- Email: [email protected]
nano-fibers
Materials Research Bulletin 41 (2006) 2303–2310 www.elsevier.com/locate/matresbu
Thermal plasma synthesis of SiC nano-powders/nano-fibers Lirong Tong, Ramana G. Reddy * Department of Metallurgical and Materials Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA Received 21 December 2005; received in revised form 2 April 2006; accepted 17 April 2006 Available online 6 May 2006
Abstract Thermal plasma synthesis of nano-powders/nano-fibers is a relatively new technology with great potential for future application in industries. The paper presents the effects of molar ratio on synthesis of SiC by thermal plasma technology. The experimental results show that SiC can be synthesized by thermal plasma technology. The average size of SiC powders is less than 100 nm and SiC fiber-like microstructure were observed. # 2006 Published by Elsevier Ltd. Keywords: A. Carbides; A. Nanostructures; B. Chemical synthesis
1. Introduction Silicon carbide is one of the most important non-oxide ceramic materials which are produced on a large scale in the form of powders, molded shapes and thin film [1,2]. It has wide industrial application due to its excellent mechanical properties, high thermal and electrical conductivity, excellent chemical oxidation resistance, and it has potential application as a functional ceramic or a high temperature semiconductor. The main synthesis method of SiC is a carbothermic reduction known as the Acheson process. The general reaction [2] is: SiO2 þ 3C ! SiC þ 2COðgÞ
(1)
A conventional method for the synthesis of pure SiC powders involves many steps and is an energy-intensive process. Further, the SiC particle size is relatively coarse [3]. It is well known that materials with fine microstructures exhibit markedly improved properties without change in their physical appearance. These characteristics include improved strength, stiffness, wear resistance, fatigue resistance and lower ductility and toughness. A great variety of alternate methods like sol–gel [4], plasma [5], laser [6] and microwave [3] have been reported in the literature for the synthesis of fine SiC powders. Nanoscale SiC fibers have important potential applications in nanoelectronics, field emission devices and nanocomposites [7], therefore, efforts have been made by many research groups to fabricate SiC nano-fibers by methods like carbon nano-tubes (as the template) confined reactions [8,9], catalyst assisted chemical vapor deposition (CVD) via the vapor–liquid–solid (VLS) mechanism [10] and template/catalyst-free processes [11]. These methods involve multi-steps and have the difficulty in establishing commercial viability.
* Corresponding author. Tel.: +1 205 348 4982; fax: +1 205 348 2164. E-mail address: [email protected] (R.G. Reddy). 0025-5408/$ – see front matter # 2006 Published by Elsevier Ltd. doi:10.1016/j.materresbull.2006.04.021
2304
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
Among the advantages of thermal plasma processing (TPP) techniques in the synthesis of powders are high enthalpies to vaporize all reactants to monatomic gaseous state and to enhance the reaction kinetics by several orders of magnitude, and a clean reaction atmosphere which is required to process high-purity products. Owing to the impurity-free environment and the rapid quenching that is achievable in a plasma reactor, plasma-processing technique ensures synthesis of clean and ultrafine metal, ceramic or composite powders. This is a very suitable method to synthesize refractory materials, such as carbides and nitrides, which are commonly used in high temperature applications. In the thermal plasma processing of nano-powders, the high enthalpy of thermal plasma is used to generate high-density vapor-phase precursors which are then quenched rapidly for synthesis of nano-powders. Due to shock quenching, the supersaturation of vapor species enhances the driving force for particle nucleation to lead to the production of ultrafine particles (down to nanoscale size) by homogeneous nucleation. Several investigations [5,12–15] on the thermal plasma synthesis of metals, ceramics, or composites in the recent past have confirmed that thermal plasma synthesis is one of the most promising methods for producing nano-powders that have high potential in the aerospace industry. The paper reports the effects of molar ratio on synthesis of SiC nano-powders. The experimental results confirmed that SiC can be synthesized by thermal plasma technology. SEM and TEM observations show that the product is the mixture of nano-powders and fiber-like structure. 2. Experimental setup The experimental setup of plasma unit includes water cooling system, plasma generating system, thermal plasma reactor, particle feeder system and data acquisition system. The reactor utilizes a non-transferred PT-50C plasma torch (with a maximum power of 45 kW). Fig. 1 shows the photograph and a schematic diagram of the thermal plasma reactor, which consists of three zones: (i) the combustion/synthesis/reaction zone, (ii) the quenching zone and (iii) the collection zone. Outer shell of the reactor is made of 316 L stainless steel. A graphite tube is used for lining inside the reactor and is insulated with several layers of alumina felt placed between the graphite tube and the inner wall of the water-cooled reactor. The quenching chamber consists of two cone-shape water-cooled copper quenching tubes. Two cone-shaped copper tubes cool the outgoing gas to temperatures in the range of 100–160 8C. The collection chamber has a cloth filter to collect product powders at the exit. The experimental procedure was described in detail in one earlier publication [13].
