J. Chil. Chem. Soc., 48, N 2 (2003)

ELECTROCATALYTIC REDUCTION OF CARBON DIOXIDE AT POLYMERIC
COBALT TETRA (3-AMINO (PHENY1) PORPHYRIN GLASSY
CARBON-MODIFIED ELECTRODES".

M.A. RIQUELME, M. ISAACS, M. LUCERO, E. TROLLUND, M.J. AGUIRRE*

Departamento de Química de los Materiales, Facultad de Química y Biología, Universidad de Santiago de Chile,
Casilla 40, Correo 33, Santiago, Chile.
*Corresponding author: Tel: 56-2-6812575, Fax: 56-2-6812108, e-mail: maguirre@lauca.usach.cl.

J. CANALES

Departamento de Ciencias Químicas, Facultad de Ingeniería, Ciencias y Administración, Universidad de La Frontera,
Casilla 54-D, Temuco, Chile.
( Received : November 25, 2002 ­ Accepted : March 28, 2003 )

ABSTRACT

Electrocatalytic carbon dioxide reduction was studied in aqueous media on glassy carbon electrode coated with polymeric Co-tetra-3-aminophenylporphyrin. The polymeric complex-modified electrode catalyses the carbon dioxide reduction to carbon monoxide and formic acid. The monomeric system does not catalyze the reduction probably because it is not stable. The polymer, instead, presents higher activity at lower potentials and loses its activity only after more than four hours of electrolysis at constant potential. The redox couple responsible for the catalysis is Co(II)/Co(I). Kinetic electro-reduction parameters for the polymer-modified electrode were obtained and a probable mechanism operating in aqueous media containing sodium perchlorate as electrolyte is proposed.

Key Words: Electrocatalytic CO₂ Reduction; Poly-Co-tetra-3-aminophenylporphyrin, Modified Electrodes, Electrocatalytic Mechanism.

I. INTRODUCTION

Transition-metal aza-macrocycles modified electrodes have been widely studied due to their numerous applications on electrocatalysis such as O₂ (1-3), CO₂, (4-6) and peroxide reduction (7), thiols (8) and hydrazine oxidation (9). In the last years, the possibility of electropolymerizing macrocyclic complexes on the electrode surface has also been studied, with the purpose of increasing stability and electrocatalytic activity of these systems (10-12). The reduction of CO₂ has been extensively studied due to the increasing "greenhouse effect" that causes undesirable changes in the environment. For that reaction, many catalysts have been used, Cu cathodes between others (13, 14). For these cathodes, one of the reaction products is methane that involves an 8-electrons reduction. However, metal cathodes suffer corrosion or passivation problems that avoid its use as electrocatalysts. For that reason and looking for cheaper and versatile systems, metal transition azamacrocycles have been investigated as catalysts for the reduction of carbon dioxide also in homogeneous and heterogeneous media (4-6). Among them, Ni cyclam has shown high activity and selectivity for a unique product, carbon monoxide (6). Several cobalt azamacrocycles has been used for the reduction of carbon dioxide, showing good catalytic behavior. Porphyrins and phthalocyanines of different transition metals have been used for this reaction. In this work we present a study of glassy carbon electrodes modified by poly-cobalt-tetra-3-aminophenyLporphyrin as an electrocatalyst for carbon dioxide reduction reaction in aqueous media. These studies show that this polymerized complex is very active to the formation of carbon monoxide and formic acid while the monomer modified electrode does not catalyzes the reaction, probably because it is not stable.

