- Email: [email protected]
Barrier properties of plastic films coated with an Al2O3 layer by roll-to-toll atomic layer deposition
Thin Solid Films 550 (2014) 164–169
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Barrier properties of plastic films coated with an Al2O3 layer by roll-to-toll atomic layer deposition Terhi Hirvikorpi a,⁎, Risto Laine a, Mika Vähä-Nissi b, Väinö Kilpi a, Erkki Salo b, Wei-Min Li a, Sven Lindfors a, Jari Vartiainen b, Eija Kenttä b, Juha Nikkola c, Ali Harlin b, Juhana Kostamo a a b c
Picosun Oy, Tietotie 3, FI-02150 Espoo, Finland VTT Technical Research Centre of Finland, Biologinkuja 7, Espoo, P.O. Box 1000, FI-02044 VTT, Finland VTT Technical Research Centre of Finland, P.O. Box 1300, FI-33101 Tampere, Finland
a r t i c l e
i n f o
Article history: Received 26 June 2013 Received in revised form 18 September 2013 Accepted 24 October 2013 Available online 1 November 2013 Keywords: Atomic layer deposition Barrier Roll-to-roll Flexible material Aluminum oxide Polymer
a b s t r a c t Thin (30–40 nm) and highly uniform Al2O3 coatings have been deposited at relatively low temperature of 100 °C onto various polymeric materials employing the atomic layer deposition (ALD) technique, both batch and roll-toroll (R2R) mode. The applications for ALD have long been limited those feasible for batch processing. The work demonstrates that R2R ALD can deposit thin films with properties that are comparable to the film properties fabricated by in batch. This accelerates considerably the commercialization of many products, such as flexible, printed electronics, organic light-emitting diode lighting, third generation thin film photovoltaic devices, high energy density thin film batteries, smart textiles, organic sensors, organic/recyclable packaging materials, and flexible displays, to name a few. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Atomic layer deposited (ALD) Al2O3 has proven to be effective in enhancing the moisture and gas barrier properties of various plastic films and coatings [1–7]. The key challenge in several applications is to find a flexible, reliable, and cost efficient material to protect sensitive goods from ambient atmosphere. The demand for continuous ALD processing stems from the fields of flexible, printed electronics, organic lightemitting diode lighting, third generation thin film photovoltaic devices, high energy density thin film batteries, smart textiles, organic sensors, organic/recyclable packaging materials, and flexible displays, to name a few. “Barrier property” refers to a material's capability to resist the diffusion of a specific species (molecule, atom or ion) into and through the material. To be a good gas and vapor barrier, the material needs to be pore-free. When considering polymeric materials, the water vapor transmission rate (WVTR) is affected by e.g. the thickness of the polymer film as well as the temperature and humidity of the surroundings [8–10]. The ⁎ Corresponding author. Tel.: +358 40 770 6648; fax: +358 20 722 7012. E-mail addresses: [email protected] (T. Hirvikorpi), [email protected] (R. Laine), Mika.Vaha-Nissi@vtt.fi (M. Vähä-Nissi), [email protected] (V. Kilpi), Erkki.Salo@vtt.fi (E. Salo), [email protected] (W.-M. Li), [email protected] (S. Lindfors), Jari.Vartiainen@vtt.fi (J. Vartiainen), Eija.Kentta@vtt.fi (E. Kenttä), Juha.Nikkola@vtt.fi (J. Nikkola), Ali.Harlin@vtt.fi (A. Harlin), [email protected] (J. Kostamo). 0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tsf.2013.10.148
common polymers used in flexible electronics include e.g. polyethylene terephthalate and polyethylene naphtalene. Hygroscopic materials, such as many biopolymers, typically lose their barrier properties at high relative humidity due to water absorption [11]. The ALD technique is a surface-controlled layer-by-layer deposition process based on self-limiting gas–solid reactions. It is well suited to produce inorganic gas barrier coatings on various materials [12]. The most common ALD-grown gas and water vapor barrier material has been Al2O3 [1–3,13–17]. In these studies the Al2O3coatings have mainly been fabricated using the TMA-H2O process, but studies also show that O3 can also be used as the oxygen source when depositing on polymers [7,18]. The advantages of ALD-grown Al2O3 coating are superior moisture protection and relatively low deposition temperature. For the purposes of protecting electronic parts, water vapor transmission rates of the order of 1*10−3 g/m2/day and oxygen transmission rates below 5*10−3 cm3/m2/105 Pa/day have been reported for less than 25 nm thick Al2O3 coatings on plastic [2]. In addition, Park et al. [14] reported a water vapor transmission rate of 0.03 g/m2/day at 38 °C and 100% relative humidity for an ALD-grown Al2O3 barrier that was 30 nm thick and deposited on both sides of a poly(ethersulfone) substrate, whereas Carcia et al. [17] showed that 25-nm thick Al2O3 barrier coatings on poly(ethylene naphthalene) substrates can have a water vapor transmission rate of less than 1*10−5 g/m2/day. We demonstrate in this paper the deposition trials on plastic films with a roll-to-roll (R2R) process. This study demonstrates that a thin
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169
165
Al2O3 layer deposited with continuous ALD process enhances the barrier performance as much as does the Al2O3 layer fabricated with the batch ALD process. 2. Experimental details The substrates used were commercial plastics films presented in Table 1. From our previous thermogravimetric study performed for most of the present substrate materials [5] we may conclude that the materials do not degrade thermally at temperatures employed in our low-temperature ALD-Al2O3 process. Biaxially oriented polylactic acid (PLA) film had a three-layer structure: one thick core layer sandwiched between two thin skin layers. Polylactic acid was thermoplastic aliphatic polyester in this case supplied by NatureWorksTM. One side of the film was corona treated, which typically increases the number of oxygen groups and surface energy. Cellophane film (CEL) was a non-commercial uncoated film. Polyimide film (PI) was an aromatic all-polyimide film with a wide processing temperature window. Cellophane is antistatic and dimensionally stable. PLA has glass transition temperature between 60 and 65 °C, while for cellophane this is typically higher and undefined. A second order transition occurs in PI between 360 and 410 °C assumed to be the glass transition temperature. PLA will likely have the highest strain at elevated temperatures. Substrates were obtained as small reels. Top layers were simply disregarded before sampling the reels in order to avoid possible dirt particles, finger prints and other contaminants. No other cleaning procedure was carried out before deposition. Silicon wafers, cellophane, polylactide, and polyimide film substrates were deposited onto with Al2O3 at 100 °C using both a batch PICOSUN™ reactor chamber and a PICOSUN™ R2R chamber. The chambers were used in the same PICOSUN™ R-200 ALD reactor. The precursors were trimethyl aluminum (TMA, electronic grade purity, SAFC Hitech) and H2O. High purity nitrogen (99.9999% N2) was used as a carrier and purge gas. The operating pressure was 5–10 hPa (mbar) in both reactor chambers. The precursor pulsing sequence for batch process was: 0.1 s TMA pulse, 6 s N2 purge, 0.1 s H2O pulse, and 6 s N2 purge. For the R2R ALD process the precursor pulsing sequence was: 0.1 s TMA pulse, 4 s N2 purge, 0.1 s H2O pulse, and 4 s N2 purge. The number of ALD cycles used for both batch and R2R process run was 500. The R2R add-on was specially designed to fit to the basic ALD reactor for achieving fully comparable results than with the batch process (Fig. 1). Fig. 2 presents the configuration of the R2R mode. The substrate, unwinding and winding units and an R2R coating channel are inside a batch chamber. The precursors are pulsed into the coating channel similarly to a batch process, while the space with the winding units is purged with nitrogen. The substrate is moved in and out of the coating channel through narrow slits. The narrow slits keep the precursors in the coating channel, and purging gases are fed into a vacuum line. Only the reaction chamber, reaction chamber lid and vacuum chamber lid of the basic ALD reactor were changed. The substrates coated in batch chamber were ca. 100 × 100 mm2 in size, while the substrates coated by R2R ALD were 70 mm wide and continuously moving at speed of 200 mm/h, from the unwinding roll to the winding roll. The servo motors which rotated the unwinding and winding rolls were located outside of the vacuum chamber. The sizes of the reaction chamber and vacuum chamber lids were the same in both batch and R2R
Fig. 1. R2R setup assembled to a batch ALD reactor (A) and the R2R mechanism (B).
