Elsevier

Electrochimica Acta

Volume 159, 20 March 2015, Pages 242-251
Electrochimica Acta

Tinned graphite felt cathodes for scale-up of electrochemical reduction of aqueous CO2

https://doi.org/10.1016/j.electacta.2015.01.209Get rights and content

Abstract

Three-dimensional cathodes with large volumetric surface areas for CO2 reduction in aqueous solutions were fabricated by depositing Sn on graphite felt. CO2 was reduced electrochemically in aqueous solutions of 1 M NaClO4 + 0.5 M NaOH saturated with CO2 to pH ca. 7.8 and circulated though the tinned graphite felt cathodes to enhance mass transport rates. Gaseous product bubbles, presumably of CO and H2, generated within the three-dimensional cathodes decreased effective conductivities of the interstitial electrolyte solutions, increasing potential drops in the ionically conducting phase. This caused reaction rates to decay within the cathodes in the direction of current flow. Hence, the effect of increasing solution flow rate was not only to enhance transport rates, but also to decrease gas fractions within the cathode, increasing effective conductivities of the electrolyte solution and so changing the spatial distribution of potential and current density. An optimal superficial current density and charge yield of 971 A m−2 and 0.58, respectively, were obtained at −1.62 V (AgCl∣Ag) and 99 ml min−1 solution flow rate; current densities were increased by a factor of 27 compared with the behaviour of a 2D electrode. A one-dimensional mathematical model was developed that was able to predict with adequate accuracy the effects of electrode potential and electrolyte solution flow rates on cross-sectional current densities, charge yields, and potential drops within the three-dimensional cathodes in the direction of current flow.

Introduction

When powered by renewable energy sources, electrochemical reduction of CO2 in aqueous solutions is a promising clean process for producing energy-rich chemicals/fuels. Methane and ethylene can be generated at Cu electrodes with significant charge yields at room temperature [1] and a large set of minor hydrocarbon products, not detected hitherto, was reported by Kuhl et al. [2]. Formate formation is predominant at Hg [3], [4], [5], Pb [6], Sn [7], [8], [9], and recently Ga2O3 [10]. CO2 reduction to CO predominates at Ag [11], [12], Au [12], and Au nanoparticle [13] cathodes. Methanol can also be produced, but appreciable charge yields are obtained only at small overpotentials and current densities (ca. 10 μA cm−2) [2], [14], [15], [16].

The mechanism of reduction of aqueous CO2 appears to involve several pathways, and is the subject of on-going research. Recently, we proposed a reaction scheme for CO:formate formation ratios, which also depended on solution pH [17]. The formation of hydrocarbons notably at Cu cathodes is believed to result from reduction of CO in a subsequent step; complex overall reaction pathways were proposed [18], [19]. Experimental parameters influencing the product distributions of CO2 reduction have been reported: local pH [20], supporting electrolyte cations [21], crystal surfaces [22], and cathode morphologies [23]. In addition, deposition of trace metals (Fe2+ and/or Zn2+) present in reagents appeared to poison cathodes and hence suppressed CO2 reduction charge yields [24]. Tin cathodes also undergo specific cathodic deactivation mechanisms involving the formation of distinct intermetallic compounds with alkali metal ions (Na+ and/or K+) in electrolyte solutions and the formation of tin hydride [25].

Apart from literature on electrode kinetics, alternative electrode designs with high volumetric surface areas to increase current densities have also been reported. Packed bed electrodes of Pb [26] and Sn [7], [8], [9] were fabricated for CO2 reduction; the latter offered formate charge yields above 60% at superficial current densities as large as 3 kA m−2. Pb/In [27] and Ag [28] powders bonded to ionic membranes to form membrane-electrode assemblies (MEA) were used with catholytes and anolytes flowing across respective surfaces of MEAs. MEAs can also function with only gaseous mixtures on the cathode sides [29]. Another variant of MEAs have been reported by electroless plating of metals on ionic membranes: Au [30], Cu [31], [32] and Ag [33]. Porous layers of PTFE-bonded catalysts (Pb, In, and Sn) fabricated without any ion-permeable membrane can operate as gas-diffusion electrodes (GDE) for CO2 reduction [34].

Graphite felt, a material with high volumetric surface areas, chemical resistance and moderate electrical conductivity, has hitherto not been used as a cathode for CO2 reduction. The material could serve as an inert matrix on which a metal could be deposited to introduce catalytic activity for CO2 reduction and removed to renew the surface after severe poisoning. We describe below the fabrication and performance for CO2 reduction to formate of a novel three-dimensional cathode made of tinned graphite felt, the large volumetric surface area of which increased current densities greatly compared to its planar counterpart. A mathematical model predicting the effects of electrode potential and electrolyte solution flow rates on current densities, charge yields and potential drops within the three-dimensional cathodes along the flow direction of superficial current densities was also developed and showed adequate accuracy for design purposes.

