Liquefied petroleum gas (LPG) is a common household fuel used for cooking purpose in India. LPG is very rich in its carbon content because of its specific mixing components of predominantly C3 alkane (Propane – C3H8) or C4 alkane (Butane – C4H10) which provides a better chance of producing strong and good quality nano products like nanotubes, nanotubes nanowires, nanoparticles etc. In our laboratory a lab scale flame reactor is designed and developed for producing carbon nanotubes using LPG as the carbon source in the presence of air as an oxidant under atmospheric conditions. The design aspects and the best operational conditions of the flame reactor for producing carbon nanotubes are discussed. The nanotubes obtained were purified and were further characterized using SEM, TEM XRD and Raman.
Carbon Nanotubes (carbon); TEM (Transmission electron microscopy); LPG (alkanes); Raman (Raman spectroscopy); XRD; Flame Synthesis;
Liquified petroleum gas (also called as LPG or Autogas) is a mixture of hydrocarbon gases used as a fuel in heating appliances and vehicles and it is increasingly replacing chlorofluorocarbons as an aerosol propellant and a refrigerant to reduce the damage and degeneration of the ozone layer. LPG is a clean, convenient energy source, which can be stored as a liquid under moderately high pressure and used as a gas in commercial and residential heating applications. It is a common household fuel used for cooking purpose in India, LPG is rich in its carbon content because of its specific mixing components of predominantly C3 alkane (Propane – C3H8) or C4 alkane (Butane – C4H10) which provides a better chance of producing strong and good quality nano products like nanotubes, nanotubes nanowires, nanoparticles etc.
Carbon nanotubes (CNTs) are among the amazing objects that science sometimes creates by accident, without meaning to, but that will likely revolutionize the technological landscape of the century ahead. Our society stands to be significantly influenced and shaped by carbon nanotube applications in every aspect, Carbon nanotubes have been synthesized for a long time as products from the action of a catalyst over the gaseous species originating from the thermal decomposition of hydrocarbons . Since their discovery by Sumio Ijima  several ways of preparing them have been explored. The CNTs have been synthesized by various methods e.g. electric arc discharge, laser evaporation and chemical vapor deposition (CVD) [3-5].
Though researchers have been successful to synthesize multi-wall nano tubes they can produce only in milligram to gram quantities in a few hours. However as many potential applications [6-7] of CNTs require kilogram to ton quantities.
Apisit Songsasen et al  have synthesized CNTs by means of catalytic decomposition of LPG on a Zeolite-supporting Nickel catalyst. Qian et al  have reported the formation of CNTs by the decomposition of liquefied petroleum gas (LPG) containing sulfur in the presence of Fe/Mo/Al2O3 catalyst, Since this contains sulfur of a few to several hundred ppm, which can lead to poisoning the catalyst heavily, few reports currently exist on using LPG or natural gas directly for production of carbon nanomaterials, only Prokudina et al  has reported CNT synthesis from LPG by CVD method, but till date no information and literature is reported on direct flame synthesis of CNTs by LPG. The main challenge in this field is to develop methods to produce nanotubes on a large scale and at low cost. As Flame synthesis of nano carbons being a continuous flow method, in which flowing gaseous feedstock mixture could produce CNTs in large quantities it has several advantages like easy scale up, particle size control, dual role of feed gas which serves both as carbon source and fuel, and in-situ generation of catalyst. Hence it is one of the preferred methods for bulk production of not only CNTs but also other nano particles and nano metal oxides. This method is very useful and is of widespread importance.
Many groups have investigated gas-phase continuous-flow production of carbon nanomaterials using other hydrocarbons. These studies typically involve passing a mixture of carbon source gas and organo metallic catalyst precursor molecules through a heated furnace. In this paper we report the direct flame synthesis of carbon nanotubes using LPG and air as our gaseous feedstock in a diffusion type burner without any external use of a catalyst and synthesis at optimum process parameters.
The flame reactor (Fig.1) has been indigenously designed to produce carbon nanotubes at our university. The detailed setup and process instrument and diagram (PID) of the reactor (Fig. 2) has been discussed in detail in our previous work . In general our reactor operates under atmospheric pressure. The measured quantity of the LPG and the oxidant reaches the ignition chamber where the partial combustion process occurs where the CNT’s are produced. During the process we have observed the dark orange flame color which is perfectly in a spindle form. Along the entire length of the flame, its temperature was recorded using a K-type thermocouple where this temperature can provide some data regarding the growth of nanotubes. The soot thus produced is captured on a glass fiber filter (Axiva GF/A) with the aid of a vacuum pump and the collected soot is scrapped carefully and weighed and later heat treated and oxidized at 550 OC in the presence of air for 60 minutes to remove any traces of amorphous carbon impurities and then the sample is reweighed in order to estimate the loss of amorphous carbon as an impurity then the samples are later characterized by SEM, TEM, XRD and Raman for their quality. The total amount of thermally oxidized and purified sample from the experiment (for 30 minute run) weighed only 0.8g.
The samples were analyzed using Phillips XL 30 series Scanning Electron Microscope (SEM) from National Center for Compositional Characterization of Materials (NCCCM), Hyderabad. From the Figs (3a – 3d) we can see a dense growth of carbon nanotubes at various flow rates with respect to the oxidant to fuel (O/F) ratios between 0.7 – 1.0 slpm/slpm (standard litre per minute). The average diameter range of the CNTs from the SEM image was found to be around 200 nm -1000 nm and lengths greater than 40 μm.
