Journal of Non-Crystalline Solids 386 (2014) 14–18

Contents lists available at ScienceDirect

Journal of Non-Crystalline Solids

journal homepage: www.elsevier.com/ locate/ jnoncrysol

A study of diamond like carbon/chromium films deposited by microwave

plasma activated chemical vapor deposition
Sambita Sahoo a, S.K. Pradhan a,1, M. Jeevitha b, Sharmistha Bagchi a, P.K. Barhai c
a
Advanced Materials Technology, Institute of Minerals and Materials Technology, Bhubaneswar 751013, India
b
Surface Engineering Department, Institute of Minerals and Materials Technology, Bhubaneswar 751013, India
c
Department of Applied Physics, Birla Institute of Technology, Mesra 835215, India

a r t i c l e i n f o a b s t r a c t

Article history: Since metal interlayer improves the adhesion, a sequential method of deposition of diamond like carbon (DLC) on
Received 27 August 2013 Cr has been tried. The Cr interlayer and DLC grown sequentially on Si (100) by pulsed DC magnetron sputtering
Received in revised form 19 November 2013 and microwave plasma CVD (MPCVD), respectively, which resulted in intermixing of DLC and Cr on the upper
Available online 8 December 2013
part of Cr interlayer. The intermixed layer contains nanocrystals of CrC which most likely contributes to formation
of a relatively hard (18 GPa) DLC with high modulus (up to 320 GPa) and improved adhesion.
Keywords:
Diamond like carbon;
© 2013 Elsevier B.V. All rights reserved.
Chromium;
CrC;
Microwave plasma CVD

1. Introduction been used successfully as electroplated coatings for corrosion, fatigue,

and wear applications. To improve the adhesion and tribological prop-
In the last few decades, many different types of coatings have been erties of DLC film, DLC/Cr multilayer films are grown either by
developed keeping in view the industrial requirements for materials sputtering or by combination of sputtering and plasma CVD [13,14].
and coatings with high wear resistance, mechanical toughness, and The film deposition by sputtering and plasma CVD carried out in the
modified coefficient of friction. Over the years PVD and CVD methods same chamber leads to target poisoning due to the high concentration
have developed considerably for this reason, helping to achieve higher of methane in the plasma. To avoid target poisoning and unnecessary
performing applications in tribology, electronics, and biotechnology cleaning of the sputtering chamber, the sequential sputtering of Cr
[1,2]. Diamond like carbon (DLC) deposited by sputtering or by RF layer and then plasma CVD of DLC layer in separate chambers has
plasma CVD is commonly used for coatings of stents, hard disks, blades, been tried.
cutting tools etc. DLC is an amorphous material containing a mixture of Usually it is difficult to grow DLC by conventional microwave plasma
sp3 and sp2 coordinated amorphous hydrogenated carbon atoms which CVD (MPCVD), but there are processing windows where the conditions
show excellent combination of properties such as high hardness, for growth of moderately hard and polymeric DLC on Si substrates are
low friction, high chemical inertness, high optical transparency and possible [15]. In this paper, a study of DLC films deposited on Cr inter-
excellent biocompatibility [3,4]. One of the drawbacks of DLC films is layer by sequential sputtering and MPCVD has been reported. By
that they tend to delaminate once the thickness reaches micrometer using this sequential method, the issue of target poisoning is completely
range, primarily due to built up of high residual stress, thereby limiting avoided and the deposition of lump-free DLC layer is achieved.
their applications. Nanocomposite metal containing DLC films are the
new generation DLC coatings with improved properties, which have
tremendous potential for applications [5,6]. For the synthesis of metal- 2. Experimental
containing DLC (Me–DLC) films (both multilayer/superlattice and
metal particle-composite) many techniques have been used including For this study thin layer of Cr was deposited on Si (100) substrates in
reactive magnetron sputtering [7], hybrid ion beam comprising magne- a pulsed DC sputtering system using a 99.99% pure Cr target. Before film
tron sputtering and ion source [8] and cathodic arc technique [9]. In deposition, the substrates were cleaned ultrasonically in acetone and
Me–DLC, the commonly used metals are the carbide-forming metals deionized water, and then dried with air blower. During film deposition
such as Ti, Cr, V, Ni, Mo, and Co [10,11]. It is known that metallic inter- the working pressure was kept at 3.5 × 10−3 Torr, and the voltage and
layer of Ti and Cr improves adhesion to metal substrates [12], and has current of the pulsed DC source were maintained at 360 V and 0.25 A,
respectively. Gas flow of argon was 12 sccm and the deposition time
E-mail address: [email protected] (S.K. Pradhan). was for 10 min. The thickness and hardness of the Cr film prepared
1
Tel.: +91 674 2584091x218; fax: +91 674 2581637. under the given condition were 60 nm and 3 GPa, respectively.

