the filler (ηfiller), each TFE layer (ηi), and air (ηext) for the 3D-RP + TFE structure, the 3D-RP + ideal ASOF + TFE structure, and the 3D-RP + DBR ASOF + TFE structure incorporating (a) the OLED Device 1 and different higher-index 1st TFE layer thickness, and (b) the OLED Device 2 and different higher-index 1st TFE layer thickness.
Figure 6(a) shows φ(θint) as a function of θint for device 1 in the 3D-RP pixel having no TFE or having TFE (with 1st TFE thickness X = 900 nm or 0 nm). It represents the filler-to-air optical out-coupling efficiency for the light within every 5° θint range (i.e., efficiency for light in 0-5°, 5-10° etc.). The two TIR critical angles θc,1 and θc,2 marked in Fig. 6(a) represent the TIR critical angles for the higher-index 5th TFE layer/air interface (32.7°≈θc,0 of filler-air interface) and that for the higher-index 1st TFE layer/lower-index 2nd TFE layer interface (53.5°). In the 3D-RP pixel having no TFE, the light with θint < θc,0≈θc,1 (∼33°, i.e., within the direct escape cone) can be directly out-coupled with nearly 100% efficiency (except for slight Fresnel loss); the remaining light with θint > θc,0≈θc,1 also show relatively high φ(θint) of ∼70-80%, since they can be reflected back/guided in the filler by TIR and be redirected by the reflective bank for out-coupling (except for metal absorption loss during multiple reflections). The valley around θint∼30°-45° is mainly due to the larger number of reflections and absorption loss. For the 3D-RP + TFE pixel (X = 900 nm or 0 nm), characteristics of φ(θint) can be divided into three regions I to III, corresponding to θint < θc,0≈θc,1∼33°, 33°∼θc,0≈θc,1 < θint < θc,2∼53.5°, and θint > θc,2∼53.5°, respectively. φ(θint) remains high (>90%) for θint < θc,0≈θc,1∼33° (region I), only slightly lower than the no-TFE case due to slightly higher filler-to-TFE reflectance (and corresponding loss) than the filler-to-air reflectance (see Fig. S7 in Supplement 1). In region II (33°∼θc,0≈θc,1 < θint < θc,2∼53.5°), φ(θint) drops substantially (even down to 10-15%) for both cases of X = 900 nm and 0 nm. The light in such θint range can mostly propagate to upper TFE layers (as ray (2) in Fig. 1(b)), but would be reflected by TIR at the 5th TFE layer/air interface. Most of such light would no longer return to the concave 3D-RP pixel region for redirection and out-coupling. Instead, they would be guided by the whole TFE stack, leading to substantial optical leakage/losses. As for the region III (θint > θc,2∼53.5°), while some light (like rays (4) and (5) in Fig. 1(b)) would be lost due to TIR and corresponding light leakage in the TFE layer(s), some light (like rays (3) and (6) in Fig. 1(b)) can still be out-coupled through re-direction by the reflective bank slope, still giving decent φ(θint) for some high θint range. Reducing the higher-index 1st TFE layer thickness (e.g., X = 0 nm) can effectively suppress the loss case of ray (4). It somewhat raises and widens the higher φ(θint) range in the high θint region, and increases ηext with reducing X (Fig. 4 and Table 1). However, even with X = 0, φ(θint) of 3D-RP + TFE at θint > θc,2∼53.5° is still apparently poorer than that of the 3D-RP (no TFE) case. With TFE (either with or without higher-index 1st TFE), some original region-III light (θint > θc,2∼53.5°, like ray (5) in Fig. 1(b)) would be re-directed (by the reflective bank slope) into the region II (i.e., 33°∼θc,0≈θc,1<θF→TFE <θc,2∼53.5°; as illustrated in Fig. 7), propagate through TFE layers, and again see TIR at 5th TFE layer and corresponding light leakage/loss (i.e., suffering non-ideal/poorer φ(θint) in this θint/θF→TFE range). Meanwhile, for the simple 3D-RP structure (no TFE), all light initially with θint > θc,0≈θc,1∼33° would be repeatedly re-directed until entering the direct escape cone (θF→TFE <θc,0≈θc,1∼33°) for out-coupling into air.
