Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix

Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix

Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎ Contents lists available at ScienceDirect Tribology International journal homepage: www.elsevier.com/locate...

4MB Sizes 0 Downloads 2 Views

Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Tribology International journal homepage: www.elsevier.com/locate/triboint

Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix Lukasz Kolodziejczyk n, Piotr Kula, Witold Szymanski, Radomir Atraszkiewicz, Konrad Dybowski, Robert Pietrasik Institute of Materials Science and Engineering, Lodz University of Technology, Stefanowskiego 1/15, 90-924 Łódź, Poland

art ic l e i nf o

a b s t r a c t

Article history: Received 17 July 2014 Received in revised form 10 October 2014 Accepted 2 December 2014

In this work friction characteristics of polycrystalline graphene grown on the metallic matrix from the liquid phase via new metallurgical method were investigated using atomic force microscopy under the range of applied normal loads. Polycrystalline graphene films were transferred onto silicon oxide using a modified polymer supported transfer process. Lateral force microscopy mode with both silicon and diamond-like carbon coated probes was used to measure the friction properties. Ogletree method was employed in order to calibrate the lateral force and estimate the average friction coefficient of polycrystalline graphene. As a comparison the friction of SiO2 surface, commercially available CVD grown graphene monolayers transferred onto SiO2 as well as highly oriented pyrolytic graphite were also measured under the same experimental conditions. & 2014 Elsevier Ltd. All rights reserved.

Keywords: Polycrystalline graphene Friction Lateral force microscopy

1. Introduction During last several years the number of methods for graphene production has significantly increased. Some of these include mechanical [1–3] or chemical [4] exfoliation, CVD [5,6] or epitaxial growth [7], arc discharge, the reduction of graphene oxide [8,9], and many others. Li et al. [10] found that low-pressure CVD synthesis of graphene on Cu foil provides a good way of fabricating uniform single-layer graphene films. They showed that the continuous films are formed by connecting randomly oriented, irregular-shaped, and micrometre-sized graphene flakes. Employing a liquid matrix has been proposed as effective means for controlling the nucleation process in CVD graphene production [11]. It completely removes the grain boundaries in polycrystalline solid, allowing a uniform distribution of graphene nucleation sites, and enabling self-assembly of graphene flakes into compact and ordered structures. Based on this phenomena a new method of manufacturing of polycrystalline graphene from liquid matrix via metallurgical process has been proposed [12,13]. Frictional behaviour of graphene has gained recent attention [14], with several studies demonstrating the frictional anisotropy

n

Corresponding author. Tel.: þ 48 42 631 30 48. E-mail addresses: [email protected] (L. Kolodziejczyk), [email protected] (P. Kula), [email protected] (W. Szymanski), [email protected] (R. Atraszkiewicz), [email protected] (K. Dybowski), [email protected] (R. Pietrasik).

of graphene grains primarily due to the development of wrinkles. Graphene has been found to lower friction when compared to bare substrates. This general result was found in case of exfoliated, epitaxially-grown, or CVD grown graphene supported on several substrates (e.g. copper, nickel, silicon dioxide). However, compared with the works on general properties of graphene [15], the tribological investigation of graphene is still quite limited. Additionally to wrinkles formation the friction of graphene is greatly affected by graphene-to-substrate adhesion and number of graphene layers [15,16]. Theoretical simulations predicted that the low friction of graphene is highly dependent on the sliding interfaces (whether the tip is sliding on the graphene surface or two graphene surfaces are sliding against each other) and on the number of graphene layers. The modelling of the AFM tipgraphene interactions [17] demonstrated that the friction force decreases as the number of layers increases. The great majority of experimental studies of graphene tribology at the nano-scale have used lateral force microscopy (LFM) [18–22]. The LFM measurements demonstrated the friction dependence on the number of graphene layers [19,21,23] in agreement with the predicted simulation results. Li et al. [23] performed LFM studies in ambient conditions using silicon probes. Both free-standing graphene and graphene supported on SiO2 showed a decrease in friction with number of layers due to the puckering effect. However, observed thickness dependent friction is not present for graphene strongly bonded to the substrate (e.g. mica) [23]. The main idea of the presented work is the evaluation of frictional characteristics of the metallurgical graphene obtained

http://dx.doi.org/10.1016/j.triboint.2014.12.003 0301-679X/& 2014 Elsevier Ltd. All rights reserved.

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2

by new technology of graphene growth on liquid metallic matrix. The evaluation has been done with reference to the graphene produced by commonly known CVD method. Additionally, complete characterisation of metallurgical graphene has been conducted, including qualitative and quantitative analysis with the use of optical, scanning and atomic force microscopy and Raman spectroscopy.

Instruments systems was used for analysis of the possible AFM tip wear. Hitachi scanning electron microscope S-3000M, working in SE mode at accelerating voltage of 5 kV, was used for qualitative analysis of metallurgical graphene layers morphology. The quality of graphene layers after the transfer process onto SiO2/Si substrate was studied using an optical microscope Nikon Eclipse MA 200. Acquisition of the optical images was obtained using Nis Elements BR software.

2. Materials and methods 2.1. Samples preparation In the presented work following materials were used: highly oriented pyrolytic graphite (HOPG) grade ZYH (Bruker Corporation, USA), silicon dioxide on 〈1 0 0〉 p-doped silicon with thermal oxide thickness of 300 nm (Semiconductor Wafer Inc., Taiwan), single layer CVD graphene (latter in the text abbreviated as CVDG/SiO2) already transferred onto 285 nm thick SiO2 (Graphene Laboratories Inc., USA) and multilayer polycrystalline graphene grown on liquid metallic matrix (referred as metallurgical graphene and abbreviated as M–G/SiO2) [24]. CVD graphene films were transferred onto SiO2 using a PMMA assisted transfer method [6,25]. Polycrystalline graphene production consisted of following steps. First, the multilayer metallic substrate, the copper/ nickel composite (72% Cu, 28% Ni) [24,26], was heated to the temperature of 1200–1250 1C in argon protective atmosphere at the constant pressure of 100 kPa. Then, the substrate was kept within this range of temperature for 1 min. Afterwards, the plate was cooled to the temperature of 1050 1C at the cooling speed of 0.5 1C/min, also in argon atmosphere at the constant pressure of 100 kPa. During the first two aforementioned thermal stages, the mixture of acetylene, hydrogen and ethylene in the proportions of 2:1:2 [27], at the partial pressure of 3 kPa, has been simultaneously introduced into the process chamber [12,28]. For the polycrystalline graphene grown on liquid matrix the modified method of graphene transfer from metallic substrates on thin PMMA film has been employed. This method and its variations are most often utilized in procedures for graphene transfer on any substrate (Si/SiO2, glass, PE, etc.) [6,29–31]. Graphene on nickel/copper substrate was coated with PMMA in chlorobenzene (46 mg/ml). The thickness of the etched Ni/Cu substrate was 0.3 mm. Surface prepared in this way was dried at 50 1C for 1 h. PMMA/Ni/Cu stack was placed in a FeCl3 solution (1 M) for 24 h. Occasionally, prepared PMMA foil was covered with the residues of FeCl3 on the free side (without graphene). In that case PMMA foil (only free side) was rinsed in concentrated HF solution and 5% H2O2 and then in deionized water. Prepared PMMA foil with graphene, prepared in this way, was ready to follow the procedures for the transfer of graphene to any substrate. Selected piece of dH2O wetted foil was placed on the SiO2 target substrate. The entire part was placed on a hot plate and heated to a temperature of about 70 1C. After stretching the foil on the substrate the setup substrate-foil-graphene was immediately transferred into the atmosphere of boiling acetone vapour for about 10 min. Then, the sample was removed and cooled down. Such a procedure ensured good adhesion of graphene to the substrate. Next, the PMMA foil was dissolved by submerging it in acetone for 24 h (four times exchanging the solvent). Polycrystalline graphene samples prepared using above described transfer method were then ready for investigation. 2.2. Materials characterization 2.2.1. Optical and scanning microscopy JEOL JSM-6610LV scanning electron microscope integrated with MiniCL-GATAN Cathodoluminescence Imaging and Oxford

