Abstract
Dependences of gas-barrier performance on the deposition temperature of atomic-layer-deposited (ALD) Al2O3, HfO2, and ZnO films were studied to establish low-temperature ALD processes for encapsulating organic light-emitting diodes (OLEDs). By identifying and controlling the key factors, i.e. using H2O2 as an oxidant, laminating Al2O3 with HfO2 or ZnO layers into AHO or AZO nanolaminates, and extending purge steps, OLED-acceptable gas-barrier performance (water vapor transmission rates ∼ 10−6 g m−2 d−1) was achieved for the first time at a low deposition temperature of 50 °C in a thermal ALD mode. The compatibility of the low-temperature ALD process with OLEDs was confirmed by applying the process to encapsulate different types of OLED devices, which were degradation-free upon encapsulation and showed adequate lifetime during accelerated aging tests (pixel shrinkage <5% after 240 h at 60 °C/90% RH).
Export citation and abstract BibTeX RIS
1. Introduction
The organic light-emitting diodes (OLEDs) are an important technology for displays and lighting applications, offering significant advantages over competing technologies including high energy efficiency, excellent image/light quality, mechanical flexibility/bendability, lightweight, and compatibility with roll-to-roll manufacturing process [1–6]. One major hindrance to the commercialization of OLEDs is their extreme sensitivity to moisture and oxygen, which necessitates OLED devices to be encapsulated with exceptional gas-permeation barriers that are capable of reducing water vapor and oxygen transmission rates (WVTR and OTR, respectively) into the devices down to extremely low levels. Specifically, OLEDs require WVTR and OTR values of ∼10−6 g m−2 d−1 and ∼10−3 cc m−2 d−1 [7–9], respectively, to obtain adequate lifetimes, which far exceed the gas-barrier performance of typical encapsulant materials. Similar requirements are also imposed by other type of organic electronics such as organic photovoltaics, thin-film transistors, memory, and sensors [10]. To meet the stringent encapsulation demands of OLEDs as well as other organic electronics, atomic layer deposition (ALD) has emerged as a particularly advantageous technique owing to its capabilities of defect-free deposition, low-temperature processing, and ease of scale-up production [11–18]. Ultra-thin (down to 20 nm) flexible gas-barrier films meeting OLEDs' requirements have been fabricated with ALD, and successful encapsulation of OLEDs and various types of organic electronics with such ALD films have been demonstrated [19–26]. These results have triggered an industries-wide scramble toward developing and adopting ALD encapsulation technologies.
Despite the promises of ALD, there are still significant hurdles needed to be overcome for its full deployment in OLEDs and organic electronics to materialize. One such hurdle is reduction of ALD processing temperatures to improve the compatibility of ALD with organic electronic materials, many of which have Tg below 80 °C [27, 28]. Such temperature-sensitive materials can be severely damaged by an elevated-temperature encapsulation process, therefore requiring process temperatures well below 80 °C. Unfortunately, reducing ALD temperatures to such a low range tends to significantly degrade the barrier performance of the resultant films, due in large part to inadequate reactivity of ALD precursors at low temperatures, which creates defects associated with unreacted residues [13–15]. A common method to improve the reactivity of ALD precursors at low temperatures is to operate in a plasma-enhanced mode (PEALD), but this compromises the unique conformal-coating capability of ALD, resulting in poorer barrier properties [29]. Consequently, OLED-acceptable barrier performance (WVTR ∼ 10−6 g m−2 d−1) has yet to be demonstrated at deposition temperatures <80 °C. Moreover, the change in the barrier property as the deposition temperature gets to the lower range has not been systematically studied for the various types of ALD oxides commonly used for gas barrier applications, and therefore it has been difficult to formulate a general strategy for developing low-temperature ALD barrier films.
