Energy & Fuels 2006, 20, 409-414
409
Electrochemical Reduction of CO2 to Methane at the Cu Electrode in Methanol with Sodium Supporting Salts and Its Comparison with Other Alkaline Salts Satoshi Kaneco,*,† Hideyuki Katsumata,† Tohru Suzuki,‡ and Kiyohisa Ohta† Department of Chemistry for Materials, Faculty of Engineering, Mie UniVersity, Tsu, Mie 514-8507, Japan, and EnVironmental PreserVation Center, Mie UniVersity, Tsu, Mie 514-8507, Japan ReceiVed August 23, 2005. ReVised Manuscript ReceiVed October 30, 2005
The electrochemical reduction of CO2 at the Cu electrode was investigated in methanol-based electrolyte using various sodium supporting salts, such as acetate, chloride, bromide, iodide, thiocyanate, and perchlorate, at a low temperature (243 K). The main products from CO2 by the electrochemical reduction were methane, ethylene, carbon monoxide, and formic acid. The formation of methane from CO2 predominated in all sodium supporting salts, and the efficiencies were relatively high. A maximum faradaic efficiency of methane was 70.5% in NaClO4/methanol-based electrolyte at -3.0 V versus Ag/AgCl-saturated KCl. In the sodium salts/ methanol-based electrolyte system, except for the case of acetate, the efficiency of hydrogen formation, being a competitive reaction against CO2 reduction, was depressed below 18%. Especially, the efficiency was below 1% when sodium thiocyanate salt was used as a supporting salt. On the basis of this work, the high-efficiency electrochemical CO2-methane conversion method is achieved.
Introduction From the standpoint of environmental protection and resource utilization, the development of a truly environmentally benign process utilizing carbon dioxide (CO2), which is the largest single source of greenhouse gas, has drawn current interest in industrial chemistry and biotechnology. The electrochemical method appears to become one of very promising methods for the conversion and reduction of CO2.1-3 A great number of studies dealing with electrochemical fixation of CO2 are of both fundamental and preparative interest.1-3 Recently, many investigators have actively investigated the electrochemical reduction of CO2 using various metallic electrodes in organic solvents, because organic aprotic solvents dissolve much more CO2 than water.4-7 It has been described that low reduced products containing carbon monoxide, oxalic acid, and formic acid were produced by the electroreduction of CO2 in dimethyl sulfoxide, N,N-dimethyl formamide, propylene carbonate, and acetonitrile. However, even at a copper electrode, * To whom correspondence should be addressed. Telephone: +81-59231-9427. Fax: +81-59-231-9442, 9471, or 9427. E-mail: kaneco@ chem.mie-u.ac.jp. † Department of Chemistry for Materials, Faculty of Engineering, Mie University. ‡ Environmental Preservation Center, Mie University. (1) Halmann, M. M.; Steinberg, M. Greenhouse Gas Carbon Dioxide Mitigation, Science, and Technology, Lewis Publishers: Boca Raton, FL, 1999. (2) Scibioh, M. A.; Viswanathan, B. Photo-/photoelectro-catalytic pathways for carbon dioxide reduction. In Photo/Electrochemistry and Photobiology in the EnVironment, Energy, and Fuel; Kaneco, S., Ed.; Research Signpost: Kerala, India, 2002; pp 1-46. (3) Dey, G. R.; Kishore, K. Carbon dioxide reduction: A brief review. In Photo/Electrochemistry and Photobiology in the EnVironment, Energy, and Fuel; Kaneco, S., Ed.; Research Signpost: Kerala, India, 2005; pp 357388. (4) Amatore, C.; Save´ant, J. M J. Am. Chem. Soc. 1981, 103, 5021. (5) Ito, K.; Ikeda, S.; Iida, T.; Nomura, A. Denki Kagaku 1982, 50, 463. (6) Ito, K.; Ikeda, S.; Yamauchi, N.; Iida, T.; Takagi, T. Bull. Chem. Soc. Jpn. 1985, 58, 3027. (7) Ikeda, S.; Takagi, T.; Ito, K. Bull. Chem. Soc. Jpn. 1987, 60, 2517.