2.2.2. Raman spectroscopy For every graphene sample on SiO2 substrate the Raman spectroscopy using Ntegra Spectra Solar spectrometer (NT-MDT, Russia) was conducted. The Raman scattering in the range of 1000– 3000 cm  1 using visible 2.33 eV (532 nm) excitation of Ar þ laser was studied. The laser power held at 3.2 mW did not cause any changes on the surface (local heating and damage of the samples). The spectra were registered using NovaPX software (NT-MDT). Data processing was performed with use of PeakFit software. The spectra were deconvoluted using Gauss–Lorentz curve. Obtained peaks intensities and half-widths were used for the calculation of characteristic peak ratios of graphene. 2.2.3. Atomic and lateral force microscopy Surface morphology, topography and lateral force measurements were measured using Multimode atomic force microscope equipped with Nanoscope V controller (Bruker Corporation, USA). All investigations were performed under ambient conditions. Prior to topography measurements and LFM experiments the areas of 500  500 nm were scanned (up to several times in one scanning direction along slow scan axis) in contact mode at the very small loads (2–3 nN) to clean the surface from mainly polymer deposits remaining after graphene transfer process and move them outside measurements’ area. Topography measurements were made in contact mode at applied load of ca. 2 nN and the size of the images were 500  500 nm. The images were shaded in to enhance contrast. Commercial silicon cantilevers type HQ:CSC15 (MicroMasch, Estonia) with nominal tip radius  8 nm and nominal cantilever spring constant of 0.3–0.8 N/m were used. Image acquisition was performed with use of Nanoscope 7.3 software and further image processing was done using Nanoscope Analysis 1.5 (Bruker Corporation, USA) and MountainsMap Premium 5.0 (Digital Surf, France) software. From the topography images commonly used roughness parameters – the average roughness (Ra) and root-mean square roughness (RMS) – of the samples were defined (average values taken from 512 surface profiles). The error was calculated as the standard deviation among all surface profiles. For friction measurements the smaller areas of 100  100 nm within previously cleaned zone were used. The average friction measurements were done at increasing loads up to  20 nN (10 normal loads ramping over the single image) at 1 Hz scan rate. For friction tests two different types of AFM tips: commercial silicon cantilevers type HQ:CSC15 (same as for topography images) and diamond-like carbon (DLC, hydrogen-free ta-C) coated probe cantilevers type HQ:XSC11 (MicroMasch, Estonia) with resulting tip radius r20 nm and nominal cantilever spring constant of 0.2 N/m were used. Precise value of cantilever spring constant was specified using thermal tune method before each measurement. Both, deflection sensitivity of the cantilever and adhesion of the tip to the scanned surface were determined from force–distance measurements. The friction tests were carried out in air at temperature 21–23 1C and relative humidity 39–42%. At least seven results from measurements under same load were taken to calculate friction coefficient for single LFM test, and at least six LFM tests using same type AFM tip were performed on each

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

sample. Based on these results the average friction coefficient for each sample was calculated. The error was calculated as the standard deviation among all single LFM tests. To do the lateral force calibration by the Ogletree method [32], a commercial TGG1 silicon calibration grid (NT-MDT, Russia) with 1-D array of triangular steps was used.

3. Results and discussion 3.1. Qualitative analysis of metallurgical graphene morphology The SEM studies have shown that the polycrystalline graphene formation on the liquid metal phase surface is initiated by the nucleation and growth of single flakes of the hexagon-shaped graphene (Fig. 1a). Growth of the individual flakes is of competitive nature, which means that during the process the amount of flakes is still increasing (nucleation), and the ones already existing grow. This leads to the arrangement in which, at some point, a single graphene flakes floating on the liquid metal surface contact with their borders thus establishing continuous layers of larger dimensions (Fig. 1b and c). In some cases, under appropriate thermodynamic conditions of the process, it is possible to obtain

3

multilayers of graphene (Fig. 1d). As reported in [33] CVD graphene is also of a polycrystalline nature, with multiple small graphene domains growing to merge into a continuous film. Two ways for connection of graphene domains into a continuous film are possible: by direct atomic bonding at the interface of two graphene flakes or by overlapping of two graphene flakes to form a bilayer boundary separated by van der Waals forces only. It has been shown that direct atomic bonding at the interface occurs in mono- layer and few-layer graphene (FLG) films, whilst the overlapping of flakes has only been noticed for FLG films [33]. Moreover, the method of nucleation and growth of polycrystalline graphene on a liquid metallic substrate is characterized by the zone cleaning effect, e.g. pushing the contaminations outside of graphene flakes. Another morphological feature can be formation of wrinkles as a result of the difference in thermal expansion coefficients between the graphene and the forming metallic matrix (Fig. 1e). The coalescence occurring during graphene growth is also dependent on the degree of coherence of the graphene flakes. Low coherence can result in the layers discontinuity, connecting partially between the sides or even overlapping one flake onto another (Fig. 1f). As a result, formation of monoor multilayers of metallurgical graphene on liquid metallic matrix, characterized by a certain degree of disordering and defects,

Fig. 1. SEM images showing polycrystalline graphene formation on the liquid metal matrix: (a) single graphite flake, (b) nucleation and growth of the flakes, (c) creation of continuous layer, (d) multilayer formation, (e) zone cleaning effect and (f) layers discontinuities and overlaps due to low coherence between individual flakes.