0022-3093/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jnoncrysol.2013.11.033
 S. Sahoo et al. / Journal of Non-Crystalline Solids 386 (2014) 14–18 15

The DLC film on top of the Cr layer was deposited in an Astex type 1337 cm−1 indicates that more numbers of the graphitic clusters are
MPCVD system using a mixture of CH4 (3 sscm), H2 (3 sccm) and Ar able to “breadth” easily which implies that short chains might be
(93 sccm). The reactor pressure was maintained at 21 Torr and the mi- attached to these clusters [3].
crowave power was set at 300 W. The grown films were characterized
for their structural, morphological and mechanical properties using a 3.2. Film morphology
micro-Raman spectrometer (STR 500, Seki Technotron), a nanoindenta-
tion system (UMIS, Fischer-Cripps), TEM (Technai, FEI) and field emis- The cross-section images of the DLC/Cr films are shown in Fig. 2.
sion scanning electron microscope (FESEM) (Supra 55, Carl Zeiss). The Here it is observed that with increase in the DC bias up to −100 V, thin-
excitation wavelength used for Raman spectrometry was 514 nm ner DLC coatings form, whereas almost no coating forms at − 150 V
(Argon ion laser) having beam size of about 1 μ (at 500×) with power bias. This happens perhaps due to etching by atomic hydrogen and by
of 40 mW. To avoid systematic errors in the data during the nanoinden- high energy ion bombardment leading to re-sputtering (by Ar ions)
tation test, area function calibration and standard tests have been done and ion-etching of the film. Generally, DLC films are deposited by RF
on fused silica before performing tests on the Cr–DLC films. To random plasma CVD where the plasma is struck near the substrate which attains
errors arising from different sample positions are shown in the plots as a negative self-bias, the self-bias developed at the substrate increases
error bars. the energy and flux of hydrocarbon ions (and radicals by collision
with ions) to the substrate, both of the above contribute to the increase
3. Result and discussion in DLC growth rate [18,19]. However, in case of microwave activation,
where the ion energy and ionization are much higher in the plasma, in
3.1. Raman this case the likelihood of etching of the film increases when negative
bias is applied to the substrate because of the increased hydrogen and
The Raman spectrum of the grown DLC/Cr film is shown in Fig. 1. The argon ion flux. As shown in Fig. 2, when bias was applied to the sub-
presence of D band at 1337 cm−1 and G band at 1610 cm−1 confirms strate, initially the film thickness did not change (up to − 50 V) and
the presence of sp2 carbon along with disorder in the graphitic structure then it became thinner and powdery, and finally at − 150 V no film
(as revealed by the breathing mode D band). Further, as reported previ- deposition was noticed.
ously, we have found that these DLC films contain considerable amount Fig. 3 shows the cross-sectional TEM image of DLC/Cr film and the
of hydrogen and are polymeric in nature which was confirmed by FTIR inset shows the corresponding SAED pattern from the interface. The
and transmission spectra [15]. Here it is noticed that the G peak has presence of nanocrystalline particles is evident from the ring patterns
shifted to higher wavenumber while the D peak shift is towards lower of the SAED. The measured d-spacings from the SAED pattern are
side. 2.344 Å, 2.0170 Å and 1.4867 Å corresponding to the (111), (002) and
Usually, high-quality diamond-like carbon films exhibit a rela- (022) reflections of CrC, respectively. Though there are CrC nanoparti-
tively broad symmetrical G-band (1580 cm− 1) and a lesser D-band cles on the interface, there is no sharp interface between the Cr and
(1350 cm−1) signifying a low amount of graphitic clusters. As reported DLC layers indicating the intermixing of DLC with Cr with formation of
by others, when there is a metal interlayer in the DLC film, the G-peak some CrC nanoparticles along it. The formation and distribution of CrC
shifts to lower wavenumber and a slight increase in the D-peak height could be linked with the formation of graphitic nanoclusters in the
is observed [16,17]. In contrast, in our case, the G peak position has DLC. Perhaps, the fine intermixing is the reason for improved adhesion
shifted to the higher wavenumber. In fact, as the G peak is beyond between DLC and Cr layers.
1600 cm−1, then according to established models [3], the only way to
describe the current situation is to believe that the film is composed 3.3. Nanoindentation
of ever smaller graphitic nanoclusters in an amorphous DLC matrix.
This is the best way to account for the G band shift to higher side. More- Nanoindentation results of the DLC/Cr films are shown in Table 1
over, this film contains considerable amount of hydrogen in the matrix and Fig. 4. For these tests, the normal load was chosen so that the mea-
which is bonded with carbon in sp3 configuration. Because the D peak is sured hardness value would not be affected much by the Si substrate.
significant, it proves the ring nature of the sp2 bonded carbon, however The hardness of the DLC/Cr film reached highest value of 18 GPa
due to smaller cluster size the G shifts to higher value even if large which was deposited at low substrate bias (−50 V), whereas films de-
amount of hydrogen is in the matrix, which otherwise should down posited at higher bias were powdery and consequently softer (hardness
shift G band. Further, the down shift of D band from 1350 cm− 1 to 4–7 GPa). The dotted line shows the hardness of 3 GPa for the Cr film
interlayer for all the films. Fig. 5 shows the variation of Young's modulus
(E) with bias voltage. The Young's modulus (E) and hardness (H) were
determined using the Oliver and Pharr method of analysis [20,21]. The
equations used to calculate the reduced modulus Er, E and H are given
as follows:
rffiffiffiffiffiffi
dP π
Er ¼  ;
dh 4A
2
1−v 1−v2i
Er ¼ þ ;
E Ei
P
H¼ ;
A