Fig. 1. The photograph and a schematic of the thermal plasma reactor [16].
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
2305
Silicon dioxide (SiO2) powder was supplied by Alfa Aesar1. Particle purity was higher than 99% and its size was less than 45 mm. The argon (used as the plasma gas and carrier gas) and the methane (used as the gaseous reactant and carrier gas) used in experiments were more than 99.99% pure to ensure a pure and clean reacting environment. Philips XL30 scanning electron microscope equipped with an EDAX DX41 energy-dispersive spectrometerand, Hitachi H-8000 transmission electron microscopy (TEM) and Philips X-ray diffractometer of model PW-1710 were used to characterize product particles. TEM and SEM samples was prepared by stirring a little amount of product sample with the help of ethyl alcohol by Bransonic1 ultrasonic cleaner Model B220R-4, and then one drop of this stirred liquid was dispersed on the standard copper grid (mesh 300) with carbon support film and allowed to dry. Product yield was estimated by means of direct comparison method using the X-ray diffraction pattern of the product particles. A Quantachrome NOVA 1200E surface area analyzer using the BET method was used to measure the average size of product particles. 3. Experimental results and discussions In the synthesis technology of nanoscale SiC, the kinetics of the SiC synthesizing reaction (1) is critical. If injected SiO2 particles cannot be fully vaporized in the plasma reactor, the diffusion of C from a SiO2 particle surface into its interface could be a key controlling step and SiO2 particles will not be completely carbonized to SiC. Therefore, the key step is to fully vaporize raw SiO2 particles for plasma synthesis of nanoscale SiC [12,13]. Simulation results [17] showed that the residence time of a SiO2 particle in plasma reactor is less than 7 ms and the final temperature of a SiO2 particle decreases with the SiO2 particle size. When the SiO2 particle size is smaller than 45 mm, the final temperature of the particle is higher than SiO2 boiling point. The simulation results also showed that the final temperature of SiO2 particles decreases with injected velocity. Therefore, we chose SiO2 powders less than 45 mm as the starting material and kept the injection velocity of SiO2 particles at 10 m/s for our experiments [17]. Table 1 lists the main experimental parameters. The plasma power and the feeding rate of SiO2 were kept at 25.6 kW and 5 g/min, respectively; the effect of molar ratio of CH4:SiO2 on synthesis of SiC nano-powders was investigated. The other experimental conditions, such as the pressure and flow rate of plasma gas, were kept constant. The direct comparison method [17,18] was used for estimating the volume percent of each phase from the XRD pattern of the product powders and then transferred volume percent to mole percent. The estimated mole percent of each phase in product powders is shown in Fig. 2. The solid lines in the figure are aid to the eye and have no physical significance. When the molar ratio of CH4:SiO2 = 0.8:1, there is more than 25 mol.% of SiO2 phase in product powders. Carbon peaks were not observed in the XRD pattern of the product powders. More than 25 mol.% of SiO2 phase in product powders may result from excess SiO2 input. On the contrary, when the molar ratio of CH4:SiO2 = 1.5:1, carbon peaks is significant in the product powders. SiO2 peaks were not observed in the XRD pattern of the product powders. The carbon peaks indicate that excess CH4 was input. When the molar ratio of CH4:SiO2 = 1:1, the product powders are composed of more than 91 mol.% of SiC phase and less than 9 mol.% of SiO2 phase. The SiO2 phase in product phase may be due to the flow rate vibration of CH4. Therefore, the total synthesis reaction of SiC can be expressed in the following equation: SiO2 þ CH4 ¼ SiC þ 2H2 O
(2)
Fig. 3 is a typical XRD pattern of product powders which clearly confirmed the formation of b-SiC using SiO2 and CH4 as starting materials. An electronic diffraction pattern (Fig. 4) of the sample in Fig. 5 appears to be that of a FCC lattice with alternate pattern of a pair and a single peak confirming that the diffraction pattern is that of b-SiC. The calculated results from XRD pattern indicated that the lattice parameter (0.43635 nm) of product b-SiC powders is little bigger than its standard value (0.435853 nm from JSPDS card 74–2307). Table 1 Experimental conditions Run no.
