1. Introduction

In the past few years, organic light-emitting diode (OLED) displays have become one of the most important display technologies and have gained wide applications in small-to-medium size applications, such as mobile or wearable devices. Thanks to the development of reliable materials and encapsulation technologies, the reliability of OLED displays has been acceptable for many indoor uses for years. Aside from general indoor and consumer applications, OLEDs are being extended to outdoor applications, such as wearable devices and car electronics. However, in current OLED display architectures, OLED displays would still suffer deterioration under long-term outdoor uses and sunlight exposure, due to degradation induced by the ultraviolet (UV) light (with wavelength λ<400 nm) and the high energy visible (HEV) light (with wavelength at 400-450 nm) [1–15]. According to the previous reports, UV/HEV light might cause chemical bond dissociation of OLED materials [5–7,14], or deteriorate metal/organic contacts [8,9,14]. The by-products generated by UV/HEV-induced photochemical reaction may also yield luminescent quenchers, nonradiative recombination centers, and deep charge traps in the emission layer [14], leading to degradation of device performance. Besides, UV/HEV irradiation might also influence some electron injection layers/materials (EIL) of OLEDs and/or induce migration of some EIL materials into the organic layer [11]. Such UV/HEV-induced degradation would heavily affect the OLED panels’ performance, such as brightness, gray scale, color performance, lifetime, pixel shrinkage, and luminance-current-voltage(L-I-V) characteristics shift, etc. [15] Thus, an effective and large-area anti-UV/HEV technique and structure, that can protect OLEDs from UV/HEV-induced degradation and meanwhile maintain display performance, is strongly desired.

In order to consider how to integrate the UV/HEV-resistant (or blocking) structure into OLED panels without influencing their optical properties, it is rational to first consider the actual OLED panel architecture. As shown in Fig. 1(a), current OLED display panels are generally comprised of four different parts: the active-matrix backplanes, top-emitting OLEDs, the encapsulation layer, the circular polarizer, and the cover lens, etc. The active matrix backplanes are usually fabricated with silicon or metal oxide thin film transistors (TFTs) for driving OLEDs. Encapsulation is used to protect OLED displays from oxygen/moisture degradation, and is usually composed of a cover glass or thin-film layers of organic and inorganic materials. The circular polarizer is needed to reduce the ambient light reflection associated with metal electrodes and metal lines in OLED displays, and thus the contrast ratio of OLED panels can be substantially enhanced. Meanwhile, cover lens on the top of OLED panels, usually made of tempered glasses, provide the outmost protection (such as scratch resistance, anti-fouling, etc.) of panels and some other optical functionalities (such as anti-glare, anti-reflection, etc.). Accordingly, the UV/HEV resistance functionality may be incorporated in or on these different parts of OLED panels. First, direct realization of UV/HEV resistance functionality in OLEDs is less feasible since the thicknesses of the organic layers are in the nm scale, introducing additional layers directly on top of OLEDs could substantially impact optical performances of OLED displays, such as color, efficiency, and viewing angles, etc. Polarizers may provide partial UV/HEV resistance (see transmission spectrum of the circular polarizer in Fig. 1(b)), but the UV/HEV irradiation may also degrade polarizers themselves as they are often made of organic materials. Thus, for indoor or short-term outdoor applications, the polarizer may be sufficient to protect OLEDs. Yet, further protection is needed for long-term outdoor applications. Therefore, in order to prevent both polarizers and OLEDs from UV/HEV degradation, cover lens may be the ideal location to integrate UV/HEV resistance structures. Since the distance between the active OLEDs and the cover lens is much longer than the OLED coherence length, the additional optical functional layers shall not cause the extra interference with the already designed OLEDs or impact emission properties, making it easier to design UV/HEV resistance structure. In this study, we investigate designs of UV/HEV-resistant (or blocking) thin-film stack structures on OLED cover lens using more robust inorganic materials for protecting OLED displays from UV/HEV degradation.

 

Fig. 1. (a) Schematic structure of the typical top-emitting OLED panel. (b) Measured transmission spectrum of a typical circular polarizer.