few hydrocarbons such as methane and ethylene have been obtained in these organic solvents. Methanol is a better solvent of CO2 than water, particularly at a low temperature. Literature data8-10 for the solubilities of CO2 in pure methanol and water, at 288 K, were of 4.6 and 1.07 cm3 cm-3, respectively. Moreover, it has been reported in the previous study26 that the solubility of CO2 in a solution of 80 mmol dm-3 NaOH/methanol, at 243 K, was approximately 15 cm3 cm-3 (about 670 µmol of CO2/cm3 of methanol). Therefore, the solubility of CO2 in methanol is about 5 times that in water, at ambient temperature, and more than 14 times that in water, at temperatures of 243 K. Hence, methanol has been industrially used as a physical absorbent of CO2 in the Rectisol method, at 243-263 K.10 Currently, over 70 large-scale plants apply the Rectisol process all over the world. Therefore, the direct electrochemical reduction of CO2 in methanol is an advantageous choice, especially when the process is performed under energetically efficient conditions. Recently, many research groups (our group,11-26 Fijishima et al.,27 Koleli et al.,28 Schrebler et al.,29 Ortiz et al.,30 and Eggins et al.31) have brought (8) Lide, D. R., Ed.; Handbook of Chemistry and Physics, 72nd ed.; CRC Press: Boca Raton, FL, 1991; pp 6-4, 3-321. (9) Chihara, H., Ed.; Kagaku binran-kiso (Handbook of Basic Chemistry, in Japanese), 3rd ed.; Maruzen: Tokyo, Japan, 1984; Vol. 2, pp 158, 165. (10) Hochgesand, G. Ind. Eng. Chem. 1970, 62, 37. (11) Kaneco, S.; Katsumata, H.; Suzuki, T.; Ohta, K. Electrochemical reduction of CO2 on Cu electrode in methanol at low temperature. In Utilization of Greenhouse Gases; Liu, C., Mallinson, R. G., Aresta, M., Eds.; ACS Symposium Series 852; American Chemical Society: Washington, DC, 2003; pp 169-182. (12) Naitoh, A.; Ohta, K.; Mizuno, T.; Yoshida, H.; Sakai, M.; Noda, H. Electrochim. Acta 1993, 38, 2177. (13) Mizuno, T.; Ohta, K.; Kawamoto, M. Energy Sources 1997, 19, 249. (14) Kaneco, S.; Iiba, K.; Ohta, K.; Mizuno, T.; Saji, A. Electrochim. Acta 1998, 44, 573. (15) Kaneco, S.; Iiba, K.; Ohta, K.; Mizuno, T.; Saji, A. J. Electroanal. Chem. 1998, 441, 221. (16) Kaneco, S.; Iiba, K.; Ohta, K.; Mizuno, T. Energy Sources 1999, 21, 643.
10.1021/ef050274d CCC: $33.50 © 2006 American Chemical Society Published on Web 12/02/2005
410 Energy & Fuels, Vol. 20, No. 1, 2006
focus into the electrochemical reduction of CO2 in methanolbased electrolyte. In our groups, the electrochemical reduction of CO2 with copper electrode in the methanol-based electrolyte has been energetically investigated at a low temperature.11-13,16-21,24-26 From these studies, it was found that the methanol at a low temperature was the best electrolyte to obtain hydrocarbons, compared to various organic solvents. A variety of supporting salts, such as lithium, potassium, and cesium salts, were tested for the electrochemical CO2 reduction at the Cu electrode in methanol. The lithium supporting salts were especially suitable for methane formation (for example, 71.8% for LiClO4 supporting salts).21 On the other hand, when KOH