Influence of the polymeric substrate on the water

Surface & Coatings Technology 337 (2018) 44–52

Contents lists available at ScienceDirect

Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat

Influence of the polymeric substrate on the water permeation of alumina barrier films deposited by atomic layer deposition

T

⁎

Karyn L. Jarvis , Peter J. Evans, Gerry Triani Institute of Materials Engineering, Australian Nuclear Science and Technology Organisation, Lucas Heights, NSW 2232, Australia

A R T I C L E I N F O

A B S T R A C T

Keywords: Al2O3 Atomic layer deposition Barrier films Polymers WVTR XPS

Atomic layer deposited (ALD) barrier films have been deposited onto a wide variety of flexible polymeric substrates to determine their effectiveness as moisture barriers for organic electronics. Little research has however been conducted on the contribution of the substrate to the barrier properties. In this study, alumina (Al2O3) barrier films have been deposited onto different polymeric substrates by ALD to investigate the effect of the substrate type and thickness on the water vapour transmission rate (WVTR). 24 nm Al2O3 films were deposited via plasma enhanced ALD onto 75 and 125 μm thick polyethylene terephthalate (PET) and polyethylene naphthalate (PEN) substrates. Half of the substrates were also O2 plasma pre-treated prior to Al2O3 film deposition to determine its effect on the WVTR. The WVTR of the substrates prior to barrier film deposition was measured using tritiated water (HTO) permeation. Prior to barrier film deposition, it was shown that the WVTR decreased as the substrate thickness increased while PEN had a lower WVTR than PET. After Al2O3 barrier film deposition, the WVTR followed the previously observed trend with lower WVTR for thicker substrates and for PEN over PET. The substrates O2 plasma pre-treated prior to barrier film deposition also showed lower WVTRs, which were attributed to surface cleaning and improved film adhesion. The lowest WVTR measured was 3.1 × 10− 3 g·m− 2/day for a 24 nm Al2O3 film deposited onto O2 plasma pre-treated 125 μm PEN. These results demonstrate that the properties of the polymer substrate influence the WVTR even after barrier film deposition and can therefore be used to improve the barrier properties.

1. Introduction Significant research into the development of organic photovoltaic cells (OPVs) and organic light emitting diodes (OLEDs) has been carried out over recent years. Polymers are typically used as the substrates for organic electronics such as OPVs and OLEDs as they are lightweight, cheap, transparent, printable and flexible. Polymer substrates however have high gas/vapour permeability. Barrier films are therefore required for both OPVs and OLEDs to achieve sufficient lifetimes for commercial application by preventing degradation from exposure to water vapour and oxygen [1]. It is generally believed that OLEDs require water vapour transmission rates (WVTR) in the range of of 10− 6 g·m− 2/day to result in sufficient lifetimes [2]. Atomic layer deposition (ALD) is an ideal technique for the deposition of barrier films as it is a self-limiting technique where gas-phase deposition is used to produce conformal pinhole-free inorganic coatings. Atomic layer growth is achieved by alternate pulsing of precursor gases and inert gases. Inert gas pulses are used to clear the reactor chamber of excess precursor and by-products [3]. Inorganic layers produced by ALD have achieved lower water

permeation with thinner layers than other techniques due to film integrity [4]. A number of metal oxide barrier films can be deposited using ALD, but alumina (Al2O3) is the most frequently used as a barrier film [5]. ALD frequently requires temperatures > 200 °C to produce dense, pinhole free films or requires plasma enhancement. Many polymers are however fragile with glass transition temperatures < 100 °C, therefore high temperature ALD is not suitable to protect organic electronic devices [6]. An advantage of Al2O3 barrier layers is the relatively high deposition rate (> 1 Å/cycle) at temperatures below 150 °C [7,8], which is make it suitable for temperature sensitive polymers. A number of studies have investigated the effect of single Al2O3 barrier films on the WVTR by varying a number of parameters such as film thickness and deposition temperature. It has been shown that increasing either the film thickness [9,10], deposition temperature [11,12] or both [13–19] decreases the WVTR. A wide range of WVTR values have been reported for single Al2O3 barrier films, which demonstrate that a number factors influence the WVTR. In particular, the WVTR is dependent on the temperature and relative humidity during measurement

⁎ Corresponding author at: ANFF-Vic Biointerface Engineering Hub, Faculty of Science, Engineering and Technology, Swinburne University of Technology, Hawthorn, VIC 3122, Australia. E-mail address: [email protected] (K.L. Jarvis).

https://doi.org/10.1016/j.surfcoat.2017.12.056 Received 23 August 2017; Received in revised form 15 December 2017; Accepted 23 December 2017 Available online 26 December 2017 0257-8972/ © 2018 Elsevier B.V. All rights reserved.


K.L. Jarvis et al.

increase the number of polar surface groups and thus the higher surface energy leads to more homogeneous films and therefore enhanced barrier properties for thinner Al2O3 films [33]. A number of studies have shown that corona, air and oxygen plasma typically increase the oxygen concentration, increase surface energy, decrease contact angle and increase surface roughness [34–39]. A few studies have investigated the effect of corona or plasma pretreatment on moisture permeation. PP substrates have been corona and/or O2 plasma treated prior to the deposition of AlOx films by vacuum web coating. O2 plasma treatment decreased surface roughness and increased the oxygen content on the surface which increased the adhesion of AlOx. For smoother surfaces, there were less shadowing effects during deposition and therefore the films were more homogeneous with fewer defects [40]. In addition, Al2O3 and silica (SiO2) films have also been deposited via ALD onto PE and PLA coated paperboard. Corona treatment decreased the contact angle and decreased some of the WVTR values, most noticeably for 25 nm Al2O3 and SiO2 films on PE coated paperboard [32]. The present research investigates the effect of substrate thickness, type and O2 plasma pre-treatment on the WVTRs of PET and PEN substrates before and after the deposition of Al2O3 barrier films by plasma enhanced ALD (PE-ALD). The effect of O2 plasma pre-treatment on the surface of PET and PEN substrates was analysed by X-ray photoelectron spectroscopy, contact angle measurements and atomic force microscopy. The WVTRs before and after the deposition of Al2O3 barrier films were determined by tritiated water permeation. The substrate thickness, substrate type and O2 plasma pre-treatment of PET and PEN substrates have all been shown to influence the WVTRs and can therefore be used to increase the effectiveness of Al2O3 barrier films.

