The preparation of activated carbon from waste biomass has a long history  -  , but still now attracting attentions from various ecological standpoints of view    and being reviewed  .
It has been reported from this laboratory that the decomposition of such biomass as sugar cane bagasse/sawdust  or cellulose/lignin  with argon plasma induced by microwave irradiation afford effectively the mixture (syngas) of hydrogen and carbon monoxide. This suggests alternative utilization of biomasses as gaseous or liquid fuels. The rate of the decomposition was extraordinarily rapid because of the locally (not in the bulk) heating of materials at a very high temperature by plasma. The comparison of the microwave plasma heating with the conventional thermal heating or the microwave-assisted heating (non- plasma) has been examined with respect to the preparation of activated carbon from phenolic fiber  . The superiority of the incidental energy consumption per specific surface area of the activated carbon by the plasma heating over the ordinary heating was elucidated in this report.
On the other hand, the preparation of activated carbon from such biomasses as bagasse, rice straw and bamboo, has been performed by alkali activation with NaOH by conventional thermal heating  . The process has been conducted by two steps, carbonization and activation. The characteristics of the derived carbonized material were to remain their original plant structures as capillaries. After the activation, they showed rather higher electric double-layer capacitances because of higher mesopore distribution.
These prompt us to try to prepare NaOH-activated carbon from sugar cane bagasse by the microwave plasma heating. The objective of the present work is to examine the possibility of the production by one step in a short time and the characteristics of the produced carbon.
Sugar cane bagasse was provided from SHOUNAN SEITOU (Okinawa, Japan). NaOH of reagent grade was purchased from Wako Chemicals Co.
Sugar cane bagasse was dried in air at 373 K for 24 h. Then it was impregnated in a solution of NaOH to sugar cane bagasse weight ratio, RNaOH, from 0.3 to 3.5. The mixture was dried at 353 K under reduced pressure for 1 h, being used as the starting material.
The carbonization of biomass and its activation were performed with one process in this project.
2.3. Apparatus of the Microwave Argon Plasma Heating
The experimental apparatus were basically the same with our former reports   , but the gaseous products were not examined in this experiment. The schematic diagram of microwave plasma heating is shown in Figure 1. The low-pressure flow type reaction apparatus, composed of gas supply unit, microwave generator, reactor, and low-pressure unit, was applied. The details of the plasma furnace are shown in Figure 2. Microwave was generated by an air- cooled magnetron (TOSHIBA 2M164) and then introduced into the plasma
Figure 1. Schematic diagram of the experimental apparatus of microwave plasma heating.
Figure 2. Details of the plasma furnace.
furnace (a quartz glass tube) through a horizontally placed waveguide (height 27 mm, width 96 mm) that complied with the appropriate specifications. A quartz glass tube (inside diameter 25 mm, length 500 mm was placed vertically in the plasma furnace. Reflected waves were reduced using a three-stub tuner and then removed by an isolator.
The starting material 1.0 g was held on silica wool bed in the middle of quartz glass tube. Argon gas was introduced from upside into the quartz glass tube at flow rate of 20 ml/min and setup to reduce pressure of 4.0 kPa. Microwave generated for 2 - 10 min at the power of 300 W was irradiated to produce the plasma state.
After the carbonization/activation of the starting material by the irradiation of prescribed time, NaOH was washed away from the resultant carbonaceous materials by using 1 M-HCl and ion-exchanged water until the pH 7 and dried at 353 K.
2.4. Characterization of Activated Carbon
The characterization of pore structure was done by measuring N2 adsorption isotherm at 77 K with an Accelerated Surface Area and Porosimetry System (Micromeritics, ASAP2020). Total surface area (Stotal), meso and exterior surface area (Smeso+ext) and micropore volume (Vmicro) were calculated by αs-plot analysis. Mesopore volume (Vmeso) was defined as the value of subtracting Vmicro from Vtotal. Average pore diameter (Wav) was calculated from the value of (Vtotal)2 devided by micropore surface area (Smicro = Stotal − Smeso+ext).
The electric double-layer capacitance (Cdl) was examined by a standard three electrodes cell. The electrode was composed of the prepared activated carbon, carbon black with a specific surface area of 130 m2/g and poly (tetrafluoroethylene) (PTFE), mixed in a weight ratio of 85:10:5. The mixture of 0.05 g was pressed onto the disk of 13φ. 1 M-H2SO4 aqueous solution was used as the electrolyte. The charge-discharge curve was measured in the voltage range from 0 V to 1.0 V at a constant current of 100 mA/g. The value of Cdl was calculated from the slope of discharge curve.
FT-IR spectra of activated carbons were recorded with a spectrometer FT/IR- 410 (JASCO Co.) by KBr pellet (activated carbon: KBr = 1:200).