II. EXPERIMENTAL

Cobalt tetra-3-aminophenyLporphyrin (Midcentury Co. Posen, Illionois) was used as received. Modification of the electrode with the monomer was performed placing a drop of monomer solution (1·10^-3M) in dimethylformamide (DMF) on the electrode glassy carbon surface. After one hour the electrodic was rinsed with DMF, ethanol and deionized, bidistilled water. Polymeric films were grown by potentiodynamically cycling the glassy carbon electrode in a 1·10^-3M Co-tetra-3-amninophenylporphyrin, 0.1M tetra butylammoniumpeacHlorate (TBAP) DMF solution, between -0.9 and +1.15V vs. Ag/AgCl during 50 cycles at 0.2 Vs^-1. After polymerization, the modified electrode was rinsed with DMF, ethanol and then with bidistilled water. Electrochemical experiments were performed in a three-compartment glass cell, one for each of the electrodes: the working electrode, a disc of glassy carbon (A = 0.071 cm²), the reference, saturated Ag/AgCl, to which all the potentials are quoted, and the counter electrode, a Pt coil (A = 14 cm²). All measurements were carried out in aqueous solutions of 0.1M NaClO₄, pH 6.9 (without CO₂), or pH 3.9 (saturated with CO₂). The solutions were bubbled with pure N₂, CO₂, or mixtures of N₂/CO₂. The cyclic voltammetry measurements were carried out in an AFCBP1 Pine bipotentiostat, connected to a rotating disk unit. Potential controlled electrolysis was performed with a Universal Programmer 175 connected to a digital coulometer 174 PAR. For electrolysis measurements an H-type cell and a reticulated glassy carbon as working electrode were used. The solvent, N, N dimethylformamide (DMF) (Merck, p.a.) and the supporting electrolyte, tetrabutylammonium perchlorate (TBAP) (J.T. Baker, p.a.), were dried before use following the procedures reported in literature (15). Analysis of the products was performed by gas chromatography. A Varian 3400 gas chromatograph equipped with capillary columns: molecular sieve connected to a TCD detector and a column DB1 connected to a FID detector were used for identifying the gas products. A UV-Vis spectroscopic method (chromotropic acid method (16)) was used to determinate formaldehyde and formic acid.

III. RESULTS AND DISCUSSION

Figure 1 shows the structure of the complex used to modify the glassy carbon electrodes. Phenyl groups, which are substituted in meta position with amino groups, are practically perpendicular to the macrocycle, but this fact does not hinder the electropolymerization to form a conductive film.

Figure 1. Structure of cobalt-tetra(x-aminopheniyl)porphyrins X= 1 (orto), 2(meta) or 3 (para).

In Figure 2, cyclic voltammograms corresponding to the electropolymerization of the Co-tetra(3-aminophenyl)porphyrin on the glassy carbon electrode from a 1·10^-3 M solution of the monomer in DMF/0.1 M TBAP are shown. The increase of the peaks after subsequent potential cycling illustrates the growth of the polymeric film. The irreversible oxidation wave that appears at ca. 1.15 V corresponds to the oxidation of the amino groups that give origin to the electropolymerization. At ca. -0.9 V, there is a reversible redox couple attributed to the Co(II)/Co(I) process. The charge of the cathodic part of this couple is higher than the anodic part, and indeed, corresponds to two electrons and the anodic part, only to one electron.

Figure 2. Electropolimerization of Co-tetra(3-aminophenyl)porphyrin (1 mM) in a solution of DMF/0.1TBAP. Potential range: -0.9 to + 1.15V versus Ag/AgCl. 50 potentiodynamic cycles. Scan rate 0.2Vs-1. Solution purged with N2. Electrodic Surface: Glassy carbon.

The reason for this difference is simple and can be appreciated in Figure 3. The cathodic part of this wave has two components: the Co(II)/Co(I) redox couple and the reduction of the oxidized free-amino groups generated during the positive scan. When the positive limit of the scan becomes less positive, as shown in the Figure, the cathodic charge decreases to become equal to the anodic charge.

Figure 3. Voltammetric response of the polymeric-modified electrode in an aqueous NaOH solution. PH = 13. One arrow indicates the change in the cathodic wave when the anodic limit is increased (other arrow). Scan rate: 0.1Vs-1.

Figure 4 depicts the voltammograms of the polymer-modified electrode in aqueous solution in the presence and in the absence of CO₂. The response under CO₂ atmosphere is not easy to interpret. There are not very large differences between N₂ and CO₂. Carbon dioxide modifies the peak corresponding to the redox couple, which becomes broader and the peak is shifted to values that are more negative. There is a current discharge in both cases, which appears at nearly the same potential (ca. -1.6 V), but in one case corresponds to hydrogen evolution (under N₂) and in the other, to the reduction of CO₂ or both processes.

Figure 4. Voltammetric response of the polymeric-modified electrode in an aqueous solution conyaining NaClO4 0.1 M saturated with N2 ( continuous line) or CO2 (dashed line).