such that increases in either will increase moisture permeation. For single Al2O3 barrier films, the WVTR measurement conditions have been varied from ambient temperature and humidity [13] to 100% RH [9] and 85 °C [20]. Thus the WVTRs for a single Al2O3 barrier film have ranged from 10 g·m− 2/day for a 25 nm thick film deposited onto a 40 μm polylactic acid (PLA) substrate with thermal roll to roll ALD at 100 °C [21] down to 1.7 × 10− 5 g·m− 2/day for a 25 nm thick Al2O3 film deposited onto a 125 μm polyethylene naphthalate (PEN) substrate with thermal ALD at 120 °C [20]. ALD barrier films have been deposited onto a number of different polymer substrates, including polyethylene terephthalate (PET) [22], polyetheretherketone (PEEK) [9], polyethersulfone (PES) [12,14,23], polycarbonate (PC) [14], PEN [13,20,24], PLA [21] and polyimide (PI) [21,25]. Prior to film deposition, these polymer substrates were found to have a wide variety of WVTRs: 1.4–3 g·m− 2/day for PET [26,27], 3.1 g·m− 2/day for PEEK [9], 34.1–92.8 g·m− 2/day for PES [12,16], 50 g·m− 2/day for PC [14], 0.5–1.3 g·m− 2/day for PEN [28,29], 39–53 g·m− 2/day for PLA [3,21] and 0.2–3 g·m− 2/day for PI [21,25]. WVTR is however not the only factor when choosing a substrate for OPVs and OLEDs. PEN has one of the lowest WVTRs but is more expensive than other substrates with a previously reported price of 22 euros per kg while PET is significantly cheaper at 8 euros per kg [30]. Substrate price is therefore an important consideration in the development of large scale OPVs and OLEDs as it can add significant costs to the device. Although a number of different polymers have been used as substrates for ALD barrier film, only a few studies have deposited the same Al2O3 films onto different substrates. 25 nm Al2O3 films were deposited on cellulose, PLA and PI by roll to roll ALD. The WVTRs for the uncoated substrates at 23 °C and 50% RH, were 144, 39 and 3 g·m− 2/day respectively. After Al2O3 film deposition, the corresponding WVTRs decreased to 15, 10 and 2 g·m− 2/day [21]. The effect of the substrate on the WVTRs of different coated papers has also been investigated. The papers were extrusion coated with low density polyethylene (LDPE), polypropylene (PP), PET and PLA which had WVTRs of 2.8, 1.1, 22 and 72 g·m− 2/day respectively at 23 °C and 50% relative humidity. The deposition of 100 nm thick Al2O3 films via spatial ALD decreased the WVTRs to 0.5, 0.3, 0.5 and 4.3 g·m− 2/day respectively [31]. Another study investigated the barrier performance of single Al2O3 layers, approximately 12 nm thick, deposited on PES, PC and PEN substrates which had uncoated WVTRs of 60, 50 and 2 g·m− 2/day respectively. These WVTRs were reduced to 4.1 × 10− 3, 4 × 10− 3 and < 4 × 10− 3 g·m− 2/day respectively after Al2O3 deposition [14]. In all studies, the highest WVTR after the deposition of an Al2O3 barrier film was reported for the substrate with the highest uncoated WVTR, thus suggesting that the substrate does play a role in the WVTR of barrier film deposited onto a polymeric substrate. Treatment of the polymer surface by corona or plasma treatment has been shown to influence the surface chemistry, roughness and contact angle. Varying these substrate properties can influence ALD film deposition and therefore also the WVTR. During corona treatment, the surface is bombarded with oxygen, free radicals of oxygen and ozone generated by electrical discharge, all of which oxidize the surface. A similar effect occurs with plasma pre-treatments where the surface is exposed to an oxygen or air plasma generated by a radio frequency discharge. Polymer films are typically chemically inert with low surface energy which results in poor wetting and adhesion of deposited films. The surface oxidation induced by corona and plasma treatment typically increases the surface energy thus improving film adhesion. The surface chemistry, roughness and contact angle of the film may affect ALD deposition, especially during the initial stages of film formation. One of the reasons behind such behaviour may be the different amount of hydrogen bonded water on the surface. Corona treatment has shown to increase both the oxygen content and CeO functionalities of polyethylene (PE) and can also remove low molecular weight contaminants [32]. Both corona and plasma pre-treatment

2. Experimental 2.1. Materials The biaxially oriented PET substrates were cut from a 75 μm thick, 300 mm wide roll (Multapex Pty. Ltd.) and from 125 μm thick, 600 mm wide roll (Goodfellow, UK). The biaxially oriented PEN pieces were cut from a 75 μm thick, 600 mm wide roll and from 125 μm thick 300 × 300 mm2 sheets (both from Goodfellow, UK). The structures of the PET and PEN repeating units are shown in Fig. 1. Trimethylaluminium (TMA) (99.999%) was purchased from Strem Chemicals. Liquid scintillation cocktail (Ultima Gold™ uLLT) and tritiated water (37 MBq/

n Polyethylene terephthalate

n Polyethylene naphthalate Fig. 1. Structure of PET and PEN repeating units.

45


K.L. Jarvis et al.

water which permeates through the substrate. The vessel port was closed and the apparatus was left for 1, 2, 3 or 7 days. A relative humidity of ~95% was measured in the lower part of the chamber during each experiment with an average ambient temperature of 25 °C. After the desired duration, the port at the top of the vessel was opened and the LiCl vial removed and sealed. To ensure the experimental duration is precisely known, the LiCl needs to be removed when there is still HTO remaining in the well. At the completion of each run, the glass vessel was removed from the stainless steel base and both the inside of the glass vessel and the top surface of the substrate were each rinsed with 50 mL of Milli-Q water to capture any HTO which may have permeated through the film but not absorbed by the LiCl. To determine the total mass of HTO permeated through the substrate, three separate samples were analysed using a liquid scintillation counter: glass rinse, substrate rinse and the LiCl from the vessel. 7 mL of Milli-Q water was added to the vial containing LiCl. In two other vials, 3 g of LiCl was weighed into each. 7 mL of the glass rinse was added to one vial and 7 mL of the film rinse was added to the other. To ensure complete dissolution, these solutions were left overnight. On the following day, 13 mL of scintillation cocktail was added to each of the three vials, followed by vigorous shaking. These solutions were once again left overnight. Finally, the samples were placed in the liquid scintillation counter (Packard Tri-Carb 2900TR) and the counts per minute for each sample were determined. Each sample was analysed a minimum of 20 times from which the average counts per minute was determined. The counts per minute were used to calculate the WVTR using a linear regression equation that was determined from the calibration of the 10 MBq/mL tritiated water working stock solution. Vials with total activities of 0.005–50 KBq were made from the 10 MBq/mL working stock solution and the counts per minutes were measured then plotted against the total activity. The WVTR (g·m− 2/day) was calculated using the following equation:

