548
Energy & Fuels 2009, 23, 548–552
Synthesis of Biodiesel Using Microwave Absorption Catalysts H. Yuan,†,‡ B. L. Yang,*,† and G. L. Zhu† Department of Chemical Engineering, State Key Laboratory of Multiphase Flow in Power Engineering, Xi’an Jiaotong UniVersity, Xi’an Shaanxi 710049, P. R. China, and School of Chemistry and Chemical Engineering, North UniVersity of Nationalities, Yinchuan 750021, P. R. China ReceiVed July 23, 2008. ReVised Manuscript ReceiVed October 2, 2008
To enhance the synthesis process for biodiesel (fatty acid methyl ester, FAME), microwave absorption solid acid catalysts were used for transesterification under microwave radiation. The H2SO4/C catalyst was prepared by an impregnation method. The surface areas of active carbon and H2SO4/C were measured by the BET method. Synthesis reactions were carried out under different conditions using castor oil and methanol as the feedstock. The amounts of FAME in the product were analyzed by high-performance liquid chromatography. Experimental results showed that the microwave radiation exhibited a notable enhanced effect for transesterification by using the microwave absorption solid acid catalysts (H2SO4/C) compared with that of the conventional heating method. When the transesterification was carried out at 338 K, with 12:1 molar ratio of methanol to castor oil, 50 wt % loading amounts of H2SO4 and 5 wt % of catalyst to castor oil, after 60 min the yield of FAME 94 wt % was obtained. This method can also be used in the case of castor oil with high free fatty acid.
1. Introduction Biodiesel, an alternative diesel fuel that is also referred to as fatty acid methyl ester (FAME),1 has been a renewed focus because of the recent rise in crude oil price and environmental concerns.2 Biodiesel could be synthesized by transesterification using basic or acid catalyst. Conventionally, NaOH and KOH are used for the industrial production of biodiesel because the alkalicatalyzed transesterification can be accomplished in a short reaction time under mild reaction conditions. However, these homogeneous alkaline catalysts could easily react with free fatty acids (FFA) in the feed oil to form the unwanted soap and water byproduct. These byproducts would adversely affect the quality of biodiesel, and an expensive separation would be required to purify the biodiesel.3,4 The small amount of water generated in the saponification reaction and moisture contained in the feedstock oil may initiate oil hydrolyzation to form FFA and glycerol during transesterification.2 Therefore, the total FFA content and moisture in the feedstock oil should not exceed 0.5 and 0.06 wt %, respectively, when the homogeneous alkaline catalysts are used.5,6 However, the most profitable raw materials * To whom correspondence should be addressed. Tel.: +86-29- 82663189. Fax: +86-29-82668789. E-mail:
[email protected]. † Xi’an Jiaotong University. ‡ North University of Nationalities. (1) Ma, F. R.; Clements, L. D.; Hanna, M. A. Bioresour. Technol. 1999, 69, 289–293. (2) Ma, F. R.; Hanna, M. A. Bioresour. Technol. 1999, 70, 1–15. (3) Chai, F.; Cao, F. H.; Zhai, F. Y.; Chen, Y.; Wang, X. H.; Su, Z. M. AdV. Synth. Catal. 2007, 349, 1057–1065. (4) Shibasaki-Kitakawa, N.; Honda, H.; Kuribayashi, H.; Toda, T.; Fukumura, T.; Yonemoto, T. Bioresour. Technol. 2007, 98, 416–421. (5) Wang, Y.; Ou, S. L.; Liu, P. P. J. Mol. Catal. A: Chem. 2006, 252, 107–112. (6) Zullaikah, S.; Lai, C. C.; Vali, S. R. Bioresour. Technol. 2005, 96, 1889–1896.