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

4

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

may directly influence its properties, e.g. electrical resistance, friction coefficient, etc. 3.2. Morphology and thickness of graphene samples after transfer onto SiO2 substrate

Apart from wrinkles, both surfaces are contaminated, mainly by the residues left after transfer procedure. Therefore, prior to friction measurements all graphene samples were cleaned mechanically, as discussed in the next subsection (Section 3.3). 3.3. Cleaning of graphene samples

Optical images presented below show CVD and metallurgical graphene samples transferred onto SiO2 substrate (Fig. 2). We can observe that the investigated structures are not perfect. Visible contaminations are result of the FeCl3 etching and the transfer process using PMMA [34]. Transfer method causes the formation of wrinkles on the surface and structure discontinuity [6]. This type of disorder present on the surface of graphene cannot be avoided. Previous methods of graphene production (growing on metallic substrates) and graphene transfer on any substrates introduce defects and impurities. We can try to reduce described defects using additional treatments for example: mechanical cleaning or annealing, which is discussed later in this work. Our experience with the transfer of graphene onto different substrates (Si, Si/SiO2, PDMS, polyimide, glass, quartz) leads to the conclusion that it is not possible to obtain discontinuity-free transferred graphene (Fig. 2b and d) over large areas. In Fig. 3 the thickness measurement of the graphene samples is shown. The images were collected in contact mode AFM in the areas where the layers are not continuous to promote direct measurement in reference to the SiO2 substrates. In case of CVDG/SiO2 sample (Fig. 3, left image and profile) acquired thickness of the graphene is ca. 0.5 nm, slightly thicker than its interlayer spacing, what may suggest that it is monolayer. For M–G/SiO2 sample (Fig. 3, right image and profile) the average thickness is 4 times higher compared to CVD graphene (ca. 1.6 nm), what may prove that it consist of at least 4–5 layers. Moreover, the spikes visible on both profiles show that the graphene layers are not completely flat, there are relatively high density of wrinkles originated from both the manufacturing and transfer processes.

Contamination is introduced on the surface of graphene during sample preparation where, for example, the use of organic solvents and other liquids cannot typically be avoided. Airborne pollutants landing on surfaces in ambient air is the other main source of contamination. The surface contaminants are typically composed of different hydrocarbons. There are several different techniques for cleaning graphene: chemical cleaning, thermal cleaning [35–37], current-induced cleaning [38,39], dry-cleaning [40], and mechanical cleaning [41–43]. Each of these can be very useful but has its own limitations. Standard cleaning using solvents cannot remove all the residues [35,36,44]. Heating treatment is a common practice for removing hydrocarbon contamination from surfaces by high temperature annealing in inert, typically Ar/H2 atmosphere [35,36,45,46]. Although annealing of graphene in vacuum or inert atmosphere reduces the amount of contamination [47] it has been shown that residuals are still present after such treatment [37,48] and the coupling between the substrate and graphene may increase, leading to mechanical deformation of the graphene [37]. Additionally, many substrates cannot sustain high temperature without oxygen atmosphere, which is incompatible with graphene. Another method, annealing by Joule heating can be done in situ in a cryostat. However, graphene is also in this case heated locally to high temperature leading to rippling or even breakage if too much current is applied. Other method for removing surface contamination from graphene is drycleaning with adsorbents (activated carbon and less efficient—alumina). The dry-cleaning treatment produced atomically clean areas in the micrometre range, however cleaning effect was observed only in single-layer graphene, whereas bi- and

Fig. 2. Optical microscopy images of transferred graphene; (a and b) CVD-G/SiO2; (c and d) M–G/SiO2; nucleus of n-layer of graphene – red arrows; discontinuity of graphene – white arrows, graphene wrinkles – black arrows; contaminations after etching and transfer process – dotted lines. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

5

Fig. 3. Topography of CVD graphene (left image) and metallurgical graphene (right image) on SiO2 substrates obtained by contact mode AFM. Below the images respective line profiles (along the yellow lines) with approximate thickness values are presented. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

multilayer graphene remained contaminated [40]. Just recently an alternative cleaning method—mechanical cleaning of graphene was suggested [41–43]. Scanning a contact mode atomic force microscope (CM AFM) tip over a graphene surface removes residues, removes doping, and improves the electronic mobility without damaging the graphene [43]. Resist residues are efficiently brushed away, piling up outside the graphene flake leading to obtaining atomically smooth graphene. To remove the residues, we scanned the samples in LFM mode with a very small constant force. We engaged the tip with the lowest force possible. When the tip made contact, we confirmed a reasonable set-point force according to force distance measurement. Then we started scanning the sample with a rate of 1 Hz. For most samples, we scanned the same area several times but without further visible effect. AFM tapping mode images taken after scanning in LFM mode show that we cleaned the graphene (Fig. 4b inside marked window) compared to as-transferred sample (Fig. 4a). The roughness of the area after cleaning was  0.55 nm, similar to the values measured before processing (ca. 0.5 nm). Clear evidence that residues were removed from the graphene are the banks of deposits that are visible in Fig. 4b and c (1 mm  1 mm scan), exactly at the boundaries of the area that was scanned several times in LFM mode (500 nm  500 nm). On the presented images (Fig. 4a) the contaminants prior to mechanical cleaning are not visible and so it is difficult to see the changes after the cleaning process (Fig. 4b). But one has to take into account that to study the tribological properties only microscopically clean areas were chosen. The material accumulated on the edges is composed of individual polymer chains, remaining after the transfer process. Many publications show that the nanometric and subnanometric residues left after wet chemistry process can be visualized using HR-TEM [48] and the preparation of graphene surface free of impurities after the transfer process with the use of PMMA is impossible [34,49]. In our case, additionally, it is difficult to observe small changes in morphology due to existing wrinkles.

3.4. Topography of the samples before LFM measurement Fig. 5 shows topography images obtained using AFM working in contact mode. Features like blisters or wrinkles can be seen on the surface of graphene. These features are not similar to those shown in Fig. 2, which are generated by crimping layers of graphene. Wrinkles visible on the presented AFM images are considerably smaller. They may be formed during the transfer process on the SiO2/Si substrate. Wrinkles formation on graphene due to the difference in thermal expansion of the metal growth substrate and graphene as well as wrinkles formed in the transfer process is described extensively in the literature [50–53]. From the presented publications it can be concluded that wrinkles are always present. However, it must be noted that their morphology and size are different, depending on the manufacturing process, its parameters as well as the transfer method, which is described in the literature to a great extent similarly, with some minor details. The formation of islands (blisters) on the surface of the graphene transferred onto different substrates is based on various research papers and doctoral dissertations [50,54,55]. The formation of blisters occurs upon the contact of the hydrophilic substrate, on which the graphene is transferred, with moisture in the air. It is also necessary to take into account the transfer process using wet chemistry. Air and water trapped during the graphene transfer process on any substrate, in particular hydrophilic, as SiO2 (CA ¼521 [56]), can cause both blistering and wrinkles formation. The literature focuses more on the water trapped between the graphene and the substrate. Our experience shows, however, that on the surface of PMMA, subjected to hot acetone vapours treatment, bubbles formed from a thin PMMA film can be observed (clearly visible when transferring of large area graphene). A diagram which supports the blisters and/or wrinkles formation hypothesis is presented in Fig. 6. It is possible that between graphene and the target substrates enclosed areas air, water or both are formed. The blister formation is illustrated in

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

Fig. 4. AFM images (a) before and (b) after the mechanical cleaning of M–G/SiO2 sample. White box on (b) shows the area of cleaning with banks of deposits on the boundaries. (c) Phase contrast image clearly showing the cleaned area.

Fig. 5. Topography images: (a) HOPG, (b) SiO2 substrate, both (c) metallurgical and (d) CVD graphene after cleaning procedure, with (e) corresponding horizontal cross profiles. Description of the features marked on the AFM images: wrinkles—dotted line areas, blisters—dashed line circles, solid line rectangles—mapped SiO2 topography. Scan size 500 nm x 500 nm.

