You are on page 1of 7

Carbon 43 (2005) 303–309

www.elsevier.com/locate/carbon

Structural investigation and gas barrier performance of


diamond-like carbon based films on polymer substrates
G.A. Abbas, S.S. Roy *, P. Papakonstantinou, J.A. McLaughlin
Nanotechnology Research Institute, School of Electrical and Mechanical Engineering, University of Ulster, Newtownabbey, BT37 0QB, UK

Received 28 June 2004; accepted 9 September 2004

Abstract

In this report, tetrahedral amorphous carbon (ta-C), hydrogenated amorphous carbon (a-C:H), silicon doped tetrahedral amor-
phous carbon (ta-C:Si:H), and silicon doped hydrogenated amorphous carbon (a-C:H:Si) films with thickness in the range 50–
370 nm have been produced by PECVD (Plasma Enhanced Chemical Vapour Deposition) and FCVA (Filtered Cathodic Vacuum
Arc) techniques on Polyethylene terepthalate (PET) and polycarbonate (PC) substrates. The paper is concerned with exploring the
links between the atomic structure, gas barrier performance in carbon based films deposited on polymer substrates. A range of tech-
niques including XRR, NEXAFS, Raman, surface profilometry, nano-indentation and water vapour permeation analysis were used
to analyze the microstructure and properties of the films. The intensity and area of p* peak at the C K (carbon) edge of the NEXAFS
spectra was lower in the FCVA films in comparison to that of PECVD ones confirming the higher sp3 content of FCVA films. The
surface of ta-C films showed a network of micro-cracks, which is detrimental for gas barrier application. However, the surfaces of
both ta-C:H:Si and a-C:H:Si silicon-incorporated films were almost free of cracks. We also found that the incorporation of Si into
both types of DLC films lead to a significant reduction of water vapour transmission rate.
 2004 Elsevier Ltd. All rights reserved.

Keywords: A. Diamond-like carbon; B. Arc discharge, Chemical vapor deposition; D. Microstructure

1. Introduction resistance, and high susceptibility to either gases or


chemicals.
Plastics are presently widely used as cheap, transpar- Alternatives have been developed for modifying the
ent, light-weight, and unbreakable substitutes for glasses polymer surfaces to have a high gas barrier property,
in several applications, including automotive and soft good adhesion and chemical stability in different hazard
drink industries as well as for organic light emitting de- environments [1–6]. The surface modification of poly-
vices (OLED) display applications. Moreover, a variety mer films with amorphous diamond-like carbon (DLC)
of medical devices made of polymers are now been used is one of these new promising methods. DLC films have
as ophthalmic lenses and as heart valves being im- all the properties needed for excellent barrier coatings
planted in human body. However, the use of polymers including: high hardness, excellent chemical and thermal
is limited because of its low hardness, poor scratch stability and it is transparent in the visible spectrum. In
addition, carbon is already present in plastics, and there
is no implication with regards to recycling.
Recently, a number of techniques have been used for
*
Corresponding author. Tel.: +44 28 80368997; fax: +44 28 903
coating polymer substrates with DLC films including
66863. PECVD [7], ion implantation (II) [8–10], ion beam as-
E-mail address: s.sinha-roy@ulster.ac.uk (S.S. Roy). sisted deposition (IBAD) [11], and pulsed laser deposition