(e.g., waste cooking oils and fats or low-value fats) usually have a high content of FFA7 and a certain content of water. The soap formation can be avoided by using acid catalysts, which can simultaneously catalyze both the transesterification of triglyceride and esterification of FFA to produce FAME. According to the acid-catalyzed reaction mechanism, a Brønsted acid catalyst is superior to a Lewis catalyst in producing biodiesel. Sulfuric acid is a good Brønsted acid whose acid strength (H0) is -11.93 and has a good catalysis performance. However, sulfuric acid will cause equipment corrosion, and it is difficult to separate from the product after reaction. Furthermore, catalyst-contaminated glycerol has little value in today’s market and is increasingly becoming a disposal issue.8 To solve these problems, the sulfuric acid could be loaded on the support to convert into a solid acid catalyst. However, if the solid acid catalyst is used, the reaction will experience a longer time, a higher reaction temperature, and a higher methanol/oil molar ratio compared with homogeneous-catalyzed reaction because of its heat and mass transfer disadvantages. Therefore, a microwave-assisted synthesis method was adopted in this research. Microwave radiation has a selective heating function,9 and some studies on microwave-assisted synthesis of biodiesel reported that microwave radiation can accelerate the reaction.10,11 For a heterogeneous catalysis reaction, the reaction takes place on the internal and external surfaces of solid catalyst. If the catalyst is a strong microwave absorption material, the radiation would directly act on the catalyst and the “microwave hot spots” (7) Vicente, G.; Martı´nez, M.; Aracil, J. Bioresour. Technol. 2004, 92, 297–305. (8) Singh, A. K.; Fernando, S. D. Energy Fuels 2008, 22, 2067–2069. (9) Thostenson, E. T.; Chou, T. W. Composites: Part A 1999, 30, 1055– 1071. (10) Hernando, J.; Leton, P.; Matia, M. P.; Novella, J. L.; Alvarez-Builla, J. Fuel 2007, 86, 1641–1644. (11) Leadbeater, N. E.; Stencel, L. M. Energy Fuels 2006, 20, 2281– 2283.
10.1021/ef800577j CCC: $40.75 2009 American Chemical Society Published on Web 12/09/2008
Synthesis of Biodiesel Using MicrowaVe Catalysts
would be formed under microwave radiation.12,13 The “microwave hot spots” are extremely favorable to the endothermal reaction. On the basis of the above considerations, the carrier of the catalyst should be a strong microwave absorbing material. Activated carbon is one of the best absorbing materials, having a high rate of temperature rise among the various materials that could be heated under microwave radiation.14 Furthermore, activated carbon has a significant surface area and steady physical properties. Therefore, it would be a proper material used as the microwave absorption carrier of the catalyst. At the same time, as a type of Brønsted acid, sulfuric acid, which is a polar material and loaded on the activated carbon, also has a strong microwave radiation absorbing function. With respect to the feedstock, methanol shows stronger microwave absorption capacity; therefore, it is better than ethanol in transesterification under the microwave condition. Also, if a nonedible oil and low-cost oil can be chosen as other feed material, the cost of biodiesel may significantly be reduced. On the basis of this consideration, castor oil was used to produce biodiesel in this work because castor is wildly planted in Brazil, China, and India and has a high yield of oil. Furthermore, castor oil contains one hydroxylated fatty acid (12-hydroxy-9-cisoktadecenoic acid), and the transesterification product shows much more effective lubricity than other oils that do not contain any hydroxylated fatty acids.15,16 This fact makes it an appealing candidate as a diesel additive.17 Castor oil also has stable properties in the air and is not easy to decay, and thus it can be stored in a storage stock for a long time. Castor oil as a biodiesel feedstock oil has been investigated by a few researchers. This work was designed to investigate the effect of the microwave radiation on transesterification when strong microwave absorbing materials such as solid acid catalyst and feed material were used. The affecting factors in the synthesis conditions including reaction temperature, reaction time, catalyst/ castor oil mass ratio, H2SO4 loading amount, and the methanol/ castor oil molar ratio were analyzed and optimized. Furthermore, castor oil contains different levels of FFA due to the different places of origin and the harvest season. To investigate the effect of FFA on the yield of FAME, a certain quantity of FFA was added to castor oil as the feedstock for biodiesel production under the microwave radiation. 2. Experimental Section 2.1. Preparation of the Catalyst. Wooden activated carbon was dipped into boiling deionized water for 3 h, then filtered and washed by deionized water several times. The activated carbon finally was dried at 393 K, and the mass was not changed. These operations could remove the ash from the internal pores of active carbon, by clearing the channels of the internal pores. The solid acid catalysts H2SO4/C were prepared by the impregnation method. Ten grams of pretreatment activated carbon was impregnated with 50 g of the sulfuric acid solution at 303 K overnight. Water was then removed in a rotary evaporator at 353 K with 0.01 MPa until dry. The obtained catalysts with different H2SO4 loading amounts were (12) Chen, C. L.; Hong, P. J.; Dai, S. S.; Zhang, C. C.; Yang, X. Y. React. Kinet. Catal. Lett. 1997, 61, 157–180. (13) Zhang, X. L.; Hayward, D. O. Inorg. Chim. Acta 2006, 359, 3421– 3433. (14) Jin, Q. H.; Dai, S. S.; Hong, K. M. MicrowaVe Chemical; Science Press: Beijing, 1999; p 84. (15) Drown, D. C.; Harper, K.; Frame, E. J. Am. Oil Chem. Soc. 2001, 78, 579–584. (16) Cvengros, J.; Paligova, J.; Cvengrosova, Z. Eur. J. Lipid Sci. Technol. 2006, 108, 629–635. (17) Goodrum, J. W.; Geller, D. P. Bioresour. Technol. 2005, 96, 851– 855.