The potentials of the electronically and the ionically conducting phases within a three-dimensional electrode are related to the superficial current densities, js, in their respective phases through their superficial conductivities, κs, according to:κsϕ=-js

The superficial current densities of both phases in equation E1 are related to the true cathodic current density, j, at tinned surfaces of graphite fibres by a charge balance at steady state:js=aj

The electrode potential determining the true current density also varies along the flow direction of superficial current densities, so leading to varying reaction rates across felt thickness.

Significant variations of potential in both phases are expected if both phases have comparable superficial conductivities. In this case, the configurations of current collectors relative to electrode matrices strongly affect the distribution of current density and potential, and hence the charge yields of electrochemical reactions occurring at ionically- | electronically-conducting phase interfaces [35]. However, as metals are exceedingly more conductive than aqueous solutions, only slight variation in potential in electronically-conducting phase is expected in consolidated 3D electrodes, especially for low current densities operation, rendering electrode configurations insignificant. Hence, a mathematical model of the spatial distribution of current densities could be simplified greatly if the potential in the electronically conducting phase is assumed to be constant, as shown schematically in Fig. 1 for a 3D cathode. The potential of the electrolyte was expected to decrease from the front to the back of the 3D electrode with diminishing rates of decrease, eventually having a zero gradient at the back of the electrode due to the superficial current density being zero; hence, the boundary condition: ϕelectrolyte/x|x=0=0. The potential at the surface of a hypothetical probing reference electrode would exhibit similar decreases but maintain a constant difference from the potential of the electrolyte were it possible to move one freely inside the felt electrode.

Therefore, electrode potentials must be corrected for the position-dependent potentials of the ionically-conducting phase:E=(μe,mF+ϕm)(μe,refF+ϕref)=(μe,mF+ϕmϕelectrolyte)(μe,refF+ϕrefϕelectrolyte)={μe,mF+ϕm|x=0(ϕelectrolyte|x=0+Δϕ)}{μe,refF+ϕref|x=0ϕelectrolyte|x=0}=(μe,mF+ϕm|x=0)(μe,refF+ϕref|x=0)Δϕ=E|x=0ΔϕAs the electrode potential E|x=0 is a constant during each batch of potentiostatic electrolysis, Δϕ is also related to the superficial current density in a 3D electrode by an equation similar to Eq. E1:κs(Δϕ)=js

Trainham and Newman’s mathematical model of a porous flow-through electrode for metal ion removal [36] incorporated two mass transport phenomena: axial diffusion and mass transfer to reaction interfaces. The former appears as the gradient term in the flux equation of a species:Ni=Dci+υsci

Therefore, a material balance equation at steady states has the form:Ni=D2ci+υsci=ari

Eqs. E5 and E6 predict the concentrations of reactants and products to drop and to rise, respectively, as the electrolyte solution flows through a 3D electrode from the back to the front. As electrochemical reaction rates depend on the concentrations of reactants and products, the variations in the concentrations also contribute to varying reaction rates along the direction of current flow. Fig. 2 shows schematic profiles of the concentrations of a reactant and a product. It is worth pointing out that axial diffusion causes the concentrations of the reactants and the products to deviate from their respective feed concentrations even before the electrolyte solution enters the 3D electrode.

Mass transfer, another mass transport resistance, of a species can be quantified through a mass transfer coefficient and, at steady states, must equal its genaration or consumption at reaction interfaces:ri=km,icici'where ci represents the concentration of species i at reaction interfaces.

The two mass transport phenomena can greatly complicate a mathematical model for a 3D electrode but could be neglected under certain conditions; axial diffusion becomes insignificant at small single-pass reactant conversions and mass transfer resistance becomes negligible at sufficiently high flow rates. If these two conditions are met, flat distributions of reactant concentrations across the thickness of a 3D electrode can be assumed.

Section snippets

Equipment

The deposition of tin on graphite felt and all electrochemical measurements were carried out under ambient temperature and pressure in a two-compartment PVDF cell separated by a piece of Nafion 415 cation-permeable membrane (Fig. 3). The working electrode compartment consists of two blocks with an electrolyte inlet/outlet port on each, between which a felt cathode assembly was placed, enabling circulation of electrolyte solutions through the working electrode to enhance mass transport. An IrO2

Deposition of tin on graphite felt

Fig. 4 shows a SEM photomicrograph of tin-plated graphite felt fabricated by the method described in Section 2.2. In general, homogeneous tin coverage was achieved, but a significant amount of tin deposits had detached from the graphite fibres, probably when the plated felt was handled before and/or during the microscopy, as the tin coverage was generally more complete underneath.