where, A is the contact area at maximum load, v and vi are the Poisson's
ratio of the film and indenter, respectively, and Ei is the elastic modulus
of the indenter. Accordingly, the calculated E and H values are plotted
versus the depth. The modulus decreased with increase in voltage and
at the highest bias of − 150 V the contribution from DLC is almost
negligible and so the rise in E value reflects the properties that of Cr
Fig. 1. Raman spectra of DLC/Cr film. film on Si. The calculated E value from nanoindentation data for Cr
16 S. Sahoo et al. / Journal of Non-Crystalline Solids 386 (2014) 14–18

Fig. 2. FESEM cross-section images: (a) Cr interlayer, and DLC/Cr deposited at (b) 0 V, (c) −50 V, (d) −100 V, and (e) −150 V substrate bias.

film is 160–170 GPa, which is much less than that of bulk Cr which is loads at which failures appeared, i.e. the critical loads are used to com-
277 GPa. In nanoindentation measurement, the modulus value is dom- pare the adhesive properties of coatings. Post scratch optical microscopy
inated by contribution from the material having low E value, so in the images of scratch lines on the DLC/Cr coatings are shown in Fig. 6. The
case of Cr film it is dominated by Si data, whereas for DLC/Cr films the critical load was highest, about 30 mN, for film deposited without any
data is dominated by DLC. However, in the present case the measured bias. At a bias of − 50 V the critical load decreased to 23 mN, and at
modulus values are much higher (up to 320 GPa), therefore it seems higher bias, the critical load recorded was about 13–14 mN. As the
that the presence of CrC, which has modulus of 380 GPa, affects the films were powdery at higher bias, the scratch tracks actually show
overall modulus of the DLC film. the scratching of the underlying Cr layer due to plastic deformation.
The sharper tip (5 μm diameter) of nanoscratch test gives lower critical
3.4. Nanoscratch test load values (in mN), while it has been reported that the microscratch
test with standard tip size (200 μm) gives higher critical load values
Adhesion property of the films was examined by the nanoindenta- (in N) for DLC films. When the tip sizes are different, the stress
tion system with the nanoscratch option (UMIS, Fischer-Cripps) using
a spherical stylus of 5 μm diameter. Progressive load scratch test was
considered for our test, wherein the load progressively increased from
1 to 50 mN at the rate of 0.5 mN/s. Four tests were done on each sample Table 1
Mechanical properties of DLC film grown on 60 nm Cr interlayer.
in order to determine the critical load and the coefficient of friction. The
Substrate bias Thickness Hardness Critical load Coefficient of
(V) (nm) (GPa) (mN) friction