mL) were purchased from Perkin-Elmer. 2.2. Plasma enhanced atomic layer deposition (PE-ALD) The deposition of PE-ALD alumina films onto PET and PEN was undertaken using a Cambridge Nanotech ALD Fiji F200 with TMA and O2 plasma as the precursors. 150 mm diameter circular substrate pieces were cut from the roll or sheets and held flat during deposition using a 1 cm wide aluminium ring. Several pieces of silicon wafer were also simultaneously coated to determine film thickness by spectroscopic ellipsometry. Prior to film deposition, half of the PET and PEN pieces were in-situ O2 plasma pre-treated for 10 min at an O2 flow of 80 sccm and plasma power of 300 W. All substrates were coated on one side with an ALD Al2O3 film at 120 °C for 191 cycles which yielded thicknesses of approximately 24 nm. In the case of the O2 plasma pre-treated samples, Al2O3 film deposition was performed directly after the pretreatment without removal from the vacuum chamber. Pulse/purge times of 0.02 s/30 s and 20 s/5 s were used for the TMA and 300 W O2 plasma respectively. Thicknesses of the Al2O3 ALD films deposited onto silicon wafers were measured using a J.A. Woollam M2000XI spectroscopic ellipsometer. The Al2O3 films were deposited onto silicon wafers rather than PET/PEN for ellipsometry measurements as the difference between the refractive indices of the Al2O3 films and PET or PEN were not sufficient to accurately determine the thicknesses. Measurements were undertaken over the wavelength range of 210–1687 nm and at incidence angles of 60°, 65°, 70° and 75°. A model was fitted to the acquired data using the CompleteEase software and was comprised of a silicon substrate with a 1.5 nm native oxide layer underneath an Al2O3 Cauchy layer. Mean squared error (MSE) values of approximately 3 were observed for Al2O3 films deposited onto both untreated and plasma pretreated silicon. Thicknesses of the PE-ALD Al2O3 films on silicon were determined to be 23.8 nm and 24.7 nm for untreated and O2 plasma pre-treated silicon substrates respectively.

WVTR =

2.3. Tritiated water (HTO) permeation

Dv = droplet volume (μL) CPM = counts per minute c = y-axis intercept of linear regression (CPM) m = slope of linear regression (CPM/KBq) A = surface area of film (m2) D = experiment duration (days) iAc = initial total activity of droplet (KBq).

A 1 mL 10 MBq/mL tritiated water working stock solution was made by adding 270.3 μL of the 37 MBq/mL stock to 729.7 μL of Milli-Q water in an Eppendorf tube. 5 μL of this 10 MBq/mL solution was then pipetted into the well in the base of the stainless steel HTO rig, shown in Fig. 2, followed by 45 μL of Milli-Q water which resulted in a total droplet activity of 50 KBq. The PET or PEN substrate was placed over the base, followed by a Teflon seal which results in available film area of 0.009 m2. For the Al2O3 coated substrates, the Al2O3 side faced down. The glass vessel was then placed on top and bolted in place. The port on the top of the vessel was opened and a vial with 3 g of LiCl was placed in the vial holder. LiCl is hydroscopic and therefore absorbs the

2.4. X-ray photoelectron spectroscopy (XPS) XPS analysis of the uncoated and Al2O3 coated PET and PEN substrates were carried out using a Thermo Scientific ESCALAB250Xi fitted with a monochromated Al Kά X-ray source (hν = 1486.6 eV) at an Xray power of 150 W and a spot size of 500 × 500 μm2. Survey spectra were collected over an energy range of 0–1354 eV and a pass energy of 100 eV and a resolution of 1 eV. High resolution C 1s and O 1s spectra were collected at a pass energy 20 eV and a resolution 0.1 eV. Surface adventitious carbon was removed from the PE-ALD alumina films by sputtering for 20 s with a monoatomic argon gun at 1 keV with an etch area of 2.5 × 2.5 mm2. The spectra were curve fitted using CASA XPS software. The binding energies of the peaks were normalized by shifting the binding energy of the C1s peak for CeC to 285 eV. The shift between 285 eV and the CeC binding energy was applied to all peaks.

LiCl

tritiated water

polymer substrate & ALD film

Dv(CPM + c) 1000mADiAc

O-ring seal

2.5. Atomic force microscopy (AFM) The surface topography of the uncoated PET and PEN substrates were analysed using a Digital Instrument D3000 AFM with a Nanoscope III controller. 5 × 5 μm2 regions were imaged in tapping mode with a scan rate of 1 Hz and a tip velocity of 10 μm/s. AFM data was processed

stainless steel base Fig. 2. Schematic diagram of the HTO permeation rig.

46


K.L. Jarvis et al.

using NanoScope Analysis version 1.40 software. Images were flattened using a 2nd order flatten. The root mean square (RMS) roughness values were calculated from a minimum of 2 images.

XPS Atomic Concentration (%)

80

2.6. Contact angle Static contact angles of the untreated and plasma pre-treated PET and PEN substrates were measured using a custom contact angle goniometer. The goniometer consisted of a sample stage, a pipette mounted over the sample and camera. A 10 μL droplet of Milli-Q water was lowered onto the surface using the pipette and an image of the droplet on the surface captured using the camera. Three droplets were imaged in three different positions on each substrate. The contact angle of each droplet was determined using the angle tool in ImageJ. The three measurements were then averaged to determine the average contact angle for each substrate.

a.