Fig. 5 shows SEM micrographs of the surface morphology and a cross section of an individual fibre of the tinned

Model improvement

The kinetic model is the key component of this model but only approximations of hydrogen evolution kinetics were used. Furthermore, interactions amongst adsorbed reaction intermediates were not considered. The accuracy of the predictions will be improved significantly if a more accurate model for the kinetics of simultaneous CO2 reduction and hydrogen evolution is employed. However, such a model does not exist at the time of writing and its development would be extremely difficult.

The lack of

Conclusions

  • a)

    Graphite felt provided an electronically conducting matrix on which tin was electrodeposited to make three-dimensional cathodes for CO2 reduction, in parallel with hydrogen evolution from CO2-saturated aqueous solutions at near neutral pHs.

  • b)

    Accumulation of bubbles within the ionically conducting phase of a three-dimensional electrode decreased the superficial conductivity of the electrolyte solution and hence led to even less homogeneous spatial distributions of reaction rates in the direction

Acknowledgements

The authors thank the UK Engineering and Physical Sciences Research Council for a grant, and the Thai government for a studentship for P.B.

References (40)

  • M. Maeda et al.

    Reduction of carbon dioxide on partially-immersed Au plate electrode and Au-SPE electrode

    J Electroanal Chem

    (1987)
  • S. Komatsu et al.

    Preparation of cu-solid polymer electrolyte composite electrodes and application to gas-phase electrochemical reduction of CO2

    Electrochim. Acta

    (1995)
  • Y. Hori et al.

    Silver-coated ion exchange membrane electrode applied to electrochemical reduction of carbon dioxide

    Electrochim. Acta

    (2003)
  • B. Delanghe et al.

    Mass transfer to a carbon or graphite felt electrode

    Electrochim. Acta

    (1990)
  • K.P. Kuhl et al.

    New insights into the electrochemical reduction of carbon dioxide on metallic copper surfaces

    Energy Environ. Sci.

    (2012)
  • J. Ryu et al.

    Electrode reduction kinetics of carbon dioxide in aqueous solution

    J. Phys. Chem.

    (1972)
  • Y. Hori et al.

    Electrolytic reduction of bicarbonate ion at a mercury electrode

    J. Electrochem. Soc.

    (1983)
  • B. Innocent et al.

    Electro-reduction of carbon dioxide to formate on lead electrode in aqueous medium

    J. Appl. Electrochem.

    (2009)
  • H. Li et al.

    The Electro-Reduction of Carbon Dioxide in a Continuous Reactor

    J. Appl. Electrochem.

    (2005)
  • H. Li et al.

    Development of a continuous reactor for the electro-reduction of carbon dioxide to formate–Part 1: Process variables

    J. Appl. Electrochem.

    (2006)
  • Cited by (25)

    • Bioelectrochemical cells as a green energy source for electrochemical treatment of water and wastewater

      2022, Journal of Water Process Engineering
      Citation Excerpt :

      Porous materials have been suggested as cathodes for MECs to generate more H2 as the porous materials provide large surface area for enhance cathodic reaction. This solution may work on laboratory scale, but it can't be used for very high current density [63]. At high current density, bubbles cover the surface of the interstitial electrolyte, which reduces the effectual conductivity and then coalesces in the pores, obstructing the electrode performance.

    • Design of membraneless gas-evolving flow-through porous electrodes

      2021, Journal of Power Sources
      Citation Excerpt :

      This configuration was already studied for water electrolysis in, e.g. Refs. [13,30,31] but only recently its potential for membraneless operation has been highlighted in Refs. [14,15,32,33]. Many studies have been devoted to modelling flow-through electrodes [19,20,34,35] focusing mostly on mass transfer [31,36,37] while far fewer include the effect of gas evolution [38–42]. All these models assume that distinct bubbles move with a velocity proportional to the liquid velocity.

    • Microbial electrolysis cell (MEC): Strengths, weaknesses and research needs from electrochemical engineering standpoint

      2020, Applied Energy
      Citation Excerpt :

      This solution may be successful at the laboratory scale, but cannot be adopted for high current density. At high current density, the bubble coverage (the fraction of the electrode surface area covered by adhering bubbles [163]) would decrease the effective conductivity of the interstitial electrolyte [164] and finally coalesce in the porous structure, blocking the electrode. Using a porous electrode to evolve gas is generally not an industrial option, unless the electrolyte is forced through it [164].

    View all citing articles on Scopus
    View full text