ESR measurements were performed by use of an ESR spectrometer JES-FA200 (JEOL Ltd., Tokyo, Japan). Powdered samples of activated carbon were transferred into ESR sample tube and the spin intensity was evaluated via signal area of Mn2+/MgO marker. Experimental conditions were as follows: modulation frequency, 9.4 GHz; magnetic field, 337 ± 7.5 mT; microwave power, 1 mV; response time, 0.03 sec, modulation width, 1 mT; amplitude, 1 × 100; Mn marker, 600; sweep time, 30 sec. All measurements were operated at room temperature.
3. Results and Discussion
3.1. Evaluation of the Production Efficiency
Activated carbon yield (Y) was calculated from the weight ratio of the prepared activated carbon to the raw material, bagasse.
The most efficient condition of RNaOH and the irradiation time for the production of surface area was evaluated by YS-Y diagram  . YS is the value of multiplied the specific surface area (S = Stotal) by the yield of activated carbon. In YS-Y diagram, the upper right corner of data indicates the most excellent process condition.
Figure 3 shows the YS-Y diagram of NaOH-activated carbon in various conditions. When the RNaOH is changed, NaOH-activated carbon with the specific surface area of 1000 m2/g and over is produced at the RNaOH ranging from 1.0 to 3.0. The highest production of surface area in these activated carbons is performed at RNaOH of 1.0. NaOH-activated carbon with the specific surface area of 1980 m2/g is the highest YS obtained at the irradiation time of 3min. As the results the most suitable condition for production of surface was found to be the irradiation time of 3 min and RNaOH of 1.0 in the production of NaOH-activated
Figure 3. Surface developing diagram with the change of the alkali ratio RNaOH and the plasma irradiation time.
carbon from bagasse using the plasma heating.
This means the energy consumption by the plasma heating should be far less from our previous process by the conventional electric heating (carbonization by 1 h and activation by 1 h)  and the microwave-assisted heating     . Comparing with our previous studies on the NaOH-activated carbon from sugar cane bagasse with the conventional heating being conducted in two steps, carbonization and activation   , the present process using plasma heating showed a comparable surface area within a very short operation time of carbonization and activation. This might be the effect of ionization and active spices in plasma on activation mechanism of NaOH  . The plasma heating might induce not only the effect of local heating of the material in the cold state in bulk, but also the activation by active spices such as cation, anion and radicals. The energy efficient process may afford another ecological utilization of biomass.
3.2. Surface Area and Pore Structure of the Produced Carbon
N2 adsorption isotherms of NaOH-activated carbons at the irradiation time 10 min are shown in Figure 4 with RNaOH ranging from 0.5 to 3.0. The N2 adsorption volume increases with the increase in RNaOH at each P/P0 (relative pressure). According to IUPAC classification, the N2 adsorption-desorption curves of NaOH-activated carbon at RNaOH ranging from 0.5 to 2.5 were Type I isotherm with a less hysteresis loop. This adsorption behavior shows that the NaOH-activation carbon has microporous structure. However, the adsorption behavior of NaOH-activated carbon at RNaOH of 3.0 increases with the increase in P/P0 of 0.4 and over with an apparent hysteresis loop, being classified into Type IV of IUPAC definition which suggests the existence of mesoporous structure.
Figure 5 shows N2 adsorption isotherms of NaOH-activated carbon (RNaOH = 3.0) varied with the irradiation time. N2 adsorption curve of NaOH-activated
Figure 4. N2 adsorption/desorption isotherms of activated carbons prepared with different RNaOH (the irradiation time: 10 min).
Figure 5. N2 adsorption/desorptin isotherms of activated carbons with different plasma irradiation time (RNaOH: 3.0).
carbons at the irradiation time of 2 min is classified into Type I of IUPAC as rich in microporous structure. However, that of NaOH-activated carbons at the irradiation time of 3, 5, 10 min were fund to increase with the increase at P/P0 0.4 and over with hysteresis loops. Their adsorption-desorption behaviors are classified into Type IV. Therefore, the increase of the irradiation time also induced the migration of the pore structure from microporous to mesoporous. The maximum adsorption volume was obtained at the irradiation time of 3 min in accordance with the maximum distribution of mesopore.
The summaries of specific surface area (total (Stotal), meso plus exterior (Smeso+ext)), pore volume (total (Vtotal), meso (Vmeso)) and average pore size (Wav)) of NaOH-activated carbon obtained at varied RNaOH and the irradiation time are listed in Table 1. The data of Stotal and Vmeso were increased with the increase in
Table 1. Characteristics of NaOH-activated carbons from sugar cane bgasse by the argon plasma heating, referring with the conventional heating.
*based on dried bagasse weight; **the previous work  .
RNaOH to 1393 m2/g and 0.49 ml/g, respectively, at RNaOH of 3.0 with the irradiation time of 10 min. With respect to the starting material of RNaOH of 3.0, the maximum Stotal and Vmeso of 1980 m2/g and 0.73 ml/g, respectively, were obtained at the irradiation time of 3 min. However, the longer irradiation time over 5 min brought in the decrease in both of specific surface area and pore volume.