0008-6223/$ - see front matter  2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbon.2004.09.016
304 G.A. Abbas et al. / Carbon 43 (2005) 303–309
 2  
(PLD) [12], techniques. To our knowledge, so far there is Es ts 1 1
rs ¼  ; ð1Þ
no work reported on ta-C coatings on polymers synthe- 6ð1  ms Þ tf R R0
sized by the FCVA technique. The high level of compres-
sive stress in ta-C films is a major drawback of these types where Es, ms, and ts are the Young modulus, Poisson ra-
of coatings. It does not only put an upper limit to the tio and thickness of the substrate respectively. The R0
thickness of the coatings over wide range of substrates, and R are the radii of curvature of the substrate before
but it also has a dramatic effect on the surface morpho- and after coating, respectively. An optical profilometer
logy on soft substrates such polymers [12]. This area is lar- [WYKO NT800] was used for surface morphology
gely unexplored. study. Nano-indentation tests were carried out using a
In this study, we demonstrate the fabrication of supe- Nano-indenter XP (Nanoinstruments Inc) with a high-
rior gas barrier, hard, transparent and stress free DLC resolution (1 nN) dynamical contact module (DCM)
coatings on different types of polymer substrates such head. In addition, the indenter was fitted with the con-
as PC and PET. We propose a strategy (incorporation tinuous stiffness mode (CSM) option. The indentations
of silicon) for releasing the compressive stress in these were performed using CSM mode at a constant rate of
films. Moreover, the silicon-induced-modification in sur- 0.05 s1 up to 100-nm depth. The Hardness (H) values
face morphology, electronic structure, and the gas bar- are averaged over eight indents.
rier performance of different DLC films is also explored. The silicon content in DLC films was determined by
X-ray photoelectron spectroscopy (XPS) in an XSAM
800 (KRATOS) spectrometer. An ISA Labram 300 Ra-
man spectroscope equipped with Ar ion laser (514.5 nm)
2. Experimental was used to examine the structure of the films. In order
to precisely estimate the sp2 contents in the film, NEX-
In this study we used untreated 75-lm thick optical AFS measurement was carried out at the Synchrotron
grade Poly(ethylene terepthalate) (PET) (Hostaphane) Radiation facility of CCLRC at Daresbury using the
and polycarbonate (PC) coated commercial CDs as sub- Beam Line 1.1. Soft X-rays generated from synchrotron
strates. The pure ta-C films were produced by FCVA source were used for scanning the films within energy
system equipped with double bend off-plane (DBOP) fil- range of 275–320 eV.
ter to screen the macro-and micro-particles. A scanning
waveform was used at the exit of the plasma duct to scan
the plasma beam before deposition onto the out-of-sight 3. Results and discussion
substrates. The substrate holder has been also rotated to
increase the thickness uniformity of the deposited films. 3.1. Density, stress and gas barrier properties of carbon
Films were prepared under floating condition, i.e. with- nano-layers on PET
out any substrate bias with 80 A arc current at room
temperature and a pressure of 1 · 106 Torr. The pure X-ray reflectivity (XRR) is a widely used technique,
a-C:H films were deposited by RF (13.56 MHz) PECVD for determining density and thickness of thin films by
using C2H2/Ar plasma with a ratio of 2:1 at electrode measuring the reflection coefficient as a function of the
self-bias voltage of 300 V. For silicon-incorporated sam- glancing angle of incidence, R = f(h). Fig. 1 shows two
ples, Tetramethylsilane (TMS) vapour was introduced representative XRR spectra of a PECVD (100-nm thick)
as a Si source inside the plasma deposition chambers and a FCVA (50-nm thick) samples. The reflectivity
for both coating systems. Details of the deposition tech- sharply decreases at the critical angle from which the
niques can be found elsewhere [13]. density (q) value can be evaluated [14]. One can notice
The XRR measurements were performed using a the difference in the critical angle of the two films, with
Philips Analytical high resolution X-ray diffractometer a lower incidence angle shift as an indication of lower
using a Cu Ka (k = 1.541 Å) for the determination of density. The ta-C film has a higher density (3.2 g/cm3)
the film density and thickness. A nano-indenter was in comparison to the PECVD film, in which the value
used to measure the hardness of the films. A water va- of density reduced to be 2.3 g/cm3.
pour permeation analysis system (Permatran-W 3/31) This is an expected result. Obviously, the energy of
was used for Water vapour transmission rate (WVTR) the carbon ions together with the hydrogen content play
measurements under conditions of (20 C and 70% a crucial role in determining the density value. For the
RH). A surface profilometer (TencorP-10) was used to FCVA system, no hydrogen was included, the carbon
measure the residual stress of DLC films on Si substrates ions were of relatively high energy, mono-energetic
by using the radius of curvature method with an average (20–30 eV) and highly ionized. This is contrast to the
of 6 measurements at different locations for each film. PECVD system, where the precursor gas (C2H2) in-
The stress, rs, in a thin film of thickness tf is given by cluded hydrogen, the ion energies were broadly distrib-
stoneyÕs equation uted around value of 5 eV leading to a lower film
G.A. Abbas et al. / Carbon 43 (2005) 303–309 305