The open circuit potential varies when N₂ is changed by CO₂. In the first case, the value is: -0.13 V and in the other ­0.278 V. This change could correspond to an open circuit-adduct formation between the CO₂ and the metal center because the change in the open circuit potential by effect of the pH is very low compared with this change. The formation of an adduct could explain the changes in the voltammetric profile when the modified electrode is in the presence of CO₂. The electrolysis at controlled potential ­1.25 V gives as product carbon monoxide, formic acid and hydrogen. Hydrogen and carbon monoxide are the main products. Depending on the time of electrolysis, the proportion of CO₂ and H₂ varies slightly around 50%. Formic acid is a minor product. From the polarization curves obtained in the presence of CO₂, (see Figure 5) the Tafel slopes were obtained. They correspond to -0.12 V/decade.

Figure5. Polarization curves of the polymeric-modified electrode in an aqueous solution (0.1M NaClO4) saturated with CO2. Scan rate 0.005Vs-1. The rotation are indicated in the figure.

A -0.12 V/decade-Tafel slope indicates that the rate-determining step of the reaction is the first electron transfer (17). In this case, the formation of the active species, [Co (I)···CO₂] from [Co(II)···CO₂]. It seems surprising that the cathodic wave corresponding to the reduction of CO₂ does not appear at the potential of the metallic redox couple. Generally, when the Tafel slope is -0.12 V/decade, the redox processes occurring simultaneously with the formation of the active species. In this case, the electrolysis were done at the potential corresponding to the redox couple of the metal and reduction products (CO, formic acid and H₂) were detected. Then, the reduction takes place at this potential and the Tafel slope of -0.12 V/decade is not an artifact. However, there is not a current discharge at this potential. A possible explanation for this fact is a poisoning of the polymer with the reduction product. The rapid scan of the potential used in the voltammograms does not allow (using this time scale) the product of the reaction to go out of the polymer to the bulk. In fact, the electrolysis gives increasing product during four hours. On the other hand, there is another process that takes place under N₂ and under mixtures of CO₂ and N₂ that does not correspond to H₂ evolution because the reaction is occurring during all the scanning of the potential, as can be seen in Figure 6.

Figure 6. Polarization curves of the polymeric-modified electrode in an aqueous solution (0.1M NaClO4) saturated with a mixture 50%v/v N2/CO2. Scan rate: 0.005Vs-1. The rotation are indicated in the figure.

The currents of the polarization curves corresponding to N₂ or N₂/CO₂ mixtures are low, do not have the typical profile corresponding to this reaction and the Tafel slopes (0.2-0.4 V/decade) do not have kinetic significance. It is possible that this unknown process takes place during the reduction of CO₂ and modifies the profile of the voltammograms. However, the explanation of the profile of the voltammogram under CO₂ and its bad-concordance with the Tafel slope is not clear. The number of transferred electrons is 2 (because the nature of the products) and an order on CO₂ of 1 is obtained from the Koutecky -Levich plots (not shown). With these data, a probable mechanism can be postulated at acids pHs and ignoring the unknown reduction process.

In this mechanism, P represents the ligand. The charge on the macrocycle was omitted for simplicity. It was represented only the electrons transferred during the reduction process. In general, for other macrocycles, there is evidence of a second electron reduction-taking place at the ligand as a required process to obtain products that are two [Co(I)-P---CO₂-] electrons reduced. In our case, there is not evidence on the participation of the ligand because the electrolysis done at the potential corresponding to the metal gives the same products. If only the single-reduced metal participates, the one-electron product will be oxalate (18)

In the step viii) a hydride was formed. In this work we do not found evidence for this formation. However, there are many examples in the literature (12, 19) where a hydride-intermediate is postulated.

CONCLUSIONS

The poly-Co-tetra(3-aminophenyl)porphyrin-modified electrode catalyses the electroreduction of carbon dioxide in aqueous media. The monomeric-modified electrode is not stable in this media. The polymeric system gives carbon monoxide, formic acid and molecular hydrogen as products. The slow step of the reaction is the first electronic transfer, when Co(I) is formed. There is another process of reduction that takes place during the reduction of CO₂ that could mask a normal behavior in the current of the voltammogram of the polymer-modified electrode in the presence of carbon dioxide.