It has been shown by Liou’s intensive work  on the utilization of biomass that more higher Stotal and Smeso+ext are obtained from sugar cane bagasse by the activation with ZnCl2 (2 - 4 multiplied, at 723 K, 1 h). The economical assessment shall be expected for further practical and ecological application. Moreover, the activated carbon obtained from sugar cane bagasse activated with ZnCl2 (at 1173 K, 1 h) has been reported by Rafford et al.  to show Vmeso up to 1.55 ml/g. They marked the double-layer capacitance with the activated carbon close to 300 F/g in 1 MH2SO4 with a two-electrode, sandwich-type cell. This value might be attributable to the high mesopore volume and also the higher existence of oxygen and nitrogen on the carbon surface affecting favorable in aqueous solvent.
3.3. Electric Double-Layer Capacitance
It is well known that the electric double-layer capacitance (EDLC) with activated carbon increases with an increase of the area of electrode, but also the presence of mesopore structure has an important role in the formation of effective double-layer  . As shown the increase of mesopore distribution with respect to the plasma-heating NaOH-activated carbons, their EDLC were examined with a standard three electrodes in 1.0 M-H2SO4 solution.
The measured data of capacitance (Cdl) are added in the list of Table 1. The highest value (201 F/g) of Cdl was obtained at the NaOH-activated carbon with Stotal of 1980 m2/g and Vmeso of 0.73 ml/g. It shows the highest specific surface area in the activated carbon obtained with various conditions in this project, but also rich in the mesopore distribution. In our previous study  , the activated carbon with a higher specific surface area (SBET 2871 m2/g) obtained from carbonized sugar cane bagasse pith using an electric furnace (carbonization 1 h and activation 1 h with RNaOH 6.0, at 1073 K) afforded the value of Cdl as much as 109 F/g. The value almost half to the present data in this project might come from the lower volume of mesopore (Vmeso 0.27 ml/g).
3.4. Surface Analysis
It could be considered that the content of oxygen functional groups, other than the surface area and the mesopore volume, would become the factor of increasing Cdl in an aqueous cell solution, because of the hydrophilic nature of the mesopore surface. As for ZnCl2-activated carbons, the existence of oxygen group on the surface has been detected by FT-IR spectra  or X-ray photoelectron spectra  .
However, the FT-IR spectra (Figure 6) of the NaOH-activated carbons showed no absorption band corresponding to oxygen groups at all without regard to the heating methods, the conventional heating and the plasma heating. This might come from the reductive process existing in the activation mechanism of carbon with NaOH, which has been proposed by Lillo-Rodenas et al.  and followed by us  . Nevertheless, the elimination of oxygen from the surface by the argon plasma bombardment would not be excluded, because it has been reported that the thermal plasma treatment of activated carbon induces the disordered structure along with the decrease of oxygen on the surface, showing a smaller impedance change of EDCL  .
Figure 6. FT-IR spectra of NaOH-activated carbons.
The observation of ESR signal (Figure 7) of the activated carbon prepared by the argon plasma heating demonstrates the existence of a stable free radical on the surface as well. Far less signal intensity (relative) was detected with respect to the activated carbon by the conventional heating of sugar cane bagasse as compared in Table 2. This suggests the possibility of the trapping of plasma radical on the surface. The generation of free radicals by γ-ray irradiation on the surface of lignite-base CO2 activated carbon has been reported by Erҫin et al.  , presenting the higher g-factor values which might relate to some oxygen groups on the surface.
Table 2 listed the g-factor value of the signal with NaOH-activated carbon by the plasma heating accompanying with the one by the conventional heating for reference. The g-value 2.0028 suggests the radical is corresponding to the carbon-centered one. This is coincident with the information from the FT-IR measurement showing no oxygen functional group.
It was elucidated in this work that, within a very short operation time, sugar cane bagasse could convert to activated carbon with NaOH by heating with the argon plasma generated by microwave. By this technique, the high efficiency to give higher specific surface area was obtained, but also the development of mesopore distribution was observed markedly in comparison to the conventional heating system with an electric furnace, previously reported.
Figure 7. ESR spectra of NaOH-activated carbons.
Table 2. ESR data of NaOH-activated carbons from sugar cane bagasse.
Distinctive characteristics of NaOH-activation with the plasma heating were the absence of oxygen group and the existence of stable carbon-centered free radical on the surface of the produced activated carbon. This might be a fact of interest, how the carbon surface structure would stabilize the free radical, and what would be the effect of the surface radical on the enhancement of Cdl value.
The further investigations are expected for the capacitance probability in organic fluid with the present NaOH-activated carbon prepared by the plasma heating procedure for the practical usage.