tent, have a high diamond-like property (sp3  70%)


with high stress values up to (8–12 GPa). On the other
10
4
hand, PECVD a-C:H films contain higher hydrogen
content (up to 30%), which resulted in less compact struc-
Reflectivity (a.u)

FCVA ( =3.1 g cm-3 ) ture, containing high percentage of voids, with lower
stress values.
There are two factors could be responsible on the
10
2
relaxation of the DLC structure. Firstly, the formation
of weaker Si–H bonds, which caused the relaxation of
PECVD (ρ =2.3 g cm-3 ) a three-dimensional rigid network and reduced the com-
pressive stress in the films [17]. Secondly, the substitu-
tion of carbon atoms of the sp3 bonds by Si atoms
0
10 and the formation of Si–C bonds can play a role in
500 1000 1500 2000 2500 3000 releasing bond distortion of the nearby strong sp3 C–C
Incident angle (arcsecond) bonds (3.7 eV; 411 kJ mol1) with large strain in the
Fig. 1. Two representative XRR spectra of a-C:H film (100-nm thick) weaker Si–C bonds (3.21 eV; 320 kJ mol1) and decreas-
and ta-C (50-nm thick) deposited on silicon substrates using PECVD ing the residual stress of the DLC structure [18,19]. We
and a FCVA, respectively. have to emphasize that using TMS precursor gas as a sil-
icon source inside the FCVA deposition chamber has
density. According to the subplantation model, the high- been resulted in incorporation of hydrogen as well as sil-
est densities occur for ion energies just above a critical icon in the ta-C films. Therefore, the Si incorporation in
value, which is necessary to penetrate the surface layers both ta-C and a-C:H films seems to relax the stressed
inducing stress, transforming sp2 states into sp3 states structure but with development of more polymer-like
[15,16]. structure in these films.
The measurements of water vapour transmission rate,
WVTR, for a-C:H, ta-C and Si doped DLC coatings on 3.2. Structural characterization of carbon nano-layers
PET substrates are listed in Table 1. The results show on PC
that, undoped DLC coatings did not show a significant
reduction in the water vapour transmission rate. More- 3.2.1. Raman spectroscopy
over, lower density PECVD film had a superior gas Depending on the sp3/sp2 ratio of carbon bonds, the
barrier property (WVTR = 0.56 g/m2 day) than the high- visible Raman spectra of DLC films always show a sin-
er-density ta-C film (WVTR = 1.4 g/m2 day). Silicon- gle G peak or two peaks (G peak and D peak) in the
incorporation into both the hydrogenated amorphous 1100–1800 cm1 region [20,21]. The G peak is connected
carbon and tetrahedral amorphous carbon films reduced to the presence of all sp2 structures, both olefinic
the water vapour transmission rate over two orders of (chains) and aromatic (rings) while the D peak is related
magnitude. to the presence of aromatic rings only [20]. Two repre-
It appears that the Si-incorporation succeeded in sentative Raman spectra are demonstrated in Fig. 2.
relaxing the stressed-structures for both types of DLC One can clearly observe the difference between the ta-
films. However, different amount of silicon was neces- C and a-C:H spectra. Significant changes occurred near
sary to release the different level of stress in both films. the D peak position. The D peak was prominent in a-
While a silicon content of 7.8 at.% was enough for relax- C:H film indicating a higher sp2 content. This shoulder
ing a stress level of 3.5 Gpa in a-C:H, a 21.5 at.% of sil- has disappeared and the spectrum becomes more sym-
icon was necessary for releasing a compressive stress of metric for ta-C film. The spikes appear in the spectra
12 GPa in ta-C films (see Table 1). due to the polymer substrate.
Both density and stress measurements are in agree- The spectra were fitted with two Gaussian peaks for
ment. Usually, ta-C films, with very low hydrogen con- G and D bands. Raman parameters calculated from