was evaluated as a supporting salt, the maximum ethylene current efficiency was 37.5%.18,20,25 However, little information on the electrochemical reduction of CO2 in methanol using sodium supporting salts has been reported.26 In this study, the electrochemical reduction of CO2 at a copper electrode in methanol with various sodium supporting salts has been investigated. Moreover, the effect of the anionic and cationic species on the reduction reaction on the electrode surface has been discussed. Future work to advance this technology may include the use of solar energy as the electric energy source. This research can contribute to large-scale manufacturing of useful organic products from readily available and cheap raw materials: CO2-saturated methanol from industrial absorbers (the Rectisol process). Experimental Procedures The apparatus and experimental conditions for the electrochemical reduction of CO2 are shown in Table 1. Electrochemical reduction of CO2 was performed in a custom-made, divided H-type cell. An Aldrich Nafion 117-type ion-exchange membrane (0.18 mm thickness) was used as the diaphragm. The cathode potential was measured with respect to a Ag/AgCl-saturated KCl electrode that was connected with the catholyte through an agar salt bridge. The experimental procedures are shown in Figure 1. (17) Kaneco, S.; Iiba, K.; Suzuki, S.; Ohta, K.; Mizuno, T. J. Phys. Chem. B 1999, 103, 7456. (18) Kaneco, S.; Iiba, K.; Hiei, H.; Ohta, K.; Mizuno, T.; Suzuki, T. Electrochim. Acta 1999, 44, 4701. (19) Kaneco, S.; Iiba, K.; Ohta, K.; Mizuno, T. J. Solid State Electrochem. 1999, 3, 424. (20) Kaneco, S.; Iiba, K.; Ohta, K.; Mizuno, T. Energy Sources 2000, 22, 127. (21) Kaneco, S.; Iiba, K.; Yabuuchi, M.; Nisho, N.; Ohnishi, H.; Katsumata, H.; Suzuki, T.; Ohta, K. Ind. Eng. Chem. Res. 2002, 41, 5165. (22) Kaneco, S.; Iwao, R.; Katsumata, H.; Suzuki, T.; Ohta, K. Photo/ Electrochem. Photobiol. EnViron. Energy Fuel 2002, 1, 69; SciFinder 2004: 281054. (23) Kaneco, S.; Iwao, R.; Katsumata, H.; Suzuki, T.; Ohta, K. Photo/ Electrochem. Photobiol. EnViron. Energy Fuel 2003, 2, 181; SciFinder 2004: 276792. (24) Kaneco, S.; Yamauchi, H.; Katsumata, H.; Suzuki, T.; Ohta, K. Electrochemical reduction of CO2 at alloy electrode in methanol. In Carbon Dioxide Utilization for Global Sustainability. Studies in Surface Science and Catalysis; Park, S. E., Chang, J. S., Lee, K. W., Eds.; Elsevier Science: Amsterdam, The Netherlands, 2004; pp 277-282. (25) Kaneco, S.; Katsumata, H.; Suzuki, T.; Ohta, K. Electrochim. Acta, in press. (26) Kaneco, S.; Katsumata, H.; Suzuki, T.; Ohta, K. Electrochim. Acta, manuscript submitted. (27) Saeki, T.; Hashimoto, K,; Fujishima, A.; Kimura, N.; Omata, K. J. Phys. Chem. 1995, 99, 8440. (28) Koleli, F.; Ropke, T.; Hamann, C. H. Electrochim. Acta 2003, 48, 1595. (29) Schrebler, R.; Cury, P.; Suarez, C.; Munoz, E.; Gomez, H.; Cordova, R. J. Electroanal. Chem. 2002, 533, 167. (30) Ortiz, R.; Marquez, O. P.; Marquez, J.; Gutierrez, C. J. Electroanal. Chem. 1995, 390, 99. (31) Eggins, B. R.; Ennis, C.; McConnell, R.; Spence, M. J. Appl. Electrochem. 1997, 27, 706.
