SciELO - Scientific Electronic Library Online

 
vol.93 número4-6Operational performance of the SAC-C satellite nickel-hydrogen batteriesConformational equilibrium of the thermal unimolecular decomposition reactions of substituted 4-fluorophenyl 1,2,4-trioxanes índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

  • No hay articulos citadosCitado por SciELO

Links relacionados

  • No hay articulos similaresSimilares en SciELO

Compartir


Anales de la Asociación Química Argentina

versión impresa ISSN 0365-0375

An. Asoc. Quím. Argent. v.93 n.4-6 Buenos Aires ene./dic. 2005

 

REGULAR PAPERS

Electrochemical Behaviour Of Carbon Dioxide On Platinum And Rhodium

Martins, M.E.

Instituto de Investigaciones Fisicoquímicas Teóricas y Aplicadas (INIFTA), Facultad de Ciencias Exactas, Universidad Nacional de La Plata, Sucursal 4, CC 16 (1900), La Plata, Argentina. Fax: +54-221-425-4642, E-mail : mmartins@inifta.unlp.edu.ar

Received August16th, 2005. In final form September 14th, 2005.
Dedicated to the memory of the late Prof. Hans J. Schumacher on the occasion of his 100th birthday

Abstract
The electroadsorption and electroreduction of carbon dioxide, and the subsequent electrooxidation of reduced carbon dioxide was studied on polycrystalline and electrofaceted Pt and on polycrystalline Rh electrodes. The influence of the adsorption potential and the adsorption time was established. Combining voltammetric charge measurements and differential electrochemical mass spectrometry (DEMS) analysis, three adsorbates were tentatively identified, namely COOH and CO in bridge and linear configurations on Pt, and COH in triple and single adsorption configurations on Rh. By applying medium frequency potential routines on the Pt-adsorbate/sulphuric acid interface, the existence of equilibrium between adsorbates was demonstrated.

Resumen
Se estudia la electroadsorción y la electrorreducción de dióxido de carbono como así también la electrooxidación del dióxido de carbono reducido sobre Pt policristalino y electrofacetado y sobre Rh policristalino. Se determina la influencia del potencial y del tiempo de adsorción. Mediante el empleo de técnicas combinadas voltamperométricas y de espectrometría de masa diferencial (DEMS), se identificaron tres adsorbatos posibles sobre Pt, COOH y CO en configuración lineal y puente, y dos sobre Rh, COH en configuraciones de adsorción simple y triple. Mediante la aplicación de perturbaciones de potencial a la interfase Pt-adsorbato/ácido sulfúrico acuoso, se demuestra la existencia de un equilibrio entre los adsorbatos.

Introduction

     The electrochemical behaviour of several C1-containing organic molecules shares some similarities that encourage the search of a molecular model for these compounds. However, the concept of molecular model should be strictly defined in order to be used for generalized purposes. For example, CO can be considered as a molecular model regarding electroadsorption processes, since, once it is adsorbed onto a metallic surface, the chemical structure remains unaltered and only acquires different surface configurations, i.e., on top, bridge and multiple states [1-5]. In contrast, those C1-containing molecules in which C atom exhibits a higher oxidation state than CO (carbon dioxide, formaldehyde), the formation of adsorbates on the metallic surface is assisted by H ad-atoms [6-8]. The chemical structure of these adsorbates is different from the original molecule, indicating that the electroadsorption process differs from that observed for CO.
     Carbon dioxide can be considered as a molecular model by itself, because is the only carbonaceous substance with the C atom in the highest oxidation state (+4). Moreover, from an electrocatalytic point of view, carbon dioxide could be also considered as a molecular model for the electroadsorption process taking into account the nature of the adsorbed residues, which are similar to those of other C1-containing compounds. There is wide experimental evidence that supports such assumptions. Several infrared spectroscopic studies allow to conclude that the main adsorbate formed from CO2 [9-13] and other C1-containing compounds [1,13-20] is CO with different surface configurations. Besides, selecting accurately the potential value of the modulation, COOH and COH adsorbates have been also identified [13-15]. However, transient electrochemical techniques such as voltammetry and coupled voltammetry/DEMS usually conclude that COH is the main adsorbate [16].
     Therefore, those results suggest that, according to the experimental conditions, that different adsorbates can be formed. The experimental set up for infrared studies includes a thin layer configuration that favours the formation of CO, which is a thermodynamically stable product. On the other hand, DEMS analysis detects the first product, i.e., the one formed under kinetic control, because the reaction species formed at the surface of the electrode are detected on a very short time scale. Moreover, voltammetry allows predetermining the adsorption time and consequently to recognize adsorbates which were formed through either thermodynamic or kinetic control. Hence, low adsorption times favour the product initially formed and for prolonged adsorption times prevails the formation of thermodynamically stable adsorbates. As a result, cyclic voltammetry comes out as a useful technique to determine the conditions under which the adsorbates are formed. The occurrence of different adsorbates suggests the existence of equilibrium among them, the equilibrium position being dependent on the experimental conditions.
     The aim of this work is to study the formation of reduced CO2 adsorbates (r-CO2) on platinum and rhodium, and to determine the possible relations between them.