Download Full Size | PDF

2. Strategies to design the UV-blocking optical structures

Considering the protection against UV/HEV exposure without influencing/lowering R/G/B color performance of OLED displays, the transmission spectrum T(λ) of the UV/HEV-blocking structure is targeted to have a low (ideally zero) transmittance around and below 410-420 nm (UV-/HEV-blocking region), and meanwhile possess high transmittance for λ>440-450 nm (visible transmissive region). Accordingly, the transmission spectrum of the UV/HEV-blocking structure would have a sharp transition between the blocking region and transmissive region. For practical applications, the UV/HEV-blocking structure should be kept as simple and cost-effective as possible, and complicated and costly materials and fabrication should be avoided. Finding a single inorganic material that can simultaneously provide effective UV/HEV blocking (absorption), high visible transmission, and also steep transmission transition at exact wavelength ranges is indeed quite difficult, if not impossible. As such, combining UV absorption materials with certain proper optical structures deems necessary to fulfill all these requirements. The role of the optical structures is then to reflect the UV-/HEV light that cannot be effectively absorbed by the UV-/HEV-absorbing materials and to provide a sharp spectral transition between UV/HEV and visible regions. One rational approach is to construct the DBR (distributed Bragg reflector) structure with alternating high-index (n_H)/low-index (n_L) UV absorbing and dielectric materials, so that very low T in UV, precise control of T transition/drop around the near-UV/deep-blue wavelengths, and very high T in the visible can be achieved at the same time. DBR is a simple and powerful optical structure that can provide nearly 100% reflectivity around target wavelength (λ_T) and sharp transmission transition/drop. Accordingly, DBR is usually applied in optical components like perfect mirrors at certain wavelength ranges in Fabry-Perot spectroscopy or in optical fibers [16–20]. Conventional DBR is usually composed of pairs of high-index and low-index materials and the thickness of each layer is set at the corresponding quarter wavelength (λ/4n, n is refractive index) of the targeted resonant wavelength as illustrated in Fig. 2(a) [21,22]. To minimize the optical transmittance (i.e., full blocking) of the DBR, the electric field should be minimized or close to zero at the exiting interface. Since the total optical thickness of a high index/low index pair in DBR is half of λ_T, the electric field at the interface between two adjacent pairs should also be close to zero (i.e., the node). Meanwhile, the antinode would occur at the high-index/low-index interface in a pair as illustrated in Fig. 2(a). As the index difference becomes larger, the higher reflection induced by the high-index/low-index interface would induce a larger bandwidth for high reflection. Furthermore, as the number of DBR pair increases, the transition edge of the reflection band would become sharper [23,24].

 

Fig. 2. Schematic illustration of (a) conventional DBR design, (b) 1^st design of the modified DBR structure, and (c) 2^nd design of the modified DBR structure. The red curve in each panel represents the schematic distribution of the electric field intensity (magnitude) at the resonant wavelength.

Download Full Size | PDF

In this study, TiO₂ and SiO_x prepared by RF sputtering were adopted as the high-index and the low-index materials, respectively, for the DBR design. Their optical properties are shown in Fig. 3(a)-(d). Figure 3(a) and Fig. 3(b) show the refractive indexes (n) and extinction coefficients (k) for TiO₂ and SiO_x, respectively, as characterized by spectroscopic ellipsometry. Sputtered TiO₂ exhibits a high refractive index above 2.4 over the 300-800 nm range and even higher n of >2.5 for HEV to UV wavelengths; in contrast, SiO_x exhibits a relatively low n of ∼1.5 over the spectral range of 300-800 nm. The index difference between TiO₂ and SiO_x is large enough to form DBR with a broad reflection bandwidth. Figure 3(c) and Fig. 3(d) show the transmittance (T), reflectance (R), and absorptance (A) spectra for a 30-nm-thick TiO₂ film and a 60-nm-thick SiO_x film on glass substrates, respectively. Both film thicknesses are roughly around λ/4n of λ_T. As seen in Fig. 3(a) and 3(c), TiO₂ not only possesses a high refractive index but also absorbs the UV light below 350 nm. Therefore, with appropriate DBR designs, it is possible to achieve the UV/HEV-blocking function by reflecting HEV light and absorbing UV light when combining high-index/UV-absorbing TiO₂ and low-index SiO₂.