70 60 50 40 30 20 10 0 C

3. Results and discussions

75 µm PET untreated 125 µm PET untreated 75 µm PEN untreated 125 µm PEN untreated

3.1. Effect of plasma pre-treatment on surface roughness and chemistry The effect of O2 plasma pre-treatment on the surface chemistry of 75 and 125 μm PET and PEN were investigated by XPS. In Fig. 3a, the atomic concentrations from the XPS survey spectra show that for all PET and PEN substrates, the oxygen concentrations increase and the carbon concentrations decrease after O2 plasma pre-treatment. The oxygen concentration of all substrates was found to increase by between 5.2 and 11% after O2 plasma pre-treatment, an effect which is well documented in the literature [37,38,41]. The C:O ratio decreased after plasma pre-treatment. Initially the PET and PEN had similar C:O ratios of between 3 and 4 which then decreased to approximately 2 and 2.5 for PET and PEN respectively. The lower C:O ratios for PET suggest that more oxygen is added to the surface of PET than PEN by plasma pre-treatment. Three peaks were fitted to the high resolution C 1s spectra of untreated and plasma pre-treated PET and PEN substrates: aromatic CeC, CeH at 285 eV, carbon singularly bonded to oxygen CeO at 286.5 eV and ester O]CeO at 288.9 eV [37], as shown in Fig. 4. Two peaks were fitted to the high resolution O 1s spectra: OeC]O at 532 eV and O]CeO at 533.5 eV [42,43], as shown in Fig. 5. In Figs. 3b, 4 and 5 it can be observed that plasma pre-treatment increased the concentration of the C 1s CeO, O]CeO and O 1s O]CeO peaks while the C 1s CeC and O 1s OeC]O concentrations decreased. The increase in the C 1s CeO and O]CeO peaks accompanied by a decrease in the CeC peaks was attributed to the increases in the overall oxygen concentration on the surface after plasma pre-treatment, which has previously been observed [41]. Prior to plasma pre-treatment, the high resolution O 1s spectra of all four surfaces show approximately 1:1 ratios for CeO and O]CeO peaks, as seen in Fig. 5, which is agreement with previous work [41]. After plasma pre-treatment, the ratio of these peaks changes with significantly more O]CeO than OeC]O groups. The combination of the atomic concentrations from the C 1s and O 1s spectra, suggest that while plasma pre-treatment increases the overall concentration of both the CeO and C]O species on the surface of PET and PEN, significantly greater concentrations of CeO are generated on the surface. Such behaviour is in contrast to a previous study which showed that both PET and PEN retained a 1:1 O 1s O]CeO to OeC]O ratio after plasma pre-treatment [41]. In that study however the maximum plasma treatment time was 2 s whereas in this work the surface was treated for 10 min, which suggests that treatment time may influence the CeO to O]CeO ratio. The contact angle of the PET and PEN were measured before and after plasma pre-treatment to investigate its effect on hydrophilicity. The PET and PEN substrates had initial contact angles ranging from approximately 68° to 80°, as shown in Fig. 6. The 75 μm PET surface had the highest initial contact angle of almost 80° in comparison to the

O 75 µm PET plasma pre-treated 125 µm PET plasma pre-treated 75 µm PEN plasma pre-treated 125 µm PEN plasma pre-treated


80

b.

70 60 50 40 30 20 10 0 C-C

C-O

C 1s

O=C-O O=C-O O=C-O

O 1s

Fig. 3. a. XPS survey and b. high resolution C 1s and O 1s atomic concentrations of untreated and O2 plasma pre-treated (300 W, 10 min) PET and PEN substrates.

approximately 70° for the other three surfaces. Such a difference suggests that the 75 μm PET surface may have more carbonaceous surface contamination than the other samples, thus increasing the contact angle [37]. Greater surface contamination on the 75 μm PET was not unexpected as this industrial grade material was received without protective packaging. In contrast, the other three substrates were research grade materials that were supplied in protective packaging and subsequently stored in a laboratory environment. Excluding the 75 μm PET, PET and PEN appear to have similar wettability. After plasma pretreatment, the contact angle of all 4 samples decreased by approximately 40°, as shown in Fig. 6. Decreases in the contact angle after O2 plasma treatment have previously been observed in the literature [37,38]. The increased wettability after plasma pre-treatment can be attributed to the increased concentration of CeO and C]O on the surface. The addition of these polar groups increases the surface energy of the surface and thus increases its wettability. AFM images of untreated and plasma pre-treated PET and PEN substrates were taken to determine the effect of plasma pre-treatment on surface roughness. The surface roughness of both PET substrates 47


K.L. Jarvis et al.

O=C-O

O=C-O

C-O 2

O-C=O

C-C, C-H

2

plasma pre-treated

plasma pre-treated

Normalised counts per second

1.8

Normalised counts per second

1.6

1.4

1.2

1

untreated 0.8

1.5

1

untreated

0.5

0.6

0.4

0 538

536

534

532

530

528

Binding energy (eV)

0.2

Fig. 5. XPS curve fitted high resolution O 1s spectra for untreated and O2 plasma pretreated (300 W, 10 min) 75 μm thick PET substrates.

0 292

290

288

286

284

282

280

90

Binding energy (eV)

Untreated

Plasma pre-treated

80

Fig. 4. XPS curve fitted high resolution C 1s and spectra for untreated and O2 plasma pretreated (300 W, 10 min) 75 μm thick PET substrates.

Contact angle (°)

70

decreased after plasma pre-treatment while in contrast, they increased for both PEN substrates, as shown in Fig. 7. The untreated PET and PEN surfaces are relatively smooth, as shown in Fig. 8, with average RMS roughness values between approximately 1 and 2.5. The increase in average RMS roughness after plasma pre-treatment for PEN but not PET is an interesting phenomenon. The reason for such behaviour at this stage is unknown. The decreases in the surface roughness after PET plasma pre-treatment suggest that the surface is cleaned and is thus reducing the surface contaminants which contribute to the surface roughness. Such decreases however indicate that the plasma pretreatment is not roughening the surface as was seen with PET in previous studies [38,39], where the surface roughness significantly increased after both air and oxygen plasma treatment. In these studies, plasma pre-treatment resulted in the formation of cigar-like structures, similar to those in the AFM images for PEN in Fig. 8. The formation of such structures has been attributed to the surface bombardment by plasma species. Thus their absence on the PET surfaces suggests that they were not significantly affected by plasma bombardment. This behaviour may be substrate dependent as the PET used herein was biaxially orientated, while the properties of PET substrates in previous studies were not stated.