Experimental

     Electrochemical measurements were made by using a conventional glass-made electrochemical cell with Pt and Rh as working electrodes, large gauzes of Pt and Rh as counter electrodes, and a reversible hydrogen electrode as the reference. Potentials in the text are given on the reversible hydrogen electrode (RHE) scale.
     Polycrystalline Pt and Rh wires were cleaned in hot H2SO4 + HNO3 mixture, and cycled at 0.05 V s-1 from 0.05 to 1.50 V for Pt and from 0.02 to 1.40 V for Rh. For some experiments, preferentially oriented (electrofaceted) Pt surfaces were used and prepared employing potential routines that are thoroughly described elsewhere [6]. The real surface area of the electrodes was evaluated from the H-atom electrosorption charge. For this purpose, the H-adatom monolayer charge density for Pt and Rh surfaces was taken from the literature to be 210 µC cm-2.
     1 M aqueous H2SO4 (base electrolyte), deoxygenated by bubbling high purity nitrogen gas, and 1 M aqueous H2SO4 saturated with CO2 (working solution) were prepared from analytical reagent quality chemicals and Millipore-Milli-Q* water. All runs were performed at 298 K under a nitrogen atmosphere.
     For electroadsorption experiments, carbon dioxide was bubbled into the supporting electrolyte for 30 min while the platinum working electrode was subjected to a potential cycling from 0.05 to 1.45 V at 0.1 Vs-1. Carbon dioxide adsorption on platinum was accomplished by holding the potential at the adsorption value Ead, (0.05 ≤ Ead ≤ 0.35) V for a period, tad, (0 ≤ tad ≤ 600 s), followed by nitrogen bubbled through the electrolyte to remove carbon dioxide. Subsequently, a negative going potential scan from Ead to the cathodic switching potential, Ecs = 0.05 V followed by a positive going potential scan to the anodic switching potential, Eas = 1.45 V at 0.1 Vs-1 were applied. Carbon dioxide adsorbates generated at Ead were oxidized during the positive going potential scan to Eas.
     On the other hand, carbon dioxide adsorption on rhodium was accomplished at Ead (- 0.35 ≤ Ead ≤ 0.05 V) for a period tad (0 ≤ tad ≤ 900 s). Taking into account that Ead was located within a potential domain where the hydrogen evolution reaction takes place, a potential step to E = 0.25 V during 5 min was applied accompanied by nitrogen bubbling within the electrolyte to displace carbon dioxide. Afterwards, the oxidation of carbon dioxide adsorbates was performed by means of a positive going potential scan up to Eas = 1.4 V at 0.1 Vs-1.
     Differential Electrochemical Mass Spectrometry (DEMS) experiments were carried out in a 2 mL flow cell. Rh working electrode was electrodeposited at a constant potential onto a substrate constructed with a hydrophobic PTFE membrane (Scimat Ltd., porous size 17 µm, 50 % porosity) covered with a gold deposit, employing a solution containing 3 % RhCl3 in 1 M HCl. The mass signal m/z = 44 ([CO2]•+) corresponding to CO2 was followed during the oxidation of the adsorbates. Details on the procedure for the calibration of DEMS systems are thoroughly described elsewhere [17].
     Medium frequency potential routine (MFPR) involving a repetitive triangular potential cycling run at a sweep rate equal to 2 Vs-1 was applied to an electrofaceted Pt surface previously covered by r-CO2 adsorbates. The potential domain of the MFPR corresponded to that of the OH-electrosorption reaction.