Kaneco et al. Table 1. Apparatus and Experimental Conditions electrochemical reduction cell potentiostat/galvanostate coulometer potential sweep XY recorder working electrode counter electrode reference electrode electolyte catholyte anolyte carbon dioxide potential temperature product analysis gas products
liquid products
H-type cell Hokuto HA-3001A Integrator 1109 (Fusou Seisakujyo, Inc., Japan) Hokuto HB-111 function generator Graphtec WX1100 Cu foil (30 × 20 mm, 0.1 mm thickness, 99.98% purity) Pt foil (30 × 20 mm, 0.1 mm thickness, 99.98% purity) Ag/AgCl-saturated KCl (Horiba, 2060A-10T) 250 mmol dm-3 CH3COONa, NaCl, NaBr, NaI, NaSCN, NaClO4 in methanol 300 mmol dm-3 KOH in methanol 99.9999% purity -3.0 versus Ag/AgCl-saturated KCl 243 ( 0.5 K gas chromatography TCD (GL Sciences GC-320, Molecular Sieve 5A; 13X-S, Ar and He carrier gas) FID (GL Sciences GC-353B, Porapak Q, N2 carrier gas) HPLC with UV detector (Hitachi L4000) TCD and FID gas chromatography
The methanol (99%, Nacalai Tesque, Inc., Japan) was purified by double distillation from metallic magnesium. Water content in the pure methanol was less than 0.1% (confirmed by the Karl Fisher test). Sodium salts such as acetate, chloride, bromide, iodide, thiocyanate, and perchlorate (Nacalai Tesque, Inc.) were used as the supporting salt in the methanol-based catholyte. The pH of the catholyte was evaluated with a glass electrode for nonaqueous solvent (Horiba, 6377-10D), calibrated in water. Mechanical processing of the Cu and Pt electrodes required polishing each surface with successively finer grades of alumina powder (Baikalox emulsion, manufactured by Baikowski International Co.) down to 0.05 µm, followed by the removal of grease with acetone. The electrodes were activated by anodic polarization at 500 mA for 100 s in 14.7 mol dm-3 phosphoric acid. Finally, both electrodes were rinsed with water and ethanol. A discontinuous electroreduction procedure was used. First, CO2 gas was bubbled into the methanol catholyte for 1 h at a rate of 30 mL min-1. Then, the CO2-saturated solution was reduced electrolytically at cathodic polarization of -3.0 V versus Ag/AgClsaturated KCl. The cell voltage (the potential between the cathode and anode) was in the range of ∼30-60 V. A cooling device was used to maintain a low temperature. Stirring of the catholyte was provided by a magnetic bar. The faradic efficiencies of formation for the main products were calculated from the total charge passed
Figure 1. Experimental procedures for the electrochemical reduction of CO2.
Electrochemical Reduction of CO2 to Methane
Energy & Fuels, Vol. 20, No. 1, 2006 411
Figure 2. Typical cyclic voltammogram on the Cu electrode in CO2saturated methanol at 243 K. Catholyte ) 250 mmol dm-3 NaClO4/ methanol. Anolyte ) 300 mmol dm-3 KOH/methanol.
during batch electrolyses, which was set to 50 C. The reactions of the main products for the calculation of faradaic efficiency were as follows: CO2 + H+ + 2e- f HCOO-
(1)
CO2 + 2H+ + 2e- f CO + H2O
(2)
CO2 + 8H+ + 8e- f CH4 + 2H2O
(3)
2CO2 + 12H+ + 12e- f C2H4 + 4H2O
(4)
2CO2 + 14H+ + 14e- f C2H6 + 4H2O
(5)
The gaseous products obtained during electroreduction were collected in a gas collector and were analyzed by gas chromatography. Products soluble in the catholyte were analyzed by using highperformance liquid chromatography and gas chromatography.