Table 1
Summary of the study of DLC nano-nalyes on PET
DLC type Deposition method Thickness (nm) Si% WVTR (g/m2 day) Density (g/cm3) Stress (GPa)
a-C:H PECVD 148 0 0.5611 2.27 3.5
a-C:H:Si PECVD 237 7.8 0.0316 1.94 0.5
ta-C FCVA 127 0 1.407 3.2 8–13
ta-C:Si:H FCVA 370 21.5 0.0892 2.2 0
Uncoated-PET – – – 1.614 – –
306 G.A. Abbas et al. / Carbon 43 (2005) 303–309

Fig. 2. (a) Typical Raman spectra of a-C:H films synthesized by PECVD. (b) Raman spectra of a ta-C films synthesized by FCVA.

Table 2
Summary of the studies of ta-C and a-C:H nano-layer coated PC substrates
DLC Type Thickness (nm) ID/IG ratio G position G width sp2 content Hardness (GPa)
a-C:H by PECVD 100 0.83 1554 150 60% 2
ta-C by FCVA 100 0.25 1565 220 35% 8

the fittings are listed in Table 2. The ID/IG ratio in- The increase in sp2 cluster size in a-C:H films is in agree-
creased and the G linewidth and G position decreased ment with a reduction in hardness as evaluated by nano-
in a-C:H films compared to those of ta-C films. In amor- indentation studies shown in Table 2.
phous carbon films the ID/IG intensity ratio, G line The measurement of the mechanical properties on
width and G peak position are the most commonly used thin films is problematic due to the effect of the underly-
indicators of any structural changes. The high ID/IG ra- ing substrate. The common approach for dealing with
tio and narrow G line width in a-C:H films indicates this problem and to get the substrate-independent meas-
higher sp2 content. This prediction was further sup- urements is to restrain the indentation depth to less than
ported by NEXAFS measurements reported later. It is 10% of the film thickness [23]. However, for films 100-
well established that the G peak dispersion of any car- nm thick as in our case, 1/10 of film thickness corre-
bon films depend on various factors such as film thick- sponds to 10 nm. Accurate measurements at this scale
ness, deposition condition, substrate, laser wavelength, are difficult to obtain due to vibration, tip rounding ef-
film stress and to some extent on instrumental precision. fects and difficulties in precise determination of the loca-
So the above factors should be taken into consideration tion of the specimen surface. Since the main purpose in
when comparing results from the literature. We found our study is to obtain a ranking in the mechanical prop-
(Table 2) the ta-C film has a higher G peak position than erties between ta-C and a-C:H films, we considered data
that of a-C:H film. Recently, Shin et al. [22] found the G obtained at a contact depth of 50% (around 50 nm in our
peak of the Raman spectrum shifts to higher wave num- case) of the coating thickness as representative of the
bers with the increase of compressive stress. The higher composite film/substrate.
stress in ta-C films may be one of the reasons for the
higher G wave number in ta-C films. 3.2.2. NEXAFS spectroscopy
The intensity of the D band is quite sensitive to the The sp2 content was quantitatively determined for ta-
size of crystalline nano-domains. The relation for esti- C and a-C:H films by near edge X-ray absorption fine
mating the size of ordered regions from the ID/IG ratio structure (NEXAFS) spectroscopy and is listed in Table
in the Raman spectra of carbon is as follows [20]: 2. The details of the sp2 content determination are pro-
ID vided in the literature [24]. Fig. 3a shows a C K-edge
¼ CðkÞL2a ; ð2Þ NEXAFS spectra of ta-C and a-C:H films on PC sub-
IG
strate along with HOPC graphite. All spectra were nor-
where C is a constant having a value of approxi- malized to the maximum peak at about 290 eV and
mately 0.0055 at 514.5 nm and La (620 Å) is a linear shifted upward for ease of comparison.
dimension of the ordered region. The sp2 cluster size A pre-edge resonance at 285.5 eV is due to transitions
estimated from Eq. (2) was much higher for a-C:H films from the C 1s orbital to the unoccupied p* orbitals ide-
(12.28 Å) than ta-C films (6.7 Å). The increase of ally originating from sp2 (C=C) sites, and also from sp
cluster size at a-C:H films leads to the formation of (C„C) sites, if present. This peak is not visible in the
structures with reduced three-dimensional cross-linking. spectrum of diamond, because diamond does not have
G.A. Abbas et al. / Carbon 43 (2005) 303–309 307