ACKNOWLEDGMENTS

Authors aknowledge Fondecyt (project 1010695) financial support. M.A. R. acknowledge a Mecesup doctoral fellowship.

REFERENCES

1. Y-H. Tse, P. Panda, H. Lam, J. Zhang, W. J. Pietro, and A. B. P. Lever, J. Porph. Phthal., 1 (1997) 3

2. C. Coutanceau, A. Rokotondrainible, P. Crougneau, J. M. Lèger, C. Lamy, J. Electroanal. Chem., 386 ( 1998) 173.

3. G. Ramírez, Licenciate Degree Thesis, Facultad de Química y Biología, Universidad de Santiago de Chile, 2001.

4. (a). M. Beley, J. P. Collin, R. Ruppert, J.P Sauvage, J. Am. Chem. Soc., 24 (1986) 7461 (b). J. P. Collin, J. P. Sauvage. Coord. Chem. Rev. , 93 (1989) 245. (c) F. Abba, G. De Santi, L. Fabrizzi, M. Licchelli, A. Manotti, P. Pallavicini, A. Poggi, F. Ugozolli. Inorg. Chem. 33 (1994) 1366.

5. J. Costamagna, J. Canales, J. Vargas, G. Ferraudi, Coord.. Chem Rev., 148 (1996) 221. (b) I. Bhugun, D. Lexa, J. M. Saveant, J. Am. Chem Soc., 118 (1996) 257.

6. Jarzebinska, P. Rowinski, I. Zawista, R. Bilewicks, L. Siegfried, T. Kaden, Anal. Chim. Acta, 396 (1999) 1

7. H. Li, T. F. Guarr, J. Chem. Soc. Commun. (1989) 832. (b) P. Vasudean, N. Phougat, A. K. Shukla, Appl. Organomet. Chem., 10 (1996) 591.

8. E.M. Tyapochkin, E.I. Kozliak, J. Porph. Phthal. , 5 (2001) 405.

9. Q. Y. Peng and T.F. Guarr, Electrochim. Acta, 39 (1994) 2629 . (b) D. W. Pang, B. H. Deng and Z.L. Wang, Electrochim. Acta, 39 (1994) 847.

10. P. Ardiles, E. Trollund, M. Isaacs, F. Armijo, J.C. Canales, M.J. Aguirre, M.J. Canales. J. Mol. Catal. A, Chem., 165 (2001) 169

11. Brett , A. M. Oliveira; "Electrochemistry. Principles, Methods and Applications", Oxford University Press Inc, 1993.

12. H. Aga, A. Aramata, Y. Hisaeda, J. Electroanal. Chem. 437 (1997) 111.

13. K. Hara, A. Kudo, T. Sakata. J. Electroanal. Chem., 391 (1995) 141. (b) R. Shiratsuchi, G. Nogami. J. Electrochem Soc., 2 (1996) 582.

14. Kaneko, K. Iba, k. Ohta, T. Mizuno, A. Saji. Electrochim. Acta, 44 (1998) 573.

15. T. Sawyer, J. L. Roberts, "Experimental electrochemistry for chemists". Cap. 4. Wiley Interscience, New York, 1974.

16. (a) J. A. Ramos Sende, C. R. Arana, L. Hernández, K. T. Potts, M. Keshevarz-K and H. D. Abruña, Inorg. Chem., 34 (1995) 3339. (b) W. M. Grant. Anal. Chem., 20 (1948) 3.

17. (a) A.J. Bard and L.R. Faulkner, "Electrochemical Methods, Fundamentals and Applications", J. Wiley, New York, 1980. (b) J.O'M. Bockris, A.K.N. Reddy, in "Electroquímica Moderna", Vol. 2, (Reverté S.A. Ed)., Spain, 1978.

18. J. Grodkowski, T. Dhanasekaran, P. Neta, P. Hambright, B.S. Brunschwing, K. Shinozaki, E. Fujita, J. Phys. Chem.A, 104 (2000) 11332.

19. H. Tanaka, A. Aramata, J. Electroanal. Chem., 437 (1997) 29.