Journal of Physics and Chemistry of Solids 73 (2012) 308–316

Contents lists available at SciVerse ScienceDirect

Journal of Physics and Chemistry of Solids

0022-36

doi:10.1

n Corr

E-m

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

Investigation of properties of Cu containing DLC films produced byPECVD process

Neeraj Dwivedi a,b, Sushil Kumar a,n, Hitendra K. Malik b, C. Sreekumar a, Saurabh Dayal a,C.M.S. Rauthan a, O.S. Panwar a

a Physics of Energy Harvesting Division, National Physical Laboratory (CSIR), K.S. Krishnan Road, New Delhi 110012, Indiab Department of Physics, Indian Institute of Technology Delhi, New Delhi 110016, India

a r t i c l e i n f o

Article history:

Received 1 June 2011

Received in revised form

12 August 2011

Accepted 13 October 2011Available online 21 October 2011

Keywords:

A. Thin films

B. Plasma deposition

C. Infrared spectroscopy

D. Electrical properties

D. Mechanical properties

97/$ - see front matter & 2011 Elsevier Ltd. A

016/j.jpcs.2011.10.019

esponding author. Tel.: þ91 11 45608650; fa

ail address: skumar@nplindia.org (S. Kumar).

a b s t r a c t

Copper containing diamond like carbon (Cu-DLC) thin films were deposited on various substrates at a

base pressure of 1�10�3 Torr using a hybrid system involving DC-sputtering and radio frequency-

plasma enhanced chemical vapor deposition (RF-PECVD) techniques. The compressive residual stresses

of these films were found to be considerably lower, varying between 0.7 and 0.94 GPa and Cu

incorporation in these films improve their conductivity significantly. Their structural properties were

studied by Raman spectroscopy, atomic force microscopy, scanning electron microscopy, X-ray

photoelectron spectroscopy and X-ray diffraction techniques that clearly revealed the presence of Cu

in the DLC structure. Raman analysis yields that Cu incorporation in DLC enhances the graphite-like sp2

bonding. However, the sp2 bonding was found to continuously reduce with the increasing C2H2 gas

pressure, this may be due to reduction of Cu nanocrystal at the higher pressure. FTIR results inferred

various bonding states of carbon with carbon, hydrogen and oxygen. In addition, hydrogen content and

sp3 and sp2 fractions in different Cu-DLC films were also estimated by FTIR spectra and were correlated

with stress, electrical, optical and nano-mechanical properties of Cu-DLC films. The effect of indentation

load (4–10 mN) on nano-mechanical properties of these films was also explored.

& 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Diamond like carbon (DLC) or hard hydrogenated amorphouscarbon (a-C:H) is recognized as a promising alternative material forvarious applications such as cutting tools, wear and hard disk due toits high hardness, low friction coefficient, high thermal conductivityand very high optical transparency in infrared (IR) region [1–7]. Theability to tune various hybridizations states of carbon in DLC bychanging process parameters has led recent research into thismaterial for exploring explore its electronic and photovoltaic proper-ties [8–11]. However, poor adhesion of DLC films with the substratedue to high compressive residual stress restricts its widespreadapplications [12,13]. It should be noted that the adhesion of DLCfilms with substrates can be improved by doping of silicon, nitrogenand fluorine (Si, N, F), though at the expense of some of their otherproperties [14–16]. In contrast, the incorporation of metals into DLCfilms is an alternative method that improves the adhesion of thesefilms with substrate without affecting their other properties [17–19].Moreover, lower conductivity in DLC films, which restricts itspotential electronic application, can also be resolved by incorporating

ll rights reserved.

x: þ91 11 45609310.

metal in DLC matrix (as its incorporation in DLC structure mayimprove the transport properties). Although some reports pertainingto decrease of residual stress with metal addition have already beeninvestigated, there is a lack of literature on the enhancement oftransport properties with incorporation of Cu in DLC film.

In this paper, the reduction in stress and improvement intransport properties of DLC films due to incorporation of Cu inDLC matrix have been explored. In addition, measured conduc-tivity of different Cu-DLC films are correlated with hydrogencontent and C–H based sp3 and sp2 fractions. Their structural,morphological and nano-mechanical properties have also beeninvestigated by X-ray photoelectron microscopy (XPS), Ramanspectroscopy, X-ray diffraction (XRD), Fourier transform infraredspectroscopy (FTIR), scanning electron microscopy (SEM), atomicforce microscopy (AFM) and nanoindentation measurementswere performed.

2. Experimental details

Cu-DLC films were deposited on various substrates usinga combined DC sputtering and RF-PECVD hybrid system. Theschematic representation of the deposition unit used for thegrowth of Cu-DLC films is shown in Fig. 1. Roots and rotary pumps

DCPower

Vacuum Pumps

Sputtering Gun

CuTarget

Plasma

Ar & C 2H2Gases

RF (13.56 MHz) Power

Substrate

Electrode

Fig. 1. Schematic representation of deposition unit used for the growth of Cu-

DLC films.

Table 1Process parameters for the growth of Cu-DLC films.

Samples Ar pre.

(mTorr)

DC power

(W)

C2H2 pre.

(mTorr)

Negative self

bias (V)

Thickness

(nm)

A 150 8 25 100 151

B 150 8 50 100 159

C 150 8 75 100 168

N. Dwivedi et al. / Journal of Physics and Chemistry of Solids 73 (2012) 308–316 309

were used to obtain base vacuum of 1�10�3 Torr. Sputtering gun,having Cu target and attached with DC power supply, was fixed onthe top of process chamber. The RF (13.56 MHz) power supply wasattached to the substrate electrode. Argon (Ar) and acetylene(C2H2) were used as working gases. The Cu target of 2 in. diameterwas kept 5 cm apart from the substrate to sputter Cu by energeticAr ions. The substrates used in the present study were n-typesilicon wafer /100S, both side polished p-type silicon wafer,corning 7059 glass and stainless steel. The substrates were cleanedin Ar plasma for 10 min prior to the film deposition for avoidingmoisture effect from the surface. Cu-DLC films were deposited atfixed negative self bias of 100 V with varying C2H2 gas pressures of25, 50 and 75 mTorr (samples A, B and C, respectively). DC powerof 8 W and Ar gas pressure of 150 mTorr were kept fixed in all theruns and were applied to sputtering unit for sputtering of Cuparticles. Deposition parameter used for the growth of Cu-DLCfilms is summarized in Table 1.