0 37 15 ± 0.5 30 ± 1 0.15 ± 0.02

−50 42 19 ± 0.6 23 ± 0.8 0.12 ± 0.02
−100 60 4 ± 0.5 14 ± 0.5 0.11 ± 0.03
−150 15 7 ± 0.4 13 ± 0.5 0.06 ± 0.02

Fig. 3. Cross-sectional TEM micrograph and SAED pattern of DLC/Cr film. Fig. 4. Nanoindentation hardness of DLC/Cr films deposited at different substrate bias.
 S. Sahoo et al. / Journal of Non-Crystalline Solids 386 (2014) 14–18 17

Fig. 5. Young's modulus DLC/Cr films deposited at different substrate bias. Fig. 7. Hardness versus critical load plot of DLC/Cr films.

distribution differs on loading. Sharper tip cracks the film earlier than correlation is not observed, however as shown in Figs. 6 and 7, it is
the blunt tip. Relatively, the observed critical load of the films is quite seen that hardness and critical load of these films are interrelated. The
high, which shows better adhesion of the DLC films on the Cr interlayer. coefficient of friction (COF) for all the DLC/Cr films is very low (less
A thorough study has been done by Li et al. on the effect of the thick- than 0.15), indicating the lubricating nature of the films.
ness and scratch performance of thin amorphous carbon films using a Only a few papers have reported the complete mechanical proper-
nanoindenter with a nanoscratch attachment [22]. They found that ties of Cr–DLC films. Table 2 shows a comparison of Cr–DLC films
the critical load increased with increasing coating thickness. The thicker grown by various methods mainly presenting the hardness studies.
films exhibited better scratch performance than the thinner ones due to Most of them are complex multilayers or nanocomposite of Cr–DLC
better load-carrying capability of the thicker films. In our case, such a films unlike the simple multilayered Cr–DLC film presented in this
paper with moderately high hardness, good adhesion and low coeffi-
cient of friction. Our experimental method, having much simple vacu-
um design of the deposition system, can be used to deposit highly
dense and scratch resistant anticorrosive Cr–DLC coatings which may
be developed into an industrially feasible method.

4. Conclusions

The DLC/Cr coatings grown by sequential deposition with sputtering

and MPCVD have moderate hardness with high modulus and show im-
proved adhesion. It seems that DLC with the presence of nanocrystalline
graphic cluster in it helps the formation of nanocrystalline CrC particles
at the interlayer. The enhanced modulus and adhesion properties of the
films can be attributed to the presence of CrC at the interface.

Acknowledgment

The authors would like to thank the Director, IMMT Bhubaneswar

for giving permission to publish this work. Financial support for this
work by DST, Govt. of India, New Delhi is acknowledged.

Table 2
Mechanical properties of Cr–DLC films deposited by various methods [23–28].