Results

General features of voltammograms of r-CO2 on Pt and Rh
     The voltammograms shown in Figure 1a, exhibit a current peak at 0.82 V and a shoulder at 0.73 V corresponding to the oxidation of r-CO2 adsorbates generated on Pt within the potential range from 0.05 to 0.35 V for an adsorption time (30 ≤ tad ≤ 600) s. For potentials higher than 0.95 V, the electro-oxidation current and that of the blank are identical, hence at this potential the complete desorption of r-CO2 has been achieved. The amount of r-CO2, which has been calculated from the electrooxidation charge density, varies with Ead (Figure 1b). For Ead ≤ 0.15 V, a maximum constant charge density of 194 µCcm-2 is found, while for Ead > 0.15 V, a fast decrease of the amount of adsorbate is observed. Likewise, the influence of tad was studied at Ead = 0.10 V (Figure 1c), and for tad = 300 s the maximum coverage is attained.

Figure 1. a) Voltammograms of pc Pt in working solution after Ead = 0.10 V for 30 ≤ tad≤ 600 s; v = 0.1 V s-1; b) Charge density vs Ead plot for tad = 300 s; c) Charge density vs tad plot at Ead = 0.10 V.

     To gain information about the r-CO2 adsorbates, the anodic part of the voltammograms performed at Ead after different tad were graphically deconvoluted. The results are depicted in Figure 2a where three contributions, namely AIII (0.68 - 0.70 V), AI (0.73 - 0.77 V) and AII (0.80 - 0.82 V) can be distinguished. The evolution of charge density with tad for each contribution (Figure 2b) shows that AI steadily increases up to 300 s, while both AIII and AII display a maximum at ca. 100 s and thereafter decrease either to a constant finite value (AII) or to a null value (AIII).

Figure 2. a) Anodic voltammograms of pc Pt in working solution after Ead = 0.10 V for tad = 30, 120 and 300 s, (O) experimental data (—) deconvoluted voltammograms showing components AI, AII, AIII; b) Charge density vs tad plot for components AI, AII, AIII depicted in a) v = 0.1 V s-1.

     On the other hand, the voltammogram for the electro-oxidation of r-CO2 on Rh is less clearly defined than that on Pt because of the overlapping of the early stages of O-electroadsorption, which occur at a potential lower than that of Pt. An electro-oxidation current peak corresponding to r-CO2 is observed at 0.78 V (Figure 3a). The maximum amount of adsorbate is achieved at high negative values of Ead and for prolonged tad values (Figure 3b).

Figure 3. a) Voltammograms of pc Rh in working solution after Ead = -0.15 V and 0.05 V for tad = 600 s; ( ……) blank; b) Charge density vs tad plot for -0.35 ≤ Ead ≤ 0.05 V. v = 0.1 Vs-1.

DEMS analysis of the electrooxidation product formed on Rh
     The electrooxidation of r-CO2 formed at Ead = 0 V for tad = 10 min in 0.5 M aqueous H2SO4 was followed by DEMS (Figure 4, dashed line), monitoring the signal corresponding to the production of CO2 (m/z = 44). It is not possible to perform this study at Ead < 0 V, since the hydrogen evolution reaction brings about instability of Rh-deposited electrode. The electrooxidation of r-CO2 begins at 0.50 V and a current peak at 0.70 V is developed. The oxidation charge density related to the latter process is equal to 141 µCcm-2, and that associated to H electrodesorption results equal to 102 µCcm-2. From these figures, a value for e.p.s. (number of electrons per site) = 1.3 was calculated. Furthermore, from the massograms for the signal m/z = 44, the value of n (number of electrons) consumed to form one molecule of CO2 is 2.97 (≈ 3). Consequently, from e.p.s. and n data, the number of Rh sites occupied was found to be 2.3.