 

Fig. 3. Refractive indexes (n) and extinction coefficients (k) of (a) TiO₂, and (b) SiO_x. Measured transmittance (T)), reflectance (R), and absorptance (A) spectra of (c) a 30-nm TiO₂ film and (d) a 60-nm SiO_x film deposited on glass substrates.

Download Full Size | PDF

Based on the transfer matrix method (TMM) and geometrical optics [25], transmittance, reflectance, and absorptance (T-R-A) spectra of conventional DBR structures (i.e., like Fig. 2(a)) consisting of 3-8 pairs of λ/4n-thick TiO₂ (32.6 nm)/SiO_x (62.1 nm) with the target reflection resonant wavelength λ_T around 370 nm were calculated. Simulation results for 6-8 DBR pairs are shown as the black curves in Fig. 4(a)-(i), while those for 3-5 DBR pairs are shown in Fig. S1(a)-S1(i) in the Supplement 1. To compare the transmittance transition gradient of the UV-blocking structures, the wavelengths at 10% transmittance and 90% transmittance (λ_L and λ_H, respectively, as illustrated in Fig. 4(a)-(c)), the difference between λ_L and λ_H (Δλ=λ_H-λ_L), and the transmittance at 460 nm (around the peak wavelength of blue pixel of the state-of-the-art OLED panel used in iPhone) for various DBR structures are also defined and are summarized in Table S1 in the Supplement 1. These results show that upon increasing the pair number, λ_L would increase (thus giving better UV-/HEV-blocking capability), λ_H would decrease (thus giving the wider transmissive region), and Δλ would drop correspondingly (thus giving the steeper transmittance transition edge). These results suggest that ≥6 DBR pairs would be necessary to achieve the sharp enough transmission edge (e.g., Δλ of 30-40 nm around the 410-450 nm range) and relatively high peak transmission in the visible range. As shown in Fig. 4(a)-(c), conventional DBR structures having 6-8 pairs can achieve nearly perfect UV/HEV-blocking properties by reflecting the 350-420 nm light (Fig. 4(d)-(f)) and absorbing the UV light below 350 nm (Fig. 4(g)-(i)). Meanwhile, fewer pairs would lead to less sharp transmittance edges, risking lower transmission for blue OLEDs and poorer UV-blocking capability in the near-UV range, whereas more pairs would apparently increase fabrication complexity and cost, a problem often seen in commercial optical coatings using tens of DBR pairs.

 

Fig. 4. (a)-(c) Calculated transmittance spectra of 6, 7, and 8-pair DBRs. (d)-(f) Calculated reflectance spectra of 6,7,8-pair DBRs. (g-)-(i) Calculated absorptance spectra of 6, 7, 8-pair DBRs.

Download Full Size | PDF

Aside from blocking the transmission at the target (UV) resonant wavelength, the DBR structure would also cause constructive interference and undesired strong ripples in T/R spectra at visible wavelengths, which are detrimental to OLED display performances such as lower transmittance/luminance/efficiency and colors, etc. To reduce or even eliminate such (chromatic) ripples in the visible region, two modified DBR designs, labeled as 1^st design and 2^nd design in Fig. 2(b)–2(c), were further devised. Generally, strong spectral ripples at longer wavelengths are associated with the higher-order interference caused by the high-n/low-n structure. Accordingly, in the 1^st modified DBR design (Fig. 2(b)), instead of the λ/4n thickness for each layer in the conventional DBR, the thicknesses of the high-index and low-index layers are made slightly smaller and larger than quarter resonant wavelength λ/4n and meanwhile keep the total optical thickness of one pair same as the conventional design (i.e., half of λ_T). Thicknesses of TiO₂ (27 nm) and SiO_x (73 nm) are determined by minimizing the undulation in transmittance in the visible range and keeping the transmittance edge sharp. By locating the standing wave peak slightly away from the high-n/low-n interface, the 1^st modified DBR design is aimed to weaken collective reflection/interference at higher-order modes but to keep the similar resonant wavelength. As shown in Fig. 4(a)–4(c) for 6-8 DBR pairs (also see Fig. S1(a)-S1(i) in Supplement 1 for 3-5 DBR pairs), the transmittance edges and peaks of the 1^st modified DBR design occur at almost same wavelengths as the conventional design, but the transmittance undulation in the transmission region is noticeably reduced.