Energy & Fuels, Vol. 23, 2009 549
Figure 1. Schematic diagram of the experimental apparatus. a, Condenser; b, control panel; c, furnace chamber switch; d, power switch; e, magnetic stirring adjustment knob; f, infrared temperature measurement; g, flask; h, display.
designated as 40, 45, 50, 55, and 60 wt % H2SO4/C, respectively. The surface areas of active carbon and H2SO4/C were measured by nitrogen adsorption experiments. 2.2. Transesterification Procedures. The acid value of castor oil was less than 2 mg of KOH g-1 from the acid value determination, and the average molecular weight of 918 g mol-1 was calculated from the saponification value (SV ) 183 mg of KOH g-1). Adding different quantities of oleic acid to castor oil made the castor oil with high FFA system. As shown in Figure 1, the microwave reactions were carried out in a microwave synthesis reactor (Shanghai Sineo Microwave Chemistry Technology Co., Ltd., MAS-1), working at 2.45 GHz rated at 200 W. Stirring was performed at 600 rpm with a magnetic nucleus. The temperature of the reaction mixture was maintained constant by the infrared detection system and a control circuit. Although some temperature variation could be observed under the influence of microwave radiation, the fluctuation range did not exceed 1 K. Methanol was analytical reagent (99.5 wt %), ethanol was also analytical reagent (99.7 wt %), and castor oil was chemical pure. The 20 g of castor oil (21 mL), 2.1-16 g of methanol (2.6-21.1 mL) or ethanol (15 mL), and 0.2-1.4 g of catalyst were added to the 250-mL three-necked round-bottomed quartz material flask equipped with a condenser. The reaction temperature was controlled in the range of 328-343 K under atmospheric pressure. The transesterification was stopped after 5-240 min. The conventional heating reactions were performed using a digital heating circulating water bath (Hangzhou Electrical Instrument Factory, D60-2F), equipped with a mechanical stirrer (Hangzhou Electrical Instrument Factory, D60-2F) and a condenser. Twenty grams of castor oil (21 mL), 8 g of methanol (10 mL), and 1.0 g of catalyst were added to the 250-mL three-necked round-bottomed flask. The transesterification reaction was done at 338 K after 5-240 min. 2.3. Product Analysis. The amounts of methyl ester in the product were analyzed by high-performance liquid chromatography (HPLC) using a Shimadzu liquid chromatograph, consisting of two LCT-10AT HPLC pumps and a UV-VIS detector. The LC column was a Shim-pack VP-ODS polymer-based column, 150 mm × 4.6 mm with 5-µm diameter particle size silica. The mobile phase was methanol (HPLC-grade) with a flow rate of 1 mL/min (1 mL/min) at a temperature of 303 K and a loop of 10 µL.
3. Results and Discussion 3.1. Comparing the Yield of FAME Using Catalyst H2SO4/C and H2SO4 under Microwave Radiation. To study the effect of different catalysts on the yield of FAME under microwave radiation, the experiments were carried out by using
550 Energy & Fuels, Vol. 23, 2009
Figure 2. Comparison of the yield of FAME using catalyst H2SO4/C and H2SO4 under microwave radiation. Reaction time: 240 min; methanol/castor oil molar ratio: 12/1; catalyst/castor oil mass ratio: 5 wt %; H2SO4/C: 55 wt %.