Fig. 6c. Such areas can be observed on Fig. 5d and e. The second possibility could be the formation of a corrugated graphene surface which is formed during the shrinking of the graphene supporting polymer. In this case multipoint contact (not the entire area) occurs between the transferred graphene and a target substrate. Air can also be trapped under the cavity of graphene waviness. After dissolving the supporting polymer it is not possible to remove wrinkles and this feature can be observed in Fig. 5c and e. Attention must be paid to the topography of the bare SiO2 substrate (Fig. 5b and e). In this case the layer of graphene may map SiO2/Si surfaces. Topography of graphene is a superposition of two phenomena: wrinkle/blister formation and mapping the topography of the substrate. In contrast, the surface of HOPG is perfectly smooth (Fig. 5a and e), compared to the other surfaces. Results of roughness value for HOPG show that there are no sudden changes in the roughness profile (high peak or deep valley). Ra and RMS parameters are almost equal (Table 1) and it can be concluded that HOPG surface is homogeneous. The commonly used roughness parameters (Ra and RMS) for all investigated surfaces are gathered in Table 1. 3.5. Raman spectra analysis Raman spectroscopy is a useful technique for characterizing the carbon structures composed of atoms of the sp2 and sp3 carbon hybridization, such as graphite, fullerenes, carbon nanotubes, and graphene. Raman spectra derived from sp2 hybridised carbon atoms are composed of many peaks. The appearance of these

peaks in the spectrum is associated with the first and higher orders scattering process [57]. The peaks used primarily in the characteristics of the single and multi-layers of graphene, up to graphite, are the most intense peaks: G (  1580 cm  1) and 2D, historically denoted by G0 (  2700 cm  1) [58]. In Fig. 7 the Raman spectra of CVD and metallurgical graphene prepared from the liquid phase are presented. Raman spectroscopy studies were performed on the graphene samples transferred onto Si/SiO2 ( 300 nm) substrate. It can be observed that the spectrum of GSM graphene has two characteristic peaks, namely G and 2D. The names of the bands with the characteristic wave number in relation to the literature reports [57] is shown in Table 2. It should be noted that the intensity of these peaks seem to be correct in relation to the graphene. 2D peak has a higher intensity than G peak, a typical ratio for n layers of graphene (n ¼1–6) [59]. I2D/IG intensity ratio is  2.6 (Table 3). According to Ferrari, the ratio for a single layer of graphene should be approximately  4 [58]. From the work of Das et al. [59], where they determined the dependence of the intensity ratio of the peaks on the number of graphene layers, is known that the IG/I2D for monolayer should be 0.2. For the CVD graphene used in the study, the ratio is 0.4 (Table 3). According to the authors of the abovementioned work, this corresponds to approximately bilayer of graphene [59]. In the presented Raman spectrum of CVD layer three additional peaks can be observed: peak D, which in a perfect graphene should not occur. Its presence indicates disorder of carbon atoms or defects such as edges, dislocations, cracks or vacancies [59].

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

7

Fig. 6. Scheme of blisters and wrinkles formation hypothesis: (a) SEM image of hexagonal graphene flake with visible wrinkles due to thermal expansion differences with corresponding scheme, (b) scheme of additional wrinkles formation due to polymer shrinking, (c) air or water trapped between graphene/PMMA and substrate may prevent complete flattening of graphene or cause formation of new wrinkles, (d) top view of graphene after removing of PMMA (consistent with AFM topography images).

Table 1 Roughness parameters for studied samples.

Ra (nm) RMS (nm)

M–G/SiO2

CVD-G/SiO2

SiO2

HOPG

0.5497 0.097 0.6807 0.116

0.366 70.086 0.468 70.112

0.1927 0.036 0.2337 0.044

0.023 7 0.004 0.028 7 0.005

Peak Gn has low intensity and its shape is similar to the n layers (2, 3, 4,…, n) with the exception of graphene n ¼1 for which it is relatively sharp [60]. The last peak, which can be distinguished in this spectrum is the peak 2D0 , which position is fixed and does not depend on the number of layers of graphene [61]. A different spectrum was obtained for metallurgical graphene. In addition to the peaks D, G, Gn, 2D and 2D0 it can be observed the appearance of two others: D0 and D þG. These peaks are related to the defects similarly as D peak. The occurrence of D0 and D þG peaks in the highly oriented crystalline structure (graphene, graphite) is not possible [62]. It can be seen from the presented spectra (Fig. 7) that the investigated structure is a multilayer graphene. This is evidenced by the intensity of the peaks G and 2D. IG/I2D intensity ratio is 1.3 and according to work of Das et al., it means that graphene consists of about 23 layers [59]. A multilayer nature of the graphene is also supported by the shifts of 2D peak towards higher and Gn peak towards lower wavenumber values relative to the CVD graphene (  11.5 and 2.9 cm  1, respectively). In the work already cited [59] in addition to changes in the relative

peak intensities the shape of the 2D peak with increasing number of layers of graphene changed. It became asymmetric and it could be decomposed into two peaks 2D1 and 2D2 [59,60]. Raman spectrum of M–G/SiO2 has a symmetric peak slightly broadened with respect to the CVD graphene spectrum of about 10 cm  1. Similar spectrum is presented in the work of Pimenta et al., where the studied structure is referred to as nanographite [62]. In the studied M–G/SiO2 the ratio of peak intensities I2D/IG and I2D0 /IG are 0.8 and 0.05, respectively, and compared to CVD-G/SiO2, this ratio decreased three times and twice, respectively (Table 3). This indicates a significant degree of defects in the tested structure [59]. Given the above, it appears that in the case of CVD graphene it is a structure composed of about 1–2 layers with a low degree of defects. However, in the case of the structure obtained in the metallurgical process from the liquid phase (M–G/SiO2) it is a multilayer crystalline graphene structure with a high degree of disorder and defects. The Raman spectroscopy results are in agreement with thickness measurement of CVD graphene described earlier in this work (Section 3.2). However the number of layers of metallurgical graphene, as calculated from specific Raman peaks ratio, is inconsistent with the AFM thickness measurement. It is rather build of lower number of layers (similar to that from AFM analysis) as the shape of the specific Raman peaks for M–G/SiO2 sample differs greatly from the “typical” CVD multilayer graphene spectra. Thus it is not possible to associate precisely the specific peaks ratio with the number of layers of

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

8

Fig. 7. Raman spectra of CVD and metallurgical graphene (right) and laser spot location during acquisition of Raman spectra (left).

Table 2 Names of Raman peaks identified in the studied graphene structures with their specific frequencies. ω [cm  1]

Peak name

D G G0 Gn 2D D þG 2D0 1

CVD-G/SiO2

M–G/SiO2

G1

1349.1 1585.4 – 2462.0 2681.6 – 3249.0

1349.0 1586.3 1622.6 2459.1 2693.1 2937.2 3241.8

1350 1585 1620 2450 2700 2935 3240

E laser¼ 2.41 eV [45].

Table 3 FWHM values and ratios of typical peaks calculated on the basis of Raman spectra deconvolution. Peak name

D G G0 Gn 2D D þG 2D0

CVD-G/SiO2

M–G/SiO2

FWHM [cm  1]

IG/ I2D

I2D/ IG

I2D0 / IG

FWHM [cm  1]

29.6 21.1 – 42 45 – 27

0.4

2.6

0.1

56 35 28 38 55 139 66

IG/ I2D 1.3

I2D/ IG

I2D0 / IG

0.8

0.05

metallurgical graphene. Another cause of the difference between AFM and Raman spectroscopy results could be relatively high thickness inhomogeneity over entire sample as well as different area of analysis of both techniques. 3.6. Friction behaviour To calculate the lateral force acting on the tip in the force units (e.g. nN) instead of the electrical signal only (e.g. mV) the calibration of the lateral response of the system was performed. Many options of in situ calibration can be found in the literature (e.g. [63–67]) and most of them are based on the method developed by Ogletree [32], which employs a standard calibration grid

with well-defined slopes. Although for comparison among samples it may be sufficient to measure the lateral forces by the electrical response (in mV), the application of this method to calibrate the lateral deflection of the tip was also attempted. The procedure for the estimation of the average friction coefficient consists of few steps. In order to get reliable information, the measurement of the lateral forces at different load is needed. In order to fulfil this requirement, the whole image was divided into 10 segments where the normal load was progressively increased. In order to evaluate the friction contribution to the total lateral force, the subtraction of lateral forces profiles recorded in both directions, forward and backward, is needed. By this procedure, the lateral forces originated from topographic effects should be minimized, as they cannot be totally avoided [68]. From the subtraction of each pair of segments, one datum of lateral force was calculated, yielding thus 10 pairs of data “normal load–lateral force” for each image. The slope of the plot of lateral force versus normal load is proportional to the friction coefficient of each sample. In Fig. 8 the images of the tips used for LFM measurements are collected. The silicon tip will wear faster than the one coated with several nanometres thick DLC. It is clearly seen on the images of the tips after the set of LFM tests. The tip radius of silicon tip is bigger compared to the new one; however, it is not clear whether the Si tip apex is worn or contaminated). Unfortunately, the magnification of the SEM image did not allow more accurate evaluation of the tip radii. Bigger tip radius may result in bigger contact area while scanning the surface, thus introducing errors in the friction force calculation. In case of DLC coated tip the wear is negligible (for this image magnification). Only some residues accumulated during LFM measurement are attached to the tip at some distance from the apex, though not influencing the contact area while scanning in contact mode. In addition to SEM imaging of the tip wear, its radius calculation based on deconvolution procedure was performed. Before and after set of friction tests we scanned a silicon calibration grid TGG1 in a contact mode and the acquired images were further used for blind tip reconstruction. The tip estimation was done with the use of SPIP 6 software (Image Metrology A/S). The obtained results showed similar wear for both tip materials. However, it should be noted that in case of Si tip there is a large discrepancy between nominal (8 nm) and estimated value before friction test (22 nm). It may suggest that the tip radius may wear faster during the first few scans. After the set of friction tests the radius was 30 nm. In case of DLC tip the initial value (before friction tests) is similar to

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

9

Fig. 8. SEM images of the AFM tips used in friction measurements, Si and DLC coated, before and after the whole set of tests.

the nominal one, 27 nm and 20 nm, respectively. The calculated radius after the friction tests equalled to 36 nm. In Fig. 9 the representation of the lateral force vs. applied normal load for the graphene samples (M–G/SiO2, CVD-G/SiO2) and their substrate (SiO2) for selected examples of single LFM test is displayed. Two different AFM probe materials, Si (Fig. 9a) and DLC (Fig. 9b) are considered. In these graphs, higher slope means higher friction coefficient. At first glance, it can be observed that the SiO2 substrate has the highest friction coefficient in both cases (0.31 and 0.23 for Si and DLC coated probe, respectively). Amongst two graphene samples the lower slope (friction coefficient) for CVD-G/SiO2 is noticed. The horizontal axis on both graphs (normal load) represents normal load applied on the AFM tip excluding adhesion force between the tip and scanned surface. In most cases the tests were conducted slightly above adhesion-only regime. The adhesion-only regime was identified by force spectroscopy measurements prior to each test. After that, knowing the value of pulloff force (according to retract curve), we conducted experiments with the normal load equal or near zero (zero cantilever deflection). According to force spectroscopy measurements, similar trend can be found independently on the tip material, i.e. the order of samples due to the increasing value of adhesion is as follows: SiO2 substrate, CVD graphene, metallurgical graphene and HOPG reference. However, many factors can influence the adhesion force, though these findings should not be overly interpreted. In Fig. 10 the summary of friction measurements for all investigated samples is presented. The results base on the average of individual LFM tests performed for each sample and two tip materials. It can be noted that the highest coefficient of friction was obtained for SiO2 substrate (0.320 70.015 and 0.2347 0.016 for Si and DLC-coated tip, respectively). Higher value of friction coefficient of SiO2 substrate for silicon tip can be explained by the fact that SiO2 has higher chemical affinity to silicon than DLC, thus stronger bonding for SiO2–Si couple may occur, leading to the increase of friction between these materials. Totally inverse trend, compared to SiO2, is observed for the friction measurements of carbon materials: both graphene samples and pyrolytic graphite

Fig. 9. The representation of the lateral force vs. applied normal load for the graphene samples and SiO2 substrate acquired using (a) Si and (b) DLC-coated tip. In the graph legend (in the brackets) the values of pull-off force are provided.

reference. This could be explained by the same phenomena— stronger affinity of carbon materials to DLC coating than to silicon, though higher friction for the graphene/graphite-DLC couples is expected. Taking into account graphene samples only, which coefficients of friction are located between SiO2 and HOPG, the metallurgical graphene COF is higher than CVD graphene. In case

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

10

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Fig. 10. Friction coefficient (bars) and roughness (dashed line) for studied graphene samples, SiO2 substrate and reference sample (HOPG), acquired for two different AFM probe materials (Si and DLC).

of silicon tip the value of the coefficient is doubled compared to CVD-G/SiO2 (0.1177 0.013 and 0.0577 0.020, respectively), while for DLC-coated tip this ratio is about 1.5 (0.16270.007 and 0.109 70.021, respectively). Finally, the lowest values obtained for HOPG (0.013 70.002 and 0.0237 0.013 for Si and DLC-coated tip, respectively) appears near to others reported in literature (0.0012 [69], 0.001–0.01 [70], 0.004 [71,72] and 0.006 [73,74]). Although the main cause of the difference in the friction coefficients between both graphene samples on SiO2 seems to be the structural disorder due to manufacturing and transfer processes, it may also originate from their different thickness. The interaction between a single graphene sheet and the substrate differs from that between two graphene layers. Lee et al. [19] reported that the friction force for graphene decreased as the number of layers increased. Another confirmation of this frictional behaviour have been found in [75]. The authors described the elastic properties and frictional characteristics of suspended graphene with various number of layers (from 1 to 4) using an AFM. The authors confirmed that the frictional force between an AFM tip and graphene does not depend on the presence of a substrate and decreases with thickness. The friction on a single-layer graphene grown on SiC was found to be two times greater than that on a bilayer film [76]. However in most of the other papers it is reported that the thickness dependent friction does not originate from the difference in structural properties or lateral contact stiffness, but the graphene adhesion to the substrate plays an important role in the frictional behaviour of graphene. Li et al. [23] reported that both graphene on SiO2 substrates and freely suspended exhibited increasing friction as the number of graphene layers decreased due to lower bending stiffness of the thinner graphene samples, leading to greater puckering effect and frictional resistance. This trend was not observed for graphene strongly bonded to the substrate (e.g. mica), where the puckering effect will be suppressed and no thickness dependence should be observed. Similar results were reported by [77]. The related models proposed by Li et al. [23] and Cho et al. [77] are consistent with the results of other studies [15,19,23,75,76,78]. In our results we cannot find such relation between monolayer CVD graphene and multilayer metallurgical graphene. In this case we observe lower friction for monolayer graphene. The explanation for this could be a large number of defects created during metallurgical graphene manufacturing, so direct comparison between samples is rather not possible. The phenomenon of increase in the coefficient of friction due to the defects in graphene is reported in literature [79,80]. Additional experiments, leading to creation

of metallurgical graphene with varying numbers of layers are needed. In the work of Penkov et al. [81] the overview of tribological properties of graphene in nano- and micro-scale is presented. In one of the summary table they gathered data of experimental results on nano-tribology of graphene. Depending on the deposition parameters, substrate material, friction test conditions (AFM tip shape and material, normal load) the variety of friction coefficient is very broad, ranging from 0.02 up to 1. Despite the relatively broad range of friction coefficients there are only single results for the similar conditions as presented in our paper, however they concern exfoliated graphene on SiO2 only. In case of graphene/SiO2 sample the friction coefficient for the Si tip ranges from 0.15 to 0.6, while for DLC tip it is from 0.4 to 1. For both configurations the results presented in our work are below these values (0.057–0.117 and 0.109–0.162 for Si and DLC-coated tip, respectively). These differences may arise mainly from the deposition process but as well it could be related to the differences in friction measurement procedure, e.g. measuring at single [15,75] or ramped load [53], this paper. Lin et al. [21] investigated the friction of exfoliated multilayer graphene. It was found that graphene films exhibited much lower friction than a bare Si surface when the applied loads ranged from 3 to 30 nN. In case of our results, the friction coefficient for metallurgical graphene was 2.7 and 1.4 times lower than bare SiO2 (for Si and DLC tip, respectively). CVD graphene exhibited even lower friction, 5.6 and 2.1 times lower than SiO2 (for Si and DLC tip, respectively). After the friction tests surface topography imaging was also performed, although no visible signs of wear were noticed. This can be explained by the fact that the applied normal loads during the friction tests of graphene samples were relatively low. There are not many reports regarding wear of graphene. Lin et al. [21] also investigated the wear characteristics of multilayer graphene coated on Si wafer. According to the results of the wear tests, they found that 5 mN of normal load after 100 cycles of sliding was considered to be the critical condition that led to the removal a single layer of graphene. In our study the applied load was 2–3 orders of magnitude lower and no cyclic tests were performed as the measured surface was scanned only once. According to Ra parameter values acquired for all samples, it can be concluded that friction properties are not directly determined by the roughness. Although all carbon material samples (M–G/SiO2, CVD-G/SiO2 and HOPG) show a clear trend, decrease of friction coefficient with decrease of roughness, such relationship is not valid for bare SiO2 substrate, displaying the second lowest roughness value. However, as mentioned earlier in this work, it should be reminded that the tip may wear during the measurement (Fig. 8) thus changing the contact area, so these values of friction coefficient should be considered with caution. Besides, it is well known that there is a large uncertainty in the calibration of the lateral forces [82].

4. Conclusions In this work friction behaviour of metallurgical and CVD graphene transferred onto SiO2 substrates were investigated using lateral force microscopy with two AFM probes materials. Conducted experiments showed substantial decrease of friction coefficient for both graphene samples compared to the bare SiO2 substrate. The differences in friction values for measurements using different AFM tip materials are related to their affinity to the scanned surface, thus for all carbon-based samples higher friction is obtained for DLC-coated compared to silicon tip. Lower value of friction for CVD graphene as compared to metallurgical

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

graphene is mainly related with large number of defects created during manufacturing process of the latter one. Additional experiments, leading to creation of metallurgical graphene with varying numbers of layers are needed, in order to establish the thickness dependent relation.

[27]

[28]

Acknowledgments The authors would like to thank Dr. Sylwia Kotarba (Lodz Regional Park of Science and Technology Ltd.) for Raman spectroscopy of the graphene samples. The study is financed by The National Centre for Research and Development as part of the program GRAF-TECH, project acronym GraphRoll, project title: “Graphene nanocomposite for the reversible hydrogen storage”, agreement number GRAF-TECH/NCBR/07/24/2013.

[29] [30]

[31]

[32]