Figure 4. a) Voltammograms of pc Rh showing the electro-oxidation of ( …) CO adsorbates after Ead = 0.02 V and (-----) CO2 adsorbates after Ead = 0 V for tad = 10 min, (—) blank. b) Simultaneous massograms of pc Rh under the same conditions as a). m/ e = 44. v = 0.01 Vs-1.

     For the sake of comparison, the electro-oxidation of CO was then performed at Ead = 0.02 V for tad = 10 min in the same supporting electrolyte (Figure 4, dotted line). In this case, the current peak also starts at 0.50 V, but the maximum is reached at 0.63 V, involving an oxidation charge density of 318 µCcm-2. The oxidation charge related to H electrodesorption was 57 µCcm-2, yielding a value of e.p.s. of 2.1. The number of electrons was found to be 2, indicating that one site of Rh is occupied by adsorbed CO.

A proposal for identification of the adsorbates
     A proposal for the identification of the adsorbates was encouraged, based specifically on e.p.s. values, the voltammetric behaviour, as well as a comparison of our results to those found in the literature concerning spectroscopic studies [15,21]. The results are resumed in Table 1.

Table 1 Identification of adsorbates

Where COL and COB correspond to linear and bridge configurations, respectively, of adsorbed CO.

Assessment of the surface equilibrium between adsorbates
     The relative contribution of each adsorbate depends on the crystallographic sites of the electrode surface [11]. At high Miller index orientations, as those obtained by electrochemical faceting, COB species are mainly formed on (100) sites, while COL is the preferred intermediate on (110) terraces [18]. It has been previously shown that, on a preferentially oriented surface exhibiting (110) sites, all three r-CO2 adsorbates are detected simultaneously [6]. Hence, the results indicate that electrofaceted surfaces can be appropriate substrates to account for the existence of a surface equilibrium between the adsorbates.
     To carry out this study, CO2 was allowed to react at Ead = 0.20 V for 300 s on a Pt electrofaceted surface exhibiting a Pt(320) structure [22]. Once the CO2 residues have been formed, a medium frequency potential routine (MFPR) was applied at f = 5 Hz in the (0.70- 0.90) V potential domain, the stripping performance being compared with that obtained without application of the routine (Figure 5a). The comparison between the deconvoluted stripping profile
without (Figure 5b) and with (Figure 5c) the use of the MFPR indicates that a redistribution of the adsorbates has taken place.

Figure 5. a) Voltammograms of electrofaceted Pt in working solution after Ead = 0.20 V for tad = 300 s; (- - - - ) with MFPR, ( …) without MFPR, (—) blank. b) (—) Deconvoluted voltammogram without MFPR, ( …) experimental data; c) (—) Deconvoluted voltammogram with MFPR, ( …) experimental data. v = 0.1 Vs-1.

Discussion

Carbon dioxide electroadsorption on Pt and Rh
     It is widely recognized that the formation r-CO2 adsorbates from CO2 is assisted by H adatoms [7,23,24] and that these adsorbates do not cover entirely the surface. For both metals, Pt and Rh, the largest amount of r-CO2 is produced at potentials more negatives than the potential of zero charge, Epzc (0.20 V for Pt and 0.00 V for Rh) [25]. In fact, Epzc plays a key role in the electroadsorption process, as for E > Epzc the amount of r-CO2 decreases remarkably. For E = Epzc, CO2 molecules approach the metallic surface without adsorption. Already adsorbed H atoms play the role of anchoring the CO2 molecules, by reducing them. The nature of the first electroreduced product is unknown, although an ensemble of CO2, H2O and Had has been postulated [7]. Thus, the electroadsorption process for Pt can be formulated as