60 50 40 30 20 10 0 75 µm PET

125 µm PET

75 µm PEN

125 µm PEN

Fig. 6. Contact angle of untreated and O2 plasma pre-treated (300 W, 10 min) PET and PEN substrates.

permeation was undertaken for test durations of 1, 2 and 3 days for both PET and PEN. A test duration of 7 days was only used for PEN as complete droplet evaporation had occurred prior to this time for the PET substrates. As can be seen in Fig. 9, the WVTRs for both thicknesses of PET and PEN were relatively constant across the different test durations. Such behaviour suggests that 1 day is sufficient for the system to reach steady state conditions and thus enabling an accurate determination of the WVTR. Average WVTR values of 1.5 g·m− 2/day (75 μm PET), 1.1 g·m− 2/day (125 μm PET), 4.6 × 10− 1 g·m− 2/day (75 μm PEN) and 3.0 × 10− 1 g·m− 2/day (125 μm PEN) were calculated. These values show that increasing the substrate thickness of both PET and PEN substrates reduces the WVTR. Such behaviour was

3.2. Effect of the substrate on WVTR of alumina barrier films The WVTRs of untreated PET and PEN substrates were measured using HTO permeation to determine their initial WVTR prior to Al2O3 barrier film deposition. To monitor the transmission kinetics, HTO 48


K.L. Jarvis et al.

5

Untreated

in this study, if the substrate is regarded as a second barrier layer, where increasing its thickness and/or decreasing its WVTR also reduces the WVTR of the entire structure. The importance of the substrate has also been demonstrated by modeling water permeation through a substrate/barrier system which concluded that the permeability of the barrier layer determines the transport time scale which the substrate defines the driving force of transport [46]. The effect of plasma pre-treatment on WVTR was determined by comparing the WVTR of untreated and plasma pre-treated substrates coated with 24 nm Al2O3 films. Plasma pre-treatment prior to barrier film deposition reduced the WVTR of all substrates, as shown in Fig. 10. The most significant reduction was for the 75 μm PET where the WVTR was reduced from 6.8 × 10− 2 to 1.3 × 10− 2 g·m− 2/day, an 80% decrease. The WVTR for 125 μm PET, 75 μm PEN and 125 μm PEN decreased by 60%, 8% and 50% respectively. For three out of the four substrates, plasma pre-treatment results in WVTRs that follows the previously observed trends of lower WVTR as the thickness is increased and for PEN relative to PET. The same behaviour may not be observed here for all substrates as the improvement in WVTR after plasma pretreatment is surface dependent rather than thickness dependent. The reduction in the WVTR for both PET and PEN is attributed to improved adhesion and removal of surface contaminants. O2 plasma pre-treatment has previously been shown to increase the adhesion of alumina films to a polymer substrate due to the increase in oxygen groups [40]. It is therefore proposed that the increase in oxygen groups on both the PET and PEN surfaces, as shown in Fig. 3, improves the adhesion of the Al2O3 barrier film thus reducing the likelihood of damage during handling. Removal of surface contaminants is also likely to play a role in the reduction of WVTR by plasma pre-treatment. The untreated PET and PEN substrates received no cleaning prior to barrier film deposition. The high conformality of ALD films means that any contaminants on the surface, such as dust particles, will be coated. The coating of these surface contaminants may then result in defects as such high points are easily abraded during handling [47]. By removing surface contaminants prior to barrier film deposition, the defect density is reduced thus lowering the WVTR. A lower level of surface contamination initially may also explain the comparatively smaller reduction in the WVTR of the plasma treated 75 μm PEN after Al2O3 film deposition. For a plasma pre-treated silicon wafer, an Al2O3 film thickness of 24.7 nm was determined via spectroscopic ellipsometry. For the untreated surface, the Al2O3 film was however slightly thinner at 23.8 nm. The increased thickness may be due to run to run variability or may suggest that the surface cleaning effect of the O2 plasma pre-treatment results in a slightly thicker film. Although plasma pre-treatment results in lower WVTR values, it is expected that the slight increase in film thickness does not significantly contribute to such reductions. The effect of O2 plasma pre-treatment on the chemistry of Al2O3 films deposited onto 75 and 125 μm PET and PEN was investigated by XPS. The surfaces of the films were analysed prior to argon sputtering and were found to have an average of 55% oxygen, 32% aluminium and 12% carbon, as shown in Fig. 11a. For a pristine Al2O3 film, it would be expected to be composed of 60% oxygen and 40% aluminium. However the carbon concentration, proposed to be due to adventitious carbon, reduced both the expected oxygen and aluminium concentrations. The surfaces were therefore sputtered for 20 s with a monoatomic argon gun to remove any adventitious carbon and therefore determine the actual aluminium and oxygen concentrations of the films. The survey atomic concentrations of oxygen, carbon and aluminium of the sputter cleaned films are shown in Fig. 11b. The Al2O3 films were composed of approximately 60% oxygen, 39% aluminium and 1% carbon, close to the expected Al:O ratio of 2:3. The low concentration of carbon within the film demonstrates the virtually complete conversion of TMA to Al2O3 at 120 °C. For a Al Kα XPS X-ray source, carbon has a sampling depth of 8.7 nm [48]. Therefore, the underlying PET or PEN substrate after the deposition of 24 nm of Al2O3 will not contribute to the spectra. Comparing the atomic concentrations of oxygen, carbon and aluminium for

Plasma pre-treated

RMS Roughness (nm)

4

3

2

1

0 75 µm PET

125 µm PET

75 µm PEN

125 µm PEN

Fig. 7. AFM RMS surface roughness of untreated and O2 plasma pre-treated (300 W, 10 min) PET and PEN substrates.