Although the transmission undulation is to some degree reduced by the 1^st modified DBR, it is, however, still not good enough for real uses. Ideally, a constant/flat T of >90% over the whole visible range is desired. To further remove transmittance ripples in the visible range, it is necessary to weaken/eliminate the constructive interference of higher-order modes (at λ>λ_T), while still keeping the UV/HEV blocking properties. The hint for the solution may be obtained by examining evolution of DBR properties upon stacking more pairs, as seen in Fig. 4 and also Fig. S1 in Supplement 1. As the phases of the reflection coefficients at interfaces of a single DBR pair is either 0^o or180^o, the stacking of DBR pairs would mainly enhance the magnitude of the total reflection coefficient without influencing the phase. Consequently, as seen in Fig. 4(d)–4(f) and Fig. S1(d)-S1(f) in Supplement 1, upon increasing the pair number, the reflection band at the target resonant wavelength are almost the same in both conventional and 1^st modified DBRs. Differently, the reflection side bands would gradually blue-shift with stacking more pairs (Fig. S1(a)-S1(f) in Supplement 1 and Fig. 4(a)–4(f)), since the side bands are induced by the whole structure, not just by a single pair. Inspired by these observations, it is intuitive to devise the 2^nd modified DBR design as in Fig. 2(c): (i) except the first and the last DBR pair in the overall 3-8 DBR pairs, the middle 1-6 pairs keep the same thicknesses as in the 1^st design to block the light around the resonant wavelength; (ii) adjust the total/optical thicknesses/layer thicknesses of the first and the last pair to deviate from those of middle pairs and to destruct the overall reflection in side bands. The layer thicknesses of the first and last pairs were adjusted individually and sequentially by fixing the thicknesses of other layers and minimizing the undulation of the transmittance in the visible range. The roughly optimized overall structure of the 2^nd design then consisted of the first pair of TiO₂ (15 nm)/SiO_x (65 nm) and the last pair of TiO₂ (5 nm)/SiO_x (40 nm), while the layer thicknesses in the middle pairs/layers were the same as those in the 1^st design (TiO₂/SiO_x:27 nm/73 nm). As seen in Fig. 4(a)–4(i) for 6-8 pairs, reflection side bands and transmission ripples in the visible region are nearly eliminated, while steep transmittance edges are maintained and high T of >91% (approaching T∼92% of glass) is obtained over the whole visible range. In particular, high transmittance of >92% at the typical blue OLED pixel wavelength (460 nm) and λ_L of >410 nm are obtained with the pair number of >5 (see Table S1 in Supplement 1). To make sure that the UV-blocking structure would not influence the emission characteristics of the OLED display, in Fig. S2 of Supplement 1, measured electroluminescence (EL) spectra of R/G/B pixel OLED devices in the state-of-the-art iPhone panel are compared with calculated transmission spectra of 2^nd modified DBR structures having 6 and 7 pairs. It is seen that even EL of the blue pixel safely falls within the high and flat transmittance region, making the 2^nd modified DBR design suitable for UV-blocking and OLED display applications.