Method of deposition Hardness Critical load

Cr/a-C multilayered nanocomposite synthesized 9–11 GPa

by e-beam physical vapor deposition
DLC with chromium interlayer synthesized 16–18 GPa 90–260 mN by
by PECVD microscratch tester
Cr-containing DLC films using a hybrid 18–20 GPa
plasma-assisted CVD/PVD process
(DLC:Cr:Cu) films by cathodic arc 7–16 GPa
evaporation process
Cr–DLC films deposited by DC magnetron ~12 GPa
sputtering unit and a linear ion source
Cr–C:H films deposited by a reactive magnetron 4–10 GPa
sputtering technique
Fig. 6. Optical images of scratch profiles of DLC/Cr samples after scratch testing.
18 S. Sahoo et al. / Journal of Non-Crystalline Solids 386 (2014) 14–18

References [13] Z.Q. Qi, E.I. Meletis, Thin Solid Films 479 (2005) 174–181.
[14] W. Dai, G. Wu, A. Wang, Diam. Relat. Mater. 19 (2010) 1307–1315.
[1] K.L. Choy, Prog. Mater. Sci. 48 (2003) 57–170. [15] S. Sahoo, S.S. Pradhan, V. Bhavanasi, S.K. Pradhan, Surf. Coat. Technol. 204 (2010)
[2] S.K. Pradhan, C. Nouveau, A. Vasin, M.A. Djouadi, Surf. Coat. Technol. 200 (2005) 2817–2821.
141–145. [16] B.K. Tay, P. Zhang, Thin Solid Films 420 (2002) 177–184.
[3] J. Robertson, Mater. Sci. Eng. R 37 (2002) 129–281. [17] V.V. Uglova, V.M. Anishchik, Y. Pauleau, A.K. Kuleshov, F. Thièry, J. Pelletier, S.N. Dub,
[4] S. Neuville, A. Matthews, Thin Solid Films 515 (2007) 6619–6653. D.P. Rusalsky, Vacuum 70 (2003) 181–185.
[5] V. Singh, J.C. Jiang, E.I. Meletis, Thin Solid Films 489 (2005) 150–158. [18] S.J. Yu, Z.F. Ding, J. Xu, J.L. Zhang, T.C. Ma, Thin Solid Films 390 (2001) 98–103.
[6] R.J. Narayan, H. Abernathy, L. Riester, C.J. Berry, R. Brigmon, J. Mater. Eng. Perform. [19] W. Dai, H. Zheng, G. Wu, A. Wang, Vacuum 85 (2010) 231–235.
14 (2005) 435–440. [20] A.C. Fischer-Cripps, Nanoindentation, 2nd ed. Springer, New York, 2004.
[7] C. Corbella, E. Bertran, M.C. Polo, E. Pascual, J.L. Andujar, Diam. Relat. Mater. 16 [21] W.C. Oliver, G.M. Pharr, J. Mater. Res. 7 (1992) 1564–1583.
(2007) 1828–1834. [22] X. Li, B. Bhushan, J. Mater. Res. 14 (1999) 2328–2337.
[8] A.Y. Wang, K.R. Lee, J.P. Ahn, J.H. Han, Carbon 44 (2006) 1826–1832. [23] Z.Q. Qia, E.I. Meletis, Thin Solid Films 479 (2005) 174–181.
[9] Y.Y. Chang, D.Y. Wang, Surf. Coat. Technol. 200 (2006) 3170–3174. [24] Chehung Wei, Jui-Ying Yen, Diam. Relat. Mater. 16 (2007) 1325–1330.
[10] M. Weber, K. Bewilogua, H. Thomsen, R. Wittorf, Surf. Coat. Technol. 201 (2006) [25] Sunil K. Pal, Jiechao Jiang, E.I. Meletis, Surf. Coat. Technol. 201 (2007) 7917–7923.
1576–1582. [26] Jui Yun Jao, Sheng Han, Li Shen Chang, Ya-Chi Chen, Chi-Lung Chang, Han Chang
[11] D.Y. Wang, Y.Y. Changa, C.L. Changa, Y.W. Huang, Surf. Coat. Technol. 200 (2005) Shih, Diam. Relat. Mater. 18 (2009) 368–373.
2175–2180. [27] Wei Dai, Peiling Ke, Aiying Wang, Vacuum 85 (2011) 792–797.
[12] C. Wei, J.Y. Yen, Diam. Relat. Mater. 16 (2007) 1325–1330. [28] Wei Daia, Guosong Wua, Aiying Wang, Appl. Surf. Sci. 257 (2010) 244–248.