processes; however, the R2R chamber lid was specially designed for the R2R mechanics. The basic controlling software was the same for both configurations, except the additional control panel for servo motors in the R2R software. Coatings were grown on one side in batch processing and on both sides of the substrates in the R2R process. The Al2O3 deposited samples were characterized for their barrier and surface characteristics. The thickness of the fabricated Al2O3 coatings was estimated by growing the Al2O3 films on silicon wafer and measured by single wave-length ellipsometer. Because of the different surface chemistries of different polymers, the actual thickness may somewhat deviate from that
Table 1 The commercial plastic materials employed as substrates. Code Description PLA PI CEL
Biaxially oriented polylactide film (EVLON (R) EV) from BI-AX International Inc., Canada, one-side treated, thickness 40 μm Kapton® HN polyimide film from DuPont, thickness 127 μm Cellophane film from Innovia Films, thickness 27 μm
Fig. 2. The R2R mechanism in the batch ALD chamber.
166
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169
determined for the Al2O3-coated silicon wafer [13,19]. It should also be mentioned that even though the aim in batch mode was to deposit only on one side of the substrate, unintentional growth could not be totally prevented on the substrate back-side, away from the primary precursor flux direction. The surface topography of uncoated and coated plastic samples was characterised using non-contact mode atomic force microscopy (NC-AFM). The NC-AFM analysis was performed using Park Systems XE-100 AFM equipment with cantilever 905 M-ACTA (purchased from ST Instruments B.V.). Typically, the scan rate was 0.4–0.6 Hz and the measured area was 2.5 × 2.5 μm2. The chemical composition of the Al2O3 films on silicon wafers was measured by Fourier transform infrared spectrometer (FTIR). The Al2O3 coatings characterized with IR were grown by the R2R process on silicon wafer, and the reference sample of thick Al2O3 film (84 nm) was grown by the batch process on the silicon wafer. The surface IR spectra of the Al2O3 films were measured with 60 degree germanium ATR (attenuated total reflectance) crystal of VariGATR™ accessory attached to the Nicolet iS50 FTIR spectrometer. Spectra were collected by averaging 128 scans at a resolution of 4 cm−1. ATR and baseline corrections were performed using Omnic software before spectrum interpretation. Three IR spectra were measured from each sample surface and an average spectrum was calculated of the parallel measurements. The surface spectra of three Al2O3 films with thicknesses of 32 nm, 20 nm and16 nm, grown by the R2R process, were then compared to the reference sample spectrum. Contact angle (CA) and surface energy (SE) measurements (KSV CAM 200 Optical Contact Angle Meter) were carried out for some of the samples in a controlled atmosphere (relative humidity 50%, temperature 23 °C) with three to eight parallel measurements and expressed as degrees (°). The CA value was determined using water as the test liquid. For the SE measurements, water and di-iodomethane were used as the test liquids. The CA values were calculated at the time of 1 s from the moment the drop contacts. The SE values were calculated from the CA data by using the OWRK method and expressed as mN/m. The OWRK (Owens, Wendt, Rabel and Kaelble) method utilizes a geometric mean equation for determination of the total surface energy and its polar and dispersive components of the solid surfaces [20]. These components of the surface energy are determined by contact angle measurements of two (or more) liquids with known surface tension components. For all the samples, the oxygen and water vapor transmission rate (OTR and WVTR, respectively) values were determined. The OTR values expressed as cm3/m2/105 Pa/day were measured (Systech M8001) from two parallel samples using humid gases at 23 °C and in 50% relative humidity. The size of samples was either 50 cm2 or with masking 5 cm2. 100% oxygen was used as a test gas. The WVTR values were measured from two parallel samples according to the modified gravimetric methods ISO 2528:1995 and SCAN P 22:68 and were expressed as g/m2/day at 23 °C and with a moisture gradient of 50% relative humidity.
the results, the thickness of the films fabricated with the R2R add-on on pieces of silicon wafers is similar to the results obtained with the batch process. For example, when the R2R add-on was used to fabricate Al2O3 coating with thickness of 25 nm on the silicon wafer, the average thickness measured on several places of the wafer was 25.26 nm, as having
3. Results and discussion The R2R add-on parts were attached to the process chamber of a standard batch ALD reactor. The R2R add-on makes the desired coating only on the surface of the substrate and on the coating channel walls. This maintains the rolls and mechanisms uncoated, which shortens the maintenance cycle considerably. During the R2R deposition, the web is moved from one roll to another roll through a coating channel. Due to the produced pressure difference and a very narrow slit at the both ends of the coating channel, the channel will keep the precursor chemicals in the channel. This design enables the fabrication of highquality thin films even on flexible substrates. For the detection of the growth per cycle (GPC) value of R2R process, pieces of silicon wafer were taped on steel foil web and the thickness of the fabricated Al2O3 film was measured with ellipsometer. According to
Fig. 3. The AFM images of uncoated and Al2O3-coated polylactide (PLA) films. The Al2O3 coatings have been fabricated by using both batch and R2R process methods.