expected as a thicker substrate will slow the permeation of water. For both substrate thicknesses, PEN had a lower WVTR than PET, which is consistent with the previously reported values for PET [26,27] and PEN [28,29]. Comparisons between different studies are however difficult due to differences in substrate thickness or WVTR measurement conditions. It is proposed that the lower WVTR values for PEN compared with PET of the same thickness is due to differences in their repeating polymer unis, as shown in Fig. 1.The repeating unit of PET only has one aromatic ring while PEN has two. It is proposed that the increases in the number of aromatic rings lead to an increased packing density of the polymer chains, thus reducing the WVTR by retarding water permeation. The WVTRs of the untreated PET and PEN substrates coated with 24 nm Al2O3 films by PE-ALD were measured to determine the effect of the Al2O3 barrier film on the different substrates. The WVTRs for the 75 and 125 μm thick PET and PEN after Al2O3 deposition are shown in Fig. 10. For the PET substrates, the WVTR decreases from 6.8 × 10− 2 to 4.7 × 10− 2 g·m− 2/day when the substrate thickness was increased from 75 to 125 μm. For the PEN substrates, the WVTR decreases from 1.2 × 10− 2 to 6.1 × 10− 3 g·m− 2/day. In comparison to the uncoated substrates, the addition of a 24 nm Al2O3 barrier film reduced the WVTR by approximately 1.5 orders of magnitude for all four samples. For both the 75 and 125 μm thick substrates, lower WVTRs were observed when the 24 nm Al2O3 films were deposited onto PEN rather than PET. Such behaviour follows the trend of the WVTR of the untreated substrates in Fig. 9, therefore indicating that the initial WVTR of a polymer substrate does influence the WVTR after the deposition of an Al2O3 barrier film. The combined effect of the polymer substrate and barrier film can be attributed to the fact that the WVTR measures water that has permeated through both layers. Although the barrier layer significantly retards water permeation, the polymer substrate also plays an important role in determining the final WVTR of the system. In particular, the present study shows that increasing the substrate thickness and/or the type of polymer can both be used to reduce the WVTR without altering the barrier film. In this respect, the present work is in general agreement with previous studies that have reported a similar trend [14,21,31]. Water permeation in a substrate/barrier system such as this can be considered similar to multilayer barrier systems [44] where the combination of the substrate and barrier film reduces the WVTR by blocking defects present in one layer which therefore inhibits the continuous pathways for moisture diffusion through the entire film [11,45]. In multilayer systems, increasing the thickness [9,10] or reducing the WVTR [14,15] of one layer decreases the overall WVTR of the entire structure. Similar behaviour is observed 49


K.L. Jarvis et al.

Untreated

Plasma pre-treated 75 µm PET

125 µm PET

75 µm PEN

125 µm PEN

Fig. 8. 3D AFM images of untreated and O2 plasma pre-treated (300 W, 10 min) PET and PEN substrates.

4. Conclusions

untreated and plasma pre-treated PET and PEN, it appears that the underlying substrate had no effect on the chemistry of the Al2O3 films. While previous research has shown that the substrate influences the initial growth dynamics of Al2O3 films deposited by PE-ALD, the XPS spectra of films with thicknesses greater than the XPS sampling depth also showed very similar surface chemistry [33]. As seen in the survey spectra, the XPS high resolution spectra for Al2O3 films deposited onto 75 and 125 μm PET and PEN show almost identical surface chemistries. Such behaviour therefore indicates that the differences in the WVTRs shown in Fig. 10 can be attributed to the differences in the substrate thickness, type and pre-treatment rather than any film differences induced by the different substrates.

Depositing 24 nm Al2O3 barrier films onto polymeric substrates of different type, thickness and surface treatment has shown that all of these parameters affect WVTR. Prior to Al2O3 deposition, lower WVTRs resulted from thicker substrates for both PET and PEN while for the same substrate thicknesses, PEN had a lower WVTR than PET. XPS characterisation showed that O2 plasma pre-treatment increased the concentration of both CeO and C]O functionalities on the surface. After the deposition of 24 nm Al2O3 barrier films, lower WVTRs resulted for 125 μm than 75 μm substrates and for PEN over PET. O2 plasma pre-treatment prior to barrier film deposition also reduced the 50


K.L. Jarvis et al.

1.8

WVTR (g.m -2/day)

1.6


1 day 2 days

1.4

3 days

1.2

7 days

1 0.8 0.6 0.4 0.2

a.

60 50 40 30 20 10

0 75 µm PET

125 µm PET

75 µm PEN

125 µm PEN

0 O

Fig. 9. WVTR of 75 μm and 125 μm thick PET and PEN substrates.

75 µm PET untreated 125 µm PET untreated 75 µm PEN untreated 125 µm PEN untreated

7 No pre-treatment Plasma pre-treated


WVTR x 10-2 (g.m -2/day)

6 5 4 3 2 1 0 75 µm PET

125 µm PET

75 µm PEN

125 µm PEN

Al

C

75 µm PET plasma pre-treated 125 µm PET plasma pre-treated 75 µm PEN plasma pre-treated 125 µm PEN plamsa pre-treated

b.

60 50 40 30 20 10 0 O

Fig. 10. WVTR of 24 nm alumina ALD films deposited onto untreated and O2 plasma pretreated (300 W, 10 min) 75 μm and 125 μm thick PET and PEN substrates.

Al

C

Fig. 11. XPS survey atomic concentrations a. before Ar sputtering and b. after 20 s of Ar sputtering for untreated and O2 plasma pre-treated (300 W, 10 min) PET and PEN substrates coated with 24 nm alumina ALD films.

WVTRs which was attributed to surface cleaning and improved film adhesion. The lowest best barrier was O2 plasma pre-treated 125 μm PEN with 24 nm of Al2O3 which had a WVTR of 3.1 × 10− 3 g·m− 2/ day. These results show that the properties of the underlying polymer substrate do contribute to the WVTR of the entire structure. The choice of underlying polymeric substrate is often overlooked but careful selection can enable the further enhancement of the barrier properties.

