Efficient electroreduction of CO₂ to CO by Ag-decorated S-doped g-C₃N₄/CNT nanocomposites at industrial scale current density

Highlights

• C₃N₄-based nanomaterial has been developed as an efficient catalyst towards electrochemical CO₂ reduction reaction.

• Systematic studies were carried out to understand the catalytic mechanism of the C₃N₄-derivates.

• The best electrochemical CO₂ reduction performance among the C₃N₄-based materials was achieved

Abstract

In recent years, the application of graphitic carbon nitride (g-C₃N₄) for electrochemical CO₂ reduction reaction (eCO₂RR) has aroused strong interest. However, this material is still facing severe activity issue towards eCO₂RR so far, and studies on its catalytic mechanism have not been sufficiently implemented either. Herein, we report an Ag-decorated sulfur-doped graphitic carbon nitride/carbon nanotube nanocomposites (Ag–S–C₃N₄/CNT) for efficient eCO₂RR to carbon monoxide (CO). The resulting Ag–S–C₃N₄/CNT catalyst exhibits a notable performance in eCO₂RR, yielding a high current density of −21.3 mA/cm² at −0.77 V_RHE and maximum CO Faradaic efficiency over 90% in H-type cell. Strikingly, when combining with flow cell configuration, the fabricated nanocomposites permit an industrial scale and cost-effective eCO₂RR, showing a current density larger than 200 mA/cm² and the Faradaic efficiency of CO over 80% in a wide potential window, delivering the best eCO₂RR performance among the C₃N₄-derivatives. Moreover, the catalytic mechanism of this nanocomposite

has been further explored through density functional theory (DFT) and electrochemical methods carefully. Our work not only sheds light on industrial scale eCO₂RR to CO but also leads to new insights on the application of C₃N₄-based composite materials in electrocatalytic processes.

Introduction

The concentration of atmospheric carbon dioxide (CO₂) has continued to rise sharply from 260 ppm to more than the deadly 400 ppm mark since the first industrial revolution [1,2]. The accumulation of the atmospheric CO₂ is deemed to be the culprit of many environmental problems, such as global warming, and erratic weather pattern [3,4]. To alleviate these climate challenges, the electrochemical CO₂ reduction reaction (eCO₂RR) for the production of value-added products (e.g. carbon monoxide [CO], methane [CH₄]), through solar and wind-derived renewable electricity, provides a near-perfect solution to curb CO₂ emission while producing useful chemicals and storing energy [5,6]. However, despite certain breakthroughs having been made in eCO₂RR, the practical applications are still limited by the low production rate, poor selectivity, and high cost of catalysts.

The eCO₂RR current density (j) signifies the production rate and activity of the catalysts. Considering the techno-economic factors, a gross margin model proposed in 2016 has pointed out the importance of high eCO₂RR current density (j > 200 mA/cm²) for practical applications [[7], [8], [9]]. However, in addition to the intrinsic activity of the catalysts, an overwhelming percentage (>95%) of current researches on eCO₂RR to CO are conducted in H-type cell [8], where the low CO₂ solubility in aqueous solution restricts the j value of eCO₂RR to ~35 mA/cm² [[10], [11], [12], [13]], and few of catalysts have demonstrated the capability for industrial-rate production [14]. Additionally, the liquid reaction environment in H-type cells also facilitates the competitive hydrogen evolution reaction (HER) against the eCO₂RR [14]. To overcome these limitations and achieve an industrial scale current density, the flow cell architecture has been introduced to the eCO₂RR research field recently because of its independency from CO₂ solubility and unique triple-phase boundary formed at the catalyst–electrolyte interfaces [15]. However, although the utilization of flow cells is widely investigated in fuel cells and water electrolysis, the studies of applying this technique in eCO₂RR are still limited so far [12].