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169
167
all the measuring points between 24.97 nm and 25.53 nm. The accuracy of the ellipsometer was ±1 nm. AFM images of uncoated and ALD coated samples revealed some differences. The ALD coating seemed to grow as particle-like structure on the polymer substrates, which is in agreement with previous studies [21–24]. The formed particles were clearly attached on the surface. The surfaces of polylactide and polyimide substrates with and without
ALD-coating are presented in Figs. 3 and 4, respectively. The R2R and batch processes formed similar structure on polylactide and cellophane substrates. However, for cellophane film, the Al2O3 coating seemed to form larger particles than observed on PI and PLA substrates. The cellophane film substrate with and without ALD-coating is presented in Fig. 5. These differences may be due to the initial surface roughness of
Fig. 4. The AFM images of uncoated and Al2O3-coated polyimide (PI) films. The Al2O3 coatings have been fabricated by using both batch and R2R process methods.
Fig. 5. The AFM images of uncoated and Al2O3-coated cellophane (CEL) films. The Al2O3 coatings have been fabricated by using both batch and R2R process methods.
168
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169 Table 3 The OTR and WVTR values for plain and Al2O3-coated plastic samples. The Al2O3-coating has fabricated by using both, batch and R2R ALD processes. Substrate
CEL PLA PI a
Fig. 6. The FTIR spectra of four Al2O3 coatings with different thicknesses. The reference sample (84 nm) was deposited with the batch ALD reactor. The spectrum of the reference sample was compared to the spectra of the three samples (16 nm, 20 nm and 32 nm) deposited by using the R2R add-on.
the uncoated substrate or to the differences in the surface chemistry between the substrates. For the FTIR analysis, four Al2O3 coatings with different thicknesses were grown by the R2R process on the silicon wafer. The reference sample thick film of Al2O3 (84 nm) was deposited with the batch ALD reactor on the silicon wafer. The spectrum of the reference sample was compared to the spectra of the three samples (16 nm, 20 nm and 32 nm) deposited on the silicon wafer by using the R2R add-on. FTIR analyses indicated Al2O3 covered surfaces after the R2R ALD. The measured surface IR spectra at region 4000 – 600 cm−1 are shown in Fig. 6. In the sample of 16 nm film the main IR band of bulk Al2O3 at the region of 1000–850 cm−1 is rather weak. The growth of aluminum oxide film is clearly seen in the IR spectra of 20 nm or 32 nm films, where the bulk absorptions of Al2O3 at 950–700 cm−1 are stronger, as the film thickness is increasing [13,25]. The band position of the strongest IR band of aluminum oxide film is transferring to lower wavenumber direction towards the band position of reference sample (883 cm−1), with 84 nm Al2O3, as aluminum oxide film is increased from 16 nm to 32 nm. This shift indicates that when the film thickness increases, the layer structure is more similar to the reference thick film aluminum oxide. Contact angle and surface energy measurements revealed some differences on the chemical composition between batch and R2R deposition methods. The results are presented in Table 2. Al2O3 thin films fabricated with the R2R on the polymer films had lower polar components and overall surface energy than the thin films obtained with the batch process. Without further analysis it can be speculated that this could be due to relatively low deposition temperature and thus an increased amount of non-polar impurities on the Al2O3 surface, i.e. hydrocarbon compounds not removed by the nitrogen purge in the roll-toroll process. The FTIR analysis of Al2O3 deposited onto silicon wafers also indicated the presence of such impurities, as weak bands of C–H
Table 2 Contact angle (CA), surface energy (SE) values for plain and Al2O3-coated polymers. The total value of surface energy (SE) is the sum of dispersive (SEd) and polar (SEp) components. Sample
CA (°)
CEL CEL + Al2O3 batch CEL + Al2O3 R2R PLA PLA + Al2O3 batch PLA + Al2O3 R2R PI PI + Al2O3 batch PI + Al2O3 R2R
40 60 74 71 48 78 70 52 102
± ± ± ± ± ± ± ± ±
1 4 3 1 3 3 2 2 2
SEp (mN/m)
SEd (mN/m)
SE (mN/m)
22.9 13.2 7.0 8.2 22.8 6.6 5.5 19.9 0.3
36.4 34.0 31.8 32.6 30.9 29.0 42.9 30.6 26.2
59.3 47.2 38.8 40.8 53.6 35.6 48.4 50.5 26.5
OTR
WVTR
Polymer
+ Al2O3 Batch
+ Al2O3 R2R
Polymer
+ Al2O3 Batch
+ Al2O3 R2R
4.0 ± 0.1 470 ± 1 30 ± 0.1
2.6 ± 2.3 0.4 ± 0.1 b0.01 a
0.3 ± 0.2 47 ± 6 b 2.2 ± 1.9 b
144 ± 19 39 ± 5 3±1
29 ± 16 b0.2 a b0.1 a
15 ± 9 10 ± 1 b 2±1b
Result was under the detection limit; b Small sample size.
stretching vibrations are seen in the region of 2920 – 2960 cm−1 of 20 and 32 nm samples, see Fig. 6. The Al2O3 surfaces as such are intrinsically hydrophilic in nature [26]. The transitions in the contact angles after initial ALD cycles can be explained by changes in surface roughness, nucleation effects and chemical bonding [22,27,28]. A hydrophobic effect has been observed on cellulose after the first TMA deposition cycles explained by the stable Al–(O–C–)3 groups with minimal hydrophilic Al–OH groups and carbon adsorption after deposition [29]. Retention of by-products into the plastic films and thus a lower polarity of the Al2O3 layer could be due to short purge/water pulses and both side deposition. In addition, winding the reel could impair purging and prevent desorption of by-products. Low deposition temperature (b100 °C) usually leads to a certain amount of impurities. The oxygen and water vapor barrier results achieved for the polymer substrates with and without of the Al2O3 coating deposited by batch and R2R ALD processes are summarized in Table 3. Al2O3 coating was deposited on one side in batch processing and on both sides of the substrates in the R2R process. Independently of the deposition process used, the Al2O3 coating improved both the oxygen and water vapor barrier properties of the polymeric samples. Thermal stability and viscoelastic response of polymer substrates to mechanical stresses during deposition can have an effect on the barrier properties after R2R deposition. Similarly, defects created during winding, such as cracking, are challenges for the industrial use of R2R ALD. The probability of such damages seems to depend on the polymer properties. For example, CEL is antistatic attracting less charged particulate contaminants. Particulate contaminants are a known cause of defects in e.g. vacuum metallizing. It has also been shown that the surface chemistry has an impact. 