Large‐Area Flexible Thin Film Encapsulation with High Barrier and Super‐Hydrophobic Property

With the development of optoelectronic devices toward miniaturization, flexibility, and large‐scale integration, conventional submillimeter rigid encapsulation techniques rarely achieve conformational functionality while blocking water and oxygen. At the same time, the sensitivity of electronic devices with organic/metal/semiconductor components to humidity and oxygen severely impairs their operational stability and lifetime. Here, a nanometer to micrometer scale organic/inorganic hybrid thin film encapsulation (TFE) with the self‐cleaning ability for flexible encapsulation is developed. The water vapor transmittance rate of polyethylene terephthalate substrate coated with the TFE is as low as 1.65 × 10−4 g m−2 day−1, and the barrier improvement factor reaches 104 at 38 °C and 90% relative humidity. This value is equivalent to 9.81 × 10−6 g m−2 day−1 at ambient conditions, sufficient to improve the lifetime of water‐sensitive electronic devices. Meanwhile, this TFE shows a super‐hydrophobic performance, with a water contact angle of 168.4°. In addition, the resulting barrier films exhibit outstanding optical properties, with an average optical transmittance of 86.88% in the visible region. This versatile TFE can promote the development of optoelectronic devices toward miniaturization and large‐scale integration in the future.


Introduction
Optoelectronic devices have attracted substantial research interest and achieved remarkable progress in optical storage, communication, and lasers. [1][2][3][4][5][6][7][8] To cope with more specific applications in the future, optoelectronic devices should be with the characteristics of miniature, intelligent, flexible, waterproof, and self-cleaning. Nonetheless, a significant drawback of these devices is their sensitivity to environmental stimuli due to their organic, metallic, and semiconductor compositions. [9][10][11][12] In particular, some devices need to work in harsh environments that could result in severe corrosion and degradation. [13,14] Surface functional encapsulation of these devices can improve their environmental adaptability and performance. [9,15] It is reported that the water vapor transmittance rate (WVTR) should be as low as 10 −4 g m −2 day −1 to maintain the stable operation of devices like solar cells and organic light-emitting diodes in the environment. [16] Therefore, the development of high-barrier encapsulation techniques for optoelectronic devices is required.
Several traditional encapsulation strategies have been applied to protect the device, such as UV-cured adhesive, glass cover plate encapsulation, and thermally cured epoxy. However, these methods may damage the material [17] and limit the large-scale flexible fabrication, thus not suitable for the requirements of future optoelectronic devices. [18] Considering the need for practical application, some researchers have conducted preliminary exploration on organic/inorganic hybrid multilayer thin film encapsulation (TFE). [19][20][21][22][23] Compared with conventional submillimeter rigid encapsulation, the TFE combines barrier and flexibility successfully. [24,25] However, most previous studies focused more on improving the barrier properties, and finite monomer selection, [20] solution preparation, [22] and elevated energy particle bombardment [23] also limit its practical applications. In addition to meet the high barrier performance, TFE also needs to have the corresponding function in some specific scenarios. In such conditions, the corresponding optoelectronic device surfaces may be stained with dust, oil, and additional contaminants, severely affecting the device performance. For example, the surface of the camera responsible for intelligent security monitoring in the mine is easily stained and out of action due to extreme conditions of high temperature, humidity, and excessive dust. Similarly, photovoltaic arrays installed in desert areas can quickly accumulate dust on their surfaces, resulting in a severe reduction in power conversion efficiency and a considerable increase in management costs. [26] In these cases, encapsulated films are required to have both high barrier and self-cleaning properties to be automatically resistant to contaminants. [27,28] Furthermore, for miniaturized semiconductor devices containing a large number of structures with a high aspect ratio (depth/width), [29] conformational properties are expected, that is, the shape and functionality of the original device should be preserved. [30] Therefore, a versatile TFE is urgently needed to meet the demand for different application scenarios of future large-scale integrated optoelectronic devices.