In this study, we examined the temperature-dependence of the gas-barrier performance as well as other properties of several ALD oxide films—including Al2O3, HfO2, ZnO, and their nanolaminates deposited with H2O or H2O2 as oxidant in a thermal ALD mode at low deposition temperatures down to 50 °C. Based on the results, we identified the key factors determining the lower deposition-temperature limit for obtaining good gas-barrier performance. By controlling these factors, we achieved OLED-acceptable gas-barrier performance (WVTR ∼ 10−6 g m−2 d−1) from ALD films for the first time at the low deposition temperature of 50 °C. Subsequently, we applied the low-temperature ALD nano-laminates as encapsulants on two types of OLED devices to confirm their compatibility with various temperature-sensitive OLED materials. The OLED devices were degradation-free upon encapsulation, after which they exhibited a sustained stability throughout an accelerated aging test.
2. Experiment
2.1. Atomic layer deposition
The barrier films analyzed in this study were deposited in a Cambridge NanoTech Savannah 100 ALD system. The ALD chamber pressure was 0.1 Torr during process with a constant 20 sccm flow of N2 throughout to carry and purge the precursor vapors. The process conditions of the Al2O3, HfO2, ZnO, and nanolaminated films are summarized in table 1 [26, 30]. Metal–organic precursors used in the ALD processes included trimethylaluminum (TMA) (Sigma-Aldrich, 97% purity, used as received) for Al2O3, tetrakis(dimethylamido) hafnium (TDMAHf) (Sigma-Aldrich, ≥99.99% purity, used as received) for HfO2, and diethylzinc (DEZn) (Sigma-Aldrich, 97% purity, used as received) for ZnO. H2O and H2O2 (35%, Sigma Aldrich, used as received) were used as oxidants. All of the precursors were used without heating except for TDMAHf, which was heated to 75 °C.
Table 1. Processing conditions for the Al2O3, ZnO, HfO2 films and their nano-laminates.
| Film | Precursor | Pulse time(s) | Soaking time(s) | Pumping time(s) | Cycles/thickness |
|---|---|---|---|---|---|
| Al2O3 (H2O) | TMA | 0.03 | 0 | 8 | 350/35 nm |
| H2O | 0.02 | 0 | 25 | ||
| Al2O3 (H2O2) | TMA | 0.03 | 0 | 8 | 350/35 nm |
| H2O2 | 0.02 | 0 | 25 | ||
| ZnO (H2O) | DeZn | 0.06 | 0 | 8 | 200/30 nm |
| H2O | 0.02 | 0 | 25 | ||
| ZnO (H2O2) | DeZn | 0.06 | 0 | 8 | 200/30 nm |
| H2O2 | 0.02 | 0 | 25 | ||
| HfO2 (H2O) | TDMAHf | 0.1 | 0 | 8 | 350/35 nm |
| H2O | 0.02 | 0 | 25 | ||
| HfO2 (H2O2) | TDMAHf | 0.1 | 0 | 8 | 350/35 nm |
| H2O2 | 0.02 | 0 | 25 | ||
| AHO | Alternating 2 nm of Al2O3 with 2 nm of HfO2 for 9 times | ||||
| AZO | Alternating 2 nm of Al2O3 with 2 nm of ZnO for 9 times | ||||
ALD films were deposited on silicon wafers and polyethylene terephthalate (PET, 100 μm, Alliance Material Co. Ltd) substrates for ellipsometry, and gas transmission rate measurements, respectively. Before deposition, the subtrates were cleaned in turn with ultrasonication in acetone, methanol, and deionized water for 10 min, and then cleaned with oxygen plasma (Harrick Scientific, Model PDC-32G) under low power for 5 min. A 0.5 nm (5 cycles) Al2O3 layer was used as the initial nucleation layer for all of the barrier films, whose process parameters were as follows: 0.1 s pulse of TMA, 30 s soak, 30 s purge, 0.1 s pulse of H2O, 30 s soak, 30 s purge. For encapsulating OLED devices, the ALD processes outlined above were applied directly to OLED devices.