providing the 75 μm PET substrate. References [1] H. Klumbies, P. Schmidt, M. Hähnel, A. Singh, U. Schroeder, C. Richter, T. Mikolajick, C. Hoßbach, M. Albert, J.W. Bartha, K. Leo, L. Müller-Meskamp, Thickness dependent barrier performance of permeation barriers made from atomic layer deposited alumina for organic devices, Org. Electron. 17 (2015) 138–143. [2] J. Lewis, Material challenge for flexible organic devices, Mater. Today 9 (2006) 38–45. [3] M. Vähä-Nissi, P. Sundberg, E. Kauppi, T. Hirvikorpi, J. Sievänen, A. Sood, M. Karppinen, A. Harlin, Barrier properties of Al2O3 and alucone coatings and nanolaminates on flexible biopolymer films, Thin Solid Films 520 (2012) 6780–6785. [4] J.-S. Park, H. Chae, H.K. Chung, S.I. Lee, Thin film encapsulation for flexible AMOLED: a review, Semicond. Sci. Technol. 26 (2011) 034001. [5] V.H. Martínez-Landeros, B. Gnade, M.A. Quevedo-López, R. Ramírez-Bon, Permeation studies on transparent multiple hybrid SiO2-PMMA coatings-Al2O3 barriers on PEN substrates, J. Sol-Gel Sci. Techn. 59 (2011) 345–351. [6] J. Meyer, H. Schmidt, W. Kowalsky, T. Riedl, A. Kahn, The origin of low water vapor transmission rates through Al2O3/ZrO2 nanolaminate gas-diffusion barriers grown by atomic layer deposition, Appl. Phys. Lett. 96 (2010) 243308-243303. [7] M.D. Groner, F.H. Fabreguette, J.W. Elam, S.M. George, Low-temperature Al2O3 atomic layer deposition, Chem. Mater. 16 (2004) 639–645. [8] D.-S. Han, D.-K. Choi, J.-W. Park, Al2O3/TiO2 multilayer thin films grown by plasma enhanced atomic layer deposition for organic light-emitting diode passivation, Thin

Acknowledgements The Cooperative Research Centre for Polymers (CRC-P) (Application No: 20110054) is gratefully acknowledged for their funding. This work was performed in part at the Melbourne Centre for Nanofabrication (MCN) and the Biointerface Engineering Hub @ Swinburne, both part of the Victorian Node of the Australian National Fabrication Facility (ANFF), a company established under the National Collaborative Research Infrastructure Strategy, to provide nano and microfabrication facilities for Australia's researchers. The authors would also like to thank Dr. Lachlan Hyde while at the Melbourne Centre for Nanofabrication for his assistance with ALD, Dr. Bill Gong at the University of New South Wales for carrying out the XPS analysis and the Centre for Organic Electronics at the University of Newcastle for 51


K.L. Jarvis et al.

Solid Films 552 (2014) 155–158. [9] T. Ahmadzada, D.R. McKenzie, N.L. James, Y. Yin, Q. Li, Atomic layer deposition of Al2O3 and Al2O3/TiO2 barrier coatings to reduce the water vapour permeability of polyetheretherketone, Thin Solid Films 591 (Part A) (2015) 131–136. [10] J.G. Lee, H.G. Kim, S.S. Kim, Defect-sealing of Al2O3/ZrO2 multilayer for barrier coating by plasma-enhanced atomic layer deposition process, Thin Solid Films 577 (2015) 143–148. [11] J. Meyer, P. Görrn, F. Bertram, S. Hamwi, T. Winkler, H.-H. Johannes, T. Weimann, P. Hinze, T. Riedl, W. Kowalsky, Al2O3/ZrO2 nanolaminates as ultrahigh gas-diffusion barriers—a strategy for reliable encapsulation of organic electronics, Adv. Mater. 21 (2009) 1845–1849. [12] H. Choi, S. Lee, H. Jung, S. Shin, G. Ham, H. Seo, H. Jeon, Moisture barrier properties of Al2O3 films deposited by remote plasma atomic layer deposition at low temperatures, Jpn. J. Appl. Phys. 52 (2013) 035502. [13] M.D. Groner, S.M. George, R.S. McLean, P.F. Carcia, Gas diffusion barriers on polymers using Al2O3 atomic layer deposition, Appl. Phys. Lett. 88 (2006) 051907051903. [14] H.G. Kim, S.S. Kim, Aluminum oxide barrier coating on polyethersulfone substrate by atomic layer deposition for barrier property enhancement, Thin Solid Films 520 (2011) 481–485. [15] P.F. Carcia, R.S. McLean, M.H. Reilly, Permeation measurements and modeling of highly defective Al2O3 thin films grown by atomic layer deposition on polymers, Appl. Phys. Lett. 97 (2010) 221901–221903. [16] S.H.K. Park, J. Oh, C.S. Hwang, Y.S. Yang, J.I. Lee, H.Y. Chu, Ultra thin film encapsulation of OLED on plastic substrate, J. Inf. Displays 5 (2004) 30–34. [17] S.-H. Park, J. Oh, C.-S. Hwang, J.-I. Lee, Y.S. Yang, H.Y. Chu, K.-Y. Kang, Ultra thin film encapsulation of OLED on plastic substrate, ETRI J. 27 (2005) 545–550. [18] S.-H.K. Park, J. Oh, C.-S. Hwang, J.-I. Lee, Y.S. Yang, H.Y. Chu, Ultrathin Film Encapsulation of an OLED by ALD, Electrochem. Solid-State Lett. 8 (2005) H21–H23. [19] P.F. Carcia, R.S. McLean, B.B. Sauer, M.H. Reilly, Atomic layer deposition ultrabarriers for electronic applications — strategies and implementation, J. Nanosci. Nanotechnol. 11 (2011) 7994–7998. [20] P.F. Carcia, R.S. McLean, M.H. Reilly, M.D. Groner, S.M. George, Ca test of Al2O3 gas diffusion barriers grown by atomic layer deposition on polymers, Appl. Phys. Lett. 89 (2006) 031915-031913. [21] T. Hirvikorpi, R. Laine, M. Vähä-Nissi, V. Kilpi, E. Salo, W.-M. Li, S. Lindfors, J. Vartiainen, E. Kenttä, J. Nikkola, A. Harlin, J. Kostamo, Barrier properties of plastic films coated with an Al2O3 layer by roll-to-toll atomic layer deposition, Thin Solid Films 550 (2014) 164–169. [22] L. Hoffmann, D. Theirich, T. Hasselmann, A. Räupke, D. Schlamm, T. Riedl, Gas permeation barriers deposited by atmospheric pressure plasma enhanced atomic layer deposition, J. Vac. Sci. Technol. A 34 (2016) 01A114. [23] S. Lee, H. Choi, S. Shin, J. Park, G. Ham, H. Jung, H. Jeon, Permeation barrier properties of an Al2O3/ZrO2 multilayer deposited by remote plasma atomic layer deposition, Curr. Appl. Phys. 14 (2014) 552–557. [24] M. Park, S. Oh, H. Kim, D. Jung, D. Choi, J.-S. Park, Gas diffusion barrier characteristics of Al2O3/alucone films formed using trimethylaluminum, water and ethylene glycol for organic light emitting diode encapsulation, Thin Solid Films 546 (2013) 153–156. [25] A.A. Dameron, S.D. Davidson, B.B. Burton, P.F. Carcia, R.S. McLean, S.M. George, Gas diffusion barriers on polymers using multilayers fabricated by Al2O3 and rapid SiO2 atomic layer deposition, J. Phys. Chem. C 112 (2008) 4573–4580. [26] R. Paetzold, A. Winnacker, D. Henseler, V. Cesari, K. Heuser, Permeation rate measurements by electrical analysis of calcium corrosion, Rev. Sci. Instrum. 74 (2003) 5147–5150. [27] S.-W. Seo, E. Jung, C. Lim, H. Chae, S.M. Cho, Water permeation through organic–inorganic multilayer thin films, Thin Solid Films 520 (2012) 6690–6694. [28] L.H. Kim, K. Kim, S. Park, Y.J. Jeong, H. Kim, D.S. Chung, S.H. Kim, C.E. Park,