Fig. 6. The filler-to-air optical out-coupling efficiency φ(θint) as a function of the initial internal angle θint of OLED emission into the filler for (a) OLED Device 1 and (b) OLED Device 2 in the no-TFE 3D-RP pixel structure, the 3D-RP + TFE structure, the 3D-RP + ideal ASOF + TFE structure, and the 3D-RP + DBR ASOF + TFE structure having different higher-index 1st TFE layer thickness. φ(θint) represents the filler-to-air optical out-coupling efficiency for the light in every 5° θint range (i.e., efficiency for light in 0-5°, 5-10° etc).
Fig. 7. Schematic illustration of some original region-III light (with θint = 70° and 80°) incident at the reflective bank slope being reflected/redirected into the region II light (with θF→TFE = 50° and 40°, respectively).
4. Enhancing light extraction of thin-film encapsulated 3D pixels with angle-selective optical films
Overall, results of Fig. 5(a) and Fig. 6(a) suggest that the drop of ηext upon adding the TFE layers is mainly caused by reduction of φ(θint) over θint > θc,0≈θc,1∼33° (either region II or III). It is associated with TIR at the 1st/2nd TFE layer interface and the 5th TFE/air interface and corresponding light leakage/loss in the TFE layer(s) when the light is coupled from the filler into the TFE layer(s) either directly or by re-direction. As such, one possible remedy for such optical losses is to ensure re-direction of all initial θint > θc,0≈θc,1 light (both regions II and III) into the region-I escape cone (θAOI <θc,0≈θc,1∼33°) before being coupled from the filler into TFE layer(s). As such, the light propagating into TFE would no longer see TIR at various TFE interfaces (and corresponding leakage/losses), ensuring efficient out-coupling from TFE to air. Such characteristics may be realized by inserting some angle-selective optical films (ASOF, having optical transmission and reflection dependent on the filler-to-ASOF incidence angle θF→ASOF) between the filler and the TFE layers to form the 3D-RP + ASOF + TFE pixel structure as shown in Fig. 1(c). Ideally, the ASOF should have 100% optical transmission (0% reflection) for θF→ASOF<θc,0≈θc,1 but 100% optical reflection (0% transmission) for θF→ASOF>θc,0≈θc,1, as shown in Fig. S8 (Supplement 1). Thus, all light initially with θint > θc,0≈θc,1 would be reflected back by the ASOF and be re-directed by the reflective bank slope until θF→ASOF<θc,0≈θc,1, as illustrated by region-II ray (2’) and region-III rays (3’), (4’), (5’) and (6’) in Fig. 1(c). With inserting such ideal ASOF (as a user-defined coating between filler and TFE in the multi-scale optical simulation), φ(θint) of the device 1-embedded 3D-RP + ideal ASOF + TFE pixel structure (with 1st TFE X = 900 nm or 0 nm, Fig. 6(a)) is much improved to resemble that of the simple (no TFE) 3D-RP structure, except for slight Fresnel losses through TFE layers etc. It is no surprise since the optical function of the ideal ASOF is quite similar to TIR at the filler-air interface. With effective re-direction of θint > θc,0≈θc,1 light into θF→ASOF<θc,0≈θc,1 in the 3D-RP + ideal ASOF + TFE pixel structure, the significant coupling efficiency drops seen from η1 to η2 (X = 900 nm case) or from η5 to ηext are substantially reduced (Fig. 5(a)). It leads to much recovered ηext of 63.0% (X = 900 nm)/64.0% (X = 0 nm), significantly higher than 43.0%/50.4% of the 3D-RP + TFE structure and closely approaching 68.0% of the no-TFE 3D-RP structure (Fig. 5(a) and Table 1).