To further investigate the detailed mechanism in UV-blocking DBR designs, Fig. 5(a) and 5(b) show the simulated total electric field intensities (|E|) at different wavelengths for the 6-pair conventional DBR and 2^nd modified DBR, respectively, with the incident intensity being set as unity. As seen in Fig. 5(a), the UV light at 300 nm would be effectively absorbed by the UV-absorbing/conventional DBR and the electric field at the exiting interface reaches a very low level (<10⁻²), giving extremely low transmittance. On the other hand, Fig. 5(b) shows that the 2^nd design can also absorb the 300-nm UV light but the electric field intensity at the exiting interface is slightly larger than the conventional DBR since the total thickness of TiO₂ is smaller. In spite of the slightly larger exiting field in the 2^nd modified DBR, the UV/HEV transmission in Fig. 4(a)-4(c) is still sufficiently low for UV-blocking applications. For the target (reflection) resonant wavelength (370 nm), Fig. 5(a) clearly manifests that in the conventional DBR, the nodes in the field intensity occur at interfaces between two adjacent pairs and the antinodes occur at the high-n/low-n interface in the pair. For the 2^nd modified DBR in Fig. 5(b), the electric field profile (for 370 nm) in the middle pairs are similar to that of the conventional DBR structure. It indicates in this structure, the middle four pairs are also similar to the conventional DBR and are responsible to fulfill the HEV-blocking properties. Finally, for the 493-nm (visible) light, which is the peak wavelength of the first ripple in the conventional DBR (Fig. 4(a)), the peak-to-valley modulation of the total electric field in the conventional DBR (Fig. 5(a)) is larger than that in the 2^nd modified DBR (Fig. 5(b)). Since the total field is composed of waves propagating toward + z and -z, larger peak-to-valley modulation indicates that more similar intensities between + z and -z propagating waves (stronger collective reflection). Meanwhile, the low peak-to-valley modulation in the 2^nd modified DBR shows that adjusting the layer thicknesses in the first pair and the last pair can effectively suppressing the intensity of the -z propagating (reflecting) wave and thus substantially enhance the overall transmittance.

 

Fig. 5. Simulated distributions of total electric field intensities in the 6-pair DBRs for wavelengths of 300 nm, 370 nm, and 493 nm: (a) the conventional DBR structure, (b) 2^nd modified DBR design.

Download Full Size | PDF

Following the 2^nd modified DBR design strategy, Fig. 6 shows calculated T-R-A spectra for 6-pair 2^nd modified DBR structures with varied λ_T of 350 nm to 390 nm (while more detailed properties are summarized in Table S2 of Supplement 1). It is seen that boundaries/edges of their transmittance and reflection spectra can be readily blue-shifted with reducing the target resonant wavelength λ_T (simply by adjusting the layer thicknesses), if desired (e.g., when deeper blue OLED pixels is desired and adopted and thus a different balance between blue EL and UV blocking is required). This (easy and precise tuning of the transition wavelength and sharpness of the transmittance spectrum) perhaps is a particular advantage and convenience of using the DBR structure for UV blocking, compared to other approaches (e.g., simply using absorption properties of materials). From these results, it is also clear that for a blue OLED pixel with EL peaking around 460 nm, λ_T of 370 nm gives a reasonable balance between blue EL and UV blocking.

 

Fig. 6. Calculated (a) transmittance spectra, (b) reflectance spectra, and (c) absorptance spectra for 2^nd modified DBR structures with 6 pairs and λ_T=350 nm, 360 nm, 370 nm, 380 nm, and 390 nm.