[29]

[30]

[31]

[32]

[33]

[34]

[35] [36] [37] [38]

[39]

[40] [41]

[42]

[43] [44]

[45]

[46] [47]

[48]

52

Al2O3/TiO2 nanolaminate thin film encapsulation for organic thin film transistors via plasma-enhanced atomic layer deposition, ACS Appl. Mater. Interfaces 6 (2014) 6731–6738. P.S. Maydannik, T.O. Kääriäinen, K. Lahtinen, D.C. Cameron, M. Söderlund, P. Soininen, P. Johansson, J. Kuusipalo, L. Moro, X. Zeng, Roll-to-roll atomic layer deposition process for flexible electronics encapsulation applications, J. Vac. Sci. Technol. A 32 (2014) 051603. J. Fahlteich, M. Fahland, W. Schönberger, N. Schiller, Permeation barrier properties of thin oxide films on flexible polymer substrates, Thin Solid Films 517 (2009) 3075–3080. K. Lahtinen, P. Maydannik, P. Johansson, T. Kääriäinen, D.C. Cameron, J. Kuusipalo, Utilisation of continuous atomic layer deposition process for barrier enhancement of extrusion-coated paper, Surf. Coat. Technol. 205 (2011) 3916–3922. T. Hirvikorpi, M. Vähä-Nissi, A. Harlin, J. Marles, V. Miikkulainen, M. Karppinen, Effect of corona pre-treatment on the performance of gas barrier layers applied by atomic layer deposition onto polymer-coated paperboard, Appl. Surf. Sci. 257 (2010) 736–740. M. Vähä-Nissi, E. Kauppi, K. Sahagian, L.-S. Johansson, M.S. Peresin, J. Sievänen, A. Harlin, Growth of thin Al2O3 films on biaxially oriented polymer films by atomic layer deposition, Thin Solid Films 522 (2012) 50–57. E. Földes, A. Tóth, E. Kálmán, E. Fekete, Á. Tomasovszky-Bobák, Surface changes of corona-discharge-treated polyethylene films, J. Appl. Polym. Sci. 76 (2000) 1529–1541. L.-A. O'Hare, S. Leadley, B. Parbhoo, Surface physicochemistry of corona-dischargetreated polypropylene film, Surf. Interface Anal. 33 (2002) 335–342. L. Yang, J. Chen, Y. Guo, Z. Zhang, Surface modification of a biomedical polyethylene terephthalate (PET) by air plasma, Appl. Surf. Sci. 255 (2009) 4446–4451. A. Vesel, M. Mozetic, A. Zalar, XPS study of oxygen plasma activated PET, Vacuum 82 (2007) 248–251. I. Junkar, A. Vesel, U. Cvelbar, M. Mozetič, S. Strnad, Influence of oxygen and nitrogen plasma treatment on polyethylene terephthalate (PET) polymers, Vacuum 84 (2009) 83–85. K.N. Pandiyaraj, V. Selvarajan, R.R. Deshmukh, M. Bousmina, The effect of glow discharge plasma on the surface properties of poly(ethylene terephthalate) (PET) film, Surf. Coat. Technol. 202 (2008) 4218–4226. U. Moosheimer, C. Bichler, Plasma pretreatment of polymer films as a key issue for high barrier food packagings, Surf. Coat. Technol. 116–119 (1999) 812–819. E. Gonzalez, M.D. Barankin, P.C. Guschl, R.F. Hicks, Remote atmospheric-pressure plasma activation of the surfaces of polyethylene terephthalate and polyethylene naphthalate, Langmuir 24 (2008) 12636–12643. R. Cueff, G. Baud, M. Benmalek, J.P. Besse, J.R. Butruille, M. Jacquet, X-ray photoelectron spectroscopy studies of plasma-modified PET surface and alumina/PET interface, Appl. Surf. Sci. 115 (1997) 292–298. A. Doren, M.J. Genet, P.G. Rouxhet, Analysis of poly(ethylene terephthalate) (PET) by XPS, Surf. Sci. Spectra 3 (1994) 337–341. G. Garnier, S. Marouani, B. Yrieix, C. Pompeo, M. Chauvois, L. Flandin, Y. Brechet, Interest and durability of multilayers: from model films to complex films, Polym. Adv. Technol. 22 (2011) 847–856. J. Oh, S. Shin, J. Park, G. Ham, H. Jeon, Characteristics of Al2O3/ZrO2 laminated films deposited by ozone-based atomic layer deposition for organic device encapsulation, Thin Solid Films 599 (2016) 119–124. V. Baukh, H.P. Huinink, O.C.G. Adan, S.J.F. Erich, L.G.J. van der Ven, Predicting water transport in multilayer coatings, Polymer 53 (2012) 3304–3312. E. Dickey, W.A. Barrow, High rate roll to roll atomic layer deposition, and its application to moisture barriers on polymer films, J. Vac. Sci. Technol. A 30 (2012) 021502. D. Briggs, Surface Analysis of Polymers by XPS and Static SIMS, Cambridge University Press, Cambridge, 1998.