Among possible approaches for implementing such ASOF characteristics [29–33], in view of the structure and fabrication compatibility, the distributed Bragg reflector (DBR) structure, composed of alternating high-index and low-index layers and capable of providing high and angle-dependent reflection at the designated wavelength ranges, is considered in this study [34–q35]. For fabrication simplicity and compatibility, the high-index and low-index materials in the DBR can use the common OLED materials like NPB (nNPB∼1.84 at 520 nm) and LiF (nLiF∼1.37 at 520 nm) or the higher-index TFE material (nH,TFE∼1.85, like SiNx) already in the TFE. Accordingly, as illustrated in Fig. S9 (Supplement 1), the DBR pairs consist of alternating LiF/NPB layers and the higher-index 1st TFE layer (as the last high-index layer of the DBR pairs). Since the filler is also a high-index material (also assume NPB), LiF is adopted as the first DBR layer above the filler. With an index even lower than the lower-index TFE layer, it shall not cause the light leakage or loss problem as the higher-index 1st TFE layer. The thickness of each layer in DBR is set at λT/4n, where n is the refractive index of the corresponding material and λT is the target central wavelength (of the DBR reflection band). Since the 1st TFE layer also serves as the last (high-index) layer of the DBR pairs, the thickness of the 1st TFE layer is adjusted according to λT/4n. In designing the DBR for ASOF here, it is desired to set the target wavelength λT/strong reflection band at wavelengths larger than major OLED emission wavelengths considered (e.g., green emission here) to ensure high optical transmission for OLED emission at the normal direction or small filler-to-ASOF incident angles (e.g., θF→ASOF <θc,0≈θc,1). Fig. S10 (Supplement 1) and Fig. 8 show detailed optical reflection properties [s- and p-polarization reflectance Rs and Rp, average reflectance Ravg = (Rs + Rp)/2] as a function of the wavelength and θAOI for DBR’s having the pair number N of 2-4 and the target central wavelength λT of 600-1100 nm (with corresponding DBR layer thicknesses listed in Table S1 of Supplement 1). As seen in these figures, the general blue shift characteristics of the DBR reflection band with increase of θAOI can be used to raise reflection for OLED emission wavelengths at larger θAOI (e.g., θAOI > θc,0≈θc,1), to reduce direct coupling of such light (larger θAOI) into TFE without re-direction by the 3D-RP structure. Furthermore, the higher index of the incident material (filler) than the exiting material (the 2nd TFE layer, considering the higher-index 1st TFE layer as part of DBR) ensures high reflection for θAOI > θc,2 (filler-to-2nd TFE or 1st TFE-to-2nd TFE TIR critical angle). Overall, although not as perfect as the ideal ASOF characteristics shown in Fig. S8, few-pair DBR with appropriate λT (e.g., N = 2-4, larger λT of 1100 nm) can give angle-dependent reflection spectra somewhat resembling that of ideal ASOF. (Fig. S8).
Fig. 8. Calculated average reflectance Ravg(λ, θAOI) as the function of the wavelength and angle of incidence (AOI) for DBRs having different DBR pairs (2-4 pairs) and different target central wavelength λT (600 nm, 1100 nm).
Figure 9(a) shows the calculated ηext of the 3D-RP + DBR ASOF + TFE pixel structure (with OLED device 1) as a function of N and λT, by inserting the DBR as a user-defined coating between filler and TFE in the multi-scale optical simulation. As λT increases, angle-dependent reflection spectra of DBR more resemble those of ideal ASOF. Thus, ηext of all cases also increases with λT and reaches ∼57.6% (Table 1) under N = 4 and λT = 1100 nm, a significant enhancement over those of the 3D-RP + TFE structure (43.0% and 50.4% for X = 900 nm and 0 nm, respectively). Interestingly, at larger λT (e.g., 1100 nm), one can reduce the DBR pair number (e.g., down to 2 pairs) and yet still achieve similar ηext under larger pair number (e.g., N = 4). Thus, if the overall DBR ASOF thickness would be an issue in the real application, one can adopt the least DBR pairs and yet keep most of the efficiency enhancement. In the 3D-RP + DBR ASOF + TFE structure with DBR λT of 1100 nm (for 2-4 pairs), the higher-index 1st TFE layer, as a part of the DBR, is set at 144 nm (Table S1). We also calculate ηext for the 3D-RP + DBR ASOF + TFE structure upon increasing the 1st TFE layer thickness to 900 nm (i.e., adding an additional 756 nm 1st TFE thickness for better encapsulation property). With such 3.5-pair DBR + 900 nm 1st TFE layer and λT = 1100 nm, ηext retains a decent value of ∼56% (Table 1), close to the full 4-pair DBR (1st TFE layer thickness = 144 nm) case and suggesting that similar efficiency enhancement can be achieved without reducing 1st TFE thickness and sacrificing the encapsulation properties.