25 nm of Al2O3 had better oxygen barrier on cellophane than on PLA, and straining the film had a less negative effect on oxygen barrier in the case of cellophane [30]. In addition, abrasion testing of PLA and cellophane films with 25 μm of Al2O3 has indicated that oxygen barrier is impaired less by abrasion with cellophane than with PLA (0–60 vs. 90–130 times higher, respectively, after abrasion depending on the test method). Cellophane has more reactive hydroxyl surface groups, while PLA has also carbonyl groups. Polyimide is more hydrophobic with carbonyl groups. Differences in surface and diffusion properties can lead to different thin film growth, structure and properties [2,31]. Similarly, an Al2O3 coating on CEL was in this study less sensitive to various defects, and the barrier values obtained with an Al2O3 coating on both sides were still better than those obtained with the same substrate coated only on one side. With PLA and PI the barrier results were impaired to a level exceeding the barrier values of one side coated substrates. In the latter two cases, the size of the samples fabricated by R2R was also smaller than the samples fabricated by the batch process. Aluminum masks had to be used with the small samples. This can lead to poorer detection limit and larger edge effect and thus affect OTR and WVTR values. 4. Conclusions The positive effect of the Al2O3 coating on the barrier properties of plastic films was evident with both processes. Thermal stability, viscoelastic properties at elevated temperatures and sensitivity to various
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169
defects during winding are parameters likely affecting the barrier properties of plastic films after the roll-to-roll (R2R) ALD. There were differences between the films also probably due to different surface chemistry. For example, the Al2O3 coatings deposited with the R2R process on both sides of the cellophane film were less sensitive to defects than coated polylactide and polyimide films, and the barrier values were impaired to a level still better than that obtained with cellophane one side coated in the batch process. However, with polylactide and polyimide films the defects during the R2R process led to barrier properties inferior to those of one side coated films. The related phenomena need to be investigated more in the future. The FTIR analyses detected Al2O3 covered surfaces after the R2R ALD. AFM images for the batch and R2R produced samples were also quite similar. There were slight indications of larger clusters with cellophane and increased roughness in the μm scale with polylactide film after the R2R ALD. The relative polarities of surface energy for Al2O3 deposited with R2R ALD on all three coatings were lower than for the batch samples. This indicates also some differences in the thin film growth between the batch and R2R processes. References [1] P.F. Carcia, R.S. McLean, M.H. Reilly, M.D. Groner, S.M. George, Appl. Phys. Lett. 89 (2006) 031915. [2] M.D. Groner, S.M. George, R.S. McLean, P.F. Carcia, Appl. Phys. Lett. 88 (2006) 051907. [3] A.A. Cameron, S.D. Davidson, B.B. Burton, P.F. Carcia, R.S. McLean, S.M. George, J. Phys. Chem. C 112 (2008) 4573. [4] T. Hirvikorpi, M. Vähä-Nissi, A. Harlin, M. Karppinen, Thin Solid Films 518 (2010) 5463. [5] T. Hirvikorpi, M. Vähä-Nissi, T. Mustonen, E. Iiskola, M. Karppinen, Thin Solid Films 518 (2010) 2654. [6] T. Hirvikorpi, M. Vähä-Nissi, J. Nikkola, A. Harlin, M. Karppinen, Surf. Coat. Technol. 205 (2011) 5088.
169
[7] T. Hirvikorpi, Thin Al2O3 Barrier Coatings Grown on Bio-based Packaging Materials by Atomic Layer Deposition, VTT, 770, VTT Publications, 2011. [8] C. Andersson, Packag. Technol. Sci. 21 (2008) 339. [9] J. Kuusipalo, K. Lahtinen, TAPPI J. 7 (2008) 8. [10] A. Savolainen, J. Kuusipalo, E. Laiho, T. Penttinen, in: A. Savolainen (Ed.), Paper and Paperboard Converting, Fapet Oy, Helsinki, 1998, p. 123. [11] Y. Leterrier, Prog. Mater. Sci. 48 (2003) 1. [12] R. Puurunen, J. Appl. Phys. 97 (2005) 121301. [13] J.D. Ferguson, A.W. Weimer, S.M. George, Chem. Mater. 16 (2004) 5602. [14] S.H.K. Park, J. Oh, C.S. Hwang, J.I. Lee, Y.S. Yang, H.Y. Chu, Electrochem. Solid-State Lett. 8 (2005) H21. [15] R. Cooper, H.P. Upadhyaya, T.K. Minton, M.R. Berman, X. Du, S.M. George, Thin Solid Films 516 (2008) 4036. [16] Y. Zhang, J.A. Bertrand, R. Yang, S.M. George, Y.C. Lee, Thin Solid Films 517 (2009) 3269. [17] P.F. Carcia, R.S. McLean, S. Hegedus, Sol. Energy Mater. Sol. Cells 94 (2010) 2375. [18] T.O. Kääriäinen, P. Maydannik, D.C. Cameron, K. Lahtinen, P. Johansson, J. Kuusipalo, Thin Solid Films 519 (2011) 3146. [19] X. Liang, L.F. Hakim, G.D. Zhan, J.A. McCormick, S.M. George, A.W. Weimer, J.A. Spencer, K.J. Buechler, J. Blackson, C.J. Wood, J.R. Dorgan, J. Am. Ceram. Soc. 90 (2007) 57. [20] D.K. Owens, R.C. Wendt, J. Appl. Polym. Sci. 13 (1969) 1741. [21] C.A. Wilson, R.K. Grubbs, S.M. George, Chem. Mater. 17 (2005) 5625. [22] M. Vähä-Nissi, E. Kauppi, K. Sahagian, L.-S. Johansson, M.S. Peresin, J. Sievänen, A. Harlin, Thin Solid Films 522 (2012) 50. [23] K. Grigoras, S. Franssila, V.M. Airaksinen, Thin Solid Films 516 (2008) 5551. [24] R.L. Puurunen, W. Vandervorst, W.F.A. Besling, O. Richard, H. Bender, T. Conard, C. Zhao, A. Delabie, M. Caymax, S. de Gendt, M. Heyns, M.M. Viitanen, M. de Ridder, H.H. Brongersma, Y. Tamminga, T. Dao, T. de Win, M. Verheijen, M. Kaiser, M. Tuominen, J. Appl. Phys. 96 (2004) 4878. [25] D.N. Goldstein, J.A. McCormick, S.M. George, J. Phys. Chem. C 112 (2008) 19530. [26] D.S. Finch, T. Oreskovic, K. Ramadurai, C.F. Herrmann, S.M. George, R.L. Mahajan, J. Biomed. Mater. Res. A 87 (A) (2008) 100. [27] R.L. Puurunen, Appl. Surf. Sci. 245 (2005) 6. [28] G.K. Hyde, G. Scarel, J.C. Spagnola, Q. Peng, K. Lee, B. Gong, K.G. Roberts, K.M. Roth, C.A. Hanson, C.K. Devine, S.M. Stewart, D. Hojo, J.-S. Na, J.S. Jur, G.N. Parsons, Langmuir 26 (2010) 2550. [29] K. Lee, J.S. Jur, D.H. Kim, G.N. Parsons, J. Vac. Sci. Technol. A 30 (2012) 01A163. [30] M. Vähä-Nissi, P. Sundberg, E. Kauppi, T. Hirvikorpi, J. Sievänen, A. Sood, M. Karppinen, A. Harlin, Thin Solid Films 520 (2012) 6780. [31] B. Gong, G.N. Parsons, J. Mater. Chem. 22 (2012) 15672.