Figure 3. Comparison of the yield of FAME using microwave radiation and conventional heating. Reaction time: 240 min; methanol/castor oil molar ratio: 12/1; catalyst amount: 5 wt %; H2SO4/C: 55 wt %.
homogeneous acid catalyst and heterogeneous acid catalyst, respectively. The homogeneous acid catalyst was only the H2SO4 liquid with the concentration of 98 wt %, without activated carbon in this case. The heterogeneous acid catalyst was H2SO4/ C. The amount of H2SO4 was equivalent in both cases. As shown in Figure 2, it was observed that a 90 wt % of yield could be obtained using heterogeneous catalyst H2SO4/C and homogeneous catalyst H2SO4 when the reaction finally reached equilibrium after 60 and 120 min, respectively. This fact indicates that the transesterification can be reached in equilibrium in a short time under the microwave radiation when H2SO4/C catalyst was used. For the case of H2SO4/C being used as catalyst, the reaction occurred on the internal and external surface of the catalyst. Microwave radiation has a selective and instantaneous heating function, so the “microwave hot spots” could be formed directly on the strong microwave absorbing catalyst H2SO4/C. These hot spot temperatures are far higher than the temperature of liquid, which then allowed the higher reaction rate and conversion to be obtained. Therefore, the decline in activity caused by the mass transfer resistance was offset by high reaction temperature from “microwave hot spots” in the case of heterogeneous acid catalyst being used. 3.2. Comparing the Yield of FAME Using Microwave Radiation and Conventional Heating. To study the influence of microwave radiation and conventional heating on the yield of FAME, experiments were carried out by using a H2SO4/C catalyst. As shown in Figure 3, the reaction could reach equilibrium after 60 min under microwave radiation, and the yield of FAME was 94 wt %. However, the reaction did not reach equilibrium until 180 minutes using conventional heating, and the yield of FAME was 70 wt %. This indicated that a
Yuan et al.
Figure 4. Comparison of the effect of methanol and ethanol on the transesterification. Reaction time: 240 min; reaction temperature: 338 K (methanol), 351 K (ethanol); methanol (ethanol)/castor oil molar ratio: 12/1; catalyst amount: 5 wt %; H2SO4/C: 55 wt %.
shorter time was needed and higher yield of FAME could be obtained under microwave radiation compared to conventional heating. The result shows that both the effect of “microwave hot spots” and the microwave absorbing character of feed are beneficial to the reaction. Methanol is a good microwave radiation absorption material. Its dipole quickly reorientates under the microwave radiation, which could destroy the two-tier structure of the interface of methanol and oil.18,19 Therefore, the solubility of methanol and oil was improved under microwave radiation, to the advantage of the reaction. 3.3. Comparing the Effect of Methanol and Ethanol on the Transesterification. To study the effect of feed on the transesterification, experiments were carried out by using methanol and ethanol under microwave radiation at 338 and 351 K, respectively. It may be seen from Figure 4 that, at less than 120 min of reaction time, the yield of FAME of the methanol system was higher than the yield of FAEE of the ethanol system. The methanol system achieved equilibrium after 60 min; however, the ethanol system needed 120 min. At equilibrium, the yield of the two systems was about 90 wt %. Despite methanol and castor oil’s immiscibility as well as lower working temperature, however, the dielectric loss angle tangent tgδ of methanol is 0.64, which is higher than 0.25 of ethanol, implying that methanol has stronger absorption ability of microwave radiation than ethanol,19 and thus the kinetics is more favorable compared to that of EtOH under the condition of microwave radiation. Hence, methanol molecules reorientate more rapidly than ethanol under the microwave radiation, which could destroy the two-tier structure of the interface of methanol and oil. Therefore, methanol is a preferable feed compared to ethanol in this experimental system. Meneghetti et al. produced biodiesel by using castor oil with methanol or ethanol, in the presence of catalysts (KOH, NaOH, KOCH3, NaOCH3, H2SO4, or HCl), under the boiling point of the respective methanol and ethanol. After 6-8 h, the highest yield of biodiesel reached about 85%.20 3.4. Effect of Reaction Temperature. To study the influence of the reaction temperature on the yield of FAME under microwave radiation, experiments using H2SO4/C catalyst were conducted respectively at 328, 333, 338, and 343 K. As shown (18) Lidstro¨m, P.; Tierney, J.; Wathey, B.; Westman, J. Tetrahedron 2001, 57, 9225–9283. (19) Jin, Q. H.; Dai, S.S.; Huang, K. M. MicrowaVe Chemistry; Science Press: Beijing, 1999; p 171. (20) Meneghetti, S. M. P.; Meneghetti, M. R.; Wolf, C. R.; Silva, E. C.; Lima, G. E. S.; de Lira Silva, L.; Serra, T. M.; Cauduro, F.; de Oliveira, L. G. Energy Fuels 2006, 20, 2262–2265.