Fig. 9. ηext of (a) OLED Device 1 and (b) OLED Device 2 in the 3D-RP + DBR ASOF + TFE structure as a function of the target central wavelength λT, under different DBR pairs (N).
φ(θint)’s of the device 1-embedded 3D-RP + DBR ASOF + TFE pixel structures, having (4 DBR pairs, λT = 1100 nm, 1st TFE X = 144 nm) or (3.5 DBR pairs, λT = 1100 nm, 1st TFE X = 900 nm), are also shown in Fig. 6(a). Although not as ideal as the ideal ASOF case, both DBR ASOF’s (either with X = 900 or 144 nm) yield similarly significant enhancement of φ(θint) over θint > θc,0≈θc,1, in comparison with the 3D-RP + TFE structures. Nevertheless, hardly any φ(θint) enhancement occurs in the 33°(∼θc,0≈θc,1) < θint < 42° range (vs. 3D-RP + TFE case), due to low reflection of the current DBR (and thus low recycling efficiency) for OLED emission wavelengths in this θAOI/θF→ASOF range. In addition, for both DBR ASOF cases, φ(θint) still show significant drops/losses at θint>∼76°, although already postponed to larger angles than both 3D-RP + TFE cases (X = 900 or 0 nm). High Ravg for θAOI (θF→ASOF) > θc,2 (Fig. S10, Fig. 8) ensures nearly complete reflection of θint (θAOI, θF→ASOF) > θc,2 light by the DBR for re-direction. However, not all region-III light would be re-directed into θAOI (θF→ASOF) < θc,0≈θc,1 for efficient out-coupling; some of such reflected light would be re-directed into θc,1 < θAOI (θF→ASOF)<θc,2 region II (as illustrated in Fig. 7) and thus be subjected to losses caused by the non-ideal/poorer φ(θint) in this θAOI (θF→ASOF) range (vs. no-TFE 3D-RP case or 3D-RP + ideal ASOF + TFE case). Overall, the above phenomena result in still significant efficiency drop from η5 to ηext in the 3D-RP + DBR ASOF + TFE cases (Fig. 5(a)). Such situation is even worse in the 3D-RP + TFE cases [i.e., low φ(θint) over the whole θc,1<θint<θc,2 range, in addition to TIR at the 1st TFE/2nd TFE interface and corresponding light leakage/loss], rendering the φ(θint) band lower and narrower in θint > θc,2 and even larger efficiency drop from η5 to ηext.
5. Further enhancing light extraction with low-index OLED capping
With inserting appropriate DBR ASOF, overall ηext of OLED device 1 in the 3D-RP + TFE pixel structure can be effectively raised from 43-50% to 56-58% (Table 1). It is a decent improvement, but one still figures the possibility to further enhance ηext. For instance, considering very similar refractive indexes of OLED organic layers and the filler material, ηfiller (the optical coupling efficiency from OLED to the filler) of 81.3% for device 1 (Table 1) appears not so ideal and one figures whether it can be further increased. In the calculated mode distribution for the device 1 + filler (assumed semi-infinite) configuration, as shown in Fig. 10, in addition to strong coupling of most OLED power into the filler mode (and very weak coupling to surface plasmon (SP) mode), there are two additional sharp peaks associated with the TE and TM waveguided modes (WGTE, WGTM) in the top ITO transparent electrode layer (having a thickness of 80 nm, and a refractive index n of ∼1.92 at 520 nm higher than those of OLED organic layers and the filler as shown in Fig. S4), as detailed in Fig. S11 and Fig. S12 (Supplement 1). These WG modes decrease ηfiller and should be eliminated/reduced for higher ηfiller. It is found that by inserting an appropriate low-index layer between ITO and the filler (see Fig. S13 and Table S2 of Supplement 1 - calculated ηfiller as a function of the LiF thickness), like device 2 having 55 nm LiF, WG modes can be suppressed (Fig. 10, Fig. S11) and ηfiller can be enhanced to 92.5% (Fig. S13, Table 1). The calculated emission properties from device 2 into the filler are shown in Fig. 2 for comparison with device 1, and are used to set up ray sources for further multi-scale optical simulation. Emission patterns in the filler shown in Fig. 2 reveal that compared to device 1, device 2 has a significantly larger power distribution over large θint of 70-90° and smaller power distribution over θc,2 < θint < 70°.