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Barrier properties of plastic films coated with an Al2O3 layer by roll-to-toll atomic layer deposition Terhi Hirvikorpi a,⁎, Risto Laine a, Mika Vähä-Nissi b, Väinö Kilpi a, Erkki Salo b, Wei-Min Li a, Sven Lindfors a, Jari Vartiainen b, Eija Kenttä b, Juha Nikkola c, Ali Harlin b, Juhana Kostamo a a b c
Picosun Oy, Tietotie 3, FI-02150 Espoo, Finland VTT Technical Research Centre of Finland, Biologinkuja 7, Espoo, P.O. Box 1000, FI-02044 VTT, Finland VTT Technical Research Centre of Finland, P.O. Box 1300, FI-33101 Tampere, Finland
a r t i c l e
i n f o
Article history: Received 26 June 2013 Received in revised form 18 September 2013 Accepted 24 October 2013 Available online 1 November 2013 Keywords: Atomic layer deposition Barrier Roll-to-roll Flexible material Aluminum oxide Polymer
a b s t r a c t Thin (30–40 nm) and highly uniform Al2O3 coatings have been deposited at relatively low temperature of 100 °C onto various polymeric materials employing the atomic layer deposition (ALD) technique, both batch and roll-toroll (R2R) mode. The applications for ALD have long been limited those feasible for batch processing. The work demonstrates that R2R ALD can deposit thin films with properties that are comparable to the film properties fabricated by in batch. This accelerates considerably the commercialization of many products, such as flexible, printed electronics, organic light-emitting diode lighting, third generation thin film photovoltaic devices, high energy density thin film batteries, smart textiles, organic sensors, organic/recyclable packaging materials, and flexible displays, to name a few. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Atomic layer deposited (ALD) Al2O3 has proven to be effective in enhancing the moisture and gas barrier properties of various plastic films and coatings [1–7]. The key challenge in several applications is to find a flexible, reliable, and cost efficient material to protect sensitive goods from ambient atmosphere. The demand for continuous ALD processing stems from the fields of flexible, printed electronics, organic lightemitting diode lighting, third generation thin film photovoltaic devices, high energy density thin film batteries, smart textiles, organic sensors, organic/recyclable packaging materials, and flexible displays, to name a few. “Barrier property” refers to a material's capability to resist the diffusion of a specific species (molecule, atom or ion) into and through the material. To be a good gas and vapor barrier, the material needs to be pore-free. When considering polymeric materials, the water vapor transmission rate (WVTR) is affected by e.g. the thickness of the polymer film as well as the temperature and humidity of the surroundings [8–10]. The ⁎ Corresponding author. Tel.: +358 40 770 6648; fax: +358 20 722 7012. E-mail addresses: [email protected] (T. Hirvikorpi), [email protected] (R. Laine), Mika.Vaha-Nissi@vtt.fi (M. Vähä-Nissi), [email protected] (V. Kilpi), Erkki.Salo@vtt.fi (E. Salo), [email protected] (W.-M. Li), [email protected] (S. Lindfors), Jari.Vartiainen@vtt.fi (J. Vartiainen), Eija.Kentta@vtt.fi (E. Kenttä), Juha.Nikkola@vtt.fi (J. Nikkola), Ali.Harlin@vtt.fi (A. Harlin), [email protected] (J. Kostamo). 0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tsf.2013.10.148
common polymers used in flexible electronics include e.g. polyethylene terephthalate and polyethylene naphtalene. Hygroscopic materials, such as many biopolymers, typically lose their barrier properties at high relative humidity due to water absorption [11]. The ALD technique is a surface-controlled layer-by-layer deposition process based on self-limiting gas–solid reactions. It is well suited to produce inorganic gas barrier coatings on various materials [12]. The most common ALD-grown gas and water vapor barrier material has been Al2O3 [1–3,13–17]. In these studies the Al2O3coatings have mainly been fabricated using the TMA-H2O process, but studies also show that O3 can also be used as the oxygen source when depositing on polymers [7,18]. The advantages of ALD-grown Al2O3 coating are superior moisture protection and relatively low deposition temperature. For the purposes of protecting electronic parts, water vapor transmission rates of the order of 1*10−3 g/m2/day and oxygen transmission rates below 5*10−3 cm3/m2/105 Pa/day have been reported for less than 25 nm thick Al2O3 coatings on plastic [2]. In addition, Park et al. [14] reported a water vapor transmission rate of 0.03 g/m2/day at 38 °C and 100% relative humidity for an ALD-grown Al2O3 barrier that was 30 nm thick and deposited on both sides of a poly(ethersulfone) substrate, whereas Carcia et al. [17] showed that 25-nm thick Al2O3 barrier coatings on poly(ethylene naphthalene) substrates can have a water vapor transmission rate of less than 1*10−5 g/m2/day. We demonstrate in this paper the deposition trials on plastic films with a roll-to-roll (R2R) process. This study demonstrates that a thin
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169
165
Al2O3 layer deposited with continuous ALD process enhances the barrier performance as much as does the Al2O3 layer fabricated with the batch ALD process. 2. Experimental details The substrates used were commercial plastics films presented in Table 1. From our previous thermogravimetric study performed for most of the present substrate materials [5] we may conclude that the materials do not degrade thermally at temperatures employed in our low-temperature ALD-Al2O3 process. Biaxially oriented polylactic acid (PLA) film had a three-layer structure: one thick core layer sandwiched between two thin skin layers. Polylactic acid was thermoplastic aliphatic polyester in this case supplied by NatureWorksTM. One side of the film was corona treated, which typically increases the number of oxygen groups and surface energy. Cellophane film (CEL) was a non-commercial uncoated film. Polyimide film (PI) was an aromatic all-polyimide film with a wide processing temperature window. Cellophane is antistatic and dimensionally stable. PLA has glass transition temperature between 60 and 65 °C, while for cellophane this is typically higher and undefined. A second order transition occurs in PI between 360 and 410 °C assumed to be the glass transition temperature. PLA will likely have the highest strain at elevated temperatures. Substrates were obtained as small reels. Top layers were simply disregarded before sampling the reels in order to avoid possible dirt particles, finger prints and other contaminants. No other cleaning procedure was carried out before deposition. Silicon wafers, cellophane, polylactide, and polyimide film substrates were deposited onto with Al2O3 at 100 °C using both a batch PICOSUN™ reactor chamber and a PICOSUN™ R2R chamber. The chambers were used in the same PICOSUN™ R-200 ALD reactor. The precursors were trimethyl aluminum (TMA, electronic grade purity, SAFC Hitech) and H2O. High purity nitrogen (99.9999% N2) was used as a carrier and purge gas. The operating pressure was 5–10 hPa (mbar) in both reactor chambers. The precursor pulsing sequence for batch process was: 0.1 s TMA pulse, 6 s N2 purge, 0.1 s H2O pulse, and 6 s N2 purge. For the R2R ALD process the precursor pulsing sequence was: 0.1 s TMA pulse, 4 s N2 purge, 0.1 s H2O pulse, and 4 s N2 purge. The number of ALD cycles used for both batch and R2R process run was 500. The R2R add-on was specially designed to fit to the basic ALD reactor for achieving fully comparable results than with the batch process (Fig. 1). Fig. 2 presents the configuration of the R2R mode. The substrate, unwinding and winding units and an R2R coating channel are inside a batch chamber. The precursors are pulsed into the coating channel similarly to a batch process, while the space with the winding units is purged with nitrogen. The substrate is moved in and out of the coating channel through narrow slits. The narrow slits keep the precursors in the coating channel, and purging gases are fed into a vacuum line. Only the reaction chamber, reaction chamber lid and vacuum chamber lid of the basic ALD reactor were changed. The substrates coated in batch chamber were ca. 100 × 100 mm2 in size, while the substrates coated by R2R ALD were 70 mm wide and continuously moving at speed of 200 mm/h, from the unwinding roll to the winding roll. The servo motors which rotated the unwinding and winding rolls were located outside of the vacuum chamber. The sizes of the reaction chamber and vacuum chamber lids were the same in both batch and R2R
Fig. 1. R2R setup assembled to a batch ALD reactor (A) and the R2R mechanism (B).