[1]
[2]

     It has been shown that H adatoms located at (110) sites have the highest activity toward CO2 electro-reduction [18]. This fact was also demonstrated by performing potential perturbations in a medium frequency range (2-20 Hz) within the potential domain for H electrosorption on (110) sites. Furthermore, H atoms adsorbed on terraces (100) can reduce CO2 at the interphase, probably forming (CO)ad in a bridge configuration,

[3]

     A different situation arises for Rh. The fact that the adsorbates formed on Rh are more reduced than CO suggests that CO2 can be dissociatively adsorbed, as already found in the solid/gas interphase [19,20]

[4]

and these adsorbates are reduced in the hydrogen potential region.

Electroreduction of the adsorbates initially formed
     H atoms participating of the ensemble built on Pt at EEpzc are able to reduce CO2 producing subsequently formiate species,

[5]

     The voltammetric results indicate that another adsorbed residue has been formed. The application of the MFPR in the OH-electrosorption potential domain allows to suggest that both adsorbates, namely (COOH)ad and (CO)ad, could be interrelated through an equilibrium involving (OH)ad species according to

[6]

The prevalence of either (COOH)ad or (CO)ad depends on the time during which the adsorption potential is applied. Thus, short times favour the formation of (COOH)ad over (CO)ad, but prolonged adsorption potentials account for the formation of most stable species, i.e., (CO)L if H atoms are adsorbed at Pt(110) sites or(CO)B if H atoms are adsorbed at Pt (100) sites.
     For Rh electrodes and according to DEMS results, a combination of three-coordinated (COH) and formyl type (COH) adsorbates can be proposed.

Relationship between the adsorbates: surface equilibrium
     The main consequence of the application of the MFPR is focused on the variation of the interatomic Pt-Pt distance, which can be almost certainly responsible of a change in the CO configuration, namely, COL changes into COB. This should indicate that (COOH)ad and (CO)L adsorbates are probably linked through an equilibrium involving (CO)B species,

[7]

and

[8]

In this sense, the formation of (CO)B is favoured after the application of the MFPR.

Electrooxidation of the adsorbates
     Carbon dioxide is the oxidation product formed both on Pt and Rh. Formiate can be oxidized at a potential more negative than that of OH-electrosorption reaction. Thus, the electrooxidation can be described as

[9]

For (CO)ad, both in bridge and lineal configurations, the electrooxidation involves the previous adsorption of OH species on the free surface sites, according to

[10]

which further oxidizes the adsorbates as follows

[11]

The electro-oxidation process involves both reactions meaning that 2 electrons are exchanged.
     On Rh the adsorbates are also oxidized by (OH)ad species,

[12]

and then, the global electrooxidation reaction involves 3 electrons.

Conclusions

     Carbon dioxide interacts with Pt and Rh electrodes in aqueous sulphuric acid through a complex mechanism that involves several stages keeping a strong dependence on the substrate.
     For Pt, CO2 interacts with H adatoms and interfacial water molecules, forming a complex that is subsequently adsorbed. The formation of this complex is assisted by H adatoms located either at (110) or (100) sites. This complex is electro-reduced to (COOH)ad which is in equilibrium with (OH)ad and (COL)ad or with (OH)ad and (COB)ad. These configurations link up with the H adatom surface location, i.e., (COL)ad is favoured by (110) sites whereas (COB)ad by (100) sites.
     For the case of Rh, CO2 is adsorbed in a dissociated fashion leaving CO as a strongly adsorbed residue which is subsequently hydrogenated to (COH)ad. This adsorbate is strongly bound to the surface. Both adsorbates, (COOH)ad on Pt or (COH)ad on Rh are electrooxidized to CO2.