2.2. Characterization of ALD films
The refractive index and thickness of the ALD films were measured by ellipsometry (Accurion Imaging Ellipsometer EP3) with λ = 633 nm. The composition of the films was determined by x-ray photoelectron spetroscopy (XPS) using an ULVAC-PHI 5000 VersaProbe XPS photometer with a monochromatic Al K-Alpha x-ray source (1486.6 eV); samples were cleaned by in situ Ar-sputtering (removing 1–3 nm of thickness) before measurement. X-ray diffraction (XRD) analyses were carried out in a Rigaku TTRAX 3 XRD system with a copper Kα line source. Water contact angles of the AHO and AZO films were measured to be 92° ± 3° using a Ramé-Hart contact angle goniometer (Model 100), where a sessile drop of 2 to 3 μl in volume was dispensed with a microsyringe and the contact angle was measured within 30 s after its formation. Root-mean-square surface roughness of the AHO and AZO films deposited on Si wafers was determined by atomic force microscopy (Digital Instrument Multimode Atomic Force Microscope) to be 0.630 nm (see figure S1). Helium transmission rate (HeTR) was measured from 5 cm × 5 cm samples using a setup that has been previously reported [31]. WVTR was determined by MOCON AQUATRAN Model 1 with 10 cm × 10 cm samples at various temperatures between 38 °C and 60 °C at 100% RH. The HeTR and WVTR of the bare PET substrate were measured to be 801 cc m−2 day−1 (at 23 °C) and 20 g m−2 day−1 (at 38 °C), respectively. Arrhenius relation between the WVTR and the measurement temperature was determined from the varied-temperature measurements, with which the room-temperaure WVTR values were calculated. The elevated-temperature WVTR results used to calculate the room-temperature values are included in table S1 of supplementary information.
2.3. OLED device fabrication and characterization
Two types of OLED devices were used to examine the effectiveness of the ALD barrier films as an encapsulation layer: a quasi-incandescent white-light device and a passive-matrix pixilated monochromatic device. The quasi-incandescent device had the following architecture: glass/ ITO/PEDOT:PSS/TAPC/(Ir(2-phq)3) doped in TCTA host/ PO-01/Ir(ppy)3 doped in TPBi host/ BmPyPb/ LiF/Al. The emission area of the devices was 9 mm2 (detailed information is provided in supplementary information). The passive-matrix device was of the model 2864HMBE manufactured by WiseChip Semiconductor Inc. with a pixel size of 0.154 mm × 0.154 mm. The luminance, CIE chromatic coordinates, and electroluminescent spectrum of the OLED devices were measured with a Photoresearch Spectrascan PR-655 spectroradiometer. A Keithley 2400 electrometer was used to measure the current–voltage (I–V) characteristics.
3. Results and discussion
The results are presented in the following sequence. Firstly, the barrier properties of the Al2O3, HfO2, and ZnO films were measured at incrementally lowered temperatures to determine the lower temperature limit for obtaining good barrier performance. The barrier performance was evaluated in terms of HeTR, whose high sensitivity ensured that the differences among the different deposition temperatures were clearly distinguished. Secondly, using the determined lower deposition temperature limits, nano-laminated films of Al2O3/HfO2 (AHO) and Al2O3/ZnO (AZO) were prepared and characterized for HeTR and WVTR. Finally, the low-temperature-deposited nano-laminated films were utilized to encapsulate OLED devices in order to verify their thermal compatibility with the devices and their effectiveness in hermetically sealing them.
3.1. Barrier properties versus deposition temperature for Al2O3, HfO2, and ZnO
The HeTR as a function of deposition temperature of the Al2O3, HfO2, and ZnO films, deposited with either H2O or H2O2 as the oxidant, is presented in figure 1. The refractive index of the films was determined to evaluate the quality of the films, and the results are shown in figure S2 of supplementary information. The results of each type of the films are discussed separately below. It should be noted that the thickness of the films, 35 nm, was chosen to obtain the optimal gas-barrier performance with the minimum amount of processing time, as shown in figure S3.