In addition to the reaction rate issue, the eCO₂RR is a complicated process with a wide variety of gas (e.g. CO, methane [CH₄], ethylene [C₂H₄], etc.) and liquid (e.g. ethanol, formic acid, n-propanol, etc.) products [[16], [17], [18], [19], [20]]. Among the various products, CO is one of the most valuable products and the conversion of CO₂ to CO holds several unique advantages. First, CO is generally more selective and convenient to be separated from the electrolyte [21]. Second, since the direct conversion of CO₂ into multicarbon (C₂₊) products is not effective, an alternative two-step strategy, where CO₂ is first reduced to CO and sequentially reduced to C₂₊ has been demonstrated to be productive [19,22]. Third, the economic cost analysis evidences that CO owns the best market compatibility and the highest net present value [23]. However, satisfactory selectivity requires suitable binding strength with the reaction intermediates, which relies on the well-designed composition and structure of the catalysts [24].

In the past few decades, carbon-based nanomaterials have been widely developed as electrocatalysts [[25], [26], [27]]. Among the various carbon materials, the low-dimensional graphitic carbon nitride (g-C₃N₄) consisting of a prototypical 2D graphitic structure, has shown promising performance in the recently reported electrocatalytic processes [28,29]. Besides its low cost, stability, easy synthesizability, and environmentally friendly merits, the g-C₃N₄ could serve as an ideal molecular scaffold for eCO₂RR because of several precious advantages: (1) the abundant pyridinic nitrogen atoms incorporated in g-C₃N₄ can lead a strong CO₂ affinity, and the selective adsorption of CO₂ is helpful for accelerating the eCO₂RR and competing with other competitive reactions [30,31]. (2) The carbon atoms in g-C₃N₄ exhibit a high oxophilicity, which is beneficial for adsorbing the complex oxygenated intermediates (e.g. *COOH, *CO) during multiple eCO₂RR steps [32]. Nevertheless, to date, the application of g-C₃N₄ derivatives in eCO₂RR has been rarely reported as it is restricted by intrinsic poor conductivity and limited exposed active sites, and the eCO₂RR mechanism on g-C₃N₄ still has not been fully investigated. Some strategies have been proposed to overcome the innate disadvantages of g-C₃N₄. Heteroatoms (like sulfur, boron, etc.) doping has been demonstrated as an effective approach in narrowing the g-C₃N₄ band gap and increasing its conductivity to tune the electrical properties of g-C₃N₄ for oxygen reduction reaction (ORR) and hydrogen evolution reaction (HER) [33,34]. In addition, R. Amal et al. have combined multiwall carbon nanotubes (MWCNTs) with g-C₃N₄ to improve the conductivity for eCO₂RR, but the current density is still lower than −8 mA/cm² at −0.8 V_RHE and maximum Faradaic efficiency for CO (FE(CO)) is only 60% [35]. X.H. Li and J.S. Chen have synthesized a 2D polarized g-C₃N₄ with a large reactive area for CO production, the catalyst achieves a maximum Faradaic efficiency of CO [FE(CO)] around 80% at −1.1 V_Ag/AgCl, whereas the reaction current density is less than −5 mA/cm2 [36]. Besides these works, although ternary Au-carbon dots-C₃N₄ [37] and Cu decorated C₃N₄ [31] have been developed toward eCO₂RR, the activity and selectivity issues still haven't been well addressed.

In this work, we synthesized Ag nanoparticles-decorated sulfur-doped C₃N₄/CNT nanocomposites (Ag–S–C₃N₄/CNT) with excellent activity and selectivity towards eCO₂RR to CO. When measuring in the traditional H-type cell, the Ag–S–C₃N₄/CNT showed a remarkable high current density of −21.3 mA/cm² at −0.77 V_RHE and the maximum FE (CO) of 91.4 ± 0.01% at −0.8 V_RHE. Moreover, to achieve an industrial scale and cost-effective CO production, the Ag–S–C₃N₄/CNT was coupled with flow cell configuration, demonstrating a remarkable maximum current density of 330 mA/cm² at a cell voltage of 3 V with a 93% maximum FE (CO) at 2.8 V. Our synthesized Ag–S–C₃N₄/CNT represents the best performance of eCO₂RR to CO with great scalability and stability among all the reported C₃N₄-derivatives.