processes; however, the R2R chamber lid was specially designed for the R2R mechanics. The basic controlling software was the same for both configurations, except the additional control panel for servo motors in the R2R software. Coatings were grown on one side in batch processing and on both sides of the substrates in the R2R process. The Al2O3 deposited samples were characterized for their barrier and surface characteristics. The thickness of the fabricated Al2O3 coatings was estimated by growing the Al2O3 films on silicon wafer and measured by single wave-length ellipsometer. Because of the different surface chemistries of different polymers, the actual thickness may somewhat deviate from that
Table 1 The commercial plastic materials employed as substrates. Code Description PLA PI CEL
Biaxially oriented polylactide film (EVLON (R) EV) from BI-AX International Inc., Canada, one-side treated, thickness 40 μm Kapton® HN polyimide film from DuPont, thickness 127 μm Cellophane film from Innovia Films, thickness 27 μm
Fig. 2. The R2R mechanism in the batch ALD chamber.
166
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169
determined for the Al2O3-coated silicon wafer [13,19]. It should also be mentioned that even though the aim in batch mode was to deposit only on one side of the substrate, unintentional growth could not be totally prevented on the substrate back-side, away from the primary precursor flux direction. The surface topography of uncoated and coated plastic samples was characterised using non-contact mode atomic force microscopy (NC-AFM). The NC-AFM analysis was performed using Park Systems XE-100 AFM equipment with cantilever 905 M-ACTA (purchased from ST Instruments B.V.). Typically, the scan rate was 0.4–0.6 Hz and the measured area was 2.5 × 2.5 μm2. The chemical composition of the Al2O3 films on silicon wafers was measured by Fourier transform infrared spectrometer (FTIR). The Al2O3 coatings characterized with IR were grown by the R2R process on silicon wafer, and the reference sample of thick Al2O3 film (84 nm) was grown by the batch process on the silicon wafer. The surface IR spectra of the Al2O3 films were measured with 60 degree germanium ATR (attenuated total reflectance) crystal of VariGATR™ accessory attached to the Nicolet iS50 FTIR spectrometer. Spectra were collected by averaging 128 scans at a resolution of 4 cm−1. ATR and baseline corrections were performed using Omnic software before spectrum interpretation. Three IR spectra were measured from each sample surface and an average spectrum was calculated of the parallel measurements. The surface spectra of three Al2O3 films with thicknesses of 32 nm, 20 nm and16 nm, grown by the R2R process, were then compared to the reference sample spectrum. Contact angle (CA) and surface energy (SE) measurements (KSV CAM 200 Optical Contact Angle Meter) were carried out for some of the samples in a controlled atmosphere (relative humidity 50%, temperature 23 °C) with three to eight parallel measurements and expressed as degrees (°). The CA value was determined using water as the test liquid. For the SE measurements, water and di-iodomethane were used as the test liquids. The CA values were calculated at the time of 1 s from the moment the drop contacts. The SE values were calculated from the CA data by using the OWRK method and expressed as mN/m. The OWRK (Owens, Wendt, Rabel and Kaelble) method utilizes a geometric mean equation for determination of the total surface energy and its polar and dispersive components of the solid surfaces [20]. These components of the surface energy are determined by contact angle measurements of two (or more) liquids with known surface tension components. For all the samples, the oxygen and water vapor transmission rate (OTR and WVTR, respectively) values were determined. The OTR values expressed as cm3/m2/105 Pa/day were measured (Systech M8001) from two parallel samples using humid gases at 23 °C and in 50% relative humidity. The size of samples was either 50 cm2 or with masking 5 cm2. 100% oxygen was used as a test gas. The WVTR values were measured from two parallel samples according to the modified gravimetric methods ISO 2528:1995 and SCAN P 22:68 and were expressed as g/m2/day at 23 °C and with a moisture gradient of 50% relative humidity.
the results, the thickness of the films fabricated with the R2R add-on on pieces of silicon wafers is similar to the results obtained with the batch process. For example, when the R2R add-on was used to fabricate Al2O3 coating with thickness of 25 nm on the silicon wafer, the average thickness measured on several places of the wafer was 25.26 nm, as having
3. Results and discussion The R2R add-on parts were attached to the process chamber of a standard batch ALD reactor. The R2R add-on makes the desired coating only on the surface of the substrate and on the coating channel walls. This maintains the rolls and mechanisms uncoated, which shortens the maintenance cycle considerably. During the R2R deposition, the web is moved from one roll to another roll through a coating channel. Due to the produced pressure difference and a very narrow slit at the both ends of the coating channel, the channel will keep the precursor chemicals in the channel. This design enables the fabrication of highquality thin films even on flexible substrates. For the detection of the growth per cycle (GPC) value of R2R process, pieces of silicon wafer were taped on steel foil web and the thickness of the fabricated Al2O3 film was measured with ellipsometer. According to
Fig. 3. The AFM images of uncoated and Al2O3-coated polylactide (PLA) films. The Al2O3 coatings have been fabricated by using both batch and R2R process methods.
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169
167
all the measuring points between 24.97 nm and 25.53 nm. The accuracy of the ellipsometer was ±1 nm. AFM images of uncoated and ALD coated samples revealed some differences. The ALD coating seemed to grow as particle-like structure on the polymer substrates, which is in agreement with previous studies [21–24]. The formed particles were clearly attached on the surface. The surfaces of polylactide and polyimide substrates with and without
ALD-coating are presented in Figs. 3 and 4, respectively. The R2R and batch processes formed similar structure on polylactide and cellophane substrates. However, for cellophane film, the Al2O3 coating seemed to form larger particles than observed on PI and PLA substrates. The cellophane film substrate with and without ALD-coating is presented in Fig. 5. These differences may be due to the initial surface roughness of
Fig. 4. The AFM images of uncoated and Al2O3-coated polyimide (PI) films. The Al2O3 coatings have been fabricated by using both batch and R2R process methods.