Synthesis of Biodiesel Using MicrowaVe Catalysts
Figure 5. Influence of the reaction temperature on the yield of FAME. Reaction time: 240 min; methanol/castor oil molar ratio: 12/1; catalyst amount: 5 wt %; H2SO4/C: 55 wt %.
Figure 6. Influence of the catalyst amount on the yield of FAME. Reaction temperature: 338 K; methanol/castor oil molar ratio: 12/1; reaction time: 60 min; H2SO4/C: 55 wt %.
in Figure 5, the reaction reached equilibrium after about 60 min at 338 and 343 K, and the yield of FAME was approximately 94 wt %. The reaction reached equilibrium after about 120 min at 333 K, and the yield of FAME was approximately 90 wt %, closer to the yield of 338 K. The higher temperature usually can lead to a higher yield; however, in this experiment, when the temperature reaches or exceeds the boiling point of methanol (341 K), the evaporation rate of methanol will increase, and the amount of retention in condenser will increase. The decreasing of methanol in liquid phase will affect the molar ratio of methanol/castor oil, and thus the yield was slightly deducted. Furthermore, since the reaction occurred on the catalyst and the temperature of catalyst was determined by both “microwave hot spots” and reaction medium, the yield of FAME depends on both the temperature of reaction medium and “microwave hot spots”. However, the temperatures of “microwave hot spots” were much higher than that of the liquid; therefore, when the temperature of liquid reactant changed from 333 to 338 K, the yield of FAME did not show an obvious variation. Comprehensive consideration, 338 K thus is chosen as the optimum temperature for the synthesis of FAME in this experimental system. 3.5. Effect of Catalyst 55 wt % H2SO4/C Loading. The experiments were performed with different amounts of catalyst 55 wt % H2SO4/C employed to investigate the influence of catalyst amount on the yield of FAME. The results are shown in Figure 6. From this figure, it may be seen that when the catalyst amount is no more than 3 wt %, an increase from 1 to 3 wt % caused a marked increase of yield of FAME from 36 to 82 wt %, respectively. Further increasing the catalyst amount from 3 to 5 wt %, we found that the yield of FAME increased slowly
Energy & Fuels, Vol. 23, 2009 551
Figure 7. Influence of H2SO4 loading amounts of the catalysts on the yield of FAME. Reaction temperature: 338 K; methanol/castor oil molar ratio: 12/1; reaction time: 60 min; catalyst amount: 5 wt %.
from 82 to 94 wt %. When the catalyst amount increased from 5 to 7 wt %, a negligible increase of FAME from 94 to 95 wt % was evident. Therefore, the 5 wt % is considered as a suitable catalyst amount in this experimental system. It is may be explained that the reaction takes place on the surface of the catalyst and when the amount of catalyst was small, the activity sites and “microwave hot spots” of catalyst were correspondingly small, leading to the lower yield of FAME. Increasing the amount of catalyst significantly increased the yield of FAME. Further increasing the catalyst loading could not obviously affect the yield of FAME when the amount of catalyst was at a high level; the reason could be considered as the equilibrium limit.21 3.6. Effect of H2SO4/C Loading Amount of the Catalysts. The H2SO4 loading amount that occurred during catalyst preparation has an effect on the yield of FAME. To illustrate this point, the transesterification experiments were carried out at 338 K by using the H2SO4/C catalyst loaded with different amounts of H2SO4, such as 40, 45, 50, 55, and 60 wt %, respectively. The results are presented in Figure 7. From this figure, it can be seen that the yield of FAME markedly increased from 56 to 94 wt % with the increase in the amount of H2SO4 in the H2SO4/C from 45 to 55 wt % during catalyst preparation. However, the yield of FAME only increased from 94 to 95 wt % when the H2SO4 amount in H2SO4/C further increased from 55 to 60 wt %. Therefore, the H2SO4/C catalyst loaded with 55 wt % H2SO4 is considered suitable in this experimental system. It may be explained that sulfuric acid is the active component of the catalyst, and with the amount of sulfuric acid increasing, the yield of FAME should also increase. However, from BET surface area measurement results, the surface areas of active carbon and 55 wt % H2SO4/C were found to be 566 and 301 m2/g, respectively. This shows that the sulfuric acid loaded on activated carbon would lead to the internal pore of active carbon being seriously blocked, reducing the surface area of catalyst. 3.7. Effect of Molar Ratio of Methanol/Castor Oil. The effect of the molar ratio of methanol/castor oil on the yield of FAME was examined at 338 K with 5 wt % of catalyst loading. The molar ratios were set at 3, 6, 12, 15, and 18 respectively, and the results are shown in Figure 8. It may be seen that an increase of molar ratio from 3 to 12 caused a marked increase in yield of FAME from 55 to 94 wt %, and a further increase in the molar ratio from 12 to 18 caused a decrease in yield of FAME from 94 to 61 wt %. Thus, the molar ratio of 12 can be selected as a suitable molar ratio for the synthesis of FAME. (21) Liu, X. J.; Piao, X. L.; Wang, Y. J.; Zhu, S. L.; He, H. Y. Fuel 2008, 87, 1076–1082.