The TEM (Technai -12, FEI) images (Fig 4a) shows the presence of thickly packed multiwalled CNT with an average diameter of 150 – 250 nm which is still surrounded by traces of carbonaceous nanoparticle aggregates possibly caused due to the dispersion of the sample in the solvent. This can be assumed that the agglomerated carbon nanoparticles were actually protected by the CNTs during the thermal treatment, as the CNTs might have formed a net like layer covering the nanoparticles and protecting it from the heat and oxidation. Fig 4b shows a thick multi walled CNT around 250 nm in its diameter with lots of traces of agglomerated carbon nanoparticles which can be accounted for the presence of C60 particles which is also in agreement with the XRD analysis in Fig 5. The broken caps of the CNTs also reveal the disorientation and a defective growth of the grapheme layers as seen in the Raman analysis in Fig 6.
The XRD (PW1830 Phillips) analysis was carried out using CuKa1 type of radiation with a wavelength (l) of 1.54060 Å. XRD (Fig. 5) of nanotubess produced using LPG-air at an O/F ratio of 0.7 slpm/slpm shows a heterogeneous crystallinity in the sample. The raw scan detected three strong peaks. The first peak at 2θ angle of 25.77O was found with (110) orientation of atoms along its plane with peak corresponding to graphite with an orthorhombic type of system and an end-centered lattice. The second peak at 2θ angle of 43.159O was found with (245) orientation of atoms along its plane with peak corresponding to C60 molecule with a cubic type of system and a primitive lattice. The third peak at 2θ angle of 83.475O was found with (112) orientation of atoms along its plane with peak corresponding to graphite with a hexagonal type of system and a primitive type lattice respectively.
Raman analysis (Horiba Jobin Yvon T64000, Raman Spectrophotometer) was carried out only on the best sample (Fig.6) which clearly shows the D band & G band respectively. The D band (the disorder band is well-known in disordered graphitic materials and located between 1330-1360 cm-1 when it is excited with a visible laser) it is expected to be observed in Multi Walled Nanotubes (MWNT). However when the D band is observed in SWNT’s , it is assumed to contain defects in the tubes. The G band or (TM- Tangential Mode) , corresponds to the stretching mode of the -C-C- bond in the graphite plane . This mode is located near 1580 cm-1. From the figure we can say that the nanotubes are in the slightly disordered graphite phase based on the D band wavelength present at 1349 cm-1. This D band also confirms the presence of amorphous state of carbon in the bulk sample. Based on the G band from the figures, there appears two peaks at 1560 and 1600 cm-1 respectively which proves the presence of multi layers of disordered graphene sheets. On analyzing the level of graphitization using the D and G band intensities ratio, we find that the sample is normally well graphitized with small degree of crystallinity and its ID/IG ratio was found to be around 0.939.
Carbon nanotube (CNT) is a versatile group of applied chemicals with high degree of applications on larger scale in various disciplines. The synthesis, purification and the cost still remains an un-doubted debate around the world hence an economical approach is to be developed in order to produce large amounts of good quality CNTs from an economical and a resourceful fuel. LPG as a general commodity plays a major role since its availability in India is high and it is a very economical source of fuel as well. Here, we were able to successfully synthesize semicrystalline, CNTs from LPG with an average diameter of 100 – 500 nm using the direct flame synthesis approach.
. Bharat Bhushan, Springer Handbook of Nanotechnology, Springer-Verlag Berlin Heidelberg, New York, 2004, Chap: 3, pp 39 – 40.
. S. Iijima, Nature 354, (1991), 56.
. T.W. Ebbesen and P.M. Ajayan, Nature 358, (1992), 220.
. T. Guo, P. Nikolaev, A.G. Rinzler, D. Tomanek, D.T. Colbert and R.E. Smalley, J. Phys. Chem. 99, (1995), 10694.
. J. Kong, A. M. Cassell and H.J. Dai, Chem.Phys. Lett. 292, (1998), 567.
. Zhou X T, Lai H L, Peng H Y, Au F C K, Liao L S, Wang N, Bello I, Lee C S, Lee S T, Chem Phys Lett 318, (2000), 58 – 62.
. Zhou X T, Wang N, Au F C K, Lai H L, Peng H Y, Bello I, Lee C S, Lee S T, Mater. Sci. Eng. A 286 (2000) 119 -124.
. Apisit Songsasen and Paranchai Pairgreethaves, the Kasetsart Journal. (Natural. Sciences) Number 3, 35, (2001), 354 – 359.
. W. Qian, H. Yu, F. Wei, Q. Zhang and W.Wang, Carbon 40, Issue 15, (2002), 2968-2970.
. N.A. Prokudina, E.R. Shishchenko, O.S. Joo, D.Y. Kim and S.H. Han, Advanced Materials, Vol. 12, Issue 19, (2000), 1444 – 1447.
. Vivek Dhand, J.S Prasad, M. Venkateswara Rao, K. Naga Mahesh, L. Anupama, V. Himabindu, Anjaneyulu Yerramilli, V.S. Raju, A.A. Sukumar Indian Journal of Engg & Mat. Sci, 14, (2007), 240-252.