Figure 1. HeTR as a function of deposition temperature of the Al2O3, HfO2, and ZnO films deposited with H2O or H2O2 as oxidant.
Download figure:
Standard image High-resolution image3.1.1. Al2O3
With the H2O process, the lower temperature limit in terms of barrier performance was 70 °C; namely, the HeTR of the Al2O3 films stayed largely constant at >70 °C, and it rose sharply at <70 °C. The refractive index of the films exhibited the same trend, i.e. constant above 70 °C, lowered markedly below 70 °C. This suggested that the reactivity of the TMA/H2O chemistry was inadequate at <70 °C, so that the lower-temperature films had a significant amount of unreacted organic residues, which lowered the refractive index and degraded the barrier performance of the films [13]. It should be noted that even though the Al2O3 films had much lower HeTR than those of the HfO2 and ZnO films, they were not functional moisture barriers because of their tendency to hydrolyze and disintegrate in a humid environment, and they required lamination with hydrolysis-resistant layers (HfO2 and ZnO in this study) to obtain adequate stability against humidity [22, 23, 26]. It was also for this reason that the WVTR of the Al2O3 films was not measureable.
The problem of low reactivity at low deposition temperatures was effectively resolved by using the more reactive H2O2 instead of H2O as the oxidant, which extended the lower temperature limit to ≤50 °C. The HeTR of the H2O2-deposited films was lower than that of the H2O-deposited films throughout the temperature range, and it remained constant from 140 °C down to 50 °C, the lower operating temperature limit of our experimental setup. Additionally, the refractive index of the H2O2-deposited films was significantly higher than that of the H2O-deposited ones throughout the temperature range, indicating a reduction in unreacted organic residues thanks to the enhanced reactivity of the H2O2 process. The ZnO and HfO2 films discussed below also showed similar rises in refractive index when switched to the H2O2 process. The amount of unreacted organic residues in the films was also examined with XPS analysis (table S2 of supplementary information), which showed that the H2O2 process indeed yielded significantly lower carbon contents than did the H2O process.
3.1.2. ZnO
The temperature-dependence of HeTR of the ZnO films was distinct from that of the Al2O3 films in that, instead of increasing monolithically with decreasing deposition temperature, the HeTR of the ZnO films decreased with decreasing temperature until it reached a minimum (at 70 °C and 60 °C for the H2O and H2O2 process, respectively), below which it began to rise. The occurrence of the minimum indicated the presence of an additional factor competing against the factor of ALD reactivity. ALD ZnO films have been shown to be crystalline at deposition temperatures down to room temperatures, with decreasing crystallinity at lower temperatures [32], which was consistent with our analysis (figure S4). Crystallinity in a barrier film tends to increase the permeability of the film by creating grain-boundary defects [30]. This can be seen in the higher HeTR of the ZnO films than that of the Al2O3 films, which were amorphous, regardless of the deposition temperature and the type of oxidant used. Therefore, as the deposition temperature was lowered, the crystallinity of the ZnO films reduced, lowering the HeTR; meanwhile, the reactivity of the ALD chemistry declined, as indicated by the ZnO films' lowered refractive index (figure S2), raising the HeTR. The balance of these two trends determined the temperature at which the minimum occurred. With the H2O2 process, the reactivity was higher than that of the H2O process, and therefore the HeTR minimum was at 60 °C as opposed to 70 °C, and the HeTR was lower than that of the H2O process across the tested temperature range.
