Advantages of CO over CO₂ as reactant for electrochemical reduction to ethylene, ethanol and n-propanol on gas diffusion electrodes at high current densities

Highlights

• Straight comparison of CO and CO₂ electrochemical reduction towards C2 and C3 products achieved using a flow cell and commercial Cu-powders deposited on gas diffusion layers.

• Higher current densities for C2 and C3 products at lower working electrode potentials with CO as the reactant, rather than with CO₂.

• High selectivity for ethylene, ethanol, and propanol using CO as the reactant with a cumulative Faradaic efficiency of 89% at −300 mA cm^-2.

• 20 hours of ethylene production with ∼ 44% Faradaic efficiency at −200 mA cm^-2.

Abstract

The electrochemical conversion of CO₂ to value-added chemicals is a technology gaining broader interest as society moves towards a carbon-neutral circular economy. Nonetheless, there are still several challenges to overcome before this technology can be applied as an industrial process. In the reaction path of the electrochemical reduction of CO₂ with Cu as an electrocatalyst, it is known that carbon monoxide is the key intermediate to chemicals such as ethylene, ethanol, and n-propanol. However, a better understanding of the electrochemical reduction of CO is still necessary to improve selectivity and efficiency at high current densities. In this work, the electrochemical reduction of CO₂ and CO towards C2 and C3 products is investigated using gas diffusion electrodes in a flow cell. Thereby the electrochemical reaction is not limited by the solubility of the feed gas in the electrolyte, and current densities of industrial relevance can be achieved. The electrodes are prepared using commercial Cu-powders consisting either of nano- or microparticles that are deposited on gas diffusion layers. Potentiostatic experiments show that with CO as the reactant, higher current densities for C2 and C3 products can be achieved at lower working electrode potentials compared to CO₂ as the reactant. Galvanostatic CO electrochemical reduction at −300 mA cm⁻² with Cu-nanoparticles (40–60 nm) results in a cumulative Faradaic efficiency of 89% for C2 and C3 products. This represents a two-fold increase in selectivity to ethylene and a three-fold increase towards ethanol and n-propanol compared to the selectivity obtained with CO₂ as the reactant. This enhancement of selectivity for C2 and C3 products at current densities of industrial relevance with CO as reactant provides a new perspective regarding a two-step electrochemical reduction of CO₂.

Introduction

The electrochemical reduction of CO₂ is an opportunity to support a low carbon economy by producing value-added chemicals out of CO₂ while utilizing low-cost renewable energy [1,2]. For decades the combustion of fossil fuels has been the least expensive means of electricity production, resulting in emission of carbon dioxide [1], a compound which has been shown to be responsible for climate change [3]. Over the last few years the cost of renewable electricity has decreased drastically and currently tends to be competitive with traditional electricity production technologies, and soon will become the least expensive source of energy [4]. However, the challenges of intermittency of energy production and grid balancing remain [1]. Within this context, electrochemical conversion of CO₂ to fuels and feedstocks using aqueous electrolytes and renewable energy is a very attractive option to close the carbon cycle [1,2,[5], [6], [7]]. In order to make that possible, high conversion at high current densities and low overpotentials, together with a selective and long term stable operation have to be achieved [2,6,8,9].

Different products can be obtained from the electrochemical reduction of CO₂ by using specific transition metals as electrocatalysts. Depending on the ability of each metal to adsorb CO, known as the main intermediate in the CO₂ reduction reaction, different products can be generated [5,10]. These electrocatalysts have been extensively studied in the last decade focusing mainly on three approaches: the first being the electroreduction of CO₂ to syngas, i.e. a mixture of CO and H₂, using Ag and Au electrodes, which is getting closer to practical applications [[11], [12], [13]]; the second is the conversion of CO₂ to formate on Sn electrodes with Faradaic efficiencies (FE) over 70% at current densities between −100 mA cm⁻² and −300 mA cm⁻² [14,15]; and the third and most challenging approach is the reduction of CO₂ to C2 and C3 hydrocarbons such as ethylene, ethanol and n-propanol, which occurs when using Cu as electrocatalyst [[16], [17], [18], [19]].