is very sensitive to the orientation of the p* orbital with


respect to the polarization vector. Thus, if the p* orbit-
als in the DLC specimen are partially oriented with re-
spect to the incident photon beam, a rotation of the
film plane with respect to the incoming photon will show
a recognisable angular dependence. Only a C–C or C–H
r* orbital orthogonal to the p* orbital will show an
opposite trend.
Fig. 3b illustrates that the p* resonance intensity de-
creases from normal to 30 incidence while the C–H r*
exhibits an opposite trend. Our results clearly reflect that
the (C=C) p* and C–H r* bonding configurations are
orthogonal to each other. The intense C–H r* reso-
nance observed at lower angle indicates the presence
of a C–H bond in the graphitic-like basal plane. This ef-
fect is being very interesting since in general amorphous
films do not show any partial orientation. Our result
indicates that the sp2 clusters are not randomly oriented
in the ta-C films on PC substrate. Intensity of the C–H
(r*) peak as depicted in Fig. 3b showed an opposite
trend to that of C=C (p*) peak. Recently Lee et al.
[26] found by a combination of NEXAFS and electron
spin resonance spectroscopies that the surface defects
or dangling bonds in ta-C films are proportional to the
content of the sp3 hybridized bond. This means that a
higher sp3 content will be accompanied by higher defect
density. This is in agreement with our NEXAFS data of
Fig. 3 which show a high rC–H contribution in ta-C
films. The rC–H contribution in a-C:H films mainly from
the intrinsic C–H bonds of the films. The defect density
Fig. 3. (a) C K-edge NEXAFS spectra of ta-C and a-C:H films on PC in a-C:H films might be lower, as the number of sp2 site
substrate along with graphite reference. (b) Angular dependence of the increases in the a-C:H films most unpaired sp2 sites do
C K-edge NEXAFS of a ta-C film on PC substrate.
not act as the defects but form carbon carbon double
bonds [26].
any sp2 (C=C) sites. Therefore, the peak intensity of this
resonance can be considered as an index of sp2 content. 3.3. Surface morphology
It is clear from Fig. 3a that the intensity and area of p*
peak (which indicates the sp2 content in the film) at the The Fig. 4a shows clearly a network of cracks spread
C K (carbon) edge decreased significantly in ta-C films in out over the whole membrane surface with crack depths
comparison to that of a-C:H films. reaching down to the film-polymer interface. By using
It is important to note that, the NEXAFS spectrum silicon incorporation in the a-C:H films, the cracks com-
solely shows information on the surface of the film and pletely disappeared (as in Fig. 4b) [13]. Surface rough-
it is not influenced by film thickness. The peak near ness of all films was smaller than 10 nm. Fig. 5
289 eV corresponds to the r* C–H transition of the C– displays optical microscopy images of various carbon-
H bonds that would originate from the inherent C–H based films coated on the PET membrane. Both pure
bonds in films and/or from the absorption of hydrocar- ta-C and a-C:H films on polymer substrate had surfaces
bon to the surface dangling bonds [25,26]. The broad covered by network of micro-cracks. However, ta-C
band between 290 and 310 eV results from overlapping films were characterised with deeper and wider micro-
C 1s ! r* transition at sp2 or/and sp3 sites. Finally the cracks in comparison to those of a-C:H, (Fig. 5) which
sp2 content in a-C:H films deposited on PC is 25% higher is most probable due to the high stress values of ta-C
than that of ta-C films, and is consistent with other spect- films (8–13 GPa) with respect to those of a-C:H films
roscopic measurements. (3.5 GPa).
The presence of local order in the DLC nano-layer is Due to the high intrinsic stress in DLC films and lower
observed (Fig. 3b) in the angular dependence of the surface tension of polymer substrates, surface crack-
NEXAFS. As the synchrotron light is linearly polarized ing has been previously observed for a-C:H [27,28] and
horizontally, the resonance intensity of the p* transition ta-C films [12,26] on different polymeric substrates.
308 G.A. Abbas et al. / Carbon 43 (2005) 303–309

