Carbon nanotubes have been discovered by Iijima [1] and gained great interest due to its unique and superb properties,such as effective field emission characteristics[2,3], capability for the storage of large amount of hydrogen [4-6], high modulus [7,8], and structural diversities that make it possible for the band gap engineering [9,10]. Potential applications includes, hydrogen storage [11], chemical sensors[12,13], nanoelectronics devices[14,15]and flat panel field emission displays [16-18]. To date, the most widely used methods for preparing nanotubes include laser vaporization of graphite [19], arcdischarge[20] and chemical vapor deposition [21]. Among these popular techniques, electric arc discharge is having some advantages over other techniques such as the temperature range of synthesis of CNT is relatively higher which make them more graphitic in nature where as in laser vaporization of graphite process require expensive laser source. In electric arc discharge technique, graphite electrodes are used and arc is generated by passing high order of current and the cathode deposit which has large amount of carbon nanotubes. However,in this process of arcing a large amount of unnecessary carbonaceous particles such as graphite particles, amorphous carbon and other carbonaceous impurities are also present in the deposit which required to be purified for various applications of CNTs. In addition, defects within a CNTs structure reduce the effective overlap and therefore, decrease the carrier density,which can lead to an increase in the materials electrical resistivity [22]. Various methods have been reported, including chemical oxidation [23], thermal oxidation [24-26], filtration [27], and chromatography [28,29]. These processes are able to remove impurities to some extent, but not completely due to certain limitations. Therefore, various purification methods were tried out. Intercalating pristine carbon nanotubes and thermal annealing the sample under controlled flow of oxygen resulted in purified CNT with minimal weight loss. There are very few papers related to purification of MWCNTs synthesized by arc discharge [30-37]. Although there are some already published papers about purification but most of the groups are working on CVD synthesized CNT for purification of SWNTs for the removal of catalysts. Thus, this work is an attempt was made to purify MWCNTs synthesized by arc discharge.

Multiwall Carbon Nanotubes (MWCNTs) were synthesized by arc discharge process as described elsewhere [38]. The double walled stainless chamber has a volume of approximately 200 liters. The electrodes and the chamber walls are cooled by the high pressure flowing water. Two different shapes of graphite were used as electrodes, one of the electrodes (anode) is rod type where as the other electrode (cathode) is flat surface of higher diameter to facilitates uniform arcing. The anode is attached to a step motor manipulator, which controls the gap between two graphite electrodes by moving forward or backward. The chamber is pumped out to a base pressure of 100mTorr and then Helium gas is introduced up to the desired pressure. Voltage was kept constant at 40 V and varied DC current was observed during the entire experiment till the anode consumed completely. The gas pressure was kept at 300 Torr [39]. The cathode deposit which is mainly multiwall carbon nanotubes is collected and processed for further analysis. For purification of MWCNTs, nanotubes were subjected to different treatments like thermal annealing, chemical treatment followed by thermal treatment without and with controlled flow of oxygen gas.

The cathode deposit which is having mainly MWCNTs and other carbonacoues impurites were subjected for different temperatures in alumina boat, as all the carbon materials including carbon nanotubes are having different oxidation temperatures. The cathode deposit is crushed and kept in a tubular furnace with a heating rate of 5ºC/min. The sample were heat treted at various temperatures ; 450ºC, 550ºC, 650ºC,750ºC & 850ºC for about 45 minutes in normal conditions i.e.without any excess supply of oxygen. Experiments were also conducted by heating the sample in controlled flow of Oxygen at various temperaures 450ºC, 550ºC, 650ºC, 750ºC & 850ºC for 45 minutes. Initially, cathode deposit is crushed and 0.5 gm of pristine CNT is taken in a stainless steel sample holder which is designed to evenly expose the sample to annealing.

Pristine CNT (0.5 g of cathode deposit powder) is taken in a mixture of sulfuric acid and nitric acid (4:1) which was sonicated for 30 min. and allowed for 45 min. The solution was diluted and filtered under vacuum by means of poly vinilidene flouride (PVDF) membrane filter with a pore size of 0.45 microns. The sample retained on the filter was washed twice with water till it reaches to the neutral pH and than subjected to different temperature for thermal annealed at 450ºC, 550ºC, 650ºC, 750ºC and 850ºC respectively with and with out oxygen supply. The samples were analysed by SEM(Model: Hitachi- S-4300SE/N), TGA (STA 409, Netzsch Germany) and BET surface area analysis (Micromeritics ASAP 2020). The chemical treated samples were analysed by FTIR[Model Spectrum GX, Perkin, Germany]. The samples for FTIR were prepared by taking 1 mg of CNT sample mixed thoroughly in 100 mg of KBr in a pestle mortar for about 15 minutes and then pressed in to pellet form in such a way that the pellet should be transparent for better results in FTIR.

Cathode deposit of arc discharge sample is subjected for

different temperature between 450~850^oC. Fig. 1 shows the typical SEM images of the sample of cathode deposit (raw sample) and heat treated at various oxidation temperatures 450~850^oC for 45 min. The raw sample seems to show no carbon nanotubes on the surface because MWNTs were mostly embedded inside the carbonaceous particles. This indicates presence of large amount of impurities, mostly amorphous carbon. With oxidation at 450^oC and 550^oC, some of the carbon particles were removed out of the surface. At 650^oC, most of the impurity particles were removed and the weight loss is about 28.5%. But some more impurities still exist, probably due to insufficient oxygen for all the impurities to completely oxidized. It was observed that oxidation at temperatures more than 650^oC, more carbon nanotubes are

visible as compared to pristine or samples heat treated at lower temperatures than 650^oC. Table 1 shows that the weight loss (%) is gradually increasing with increase in temperature.From this study it was concluded that even though different types of carbons are having different oxidation temperature,alone thermal heating of the deposits sample can not be used for purification of the carbon nanotubes.