Taking into account that it is possible to electroreduce CO₂ into more than ten products with Cu as an electrocatalyst [16,20,21], there are still a number of opportunities to explore, especially regarding the ability of copper to convert carbon monoxide into C2 and C3 products [22]. Although poor selectivity and stability, together with high overpotentials [16], seem to be the main challenges, significant improvements have been made reaching lower activation overpotentials and high Faradaic efficiencies to ethylene and ethanol [17,18,[23], [24], [25], [26], [27]]. Several studies regarding Cu catalyst structure [17,28,29] and Cu oxidation states [3,16,17] have been performed. Recently the introduction of gas diffusion electrodes with nanostructured surfaces on carbon gas diffusion layers have enabled process operation at current densities around −150 mA cm⁻² in flow cells, improving selectivity and activity towards ethylene [17,18,24,30]. To our knowledge, the best results up to now have been achieved under alkaline conditions using flow cells [18,24]. However, the formation of carbonates out of CO₂ in an alkaline medium (equations (1), (2)) [5,31] can lead to complications for up-scaling, since the non desired reaction rate between CO₂ and the electrolyte would be higher than the reaction rate of the desired electrochemical reactions [32]. The use of CO instead of CO₂ as the reactant in alkaline conditions should not represent drawbacks since CO does not react with aqueous electrolytes.

$${\text{CO}}_2+{\text{OH}}^{-}\rightleftharpoons {\text{HCO}}_3^{-}$$

$${\text{HCO}}_3^{-}+{\text{OH}}^{-}\rightleftharpoons {\text{CO}}_3^{2-}{\text{ + H}}_2\text{O}$$

Within this context, a two-step electrochemical reduction of CO₂ has been proposed as an alternative to increase the selectivity towards ethanol and acetate, in which CO₂ is reduced to CO as the first step, followed by CO reduction to hydrocarbons [22,33,34]. However, the CO electroreduction has been mainly investigated in setups where the electrochemical reaction is limited by the solubility of CO in aqueous electrolytes, reaching geometrical partial current densities for hydrocarbons lower than −1 mA cm⁻² [10,20,22,35,36]. There is scarce research on CO electrolysis performed in flow cells with gas diffusion electrodes [21,32]. Recently, Han et al. [32] reported CO electrochemical reduction using Cu-nanoparticles on gas diffusion electrodes with a maximal FE for ethylene of 17.8% and a partial current density for ethylene of −50.8 mA cm⁻². The formation of liquid products in their case has not been addressed. Furthermore, Schmid et al. [21], as part of their investigation of the CO₂ electrochemical reduction mechanisms at high current densities, performed galvanostatic CO bulk electrolysis at a current density of −170 mA cm⁻² using an in situ-grown copper nano-dendritic catalyst on a gas diffusion layer. They reported an increment in the selectivity towards acetate and ethanol, but no significant increase in the ethylene Faradaic efficiency compared to the one achieved with CO₂ as reactant, which was around 40% FE. Nevertheless, the nanodendritic Cu electrode decreases its activity dramatically after the first hour of experiment. With the progress made regarding Cu nanostructured-GDEs [24], flow cells [9] and the high Faradic efficiencies achieved in the conversion of CO₂ to CO at high current densities [11], a two-step electrochemical reduction as depicted in Fig. 1 becomes certainly worthy of further exploration.

In order to get some insights about the feasibility of a two-step electrochemical reduction of CO₂, we compare CO₂ and CO as reactants regarding selectivity towards C2 and C3 products using gas diffusion electrodes, a flow cell, and a mild electrolyte (KHCO₃). This allows a straight comparison of the electrochemical reduction of CO and CO₂. The overall cathodic half reactions using either CO₂ or CO as the reactant are shown in Table 1. For each reactant we studied four commercial Cu-powders with different particle sizes in the nano- and micro-scale, aiming to obtain further knowledge about the effect of morphology on the electrochemical activity. These Cu-powders were deposited onto carbon based gas diffusion layers, obtaining gas diffusion electrodes. We also investigated the stability of the system by performing 20 h electrolysis. Additionally, the catalyst layers were characterized by TEM and XRD before and after electrolysis. The obtained results show a remarkable increase in the selectivity towards ethylene (C₂H₄), ethanol (C₂H₅OH), and n-propanol (C₃H₇OH) using CO as the reactant instead of CO₂.