Fig. 4. The optical profilometer images of (a) ta-C and (b) a-C:H:Si films on PC substrates.

Fig. 5. Optical microscopy images with a magnification · 10 of carbon based films on PET substrate (a) a 130-nm thick a-C:H layer, (b) a 150-nm
thick hard ta-C layer and (c) 300-nm thick ta-C:Si:H layer.

Earlier, Ong et al. [29] observed a similar surface mor- particles with low energy stick to the growing surface
phology for ta-C films on polycarbonate (PC) substrate. are responsible for the higher inclusion of sp2 sites.
They showed that oxygen plasma pretreatment for 1 h The mechanical properties are determined by the sp3
was efficient in releasing the stress and improving both content as well as hydrogen content in the films. The
the surface morphology and the adhesion. They sug- dense and compact structure of the ta-C film with very
gested that plasma treatment caused an introduction of low hydrogen content (<2%) is responsible for the higher
polar groups onto the surface of the polymer, which en- hardness value.
hanced the adhesion. Moreover, Bonelli et al. [12] per- Therefore, we may conclude that both incorporation
formed residual stress measurements for similar PLD of hydrogen and silicon in ta-C and a-C:H films have re-
ta-C films on the same PC substrates. Their films exhib- sulted in relaxing the DLC structure accompanied by a
ited a residual stress value of 4.5 GPa. They showed that complete reduction in the compressive stress with con-
incorporation of nitrogen resulted in reduction of the verting the structure to more-polymer-like structure
high residual stress accompanied by a crack free surface. with reduced hardness values but with superior gas bar-
We have to emphasize that, the stress value on PC rier properties.
substrate is expected to be lower than the corresponding
value on silicon substrate due to release of stress
through surface cracking and deformation of polymer 4. Conclusions
substrate. In this study, plastic deformation of the
PET substrates was obvious with various ta-C thick- The present study of various carbon-based nano-lay-
nesses, which is an indication of the relaxation process ers on polymer substrates can be summarised as follows:
of the high residual stress status in these films. Also,
the improvement in surface morphology in both a-C:H (i) Soft a-C:H (2.3 g/cm3) and hard ta-C (3.1 g/cm3)
and ta-C films can explain the enhancement in gas bar- coatings were deposited by PECVD and FCVA
rier property by silicon addition. Moreover, it shows the techniques, respectively on PET and PC substrates.
important role of the surface morphology of the differ- (ii) A network of micro-cracks was found on all the
ent DLC films on their gas barrier performance. coatings surfaces with deeper and wider spacing
All the above-mentioned results are in agreement for ta-C films, which was suggested to be an intrin-
with each other and can also be explained in terms of sic property, correlated to the intrinsic stress (8–
deposition mechanism. The hardness was found to de- 12 GPa).
crease in a-C:H films in comparison to that of ta-C films (iii) Si-incorporated a-C:H and ta-C films have superior
(Table 2). The increase of sp2 fraction in a-C:H films is gas barrier properties and can be used as an alterna-
responsible for reduction in hardness for these types of tive barrier coating to the metallised barrier
films. PECVD deposition mechanism in which, neutral coatings.
G.A. Abbas et al. / Carbon 43 (2005) 303–309 309