Fig. 5. The AFM images of uncoated and Al2O3-coated cellophane (CEL) films. The Al2O3 coatings have been fabricated by using both batch and R2R process methods.
168
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169 Table 3 The OTR and WVTR values for plain and Al2O3-coated plastic samples. The Al2O3-coating has fabricated by using both, batch and R2R ALD processes. Substrate
CEL PLA PI a
Fig. 6. The FTIR spectra of four Al2O3 coatings with different thicknesses. The reference sample (84 nm) was deposited with the batch ALD reactor. The spectrum of the reference sample was compared to the spectra of the three samples (16 nm, 20 nm and 32 nm) deposited by using the R2R add-on.
the uncoated substrate or to the differences in the surface chemistry between the substrates. For the FTIR analysis, four Al2O3 coatings with different thicknesses were grown by the R2R process on the silicon wafer. The reference sample thick film of Al2O3 (84 nm) was deposited with the batch ALD reactor on the silicon wafer. The spectrum of the reference sample was compared to the spectra of the three samples (16 nm, 20 nm and 32 nm) deposited on the silicon wafer by using the R2R add-on. FTIR analyses indicated Al2O3 covered surfaces after the R2R ALD. The measured surface IR spectra at region 4000 – 600 cm−1 are shown in Fig. 6. In the sample of 16 nm film the main IR band of bulk Al2O3 at the region of 1000–850 cm−1 is rather weak. The growth of aluminum oxide film is clearly seen in the IR spectra of 20 nm or 32 nm films, where the bulk absorptions of Al2O3 at 950–700 cm−1 are stronger, as the film thickness is increasing [13,25]. The band position of the strongest IR band of aluminum oxide film is transferring to lower wavenumber direction towards the band position of reference sample (883 cm−1), with 84 nm Al2O3, as aluminum oxide film is increased from 16 nm to 32 nm. This shift indicates that when the film thickness increases, the layer structure is more similar to the reference thick film aluminum oxide. Contact angle and surface energy measurements revealed some differences on the chemical composition between batch and R2R deposition methods. The results are presented in Table 2. Al2O3 thin films fabricated with the R2R on the polymer films had lower polar components and overall surface energy than the thin films obtained with the batch process. Without further analysis it can be speculated that this could be due to relatively low deposition temperature and thus an increased amount of non-polar impurities on the Al2O3 surface, i.e. hydrocarbon compounds not removed by the nitrogen purge in the roll-toroll process. The FTIR analysis of Al2O3 deposited onto silicon wafers also indicated the presence of such impurities, as weak bands of C–H
Table 2 Contact angle (CA), surface energy (SE) values for plain and Al2O3-coated polymers. The total value of surface energy (SE) is the sum of dispersive (SEd) and polar (SEp) components. Sample
CA (°)
CEL CEL + Al2O3 batch CEL + Al2O3 R2R PLA PLA + Al2O3 batch PLA + Al2O3 R2R PI PI + Al2O3 batch PI + Al2O3 R2R
40 60 74 71 48 78 70 52 102
± ± ± ± ± ± ± ± ±
1 4 3 1 3 3 2 2 2
SEp (mN/m)
SEd (mN/m)
SE (mN/m)
22.9 13.2 7.0 8.2 22.8 6.6 5.5 19.9 0.3
36.4 34.0 31.8 32.6 30.9 29.0 42.9 30.6 26.2
59.3 47.2 38.8 40.8 53.6 35.6 48.4 50.5 26.5
OTR
WVTR
Polymer
+ Al2O3 Batch
+ Al2O3 R2R
Polymer
+ Al2O3 Batch
+ Al2O3 R2R
4.0 ± 0.1 470 ± 1 30 ± 0.1
2.6 ± 2.3 0.4 ± 0.1 b0.01 a
0.3 ± 0.2 47 ± 6 b 2.2 ± 1.9 b
144 ± 19 39 ± 5 3±1
29 ± 16 b0.2 a b0.1 a
15 ± 9 10 ± 1 b 2±1b
Result was under the detection limit; b Small sample size.
stretching vibrations are seen in the region of 2920 – 2960 cm−1 of 20 and 32 nm samples, see Fig. 6. The Al2O3 surfaces as such are intrinsically hydrophilic in nature [26]. The transitions in the contact angles after initial ALD cycles can be explained by changes in surface roughness, nucleation effects and chemical bonding [22,27,28]. A hydrophobic effect has been observed on cellulose after the first TMA deposition cycles explained by the stable Al–(O–C–)3 groups with minimal hydrophilic Al–OH groups and carbon adsorption after deposition [29]. Retention of by-products into the plastic films and thus a lower polarity of the Al2O3 layer could be due to short purge/water pulses and both side deposition. In addition, winding the reel could impair purging and prevent desorption of by-products. Low deposition temperature (b100 °C) usually leads to a certain amount of impurities. The oxygen and water vapor barrier results achieved for the polymer substrates with and without of the Al2O3 coating deposited by batch and R2R ALD processes are summarized in Table 3. Al2O3 coating was deposited on one side in batch processing and on both sides of the substrates in the R2R process. Independently of the deposition process used, the Al2O3 coating improved both the oxygen and water vapor barrier properties of the polymeric samples. Thermal stability and viscoelastic response of polymer substrates to mechanical stresses during deposition can have an effect on the barrier properties after R2R deposition. Similarly, defects created during winding, such as cracking, are challenges for the industrial use of R2R ALD. The probability of such damages seems to depend on the polymer properties. For example, CEL is antistatic attracting less charged particulate contaminants. Particulate contaminants are a known cause of defects in e.g. vacuum metallizing. It has also been shown that the surface chemistry has an impact. 