Preparation of carbon nanomaterials using two-group

Chemical Engineering Journal 303 (2016) 217–230

Contents lists available at ScienceDirect

Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej

Preparation of carbon nanomaterials using two-group arc discharge plasma D.L. Sun a, R.Y. Hong a,b,⇑, J.Y. Liu a, F. Wang b, Y.F. Wang c a College of Chemistry, Chemical Engineering and Materials Science & State and Local Joint Engineering Laboratory for Novel Functional Polymeric Materials, Soochow University, SIP, Suzhou 215123, China b School of Chemical Engineering, Fuzhou University, Fuzhou 350108, China c NanoComp Co. Ltd., Suzhou SND 215011, China

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

A new preparation method of carbon

materials based on two groups of plasma was achieved. One combination of AC arc plasma and DC arc plasma was adopted. One combination of non-thermal plasma and thermal plasma was adopted. Optimal operation parameters and CB generation mechanism were determined. A technical process combined a fluidized bed with plasma system was designed.

a r t i c l e

i n f o

Article history: Received 10 March 2016 Received in revised form 15 May 2016 Accepted 22 May 2016 Available online 24 May 2016 Keywords: Plasma Carbon blacks Arc discharge

a b s t r a c t Carbon nanomaterials were prepared using two-group arc discharge plasma method in this study. A combination of an alternating current (AC) arc discharge plasma and a direct current (DC) arc discharge plasma in a fluidized bed was used, in which propane was cracked into carbon blacks (CB) with controlled structure and hydrogen. The effects of parameters such as electrode type, frequency of AC arc discharge, current of DC arc discharge and gas flow ratio (propane/argon) on the synthesis and properties of CB were investigated. A systematic study of the size, morphology, microstructure and surface chemical composition of CB was conducted using various characterization techniques. By using this process, spherical and hydrophobic CB with a narrow size distribution, high degree of graphitization and large specific surface area were produced under optimal experimental conditions. Then, as-prepared CB were used to adsorb heavy-metal ions in water and to reinforce the engineering plastic. The results verified that the adsorption rate of the Cr (VI) ion was increased significantly and that the CB enhanced the electrical and mechanical properties of acrylonitrile–butadiene–styrene (ABS)/natural rubber (NR) composites. Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction

⇑ Corresponding author at: College of Chemistry, Chemical Engineering and Materials Science & State and Local Joint Engineering Laboratory for Novel Functional Polymeric Materials, Soochow University, SIP, Suzhou 215123, China. E-mail address: [email protected] (R.Y. Hong). http://dx.doi.org/10.1016/j.cej.2016.05.098 1385-8947/Ó 2016 Elsevier B.V. All rights reserved.

Carbon nanomaterials such as fullerenes, nanotubes, onions, spheres and fibers have been the subject of much research for over two decades due to their excellent physical, chemical and electrical properties [1–3]. These materials have abundant applications as semiconductor materials, hydrogen energy storage media, catalyst

218

D.L. Sun et al. / Chemical Engineering Journal 303 (2016) 217–230

supports, composite fillers and reinforcing components [4–7]. Among them, carbon blacks (CB) are one of the most common raw chemical materials and are used as reinforcing filler for rubber goods in tire manufacturing and as pigment for printing inks, coatings and plastics [8,9]. CB are formed by the incomplete combustion of fossil fuels or by thermal decomposition of hydrocarbons, and the formation of CB depends on either the presence or absence of oxygen. Agglomerated primary nanoparticles were obtained from these mechanisms of CB formation [10]. Two growth mechanisms were used to form CB: radial growth and thermal cracking [11,12]. In the industrial production, CB are mainly manufactured using the furnace and channel processes. [13,14]. The furnace process for thermal cracking, which is the predominant method for the production of CB, has been used to meet the demands of the rubber industry for such a long time. This continuous process is operated in a closed reactor and produces a large amount of polluting emissions, including COx, SOx and NOx. Because energy efficiency and environmental protection have become more important, this technology has an ambiguous future [15]. Therefore, production via plasma has emerged as a potential solution. In the laboratory, several conventional methods, such as chemical vapor deposition (CVD), laser ablation and arc discharge plasma, have been widely used for preparation of carbon nanomaterials such as CB and carbon nanotubes [16]. However, the CB produced by the CVD method are always contaminated by residual catalytic particles, and the process requires a high operating temperature [17]. Barriers to the commercial adoption of the laser ablation method include its high energy consumption and the fact that it is a non-continuous process. The plasma method is fast and efficient and is widely used at present. Thermal decomposition and plasma systems are more promising than other approaches due to their lower energy consumption [18]. There are two main categories of plasma methods used in materials engineering: thermal plasma and non-thermal plasma. Thermal plasma (thermodynamic equilibrium) is generally operated under high-currents conditions (higher than 1 A). Thermal plasma systems generate plasmas with high temperatures, and all of the systems have the same temperature [11,19]. The thermal plasma concept for CB production has led to the development of ‘‘carbon black and hydrogen” using Kvaerner’s process [20], and the three-phase arc plasma process by Fulcheri [21]. Additionally, carbon nanostructures can be continuously synthesized using induction thermal plasma technology developed by Soucy and Pristavita [22–26]. The non-thermal plasma (non-thermodynamic equilibrium) approach is generally operated under low current conditions (lower than 1 A), which can provide an electron temperature of 4000–10,000 K and heavy particle temperature of 2000–6000 K. The temperature of these heavy particles (neutrals, ions) can be lower than that of the electrons [11]. The working gas is activated to create highly energetic electrons and reactive species to initiate plasma-assisted chemical reactions [27]. Therefore, non-thermal plasma systems offer highly selective and energy-efficient of chemical reactions. Numerous researchers have attempted to improve the synthesis of carbon nanomaterials by optimizing conditions and to develop a continuous and large-scale process by using the arc discharge plasma method [28,29]. The plasma methods for the preparation of CB that exist in the current literature are listed in Table 1 [11,13,21,30–39]. However, there are some drawbacks to the conventional arc discharge plasma method that preclude its use in industrial and large-scale applications, such as being a noncontinuous process and having poor synthesis purity. On the one hand, there are significant problems to the direct current (DC) discharge method including the necessity of the maintaining the pressure in the reactor below atmospheric pressure and the difficulty of obtaining a large area and uniform discharge. Additionally, in the case of DC arc discharge, carbon vapors aggregate and drift towards

the cathode due to the temperature gradient, leading to bridging between the adjacent electrodes. On the other hand, it is still difficult to manufacture massive nano-scale materials by AC arc discharge. From the standpoints of the preparation of CB and of their chemical applications, the thermal discharge process always leads to higher gaseous temperature and has a relatively narrow tolerances for the experimental equipment. However, the gaseous temperature in the conventional non-thermal discharge process is too low for the thermal cracking of a wide range of CB. Therefore, to develop a new approach to overcome these drawbacks, we combined the advantages of both thermal and non-thermal plasma systems by developing powerful discharges, along with a method to control the discharge pattern (AC or DC arc discharge). In our previous work [34], a non-thermal AC arc discharge plasma method for the preparation of carbon materials has been designed. Here, we demonstrate a promising, large-scale and continuous method for preparing CB. Non-thermal and thermal plasma (AC and DC arc discharge) processes are introduced. Usually, the distance between the electrodes is only 2–3 mm in DC arc discharge [40,41]. Now from a number of onsite tests, the gap between the second pair of electrodes could be as much as 50 mm with the help of the successful arc starter by AC arc discharge. The purposes of this work are (i) to develop an original two-group arc discharge plasma process, based on the establishment of thermal DC discharge at high current–low voltage and non-thermal AC discharge at low current–high voltage, operating at atmospheric pressure for the gas phase synthesis of CB; (ii) to investigate the versatility of the process with respect to the synthesis of carbon nanotubes (CNTs) other than CB; and (iii) to understand CB and CNTs formation mechanisms in such twogroup discharge. Previous studies on the decomposition of propane were performed in a fixed or fluidized bed reactor [42,43]. During fluidization, the residence time could be adjusted by changing the rate of the gas flow according to the CB weight. This process was used for the generation of carbon nanomaterials with different structures in a fluidized bed reactor. This work mainly focused on the relationship between the variation of morphology and microstructure of CB and the electrode materials and then developed a description of the formation process of CB. 2. Experimental set-up and procedures 2.1. Materials Argon (Ar) (purity 99.99%, Jinhong Gas Co., Ltd., China) and propane (C3H8) was used as a working gas and carbon source for the preparation of CB, respectively. Two-group parallel electrodes (3 mm and 6 mm in diameter) were used, which were made of different kinds of metals (iron or copper) in the plasma reactor. Natural rubber (NR), recycled acrylonitrile–butadiene–styrene (ABS) and maleic anhydride (MAH) were used. 2.2. Experimental set-up A schematic of the experimental apparatus is shown in Figs. 1 and 2. The decomposition of C3H8 was carried out in a fluidized bed consisting of three sections: an AC arc discharge plasma generator, a DC arc discharge plasma generator and an electric furnace. The functions of the gas (C3H8 and Ar) were to replace air that entered the reactor, form plasma for decomposition, completely fluidize the particles and transfer the products. The flow characteristics of CB were calculated and tested in this fluidized bed to determine the minimum fluidized velocity. The temperature in the slim tube and furnace was monitored by two thermocouples installed in two bed positions.

219

D.L. Sun et al. / Chemical Engineering Journal 303 (2016) 217–230 Table 1 The outline of plasma methods for the preparation of CB. Groups

Atmosphere

Carbon source

Type of discharge

Carbon materials product

Parameter

References

M.M. Couranjou et al.

Ethylene

AC

Carbon black with ‘‘crumpled paper sheet”

5 kW

[11]

L. Fulcheri et al.

Argon Carbon dioxide Helium

AC

Carbon black

50–100 kW

[21]

L. Fulcheri et al. J. Kang et al. X. Tu et al.

Helium Benzene Hydrogen

AC AC AC

Fullerenes and carbon black Carbon black Carbon black

250 kW 1.3 kV 220 V/10 kV

[30] [31] [32]

X.Y. Liu, R.Y. Hong et al. D.L. Sun, J.J. Yuan et al. T. Zielinski

Argon Argon Argon Helium

AC AC AC

Carbon black Carbon black and carbon nanotubes Carbon black

8 kW 20 kV

[33] [13,34,35] [36]

J.J. Guo et al.

Water

Methane Ethylene Styrene Pyrolysis fuel oil Carbon powder Benzene Methane Carbon dioxide Methane Propane Methane Ethane Acetylene Graphite

DC

Carbon black and carbon nanotubes

[37]

S.H. Park et al.

Nitrogen

Methane

DC

Carbon black

W.J. Cho et al.

Methane

Methane

Carbon black and C2

Our group

Argon

Propane

Microwave Radio frequency AC + DC

28 V 30–70 A 6 kW 300 A 120 W 1.2 kW 2 + 10 kW

Carbon black and carbon nanotubes

[38] [39]

Fig. 1. Schematic illustration of experimental equipment.

The arc discharge conditions were described as follows. Two power supply apparatuses, based on different technologies, were used to generate arc discharges. Apparatus 1 comprised a high voltage transformer in AC-mode configuration, and the arc intensity was controlled by adjusting the frequency. Apparatus 2 comprised high current–low voltage (410 V) arc lines. The current could be continuously adjusted from 1 to 25 A in our experiment. (i) In the AC arc discharge plasma generator, one custom-made ceramic nozzle was used because of its excellent thermal stability. The plasma was discharged from the parallel metallic electrodes (3 mm in diameter) on the top of nozzle. The arc was initiated at the distance between the electrodes when the applied voltage reached the critical value of the gas breakdown, and the arc was pushed by the gas in the direction of its flow. The gap between the first pair of electrodes was 20 mm due to the high voltage of the power supply apparatus 1 (0.1 A and 20 kV). (ii)

In the DC arc discharge plasma generator, a new arc was created from the point-to-point metallic electrodes (6 mm in diameter). The gap between the second pair of electrodes could be up to 50 mm. (iii) The heating furnace (300 mm in diameter) was closely entwined with the installed heating wire. The main component of the furnace was silicon nitride, which effectively prevented not only the damage related to the heat generated by the plasma discharge or electric heating but also large temperature gradients. 2.3. Parameters A comprehensive parametric study mainly focusing on the following parameters: metal electrode materials, AC arc discharge frequency, DC arc discharge current and gas flow ratio (C3H8/Ar). The operating conditions were summarized in Table 2.

220


introduced into the furnace (650 °C) at the desired flow rate controlled by flow controllers and a pump. The gas velocity was adjusted as appropriate to keep particles suspended in the furnace region with the help of upward-flowing gas, to ensure a sufficient reaction. Process 4 (separation and collection): By increasing the gas velocity, particles were pushed from the furnace into a long tube that was connected to the water cooler and cyclone where all of the particles were precipitated from the gas. Next to cyclone, a leaching tower was located for final remove of the dust from the gas. All the experiments were carried out at atmospheric pressure. 2.5. Characterization

Fig. 2. Schematic illustration of plasma generator: (a) sketch map of plasma generator: 1—gas inlet; 2—electrodes of AC; 3—ceramic nozzle; 4—bakelite case; 5— electrodes of DC; (b) photo of plasma generator exit; (c) photo of multibank arc discharge in the glass tube.

Table 2 Operating conditions for synthesis of carbon nanoparticles by plasma process. Name

Parameters

Details

Power supply (AC) apparatus 1

Voltage Current Frequency Electrodes

20 kV 0.1 A 10, 20, 30 kHz Copper, iron

Power supply (DC) apparatus 2

Voltage Current Electrodes

650 V 5, 10, 15, 20, 25 A Copper, iron

Gas

Hydrocarbon Plasma gas Total flow Flow ratio (C3H8:Ar)

Propane (C3H8) Argon (Ar) 1.6 m3/h 1:8, 1:6, 1:4, 1:2, 1:1

Electronic furnace

Temperature Reaction time Pressure

650 °C 45 min Atmosphere

2.4. Process Process 1 (primary plasma decomposition stage): Ar was injected into the reaction chamber for a few minutes to purge air from it. Then the generation of the AC arc discharge began and was changed to a continuous regime at atmospheric pressure. The gas flow was controlled by flow controllers. The C3H8/Ar gas, with varying ratios, was injected tangentially from the bottom of the fluidized bed. Once this gas passed through, the Ar formed a plasma, and the chemical bonds of hydrocarbons from C3H8 were cracked. Process 2 (further plasma decomposition stage): At the same time, the DC arc was initiated at the gap between the high and low voltage of apparatus 2. The plasma that was generated that produced positive and negative ions, along with neutral particles and free radicals. According to the published research, CB and hydrogen (H2) are the primary products and hydrocarbons are the secondary products [44]. Reactions took place immediately with active species reacting with the hydrocarbons that came from the plasma region. Process 3 (further thermal decomposition and growth stage): The main products (CB and H2) and surplus hydrocarbons were

2.5.1. Methods of sample characterization The morphology and microstructure of samples were examined by the scanning electron microscopy (SEM, Hitachi S-4700, Japan) with 15.0 kV beam voltage and transmission electron microscopy (TEM, Hitachi H-600-II, Japan). Particle size distribution was analyzed by a dynamic light scattering nanometer laser granulometer (DLS, Bettersize, BT-90, China). Raman analysis is a valuable tool to characterize microstructure and phase purity of samples which was performed with an XY multichannel Raman microspectrometer (JY-HR800) employing continuous wave lasers of 514 nm wavelengths. The laser spot size was 61 lm, with a spectral resolution of 1 cm1, and the spectrum was measured in the 0–4000 cm1 range. The crystal structure of particles was recorded by X-ray diffraction (XRD, Siemens D8 Advance, Germany) using Cu-Ka radiation (k = 0.15418 nm) in the 2h range of 20–80° (scanning rate of 6°/min). Thermal decomposition of samples was characterized by thermogravimetric analyzer (TGA, EXSTAR 6300, Japan) under air and N2 flow (20 mL/min) in the temperature range of 20– 1000 °C with a heating rate of 10 °C/min. Specific surface area of the carbon nanomaterials were characterized by physical adsorption tests. This was determined by exposing the carbon nanoparticles to flowing N2 at 77 K in an automated adsorption apparatus (Quantachrome ASAP, mMK-ASAP 2020M+C, USA). The measure range is 0.01–3000 m2/g for specific surface area (BET, Brunauer, Emmett and Teller surface area measurements). Oil absorption number (OAN) of CB was measured according to ASTM D241401. Measurements were performed five times for each sample, and then the average value was calculated. Element analysis was conducted using a scanning electron microscope appurtenance (SEM-EDS, Hitachi S-4700, Japan). The X-ray photoelectron spectroscopy (XPS) was determined using the Al Ka line with VG CLAMP hemispherical analyzer. The curve fitting and data analysis software assigned the peak locations by means of XPS-peak differenating analysis. 2.5.2. Absorption of Cr (VI) Potassium dichromate and deionized water were confected into a certain concentration solution (100 ml). Initial pH of the solution was adjusted to about 3 [45]. After adding CB, the bottles were sealed with preservative film and then shaken for a given time (20 min) at 20 °C at a frequency of 90 strokes/min using a shaking bath. Then each mixture was filtered to separate the supernatants and CB. The absorbed effect of Cr (VI) was measured with a UVspectrophotometer at an absorbance of 540 nm. 2.5.3. Electrical properties of ABS/NR composites filled with CB Preparation of ABS/NR/CB composites has been reported by previous work [46]. Briefly, the NR-g-MAH sample was prepared via a single screw extruder at 100 °C and 100 rpm. ABS was grafted by maleic anhydride in twin-screw counter-rotating internal mixer at 175 °C and 30 rpm for 3 min, and then mixed with the as prepared NR-g-MAH in proportion (55%:45%, wt%) sufficiently. Finally,


the sample was compounded with/without raw ABS resin in certain CB addition (10%, wt%) in an internal mixer at 50 rpm and 220 °C for 5 min. Electrical properties of wafer-shape samples with the size of £15 4 mm were measured using an insulating resistance meter assisted by a four-point probes resistivity measurement. 2.5.4. Mechanical properties of ABS/NR composites filled with CB Notched IZOD impact test was conducted using XJJ-50 machine according to GB/T 1043-93. Tensile tests were carried out using universal tester (WDT-20) at the speed of 50 mm/min after preparing materials to dumbbell shape via universal sampling machine according to GB/T 1040-92. The GB standards were established according to the ISO ones. 3. Results and discussion 3.1. Morphology and particle size of CB under different experimental conditions 3.1.1. SEM analysis 3.1.1.1. Effect of electrode material. Fig. 3 shows SEM images of the samples which were synthesized using different collocation of metal electrodes. Raw samples had two main components: carbon nanoparticles (carbon black) and carbon nanotubes. Two samples (Fig. 3(a) and (c)) had a similar shape, but their diameters and the dispersion effects of the CB were different. However, two samples (Fig. 3(b) and (d) both of which were produced using iron electrodes in the AC arc discharge process) consisted of carbon nanotubes and carbon nanoparticles. According to typical references [47,48], the presence of small iron particles or droplets primarily promotes the nucleation, growth and elongation of the innermost carbon shell to form carbon nanotubes. Compared to other samples, fewer carbon tubes appeared using copper electrodes in AC arc discharge process, which can be explained by the fact that iron had higher catalytic activity than copper. In this experiment, we observed that the types of electrodes used in the AC arc discharge process played a dominant role in the formation

221

of the tubes while the types of electrodes used in the DC arc discharge process did not influence the growth of CB. 3.1.1.2. Effect of AC arc discharge frequency. To understand the effect of AC arc discharge frequency on the morphology of CB from the collector, SEM analysis was performed. In Fig. 4(a)–(c) show that the CB synthesized at 30 kHz have a better dispersion effect, as only a few particles were non-uniform. Nearly spherical nanoparticles with diameters of approximately 20–60 nm can be clearly observed. The CB synthesized at 10 kHz are larger and exhibit a lower degree of structural order with significant aggregation. Hence, the dispersion effect can be influenced by applying different AC discharge frequencies. 3.1.1.3. Effect of DC arc discharge current. To determine the effect of the DC arc discharge current on the morphology of the CB, the reaction was performed using Cu–Cu AC electrodes at 30 kHz and Fe– Fe (DC electrodes, with the current varied from 5 to 25 A). As shown in Fig. 4(d)–(f), there was no significant differences in the average CB size or in the dispersion effect when the current was between 5 and 15 A. Moreover, the average size of the CB increased as the current increased current (above 15 A). The discharge power influenced the temperature in the discharge zone and the plasma temperature increased with increasing energy input [49]. Thus, a higher current (25 A) enhanced C3H8 decomposition, leading to a greater non-equilibrium driving force for the carbon nanostructure to grow. Additionally, according to the previous literature [13], the particle growth rate is quicker than the rate of nucleation at the high reaction temperatures, which results in the formation of larger particles. 3.1.1.4. Effect of gas flow ratio. To determine the effect of the gas flow ratio (C3H8/Ar) on the morphology of the nanoparticles, the gas flow ratio was varied from 1:8 to 1:1 while the total flow of the gas mixture was fixed at 1.6 m3/h and the current was fixed at 15 A. Fig. 4(g)–(i) show the SEM images of CB prepared at different gas flow ratios. The micrographs of the samples show that the average size increased with the increasing C3H8 flow and decreas-

Fig. 3. The SEM images of carbon materials obtained in soot using metal electrodes: (a) Cu (AC)–Cu (DC); (b) Fe (AC)–Fe (DC); (c) Cu (AC)–Fe (DC); (d) Fe (AC)–Cu (DC).

222


Fig. 4. The SEM images of CB synthesized at different frequencies: (a) 10 kHz; (b) 20 kHz; (c) 30 kHz, at different currents: (d) 5 A; (e) 15 A; (f) 25 A and under different gas flow ratios (C3H8:Ar): (g) 1:8; (h) 1:4; (i) 1:1.

ing Ar flow. As C3H8 flow increased, the higher concentration of C3H8 in the gas mixture resulted in a higher nucleation rate, and sub-particles generated instantly in the plasma region. Numerous species nucleated, grew, joined and formed CB. Since the density of nuclei was high, these particles gathered together quickly to form larger nanostructures. The rate of nucleation was high enough to allow for the growth of large spherical carbon nanostructures. 3.1.2. TEM analysis TEM was used to investigate the morphology, texture and structure of CB. Figs. 5 and 6 show the following: (i) In Fig. 5(a), carbon nanoparticles are completely present from this sample. However, a significant difference between the use of copper electrodes and the use of iron electrodes was observed, as the CB were not entirely tangled, and some individual nanotubes could be seen in Fig. 5(b). Carbon nanotubes grew during the decomposition of hydrocarbons in the presence of iron catalysts [50]. The iron electrodes evaporated and condensed into tiny metal nanoparticles or droplets during discharge, which promoted the formation of tubes. (ii) The HRTEM image in Fig. 5(c) shows that graphitic layers on the exterior of the particle were well arranged, maintained a constant distance between neighboring layers and were found to be parallel to the concentric direction. The image shows that the CB synthesized at 30 kHz, 15 A consisted of a well-graphitized multi-layer structure, with a smooth surface and clear interlayer spacing. (iii) In Fig. 6(a) and (b), the morphologies of the CB synthesized at 30 kHz retained a spherical structure while the CB synthesized at 10 kHz have angular shapes, indicating that high frequency could be used to promote the formation of spherical nanoparticles in an AC arc discharge plasma. (iv) As shown in

Fig. 6(c) and (d), the CB had spherical morphology with relatively consistent size, and their agglomeration is shaped like a bunch of grapes. The CB consisted of many small particles that fused together in aciniform aggregate structures, which are a type of a high-structure CB according to the literature [51,52]. Additionally, the particle size increased as DC arc discharge current increased, which agreed with the SEM result. (v) The sample synthesized under a higher C3H8 flow ratio had a larger particle size, as shown in Fig. 6(e) and (f). Meantime, there was an agglomerative phenomenon accompanying with the synthetic process. Carbon particles were formed from the reaction between the carbon source and plasma, these particles quickly escaped from the plasma zone due to the effect of gas flow. A higher nucleation rate also meant more sub-particles were generated. In the furnace, they floated randomly and then became interconnected in different directions, forming a three-dimensional network that became an agglomeration structure [31]. 3.1.3. Particle size analysis Fig. 7 shows the size distributions of CB synthesized under different gas flow ratios measured by DLS. All of the samples were sonicated for 30 min before analysis. For sample 5 (C3H8: Ar = 1:1), the distribution of the particle diameters was relatively wide, and the mean particle diameter was approximately 320 nm. The distribution of particle size was narrower when the concentration of C3H8 was reduced, and the average particle diameter decreased to 134 nm (sample 4). As the concentration was further reduced, the number of relatively large agglomerates decreased, the particle size distribution was significantly narrower and the average particle diameter decreased to approximately 65 nm (sample 1).


223

Fig. 5. The TEM images of carbon materials obtained in soot using metal electrodes of AC arc discharge: (a) Cu; (b) Fe and (c) HRTEM image of CB synthesized at 30 kHz and 15 A.

Fig. 6. The TEM images of CB synthesized at different frequencies: (a) 10 kHz; (b) 30 kHz, at different currents: (c) 5 A; (d) 25 A and under different gas flow ratios (C3H8:Ar): (e) 1:8; (f) 1:1.

224


Fig. 7. The size distributions of CB synthesized under different gas flow ratios (C3H8:Ar): (a) 1:8; (b) 1:6; (c) 1:4; (d) 1:2; (e) 1:1.

3.2. Element content of CB 3.2.1. EA analysis EA provided information about the concentration of elements in our samples. The results obtained for the amount of atomic carbon (C), nitrogen (N) and hydrogen (H) for different DC arc discharge currents and gas flow ratios are presented in Table 3. It should be noted that the surface atomic N percentage of CB cannot exceed 0.03%. Various carbon-to-hydrogen (C/H) ratios were compared to exclude the influence of N. Only after raising the current to 15 A did relatively high H atom percentage (P10%) appear (as seen in Table 3). As the current increased to 15 A or more, a stable value for the H atom percentage – approximately 11% – became apparent, as indicated in the table. The increasing current enhanced C3H8 decomposition and generated more hydrogen, leading to the increase in the amounts of H in the CB due to the adsorption effect. In addition, the table shows the element content on the surface of CB varied under different gas flow ratios. The relative C atom percentage gradually increased to 93.25% under a ratio of 1:8, while the relative H atom percentage decreased from 15.60% to 6.73% because of the decreasing amount of hydrogen. Hydrogen Table 3 Element analysis of CB. Sample

1 2 3 4 5 6 7 8 9 10

Conditions

Element relative percentage (%)

Current (A)

Gas flow ratio (C3H8: Ar)

C

N

H

H/C

5 10 15 20 25 15 15 15 15 15

1:4 1:4 1:4 1:4 1:4 1:1 1:2 1:4 1:6 1:8

96.17 92.03 89.49 88.61 88.53 84.39 86.58 88.61 91.73 93.25

0.02 0.03 0.01 0.02 0.03 0.01 0.01 0.02 0.01 0.02

3.81 7.94 10.50 11.37 11.44 15.60 13.41 11.37 8.26 6.73

0.04 0.09 0.12 0.13 0.13 0.18 0.15 0.13 0.09 0.07

molecules were adsorbed on the surface due to pore structure properties. Once the CB were treated with H2 in the furnace, the adsorption of hydrogen molecules began. 3.2.2. TGA analysis Thermal gravimetric analysis (TGA) was performed in air and nitrogen (N2) atmospheres to evaluate the purity of the samples and investigate the effects of different conditions on the products. The TGA results are shown in Fig. 8(a)–(g). The curve shows the existence of amorphous carbon in CB and that the residues were catalyst metal particles from the electrodes and impurities, which came from the previous preparation process. As shown in Fig. 8 (a) and (b), there were two burning temperatures at approximately 510 and 570 °C for sample 1 (10 kHz), 610 °C for sample 2 (20 kHz) and 820 °C for sample 3 (30 kHz), which can be attributed to amorphous carbon. The temperature of sample 3 was high compared with the temperatures of samples 1 and 2, which indicates that sample 3 has a higher thermal stability. However, the differential thermogravimetric (DTG) curves demonstrate that sample 3 had some small weight losses while sample 2 has only one, and sample 1 has two, suggesting that frequency played an important role in synthesizing CB with high thermal stability. As shown in Fig. 8 (c) and (d), there was only one weight loss step – the combustion of the CB – and the change point shifted to higher temperature after adjusting the gas flow ratio from 1:1 to 1:8, indicating that raw samples became much more stable and pure. Fig. 8 (e) and (f) show the TGA traces for samples synthesized at different currents during DC arc discharge. Burning of CB began at approximately 630 °C, and as the current increased from 5 to 25 A, the burning temperature first increased (