Samples were also heat treated at various temperatures(450~850^oC) as discussed in the above case with extra flow of oxygen in a specially designed holder which facilitates uniform heating of the samples and to have more and uniform access of oxygen gas to most of the carbon particles. Table 2 shows the weight loss of the samples after the heat treatment. From the table, weight loss of the sample observed upto 30% at 650^oC and after which the weight loss increases very rapidly at higher temperatures and eventually

all the sample get oxidized. Fig. 2 shows the SEM images of these samples heated at various temperatures with controlled flow of oxygen. Annealing at temperatures more than 650oC,resulted in some part of CNT getting oxidized and showed broken CNT together with soot sitting on the surface. At higher temperatures, the impurity carbon particles are initially burned out, but as the time proceeds, more CNT are exposed on the surface and had more chance to be attacked by reactive gas species, resulting in oxidation of nanotubes and low yield of purification.

Surface area of heat treated carbon nanotube were measured and values are shown in Fig. 3. It is observed from the figure that heat treatment of pristine sample enhances the surface area. The surface area of the pristine sample was 14 m²/gm

which is enhances up to 35 m²/gm for the sample which is heat treated to 650^oC which means almost more than double the surface area. However, samples heat treated at lower temperatures does not show much increase in surface area which shows that most of the carbon nanotubes are still embedded in the other carbonaceous materials. Ultraviolet Absorption studies of these samples were also carried out in the wavelength range of 200 to 800 nm which are shown in Fig. 4. All the six samples pristine and heat treated at different temperature under control flow of oxygen are showing the absorption, however sample heat treated at 650^oC shows the maximum absorption as compared to all other samples. This could be due to the fact that sample heat treated at 650^oC under control flow of Oxygen gas has

oxidized impurities and other carbonaceous materials and now the sample contains mostly carbon nanotubes which is showing the maximum absorption of UV.

Cathode deposit sample is treated with acid mixture of H₂SO₄ & HNO₃ in ratio of 4:1 for 45 min and filtered by using a membrane filter (PVDF) of 0.45 micron size. After

filtration the solid content on the filter consists of mostly CNT as shown in Fig. 5. At the same time, CNTs were found in the suspension also as shown in Fig. 6. Therefore, both suspension and solid content on filter paper are required for getting the carbon nanotubes and should not be discarded.The chemically treated CNT is subjected to annealed at various temperatures ranging from (450~750^oC) and the SEM micrographs of these samples are shown in Fig. 7.

Samples were also heat treated under control flow of oxygen. SEM pictures of chemically treated CNTs subjected to thermal annealing in controlled oxygen atmosphere at various temperatures (450~750^oC) are shown in Fig. 8. Weight loss of the samples are given in Table 3, which shows that weight loss of the sample heat treated at 650^oC temperature is 50%. However, whereas in normal thermal annealing the weight loss was found around 30% at 650^oC. It is also observed that beyond 650^oC weight loss is very rapid and more in the case of excess supply of oxygen which is any way very obvious.

BET analysis was carried out for the samples heat treated at 550^oC and 650^oC and shown in Fig. 9. It is evident from

the figure that surface area has tremendously increases of the sample which was chemically and heat treated at 650^oC as compared to the sample which was only heat treated at the same 650^oC temperature. Similar effect is also observed in the case of UV absorption studies which is shown in Fig. 10.From the figure, it is clear that UV absorption is more in case of CNT-chemically treated and heat treated at 650^oC.

The carbon nanotubes are having inert surface and are not suitable for most of the applications such as in the area of composite etc. To overcome this problem, suitable functional groups are attached to have strong chemical bonding with matrix of the composites. The carboxylic acid functional

groups that are formed on the nanotubes surface enable them to be dispersed readily in water. The main aim of this chemical treatment is to form oxide groups on graphite such that its oxidation temperature comes down. As per literature graphite oxidation temperature is 710^oC which is similar to carbon nanotubes and therefore separating these two species is really difficult. In order to overcome it was thought of intercalating the sample with some strong oxidizing agents and it was confirmed the groups from FTIR analysis as shown in Fig. 11.

The samples were analyzed and confirms the presence of hydroxyl, carbonyl and carboxylic groups at 3440, 1627 and 2925 cm^-1 respectively. The duration of the chemical treatment is also very important if it is treated for longer duration can destroy the nanotubes which is shown in the Fig. 12.

It is well known that different structural forms of carbon

show different oxidation behaviour.

In other words, they exhibit different reactivity with oxygen which is dependent on the available reactive sites. A well graphitized structure starts to oxidize at a relatively higher temperature of ~600-700^oC. Therefore, TGA has been found to be useful tool in unraveling the nature of carbons present in the arc discharge produced deposit. Since oxidation temperature of graphite is approximately 710ºC attempts were made in reducing the oxidation temperature of graphite as it is known that the true oxidation initiation temperature of MWCNTs is ~700^oC, Fig. 13 clearly shows that the oxidation temperature has come down for the sample which was chemically treated which shows that attachment of oxide groups to graphite can bring down the oxidation temperature which will helps in oxidizing graphitic carbons and retaining the carbon nanotubes.

For the purification of MWCNTs the approach was combination of acid treatment and thermal annealing which gave a better yield of 50% at 650ºC in the presence of oxygen atmosphere. However, at higher temperatures carbon nanotubes were get damaged and oxidized. The combination of acid treatment and thermal annealing is a promising process as compared to thermal annealing alone. However, proper time of the treatment has to be optimized otherwise acid treatment for longer time will results in complete destruction of carbon nanotubes. The purification approach in the present work is quite simple and effective for the removal of unwanted carbon from the cathode deposit of the arc discharge process to get more MWCNT yield.