552 Energy & Fuels, Vol. 23, 2009
Yuan et al.
Figure 8. Influence of the molar ratio of methanol/castor oil on the yield of FAME. Reaction temperature: 338 K; catalyst amount: 5 wt %; H2SO4/C: 55 wt %; reaction time: 60 min.
Figure 9. Influence of the FFA amount on the yield of FAME. Reaction temperature: 338 K; catalyst amount: 5 wt %; H2SO4/C: 55 wt %; reaction time: 180 min.
By stoichiometric equations, 3 mol of methanol is required for each mole of triglyceride to synthesize 3 mol of FAME. Considering that the transesterification reaction is reversible, a higher molar ratio is needed for the reaction to drive the reaction toward positive direction. However, the excessive high concentration of methanol will lead to FAME and glycerol being miscible,22 and thus the recombination of FAME and glycerol to form monoglycerides is also enhanced. Furthermore, methanol is a strong microwave absorbing medium, and the excessive methanol might absorb a large portion of microwave energy, reducing the microwave field strength and power rapidly. Consequently, a little microwave radiation can reach the activated carbon catalyst, and the “microwave hot spot” thus would be reduced. Therefore, the molar ratio of methanol to triglyceride should not be excessive, and a molar ratio of 12 was considered suitable in this experimental system. 3.8. Effect of FFA Level. The effect of the FFA amount of castor oil on the yield of FAME was examined at 338 K with a 5 wt % of catalyst H2SO4/C loading. The FFA amount of castor oil was set at 1, 5, and 10 wt %, respectively, by addition of oleic acid at the acid value already present at the castor oil, and the transesterification yields are shown in Figure 9. It may be seen that the yield of FAME reached 90 mol % after 120 min when the FFA amount was 5 wt %, 60 min longer than the 1 wt % system. When the FFA content was 10 wt %, the yield of FAME increased to 82 wt % after 180 min. Thus, with an
increase in FFA amount, the yield of product was reduced. However, a high yield could be obtained using the catalyst H2SO4/C under microwave radiation for 120-180 min.
(22) Ma, F.; Hanna, M. A. Bioresour. Technol. 1999, 70, 1–15.
4. Conclusions Transesterification experiments of castor oil with methanol using solid acid catalyst (H2SO4/C), enhanced by means of microwave radiation, were conducted. The microwave radiation was proved to be an efficient method for the heterogeneous reaction system compared with conventional heating. The microwave absorption heterogeneous catalyst H2SO4/C provides equivalent yield of FAME compared with that of homogeneous catalyst H2SO4. A maximum yield of 94% was obtained using 12:1 molar ratio of methanol to castor oil, 5 wt % mass ratio of catalyst to castor oil, and 55 wt % H2SO4 loading amounts of catalyst at 338 K under microwave radiation after 60 min. The yield of FAME could increase to 82 wt % after 180 min using castor oil containing 5 and 10 wt % FFA under microwave radiation. Acknowledgment. Financial support for this work from the National Basic Research Program of China (973 Program, 2009CB219906), National Natural Science Foundation of China (20776117), and Specialized Research Fund for the Doctoral Program of Higher Education of China (20070698037) is gratefully acknowledged. EF800577J