In addition to the effects of deposition temperature and type of oxidants, the length of purging time was also a determining factor for the HeTR of the films, especially at the low end of the temperature range, as shown in table 2. At 50 °C, prolonging the purging times from 8 s/25 s for DEZn/oxidant (H2O or H2O2) to 100 s/100 s significantly reduced the HeTR of both the H2O- and H2O2-deposited ZnO films. The longer purging time also resulted in higher refractive index of the films (figure S2), an indication of more thorough removal of unreacted or physisorbed precursor molecules, which were less volatile at lower deposition temperatures and thus required longer purging times to remove. Such unreacted molecules can create defects or form particles in a film through gas-phase reactions, causing the gas-barrier performance of the film to deteriorate. Similar improvements in HeTR and increase in refractive index by the extended purging treatment were also observed for the 50 °C-deposited HfO2 and Al2O3 films, although in the case of the latter, the improvement in HeTR was slight because the Al2O3 films had a low HeTR to begin with.
Table 2. HeTR with the regular purging times (8 s/25 s for metal–organic precursors/oxidant) and with the extended purging times (100 s/100 s) of ZnO, HfO2, and Al2O3 films deposited at 50 °C.
| Material | Oxidant:H2O | Oxidant:H2O2 | ||
|---|---|---|---|---|
| Regular | Extended | Regular | Extended | |
| ZnO | 10.5 | 7.5 | 5.4 | 4.0 |
| HfO2 | 10.1 | 7.0 | 4.5 | 3.8 |
| Al2O3 | 6.9 | 6.1 | 1.7 | 1.7 |
3.1.3. HfO2
The HfO2 films showed a similar deposition-temperature dependence of HeTR to that of the ZnO films, with minimum HeTR values at 70 °C and 60 °C for the H2O and H2O2 process, respectively. The HfO2 films, similar to the ZnO films, were crystalline at the temperature range tested but with a much lower crystallinity than that of the ZnO films [32] (figure S5). Consequently, at temperatures above the minimum point, the HeTR decreased only slightly with decreasing deposition temperature.
3.2. Nano-laminated films of Al2O3/HfO2 (AHO) and Al2O3/ZnO (AZO)
Nano-laminated AHO and AZO films deposited with H2O2 as oxidant at both 50 °C and 60 °C, the latter of which was determined above to be the optimal temperature in terms of barrier performance for both the ZnO and HfO2 films, showed WVTR in the order of 10−6 g m−2 d−1, as shown in table 3. The 50 °C-deposited AZO and AHO films required the extended purging treatment discussed above to achieve the 10−6 g m−2 d−1 WVTR. Without the extended purging treatment, the 50 °C-deposited films showed much higher WVTR's at the 10−4 g m−2 d−1 level. The HeTR's of the AZO and AHO films were also shown in table 3, which were approximately equal to the combination of the HeTR contributions from the individual Al2O3, ZnO, and HfO2 layers shown above in figure 1.
Table 3. WVTR and HeTR of AHO and AZO nano-laminated films deposited at 50 °C or 60 °C with H2O2 as oxidant.
| HeTR cc m−2 d−1 | WVTR g m−2 d−1 | ||
|---|---|---|---|
| AZO | 60 °C | 1.7 | 1.0 × 10−6 |
| 50 °C | 3.7 | 2.1 × 10−4 | |
| 50 °C, extended purges | 2.4 | 1.3 × 10−6 | |
| AHO | 60 °C | 3.5 | 9.8 × 10−6 |
| 50 °C | 4.0 | 5.1 × 10−4 | |
| 50 °C, extended purges | 3.5 | 3.5 × 10−6 |
3.3. Encapsulation of OLED devices
The low-temperature-deposited ALD nano-laminated films were applied as encapsulants on a temperature-sensitive OLED device to examine the low-temperature ALD encapsulation process' compatibility with such devices. The OLED device, whose architecture was described in the experimental section, emitted quasi-incandescent white light with high efficiency and brightness (maximum luminance ∼91 000 cd m−2), as shown in figure S6 and table 4; but it was highly sensitive to elevated temperatures due to its constituents of low-Tg active materials. For instance, the device's efficiency was significantly reduced and operating voltages increased by heating at 70 °C in high vacuum for 3 h, as shown in table 4. Such thermal degradations were completely avoided with the 60 °C ALD process: as also shown in table 4, upon the application of a 35 nm AZO film deposited with the 60 °C/H2O2 process, the OLED device did not show any degradation on its OLED spectrum (shown in figure S7), but instead had moderate improvements in its efficiencies, which were likely attributable to improved interface quality by the annealing effects of the 60 °C thermal treatment, as reported previously [33, 34]. The encapsulation effectiveness of the AZO film in terms of the lifetime of the encapsulated devices, however, could not be accurately evaluated, because the OLED devices did not have adequate intrinsic stability to allow for long-term lifetime tests. Specifically, when the as-fabricated devices were stored in an O2-/H2O-free environment, they still degraded completely within hours, as shown in figure S8.
Table 4. Characteristics of the quasi-incandescent white-light OLED devices as fabricated, annealed after fabrication at 70 °C for 3 h, or encapsulated after fabrication with the 60 °C ALD AZO film.
| Operation voltage (V) | Power efficiency (lm W−1) | Current efficiency (cd A−1) | |
|---|---|---|---|
| @ 100/1000/10 000 cd m−2 | |||
| As-fabricated | 2.9/3.5/5.2 | 58.9/48.3/26.8 | 54.4/54.3/44.2 |
| Annealed | 3.0/4.0/6.1 | 53.1/35.9/21.3 | 50.9/45.7/41.4 |
| ALD-encapsulated | 2.9/3.6/5.4 | 65.3/51.5/29.0 | 60.7/58.9/50.0 |
Alternatively, the encapsulation effectiveness of the ALD films was evaluated with a commercial passive matrix OLED (PMOLED) device (Wisechip Co.; product no. 2864HMBE). The PMOLED device, when encapsulated with a standard 'glass lid and glue and getter' method [10, 35], meets a storage lifetime specification of <5% pixel shrinkage after 240 h storage in a 60 °C/90% relative humidity environment. With the 60 °C/H2O2-deposited AZO film as the encapsulating layer, the device exhibited a pixel shrinkage ratio of 4.4%, which was within the product specification of 5%; additionally, the device did not show any dark spots, as shown in figure 2. This confirmed the effectiveness of the low-temperature ALD films as OLED encapsulants. It should be noted that the initial luminance was 1003 and 1000 cd m−2 for the ALD- and glass-encapsulated OLEDs, respectively.
Figure 2. Pixel images of the passive matrix OLED devices before (0 h) and after (240 h) the accelerated aging test at 60 °C/90% RH.
Download figure:
Standard image High-resolution image
4. Conclusion
Deposition-temperature dependences of the gas-barrier performance of ALD Al2O3, HfO2, and ZnO films were studied to establish low-temperature ALD encapsulation processes for OLEDs. Key factors determining the lower deposition-temperature limit for obtaining good gas-barrier performance from the films were identified to include the type of oxidant, crystallinity of deposited films, and duration of the purge step in the ALD cycles. By controlling these factors, i.e. using H2O2 as an oxidant, laminating Al2O3 with HfO2 or ZnO layers into AHO or AZO nanolaminates, and extending purge steps, high gas-barrier performance (WVTR ∼ 10−6 g m−2 d−1) was achieved for the first time from a single layer of ultra-thin (35 nm) oxide films at the low deposition temperature of 50 °C. The compatibility of the low-temperature ALD process with OLEDs was confirmed by applying the process to encapsulate different types of OLED devices, which were degradation-free upon encapsulation and showed excellent stability during accelerated aging tests (pixel shrinkage <5% after 240 h at 60 °C/90% RH).
Acknowledgments
This work was supported by Ministry of Science and Technology (Grant Nos. 103-2221-E-002-278, 103-2623-E-002-003-ET, 102-2221-E-002-231, 102-3113-E-007-001), Industrial Technology Research Institute, and Bureau of Energy, Ministry of Economic Affairs of Taiwan. The authors thank WiseChip Semiconductor Inc. for providing the PMOLED devices.