利用原子層沉積於高分子基板上沉積氣體阻障層

Abstract

本研究成功的利用原子層沉積(atomic layer deposition, ALD)於聚亞醯胺(polyimide, PI)基板上沉積氧化鋁無機層作為氣體阻障層，可達到有機發光二極體（OLEDs）對氧氣透過率（OTR）的要求：10−3 c.c./m2．day。 由文獻中得知，表面的顆粒及缺陷會造成氣體阻障層於基板上覆蓋的不完整，而導致無法有效的阻絕水、氧氣。所以首先我們研究ALD氧化鋁薄膜對顆粒的包覆能力。由實驗結果驗證了ALD氧化鋁薄膜具有相當優越的包覆能力，其ALD薄膜對顆粒的包覆能力遠遠超過其他鍍膜方式的薄膜。我們觀察到113 Å 的ALD氧化鋁薄膜可完整包覆至少6.01μm高度的顆粒，而PI 基板上所量測到的顆粒皆小於5μm，所以113 Å的ALD氧化鋁薄膜的包覆能力已可解決PI基板上的顆粒所會造成鍍膜不完整的問題。所以於此實驗中，顆粒對於厚度大於113 Å的ALD氧化鋁氣體阻障層的影響可以完全忽略。 但是儘管我們已沈積適當的厚度用來完全包覆顆粒，ALD氣體阻障層仍未能達到預期的阻氣效果。推論是因為PI基板上缺乏ALD沈積所需的化學吸附點，而基板上化學吸附點的多寡為ALD氣體阻障層能否能於基板上沉積達到完全覆蓋的關鍵。於本實驗中，我們利用濕式處理來有效的增加PI基板表面的化學吸附點，進而顯著的改善ALD氣體阻障層的阻氣效果。 除了以上兩點：決定氣體阻障層所需最小厚度及改善基板表面性質，我們進一步將ALD氣體阻障層的製程條件最佳化來增進其阻氣效果。最佳的製程條件為沈積溫度300℃，TMA/ pulse 0.1s/ exposure 5s/ pumping 5s/Water/ pulse 0.1s/ exposure 30s/ pumping 5s及臨界厚度77 Å。 除了研究可達到OLEDs要求的ALD氣體阻障層以外，我們組裝了一台簡單、靈敏、以氦氣滲透為主的滲透量測法裝置用來評估氣體阻障層的阻氣效果。並在不同的溫度下量測氦氣滲透率，再由此換算得到氣體滲透通過阻障層所需的活化能。 在本研究中，我們使用最佳的厚度、基板表面處理和製程條件來沈積ALD氣體阻障層於PI基板上，可將PI基板的氦氣穿透率（HeTR）由1040 顯著的下降至 c.c./m2．day，且氣體滲透所需的活化能由未鍍膜的PI基板19.88 KJ/mole 增加至鍍膜後54.76 KJ/mole。由此活化能的結果我們可得知，氣體滲透通過ALD氣體阻障層的機制和其他已知氣體滲透大多是透過大型孔洞的阻障層是不同的。 目前我們所得到最佳的ALD氣體阻障層的HeTR值為13 c.c./m2．day。由於此HeTR值所相對應之OTR值低於目前我們所可使用的量測方法的極限，所以無法得知此ALD氣體阻障層的OTR值。但根據已知的孔洞薄膜的氣體傳遞機制，我們推算本實驗中所得到最佳之ALD氣體阻障層的OTR值小於1.1 × 10−3 c.c./m2．day，已可達到OLEDs對OTR值的要求。

In this study, we have successfully deposited an Al2O3 inorganic layer by atomic layer deposition (ALD) on polyimide (PI) substrates to achieve the requirement of OLEDs, 10−3 c.c./m2．day of OTR. Previous literatures have shown that transmission of gases through intrinsically impermeable films proceeds through microscopic defects in the films. The particles cause incomplete coverage of the PI surface by a barrier, which lead to ineffective water vapor and oxygen barriers. We demonstrated that the step coverage of the Al2O3-ALD barriers were far superior to those deposited by other methods. We observed that a 113 Å Al2O3-ALD barrier is adequate to fully cover particles that were up to 6.01 μm in height, which are larger than the size of particles typically found on the surface of a PI substrate. The size of particles found on the surface of a PI substrate were small than 5μm in height. Therefore, the effects of particle-induced defects on permeation can be eliminated with an Al2O3-ALD that is thicker than 113-Å. Despite having adequate thickness to fully cover particles, our ALD barriers failed to show expected barrier performance, because the PI substrate lacked chemisorption sites that are critical for the barrier to achieving complete surface coverage. We developed a wet treatment method that effectively created chemisorption sites on the PI substrate, which significantly improved the barrier performance of the resulted barriers. Having determined the minimum required thickness, and achieved desired surface properties, we optimized the deposition conditions of the ALD barriers, which are: deposition temperature at 300℃, TMA/ pulse 0.1s/ exposure 5s/ pumping 5s/Water/ pulse 0.1s/ exposure 30s/ pumping 5s, and critical thickness 77Å. Besides studying ALD barriers, we developed a simple, yet sensitive, permeation measurement method based on helium permeation to evaluate the barrier properties. We measured the helium permeability of the ALD barriers at different temperatures to determine the activation energy of permeation. Employing the optimal thickness, surface treatment, and deposition conditions, we produced ALD barriers that reduced the PI substrate’s helium transmission rate (HeTR) from 1040 to 13 c.c./m2．day. These barriers caused the substrate’s activation energy of permeation to increase from 19.88 KJ/mole to 54.76 KJ/mole, indicating that gas permeation through the ALD barrier was not due to flow mostly through macroscopic defects, which is the permeation mechanism for all other known barriers. OTR value of the optimized ALD barrier, whose HeTR is 13 c.c./m2．day, could not be measured, as it exceeded the lower limit of the methods available to us. Based on known transport mechanisms of gases through a porous membrane, we estimated the ALD barrier’s OTR to be below 1.1 × 10-3 c.c./m2．day, which satisfies the requirement of OLEDs.