Acknowledgements

     This work was partially financed by the ANPCyT PICTR /0188 (Argentina). Author acknowledges Dr. Carlos Fernando Zinola Sánchez, Dr. Eduardo Méndez Morales (Universidad de la República, Uruguay), Dr. Elena Pastor, Dr. José Luis Rodríguez and Dr. Carmen Arévalo (Universidad de La Laguna, España) for DEMS facilities and helpful discussions. MEM is a member of CONICET-UNLP.

References

[1] Corrigan D.; M.J. Weaver, J. Electroanal. Chem. 1988, 241, 143.         [ Links ]

[2] Kunimatsu K.; Lezna R.O.; M. Enyo, J. Electroanal. Chem. 1989, 258,115.         [ Links ]

[3] Chang S.C.; Weaver M.J., J. Electroanal. Chem. 1990, 285, 263.         [ Links ]

[4] Chang S.C.; Weaver M.J., Surf. Sci. 1990, 238, 142.         [ Links ]

[5] Westerhoff B.; R. Holze, Ber. Bunsenges. Phys. Chem. 1993, 97, 418.         [ Links ]

[6] Méndez E.; Martins M.E.; C.F. Zinola, J. Electroanal. Chem. 1999, 477, 41.         [ Links ]

[7] Arévalo M.C.; Gomis-Bas C.; Hahn F.; Beden B.; Arévalo A.; Arvia A.J., Electrochim. Acta 1994, 39, 793.         [ Links ]

[8] Taguchi S.; A. Aramata, Electrochim. Acta 1994, 39, 2533.         [ Links ]

[9] Nikolic B.Z.; Huang H.; Gervasio D.; Lin A.; Fierro C.; Adzic R.R.; E.B. Yeager, J. Electroanal. Chem. 1990, 295, 415.         [ Links ]

[10] Rodes A.; Pastor E.; T. Iwasita, J. Electroanal. Chem. 1994, 373, 167.         [ Links ]

[11] Rodes A.; Pastor E.; T. Iwasita, J. Electroanal. Chem. 1994, 369, 183.         [ Links ]

[12] Rodes A.; Pastor E.; T. Iwasita, J. Electroanal. Chem. 1994, 377, 215.         [ Links ]

[13] Sun S.G.; Clavilier J.; Bewick A., J. Electroanal. Chem. 1988, 240, 147.         [ Links ]

[14] Nichols R.J.; A. Bewick, Electrochim. Acta 1988, 33, 1691.         [ Links ]

[15] Iwasita T.; Nart F.C.; López B.; W. Vielstich, Electrochim. Acta 1992, 37, 2361.         [ Links ]

[16] Willsau J.; Heitbaum J., Electrochim. Acta 1986, 31, 943.         [ Links ]

[17] Silva-Chong J.; Méndez E.; Rodríguez J.L.; Arévalo M. C.; Pastor E., Electrochim. Acta 2002, 47, 1441.         [ Links ]

[18] Hoshi N.; Mizumura T.; Hori H., Electrochim. Acta 1995, 40, 883.         [ Links ]

[19] van Tol M.F.H.; Gielbert A.; B.E. Nieuwenhuys, Appl. Surf. Sci. 1993, 67, 166.         [ Links ]

[20] van Tol M.F.H.; Gielbert A.; Wolf R.M.; Lie A.B.K.; B.E. Nieuwenhuys, Surf. Sci. 1993, 287/288, 201.         [ Links ]

[21] Arévalo C.; Gomis-Bas C.; F. Hahn, Electrochim. Acta 1998, 44, 1369.         [ Links ]

[22] Furuya N.; Shibata M., J. Electroanal. Chem., 1999, 467, 85.         [ Links ]

[23] Electrochemical and Electrocatalytic Reactions of Carbon Dioxide, Elsevier, 1993.         [ Links ]

[24] Jitaru M.; Lowy D.A.; Toma M.; Toma B.C.; I. Oniciu, J. Appl. Electrochem. 1997, 27, 875.         [ Links ]

[25] Encyclopedia of Electrochemistry of the Elements, Editor A. J. Bard, Marcel Dekker, Inc., New York., 1976.         [ Links ]

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons