Synthesization of CNT on Surface of C/C Composites to Stabilize the Coefficient of Friction through a Wide Environmental Temperature Range

The purpose of this study is to stabilize the coefficient of friction of C/C composites through wide environmental temperature range by synthesizing the carbon nanotubes (CNTs) on its surface. CNTs were synthesized on the surface of carbon-carbon (C/C) composites by the thermal chemical vapor deposition method. The coefficient of mutual friction between two flat specimens of both modified and unmodified C/C composites were measured at a constant rotational speed under constant applied pressure by using disc-type test apparatus. The effect of CNT synthesization on the mutual friction coefficient and its dependence on test temperature were discussed. Frictional coefficient was high and also stable when the CNTs were synthesized on the surface while the test temperature was changed. The morphology evaluation of the frictional surface after the test suggested that synthesization of CNTs on a frictional surface of a C/C composite was effective in changing the wear mode for stability of the coefficient of friction according to the change of environmental temperature.


Introduction
A carbon-carbon (C/C) composite is known as one of the excellent composite materials because of its high mechanical properties, despite its low density, and also because of its high durability at high temperature environments (Manocha, 2003;Bussiba et al., 2008).Because of their excellent properties, C/C composites are applied not only to disc brakes of aerospace vehicles (Buchgraber, 2003;Don et al., 2009) but also to those of commercial vehicles such as motorcycles.Previous studies revealed that when the environmental temperature is low, the coefficient of mutual friction of C/C composites is relatively low than that at high temperature (Yen et al., 1995(Yen et al., , 1996;;Gouider et al., 2004;Kasem et al., 2009).From the view point of reliable and safe performance, such dependence of frictional a coefficient on temperature should be cancelld for commercial vehicle application, in order to avoid the brake torque variation during braking (Obunai et al., 2006(Obunai et al., , 2009)).In order to cancel the dependence of the frictional coefficient on temperature, metal infiltration methods have been applied to C/C composites (e.g., C/C-SiC and C/C-Al) (Krenkel et al., 2002;Wang et al., 2006;Stadler et al., 2007;Zhang et al., 2007).However, after applying metal infiltration, the coefficient of mutual friction decreases from that of original C/C composite, although its dependence on temperature was canceled.In order to improve the frictional behavior of the C/C composite, it was recognized that a novel approach would be needed.The purpose of this study is to propose an effective technique for improving the coefficient of friction between the surfaces of a C/C composite at low environmental temperatures so that this coefficient will be stable over a wide range of environmental temperatures.
To modify their surfaces, in this study, carbon nanotubes (CNTs) were, first, synthesized on C/C composites by thermal chemical vapor deposition (CVD).A commercially available C/C composite was used as the base material for the synthesization of CNTs.The thin layers of CoO on MoO 3 were prepared on the surface C/C composite by a dip-coat method to promote CNT synthesization.The heated C/C composite with thin layers was exposed to alcohol vapor.In this study, the time taken for introducing alcohol vapor into the oven was defined as the process time.The coefficient of mutual friction between two flat specimens of the modified and unmodified C/C composites was measured at a constant number of revolutions under constant pressure by the disc-type test.The effect of CNT synthesization on the coefficient of mutual friction and its dependence on test temperature was discussed.In order to explain the reason for the improvement in the coefficient of mutual friction of modified C/C composites by synthesization of CNTs, the wear mode of modified C/C composites was investigated.

Material
In order to confirm the effectiveness of proposed technique, the commercially available C/C composite (AC250: Across) was used as the base material for the synthesization of CNTs.Small plates, 20 mm square and 3 mm thick, were cut out.

Support of Catalytic Metal
In this study, a thin layer of CoO was prepared as the top surface of catalytic metal for CNT synthesization.Before the CoO layer was prepared, a base catalytic metal layer consisting by MoO 3 was formed on the C/C composite.The small plates of the C/C composite described in the previous section were dipped into a solution of Mo-acetate, after heating at 400°C for 1 h in an electric oven filled with air, to eliminate impurities on the surface.To uniformly spread the catalytic metal on the C/C composite, the plate was carefully lifted from the solution at a constant speed of 30 mm/min.Then, the plates were heated again in the oven at 400°C for 1 h to form a thin layer of MoO 3 by oxidizing Mo-acetate on the surface.The same procedures were executed to support CoO on the surface of MoO 3 .The thicknesses of CoO and MoO 3 were around 1 m.

Thermal CVD Procedure
In this study, CNTs were synthesized by the thermal CVD method.Figures 1 and 2 show the thermal CVD equipment and the thermal CVD process used in this study, respectively.The specimens were heated in reductive gas (Ar:H 2 -3%) at the quartz tube, shown in the figure, from 25 (i.e., RT ) to 900°C at constant pressure of 5 kPa(abs), to deoxidize the catalytic metals on the surface of the specimen.When the temperature reaches 900°C, alcohol vapor (C 2 H 5 OH) was introduced into the quartz tube and pressure was kept to 15 kPa, to promote the synthesization of CNTs (Mukul et al., 2010).After synthesization, the specimens were cooled again to RT in an inert atmosphere of Ar gas.In this study, the time spent introducing alcohol vapor into the oven was defined as the process time.

Friction Test
Figure 3 shows the configuration of a disc-type frictional tester.The specimen was set to the groove of the jig and rotated by the AC motor and another fixed specimen was pushed to the disc.The number of rotations of the specimen was controlled by an inverter at 100 rpm (0.1 m/s of sliding speed at effective radius).Constraint displacement was uniformly applied by the screw to apply 400 N (1.0 MPa) of normal force to the fixed specimen; the resultant force was measured by a load cell.The resultant torque was also measured to determine the coefficient of mutual friction between the two specimens.

Micro Hardness Test
The micro hardness of the CNT was measured by a micro hardness tester (SPH-1: Shimadzu) with a triangular pyramidal indenter (Murakami et al., 1997).In this study, the micro hardness was estimated by the diamond indenter applied to the thin film that was formed with the commercially available CNTs pressed under 9.5 MPa at 25°C for 3 h.The indentation force was applied to the CNT film by a diamond indenter until the reaction force increased to 1500 N and maintaining a constant force speed of 85.60 N/s.The maximum load was held for 5 s and then released to zero at 85.60 N/s force speed in the opposite direction.The micro hardness was determined by the following equation: Here, HT 115 , F and d denote the micro-hardness, applied force and diagonal length of the indentation size on the specimen, respectively.To evaluate the result of synthesization, in this study, an effective parameter of C CNT was introduced by authors, as described by the following equation:

Characterization of Synthesized CNTs on C/C Composites
Here, C CNT denotes concentration of the synthesized CNTs per unit surface area, m CVD is mass of specimen after CVD, m R is the mass before CVD, C y is the dissipated mass by CVD process, and A S denotes the surface area of the specimen.The dissipated mass of C y was depend on the hermetic performance of the CVD apparatus.In this study, the C y value was determined by measuring the mass loss during the CVD process without introducing of alcohol vapor.Figure 5 shows the concentration of the synthesized CNTs per unit surface area with respect to process time for synthesization.The Concentration of the synthesized CNTs increased from about 0.0033 to 0.0044 mg/mm 2 with the process time.CNT diameter and length increased with the concentration of synthesized CNTs when the process time of synthesization was extended.Test result showed that when the surface of C/C composite was modified with excessive CNTs (C CNT =0.0044 mg/mm 2 ), the coefficient mutual friction was rapidly decreased as same value as unmodified composites.However, when moderate CNTs (C CNT =0.0039 mg/mm 2 ) were synthesized on surface, the improvement of coefficient of mutual friction was still remaining even after 300 sec.Figure 11 shows the change in the observed average area of the synthesized CNTs in the sliding direction with respect to the concentration.The observed average area increased from 223 to 1154 m 2 with an increase in the concentration of the synthesized CNTs per unit area.These results suggest that the frictional force due to scratching by the CNTs increases with an increase in the concentration of the synthesized CNTs.This study found that the synthesization of CNTs on the frictional surface of C/C composites was effective for changing the wear mode for stability of the coefficient of friction according to the environmental temperature changes.

Conclusions
In this study, the carbon nano tubes (CNTs) were synthesized on the surface of commercially available C/C composites and its effect of mutual friction was investigated.The following conclusions were given.
1) The diameter and length of spike-like CNTs increased according to the concentration of synthesized CNTs per unit surface area when the process time was extended for synthesization.
2) The dependence of the coefficient of mutual friction on the test temperature was almost canceled with an increase in the concentration of synthesized CNTs per unit surface area.
3) The durability of synthesized CNTs were depending on the concentration of synthesized CNTs per unit surface area.
4) The synthesization of CNTs on the frictional surface of C/C composites was effective in changing the wear mode for stability of the coefficient of friction according to change in environmental temperature.

Figure 4
Figure4(a)-(c) shows the state of the CNTs synthesized on the surface of the C/C composites, with each specimen having a different process time.It was confirmed that the CNTs (indicated by white arrows) were synthesized on the C/C composite by the thermal CVD process.SEM observations suggested that both the diameter and length of spike-like CNTs increased according to the process time for the synthesization of the CNTs.The diameters of CNTs were increased from around 80 to 280 nm according to the process time from 15 to 45 min.The lengths of CNTs were also increased from around 2.7 to 4.8 m.

Figure 4 .
Figure 4. SEM photographs of synthesized CNTs with different synthesized process times

Figure 5 .
Figure 5. Concentration of synthesized CNTs per unit area with respect to synthesized time3.2Effect of CNT Synthesization on Coefficient of FrictionFigure6shows the coefficient of mutual friction of the modified C/C composites with respect to the concentration of the synthesized CNTs per unit surface area at RT condition.For comparison, the figure also shows basic data of the coefficient of mutual friction for unmodified C/C composites, indicated by a dashed line (=0.11).At RT, it was confirmed that the coefficient of mutual friction was improved from 0.17 to 0.32 by the synthesization of the CNTs on the frictional surface of the C/C composite.

Figure 6 .
Figure 6.Coefficient of mutual friction of modified C/C composites with respect to concentration of synthesized CNTs per unit area

Figure 7 .
Figure 7. Coefficient of mutual friction of C/C composites with respect to test temperature

Figure 9 .
Figure 9.Comparison of hardness of CNT and C/C composites

Figure 10 .
Figure 10.Worn surface of unmodified/modified C/C composites after test performed at RT/HT