(iv) NEXAFS result confirm that the DLC films on pol- [12] Bonelli M, Miotello A, Mosaner P, Casiraghi C, Ossi PM. Pulsed
ymer creates a locally ordered sp2 bonded network laser deposition of diamondlike carbon films on polycarbonate. J
Appl Phys 2003;93:859–65.
on the film surface and the sp2 content in a-C:H [13] Abbas GA, McLaughlin JA, Harkin-Jones E. A study of ta-C,
films are higher than that of ta-C films. a-C:H and Si-a:C:H thin films on polymer substrates as a gas
(v) Thin ta-C-films have improved the hardness values barrier. Diamond Relat Mater 2004;13:1342–5.
of PC substrate by two orders of magnitude. [14] Abbas GA, Papakonstantinou P, McLaughlin JA. X-ray reflec-
tivity, photoelectron and nanoindentation studies of tetrahedral
amorphous carbon (ta-C) films synthesized by double bend
cathodic arc. Diamond Relat Mater 2004;13:1386–490.
[15] Robertson J. Deposition mechanisms for promoting sp3 bonding
Acknowledgment in diamond-like carbon. Diamond Relat Mater 1993;2:984–9.
[16] Davis CA. A simple model for the formation of compressive
The NEXAFS measurements were performed at stress in thin films by ion bombardment. Thin Solid Films
1993;226:30–4.
CCRLC facility at Daresbury and the XRR measure- [17] Ban M, Hasegawa T. Internal stress reduction by incorporation of
ments at Seagate Technologies, N. Ireland. silicon in diamond-like carbon films. Surf Coat Technol 2002;
162:1–5.
[18] Lee CS, Lee KR, Eun KY, Yoon KH, Han JH. Structure and
References properties of Si incorporated tetrahedral amorphous carbon films
prepared by hybrid filtered vacuum arc process. Diamond Relat
Mater 2002;11:198–203.
[1] Moosheimer U, Bichler Ch. Plasma pretreatment of polymer films
[19] Jung HS, Park HH. Determination of local bonding configuration
as a key issue for high barrier food packagings. Surf Coat Technol
and structural modification in amorphous carbon with silicon
1999;116–119:812–9.
incorporation. Diamond Relat Mater 2003;12:1373–7.
[2] Haas KH, Amberg-Schwab, Rose SK, Schottner G. Functional-
[20] Ferrari AC, Robertson J. Interpretation of Raman spectra
ized coatings based on inorganic–organic polymers (ORMO-
of disordered and amorphous carbon. Phys Rev B 2000;61:
CERs) and their combination with vapor deposited inorganic
14095–107.
thin films. Surf Coat Technol 1999;111:72–9.
[21] Roy SS, Papakonstantinou P, McCann R, Abbas G, Quinn JP,
[3] Walther M, Heming M, Spallek M. Multilayer barrier coating
McLaughlin J. Bonding configurations in DBOP-FCVA nitro-
system produced by plasma-impulse chemical vapor deposition
genated tetrahedral amorphous carbon films studied by Raman
(PICVD). Surf Coat Technol 1996;80:200–2.
and X-ray photoelectron spectroscopies. Diamond Relat Mater
[4] Schiller N, Reschke J, Goedicke K, Neumann M. Application of
2004;13:1459–63.
the magnetron activated deposition process (MAD-process) to
[22] Shin JK, Lee CS, Lee KR, Eun KY. Effect of residual stress on the