References [1] Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y, Dubonos SV, et al. Electric field effect in atomically thin carbon films. Science 2004;306:666–9. [2] Novoselov KS, Jiang D, Schedin F, Booth TJ, Khotkevich VV, Morozov SV, et al. Two-dimensional atomic crystals. Proc Natl Acad Sci USA 2005;102:10451–3. [3] Chen D, Tang L, Li J. Graphene-based materials in electrochemistry. Chem Soc Rev 2010;39:3157–80. [4] Soldano C, Mahmood A, Dujardin E. Production, properties and potential of graphene. Carbon NY 2010;48:2127–50. [5] Kim KS, Zhao Y, Jang H, Lee SY, Kim JM, Kim KS, et al. Large-scale pattern growth of graphene films for stretchable transparent electrodes. Nature 2009;457:706–10. [6] Li X, Zhu Y, Cai W, Borysiak M, Han B, Chen D, et al. Transfer of large-area graphene films for high-performance transparent conductive electrodes. Nano Lett 2009;9:4359–63. [7] Sutter PW, Flege J-I, Sutter EA. Epitaxial graphene on ruthenium. Nat Mater 2008;7:406–11. [8] Lee SW, Mattevi C, Chhowalla M, Sankaran RM. Plasma-assisted reduction of graphene oxide at low temperature and atmospheric pressure for flexible conductor applications. J Phys Chem Lett 2012;3:772–7. [9] Rümmeli MH, Rocha CG, Ortmann F, Ibrahim I, Sevincli H, Börrnert F, et al. Graphene: piecing it together. Adv Mater 2011;23:4471–90. [10] Li X, Cai W, An J, Kim S, Nah J, Yang D, et al. Large-area synthesis of highquality and uniform graphene films on copper foils. Science 2009;324:1312–4. [11] Geng D, Wu B, Guo Y, Huang L, Xue Y, Chen J, et al. Uniform hexagonal graphene flakes and films grown on liquid copper surface. Proc Natl Acad Sci USA 2012;109:7992–6. [12] Kula P, Atraszkiewicz R, Dybowski K, Pietrasik R, Rzepkowski A, Kazimierski D. Method of graphene fabrication from liquid metal. PL-399096; 2012. [13] Kula P, Pietrasik R, Atraszkiewicz R, Dybowski K, Modrzyk W, Zawistowski PZ, et al. Method for separating graphene from the liquid forming matrix. WO 035264 A1, 2014 2014. [14] Berman D, Erdemir A, Sumant AV. Graphene: a new emerging lubricant. Mater Today 2014;17:31–42. [15] Lee C, Li Q, Kalb W, Liu X-Z, Berger H, Carpick RW, et al. Frictional characteristics of atomically thin sheets. Science 2010;328:76–80 (80-). [16] Ye Z, Tang C, Dong Y, Martini A. Role of wrinkle height in friction variation with number of graphene layers. J Appl Phys 2012;112:116102. [17] Smolyanitsky A, Killgore JP. Anomalous friction in suspended graphene. Phys Rev B: Condens Matter 2012;86:125432. [18] Feng X, Kwon S, Park J, Salmeron M. Superlubric sliding of graphene nanoflakes on graphene. ACS Nano 2013:1718–24. [19] Lee H, Lee N, Seo Y, Eom J, Lee SS. Comparison of frictional forces on graphene and graphite. Nanotechnology 2009;20:325701. [20] Filleter T, McChesney J, Bostwick a, Rotenberg E, Emtsev K, Seyller T, et al. Friction and dissipation in epitaxial graphene films. Phys Rev Lett 2009;102:086102. [21] Lin L-Y, Kim D-E, Kim W-K, Jun S-C. Friction and wear characteristics of multilayer graphene films investigated by atomic force microscopy. Surf Coat Technol 2011;205:4864–9. [22] Deng Z, Smolyanitsky A, Li Q, Feng X, Cannara RJ. Adhesion-dependent negative friction coefficient on chemically modified graphite at the nanoscale 2012:11. [23] Li Q, Lee C, Carpick RW, Hone J. Substrate effect on thickness-dependent friction on graphene. Phys Status Solidi 2010;247:2909–14. [24] Kula P, Pietrasik R, Dybowski K, Atraszkiewicz R, Szymanski W, Kolodziejczyk L, et al. Single and multilayer growth of graphene from the liquid phase. Appl Mech Mater 2014;510:8–12. [25] Liang X, Sperling BA, Calizo I, Cheng G, Hacker CA, Zhang Q, et al. Toward clean and crackless transfer of graphene. ACS Nano 2011;5:9144–53. [26] Kula P, Pietrasik R, Dybowski K, Atraszkiewicz R, Kaczmarek L, Szymanski W, et al. The growth of a polycrystalline graphene from a liquid phase. In: Tech.

[33]

[34]

[35] [36] [37]

[38] [39] [40] [41] [42]

[43] [44]

[45]

[46]

[47] [48] [49]

[50]

[51]

[52] [53]

[54]

[55] [56]

11

Proc. 2013 NSTI Nanotechnol. Conf. Expo, NSTI-Nanotech 2013, 1; 2013, p. 210–212. Kula P, Olejnik J, Heilman P, Olejnik J.O.Z. Hydrocarbon gas mixture for underpressure carburizing steel products, e.g. parts of machines, contains carbon carrier in the form of two unsaturated hydrocarbons. WO2004035853-A1; EP1558780A1; US2006102254-A1; EP1558780-B1; DE60315693-E; ES2291745-T3; DE60315693-T2; US7513958-B2; 2002. Atraszkiewicz R, Dybowski K, Klinowski B.M, Kula P, Modrzyk W, Pietrasik R, et al. Separating graphene film from liquid graphene-forming matrix involves upon completion of heating/adding hydrocarbon mixture, and during cooling, sliding mould frame over surface of liquid forming matrix, and lifting mould frame above matrix. WO2014035264-A1; WO2014035264-A8; PL400511-A1; 2012. Her M, Beams R, Novotny L. Graphene transfer with reduced residue. Phys Lett A 2013;377:1455–8. Suk JW, Kitt A, Magnuson CW, Hao Y, Ahmed S, An J, et al. Transfer of CVD-grown monolayer graphene onto arbitrary substrates. ACS Nano 2011;5:6916–24. Gomez De Arco L, Zhang Y, Schlenker CW, Ryu K, Thompson ME, Zhou C. Continuous, highly flexible, and transparent graphene films by chemical vapor deposition for organic photovoltaics. ACS Nano 2010;4:2865–73. Ogletree DF, Carpick RW, Salmeron M. Calibration of frictional forces in atomic force microscopy. Rev Sci Instrum 1996;67:3298. Robertson AW, Bachmatiuk A, Wu YA, Schäffel F, Rellinghaus B, Büchner B, et al. Atomic structure of interconnected few-layer graphene domains. ACS Nano 2011;5:6610–8. Pirkle a, Chan J, Venugopal a, Hinojos D, Magnuson CW, McDonnell S, et al. The effect of chemical residues on the physical and electrical properties of chemical vapor deposited graphene transferred to SiO2. Appl Phys Lett 2011;99:122108. Ishigami M, Chen JH, Cullen WG, Fuhrer MS, Williams ED. Atomic structure of graphene on SiO2. Nano Lett 2007;7:1643–8. Dan Y, Lu Y, Kybert NJ, Luo Z, Johnson ATC. Intrinsic response of graphene vapor sensors. Nano Lett 2009;9:1472–5. Cheng Z, Zhou Q, Wang C, Li Q, Fang Y. Toward intrinsic graphene surfaces: a systematic study on thermal annealing and wet-chemical treatment of SiO2supported graphene devices. Nano Lett 2011;11:767–71. Bolotin KI, Sikes KJ, Jiang Z, Klima M, Fudenberg G, Hone J, et al. Ultrahigh electron mobility in suspended graphene. Solid State Commun 2008;146:351–5. Moser J, Barreiro a, Bachtold a. Current-induced cleaning of graphene. Appl Phys Lett 2007;91:163513. Algara-Siller G, Lehtinen O, Turchanin A, Kaiser U. Dry-cleaning of graphene. Appl Phys Lett 2014;104:153115. Liu Z, Bøggild P, Yang J, Cheng Y, Grey F, Liu Y, et al. A graphite nanoeraser. Nanotechnology 2011;22:265706. Jalilian R, Jauregui La, Lopez G, Tian J, Roecker C, Yazdanpanah MM, et al. Scanning gate microscopy on graphene: charge inhomogeneity and extrinsic doping. Nanotechnology 2011;22:295705. Goossens a M, Calado VE, Barreiro a, Watanabe K, Taniguchi T, Vandersypen LMK. Mechanical cleaning of graphene. Appl Phys Lett 2012;100:073110. Reina A, Son H, Jiao L, Fan B, Dresselhaus MS, Liu Z, et al. Transferring and identification of single- and few-layer graphene on arbitrary substrates. J Phys Chem C 2008;112:17741–4. Dean CR, Young a F, Meric I, Lee C, Wang L, Sorgenfrei S, et al. Boron nitride substrates for high-quality graphene electronics. Nat Nanotechnol 2010;5: 722–6. Nourbakhsh A, Cantoro M, Klekachev A, Clemente F, Sorée B, van der Veen MH, et al. Tuning the Fermi level of SiO2-supported single-layer graphene by thermal annealing. J Phys Chem C 2010;114 (6894–900). Paredes JI, Villar-Rodil S, Martínez-Alonso A, Tascón JMD. Graphene oxide dispersions in organic solvents. Langmuir 2008;24:10560–4. Lin Y-C, Lu C-C, Yeh C-H, Jin C, Suenaga K, Chiu P-W. Graphene annealing: how clean can it be? Nano Lett 2012;12:414–9. Kumar K, Kim Y-S, Yang E-H. The influence of thermal annealing to remove polymeric residue on the electronic doping and morphological characteristics of graphene. Carbon N Y 2013;65:35–45. Calado VE, Schneider GF, AMMG Theulings, Dekker C, Vandersypen LMK. Formation and control of wrinkles in graphene by the wedging transfer method. Appl Phys Lett 2012;101:103116. Ismach A, Druzgalski C, Penwell S, Schwartzberg A, Zheng M, Javey A, et al. Direct chemical vapor deposition of graphene on dielectric surfaces. Nano Lett 2010;10:1542–8. Zhu W, Low T, Perebeinos V, Bol AA, Zhu Y, Yan H, et al. Structure and electronic transport in graphene wrinkles. Nano Lett 2012;12:3431–6. Park H, Brown PR, Bulović V, Kong J. Graphene as transparent conducting electrodes in organic photovoltaics: studies in graphene morphology, hole transporting layers, and counter electrodes. Nano Lett 2012;12:133–40. Temmen M, Ochedowski O, Schleberger M, Reichling M, Bollmann TRJ. Hydration layers trapped between graphene and a hydrophilic substrate. New J Phys 2014;16:053039. He K. Characterization of graphene-substrate interactions using scanning tunneling microscopy. University of Illinois at Urbana-Champaign; 2013. Zhang F, Kibria MG, Cormier K, Howlader M. Surface and interface characterization of sequentially plasma activated silicon, silicon dioxide and germanium wafers for low temperature bonding applications. ECS Trans 2010;33:329–38.