Fig. 10. Comparison of the mode distributions (power dissipation of OLED emission into different modes as a function of kt/ko; kt is the in-plane wavevector and ko is the free-space wavevector) of Device 1 and Device 2.
Table 1, Fig. 4, Fig. 5(b), Fig. 6(b) and Fig. 9(b) show ηext, ηi, and φ(θint) for various pixel structures incorporating device 2. Since both OLED device structures (device 1, 2) give similar optical reflection properties for bottom and bank slope surfaces of the 3D concave reflector (see Fig. S5 and Fig. S6), φ(θint) for a specific pixel type incorporating device 2 is nearly same as φ(θint) of the same pixel type incorporating device 1, as can be seen in Fig. 6. With higher ηfiller of 92.5%, ηext of the device 2 + 3D-RP (no TFE) structure now increases to 76.5% [vs. ηfiller = 81.3%, ηext = 68.0% for device 1 + 3D-RP (no TFE), Table 1]. Yet upon adding TFE layers (1st TFE X = 900 nm) over the 3D-RP pixel, ηext of the device 2 + 3D-RP + TFE structure drops more significantly than the device 1 case, giving ηext = 43.3% almost the same as ηext = 43.0% of the device 1 case. Larger TFE-induced optical losses in the device 2 case are presumably associated with its larger power distribution over large θint of 70-90° in the filler and corresponding partial re-direction into θc,0≈θc,1<θF→TFE < θc,2 having non-ideal/poor φ(θint) [see Fig. 6(b) for φ(θint) of the device 2 case], leading to larger efficiency drops from η1 to η2 and from η5 to ηext seen in Fig. 5(b). As in the device 1 + 3D-RP + TFE case (Fig. 4), reducing the 1st TFE layer thickness X can also monotonically increase ηext for device 2 + 3D-RP + TFE, approaching higher ηext = 55.5% (vs. 50.4% for device 1) at X = 0 nm (also shown in Fig. 4, Table 1). Again, this is mainly attributed to reduced 1st TFE-induced optical losses, as manifested by reduced/eliminated efficiency drop from η1 to η2 (Fig. 5(b)) and improved φ(θint) in θint > θc,2 (Fig. 6(b)). Inserting an ASOF between the filler and TFE (i.e., device 2 + 3D-RP + ASOF + TFE structure) can reduce losses induced by TIR at the 1st TFE/2nd TFE interface and the 5th TFE/air interface. With the ideal ASOF giving improved φ(θint) over θint > θc,1≈θc,0 (Fig. 6(b)) and almost eliminated efficiency drops from η1 to η2 and from η5 to ηext (Fig. 5(b)), ηext can be raised to 70.9% (1st TFE X = 900 nm) and 71.7% (X = 0 nm), as listed in Table 1. With appropriate DBR ASOF designs (e.g., 2-4 DBR pairs, larger λT = 1100 nm, Fig. 9(b)), ηext can be raised to 63.5% (3.5-pair DBR + 1st TFE X = 900 nm) and 65.0% (4-pair DBR including 144 nm 1st TFE in DBR) as shown in Fig. 9(b) and Table 1. Similarly, ηext enhancement with the DBR ASOF is less significant than the ideal ASOF mainly due to less ideal reflection properties over 33°(∼θc,0≈θc,1) < θAOI < 42° (Fig. 8) and corresponding poorer φ(θint) over θint > θc,1≈θc,0 (Fig. 6(b)), leading to still significant efficiency drop from η5 to ηext (Fig. 5(b)).
6. Conclusion
Based on the recently proposed reflective 3D pixel structure (having OLED embedded in the concave 3D reflector and patterned high-index filler) for light extraction enhancement of AMOLED displays, in this work, we conducted simulation studies of influences of thin film encapsulation (TFE) on light extraction of such reflective 3D OLED pixels. Due to TIR (total internal reflection) between TFE layers and TFE/air, TFE induces significant light confinement/leakage/loss and degrades light extraction substantially (by 25-33%). Optimizing TFE layer thicknesses (e.g., thickness of the higher-index 1st TFE layer) can partially reduce such efficiency drop (by a few to 10%), but the improvement is still limited. Accordingly, an additional angle-selective optical film structure (ASOF) is introduced between the reflective 3D pixel (high-index filler) and the TFE to control the angular distribution of the light coupled into the TFE, allowing much more efficient light out-coupling even with TFE. With realistic DBR-based ASOF, light extraction efficiency can be raised by 13-20% (absolute value), while even larger enhancement of 20-27% may be expected with more ideal ASOF. Finally, we find that by inserting a low-index capping between OLED and the high-index filler, the optical coupling efficiency from OLED to the high-index filler and thus the overall pixel light extraction can be further enhanced by 7-8%. The ASOF approach reported here has the general advantage of freeing OLED pixels from optical losses caused by layer architectures above the pixel, rendering larger freedom in designs of OLED displays. It shall also be similarly useful for other self-emitting displays. The results of this study are believed to provide useful insights and guides for developing even more efficient and power-saving AMOLEDs or other self-emitting displays.