Input power (kW)
Feeding rate (g/min)
Molar ratio (CH4:SiO2)
1 2 3
25.6 25.6 25.6
5 5 5
1:1 1.5:1 0.8:1
2306
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
Fig. 2. Relationship of SiC, SiO2 and C in product powders with molar ratio of CH4:SiO2.
Fig. 6 is a typical SEM image of SiC product. Fig. 7 is a magnified version of the micrograph in Fig. 6. The SiC product is the mixture of particles and fibers as shown in Fig. 6. Most particles are less than 100 nm. The fibers’ length is around from 50 to 900 nm and their diameter is around 30 nm as shown in Fig. 7. A Quantachrome NOVA 1200E surface area analyzer was used to measure the average surface area of product particles. Assuming the sample particles to be spherical, the average diameter of the product particles can be obtained from the following equation: 6 d¼ (3) Ar where d is the average diameter of SiC powders, A the surface area of sample powders per gram and r is SiC density. The BET results showed that the average diameter of product particles is around 50 nm. The result is less than what was observed and measured in SEM images of the product powders. As shown in Fig. 6, SiC particles are not perfectly spherical and the SiC fiber surface exhibits a corrugated morphology. These two reasons increased total measured surface area of the sample particles and thus result in the small calculated value of powder diameter. As shown in Figs. 6 and 7, a group of fibers nucleated on a substrate of a condensed particle, and grew in outward direction. The fibers could not keep the exact same growth direction and the SEM images indicate that fibers interwoven in an irregular manner. The SiC fibers appear to grow along a direction [1 1 1] in a stacking manner as
Fig. 3. XRD pattern of product powders from run no. 1 with molar ratio of CH4:SiO2 = 1:1.
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
2307
Fig. 4. An electronic diffraction pattern of the sample in Fig. 5.
shown in Figs. 5–7. High–density stacking faults observed by Wu et al. [18] confirmed the growth mechanism. The relationship of these two directions is unclear at present. Some researcher groups [18,19] reported that there is a tilt angle between the normal of stacking faults and the axial direction of SiC nano-fiber. Most of literatures [20,21] reported that the normal of stacking faults is parallel to the axial direction of SiC nano-fiber. It is well worth notice that
Fig. 5. TEM image of product powders from run 1 with input power 23.4 kW, feeding rate 5 g/min and molar ratio of CH4:SiO2 = 1:1.
2308
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
Fig. 6. SEM image of product powders from run 1 with input power 23.4 kW, feeding rate 5g/min and molar ratio of CH4:SiO2 = 1:1.
the fiber surface is not very smooth and exhibit a corrugated morphology. Wu [19] explained the corrugated surface is attributed to the relative shift of two adjacent stacking layers. When SiC supersaturated gas phase was quenched, it is assumed that SiC homogeneously nucleated and grew in the supersaturated gases in the plasma reactor as shown in Fig. 8(a). Due to the surface tension effect, spherical powders could be obtained. On the other hand, if a group of SiC heterogeneously nucleated on the surface of condensed SiC on the quench coil as shown in Fig. 8(b), nano-fibers could be formed due to directional cooling. By redesigning the quenching system, SiC powders could be avoided. The reasons why the product is the mixture of nano-powders and nano-fibers and why the fibers appear entangled are not yet clear. Three of experimental runs indicated that the mixture products of nano-powders and nano-fibers can be obtained at three different molar ratios. These experimental results imply nano-fibers can be easily synthesized using SiO2 and CH4 as starting materials by thermal plasma technology. By optimizing experimental parameters of the thermal plasma technique and redesigning the quenching system, pure nano-fibers of SiC could be synthesized. This technique can avoid using expensive carbon sources or catalysts for SiC nano-fiber or nano-wire synthesis. The most important is that this technique is promising for SiC nano-fibers synthesis in higher production rate because the feeding rate of starting material can go up to 10 g/min in the experimental setup as mentioned in experimental conditions.
Fig. 7. Micrograph in Fig. 6 at higher magnification.
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
2309
Fig. 8. Models of SiC nucleation and growth of: (a) SiC nano-powders and (b) nano-fibers in plasma reactor.
4. Summary and conclusions 1. If SiO2 and CH4 were used as the starting materials, the experimental results showed that the total synthesizing reaction of SiC can be expressed in the following equation: SiO2 þ CH4 ¼ SiC þ 2H2 O The experimental results did not indicate the obvious dependence of product shape on mole ratios of CH4:SiO2. 2. The experimental results showed that the mixture of SiC nano-powders and nano-fibers can be synthesized using SiO2 and CH4 as the starting materials by thermal technology. The average diameter of nano-powders is smaller than 100 nm. The fibers’ length is from 50 to 900 nm and their diameter is around 30 nm. 3. The SiC fibers appear to grow along a direction [1 1 1] in a stacking manner. The SiC nano-fiber surface is not very smooth and exhibit a corrugated morphology. 4. Experimental results imply nano-fibers can be easily synthesized using SiO2 and CH4 as starting materials by thermal plasma technology. By optimizing experimental parameters of the thermal plasma technique pure nanofibers of SiC may be synthesized in higher production rate.
Acknowledgements The authors are pleased to acknowledge Dr. Divakar Mantha for discussing and reviewing this paper and the financial support for this research by the U.S. Department of Defense (Grant number DAAD 19-01-1-0137). References [1] M.I. Boulos, P. Fauchais, E. Pfender, Thermal Plasmas Fundamentals and Applications, vol. 1, Plenum Press, New York and London, USA, 1994, p. 37. [2] H.O. Pierson, Handbook of Refractory Carbides and Nitrides, William Andrew, Noyes, 1996, p. 137. [3] L.N. Satapathy, P.D. Ramesh, D. Agrawal, R. Roy, Mater. Res. Bull. 40 (2005) 1871. [4] I.S. Seong, C.H. Kim, J. Mater. Sci. 28 (1993) 3277. [5] J.Y. Guo, F. Gitzhofer, M.I. Boulos, J. Mater. Sci. 30 (1995) 5589. [6] Y. Li, Y. Liang, Z.J. Hu, J. Am. Ceram. Soc. 77 (1994) 1662. [7] W. Yang, H. Araki, S. Thaveethavorn, H. Suzuki, T. Noda, Appl. Surf. Sci. 241 (2005) 236. [8] H. Dai, E.W. Wong, Y.Z. Liu, S. Fan, C.M. Lieber, Nature 375 (1995) 769. [9] C.C. Tang, S.S. Fan, H.Y. Dang, J.H. Zhao, C. Zhang, P. Li, Q. Gu, J. Cryst. Growth 210 (2000) 595. [10] X.T. Zhou, N. Wang, C.K.A. Frederick, H.L. Lai, H.Y. Peng, I. Bello, C.S. Lee, S.T. Lee, Mater. Sci. Eng. A 286 (2000) 119. [11] G.W. Meng, L.D. Zhang, C.M. Mo, S.Y. Zhang, Y. Qin, S.P. Feng, H.J. Li, J. Mater. Res. 13 (1998) 2533. [12] P.R. Taylor, S.A. Pirzada, Metall. Trans. B 23B (1992) 443. [13] L. Tong, R.G. Reddy, Scr. Mater. 52 (2005) 1253. [14] S. Niyomwas, B. Wu, R.G. Reddy, in: B.S. Mishra, S.L. Semiatin, C. Suryanarayana, N.N. Thandhani, T.C. Lowe (Eds.), Ultrafine Grained Materials, TMS, Warrendale, PA, 2000, 89. [15] P.R. Taylor, M. Manrique, JOM 6 (1996) 43. [16] R.G. Reddy, Metall. Mater. Trans. B 34 (2003) 137.
2310
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
[17] L. Tong, R.G. Reddy, Mathematical modeling for processing of fine powders by thermal plasma, in: H. Mahfuz, M.V. Hosur (Eds.), Developments in Theoretical and Applied Mechanics, vol. XXII, 2004, p. 656. [18] Y.J. Wu, J.S. Wu, W. Qin, D. Xu, Z.X. Yang, Y.F. Zhang, Mater. Lett. 58 (2004) 2295. [19] Z.L. Wang, Z.R. Dai, P.R. Gao, Z.G. Bai, Appl. Phys. Lett. 77 (2000) 3349. [20] Y.H. Gao, Y. Bando, K. Kurashima, T. Sato, J. Mater. Sci. 37 (2002) 2023. [21] Y.J. Zhang, N.L. Wang, P.R. He, X.H. Chen, J. Zhu, J. Solid State Commun. 118 (2001) 595.
Thermal plasma synthesis of SiC nano-powders/nano-fibers Lirong Tong, Ramana G. Reddy * Department of Metallurgical and Materials Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA Received 21 December 2005; received in revised form 2 April 2006; accepted 17 April 2006 Available online 6 May 2006
Abstract Thermal plasma synthesis of nano-powders/nano-fibers is a relatively new technology with great potential for future application in industries. The paper presents the effects of molar ratio on synthesis of SiC by thermal plasma technology. The experimental results show that SiC can be synthesized by thermal plasma technology. The average size of SiC powders is less than 100 nm and SiC fiber-like microstructure were observed. # 2006 Published by Elsevier Ltd. Keywords: A. Carbides; A. Nanostructures; B. Chemical synthesis
1. Introduction Silicon carbide is one of the most important non-oxide ceramic materials which are produced on a large scale in the form of powders, molded shapes and thin film [1,2]. It has wide industrial application due to its excellent mechanical properties, high thermal and electrical conductivity, excellent chemical oxidation resistance, and it has potential application as a functional ceramic or a high temperature semiconductor. The main synthesis method of SiC is a carbothermic reduction known as the Acheson process. The general reaction [2] is: SiO2 þ 3C ! SiC þ 2COðgÞ
(1)
A conventional method for the synthesis of pure SiC powders involves many steps and is an energy-intensive process. Further, the SiC particle size is relatively coarse [3]. It is well known that materials with fine microstructures exhibit markedly improved properties without change in their physical appearance. These characteristics include improved strength, stiffness, wear resistance, fatigue resistance and lower ductility and toughness. A great variety of alternate methods like sol–gel [4], plasma [5], laser [6] and microwave [3] have been reported in the literature for the synthesis of fine SiC powders. Nanoscale SiC fibers have important potential applications in nanoelectronics, field emission devices and nanocomposites [7], therefore, efforts have been made by many research groups to fabricate SiC nano-fibers by methods like carbon nano-tubes (as the template) confined reactions [8,9], catalyst assisted chemical vapor deposition (CVD) via the vapor–liquid–solid (VLS) mechanism [10] and template/catalyst-free processes [11]. These methods involve multi-steps and have the difficulty in establishing commercial viability.
* Corresponding author. Tel.: +1 205 348 4982; fax: +1 205 348 2164. E-mail address: [email protected] (R.G. Reddy). 0025-5408/$ – see front matter # 2006 Published by Elsevier Ltd. doi:10.1016/j.materresbull.2006.04.021
2304
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
Among the advantages of thermal plasma processing (TPP) techniques in the synthesis of powders are high enthalpies to vaporize all reactants to monatomic gaseous state and to enhance the reaction kinetics by several orders of magnitude, and a clean reaction atmosphere which is required to process high-purity products. Owing to the impurity-free environment and the rapid quenching that is achievable in a plasma reactor, plasma-processing technique ensures synthesis of clean and ultrafine metal, ceramic or composite powders. This is a very suitable method to synthesize refractory materials, such as carbides and nitrides, which are commonly used in high temperature applications. In the thermal plasma processing of nano-powders, the high enthalpy of thermal plasma is used to generate high-density vapor-phase precursors which are then quenched rapidly for synthesis of nano-powders. Due to shock quenching, the supersaturation of vapor species enhances the driving force for particle nucleation to lead to the production of ultrafine particles (down to nanoscale size) by homogeneous nucleation. Several investigations [5,12–15] on the thermal plasma synthesis of metals, ceramics, or composites in the recent past have confirmed that thermal plasma synthesis is one of the most promising methods for producing nano-powders that have high potential in the aerospace industry. The paper reports the effects of molar ratio on synthesis of SiC nano-powders. The experimental results confirmed that SiC can be synthesized by thermal plasma technology. SEM and TEM observations show that the product is the mixture of nano-powders and fiber-like structure. 2. Experimental setup The experimental setup of plasma unit includes water cooling system, plasma generating system, thermal plasma reactor, particle feeder system and data acquisition system. The reactor utilizes a non-transferred PT-50C plasma torch (with a maximum power of 45 kW). Fig. 1 shows the photograph and a schematic diagram of the thermal plasma reactor, which consists of three zones: (i) the combustion/synthesis/reaction zone, (ii) the quenching zone and (iii) the collection zone. Outer shell of the reactor is made of 316 L stainless steel. A graphite tube is used for lining inside the reactor and is insulated with several layers of alumina felt placed between the graphite tube and the inner wall of the water-cooled reactor. The quenching chamber consists of two cone-shape water-cooled copper quenching tubes. Two cone-shaped copper tubes cool the outgoing gas to temperatures in the range of 100–160 8C. The collection chamber has a cloth filter to collect product powders at the exit. The experimental procedure was described in detail in one earlier publication [13].
Fig. 1. The photograph and a schematic of the thermal plasma reactor [16].
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
2305
Silicon dioxide (SiO2) powder was supplied by Alfa Aesar1. Particle purity was higher than 99% and its size was less than 45 mm. The argon (used as the plasma gas and carrier gas) and the methane (used as the gaseous reactant and carrier gas) used in experiments were more than 99.99% pure to ensure a pure and clean reacting environment. Philips XL30 scanning electron microscope equipped with an EDAX DX41 energy-dispersive spectrometerand, Hitachi H-8000 transmission electron microscopy (TEM) and Philips X-ray diffractometer of model PW-1710 were used to characterize product particles. TEM and SEM samples was prepared by stirring a little amount of product sample with the help of ethyl alcohol by Bransonic1 ultrasonic cleaner Model B220R-4, and then one drop of this stirred liquid was dispersed on the standard copper grid (mesh 300) with carbon support film and allowed to dry. Product yield was estimated by means of direct comparison method using the X-ray diffraction pattern of the product particles. A Quantachrome NOVA 1200E surface area analyzer using the BET method was used to measure the average size of product particles. 3. Experimental results and discussions In the synthesis technology of nanoscale SiC, the kinetics of the SiC synthesizing reaction (1) is critical. If injected SiO2 particles cannot be fully vaporized in the plasma reactor, the diffusion of C from a SiO2 particle surface into its interface could be a key controlling step and SiO2 particles will not be completely carbonized to SiC. Therefore, the key step is to fully vaporize raw SiO2 particles for plasma synthesis of nanoscale SiC [12,13]. Simulation results [17] showed that the residence time of a SiO2 particle in plasma reactor is less than 7 ms and the final temperature of a SiO2 particle decreases with the SiO2 particle size. When the SiO2 particle size is smaller than 45 mm, the final temperature of the particle is higher than SiO2 boiling point. The simulation results also showed that the final temperature of SiO2 particles decreases with injected velocity. Therefore, we chose SiO2 powders less than 45 mm as the starting material and kept the injection velocity of SiO2 particles at 10 m/s for our experiments [17]. Table 1 lists the main experimental parameters. The plasma power and the feeding rate of SiO2 were kept at 25.6 kW and 5 g/min, respectively; the effect of molar ratio of CH4:SiO2 on synthesis of SiC nano-powders was investigated. The other experimental conditions, such as the pressure and flow rate of plasma gas, were kept constant. The direct comparison method [17,18] was used for estimating the volume percent of each phase from the XRD pattern of the product powders and then transferred volume percent to mole percent. The estimated mole percent of each phase in product powders is shown in Fig. 2. The solid lines in the figure are aid to the eye and have no physical significance. When the molar ratio of CH4:SiO2 = 0.8:1, there is more than 25 mol.% of SiO2 phase in product powders. Carbon peaks were not observed in the XRD pattern of the product powders. More than 25 mol.% of SiO2 phase in product powders may result from excess SiO2 input. On the contrary, when the molar ratio of CH4:SiO2 = 1.5:1, carbon peaks is significant in the product powders. SiO2 peaks were not observed in the XRD pattern of the product powders. The carbon peaks indicate that excess CH4 was input. When the molar ratio of CH4:SiO2 = 1:1, the product powders are composed of more than 91 mol.% of SiC phase and less than 9 mol.% of SiO2 phase. The SiO2 phase in product phase may be due to the flow rate vibration of CH4. Therefore, the total synthesis reaction of SiC can be expressed in the following equation: SiO2 þ CH4 ¼ SiC þ 2H2 O
(2)
Fig. 3 is a typical XRD pattern of product powders which clearly confirmed the formation of b-SiC using SiO2 and CH4 as starting materials. An electronic diffraction pattern (Fig. 4) of the sample in Fig. 5 appears to be that of a FCC lattice with alternate pattern of a pair and a single peak confirming that the diffraction pattern is that of b-SiC. The calculated results from XRD pattern indicated that the lattice parameter (0.43635 nm) of product b-SiC powders is little bigger than its standard value (0.435853 nm from JSPDS card 74–2307). Table 1 Experimental conditions Run no.
Input power (kW)
Feeding rate (g/min)
Molar ratio (CH4:SiO2)
1 2 3
25.6 25.6 25.6
5 5 5
1:1 1.5:1 0.8:1
2306
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
Fig. 2. Relationship of SiC, SiO2 and C in product powders with molar ratio of CH4:SiO2.
Fig. 6 is a typical SEM image of SiC product. Fig. 7 is a magnified version of the micrograph in Fig. 6. The SiC product is the mixture of particles and fibers as shown in Fig. 6. Most particles are less than 100 nm. The fibers’ length is around from 50 to 900 nm and their diameter is around 30 nm as shown in Fig. 7. A Quantachrome NOVA 1200E surface area analyzer was used to measure the average surface area of product particles. Assuming the sample particles to be spherical, the average diameter of the product particles can be obtained from the following equation: 6 d¼ (3) Ar where d is the average diameter of SiC powders, A the surface area of sample powders per gram and r is SiC density. The BET results showed that the average diameter of product particles is around 50 nm. The result is less than what was observed and measured in SEM images of the product powders. As shown in Fig. 6, SiC particles are not perfectly spherical and the SiC fiber surface exhibits a corrugated morphology. These two reasons increased total measured surface area of the sample particles and thus result in the small calculated value of powder diameter. As shown in Figs. 6 and 7, a group of fibers nucleated on a substrate of a condensed particle, and grew in outward direction. The fibers could not keep the exact same growth direction and the SEM images indicate that fibers interwoven in an irregular manner. The SiC fibers appear to grow along a direction [1 1 1] in a stacking manner as
Fig. 3. XRD pattern of product powders from run no. 1 with molar ratio of CH4:SiO2 = 1:1.
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
2307
Fig. 4. An electronic diffraction pattern of the sample in Fig. 5.
shown in Figs. 5–7. High–density stacking faults observed by Wu et al. [18] confirmed the growth mechanism. The relationship of these two directions is unclear at present. Some researcher groups [18,19] reported that there is a tilt angle between the normal of stacking faults and the axial direction of SiC nano-fiber. Most of literatures [20,21] reported that the normal of stacking faults is parallel to the axial direction of SiC nano-fiber. It is well worth notice that
Fig. 5. TEM image of product powders from run 1 with input power 23.4 kW, feeding rate 5 g/min and molar ratio of CH4:SiO2 = 1:1.
2308
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
Fig. 6. SEM image of product powders from run 1 with input power 23.4 kW, feeding rate 5g/min and molar ratio of CH4:SiO2 = 1:1.
the fiber surface is not very smooth and exhibit a corrugated morphology. Wu [19] explained the corrugated surface is attributed to the relative shift of two adjacent stacking layers. When SiC supersaturated gas phase was quenched, it is assumed that SiC homogeneously nucleated and grew in the supersaturated gases in the plasma reactor as shown in Fig. 8(a). Due to the surface tension effect, spherical powders could be obtained. On the other hand, if a group of SiC heterogeneously nucleated on the surface of condensed SiC on the quench coil as shown in Fig. 8(b), nano-fibers could be formed due to directional cooling. By redesigning the quenching system, SiC powders could be avoided. The reasons why the product is the mixture of nano-powders and nano-fibers and why the fibers appear entangled are not yet clear. Three of experimental runs indicated that the mixture products of nano-powders and nano-fibers can be obtained at three different molar ratios. These experimental results imply nano-fibers can be easily synthesized using SiO2 and CH4 as starting materials by thermal plasma technology. By optimizing experimental parameters of the thermal plasma technique and redesigning the quenching system, pure nano-fibers of SiC could be synthesized. This technique can avoid using expensive carbon sources or catalysts for SiC nano-fiber or nano-wire synthesis. The most important is that this technique is promising for SiC nano-fibers synthesis in higher production rate because the feeding rate of starting material can go up to 10 g/min in the experimental setup as mentioned in experimental conditions.
Fig. 7. Micrograph in Fig. 6 at higher magnification.
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
2309
Fig. 8. Models of SiC nucleation and growth of: (a) SiC nano-powders and (b) nano-fibers in plasma reactor.
4. Summary and conclusions 1. If SiO2 and CH4 were used as the starting materials, the experimental results showed that the total synthesizing reaction of SiC can be expressed in the following equation: SiO2 þ CH4 ¼ SiC þ 2H2 O The experimental results did not indicate the obvious dependence of product shape on mole ratios of CH4:SiO2. 2. The experimental results showed that the mixture of SiC nano-powders and nano-fibers can be synthesized using SiO2 and CH4 as the starting materials by thermal technology. The average diameter of nano-powders is smaller than 100 nm. The fibers’ length is from 50 to 900 nm and their diameter is around 30 nm. 3. The SiC fibers appear to grow along a direction [1 1 1] in a stacking manner. The SiC nano-fiber surface is not very smooth and exhibit a corrugated morphology. 4. Experimental results imply nano-fibers can be easily synthesized using SiO2 and CH4 as starting materials by thermal plasma technology. By optimizing experimental parameters of the thermal plasma technique pure nanofibers of SiC may be synthesized in higher production rate.
Acknowledgements The authors are pleased to acknowledge Dr. Divakar Mantha for discussing and reviewing this paper and the financial support for this research by the U.S. Department of Defense (Grant number DAAD 19-01-1-0137). References [1] M.I. Boulos, P. Fauchais, E. Pfender, Thermal Plasmas Fundamentals and Applications, vol. 1, Plenum Press, New York and London, USA, 1994, p. 37. [2] H.O. Pierson, Handbook of Refractory Carbides and Nitrides, William Andrew, Noyes, 1996, p. 137. [3] L.N. Satapathy, P.D. Ramesh, D. Agrawal, R. Roy, Mater. Res. Bull. 40 (2005) 1871. [4] I.S. Seong, C.H. Kim, J. Mater. Sci. 28 (1993) 3277. [5] J.Y. Guo, F. Gitzhofer, M.I. Boulos, J. Mater. Sci. 30 (1995) 5589. [6] Y. Li, Y. Liang, Z.J. Hu, J. Am. Ceram. Soc. 77 (1994) 1662. [7] W. Yang, H. Araki, S. Thaveethavorn, H. Suzuki, T. Noda, Appl. Surf. Sci. 241 (2005) 236. [8] H. Dai, E.W. Wong, Y.Z. Liu, S. Fan, C.M. Lieber, Nature 375 (1995) 769. [9] C.C. Tang, S.S. Fan, H.Y. Dang, J.H. Zhao, C. Zhang, P. Li, Q. Gu, J. Cryst. Growth 210 (2000) 595. [10] X.T. Zhou, N. Wang, C.K.A. Frederick, H.L. Lai, H.Y. Peng, I. Bello, C.S. Lee, S.T. Lee, Mater. Sci. Eng. A 286 (2000) 119. [11] G.W. Meng, L.D. Zhang, C.M. Mo, S.Y. Zhang, Y. Qin, S.P. Feng, H.J. Li, J. Mater. Res. 13 (1998) 2533. [12] P.R. Taylor, S.A. Pirzada, Metall. Trans. B 23B (1992) 443. [13] L. Tong, R.G. Reddy, Scr. Mater. 52 (2005) 1253. [14] S. Niyomwas, B. Wu, R.G. Reddy, in: B.S. Mishra, S.L. Semiatin, C. Suryanarayana, N.N. Thandhani, T.C. Lowe (Eds.), Ultrafine Grained Materials, TMS, Warrendale, PA, 2000, 89. [15] P.R. Taylor, M. Manrique, JOM 6 (1996) 43. [16] R.G. Reddy, Metall. Mater. Trans. B 34 (2003) 137.
2310
L. Tong, R.G. Reddy / Materials Research Bulletin 41 (2006) 2303–2310
[17] L. Tong, R.G. Reddy, Mathematical modeling for processing of fine powders by thermal plasma, in: H. Mahfuz, M.V. Hosur (Eds.), Developments in Theoretical and Applied Mechanics, vol. XXII, 2004, p. 656. [18] Y.J. Wu, J.S. Wu, W. Qin, D. Xu, Z.X. Yang, Y.F. Zhang, Mater. Lett. 58 (2004) 2295. [19] Z.L. Wang, Z.R. Dai, P.R. Gao, Z.G. Bai, Appl. Phys. Lett. 77 (2000) 3349. [20] Y.H. Gao, Y. Bando, K. Kurashima, T. Sato, J. Mater. Sci. 37 (2002) 2023. [21] Y.J. Zhang, N.L. Wang, P.R. He, X.H. Chen, J. Zhu, J. Solid State Commun. 118 (2001) 595.