Results and Discussion The suitable sodium supporting salt was explored for the electrochemical reduction of CO2 in methanol. Therefore, the solubility in methanol at 243 K was evaluated for various sodium salts such as NaNO3, NaH2PO4, NaHCO3, NaCl, NaBr, NaI, NaF, Na2SO4, NaSCN, NaClO4, and CH3COONa. Consequently, these compounds could be classified into three groups according to solubility: (1) insoluble, NaH2PO4, NaHCO3, NaF, and Na2SO4; (2) slightly soluble, NaNO3, NaCl, and CH3COONa; and (3) soluble, NaBr, NaI, NaSCN, and NaClO4. Sodium nitrate could easily decompose in the electrolysis. Consequently, sodium salts such as acetate, chloride, bromide, iodide, thiocyanate, and perchlorate were selected as the supporting salts from the viewpoint of the solubility into methanol. The electrolysis was performed at 243 K because the temperature in the Rectisol process was mainly in the range of 243-263 K.10 Onset Potential. The effect of the anionic species on the electrochemical reduction of CO2 on Cu in Na+/methanol electrolyte was evaluated using the onset potentials of the cathodic current. The onset potentials were obtained by cyclic voltammetric measurements at the scanning rate of 50 mV s-1 in CO2-saturated methanol. The typical cyclic voltammogram is illustrated in Figure 2. In any supporting salts, no voltammetric peak was obtained in the potential range studied. Table 2 shows the onset (appearance) potentials for the cathodic current, i.e., those potential values at which a current density of 0.1 mA cm-2 is observed. Also, some onset data obtained in the cyclic voltammetry in CO2-saturated methanol with lithium, potassium, rubidium, and cesium supporting salts, which were previously
published,17-21,25,26 are summarized to discuss the effect of the supporting salts on the onset potentials. In the electrochemical reduction of CO2, the onset potentials appear to be roughly associated with the magnitude of bias voltage required for the operation of the system. Ortiz et al.30 has presented that the onset potential recorded with the Cu electrode in NaClO4/ methanol electrolyte at room temperature was -0.6 V versus Ag/Ag+. The difference of their onset potentials from the present works may be attributed to the experimental conditions, such as the temperature, the cathode shape, and the concentration of supporting salt. One characteristic aspect of the onset potentials in methanol is that the potential with sodium supporting salts seems to be relatively more negative compared to other alkaline salts. Roughly, the onset potentials for CO2 reduction in methanol with halide and thiocyanate supporting salts tended to increase with an increasing size of the cation, a very similar trend to that obtained in the aqueous solutions.32 The onset potentials may be affected by a number of parameters including the pH of catholyte, the kind of cation and anion for supporting salts, and their concentrations. These effects are still being examined. Once the onset potentials were determined from polarization experiments, we attempted to investigate the electrochemical reduction of CO2 in CO2-saturated methanol at cathodic polarizations exceeding the onset potential. Therefore, the potential was set to -3.0 V versus Ag/AgCl-saturated KCl because it was an effective potential for the formation of hydrocarbons in the previous work.21 Faradic Efficiency of the Products. Table 3 shows current efficiencies of the products by electrochemical reduction of CO2 in methanol-based catholyte using various sodium supporting salts at 243 K. The reproducibility of the electrolysis [relative standard deviation (RSD), for faradic efficiency of the reduction products] was better than RSD 10% for five repeated measurements. Also, some experimental data obtained in the electrochemical CO2 reduction in methanol-based electrolyte with lithium, potassium, rubidium, and cesium supporting salts and in water, which were previously published,17-21,25,26 are summarized to discuss the effect of the anionic and cationic species on the reduction reaction. Methane, ethylene, CO, and formic acid are the principal products identified by the analytical system used in this study. Oxygen was detected from the anodic compartment. Methane formation tends to increase in the order Cs+, Rb+, K+, Na+, and Li+, identical to the decreasing order of the cation size. A high faradic efficiency of methane was observed in methanol with LiClO4 and NaClO4 supporting salts. In NaClO4/methanol-based electrolyte, the methane efficiency was 70.5%. In this case, the conversion rate of CO2 was 26 µmol min-1. In all sodium supporting salts tested, the methane current efficiency was relatively high (g43.4%). In contrast, potassium, rubidium, and cesium supporting salts were generally preferred for the formation of ethylene. The best ethylene faradic efficiency was 35.7% in the methanol-based catholyte with potassium hydroxide supporting salt. The reduction system containing hydroxide salts seems to be more efficient for ethylene formation. In halide/methanol-based catholyte, there was a tendency that the current efficiency for CO increased with an increasing cation size. The same trends were obtained in thiocyanate/methanol-based electrolyte. In Na+/methanol electrolyte, the maximum CO formation (19.3%) was observed with chloride supporting salt. The faradic efficiencies for formic acid in the electrolyte with sodium salts were e8% and smaller than those obtained in methanol containing potassium, rubidium, and (32) Eggins, B. R.; Bennett, E. M.; McMullan, E. A. J. Electroanal. Chem. 1996, 408, 165.
412 Energy & Fuels, Vol. 20, No. 1, 2006
Kaneco et al.