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i

12

L. Kolodziejczyk et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

[57] Jorio A. Raman spectroscopy in graphene-based systems: prototypes for nanoscience and nanometrology. ISRN Nanotechnol 2012 (2012:1–16). [58] Ferrari AC, Meyer JC, Scardaci V, Casiraghi C, Lazzeri M, Mauri F, et al. Raman spectrum of graphene and graphene layers. Phys Rev Lett 2006;97:187401. [59] Das A, Chakraborty B, Sood AK. Raman spectroscopy of graphene on different substrates and influence of defects. Bull Mater Sci 2007;31:6. [60] Yoon D, Moon H, Cheong H, Choi JS, Choi JA, Park BH. Variations in the Raman spectrum as a function of the number of graphene layers. J Korean Phys Soc 2009;55:1299–303. [61] Graf D, Molitor F, Ensslin K, Stampfer C, Jungen A, Hierold C, et al. Spatially resolved Raman spectroscopy of single- and few-layer graphene. Nano Lett 2007;7:238–42. [62] Pimenta MA, Dresselhaus G, Dresselhaus MS, Cançado LG, Jorio A, Saito R. Studying disorder in graphite-based systems by Raman spectroscopy. Phys Chem Chem Phys 2007;9:1276–91. [63] Jones R, Pollock HM, Geldart D, Verlinden-Luts A. Frictional forces between cohesive powder particles studied by AFM. Ultramicroscopy 2004;100:59–78. [64] Tocha E, Schönherr H, Vancso GJ. Quantitative nanotribology by AFM: a novel universal calibration platform. Langmuir 2006;22:2340–50. [65] Varenberg M, Etsion I, Halperin G. An improved wedge calibration method for lateral force in atomic force microscopy. Rev Sci Instrum 2003;74:3362. [66] Liu E, Blanpain B, Celis JP. Calibration procedures for frictional measurements with a lateral force microscope. Wear 1996;192:141–50. [67] Martínez-Martínez D, Kolodziejczyk L, Sánchez-López JC, Fernández A. Tribological carbon-based coatings: an AFM and LFM study. Surf Sci 2009;603:973–9. [68] Bhushan B. Nanoscale tribophysics and tribomechanics. Wear 1999;225229:465–92. [69] Schwarz U, Zwörner O, Köster P, Wiesendanger R. Quantitative analysis of the frictional properties of solid materials at low loads. I. Carbon compounds. Phys Rev B: Condens Matter 1997;56:6987–96. [70] Liu E, Blanpain B, Celis J-P, Roos JR. Comparative study between macrotribology and nanotribology. J Appl Phys 1998;84:4859.

[71] Arias DF, Marulanda DM, Baena AM, Devia A. Determination of friction coefficient on ZrN and TiN using lateral force microscopy (LFM). Wear 2006;261:1232–6. [72] Buzio R, Gnecco E, Boragno C, Valbusa U. Friction force microscopy investigation of nanostructured carbon films. Carbon N Y 2002;40:883–90. [73] Ruan J-A, Bhushan B. Atomic-scale friction measurements using friction force microscopy: Part I—General principles and new measurement techniques. J Tribol 1994;116:378. [74] Ruan J-A, Bhushan B. Frictional behavior of highly oriented pyrolytic graphite. J Appl Phys 1994;76:8117. [75] Lee C, Wei X, Li Q, Carpick R, Kysar JW, Hone J. Elastic and frictional properties of graphene. Phys Status Solidi 2009;246:2562–7. [76] Filleter T, Bennewitz R. Structural and frictional properties of graphene films on SiC(0 0 0 1) studied by atomic force microscopy. Phys Rev B: Condens Matter 2010;81:155412. [77] Cho D-H, Wang L, Kim J-S, Lee G-H, Kim ES, Lee S, et al. Effect of surface morphology on friction of graphene on various substrates. Nanoscale 2013;5:3063–9. [78] Sasaki N, Okamoto H, Itamura N, Miura K. Atomic-scale friction of monolayer graphenes with armchair- and zigzag-type edges during peeling process. E-J Surf Sci Nanotechnol 2010;8:105–11. [79] Shin YJ, Stromberg R, Nay R, Huang H, Wee ATS, Yang H, et al. Frictional characteristics of exfoliated and epitaxial graphene. Carbon NY 2011;49:4070–3. [80] Won M-S, Penkov OV, Kim D-E. Durability and degradation mechanism of graphene coatings deposited on Cu substrates under dry contact sliding. Carbon N Y 2013;54:472–81. [81] Penkov O, Kim H-J, Kim H-J, Kim D-E. Tribology of graphene: a review. Int J Precis Eng Manuf 2014;15:577–85. [82] Riedo E, Chevrier J, Comin F, Brune H. Nanotribology of carbon based thin films: the influence of film structure and surface morphology. Surf Sci 2001;477:25–34.

Please cite this article as: Kolodziejczyk L, et al. Frictional behaviour of polycrystalline graphene grown on liquid metallic matrix. Tribology International (2014), http://dx.doi.org/10.1016/j.triboint.2014.12.003i