Download Full Size | PDF

It is also worthy of stressing other advantages of the current modified DBR design compared to conventional DBR designs. Although DBR is a common technology, it is still generally difficult to achieve all the required optical properties for the current purpose, such as effective UV-/HEV-blocking capability, sharp transition of the transmittance spectrum, and high and flat transmittance in the visible light region without influencing the emission characteristics of the display. In the conventional DBR structure, usually the low index difference between the high-/low-index materials may be used to achieve high and flat transmittance (i.e., small modulation ripples) out of the target reflection band. However, tens of DBR pairs may be needed to achieve sharp transmittance transition and complete reflection in the reflection band (see Fig. S3(a) of Supplement 1). In addition, the narrow reflection band cannot cover the whole HEV or UV light region. On the other hand, with large index difference, the sharp transmittance transition and complete reflection around the target wavelength range can be achieved with much fewer DBR pairs (see Fig. S3(b) of Supplement 1), but strong ripples occurring in the transmission region (or strong reflection side bands) would strongly affect display color and efficiency and are not acceptable for display applications. In this work, based on the conventional high-index-contrast DBR structure, with carefully adjusting layer thicknesses of high-index and low-index layers to form the modified DBR structure, wide-band and full reflection in the target wavelength range, sharp transmittance transition, precise control of the transition location, and high and flat transmittance in the visible region can be all simultaneously achieved with only few DBR pairs. Such simple and practical optical thin-film design strategy is not commonly reported. In addition to display application (like this work), such optical thin-film design may also find wide use in other areas requiring high-performance optical thin-film coatings.

3. Fabrication and characterization of UV-blocking structures

Promising UV/HEV blocking DBRs having the 2^nd modified DBR design and 6-7 DBR pairs were then fabricated and tested for confirming the simulation and design. Thin film stacks were deposited on glass substrates (Corning Eagle 2000) by RF sputtering using TiO₂ (99.99% purity) and SiO_x (99.999% purity) targets. The sputtering conditions were fixed at RF power of 150 W, Ar flow rate of 40 sccm and working pressure of 1.5 mTorr. Before the actual deposition, all targets were pre-sputtered for cleaning target surfaces. After alternately depositing TiO₂ and SiO_x for 6-7 times, transmittance spectra of DBR samples were measured by UV-Vis spectrophotometer (V-670, Jasco). Figure 7(a) and Fig. 7(b) show the cross-section SEM images of the fabricated 6- and 7-pair 2^nd modified DBR structures, respectively, clearly confirming successful fabrication of precisely controlled DBR stacks.

 

Fig. 7. Cross-section SEM images of the UV-blocking, 2^nd modified DBRs: (a) 6 pairs, and (b) 7 pairs.

Download Full Size | PDF

 Figure 8(a) and Fig. 8(b) show the transmittance spectra of the fabricated 6- and 7-pair 2^nd modified DBR structures, respectively. The high and rather constant transmittance in the visible range and the rather sharp transmittance transition around 420-450 nm (HEV) are consistent with the simulation and design, confirming the usefulness of the DBR design strategy for OLED display applications. Photos in the insets of Fig. 8(a)–8(b) show photos of the corresponding DBR samples taken with a color image background (top) or a black image background (bottom), both under the normal lighting condition. These photos unambiguously demonstrate un-degraded color images through the UV/HEV-blocking structures and their low visible reflection under the black background, not compromising display viewing performances at all. In Fig. S4 of Supplement 1, original electroluminescence (EL) spectra of R/G/B pixel OLEDs in the state-of-the-art iPhone OLED panel are compared with those through the 6-pair and 7-pair 2^nd modified DBR structures. Nearly same EL spectra and Commission internationale de l'éclairage (CIE) 1931 color coordinates (Table S2 of Supplement 1) confirm that the current UV/HEV-blocking DBR structures hardly influence EL spectra and colors of current OLED displays. As the transmittance in the EL color range is nearly same as the conventional cover glass (Fig. 1(a)), it would hardly affect the EL efficiency when assembled with the OLED display panel either.

 

Fig. 8. Transmittance spectra of the UV-blocking, 2^nd modified DBRs: (a) 6 pairs, and (b) 7 pair. Insets in (a) and (b) show photos of the corresponding DBR samples taken under a color image background and the normal lighting condition (top) and under a black background and the normal lighting condition (bottom).