Funding
Ministry of Science and Technology, Taiwan (MOST 108-2221-E-002-148-MY3, MOST 111-2221-E-002-031-MY3); Applied Materials.
Acknowledgments
The authors gratefully acknowledge financial support from the Ministry of Science and Technology of Taiwan and Applied Materials, Inc. W.-K. Lee also acknowledges the post-doctoral fellowship from Ministry of Education (MOE) and Ministry of Science and Technology (MOST) of Taiwan. Authors also would like to acknowledge helpful discussion with Prof. Guo-Dong Su and Prof. Hoang Yan Lin during the study.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. M. A. Baldo, D. F. O’Brien, Y. You, A. Shoustikov, S. Sibley, M. E. Thompson, and S. R. Forrest, “Highly efficient phosphorescent emission from organic electroluminescent devices,” Nature 395(6698), 151–154 (1998). [CrossRef]
2. H. Uoyama, K. Goushi, K. Shizu, H. Nomura, and C. Adachi, “Highly efficient organic light-emitting diodes from delayed fluorescence,” Nature 492(7428), 234–238 (2012). [CrossRef]
3. S. Möller and S. Forrest, “Improved light out-coupling in organic light emitting diodes employing ordered microlens arrays,” J. Appl. Phys. 91(5), 3324–3327 (2002). [CrossRef]
4. C. Madigan, M. H. Lu, and J. Sturm, “Improvement of output coupling efficiency of organic light-emitting diodes by backside substrate modification,” Appl. Phys. Lett. 76(13), 1650–1652 (2000). [CrossRef]
5. C.-J. Yang, S.-H. Liu, H.-H. Hsieh, C.-C. Liu, T.-Y. Cho, and C.-C. Wu, “Microcavity top-emitting organic light-emitting devices integrated with diffusers for simultaneous enhancement of efficiencies and viewing characteristics,” Appl. Phys. Lett. 91(25), 253508 (2007). [CrossRef]
6. J. J. Shiang, T. J. Faircloth, and A. R. Duggal, “Experimental demonstration of increased organic light emitting device output via volumetric light scattering,” J. Appl. Phys. 95(5), 2889–2895 (2004). [CrossRef]
7. C.-C. Liu, S.-H. Liu, K.-C. Tien, M.-H. Hsu, H.-W. Chang, C.-K. Chang, C.-J. Yang, and C.-C. Wu, “Microcavity top-emitting organic light-emitting devices integrated with diffusers for simultaneous enhancement of efficiencies and viewing characteristics,” Appl. Phys. Lett. 94(10), 103302 (2009). [CrossRef]
8. H.-W. Chang, J. Lee, S. Hofmann, Y.-H. Kim, L. Müller-Meskamp, B. Lüssem, C.-C. Wu, K. Leo, and M. C. Gather, “Nano-particle based scattering layers for optical efficiency enhancement of organic light-emitting diodes and organic solar cells,” J. Appl. Phys. 113(20), 204502 (2013). [CrossRef]
9. Y. R. Do, Y.-C. Kim, Y.-W. Song, C.-O. Cho, H. Jeon, Y.-J. Lee, S.-H. Kim, and Y.-H. Lee, “Enhanced Light Extraction from Organic Light-Emitting Diodes with 2D SiO2/SiNx Photonic Crystals,” Adv. Mater. 15(14), 1214–1218 (2003). [CrossRef]
10. S. Jeon, J.-H. Jeong, Y. S. Song, W.-I. Jeong, J.-J. Kim, and J. R. Youn, “Vacuum nano-hole array embedded organic light emitting diodes,” Nanoscale 6(5), 2642–2648 (2014). [CrossRef]