25 nm of Al2O3 had better oxygen barrier on cellophane than on PLA, and straining the film had a less negative effect on oxygen barrier in the case of cellophane [30]. In addition, abrasion testing of PLA and cellophane films with 25 μm of Al2O3 has indicated that oxygen barrier is impaired less by abrasion with cellophane than with PLA (0–60 vs. 90–130 times higher, respectively, after abrasion depending on the test method). Cellophane has more reactive hydroxyl surface groups, while PLA has also carbonyl groups. Polyimide is more hydrophobic with carbonyl groups. Differences in surface and diffusion properties can lead to different thin film growth, structure and properties [2,31]. Similarly, an Al2O3 coating on CEL was in this study less sensitive to various defects, and the barrier values obtained with an Al2O3 coating on both sides were still better than those obtained with the same substrate coated only on one side. With PLA and PI the barrier results were impaired to a level exceeding the barrier values of one side coated substrates. In the latter two cases, the size of the samples fabricated by R2R was also smaller than the samples fabricated by the batch process. Aluminum masks had to be used with the small samples. This can lead to poorer detection limit and larger edge effect and thus affect OTR and WVTR values. 4. Conclusions The positive effect of the Al2O3 coating on the barrier properties of plastic films was evident with both processes. Thermal stability, viscoelastic properties at elevated temperatures and sensitivity to various
T. Hirvikorpi et al. / Thin Solid Films 550 (2014) 164–169
defects during winding are parameters likely affecting the barrier properties of plastic films after the roll-to-roll (R2R) ALD. There were differences between the films also probably due to different surface chemistry. For example, the Al2O3 coatings deposited with the R2R process on both sides of the cellophane film were less sensitive to defects than coated polylactide and polyimide films, and the barrier values were impaired to a level still better than that obtained with cellophane one side coated in the batch process. However, with polylactide and polyimide films the defects during the R2R process led to barrier properties inferior to those of one side coated films. The related phenomena need to be investigated more in the future. The FTIR analyses detected Al2O3 covered surfaces after the R2R ALD. AFM images for the batch and R2R produced samples were also quite similar. There were slight indications of larger clusters with cellophane and increased roughness in the μm scale with polylactide film after the R2R ALD. The relative polarities of surface energy for Al2O3 deposited with R2R ALD on all three coatings were lower than for the batch samples. This indicates also some differences in the thin film growth between the batch and R2R processes. References [1] P.F. Carcia, R.S. McLean, M.H. Reilly, M.D. Groner, S.M. George, Appl. Phys. Lett. 89 (2006) 031915. [2] M.D. Groner, S.M. George, R.S. McLean, P.F. Carcia, Appl. Phys. Lett. 88 (2006) 051907. [3] A.A. Cameron, S.D. Davidson, B.B. Burton, P.F. Carcia, R.S. McLean, S.M. George, J. Phys. Chem. C 112 (2008) 4573. [4] T. Hirvikorpi, M. Vähä-Nissi, A. Harlin, M. Karppinen, Thin Solid Films 518 (2010) 5463. [5] T. Hirvikorpi, M. Vähä-Nissi, T. Mustonen, E. Iiskola, M. Karppinen, Thin Solid Films 518 (2010) 2654. [6] T. Hirvikorpi, M. Vähä-Nissi, J. Nikkola, A. Harlin, M. Karppinen, Surf. Coat. Technol. 205 (2011) 5088.
169
[7] T. Hirvikorpi, Thin Al2O3 Barrier Coatings Grown on Bio-based Packaging Materials by Atomic Layer Deposition, VTT, 770, VTT Publications, 2011. [8] C. Andersson, Packag. Technol. Sci. 21 (2008) 339. [9] J. Kuusipalo, K. Lahtinen, TAPPI J. 7 (2008) 8. [10] A. Savolainen, J. Kuusipalo, E. Laiho, T. Penttinen, in: A. Savolainen (Ed.), Paper and Paperboard Converting, Fapet Oy, Helsinki, 1998, p. 123. [11] Y. Leterrier, Prog. Mater. Sci. 48 (2003) 1. [12] R. Puurunen, J. Appl. Phys. 97 (2005) 121301. [13] J.D. Ferguson, A.W. Weimer, S.M. George, Chem. Mater. 16 (2004) 5602. [14] S.H.K. Park, J. Oh, C.S. Hwang, J.I. Lee, Y.S. Yang, H.Y. Chu, Electrochem. Solid-State Lett. 8 (2005) H21. [15] R. Cooper, H.P. Upadhyaya, T.K. Minton, M.R. Berman, X. Du, S.M. George, Thin Solid Films 516 (2008) 4036. [16] Y. Zhang, J.A. Bertrand, R. Yang, S.M. George, Y.C. Lee, Thin Solid Films 517 (2009) 3269. [17] P.F. Carcia, R.S. McLean, S. Hegedus, Sol. Energy Mater. Sol. Cells 94 (2010) 2375. [18] T.O. Kääriäinen, P. Maydannik, D.C. Cameron, K. Lahtinen, P. Johansson, J. Kuusipalo, Thin Solid Films 519 (2011) 3146. [19] X. Liang, L.F. Hakim, G.D. Zhan, J.A. McCormick, S.M. George, A.W. Weimer, J.A. Spencer, K.J. Buechler, J. Blackson, C.J. Wood, J.R. Dorgan, J. Am. Ceram. Soc. 90 (2007) 57. [20] D.K. Owens, R.C. Wendt, J. Appl. Polym. Sci. 13 (1969) 1741. [21] C.A. Wilson, R.K. Grubbs, S.M. George, Chem. Mater. 17 (2005) 5625. [22] M. Vähä-Nissi, E. Kauppi, K. Sahagian, L.-S. Johansson, M.S. Peresin, J. Sievänen, A. Harlin, Thin Solid Films 522 (2012) 50. [23] K. Grigoras, S. Franssila, V.M. Airaksinen, Thin Solid Films 516 (2008) 5551. [24] R.L. Puurunen, W. Vandervorst, W.F.A. Besling, O. Richard, H. Bender, T. Conard, C. Zhao, A. Delabie, M. Caymax, S. de Gendt, M. Heyns, M.M. Viitanen, M. de Ridder, H.H. Brongersma, Y. Tamminga, T. Dao, T. de Win, M. Verheijen, M. Kaiser, M. Tuominen, J. Appl. Phys. 96 (2004) 4878. [25] D.N. Goldstein, J.A. McCormick, S.M. George, J. Phys. Chem. C 112 (2008) 19530. [26] D.S. Finch, T. Oreskovic, K. Ramadurai, C.F. Herrmann, S.M. George, R.L. Mahajan, J. Biomed. Mater. Res. A 87 (A) (2008) 100. [27] R.L. Puurunen, Appl. Surf. Sci. 245 (2005) 6. [28] G.K. Hyde, G. Scarel, J.C. Spagnola, Q. Peng, K. Lee, B. Gong, K.G. Roberts, K.M. Roth, C.A. Hanson, C.K. Devine, S.M. Stewart, D. Hojo, J.-S. Na, J.S. Jur, G.N. Parsons, Langmuir 26 (2010) 2550. [29] K. Lee, J.S. Jur, D.H. Kim, G.N. Parsons, J. Vac. Sci. Technol. A 30 (2012) 01A163. [30] M. Vähä-Nissi, P. Sundberg, E. Kauppi, T. Hirvikorpi, J. Sievänen, A. Sood, M. Karppinen, A. Harlin, Thin Solid Films 520 (2012) 6780. [31] B. Gong, G.N. Parsons, J. Mater. Chem. 22 (2012) 15672.