Thickness of Cu-DLC films were estimated by Taylor–Hobsontalystep instrument. Residual stresses in these films deposited onsilicon wafer were determined from the change in the radius ofcurvature of the wafer, before and after deposition, using 500TCtemperature controlled film stress measurement system (M/s FSMFrontier Semiconductor, USA). The substrate curvature method gen-erally relies on the Stoney formula relating the film average stress (S)to the substrate curvature under the assumption that the film ismuch thinner than the underlying substrate. The Stoney formula,which was used to estimate S is given in Eq. (1)

S¼Esds

2

6ð1�nsÞdf

1

Rf�

1

R0

� �ð1Þ

where Es, ns, df and ds are Elastic modulus, Poisson ratio, thicknessof the film and the thickness of the substrate, respectively, and R0

and Rf are the radii of substrate curvature before and after filmdeposition, respectively. Temperature dependent conductivitymeasurements were carried out using Keithley 610C solid state

electro-meter in coplanar configuration with gap of 0.078 cmwhere aluminum was evaporated in a vacuum better than10�5 Torr. AFM measurements were performed by NanoscopeVeeco V (USA) instruments to analyze surface morphology andsurface roughness of the films. Morphologies of the depositedfilms were studied also by SEM, model: LEO Electron Microscope7060. An XPS spectrum was obtained from Perkin Elmer 1257instrument by X-ray radiation of MgKa of 1253.6 eV to analyzethe composition and structural properties of the films. XRDmeasurement in the 2y of 20–601 was carried out by RigakuMiniflex II diffractometer to investigate the crystallinity andphase composition of the films. Transmission measurement wascarried out in the UV and near IR range by Shimadzu UV-Vis 1601instrument. FTIR measurement was carried out by Perkin ElmerSpectrum Bx instrument in the wave number between 2700 and3700 cm�1. Nanoindentation measurement was performed byusing IBIS nanoindentation, Fisher-Cripps Laboratories Pvt. Lim-ited, Australia under applying maximum indentation loads of4, 5 and 10 mN.

3. Results and discussion

3.1. Film thickness and deposition rate

Thickness of Cu-DLC films grown at C2H2 gas pressures of 25, 50and 75 mTorr was found to be 151, 159 and 168 nm, respectively.Since depositions were performed for different timings at differentC2H2 gas pressures, the deposition rates were also calculated inorder to clarify the role of C2H2 gas pressures on thickness. Thedeposition rates for Cu-DLC films at different C2H2 gas pressures of25, 50 and 75 mTorr were found to be 1.21 nm/s, 1.77 nm/s and2.3 nm/s, respectively. Therefore, enhanced C2H2 gas pressuresleads to higher deposition rates.

3.2. AFM and SEM analyses

Atomic force microscopy (AFM) and scanning electron microscopy(SEM) analyses were carried out to investigate the surface morphol-ogy of the films. Typical AFM and SEM micrographs of Cu-DLC filmdeposited at C2H2 gas pressure of 25 mTorr are shown in Fig. 2(a) and(b), respectively. AFM micrograph clearly shows the spreading ofnanocrystalline Cu particles in amorphous DLC matrix. Pauleau et al.[20] and Gerhard et al. [21,22] have also reported the Cu nanocrystalformation in DLC matrix. The mean roughness Ra of Cu-DLC filmdeposited at C2H2 gas pressure of 25 mTorr was found to be as low as0.22 nm. Observed SEM micrograph also followed similar surfacemorphology and showed the presence of nanocrystalline Cu particlesin amorphous DLC matrix. Recently, Mousinho et al. [23] have alsoreported the growth of nanocrystalline structure in amorphous DLCmatrix. However, instead of incorporation of foreign elements in DLCfilms they performed some pretreatments on substrates and modifiedtheir surface topography by diamond, graphite powders and wetchemical, plasma etchings before the DLC deposition. After depositionthey obtained different nanocrystalline DLC films whose surfacefeature was dependent on the pretreated surface features of thesubstrate. They had observed high sp2 bonding with graphite powdermodified substrate, though in contrast we obtained high sp2 bondingby Cu incorporation.

3.3. XPS analysis

X-ray photoelectron spectroscopy (XPS) is a very reliable techni-que to investigate chemical composition as well as bonding of thestructures. XPS spectrum was recorded for Cu-DLC film deposited atC2H2 gas pressure of 25 mTorr (sample A). Three different core level

25 mTorr

25 mTorr

2.5

2.0

1.5

1.0

0.512.0 nm0.0 nm

2.52.0

1.51.0

0.5

0.5

1.0

1.5

2.0

2.5

0.51.0

1.52.0

2.5

μm

μm

μm

μm

Fig. 2. (a) Typical AFM image and (b) SEM picture of Cu-DLC film deposited at

C2H2 gas pressure of 25 mTorr.

N. Dwivedi et al. / Journal of Physics and Chemistry of Solids 73 (2012) 308–316310

spectra such as C 1s, Cu 2p and O 1s are shown in Fig. 3(a), (b) and (c),respectively. C 1s core level spectrum shows sharp peak at 283.8 eV.Comparing with standard result [24] (where C 1s peak is obtained at284.6 eV), a nominal shifting of C 1s peak (�0.8 eV) towards lowerbinding energy side was observed in the present study. This shift maybe due to Cu particles because its incorporation can enhance thegraphite-like sp2 bonding and in XPS spectra the binding energy ofsp2C bonding is always lower than sp3C bonding. Ji et al. [25] alsosuggested that metal incorporation in DLC matrix enhances thegraphite-like sp2C bonding. The Cu 2p core level spectrum in thebinding energy range of 925 to 975 eV is also presented in Fig. 3(b),which shows the incorporation of Cu particles in the Cu-DLC film.Since deposition of Cu-DLC films was carried out at higher basepressure (1�10�3 Torr), CuO peaks were found instead of pure Cupeak, that were centered at 934.1 eV and 954.1 eV. It is to be notedgenerally pure Cu peak is observed at 932.4 eV. The possibility ofexistence of Cu2O was also found to be very less because the bindingenergy of Cu2O is lower than Cu and CuO [26,27]. The O 1s, core levelspectrum is shown in Fig. 3(c), where the observed peak at 530.5 eVcorresponds to the presence of oxygen; that arises due to high basepressure growth of Cu-DLC films. Some of oxygen may also be due tothe fact that prior to loading of sample in XPS measurement chamber,it was exposed in the ambient air.

3.4. XRD analysis

Typical X-ray diffraction (XRD) spectrum for Cu-DLC filmdeposited at C2H2 gas pressure of 25 mTorr is shown in Fig. 4. Itis evident from the spectrum that except peaks obtained at 2y of

33.11, 42.41 and 51.41 rest of the structure of the films was foundto be amorphous in nature. An intense peak found at 2y value of33.11 represents the Si substrate /200S [28]. A broad hump likeshape obtained in the 2y range of 251 to 401 may be due to theamorphous nature of carbon. Observed hump between 251 and401 may also be due to Compton modified scattering, which arisesdue to the interaction of high energy X-rays with lower atomicnumber (Z¼6) carbon elements [29]. Observed small peak in thespectrum at 2y value of 42.41 represents the CuO /111S [30].However, intensity of this peak was found to be low that may bedue to the presence of low Cu concentration. A peak observed at2y value of 51.41 may be due to the oxidation state of Cu. Anotherpossible explanation is that there may be a possibility of peakshifting as compared to bulk material [31].

3.5. Fourier transform infrared spectroscopy (FTIR) studies

Fig. 5 shows the FTIR spectra recorded for Cu-DLC films grownon both side polished p-type silicon wafer at different C2H2 gaspressures of 25, 50 and 75 mTorr. It is important to study the C–Hbonding in the carbon network, since it affects various propertiesof DLC films. In this context, it is evident from the spectra that Cuincorporation in DLC matrix did not distort various C–H bondings.Titantah et al. [32] suggested that the presence of certain amountof hydrogen is also very essential to obtain diamond-like carbonfilm. Hydrogen stabilizes the diamond-like structure by passivatingdangling bond by maintaining sp3 content. FTIR spectra were usedfor the estimation of hydrogen concentration in all the depositedCu-DLC films. Since FTIR shows bonded hydrogen with carbon,therefore estimated hydrogen concentration may refer as boundhydrogen concentration (BHC). The BHC was calculated by employ-ing the following relation [33]

BHC½ � ¼ AsðCHxÞ

Z aðkÞk

dk ð2Þ

where a(k) is the absorption coefficient at wave number k, As(CHx)is the proportionality constant. We used As(CHx)¼1�1021 cm�2

for estimation of hydrogen concentration [33–35]. Recently, Singhet al. [33] have used the FTIR analysis for the estimation ofhydrogen concentration in DLC films. Further, the area between2700 and 3100 cm�1 was considered to evaluate hydrogen con-centration. The BHC was found to enhance with the increasing C2H2

gas pressures. The values of BHC in Cu-DLC films grown at C2H2 gaspressures of 25, 50 and 75 mTorr were found to be 26.5�1021,28�1021 and 29.5�1021 cm�3, respectively. Such an amount ofhydrogen seems to be sufficient to obtain diamond-like propertiesfrom Cu-DLC films. Estimated BHC was correlated with conductiv-ity results.

In DLC films there are two sources of sp3 and sp2 bonding, oneis due to ions and another is due to C–H bonding. The ion basedsp3 and sp2 bonding and their correlation with electrical, opticaland mechanical properties of DLC films have been exploredextensively. However, correlation of various properties of DLCfilms with C–H based sp3 and sp2 bonding has not been investi-gated in greater detail. However, recently Paul et al. [36] dis-cussed the role of C–H based sp3 and sp2 bonding. Therefore, inpresent study C–H based sp3 and sp2 bonding were also estimatedand correlated with electrical, optical and mechanical propertiesof Cu-DLC films. Deconvoluted FTIR spectra of Cu-DLC filmsdeposited at C2H2 gas pressures of 25, 50 and 75 mTorr is shownin Fig. 6(a), (b) and (c), respectively. In each spectrum 6 peakswere fitted (fitting was 99.9%) by considering Gaussian compo-nents. The film deposited at 25 mTorr show various peaks at2865.5, 2898, 2920.7, 2936.3, 2954.3 and 2971.5 cm�1, which asrepresented by 1, 2, 3, 4, 5 and 6, respectively correspond tosp3CH3 (sym), sp3CH, sp3CH2 (asym), sp3CH2 (asym), sp3CH3

275 280 285 290 295

25 mTorr 283. 8 eVC 1s

Inte

nsity

(a.u

.)

Binding Energy (eV)

25 mTorr

954.1 eV

934.1 eV

Cu 2p

Inte

nsity

(a.u

.)

Binding Energy (eV)930 945 960 975

520 525 530 535 540

25 mTorrO 1s

Inte

nsity

(a.u

.)

Binding Energy (eV)

Fig. 3. Core level XPS spectra of (a) C 1s, (b) Cu 2p and (c) O 1s of Cu-DLC film deposited at C2H2 gas pressure of 25 mTorr.

2600 2800 3000 3200 3400 3600 3800

Cu-DLC

75 mTorr

50 mTorr

25 mTorr

Inte

nsity

(a.u

.)

Wave Number (cm-1)

Fig. 5. FTIR spectra in the range of 2700–3700 cm�1 of Cu-DLC films deposited at

C2H2 gas pressure of 25, 50 and 75 mTorr.

15 20 25 30 35 40 45 50 55 60

Cu-DLC25 mTorr

Inte

nsity

(a.u

.)

2θ (Degree)

Fig. 4. Typical XRD spectrum of Cu-DLC film deposited at C2H2 gas pressure of

25 mTorr.

N. Dwivedi et al. / Journal of Physics and Chemistry of Solids 73 (2012) 308–316 311

2800 2850 2900 2950 3000 3050

65

4321

75 mTorr

Inte

nsity

(a.u

.)

Wave Number (cm-1)

5o mTorr

65

4

3

2

1

25 mTorr

6

5

432

1

Fig. 6. Deconvoluted FTIR spectra of Cu-DLC films grown at C2H2 gas pressures of

(a) 25, (b) 50 and (c) 75 mTorr.

800 1200 1600 2000

G

D

Cu-DLC75 mTorr

Inte

nsity

(a.u

.)Raman Shift (cm-1)

G

D

Cu-DLC50 mTorr

GD

Cu-DLC25 mTorr

Fig. 7. Raman spectra of Cu-DLC films deposited at C2H2 gas pressures of 25, 50

and 75 mTorr.

N. Dwivedi et al. / Journal of Physics and Chemistry of Solids 73 (2012) 308–316312

(asym) and sp2CH (olef), respectively [37–40]. However, positionsof these peaks were changed by varying the C2H2 gas pressures to50 and 75 mTorr. Although estimation of % sp3 and sp2 fractionsby FTIR analysis was found to be a subject of debate as it showsC–H based bonding. Since certain amount of hydrogen is essen-tially required to stabilize diamond-like structure [32], C–H basedsp3 and sp2 fractions may also have great importance [36].Walters et al. [40] have extensively discussed the role of sp3

and sp2 fractions estimated by FTIR spectra. Sharon et al. [41]have also used FTIR spectra for calculation of sp3 and sp2

fractions. The estimation of C–H based % sp3 and sp2 fractionswere performed by considering peak area ratio method. Thus

% sp3 ¼Area of sp3 CHn bonding

Total area of all peaks� 100

% sp2 ¼Area of sp2 CHn bonding

Total area of all peaks� 100

where n¼1, 2, 3.The % sp3 fraction was found to continuously increase with the

increasing C2H2 gas pressure. The values of sp3 fraction for Cu-DLCfilms grown at different C2H2 gas pressure of 25, 50 and 75 mTorrwere found to be 66.9, 69.7 and 81.7%, respectively. In contrast, sp2

fraction was found to continuously decrease with the increasingC2H2 gas pressures from 25 to 75 mTorr. The values of sp2 fractionat 25, 50 and 75 mTorr were found to be 33.1, 30.3 and 18.3%,respectively. Observed increase in % sp3 fraction with the enhancedC2H2 gas pressure may be due to the increase in hydrogenconcentrations in Cu-DLC films. Measured C–H based sp3 and sp2

fractions were further correlated with conductivity.

3.6. Raman analysis

Raman shift of Cu-DLC films was analyzed to investigate thestructural behavior of these films grown under varied C2H2 gaspressure from 25 mTorr to 75 mTorr and the corresponding spectraare shown in Fig. 7(a)–(c). The main feature of Raman spectra of

DLC is the origination of D and G peaks. The G stands for graphiteand this peak is commonly observed at �1550 cm�1 due to highlyordered pyrolitic graphite, whereas D stands for disorder and thispeak is generally obtained at �1350 cm�1 because of disorder inthe graphite-like sp2 bonding. The square shaped peak obtainedat�960 cm�1 is the Raman second order peak of substrate Si. Thispeak is mostly observed for the films with high sp3 fraction [42].Thus, this peak also reveals the reduction of sp2 fraction withincrease in C2H2 gas pressure. Further, depending upon structuralrearrangement, the D and G peaks are deviated from their usualpositions. This can be seen from the figures that an enhancement inC2H2 gas pressure shifts the position of broad peak towards lowerwave numbers, as indicated by dotted line. The shifting of this peaktowards lower wave number clearly confirms the reduction in sp2

fraction of the film that may be due to the reduction of nanocrys-talline Cu particles into the film (as metal incorporation generallyenhances the sp2 bonding). Gerhard et al. [21,22] and Chen andHong [43] had also observed same behavior of Raman shift incopper containing DLC films. Further, the peak fitting of Ramanspectra (not given here) was performed to investigate the varia-tions of D and G peaks and the ID/IG ratios with changing C2H2 gaspressure from 25 to 75 mTorr. The G peaks at C2H2 gas pressures of25, 50 and 75 mTorr were found to be at 1569, 1558.7 and1553.4 cm�1, respectively, whereas the D peaks were located at1417, 1395.8 and 1386.4 cm�1, respectively, at the same pressure.The observed ID/IG ratios also followed the similar trend and gotcontinuously decreased with the increasing C2H2 gas pressure from25 to 75 mTorr. The values of ID/IG ratios at 25, 50 and 75 mTorrwere found to be 1.2, 0.76 and 0.65, respectively.

3.7. Residual stress

The high stress in the DLC films caused by distortion in the bondangle and bond length, leads to their poor adhesion to thesubstrate [44]. However, incorporation of Cu in DLC matrix wasfound to be an appropriate approach to minimize the stress.Fig. 8 shows the variation of residual stress as a function of C2H2

25 50 750.6

0.7

0.8

0.9

1.0

Cu-DLC

C2H2 Gas Pressure (mTorr)

Res

idua

l Stre

ss (G

Pa)

Fig. 8. Variation of residual stress with C2H2 gas pressure for different Cu-

DLC films.

10-11

1x10-10

1x10-9

1x10-8

1x10-7

1x10-6

1x10-5

Con

duct

ivity

(Ω-1

cm-1

)

1000/T (K-1)

Cu-DLC (25 mTorr) Cu-DLC (50 mTorr) Cu-DLC (75 mTorr) DLC (25 mTorr) DLC (50 mTorr) DLC (75 mTorr)

4.0x10-9

6.0x10-9

8.0x10-9

2.6 2.7 2.8 2.9 3.0 3.1 3.2 3.3 3.4

26.5 27.0 27.5 28.0 28.5 29.0 29.50.0

8.0x10-7

1.6x10-6

2.4x10-6

Con

duct

ivity

(Ω-1

cm-1

)

At 300 K

Cu-DLC

Con

duct

ivity

(Ω-1

cm-1

)

BHC (x1021 cm-3)

At 373 K

Fig. 9. (a) Variation of conductivity with inverse of temperature of pure DLC and

Cu-DLC films deposited at C2H2 gas pressures of 25, 50 and 75 mTorr, (b) variation

of conductivity at 300 K and 373 K with BHC for different Cu-DLC films.

N. Dwivedi et al. / Journal of Physics and Chemistry of Solids 73 (2012) 308–316 313

gas pressure for Cu-DLC films. The residual stress was foundcompressive in nature, which gets increased with C2H2 gaspressure. The values of residual stress for these films depositedat C2H2 gas pressure of 25, 50 and 75 mTorr were found to be0.7, 0.77 and 0.94 GPa, respectively. Hence it is inferred that thereis a nominal increase in residual stress with the increase in C2H2

gas pressure. This may be due to the fact that increased C2H2 gaspressure reduces the Cu concentration and hence minimizes thesp2 bonding. The Raman results also support the said statementbecause at C2H2 gas pressure of 25 mTorr the G peak position wasat higher wave number and the ID/IG ratio was significantly high.However, the G peak position was shifted to the lower wavenumber and the ID/IG was reduced with increasing C2H2 gaspressure to 75 mTorr. The reduction in Cu concentration at highC2H2 gas pressure was due to the increased density of C atoms aswell as target poison problem as suggested by Pauleau et al. [20]. Itis worth noting that Cu-DLC films exhibit extremely low residualstress (o1 GPa) than pure DLC films (which generally show �2–4 GPa). Chen and Hong [43] have suggested that weak bondingbetween Cu and carbon (C) that allows sliding at the grain (Cu)–matrix (C) interface may also be a possible mechanism of reducedstress in these films. Stress in Cu-DLC films was also correlatedwith C–H based sp3 and sp2 bonding. This can be seen that stresscontinuously increases with increasing % sp3 fraction. Wang et al.[45] have observed a reduction in stress by metal incorporation inDLC. Recently, we have also observed low residual stress byemploying Cu/a-C:H multilayer structure [26].

3.8. Temperature dependent conductivity (TDC) studies

Transport properties of DLC films are governed by the presenceof graphite like sp2 clusters, which are associated with p–pn weakVander wall’s bonding [5,46,47]. DLC films contain both diamondlike sp3 as well as graphite like sp2 sites and they show very lowconductivity [46,48]. We have estimated the dark conductivity ofCu-DLC films above room temperature using the following expres-sion

s¼ s0e�DE=kT ð3Þ

where s0 is the conductivity pre-exponential factor DE is theactivation energy, k is the Boltzmann constant and T is the workingtemperature (in Kelvin).

The typical variation of dark conductivity as a function oftemperature for Cu-DLC and pure DLC films grown on corning7049 glass substrate is shown in Fig. 9(a). The conductivity at thetemperatures of 300 and 373 K of Cu-DLC film deposited at25 mTorr was found to be 8.9�10�9 and 2.6�10�6 O�1 cm�1,respectively. When the pressure was increased to 50 mTorr, theconductivity got changed to 7.6�10�9 and 6.8�10�7 O�1 cm�1

at the same temperatures. The conductivity of Cu-DLC filmdeposited at 75 mTorr was found to be 4.3�10�9 and4.04�10�7 O�1 cm�1at 300 and 373 K, respectively. In order tovisualize the role of Cu particles on the conductivity of DLC films,some pure DLC films were also grown at identical C2H2 gaspressures and then characterized for conductivity, as also shownin Fig. 9(a). The values of conductivity of DLC film grown at25 mTorr was found to be 4.1�10�10 and 6.6�10�9 O�1 cm�1

at 300 and 373 K, respectively, which changed to 5.7�10�11 and5.3�10�10 O�1 cm�1 at same temperatures for the film depositedat 50 mTorr. However, conductivity at 300 and 373 K of DLC filmgrown at 75 mTorr was found to be 5.0�10�11 and8.7�10�11 O�1 cm�1, respectively. This is to be noted that theconductivity of both Cu-DLC and DLC films was decreased withincreasing C2H2 gas pressure but the magnitude of conductivity inCu-DLC films was higher than DLC films. The conductivity at 373 Kof Cu-DLC film grown at 25, 50 and 75 mTorr was about threeorders higher than that of conductivity at 373 K of DLC film grownat the same C2H2 gas pressures. Observed remarkable results ofconductivity clearly reveal that Cu played vital role for theimprovement of conductivity of DLC films. Further, the reduction

N. Dwivedi et al. / Journal of Physics and Chemistry of Solids 73 (2012) 308–316314

in conductivity with enhanced C2H2 gas pressure in Cu-DLC films isattributed to the target poisoning. When C2H2 gas pressure isincreased, carbon species in the Ar–C plasma also get increasedwhich in turn adhere over the entire Cu target in a very short timeand subsequently hinder the ejection of Cu particles from thetarget. Thus, it is realized that ejection of Cu particles from thetarget took place for very short time. Similar target poisoningproblem has been previously reported by Pauleau et al. [20] duringthe deposition of copper containing DLC films. Recently Ji et al. [25]have reported that incorporation of metal in DLC enhance thegraphite-like sp2 bonding. Thus, in the present case, observedincrease of conductivity with incorporation of Cu particles mayalso be due to an enhancement of sp2 bondings. The conductivityresults may also be corroborated with C–H based sp3 and sp2

bonding. The conductivity in all types of DLC films followed inverserelation with sp3 fraction i.e. it decreased continuously withincreasing sp3 fraction. Observed conductivity results are in goodagreement with said statement because conductivity in presentstudy was decreased with the increasing sp3 fraction. It is to benoted that with increase of C2H2 gas pressure the BHC in Cu-DLCfilms was also increased (confirmed by FTIR analysis). Therefore, itis very important to express conductivity as a function of BHC

because even small change in such hydrogen may influence theconductivity significantly. The variation of conductivity of Cu-DLCfilms at 300 K and 373 K versus BHC is depicted in Fig. 9(b). Thevalue of conductivity at both the temperatures was found todecrease with increasing BHC because presence of hydrogen inthe structure restricts the movement of charge carriers and alsoenhances the sp3 bonding and hence, reduces the conductivity. Incontrast, hydrogen in hydrogenated amorphous silicon (a-Si:H) isusually used for passivation of dangling bond to improve itselectronic properties.

3.9. Transmission studies

Higher optical transparency is a versatile characteristic of DLCespecially in the IR region [49]. However, metal incorporation inDLC film affects its optical transparency [50]. The transmissionspectra of Cu-DLC films are shown in Fig. 10. It should be notedthat transmission varies inversely with the thickness. Therefore,observed decrease of transmission with the increase in C2H2 gaspressures is due to the increase of thickness. It is evident from the

300 400 500 600 700 800 900 1000 1100

0

20

40

60

80

100 Cu-DLC 25 mTorr

C

B

A75 mTorr

50 mTorr

% T

rans

mis

sion

Wavelength (nm)

Fig. 10. Transmission spectra of Cu-DLC films deposited at C2H2 gas pressure of

25, 50 and 75 mTorr.

spectra that Cu-DLC films grown at C2H2 gas pressures of 25, 50and 75 mTorr exhibit transmission �80–90% in the near IR region.Presently, plasmonics for photovoltaic cells is a hot research topicin which metal nanoparticles present in the material help inenhancing the absorbance of light within the material and there-fore improve the efficiency of photovoltaic cells [51]. Our Cu-DLCstructure seems to be more or less same. However, if one play withthe size of incorporating Cu nanoparticles further, then these Cu-DLC structures can also be useful for photovoltaic applicationbecause Cu-DLC films show comparatively better conductivity thanpure DLC films.

3.10. Nano-mechanical properties

Recently, nanoindentation is recognized as an advanced tech-nique to measure the mechanical properties of thin films [52].Load versus displacement curves of Cu-DLC films, grown at C2H2

gas pressure of 25 mTorr, at indentation loads of 4, 5 and 10 mN isshown in Fig. 11. These load versus displacement curves wereemployed to estimate hardness (H), elastic modulus (E), elasticrecovery (ER) and plastic deformation energy (Ur). The hardness ofCu-DLC film was estimated using the composite hardness model[53]. The variations of H, E, ER and Ur with the indentation load forCu-DLC film grown at C2H2 gas pressure of 25 mTorr is depicted inFig. 12(a), (b), (c) and (d), respectively. This can be seen that H

varies inversely with the load. The value of H at 4, 5 and 10 mNwas found to be 20. 7, 17.7 and 10.2 GPa, respectively. Observed Evalues followed the similar trend and decreased from 260 to143 GPa with the increase in load from 4 to 10 mN. It is to benoted that mechanical properties of thin films strongly depend onthe penetration depth. According to Hook’s law depth varies indirect proportional with force (load). Thus, with increase in loadenhances the penetration depth and then beside film, substratealso contributes to the mechanical properties. Recently, Zhangand Komvopoulos [54,55] have reported the effect of indentationloads on the mechanical properties of tetrahedral amorphouscarbon films. Chen and Vlassak [56] and Saha and Nix [57] havealso systematically studied the effect of substrate on mechanicalproperties of thin films. They performed series of measurementsand observed that when hard film is grown on soft substrate thenhardness decrease with increasing penetration depth. However,when soft film is grown on hard substrate then hardnessincreases with the increase in penetration depth. The H and E of

0.00 0.04 0.08 0.12 0.16 0.20

0

2

4

6

8

10 25 mTorrCu-DLC

Load

(mN

)

Displacement (μm)

4 mN 5 mN 10 mN

Fig. 11. Load versus displacement curves at 4, 5 and 10 mN of Cu-DLC film grown

at C2H2 gas pressure of 25 mTorr.

4 5 6 7 8 9 10

4.0x10-10

8.0x10-10

1.2x10-956

60

64

68150

200

250

12

16

20

Cu-DLC

Ur (

Joul

e)

Load (mN)

Cu-DLC

ER

(%)

E (G

Pa) Cu-DLC

Cu-DLCH

(GP

a)

Fig. 12. Variation of (a) H, (b) E, (c) ER and (d) Ur with indentation loads for Cu-

DLC film grown at C2H2 gas pressure of 25 mTorr.

10

15

20

2525 mTorr

H (G

Pa)

Depth (nm)80 120 160 200

80 120 160 200

150

200

250

30025 mTorr

E (G

Pa)

Depth (nm)

Fig. 13. Variation of (a) H and (b) E with depth for Cu-DLC film grown at C2H2 gas

pressure of 25 mTorr. The insets represent the variations of H and E versus depth/film

thickness, respectively, for Cu-DLC film grown at C2H2 gas pressure of 25 mTorr.

N. Dwivedi et al. / Journal of Physics and Chemistry of Solids 73 (2012) 308–316 315

Cu-DLC film were also discussed as a function of depth. Here, it isto be noted that Cu-DLC is harder than Si. Since Cu-DLC filmsstudied for nano-mechanical properties are grown on Si, in thiscase hard film on soft substrate model is applicable. Plot of H andE versus depth for Cu-DLC film grown at 25 mTorr is shown inFig. 13(a) and (b). The values of H and E were found tocontinuously decrease with increasing depth because an increasein depth enhances the substrate effect. To visualize the effect ofdepth on H and E more precisely, these parameters were alsoexplored as a function of depth/film thickness. The insets ofFig. 13(a) and (b) show the variation of H and E with depth/filmthickness (h/t). It is evident that maximum H and E were observedat the h/t of 0.59. The values of H and E were found to decreasegradually when 0.59oh/to0.74 due to small substrate effect.However, when indenter penetrates the film completely (h/t41)then H and E decreased considerably and results show onlysubstrate property instead of film property. This was the reasonwhy at 10 mN, when indenter penetrates the film completely, theH was obtained just 10.2 GPa i.e. almost equal to H of substrate Si.This is to be noted that low values of hardness and elasticmodulus were previously reported by Lejeune et al. [58] fornitrogen doped DLC films and Gulbiski et al. [59] for copper/carbon composite film. However, we observed sufficient H and E

in Cu-DLC film.ER and Ur were also expressed as a function of load [2,9,60].

The % ER of Cu-DLC film was calculated using the followingrelation

% ER¼ðhmax�hresÞ

hmax� 100 ð4Þ

where hmax and hres are the displacements at the maximum loadand residual displacement after load removal, respectively. Thiscan be seen from the Fig. 12(c) that ER was continuouslydecreased with increasing indentation loads. The values of ER atindentation loads of 4, 5 and 10 mN were found to be 67, 65.2 and57.1%, respectively. The decrease of ER with the load was due toan increase of penetration depth which induces more plasticitybecause of substrate effect. Furthermore, ER results were found tobe in good agreement with the H results because ER varies inproportion with H. Plastic deformation energy (Ur) was also foundto be very important parameter to study the elastic-plasticproperties of thin films. The Ur possesses a relation with H, asgiven below

Ur ¼1

3

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

o0 tan2C

s" #1ffiffiffiffiHp P3=2

ð5Þ

where o0 is the geometry constant and attains that value of1.3 for pyramid indenter, P is the load and C is the half angle ofBerkovich indenter that has the value of 65.31. The Ur was foundto increase with increasing indentation load (Fig. 12(d)). Thevalues of Ur at indentation loads of 4, 5 and 10 mN were foundto be 2.4�10�10, 3.6�10�10 and 1.3�10�9 J, respectively.Observed enhancement in Ur with the increase of indentationload was due to enhancement in plasticity. The Ur results werealso in good agreement with H and ER results because Ur variesinversely with H and ER.

4. Conclusions

In the present study, the growth and characterizations of Cuincorporated DLC (Cu-DLC) films was explored. The Cu-DLC filmswere grown by hybrid system involving DC-sputtering and RF-PECVD units, under varied C2H2 gas pressure from 25 to 75 mTorr.These films were studied for structural, electrical, optical andnano-mechanical, properties. XPS, XRD, SEM and AFM results

N. Dwivedi et al. / Journal of Physics and Chemistry of Solids 73 (2012) 308–316316

confirmed the incorporation of Cu in DLC structure. XPS analysisreveals shifting of C 1s peak towards lower binding energy due topresence of Cu nanoparticles. Raman analysis inferred the shiftingof G and D peaks towards lower wave numbers and decrease in ID/IG ratios with the increase in C2H2 gas pressure from 25 to75 mTorr. This further confirmed the decrease in sp2 bondingwith the increase in C2H2 gas pressure. The experimental resultsshowed that copper incorporation in DLC matrix relaxed theresidual stress and improved the temperature dependent elec-trical transport considerably. FTIR was used to calculate boundhydrogen content and sp3 and sp2 fractions. The sp3 and sp2

fractions were correlated fairly with stress and electrical proper-ties. Estimated bound hydrogen content also correlated withelectrical properties. The nano-mechanical properties of Cu-DLCfilm were found to degrade with the increasing indentation loadfrom 4 to 10 mN due to an increase of substrate effect. Insummary, addition of Cu in DLC films minimizes stress andimproves the electrical and optical properties.

Acknowledgment

The authors are grateful to the Director, National PhysicalLaboratory, New Delhi, for his kind support. The authors also wishto thank Dr. S.K. Dhawan and Dr. Govind for providing UV–Vistransmission and XPS measurements facilities, respectively.Authors also wish to thank Mr. Ishpal for providing Ramanmeasurement facility. ND acknowledges CSIR, Govt. of India, forproviding SRF fellowship. CSIR, Govt. of India, is also acknowl-edged for sponsoring Network Project NWP-0027 and for theirfinancial support.

References

[1] J. Robertson, Phys. Rev. Lett. 68 (1992) 220.[2] N. Dwivedi, S. Kumar, C.M.S. Rauthan, O.S. Panwar, Appl. Phys. A 102 (2011)

225.[3] J.X. Liao, E.Q. Li, Z. Tian, X.F. Pan, J. Xu, L. Jin, H.G. Yang, J. Phys. D. Appl. Phys.

41 (2008) 055305.[4] T. Ohana, M. Suzuki, T. Nakamura, A. Tanaka, Y. Koga, Diamond Relat. Mater.

13 (2004) 2211.[5] S. Bhattacharyya, S.R.P. Silva, Thin Solid Films 482 (2005) 94.[6] L. Huang, Y. Hung, S. Chang, IEEE Trans. Magn. 33 (1997) 4551.[7] B. Gupta, P.K. Shishodia, A. Kapoor, R.M. Mehra, T. Soga, T. Jimbo, M. Umeno,

Sol. Energy Mater. Sol. Cells 73 (2002) 261.[8] S. Kumar, N. Dwivedi, C.M.S. Rauthan, O.S. Panwar, Vacuum 84 (2010) 882.[9] N. Dwivedi, S. Kumar, C.M.S. Rauthan, O.S. Panwar, Plasma Process. Polym. 8

(2011) 100.[10] C. Godet, S. Kumar, V. Chu, Philos. Mag. B 83 (2003) 3351.[11] K.M. Krishna, M. Umeno, Y. Nukaya, T. Soga, T. Jimbo, Appl. Phys. Lett. 77

(2000) 1472.[12] P.J. Fallon, V.S. Veerasamy, C.A. Davis, J. Robertson, G.A.J. Amaratunga,

W.I. Milne, J. Koskinen, Phys. Rev. B 48 (1993) 4777.[13] D.R. Mckenzie, D. Muller, B.A. Pailthorpe, Phys. Rev. Lett. 67 (1991) 773.[14] S.E. Ong, S. Zhang, H. Du, D. Sun, Diamond Relat. Mater. 16 (2007) 1628.[15] A. Champi, F.C. Marques, F.L. Freire, Diamond Relat. Mater. 13 (2004) 1538.[16] F.L. Freire Jr., J. Non-Cryst. Solids 304 (2002) 251.[17] J.F. Lin, Z.C. Wan, P.J. Wei, H.Y. Chu, C.F. Ai, Thin Solid Films 466 (2004) 137.

[18] H.W. Choi, J.H. Choi, K.R. Lee, J.P. Ahn, K.H. Oh, Thin Solid Films 516 (2007)248.