Table 2. Comparison of Onset Potentials in the Current-Potential Curves at the Cu Electrode in Methanol-Based Catholyte with Various Supporting Salts onset potential (/V versus Ag/AgCl-saturated KCl) cation of supporting salts
hydroxide
chloride
bromide
iodide
Lia Nab Kc Rbe Csf
-0.5 -1.5 -1.4 -1.3 -1.5
-1.3 -1.5
-1.5 -1.8 -1.4
-1.5 -1.8 -1.4, -1.5d
-1.5 -1.5, -1.6d
-1.3
-1.3
-1.1
a
b
-1.3
c
thiocyanate
d
perchlorate -1.4 -1.8
acetate -1.0 -0.8 -1.3 -1.3, -1.4d
e
f
References 17 and 21. This paper. References 19 and 25. Different concentration of supporting salt. Reference 25. References 18 and 20.
Table 3. Comparison of Faradic Efficiencies of the Products by Electrochemical Reduction of CO2 at the Cu Electrode in Methanol-Based Catholyte with Various Supporting Saltsa salts (mmol dm-3) methanol CH3COOLi (500) CH3COONa (250) CH3COOK (500)d CH3COOCs (500)d LiCl (500) NaCl (250) CsCl (30)d LiBr (500) NaBr (250) KBr (50)d CsBr (25) LiI (500) NaI (250) KI (50)d CsI (30)d NaSCN (250) KSCN (500) CsSCN (30)d LiClO4 (500) NaClO4 (250) LiOH (80)f NaOH (80)f KOH (80)f RbOH (80)g CsOH (80)h water LiHCO3 (500) NaHCO3 (500) KHCO3 (500) CsHCO3 (500)
faradaic efficiency (%) HCOOH H2
current density (mA cm-2)
CH4
C2H4
CO
9.7 7.6 12.9 5.5 16.4 5.6 1.7 17.0 21.7 3.2 1.5 19.3 16.7 4.2 1.3 8.9 15.3 1.3 27.0 22.7 10.1 7.3 8.9 14.4 13.4
56.2 43.4 27.0 12.3 68.4 54.7 2.8 68.1 53.1 6.4 0.1 65.5 63.1 15.8 1.0 49.8 16.7 2.4 71.8 70.5 63.0 63.0 16.0 4.6 4.1
4.4 5.1 7.1 3.3 4.0 8.6 8.8 6.8 8.0 9.8 5.9 9.0 5.2 19.9 9.9 7.9 16.5 6.0 1.6 3.1 14.7 17.6 37.5 31.0 32.3
12.0 1.1 6.5 3.1 15.7 19.3 35.6 19.2 5.4 35.0 54.1 14.9 2.0 15.2 40.0 2.7 11.1 36.1 14.3 3.2 14.2 17.3 13.6 6.1 9.0
7.7 4.6 13.0 15.6 6.8 7.4 14.6 6.3 6.7 17.4 13.0 e 7.2 e 4.8 15.2 e 16.3 3.7 4.2 5.2 4.6 7.9 10.0 13.1
26 19 16 15
4 11 14 13
2 3 4 5
3 4 5 6
i i i i
total
HCb
ratioc
reference
17.9 34.0 32.5 30.1 5.8 2.4 8.4 2.8 12.6 3.6 7.0 0.95 9.1 3.5 6.2 1.0 8.1 7.2 12.0 17.9 1.3 0.90 3.3 9.4 18.4
98.2 88.2 86.1 64.1 100.7 92.4 70.2 103.2 85.9 72.2 80.1 90.4 86.6 54.6 61.9 77.4 52.7 68.0 103.4 99.0 98.4 103.4 78.3 61.1 76.9
60.6 48.5 34.1 15.6 72.4 63.3 11.6 74.9 61.1 16.2 6.0 74.5 68.3 35.7 10.9 57.7 33.2 8.4 73.4 73.6 77.7 80.6 53.5 35.6 36.4
12 8.5 3.8 3.7 17 6.4 0.32 10 6.6 0.65 0.017 7.3 12 0.79 0.10 6.3 1.0 0.40 44 22 4.3 3.6 0.43 0.15 0.13
21
68 62 59 56
103 99 98 95
30 30 30 28
6.5 1.7 1.1 1.2
31 31 31 31
19 20 21 20 21 19 20 21 19 20 19 20 21 17 26 25 25 18
a Temperature of 243 K and potential of -3.0 V versus Ag/AgCl-saturated KCl. b Total faradaic efficiency for hydrocarbons (methane and ethylene). Ratio of faradaic efficiency of methane and ethylene, rf(CH4)/rf(C2H4). d Trace ethane (ca. 0.1%) was obtained. e Not measured. f Potential of -4.0 V. g Potential of -4.5 V. h Potential of -3.5 V. i Not measured. The electrolysis was performed at -1.8 V versus SCE and 298 K. c
cesium supporting salts. In the electrochemical CO2 reduction with the Cu electrode in water,33 the formation efficiencies for methane were 26, 19, 16, and 15% in LiHCO3, NaHCO3, KHCO3, and CsHCO3 aqueous solutions, respectively. As a consequence, the methane faradic efficiencies in the electrochemical CO2 reduction system at a copper electrode in methanol-based electrolyte with sodium supporting salts were much better than those obtained in water. Generally, for the electrochemical reduction of CO2 in water, hydrogen formation competes with the CO2 reduction reaction. Therefore, the depression of hydrogen formation is very important because the applied energy is wasted on hydrogen evolution instead of being used for the reduction of CO2. In a Na+/methanol-based electrolyte, the faradic efficiency for hydrogen formation on the Cu electrode at 243 K was suppressed to e17.9%, except for that in the case of acetate. Especially, the efficiency was below 1% when sodium hydroxide and thiocyanate salts were used as supporting salts. Only in CH3COONa/electrolyte was hydrogen-formation efficiency relatively high (34.0%), and in the other acetates, also, the efficiency was very high. The hydrogen formation roughly