Here, a novel hybrid organic/inorganic multilayer is designed and achieves high barrier performance. Using the direct test method, the WVTR of TFE on a large area (100 cm 2 ) of polyethylene terephthalate (PET) substrate is as low as 1.65 × 10 −4 g m −2 day −1 , and this value is approximate to the limit of instrument measurement (5 × 10 −5 g m −2 day −1 ). Compared with the bare PET, the WVTR of the best sample was reduced by four orders of magnitude, namely the barrier improvement factor (BIF) of 10 4 . Under environmental conditions, this value is equal to 9.81 × 10 −6 g m −2 day −1 , which could considerably improve the environmental adaptability of optoelectronic devices. In particular, the TFE with a micro-nano structure top layer make it possesses super-hydrophobic, oil-phobic, and self-cleaning capabilities. The average transmittance of TFE in the visible range reaches 86.88%, showing excellent optical properties. It is worth noted that films deposited by the initiated chemical vapor deposition (iCVD) technique can be completely and uniformly encapsulated on the device surface with high aspect ratio channel. The TFE technology we provided is highly essential for future integrated and miniaturized optoelectronic devices.

TFE Structure
We prepared organic-inorganic hybrid multilayer TFE on silicon and PET substrates using single chamber continuum deposition initiated chemical vapor deposition and atomic layer deposition (iCVD-ALD) techniques. The structure of the TFE, for both multilayer and micro-nano structures, is shown in Figure 1a. The chemical formula and the detailed reaction parameters of poly(ethylene glycol diacrylate) (pEGDA) homopolymer and poly-(ethylene glycol diacrylateco-1H,1H,2H,2H-perfluorodecyl acrylate) (p(EGDA-co-PFDA)) crosslinked copolymer are discussed in the Supporting Information Part I. As shown in Figure S1 (Supporting Information), the pEGDA and pEGDA/Al 2 O 3 films have uniform thickness on the substrate and smooth surfaces, the organic layer and inorganic layer are successfully stacked. The cross-section SEM image in Figure 1b shows a TFE with pEGDA/Al 2 O 3 /pEGDA/Al 2 O 3 / pEGDA/Al 2 O 3 /pEGDA/p(EGDA-co-PFDA) 3-dyad structure and a total thickness about 2 μm. The iCVD region is darker than the ALD region due to the higher conductivity typical of carbonrich organic layers. In TFE, the organic and inorganic layers are stacked alternately without stratification, which ensures the high barrier performance of the film. From bottom to top, the thicknesses of the pEGDA layers are 200, 100, 100, and 140 nm, respectively. The thickness of the Al 2 O 3 layer deposited by 200 ALD cycles is 20 nm. As a top layer, the thickness of p(EGDA-co-PFDA) crosslinked film on the top pEGDA layer is 1.2-1.6 μm. The above results indicate that the TFE has excellent continuity, which is a prerequisite for blocking water and oxygen.
To further test the overall performance of TFE, Fourier transform infrared (FTIR), spectroscopy and X-ray electron spectroscopy (XPS) were performed. Figure 1c shows the FTIR spectra of different films deposited on the Si substrate. The absorption peaks at 1732 and 1150 cm −1 are assigned to C=O and C-O stretching, [31] respectively. The peaks at 1238 and 1201 cm −1 are caused by the asymmetric stretching and symmetric stretching of the -CF 2moiety, [32] respectively. This result indicates that the functional group in the monomer is fully preserved in the polymer. Figure 1d shows an XPS scan of a sample with a p(EGDAco-PFDA) film as the top layer. The percentage of atomic concentration detected by XPS (Table S1, Supporting Information), is F: 54.68%, C: 38.76%, O: 6.09%, which is close to the theoretical value of the atomic proportion of pPFDA film (F: 53.13%, C: 40.63%, O: 6.25%). [33] The slightly increased concentration of fluorine atoms at the surface is presumably attributed to the introduction of the crosslinker EGDA and the supersaturation of PFDA. The inset shows the high-resolution spectrum of C1s on the p (EGDA-co-PFDA) surface. The high intensity of -C*F 2and -C*F 3 peaks indicate that the surface is mainly composed of PFDA components. [32,34] The XPS study also shows a very high retention rate for the functional group of films deposited by the iCVD technique.

Barrier Performance of TFE
WVTR is defined as the weight of water vapor that permeates a material at a given time, temperature, and humidity. The lower  the value of the WVTR, the stronger the barrier capacity of the film. The barrier properties of the TFE are mainly affected by the quality and thickness of the inorganic layer. Specifically, the main factor affecting the inorganic film quality in this study is the deposition temperature of Al 2 O 3 during ALD. In Figure 2a, a 20 nm thick Al 2 O 3 film is deposited on a PET substrate at various temperatures. It is shown that the WVTR of the film decreases significantly as the substrate deposition temperature increases. When the substrate temperature rises to 110°C, the mean WVTR (0.141 g m −2 day −1 ), is nearly an order of magnitude lower than that in 60°C (0.877 g m −2 day −1 ). Therefore, for Al 2 O 3 in TFE, we normally choose the highest deposition temperature provided that the high barrier properties are simultaneously guaranteed and the substrate device is not compromised (PET does not deform below 120°C). It can also be seen from Figure 2b that the particles of ALD Al 2 O 3 are finely homogeneous at high temperatures. The surface root-mean-square (RMS) roughness reaches 0.8 nm. To confirm the crystalline form of ALD Al 2 O 3 , 20 nm of Al 2 O 3 was deposited on Si for XRD testing. As shown in Figure 2c, ALD Al 2 O 3 is -Al 2 O 3 (ternary crystal system), which is the most stable phase of alumina, with high melting point, high hardness, wear resistance, high mechanical strength, superior electrical insulation, corrosion resistance, and other excellent properties. The XRD pattern is also consistent with that reported by the literature. [35] Although the inorganic layer plays a major role in the TFE, we still need to choose an appropriate thickness for the organic layer. Figure 2d shows that the WVTR of the organic layer is indeed insensitive to the thickness. The pEGDA film was deposited on PET, and the thickness of the film increased from 100 nm to 1.7 μm. However, the WVTR of samples deposited with pEGDA films of different thicknesses is not significantly different from that of uncoated samples. In addition, for the bilayer organic film (pEGDA/p(EGDA-co-PFDA)), the pEGDA film thickness was fixed at 200 nm, and the thickness of p(EGDA-co-PFDA) cross-linked polymer film was monotonically increased from 100 nm to 2 μm. The WVTR of the PET/pEGDA/p(EGDAco-PFDA) samples was almost unchanged with the p(EGDA-co-PFDA) thickness. The result indicates the addition of p(EGDAco-PFDA) cannot improve the WVTR of PET/pEGDA samples. The WVTR of coated PET is also similar to that of uncoated PET, in agreement with the results obtained for single-layer organic films. Therefore, the organic layer alone does not contribute significantly to the overall WVTR improvement. The effect of the thickness of the organic layer on the substrate stability and overall flexibility is mainly considered. Previous reports have shown that the thickness of organic layers in TFE is typically 5-10 times that of inorganic layers for flexible and reliable encapsulation. [36,37] Since the thickness of the inorganic layer has been determined to be 20 nm, the appropriate thickness of the organic layer is 100-200 nm. In addition, the thickness of the organic layer in direct contact with the substrate was set at 200 nm to prevent the influence of H 2 O as an oxygen source in the ALD Al 2 O 3 process on water-sensitive optoelectronic devices. The other intermediate organic layers are 100 nm thick.
Adding an inorganic layer significantly improves the barrier properties of the TFE compared to increasing the thickness of the organic layer. As shown in Figure 2e, the WVTR of the uncoated PET sample was 2.66 g m −2 day −1 , and the WVTR of TFE was reduced to 0.18 g m −2 day −1 after the addition of 20 nm Al 2 O 3 (one organic-inorganic dyad). Moreover, when the number of TFE dyads increased to 3 and 4, WVTR reached 7.47 × 10 −4 and 1.65 × 10 −4 g m −2 day −1 , respectively. Compared to uncoated PET, the barrier performance improves by four orders of magnitude, with the best BIF reaching 10 4 , demonstrating the superior water vapor barrier capability of TFE. Here, the BIF is defined as the ratio of the WVTR of the uncoated film to that of the coated barrier. [38] Notably, the test was performed on the PET sample with large size of 10 × 10 cm 2 and using a commercial MOCON water vapor transmittance tester. Since the effective test area is 50 cm 2 , and the optimal WVTR (1.65 × 10 −4 g m −2 day −1 ) is close to the test limit of the MOCON instrument (5 × 10 −5 g m −2 day −1 ) at an accelerated reaction condition of 38°C and 90% RH. This value is equivalent to 9.81 × 10 −6 g m −2 day −1 at ambient condition (20°C/50% RH) according to the acceleration relationship reported by the literature. [37,39] Although the TFE with 4-dyad has a lower WVTR, it is possible to damage the device as the ALD process time increases. Figure 2f shows the instrumental test times and the corresponding WVTR for films with different dyad numbers. The WVTR-test time curve for the TFE with 3 dyads is smoother after the WVTR reaches the steady state (54-72 h). The WVTR curve ends in a line almost parallel to the time axis, indicating a higher stability barrier performance for the TFE in this condition. Moreover, depositing the TFE with three and four dyads takes about 12 and 16 h, respectively. Thus, considering the trade-off between barrier performance and preparation time, the TFE with 3 dyads has the best cost performance.
It is worth mentioning that, unlike the calcium detection method used in most studies, this is one of the best WVTRs available for direct instrumentation of large area samples. The conventional calcium test, an indirect measurement technique in nature, involves Ca film deposition and barrier growth in a glove box filled with N 2 , which limits the application of multiple deposition techniques to the barrier. [40,41] At the same time, the calcium sample transfer process and the barrier film deposition process also lead to oxidation of calcium, making the test results inaccurate. The differences between the calcium and instrumen- tal tests are detailed in the Supporting Information Part II. In particular, compared to large-area samples, small-area samples are easier to prepare uniformly to control the average defect density and achieve higher barrier performance. In fact, the area of the calcium test sample is commonly 1 × 1-2.5 × 2.5 cm 2 . [42] If extended to a large area, the WVTR will inevitably increase with the rapid growth of defects. In our measurements, the effective test area is up to 50 cm 2 . Hence, the test results better reflect the homogeneity and ultralow defect density of the barrier film.

Super-Hydrophobic Self-Cleaning Performance of TFE
In a real environment, dust and other contaminants on the surface of a photoconversion device can severely reduce the light absorption and thus the energy conversion efficiency. At the same time, sunlight is reflected due to the smooth glass surface, reducing light absorption even if the photovoltaic panel surface is immaculate. [43] As a result, self-cleaning antireflective coatings on the surfaces of these devices can reduce the impact of natural contaminants on the devices. [44] Moreover, antireflection due to the rough surface will enhance the transmittance and hence the absorption of light. Here, we apply the crystalline nature of the fluoroalkyl side chain of pPFDA to induce island growth during iCVD, resulting in highly textured micro-nano surfaces. Combined with the low surface energy properties of long-chain fluoroalkyl polyester, we have successfully functionalized the TFE surface, giving it superhydrophobic self-cleaning properties.
As shown in Figures 1b and 3a, the top layer p(EGDA-co-PFDA) cross-linked copolymer film of TFEs shows a distinct micro-nano structure with the axial direction vertical to the surface. The formation of the micro-nanocone is presumably due to the partial pressure of monomer (P m ) of the PFDA monomer vapor exceeding its saturation pressure (P sat ) near the substrate (P m /P sat >1). [45][46][47] The supersaturated PFDA monomer vapor condenses into tiny droplets on the surface of the pEGDA film and becomes the nucleation center. The monomer vapor continues to adsorb at the gas-liquid interface and polymerizes at the liquid-solid interface, resulting in the growth of the micronanocones. [34,48] Figure 3a shows that the micro-nanocones are randomly distributed without significant periodicity. In addition, controlling the deposition time can regulate the size and density of micro-nanocones, leading to super-hydrophobic surfaces. As shown in Figure S3 (Supporting Information), the size and density of the nanocone structure continuously increase with the deposition time. Finally, when the deposition time was 30 min, the diameter of the micro-nanocones reached 500 nm-1 μm, the height reached 1.2-1.6 μm, and the static WCA reached 168.4°. The droplet exhibits super-hydrophobicity, probably due to the tendency of the micro-nanocones to form Cassie contacts with the droplet. [49,50] In this regime, the droplet is suspended over the structure, forming a solid-liquid-gas composite interface. Figure 3b shows the dependence of the surface wettability properties for different films. The static water contact angle (WCA) of the surface of the uncoated PET film is 67.2±1.5°, and the water droplets on the film surface will not drop even if the sample is 90°vertically in Figure S2 (Supporting Information). After deposition of the single-layer pEGDA film, the WCA increases to 104 ± 2.5°because the pEGDA surface energy is lower than that of PET. In particular, the PET/pEGDA-Al 2 O 3 surface exhibits hydrophilicity (WCA = 54.8 ± 3.2°), which may be caused by the low roughness and high surface energy of the Al 2 O 3 surface. The optimal static WCA and water sliding angle (WSA) of p (EGDA-co-PFDA) cross-linked copolymer films reached 168.4°and 1.1°, respectively, indicating excellent super-hydrophobic property (WCA>150°, WSA<10°). A slight tilt causes the droplet to completely roll off the film. It is because p(EGDA-co-PFDA) films resemble lotus leaves, which have Deep silicon grooves with high depth-to-width ratio were fabricated using the MEMS process. Images of b1), b2), b3) are partially amplified morphology of the pEGDA film deposited on the surface of the structure. The above image was taken with a scanning electron microscope (SEM) attachment on a focused ion beam (FIB) facility. The sample was sprayed with gold (1 nm) before scanning. micro-nano structures on the surface. Moreover, the large number of fluorine atoms in the pPFDA structure is arranged close to the surface, which reduces the surface energy and thus exhibits super-hydrophobicity. When this surface is used as the surface layer of a TFE, it has self-cleaning capability due to the lotus leaf effect, which can help optoelectronic devices to reduce surface pollution and increase their efficiency and environmental adaptability.
Hexadecane ( = 27.7 mN m −1 ) is an organic liquid widely used to evaluate the degree of oil phobicity. [51] As shown in Figure 3c, when the TFE with pEGDA/p(EGDA-co-PFDA) is deposited on the different substrates, the static contact angle of hexadecane is 120±2°, and the oleophobicity of the barrier layer is significantly improved. This super-hydrophobic and oleophobic behavior can be attributed to the low surface energy of the outermost p(EGDA-co-PFDA) film coupled with its irregular surface micro-nanocone structure. The oil contact angle of pEGDA/p(EGDA-co-PFDA) films on different substrates showed practically no difference, indicating that the surface wettability of the flat substrates without micro/nano structure was mainly determined by the coating and had nothing to do with the substrate. Furthermore, it has been shown that fabricated coatings are highly substrate adaptive. The coating can undergo predetermined functional modifications for different substrate surfaces.
Moreover, for practical applications, the durability of superhydrophobic surfaces against contaminants is also very significant. [52,53] For this purpose, durability tests were conducted on Si samples deposited with pEGDA/p(EGDA-co-PFDA) films. The results showed that the super-hydrophobic surface had stable and durable antifouling ability. At first, surface WCA tests were performed on a batch of Si/pEGDA/p(EGDA-co-PFDA) samples stored in laboratory environment (23 ± 2°C, 50 ± 5% RH) for up to 2 years. The WCA of these samples was found to be 166.3 ± 2.5°, nearly unchanged from 166.0 ± 2.3°2 years earlier. The results indicate that pEGDA/p(EGDA-co-PFDA) films can maintain super-hydrophobic properties for a long time in the natural state. In addition, for the same batch of samples prepared 2 years ago, we additionally conducted accelerated aging test at 85°C/85% RH. As shown in Figure S4  As the film is used as the top layer of the TFE, it will make the TFE practical and durable for antifouling properties.

Optical Transmittance and Conformability
After obtaining the TFE with high barrier properties, we also investigated its optical properties for use in some cases where high optical transmittance is required. The formation of micronanocones also helps to reduce the surface reflectivity of the TFE and improve the optical transmittance. As shown in Figure 4a, we test the UV-vis-NIR spectra of different films on quartz substrates. Instead of studying the transmittance of individual films, we focus on the overall transmittance of the coated sample. This result clearly shows that after coating the 170 nm pPFDA and 220 nm p(EGDA-co-PFDA), the transmittance of the sample is significantly enhanced compared to the bare quartz glass. This is caused by the reduced reflected light and enhanced transmitted light with the help of the coarse nanostructure of the pPFDA surface. As shown in Figure S5 (Supporting Information), compared with the quartz glass deposited with 20 nm Al 2 O 3 , the light transmittance of the quartz glass deposited with 240 nm Al 2 O 3 -p(EGDA-co-PFDA) film is significantly enhanced, especially in the near-infrared light region, which again confirms that the enhanced light transmittance comes from the antireflection effect generated by the rough surface structure of the film. [54] In the meantime, the 2 μm thick TFE deposited quartz glass also shows excellent light transmittance. As shown in Figure  S6a  tance test on the two transparent substrates showed that the surface of p(EGDA-co-PFDA) with the coarse micro-nano structure achieved superhydrophobic properties and enhanced the visible light transmittance of the sample. Considering the excellent WVTR and surface self-cleaning properties of the TFE, a slight reduction in the transmittance is negligible when applied to the device. In addition, to rule out differences in the transmittance of individual quartz glass substrate, we test the transmittance for multiple bare quartz glasses and PET in Figure S7 (Supporting Information). Almost no change in the transmittance can be observed in different samples.
In addition to the advantages of solvent-free, normaltemperature polymerization, strong substrate adaptation, complete retention of monomeric functional groups, and precise control of film thickness, iCVD coatings are also characterized by excellent conformability, that is, keeping the original device microstructure and shape unchanged. [31] This is important for applications in miniaturized integrated optoelectronic devices with complex structures. To demonstrate the ability, we use microelectromechanical systems (MEMS) deep silicon etching technology to prepare high depth-aspect ratio channel structures on a silicon substrate ( Figure S8, Supporting Information). As shown in Figure 4b, the channel is 20 μm wide and 153 μm deep, indicating the depth-width ratio is close to 1:8. After the pEGDA film is deposited on the channel surface, the channel cross-section of the film is obtained using a laser cutting technique. The SEM crosssection shows that the pEGDA film completely covers the entire inner surface of the channel, demonstrating excellent conformational capability. The zoom-in parts (Figure 4b1-b3) show that the films on the channel surface, side and bottom are uniformly coated, perfectly preserving the overall structure of the channel device. Depending on the nature of the conformational deposition and the broad applicability of the substrate, the iCVD-ALD technique will have a unique role and broad applications in micro and nanofabricated surface functional coatings of semiconductor materials.

Conclusions
In summary, we have developed a multifunctional hybrid organic/inorganic TFE technique for flexible encapsulation of devices to enhance their water resistance stability. WVTR of the film with a large area (50 cm 2 ) TFE is as low as 1.65 × 10 −4 g m −2 day −1 at 38°C and 90% RH, and the BIF reaches 10 4 . Under normal conditions, the value is estimated to 9.81 × 10 −6 g m −2 day −1 , which is sufficient to improve the lifetime of water-sensitive devices. To the best of our knowledge, this is one of the lowest WVTR achieved by direct instrumentation of a large-area flexible TFE. A cross-linked copolymerized film is added to the top layer of the TFE. The static WCA and WSA reached 168.4°and 1.1°, respectively, enabling TFE to have the super-hydrophobic and self-cleaning ability. Moreover, the rough micro-nano surface is able to reduce the reflection of light, achieving an average light transmittance of 86.88% in the visible region. We believe that this research can improve the service performance and stability of water-oxygen sensitive optoelectronic devices and broaden their application scenarios. This is essential for practical applications in the near future of optoelectronic devices toward flexibility, miniaturization, and large-scale integration.
Depositing Organic Layer via the iCVD: The organic films were deposited by the iCVD simultaneously on Si wafers (1.5 × 1.5 cm 2 ) for structural and chemical characterization, quartz glass (2.5 × 2.5 cm 2 ) for light transmittance test and 190 ± 5 μm thick PET substrates for the WVTR test and the study of mechanical properties. All Si wafers and quartz glasses were successively washed with deionized water, acetone, isopropyl alcohol, and ethyl alcohol in an ultrasonic bath for 20 min each, and dried in an N 2 flow.
As for iCVD process, the monomer EGDA was heated to 75°C. The initiator TBP volatilizes at room temperature. The EGDA and TBP are evaporated into custom reaction chambers with a flow rate of ≈0.3 standard cubic centimeter per minute (sccm) and 0.6 sccm, respectively, regulated by needle valves, being polymerized to a pEGDA base layer. The substrate temperature was maintained at 40°C and the nickel/chromium resistance filament was heated to 200°C during deposition. The chamber pressure is maintained at 200 mTorr and controlled by a VAT pressure controller valve. The monomer PFDA was heated to 80°C. The flow rates were set to 0.5 and 0.1 sccm for PFDA and EGDA, respectively. For p(EGDA-co-PFDA) crosslinked polymer film, after growing the pEGDA films, TBP and EGDA were stopped and filament heating was switched off. Only the PFDA monomer was fed at a flow rate of 0.5 sccm. This state was maintained for 1 min after the chamber pressure exceeded the saturation vapor pressure of the PFDA monomer, which was maintained at 180 mTorr. The filament was then switched on for heating, and TBP and EGDA were injected at a flow rate of 0.6 and 0.1 sccm, respectively. The total pressure in the control chamber is 200 mTorr. p(EGDA-co-PFDA) was obtained as a top layer. In addition, the iCVD was equipped with a 633 nm helium-neon laser (JDS Uniphase) to monitor the thickness of the film in real time.
Depositing Al 2 O 3 via the ALD: Al 2 O 3 films of various thicknesses were deposited in a custom-built stationary iCVD-ALD system, using TMA and deionized water as precursor. All the depositions of Al 2 O 3 were performed at 110°C.
The deposition pulse/exposure/purge parameters of precursor were set to 0.2 s/10 s/20 s for TMA and 0.4 s/20 s/20 s for H 2 O, respectively. Both the precursor pulse and the purge step were performed with a nitrogen flow of 30 sccm as the carrier gas.
Film Characterization: The cross section and surface image of the organic/inorganic hybrid multilayer was observed by a scanning electron microscope (SEM, Nano SEM 650, FEI, USA). Each sample was coated with a gold film with a thickness of about 1 nm and scanned at a voltage of 5-10 kV. The surface morphology and the roughness of the deposited films were examined with atomic force microscopy (AFM, Bruker, Dimension ICON, USA) www.advancedsciencenews.com www.advmatinterfaces.de under noncontact mode using NCHR-50 (Nanosensors) as cantilever. WCA and water WSA data were collected using a contact angle goniometer (Kruss DSA 100S, Germany) equipped with an automatic liquid dispenser. The WCA of each sample was measured using 8 μL of deionized water droplets, and the final WCA values were averaged from measurements of five different spots on the sample surface. The WSA data were obtained by measuring the tilt angle of the substrate at which a droplet placed on it could slide off via a manually tilted stage. The chemical bonds of the pEGDA, pPFDA, and p(EGDA-co-PFDA) polymer film were analyzed by Fourier transform infrared (FTIR) spectrometer (Varian, Excalibur 3100, USA). Spectra were acquired from 4000 to 600 cm −1 with a resolution of 4 cm −1 repeating 64 scans. The chemical compositions and states of the deposited films were measured by X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific, ESCALAB 250Xi, USA) analysis. The XPS analysis were performed on PET samples coated with TFE film and the sample size was 3 × 3 mm 2 . The XPS was equipped with a monochromatic Al K X-ray source (1486.60 eV). The survey scans and the high-resolution scans were analyzed with a pass energy of 100 and 30 eV, respectively, both under high vacuum (P < 10 −7 mbar). The survey scans started from 1350 to −10 eV taking 1 eV steps with a dwell time of 50 ms. The high-resolution scan has a step size of 0.1 eV. And all measured spectra were calibrated using the C1s spectrum at 284.80 eV. Atomic concentrations were calculated in Thermo Scientific Avantage software version 5.965. The optical transmittances of the films were measured using a UV-vis-NIR spectrophotometer (Varian Cary 5000, USA). The spectra were taken from 190 to 2500 nm, and the baseline was set to air so that the transmittance data included the transparency reduction due to the 0.5 mm thick bare quartz glasses. To analyze the crystalline behavior of the thin Al 2 O 3 film, the material was deposited by ALD on an n-type (crystal orientation [100]) silicon wafer. The crystal pattern of ALD Al 2 O 3 was analyzed by X-ray single crystal diffractometer (XRD, Rigaku Smart-Lab, Japan), where copper target (Cu K X-rays, = 0.15 405 nm) working at 40 kV and 40 mA and scattering 2 scanning from 3°t o 90°with a scanning speed of 2°min −1 was used. The SEM attachment of focused ion beam (FIB, Hitachi FB-2100, Japan) system was used to observe the cross-section morphology of 3 × 3 mm 2 silicon sample with high aspect ratio structure. The electron beam energy was 10 kV and the sample surface was sputtered with a gold film of about 1 nm to increase the conductivity.
Barrier Property Measurement: Different dyads of TFE were deposited onto the PET film and the WVTR of the film was directly tested by a water vapor transmittance analyzer (MOCON AQUATRAN 2, USA). The MOCON AQUATRAN 2 instrument used a high sensitive Coulomb sensor to measure the quality of water vapor permeating from the coated PET sample and provides a real-time WVTR-Time graph with specialized software. The PET sample size was 10 cm × 10 cm × (190 ± 5 μm), and the effective test area was 50 cm 2 . Due to the high permeability of the PET film with respect to the TFE, the WVTR of the TFE is believed to be approximately the WVTR of the entire coated PET. [55] The temperature, humidity, and ambient pressure of front chamber were set to 38°C, 90% RH and 0.1 MPa, respectively. The instrument underwent 2.5 h of calibration and environmental preparation prior to each measurement. The sample WVTR values are taken from the steady state. Namely, the measurement ends when the curve approaches a line parallel to the time axis, and the result of the last measurement is taken as the sample WVTR.
Durability Measurement: The film's durability was characterized by accelerated aging tests in a constant temperature and humidity test chamber (EYELA KCL-1000, Tokyo Rikakikai, Japan). The surface static WCA of the Si sample was measured before the aging test. The sample to be tested was placed after the test chamber had been operated for 6 h to keep the temperature and humidity (85°C/85% RH) constant. Samples were subsequently taken out every 24 h for testing. The surface WCA was measured after natural cooling of the sample for 20 min. The final WCA values were averaged from measurements of five spots on the sample surface. Immediately after the measurement, the sample is put back into the test chamber and retimed. The accelerated aging test was performed for eight consecutive days.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.