11. C.-Y. Chen, W.-K. Lee, Y.-J. Chen, C.-Y. Lu, H. Y. Lin, and C.-C. Wu, “Enhancing Optical Out-Coupling of Organic Light-Emitting Devices with Nanostructured Composite Electrodes Consisting of Indium Tin Oxide Nanomesh and Conducting Polymer,” Adv. Mater. 27(33), 4883–4888 (2015). [CrossRef]
12. S. M. Jeong, F. Araoka, Y. Machida, K. Ishikawa, H. Takezoe, S. Nishimura, and S. Suzaki, “Enhancement of normally directed light outcoupling from organic light-emitting diodes using nanoimprinted low-refractive-index layer,” Appl. Phys. Lett. 92(8), 083307 (2008). [CrossRef]
13. W. H. Koo, S. M. Jeong, F. Araoka, K. Ishikawa, S. Nishimura, T. Toyooka, and H. Takezoe, “Light extraction from organic light-emitting diodes enhanced by spontaneously formed buckles,” Nat. Photonics 4(4), 222–226 (2010). [CrossRef]
14. Q.-D. Ou, L. Zhou, Y.-Q. Li, S. Shen, J.-D. Chen, C. Li, Q.-K. Wang, S.-T. Lee, and J.-X. Tang, “Extremely Efficient White Organic Light-Emitting Diodes for General Lighting,” Adv. Funct. Mater. 24(46), 7249–7256 (2014). [CrossRef]
15. Y.-J. Chen, W.-K. Lee, Y.-T. Chen, C.-Y. Lin, S.-W. Wen, M. Jiao, G.-D. Su, H. Y. Lin, R. J. Visser, B. L. Kwak, C.-C. Chen, W.-Y. Lin, S. Wang, C.-P. Chang, and C.-C. Wu, “A Vision toward Ultimate Optical Out-Coupling for Organic Light-Emitting Diode Displays: 3D Pixel Configuration,” Adv. Sci. 5(10), 1800467 (2018). [CrossRef]
16. W-K Lee, Y-T Chen, S-W Wen, P-H Liao, M-C Lee, T-S Hsu, Y-J Chen, G-D Su, H. Y. Lin, C-C Chen, W-Y Lin, L. Xu, G. Yu, B. L. Kwak, R. J. Visser, and C-C Wu, “Three-dimensional pixel configurations for optical outcoupling of OLED displays—optical simulation,” J. Soc. Inf. Disp. 27(5), 273–284 (2019). [CrossRef]
17. Y-T Chen, S-W Wen, P-H Liao, W-K Lee, C-C Lee, C.-W. Huang, Y.-H. Yang, K.-C. Lin, C.-J. Chang, G.-D. Su, H.-Y. Lin, C.-C. Chen, W-Y Lin, B. L. Kwak, R. J. Visser, and C-C Wu, “Reflective 3D pixel configuration for enhancing efficiency of OLED displays,” Org. Electron. 103, 106451 (2022). [CrossRef]
18. J.-S. Park, H. Chae, H. K. Chung, and S. I. Lee, “Thin film encapsulation for flexible AM-OLED: a review,” Semicond. Sci. Technol. 26(3), 034001 (2011). [CrossRef]
19. P. van de Weijer, P. C. P. Bouten, S. Unnikrishnan, H. B. Akkerman, J. J. Michels, and T. M. B. van Mol, “High-performance thin-film encapsulation for organic light-emitting diodes,” Org. Electron. 44, 94–98 (2017). [CrossRef]
20. Y. C. Han, E. Kim, W. Kim, H. G. Im, B. S. Bae, and K. C. Choi, “A flexible moisture barrier comprised of a SiO2-embedded organic-inorganic hybrid nanocomposite and Al2O3 for thin-film encapsulation of OLEDs,” Org. Electron. 14(6), 1435–1440 (2013). [CrossRef]
21. D. Yu, Y. Q. Yang, Z. Chen, Y. Tao, and Y. F. Liu, “Recent progress on thin-film encapsulation technologies for organic electronic devices,” Opt. Commun. 362, 43–49 (2016). [CrossRef]