Download Full Size | PDF

4. Solar test and OLED’s characterization

Display makers develop OLED panels using I-V characteristics of OLEDs to determine panel’s gray scale, brightness, color performance, and power consumption, etc. Thus, if the OLEDs’ I-V characteristics are changed by solar irradiation/degradation, the panel’s display performances would deviate/degrade from the initially set specifications. One typical way to test the OLED panel’s resistance against sunlight degradation is to measure the voltage shifts (ΔV) at a specified current density vs. solar exposure cycles, following the rigorous IEC60068-2-5 standard with specified spectral light source intensities in Table 1. Figure 9(a)–9(c) show the three solar exposure test architectures conducted in this study: (a) architecture A having no polarizer and no UV-blocking structure, (b) architecture B having polarizer but no UV-blocking structure (i.e., the typical OLED panel structure), and (c) architecture C having both polarizer and UV-blocking DBR structure (6 and 7 DBR pairs, 2^nd modified DBR in this study) on the cover lens. In this work, red top-emitting OLEDs obtained from AU Optronics Corporation were used for the solar irradiation tests; the organic materials and OLEDs were those used in real commercial wearable devices. During each solar exposure test cycle, samples were exposed to the simulated sunlight for 8 hours under a chamber temperature of 42°C and then stood in dark state for 16 hours at 25°C. Such test cycles were then repeated. Between repeated test cycles, the voltage shifts (ΔV) of OLEDs under the current density of 10 mA/cm² were monitored by every five rounds.

 

Fig. 9. Three solar testing architectures with top-emitting OLEDs: (a) A having no polarizer and no UV-blocking structure, (b) B having polarizer but no UV-blocking structure, and (c) C having both polarizer and UV-blocking structure on the cover lens.

Download Full Size | PDF

Table 1. Spectral energy distribution of IEC 60068-2-5.

View Table

 

Fig. 10. Voltage shifts (ΔV) of OLEDs vs. solar test cycles under the current density of 10 mA/cm²: for various test architectures and UV-blocking structures.

Download Full Size | PDF

Figure 10 shows ΔV vs. test cycles for the three testing architectures, in which 6-pair and 7-pair 2^nd modified DBRs were used as the UV-blocking structure. OLEDs without any UV protection (architecture A) suffer large voltage shift (ΔV=2.62 V after 25 test cycles), clearly manifesting detrimental effect of sunlight irradiation on OLEDs. The polarizer in architecture B (conventional OLED panel structure) provides partial UV protection and thus reduces ΔV (ΔV=1.3∼1.5 V after 25 cycles). Finally, combining the polarizer and UV-blocking DBR (architecture C) provides further protection of OLED panels against solar degradation, and ΔV after 25 test cycles is further reduced down to ΔV∼ 0.8 V, confirming effectiveness of the designed UV-blocking DBRs herein. It is also noticed that both 6-pair and 7-pair UV-blocking DBRs structures provide similar protection against UV/HEV light. Table S4 in Supplement 1 further provides a summary of the luminance of the OLED device under a same operation voltage (4 V) before and after being subjected to the solar test architectures A, B, and C (with 6-pair 2^nd modified DBR structure), again confirming effectiveness of the designed UV-blocking DBR structures.

5. Conclusion

In summary, we developed modified DBR structures composed of UV-absorbing and dielectric materials for protecting OLED panels against UV/HEV irradiation and degradation, thus strengthening their outdoor applications. By tuning thicknesses of high-index/low-index layers in the DBR pairs, we designed and fabricated 6-pair and 7-pair UV-blocking structures with several significant merits, such as nearly constant high transmittance of up to 91% over the whole spectral range, very sharp drop of transmittance around the deep-blue/near-UV region, and effective protection against UV/HEV light, making them suitable for incorporation into OLED display panels with minimal impacts on display colors and efficiencies, etc. Furthermore, we also studied the total electric fields in the modified DBR structure. The simulation results reveal that in the modified UV-blocking DBR structure, the reflection at visible wavelengths can be effectively suppressed by adjusting the thicknesses of the first and the last pair in the DBR structure. Such UV/HEV-blocking structures with high and flat transmittance in the visible range are applicable not only in displays but maybe also in the other areas requiring high-performance optical thin-film coating. Finally, solar test experiments of OLED panels verified usefulness of developed UV-blocking DBRs for reducing UV/HEV/solar-induced degradation and voltage shifts in OLED devices.