Titanium carbide/carbon composite nanofibers prepared by a plasma process

Home

Search

Collections

Journals

About

Contact us

My IOPscience

Titanium carbide/carbon composite nanofibers prepared by a plasma process

This article has been downloaded from IOPscience. Please scroll down to see the full text article. 2010 Nanotechnology 21 435603 (http://iopscience.iop.org/0957-4484/21/43/435603) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 193.190.193.2 The article was downloaded on 24/01/2012 at 15:54

Please note that terms and conditions apply.

IOP PUBLISHING

NANOTECHNOLOGY

Nanotechnology 21 (2010) 435603 (8pp)

doi:10.1088/0957-4484/21/43/435603

Titanium carbide/carbon composite nanofibers prepared by a plasma process A A El Mel1 , E Gautron1 , C H Choi2 , B Angleraud1, A Granier1 and P Y Tessier1 1 Universit´e de Nantes, CNRS, Institut des Mat´eriaux Jean Rouxel, UMR 6502, 2 rue de la Houssini`ere BP 32229-44322 Nantes cedex 3, France 2 Department of Mechanical Engineering, Stevens Institute of Technology, Hoboken, NJ 07030, USA

Received 21 June 2010, in final form 30 August 2010 Published 4 October 2010 Online at stacks.iop.org/Nano/21/435603 Abstract The incorporation of metal or metal carbide nanoparticles into carbon nanofibers modifies their properties and enlarges their field of application. The purpose of this work is to report a new non-catalytic and easy method to prepare organized metal carbide–carbon composite nanofibers on nanopatterned silicon substrates prepared by laser interference lithography coupled with deep reactive ion etching. Titanium carbide–carbon composite nanofibers were grown on the top of the silicon lines parallel to the substrate by a hybrid plasma process combining physical vapor deposition and plasma enhanced chemical vapor deposition. The prepared nanofibers were analyzed by scanning electron microscopy, x-ray photoelectron spectroscopy, Raman spectroscopy and transmission electron microscopy. We demonstrate that the shape, microstructure and the chemical composition of the as-grown nanofibers can be tuned by changing the plasma conditions. (Some figures in this article are in colour only in the electronic version)

of carbon nanotubes. Hence, another strategy, based on coating TiC particles on the surface of CNFs has been developed [14–17]. However, the bad adhesion due to the presence of an interface between the TiC particles and carbon fibers is the origin of reduced rigidity and strength of the carbon fibers. Recently, Zhu et al have succeeded in incorporating TiC nanoparticles into CNFs by means of the electrospinning technique and via a carbonization and high temperature pyrolysis [10]. The prepared nanofibers have shown homogeneous distribution of TiC nanoparticles in the CNFs. Compared to the previous reported method, no adhesion problems between TiC nanoparticles and CNFs were recorded. The prepared c-TiC/CNFs have shown a high tensile strength and a high Young’s modulus. Despite the success of preparation of c-TiC/CNFs, long procedures under high temperature were needed. In addition, the synthesized nanofibers were not organized. In order to prepare nanofibers with a high length at temperature close to room temperature, two different approaches can be used. The first one, shown in figure 1 (left column), consists of a first step corresponding to a thin film deposition followed by a lithography step for which

1. Introduction Over the last few years, carbon nanofibers (CNFs) have attracted huge interest due to their specific physical properties such as high tensile strength, high elastic modulus and high thermal and electrical conductivity [1, 2]. These one-dimensional (1D) nanostructured materials are used as reinforcements for polymers and metals in order to improve their mechanical and electrical properties [3, 4]. Recently, it has been found that the incorporation of metallic nanoparticles into CNFs modifies their properties and enlarges their field of application. Depending on the chosen metal, metal/CNFs composite can be used for catalytic [5–9] or mechanical applications [10]. Titanium carbide/carbon composite nanofibers (c-TiC/CNFs) have recently been considered as a 1D nanostructured hard material [10]. Several works have reported on the synthesis of such 1D nanostructured hard material composites [10–17]. One of the strategies to prepare such a composite is to use carbon nanotubes as templates [11–13]. This strategy still has its limitations and disadvantages due to the low product yield, the very short length of the 1D nanocomposite and the high cost 0957-4484/10/435603+08$30.00

1

© 2010 IOP Publishing Ltd Printed in the UK & the USA

Nanotechnology 21 (2010) 435603

A A El Mel et al

The organization and the alignment of the nanofibers facilitate the manipulation, the characterization and the study of their properties before integrating them in a more complex system. The deposition method is based on the use of a hybrid plasma process combining physical vapor deposition and plasma enhanced chemical vapor deposition (PVD/PECVD) which consists of a simultaneous deposition of carbon and metal. The authors have previously studied the deposition of TiC/C composite thin films using the same PVD/PECVD process [19]. It was made clear that it was possible to control the microstructure and the chemical composition of TiC/C composite thin films by varying the deposition conditions. In this work the deposition of c-TiC/CNFs is studied using the same hybrid PVD/PECVD plasma process on nanopatterned silicon template substrates, prepared by laser interference lithography and followed by deep reactive etching [20], in order to prepare c-TiC/CNFs. The microstructure and the chemical composition of the as-grown nanofibers were studied by scanning electron microscopy (SEM), x-ray photoelectron spectroscopy (XPS), Raman spectroscopy and transmission electron microscopy (TEM).

Figure 1. Fabrication processes of nanofibers. By combining etching and deposition techniques, organized nanofibers with high length and parallel to the substrate can be produced at low temperature.

the film has to be etched in order to obtain nanofibers parallel to the substrate [18]. Nevertheless, this approach remains hard to perform on nanocomposite material because the plasma etching of such a heterogeneous material remains difficult to control. Moreover, a lithography mask should be developed specifically for this type of material. Finally, for each preparation of nanofibers an etching step will be needed. All these difficulties complicate this approach and make it expensive. We propose another approach, shown in figure 1 (right column), in order to achieve the nanofibers synthesis without facing any of the difficulties mentioned previously. At first, a nanopatterned silicon template substrate can be prepared by lithography, and then the nanofibers can be deposited by a plasma process on the top of the structural ridges. The lithographical patterning of silicon enables us to control the dimensions of the etched substrate at a very low dimension. A main important key to be mentioned is that by adding a sacrificial layer between the nanofibers and their silicon support, the nanofibers can be collected after a wet etching of this layer in a definite solvent. Thus, the nanopatterned template substrate can be re-used to prepare nanofibers multiple times. For all these reasons, the work reported in this paper is focused on the study of the preparation of organized c-TiC/CNFs using the second approach.

2. Experimental details 2.1. Preparation of the nanopatterned substrates The deposition was made on a nanopatterned silicon substrate (figure 2) which served as a template to grow well-aligned cTiC/CNFs along the pre-defined line patterns. The detailed preparation method was studied elsewhere [20]. In brief, it consists of coupling laser interference lithography with deep reactive ion etching in one process flow, in order to prepare nanopatterned silicon substrates of excellent regularity (230 nm of pitch in this study) and uniform coverage over a large sample area (2 × 2 cm2 in this study). As can be seen in figure 2, the silicon lines are approximately 115 nm in width, 500 nm in height and they are periodically separated by 115 nm wide nanotrenches. 2.2. Hybrid PVD/PECVD system The c-TiC/CNFs were prepared in the hybrid PVD/PECVD system presented in figure 3. A radio frequency (RF) generator at 13.56 MHz was connected to the unbalanced magnetron sputtering source via a matching box. A titanium disc

Figure 2. SEM images of the cross-section of the nano-grated silicon substrate patterned by laser interference lithography followed by deep reactive ion etching.

2

Nanotechnology 21 (2010) 435603

A A El Mel et al

Table 1. Summary of deposition parameters for the samples prepared in the present study. CH4 % = CH4 fraction in CH4 /Ar plasma, PCoil = RF power applied to the coil, PTarget = RF power applied to the magnetron, F = floating potential, G = grounded. The pressure during the deposition was maintained at 0.67 Pa for all the samples.

PCoil PTarget Deposition Substrate Type of Condition CH4 % (W) (W) time (min) bias material (1) (2) (3) (4)

Figure 3. Sketch of the hybrid PVD/PECVD system used to synthesize the nanofibers.

150 0 150 150 150 150 150 150

6 6 6 6

F F F G

Pure carbon Pure titanium TiC/C TiC/C

depends on the electron temperature Te of the plasma and the ion to electron mass ratio by the following relation [25]:   1 kTe e mi VP − Vf = ln (2) . 2 q 2π m e

(99.995% pure, 5 cm in diameter) was used as a target. The distance between the titanium target and the substrate was 8 cm. Another RF generator was connected through a matching box to a one turn stainless coil, located at equal distance from the target and the substrate, in order to generate an additional plasma. In pure argon atmosphere, the reactor is operated in PVD mode. The addition of a flow of CH4 gas leads to the deposition of hydrocarbon species by PECVD. The role of the additional plasma created by the coil is then to generate PECVD conditions for carbon deposition, which are independent of magnetron sputtering. A previous study has demonstrated that this hybrid process allows us to extend the range of deposition conditions and also leads to a stabilization of the reactive magnetron discharge [21]. The pressure during the deposition was maintained at 0.67 Pa. No substrate heating was applied during the deposition, but an increase of the temperature was observed due to plasma conditions. This temperature remains lower than 120 ◦ C. The total gas flow of the injected argon and methane gas into the deposition chamber was fixed at 12 sccm. The RF powers coupled to the coil ( PCoil ) and to the magnetron ( PTarget ) were controlled, depending on the deposition conditions. In addition, the substrate can be electrically grounded or kept at floating potential during the deposition which allows modifying the ion energy. In order to obtain a rough estimation of the ion energy on the sample surface, the floating potential Vf was measured during the deposition of the samples and Vf was found to be around 200 V. This high floating potential Vf results from the high value of the plasma potential generated by the RF powered coil. It was previously shown that for this value of Vf a high energy ion bombardment could be generated when grounding the substrate. It was even shown that grounding the substrate could be used for etching instead of depositing thin films [22–24]. Indeed, if we assume a non-collisional sheath in the plasma, the ion energy E ion is equal to the difference between the plasma potential, Vp and the substrate potential, VS :

E ion = q(VP − VS ).

24 0 24 24

For a typical value of kTe equal to 5 eV and for Ar+ ions, the potential difference (VP −Vf ) is evaluated at about 26 V. Hence, for a deposition at floating potential the Ar+ ion will strike the substrate with an energy of about 26 eV. Whereas, in the case of a grounded substrate the ion energy is given by:

E ion = q VP .

(3)

Therefore, the ion energy can be given by the relation:

E ion

  e mi 1 kTe = q Vf + q(VP − Vf ) = q Vf + ln . (4) 2 q 2π m e

In this case, the Ar + ion will strike the surface of the sample with an energy of about 226 eV which is much higher than the energy of a sample at floating potential (26 eV). 2.3. Deposition conditions In order to perform this study, four deposition conditions were used (table 1). For condition (1), the magnetron is off ( PTarget = 0 W), which means that there is no Ti sputtering and that only carbon is deposited. For condition (2), no methane is injected into the chamber (pure Ar plasma) which leads to a pure titanium deposition. For conditions (3) and (4), the magnetron and the coil are turned on ( PCoil = PTarget = 150 W) and a methane/argon mixture is injected into the chamber. Thus, carbon and titanium are simultaneously deposited. We used conditions which were previously shown to lead to the deposition of film made of TiC nanoparticles embedded in an amorphous carbon matrix [19]. These conditions are the following: an RF power of 150 W coupled to the RF coil and to the magnetron cathode and a Ar/CH4 76:24 mixture. The deposition time was fixed at 6 min for all the experiments. Samples prepared under conditions (1), (2) and (3) were at floating potential, whereas samples deposited under condition (4) were at ground potential. Hereafter, the samples deposited under conditions (1) to (4) are denoted S1–S4.

(1)

When the substrate is not grounded (i.e. when it is at the floating potential Vf ), the ion energy is equal to Vp − Vf which 3

Nanotechnology 21 (2010) 435603

A A El Mel et al

Figure 4. SEM image (a) and SEM backscattered electron image (b) of the trench cross-section of a sample prepared under condition (1). SEM image (c) and SEM backscattered electrons image (d) of the trench cross-section of a sample prepared under condition (2).

for the deposition of carbon, c-TiC/CNFs were grown on the nanopatterned silicon substrates at floating potential (S3), as shown in figure 5. Similar to those of S1 and S2, the SEM image of the trench cross-section of S3 (figure 5(a)) shows well-ordered nanofibers grown on the top of the silicon lines. The c-TiC/CNFs have a mean diameter around 150 nm. Their length corresponds to the silicon lines length, i.e. about 2 cm. The synthesized nanofibers are aligned, parallel to the substrate and separated from each other by about 40 nm, as shown in the top SEM image (figure 5(b)). Figures 5(c) and (d) show SEM images of the nanofibers disengaged from the template substrate after the etching of the native silicon oxide layer in hydrofluoric acid. In addition, they present a smooth surface, which is a characteristic of amorphous carbon deposited by PECVD. The nanofibers were collected on a silicon substrate and analyzed by SEM. As can be seen in the SEM images shown in figures 5(e) and (f), the long nanofibers exhibit a good flexibility which facilitates their manipulation. The CNFs and c-TiC/CNFs were analyzed by XPS in order to study their chemical composition. As for TiC/C composite thin films, Ti, O and C are detected [19]. The presence of oxygen can be understood as an oxidization of the surface of the nanofibers due to their exposure to the air. Special attention was paid to the C 1s spectrum. The XPS spectra of pure CNFs deposited at floating potential (S1) and c-TiC/CNFs deposited at floating potential (S3) are plotted in figure 6. For S1 we can identify one main peak located around 285 eV, whereas for S3 two main peaks located at about 282 and 285 eV are present in the spectrum. These peaks are attributed to Ti–C bonds in TiC phase and C–C bonds in amorphous carbon phase, respectively [28–32]. This led us to the conclusion that the nanofibers of S1 contain only an amorphous carbon phase, while for S3 the nanofibers are composed of titanium carbide and amorphous carbon. Micro-Raman scattering spectroscopy was used to explore the structural information of the amorphous carbon phase. The nanofibers were analyzed after having been dispersed on

2.4. Material characterization SEM images of the nanofibers were obtained by using a JEOL JSM 7600 F microscope operating at 5 kV. MicroRaman spectra were recorded using a Jobin-Yvon T64000 spectrometer under a microscope allowing a spatial resolution of about 2 μm. TEM imaging was performed on a Hitachi HNAR9000 microscope (LaB6 filament, 300 kV, Scherzer resolution: 0.18 nm). Surface analysis was performed by ex situ XPS (x-ray photoelectron spectroscopy) using a Kratos axis ultra spectrometer and monochromatic Al K radiation (1486.6 eV). The C 1s XPS region was acquired at a 20 eV pass energy, leading to an energy resolution of about 0.9 eV.

3. Results and discussion 3.1. Deposition of nanofibers at floating potential First, the deposition of pure carbon by PECVD (S1) and pure titanium by PVD (S2) on nanopatterned silicon substrates was studied. The SEM images of pure carbon and pure titanium deposited on the top of the silicon lines are shown in figures 4(a) and (c), respectively. These results demonstrate the possibility of synthesizing horizontally aligned and parallel pure carbon and pure titanium nanofibers on the top of the silicon lines. On the basis of the SEM backscattered electron images of both samples (figures 4(b) and (d)), it cannot be quantified how much deposition takes place at the bottom and on the sidewall of the trenches. The nanotrench filling depends on many parameters such as the ion energy, ion and neutral fluxes, the sputtering yield, the shadow effect of the material, and the aspect ratio of the patterned substrate [26, 27]. Qualitatively, no significant deposition was observed on the sidewall and at the bottom of the nanotrenches under the deposition conditions that we are studying here. By applying the two deposition techniques simultaneously, i.e. PVD for the deposition of titanium and PECVD 4

Nanotechnology 21 (2010) 435603

A A El Mel et al

Figure 5. SEM images of S3: trench cross-section (a), top image of aligned c-TiC/CNFs (b), c-TiC/CNFs on the patterned substrate after etching the native silicon oxide using fluoridric acid (c) and (d), c-TiC/CNFs collected and dispersed on silicon substrate (e) and (f). SEM images of the trench cross-section of S4 (g) and (h).

a silicon substrate. Figure 7 (insets) shows images of the analyzed region of S3 nanofibers, taken by the CCD camera of the micro-Raman system (top-left) and by SEM (top-right), respectively. The Raman spectra of S1, S3 and S4 recorded at 514 nm are shown in figure 7 (bottom). The spectra reveal an asymmetrical broad G-peak around 1580 cm−1 and a Dpeak around 1370 cm−1 . A difference can be noted between the Raman spectra of S1 and S3. The spectrum of S3 exhibits a G band more intense than the D band. As reported in the literature, this spectrum is typical of amorphous carbon with a high sp2 /sp3 ratio [28–34]. It is hence concluded that, beside the TiC phase detected by XPS, an amorphous carbon phase also exists in the samples. Figure 8(a) shows a low magnification TEM image of a nanofiber as-grown on a silicon line after being positioned on a copper grid covered with a thin holey carbon film. The high resolution TEM image of S3 is presented in figure 8(b). Crystalline nanoparticles with diameters between 2 and 4 nm embedded in an amorphous matrix are observed. The amorphous matrix can be attributed to the carbon

Figure 6. C 1s XPS spectra of S1, S3 and S4. A component around 282 eV, attributed to Ti–C bonds present in TiC phase, was detected in S3 and S4.

phase previously detected by XPS and Raman spectroscopy. Moreover, the lattice spacing measured on the high resolution ˚ TEM image of S3, shown in figure 8(b), is found to be 2.1 A. 5

Nanotechnology 21 (2010) 435603

A A El Mel et al

3.2. Ion bombardment effect during the growth of c-TiC/CNFs In order to modify the growth of the c-TiC/CNFs, the substrate has been electrically connected to the ground during the growth (S4). A modification of the shape of the nanofibers, as shown in SEM images of the trench cross-section of S4 in figures 5(g) and (h), clearly appears. The nanofibers are still aligned and parallel to the substrate but they have a triangular-like shape. This modification in shape can be related to the resputtering of material during the deposition when increasing the ion energy [26]. The C 1s XPS spectrum of S4 is presented in figure 6. An interesting result is that the intensity of the peak related to Ti–C bonds is much higher for S4 than for S3. This increase in intensity can be attributed to the increase in the amount of the TiC phase present in the nanofibers when increasing the ion energy. Crystalline nanoparticles, with diameters between 8 and 12 nm, can be observed in the high resolution TEM image (figure 8(c)). Using the SAED pattern of S4 shown in figure 8(d), the lattice spacings are found to ˚ These values can be attributed to be 2.43, 2.1 and 1.5 A. the (111), (200) and (220) lattice planes in fcc TiC crystal, as demonstrated by Lengauer et al [35]. Therefore, using high resolution TEM imaging we demonstrate that the nanoparticle size increases when grounding the substrate. This increase of nanoparticle size can be correlated to the increase of the TiCpeak intensity seen previously by XPS. Hence, the increase of ion energy, when grounding the substrate, can modify the

Figure 7. Raman spectra of samples prepared under conditions (1), (3) and (4) recorded at wavelength 514 nm. Image of dispersed nanofibers prepared under condition (3) taken by: the CCD camera of the micro-Raman system (top-left) and by SEM (top-right). The circle corresponds to the Raman analysis region.

This value, confirmed using selective area electron diffraction (SAED) (not presented in this paper), corresponds with (200) lattice planes in fcc TiC crystal as demonstrated by Lengauer et al [35].

Figure 8. TEM image showing a nanofiber as-grown on a silicon line positioned on a copper grid (a) and a high resolution TEM image of c-TiC/CNFs of S3 (b); a high resolution TEM image (c) and associated SAED pattern (d) of S4.

6

Nanotechnology 21 (2010) 435603

A A El Mel et al

Figure 9. SEM images of the trench cross-section of the sample prepared under condition (3), after 5 min of sputter-etching in pure argon (a) and (b) SEM images of the trench cross-section of the sample prepared under condition (2), after 5 min of sputter-etching in pure argon (c) and (d). The PCoil was maintained at 150 W and the substrate was grounded.

In this case, not only the shape is changed, but also a deposition of pure titanium takes place between the nanofibers. The incidence angle θ2 is found to be equal to 60◦ which indicates that the maximum sputtering yield of titanium can be found for an angle around 60◦ .

shape, chemical composition and the microstructure of the asgrown c-TiC/CNFs. 3.3. Ion bombardment effect on the deposited nanofibers As explained previously, the origin of the modification of the shape, chemical composition and microstructure of the nanofibers should be related to the ion energy which increases when the substrate is at ground potential. In order to demonstrate this, the same reactor used to synthesize the nanofibers was used to sputter, in pure argon plasma, samples prepared under conditions (2) and (3). The pressure and the argon flow during etching were, respectively, kept at 0.67 Pa and 12 sccm. The substrate was grounded and the electrical power of the coil PCoil was maintained at 150 W. The electrical power was maintained at PTarget = 0 W and the shutter was closed. An SEM image of the trench cross-section of the sample prepared under condition (3) after 5 min of sputteretching is presented in figures 9(a) and (b). Comparing figure 9(b) with 5(a), a modification of the shape of the nanofibers can be clearly seen. The triangular-like shape obtained can be related to the sputter-etching yield which depends on the value of the incidence angle θ1 of the Ar ions indexed on figure 9(b) [36]. The mean value of θ1 , after the sputter-etching of this sample, is found to be equal to 48◦ . Similarly, the mean value of θ found for S4 (nanofibers with a triangular-like shape shown in figures 5(g) and (h)) is found to be 49◦ . These values of the incidence angle θ found in the two cases confirm that the modification of the shape of the nanofibers obtained under condition (4) can be explained as a re-sputtering of the material on the substrate during the growth of the nanofibers. To confirm this result, the same experiment was performed on a sample prepared using condition (2) (pure titanium nanofibers). After 5 min of sputter-etching, the nanofibers are no longer separated and they start taking a triangular-like shape (figures 9(c) and (d)).

4. Conclusion Aligned TiC/C composite nanofibers parallel to the substrate, 150 nm in diameter and with a length that reaches up to several millimeters, were elaborated using patterned silicon substrates prepared by laser interference lithography. The hybrid plasma PVD/PECVD process used for the nanofiber synthesis is a low temperature process (T < 120 ◦ C). In 6 min and without any treatment at high temperature the nanofibers can be produced. The plasma parameters enable us to control the shape, the chemical composition and the microstructure of the deposited nanofibers. The good organization of the nanofibers on the top of the silicon lines facilitates their manipulation and the study of their different physical properties. Depending on their applications, these nanofibers can be used as-grown on the silicon substrate or they can be collected after wet etching of the native silicon oxide considered as a sacrificial layer. This process is not destructive for the nanopatterned silicon substrate which can be used again and again. More studies will be performed in order to investigate the different properties of these objects. Moreover, this method can also be extended in order to synthesize any other carbide/carbon or metal/carbon nanofibers by replacing the titanium target with another metal target.

Acknowledgments The authors would like to thank F Massuyeau and J M Lorcy for the useful discussions on the nanofibers, N Stephant and S Grolleau for their assistance on SEM imaging, J Y Mevellec 7

Nanotechnology 21 (2010) 435603

A A El Mel et al

for his useful advice on Raman spectroscopy and F Lari for her help with the manipulation of the nanofibers.

[17] Dong Z J, Li X K, Yuan G M, Cong Y, Li N, Jiang Z Y and Hu Z J 2009 Thin Solid Films 517 3248 [18] Francioso L and Siciliano P 2006 Nanotechnology 17 3761 [19] El Mel A A, Angleraud B, Gautron E, Granier A and Tessier P Y 2009 Surf. Coat. Technol. 204 1880 [20] Choi C H and Kim C J 2006 Nanotechnology 17 5326 [21] Tessier P Y, Issaoui R, Luais E, Boujtita M, Granier A and Angleraud B 2008 Solid State Sci. 11 1824 [22] Angleraud B, Mubumbila N and Tessier P Y 2003 Diam. Relat. Mater. 10 1024 [23] Angleraud B and Tessier P Y 2004 Surf. Coat. Technol. 180 59 [24] Ghamouss F, Luais E, Thobie-Gautier C, Tessier P Y and Boujtita M 2009 Electrochim. Acta 54 3026 [25] Lieberman M A and Lichtenberg A J 1994 Principles of Plasma Discharges and Materials Processing (New York: Wiley) chapter 6 [26] Hamaguchi S and Rossnagel S M 1995 J. Vac. Sci. Technol. B 13 183 [27] Radzimski Z J, Posadowski W M, Rossnagel S M and Shingubara S 1998 J. Vac. Sci. Technol. B 16 1102 [28] Voedvodin A A and Zabinski J S 1998 J. Mater. Sci. 33 319 [29] Zehnder T and Patscheider J 2000 Surf. Coat. Technol. 138 133 [30] Li G and Xia L F 2001 Thin Solid Films 396 16 [31] Gulbi´nski W, Mathur S, Shenb H, Suszkoa T, Gilewicza A and Warcholi´nski B 2005 Appl. Surf. Sci. 239 302 [32] Lewin E, Wilhelmsson O and Jansson U 2006 J. Appl. Phys. 100 054303 [33] Ferrari A C and Robertson J 2000 Phys. Rev. B 61 14095 [34] Casiraghi C, Ferrari A C and Robertson J 2005 Phys. Rev. B 72 085401 [35] Lengauer W, Blinder S, Aigner K, Ettmayer P, Guillou A, Debuigne J and Groboth G 1995 J. Alloys Compounds 217 137 [36] Sigmund P 1969 Phys. Rev. 184 383

References [1] Lawrence J G, Berhan L M and Nadarajah A 2008 ACS Nano 2 1230 [2] Yu C, Saha S, Zhou J, Shi L, Cassell A M, Cruden B A, Ngo Q and Li J 2006 J. Heat Transfer 128 234 [3] Duszova A, Dusza J, Tomaoek K, Morgiel J, Blugan G and Kuebler J 2008 Scr. Mater. 58 520 [4] Hirota K, Nakayama Y, Kato M, Nakane S and Nishimura N 2008 J. Appl. Ceram. Technol. 6 607 [5] Kim C, Kim Y A, Kim J H, Kataoka M and Endo M 2008 Nanotechnology 19 145602 [6] Sen S and Puri K I 2004 Nanotechnology 15 264 [7] Zheng J S, Wang M X, Zhang X S, Wu Y X, Li P, Zhou W G and Yuan W K 2008 J. Power Sources 175 211 [8] Lin Z, Ji L and Zhang X 2009 Mater. Lett. 63 2115 [9] Zhang L, Cheng B and Samulski E T 2004 Chem. Phys. Lett. 398 505 [10] Zhu P, Hong Y, Liu B and Zou G 2009 Nanotechnology 20 255603 [11] Taguchi T, Yamamoto H and Shamoto S 2007 J. Phys. Chem. C 111 18888 [12] Kim Y, Kim J, Park C, Yun S, Kim W, Heu S and Park J 2008 Thin Solid Films 517 1156 [13] Wong E W, Maynor B W, Burns L D and Lieber C M 1996 Chem. Mater. 8 2041 [14] Baklanova N I, Zima T M, Titov A T, Naimushina T M and Berveno V P 2008 Inorg. Mater. 44 121 [15] Baklanova N I, Zaitsev B N, Titov A T and Zima T M 2008 Carbon 46 261 [16] Li X, Dong Z, Westwood A, Brown A, Zhang S, Brydson R, Li N and Rand B 2008 Carbon 46 305

8