decreased with a decreasing cation size. Therefore, hydrogen evolution may be affected by both anionic and cationic species. In the electrochemical reduction of CO2 on Cu in water,33 the hydrogen current efficiencies were 68, 62, 59, and 56% in LiHCO3, NaHCO3, KHCO3, and CsHCO3 aqueous solutions, respectively. Consequently, the methanol-based electrolyte with sodium supporting salts at a low temperature was suitable for suppression of hydrogen formation in the electrochemical reduction of CO2. Reaction Mechanism. The mechanism of the electrochemical reduction of CO2 with a copper electrode in methanol was studied for various sodium supporting salts. A GC-MS study with deuterated methanol catholyte demonstrated that no reduction product was produced from methanol.12 When the electrolysis was conducted under a nitrogen atmosphere, with any sodium supporting salts, exclusively hydrogen was evolved, with no other product being produced. Eggins et al.31 stated that the reduction products were produced from CO2 in the CO2 (33) Kyriacou, G. Z.; Anagnostopoulos, A. K. J. Appl. Electrochem. 1993, 23, 483.
Electrochemical Reduction of CO2 to Methane
Energy & Fuels, Vol. 20, No. 1, 2006 413
Figure 3. Reaction mechanism of the electrochemical reduction of CO2 at the Cu electrode in methanol.
Figure 4. Schematic cross-sections on the chemical species onto the Cu electrode.
electrochemical reduction in methanol. Consequently, the needed products were not formed from the decomposition of methanol and supporting salts at the cathode and were produced by the electrochemical reduction of CO2, and the reaction path of the reduction products seems to be identical for all supporting salts used. Amatore and Saveant4 investigated the electrochemical reduction of CO2 in media of low proton availability, in which three competing pathways were proposed: (1) oxalate production through self-coupling of ‚CO2-, (2) CO production via oxygencarbon coupling of ‚CO2- with CO2, and (3) formate production through protonation of ‚CO2- by residual or added water followed by an electron transfer occurring in the solution, with the electron source being ‚CO2- itself. The CO production mechanism was thought to contain the coupling of CO2 and ‚CO2- radical anion, where the complex radical anion can be shown as (CO2)2‚-. These species can then undergo either the electrochemical reduction or chemical reduction via ‚CO2-, and the resulting dianion can then decompose into CO and carbonate. The electroreduction of CO2 is, apparently, a very complex process whose mechanism is not yet fully understood.1-3 The present experimental data and literature reports1-7,11-34 suggest that the pathway by which methane, ethylene, carbon monoxide, and formic acid are formed is as follows (Figure 3). Hydrocarbons are yielded by a series of simultaneous or consecutive electronation/protonation steps. The adsorbed ‚CO2radical anion formed in the first electronation step undergoes a second electronation/protonation to yield adsorbed CO as the key intermediate. From a succession of four electronation/ (34) Murata, A.; Hori, Y. Bull. Chem. Soc. Jpn. 1991, 64, 123.
protonation steps, an adsorbed reactive methylene group forms, and this may either stabilize as a methane molecule by a subsequent double electronation/protonation sequence or dimerize to form ethylene and ethane. For the formation of formic acid and CO, we assume the usual pathway, which involves a one-electron reduction followed by a second electronation/ protonation to yield formic acid and by the disproportionation of ‚CO2- radical anions to neutral CO molecules and dinegative carbonate ions. Several researchers33,34 have presented that the methane/ ethylene faradic efficiency ratio increased as the radius of the cation decreased in the electrochemical reduction of CO2 on Cu in water. Hence, a similar tendency may be obtained in the electrochemical reduction of CO2 at a Cu electrode in methanol. The schematic cross-sections on the chemical species onto the copper electrode are shown in Figure 4. Small cations such as Li+ and Na+ were not susceptible to being adsorbed at the electrode surface as compared to large ones, because they have a larger hydration power. In addition, small cations carry a large part of the water molecules in methanol to the cathode and thus supply protons for the electroreduction. The conversion of intermediate CudCH2 to methane require the presence of adsorbed hydrogen. Thus, this reaction may be favored at larger surface hydrogen coverage, as is the case of Na+. In the electrochemical reduction of CO2 in methanol under high pressure,27 the cationic species had a large effect on the reaction, and in contrast, the effect of anionic species was small. However, in the present system, the product selectivity was greatly affected by both cationic and anionic species. The effects of the cationic and anionic species on the electrochemical
414 Energy & Fuels, Vol. 20, No. 1, 2006
reduction of CO2 in methanol-based electrolyte are summarized as follows. (1) The faradic efficiency of methane increases in the order Cs+, Rb+, K+, Na+, and Li+, identical to the decreasing order of the cation size. (2) When Na salts are used, high current efficiency for methane can be obtained, and these phenomena were very similar to those in Li+/methanol-based electrolyte. (3) The hydrogen formation roughly decreases with a decreasing cation size. (4) In the presence of acetate, the formation efficiency of hydrogen is relatively high. Conclusion The electrochemical reduction of CO2 with a Cu electrode in methanol was investigated for various sodium supporting salts. The best current efficiency for methane was 70.5% in NaClO4/methanol-based electrolyte. In all sodium supporting salts tested, the methane formation efficiencies were relatively large (g43.4%). In Na+/methanol-based electrolyte, the faradic efficiency for hydrogen formation on the Cu electrode at 243 K was suppressed to e17.9%, except for that in the case of acetate. Especially, the efficiency was below 1% when sodium thiocyanate salt was used. The selectivity of the electrochemical reduction products of CO2 on Cu in methanol depended
Kaneco et al.
remarkably upon both cationic and anionic species. Because methanol is widely used industrially as a CO2 absorbent at a low temperature in the Rectisol process,10 this research may contribute to applications in the conversion of CO2-saturated methanol into useful products. The present conversion efficiencies of single crystalline, multicrystalline, and amorphous silicon solar cells are at the levels of about 24, 18, and 13%, respectively.35 Hence, one possible system to industrially convert CO2 to useful products may be a hybrid method of the solar cell and the CO2 electrochemical reduction cell. Thus, the synthesis of hydrocarbons by the electrochemical reduction of CO2 might be of practical interest in fuel production, storage of solar energy, and production of intermediate materials for the petrochemical industry. Acknowledgment. The present research was partly supported by Grant-in-Aid for Young Scientists (A) 16681006 from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. EF050274D (35) Gratzel, M. Nature 2001, 414, 338.