22. Y.-H. Huang, C.-Y. Lu, S.-T. Tsai, Y.-T. Tsai, C.-Y. Chen, W.-L. Tsai, C.-Y. Lin, H.-W. Chang, W.-K. Lee, M. Jiao, and C.-C. Wu, “Enhancing light out-coupling of organic light-emitting devices using indium tin oxidefree low-index transparent electrodes,” Appl. Phys. Lett. 104(18), 183302 (2014). [CrossRef]
23. J.-B. Kim, J.-H. Lee, C.-K. Moon, S.-Y. Kim, and J.-J. Kim, “Highly Enhanced Light Extraction from Surface Plasmonic Loss Minimized Organic Light-Emitting Diodes,” Adv. Mater. 25(26), 3571–3577 (2013). [CrossRef]
24. G. Li, T. Fleetham, E. Turner, X.-C. Hang, and J. Li, “Highly Efficient and Stable Narrow-Band Phosphorescent Emitters for OLED Applications,” Adv. Opt. Mater. 3(3), 390–397 (2015). [CrossRef]
25. S.-K. Kwon, E.-H. Lee, K.-S. Kim, H.-C. Choi, M. J. Park, S. K. Kim, R. Pode, and J. H. Kwon, “Efficient micro-cavity top emission OLED with optimized Mg:Ag ratio cathode,” Opt. Express 25(24), 29906–29915 (2017). [CrossRef]
26. J.-H. Lee, S. Lee, J.-B. Kim, J. H. Jang, and J.-J. Kim, “A high performance transparent inverted organic light emitting diode with 1,4,5,8,9,11-hexaazatriphenylenehexacarbonitrile as an organic buffer layer,” J. Mater. Chem. 22(30), 15262 (2012). [CrossRef]
27. S. Lee, J.-H. Lee, J.-H. Lee, and J.-J. Kim, “The Mechanism of Charge Generation in Charge- Generation Units Composed of p-Doped Hole-Transporting Layer/HATCN/n-Doped Electron-Transporting Layers,” Adv. Funct. Mater. 22(4), 855–860 (2012). [CrossRef]
28. Y.-L. Wu, C.-Y. Chen, Y.-H. Huang, Y.-J. Lu, C.-H. Chou, and C.-C. Wu, “Highly efficient tandem organic light-emitting devices utilizing the connecting structure based on n-doped electron-transport layer/HATCN/hole-transport layer,” Appl. Opt. 53(22), E1–E6 (2014). [CrossRef]
29. Y. Shen, C.-W. Hsu, Y.-X. Yeng, J. D. Joannopoulos, and M. Soljačić, “Broadband angular selectivity of light at the nanoscale: Progress, applications, and outlook,” Appl. Phys. Rev. 3(1), 011103 (2016). [CrossRef]
30. Y. Shen, D. Ye, I. Celanovic, S. G. Johnson, J. D. Joannopoulos, and M. Soljačić, “Optical Broadband Angular Selectivity,” Science 343(6178), 1499–1501 (2014). [CrossRef]
31. Y. Shen, D. Ye, L. Wang, I. Celanovic, L. Ran, J. D. Joannopoulos, and M. Soljačić, “Metamaterial broadband angular selectivity,” Phys. Rev. B 90(12), 125422 (2014). [CrossRef]
32. E. D. Kosten, J. H. Atwater, J. Parsons, A. Polman, and H. A. Atwater, “Highly efficient GaAs solar cells by limiting light emission angle,” Light: Sci. Appl. 2(1), e45 (2013). [CrossRef]
33. R. E. Hamam, I. Celanovic, and M. Soljačić, “Angular photonic band gap,” Phys. Rev. A 83(3), 035806 (2011). [CrossRef]
34. T. Fleetham, J. Ecton, G. Li, and J. Li, “Improved out-coupling efficiency from a green microcavity OLED with a narrow band emission source,” Org. Electron. 37, 141–147 (2016). [CrossRef]
35. K.-C. Lin, W-K Lee, B.-K. Wang, Y.-H. Lin, H.-H. Chen, Y.-H. Song, Y.-H. Haung, L.-W. Shih, and C-C Wu, “Modified distributed Bragg reflector for protecting organic light-emitting diode displays against ultraviolet light,” Opt. Express 29(5), 7654–7665 (2021). [CrossRef]