coat polymer films with alumina in web coaters. Surf Coat
Raman-spectrum analysis of tetrahedral amorphous carbon films.
Technol 1996;86–87:776–82.
Appl Phys Lett 2001;78:631–3.
[5] Feichtinger J, Galm R, Walker M, Baumgartner KM, Schulz A,
[23] Papakonstantinou P, Lemoine P, McLaughlin J, MacKay K,
Rauchle E, et al. Plasma polymerized barrier films on membranes
Dodd PM, Pollard RJ, et al. Nanoindentation studies of FeXN
for direct methanol fuel cells. Surf Coat Technol 2001;142–
(X = Ta, Ti) soft magnetic films. J Appl Phys 2000;87:6170–2.
144:181–6.
[24] Lenardi C, Piseri P, Briois V, Bottani CE, Li Bassi A, Milani P.
[6] Borcia G, Brown NMD, Dixon D, McIlhagger R. The effect of an
Near-edge X-ray absorption fine structure and Raman character-
air-dielectric barrier discharge on the surface properties and peel
ization of amorphous and nanostructured carbon films. J Appl
strength of medical packaging materials. Surf Coat Technol
Phys 1999;85:7159–67.
2004;179:70–7.
[25] Gutierrez A, Diaz J, Lopez MF. X-ray absorption spectroscopy
[7] Park KJ, Chin EY. Effect of diamond-like carbon thin film
study of pulsed-laser-evaporated amorphous carbon films. Appl
deposition on the resistance of polycarbonate to radiation-
Phys A 1995;61:111–4.
induced degradation. Polym Degrad Stabil 2000;68:93–6.
[26] Lee CS, Shin JK, Eun KY, Lee KR, Yoon KH. Defect density
[8] Yoshida M, Tanaka T, Watanabe S, Shinohara M, Lee JW,
and atomic bond structure of tetrahedral amorphous carbon (ta-
Takagi T. Improvement of oxygen barrier of PET film with
C) films prepared by filtered vacuum arc process. J Appl Phys
diamond-like carbon film by plasma-source ion implantation. Surf
2004;95:4829–32.
Coat Technol 2003;174–175:1033–7.
[27] Vasquez S, Achete CA, Borges CB, Franceschini Jr DF, Freire
[9] Yoshida M, Watanabe S, Tanaka T, Takagi T, Shinohara M, Lee
FL, Zanghellini E. Structure and properties of a-C:H films
JW. Investigation of diamond-like carbon formed on PET film by
deposited onto polymeric substrates. Diamond Relat Mater
plasma-source ion implantation using C2H2 and CH4. Nucl Instr
1997;6:551–4.
Methods Phys Res B 2003;206:712–6.
[28] Vasquez-Borucki S, Jacob W, Achete CA. Amorphous hydro-
[10] Yoshida M, Tanaka T, Shinohara M, Watanabe S, Lee JW,
genated carbon films as barrier for gas permeation through
Takagi T. Improvement of oxygen barrier of polyethylene
polymer films. Diamond Relat Mater 2000;9:1971–8.
terepthalate film by plasma-source ion implantation of carbon. J
[29] Ong HC, Chang RPH, Baker N, Oliver WC. Improvement of
Vac Sci Technol A 2002;20:1802–7.
mechanical properties of amorphous carbon films deposited on
[11] Li DJ, Cui FZ, Gu HQ. Studies of diamond-like carbon films
polycarbonate plastics. Surf Coat Technol 1997;89:38–46.
coated on PMMA by ion beam assisted deposition. Appl Surf Sci
1999;137:30–7.

You might also like