[19] A. Grill, IBM J. Res. Develop. 43 (1999) 147.[20] Y. Pauleau, F. Thiery, P.B. Barna, F. Misjak, A. Kovacs, S.N. Dub, V.V. Uglov,

A.K. Kuleshov, Rev. Adv. Mater. Sci. 6 (2004) 140.[21] I. Gerhards, C. Ronning, H. Hofsaas, M. Seibt, H. Gibhardt, J. Appl. Phys. 93

(2003) 1203.[22] I. Gerhards, H. Stillrich, C. Ronning, H. Hofsaas, M. Seibt, Phys. Rev. B 70

(2004) 245418.[23] A.P. Mousinho, R.D. Mansano, M.C. Salvadori, J. Alloys Compd. 495 (2010)

620.[24] C.D. Wagnar, W.M. Wiggs, L.E. Davis, J.F. Moulder, G.E. Muilenberg, Perkin-

Elmer Corp., 1979.[25] L. Ji, H. Li, F. Zhao, J. Chen, H. Zhou, Diamond Relat. Mater. 17 (2008) 1949.[26] N. Dwivedi, S. Kumar, Ishpal, S. Dayal, Govind, C.M.S. Rauthan, O.S. Panwar, J.

Alloys Compd. 509 (2011) 1285.[27] S. Paulston, P.M. Parlett, P. Stone, M. Bowker, Surf. Interface Anal. 24 (1996)

811.[28] S.H. Jeong, D.C. Kim, H.G. Jee, O.M. Moon, C.K. Jung, J.S. Moon, S.K. Kim,

S.B. Lee, J.H. Boo, J. Vac. Sci. Technol. B 22 (2004) 2216.[29] B.D. Cullity, Element of X-ray Diffraction, second edition, Addison-Wesley

Publishing Company Inc., 1978.[30] Z. Jiang, Q. Niu, W. Deng, Nanoscience 12 (2007) 40.[31] S. Hussain, A.K. Pal, Bull. Mater. Sci. 29 (2006) 553.[32] J.T. Titantah, D. Lamoen, E. Neyts, A. Bogaerts, J. Phys. Condens. Matter 18

(2006) 10803.[33] S.V. Singh, T. Zaharia, M. Creatore, R. Groenen, K.V. Hege, M.C.M. Van De

Sanden, J. Appl. Phys. 107 (2010) 13305.[34] W. Jacob, W. Moller, Appl. Phys. Lett. 63 (1993) 1771.[35] B. Dischler, A. Bubenzer, P. Koidl, Solid State Commun. 48 (1983) 105.[36] R. Paul, S.N. Das, S. Dalui, R.N. Gayen, R.K. Roy, R. Bhar, A.K. Pal, J. Phys. D

Appl. Phys. 41 (2008) 55309.[37] X.B. Yan, T. Xu, S.R. Yang, H.W. Liu, Q.J. Xue, J. Phys. D Appl. Phys. 37 (2004)

2416.[38] G. Lazar, K. Zellama, I. Vascan, M. Stamate, I. Lazar, I. Rusu, J. Optoelectron.

Adv. Mater. 7 (2005) 647.[39] N. Hand-Zoubir, F. Boukhors, R. Baghdad, K. Zellama, Sci. Technol. A 28 (2008)

9.[40] J.K. Walters, R.J. Newport, S.F. Parker, W.S. Howells, G. Bushnell-Wye, J. Phys.

Condens. Matter 10 (1998) 4161.[41] M. Sharon, S. Jain, P.D. Kichambare, M. Kumar, Mater. Chem. Phys. 56 (1998)

284.[42] K.W.R. Gilkes, S. Prawer, K.W. Nugent, J. Robertson, H.S. Sands, Y. Lifshitz,

X. Shi, J. Appl. Phys. 87 (2000) 7284.[43] C.C. Chen, F.C.N. Hong, Appl. Surf. Sci. 242 (2005) 261.[44] J.H. Choi, H.S. Ahn, S.C. Lee, K.R. Lee, J. Phys. Conf. Ser. 29 (2006) 155.[45] A.Y. Wang, H.S. Ahn, K.R. Lee, J.P. Ahn, Appl. Phys. Lett. 86 (2005) 111902.[46] J. Robertson, Semicond. Sci. Technol. 18 (2003) S12.[47] G.A.J. Amaratunga, M. Chhowalla, D.A.I. Munindradasa, K.G. Lim, C.J. Kiely, I.

Alexandrou, in: Proceedings of the First International Specialist Meeting onAmorphous Carbon, 1997, p. 311.

[48] D.R. McKenzie, E.G. Gerstner, in: Proceedings of the First InternationalSpecialist Meeting on Amorphous Carbon, 1997, pp. 3.

[49] S.S. Tinchev, Y. Dyulgerska, P. Nikolova, D. Grambole, U. Kreissig, T. Babeva, J.Optoelectron. Adv. Mater. 8 (2006) 308.

[50] H.W. Choi, M.W. Moon, T.Y. Kim, K.R. Lee, K.H. Oh, Mater. Sci. Forum 475–479(2005) 3619.

[51] H.A. Atwater, A. Polman, Nat. Mater. 9 (2010) 205.[52] J.F. Lin, P.J. Wei, J.C. Pan, C.F. Ai, Diamond Relat. Mater. 13 (2004) 42.[53] B. Johnson, S. Hogmark, Thin Solid Films 114 (1984) 257.[54] H.S. Zhang, K. Komvopoulos, J. Appl. Phys. 105 (2009) 083305.[55] H.S. Zhang, K. Komvopoulos, J. Appl. Phys. 106 (2009) 093504.[56] X. Chen, J.J. Vlassak, J. Mater. Res. 16 (2001) 2974.[57] R. Saha, W.D. Nix, Acta Mater. 50 (2002) 23.[58] M. Lejeune, M. Benlahsen, P. Lemoine, Solid State Commun. 135 (2005) 434.[59] W. Gulbiski, S. Kukielka, Y. Pauleau, F. Theiry, Surf. Coat. Technol. 200 (2005)

2146.[60] M. Sakai, Acta Met. Mater. 41 (1993) 1751.