Upcycling waste protein and heavy metal into single-atom catalytic gas diffusion electrode for CO₂ reduction

Abstract

The global production of organic wastes and heavy metals (HMs) poses significant environmental risks, along with considerable carbon emissions from waste decomposition. This highlights the significance of synergistic management of both wastes and CO₂, which is a vital strategy for mitigating environmental pollution and climate change. Herein, we employed waste protein from wastewater produced during soybean peptide (SP) processing as a carbon matrix to anchor HMs Ni from electroplating wastewater. This mixture was electrospun into a gas diffusion electrode (GDE). This unique GDE design eliminates the need for a separate gas diffusion layer (GDL) and simplifies catalyst production. This versatile GDE consists of nanofibers with uniformly dispersed Ni single atom catalysts (SACs) on the fiber surface. Therefore, boasts a porous structure that facilitates CO₂ diffusion and storage. The homogeneous distribution of Ni SACs within the GDE fosters high activity in the electrochemical conversion of CO₂ to CO. At 50 mA/cm² and 2.5 V cell voltage, Ni SACs achieved an excellent Faradaic efficiency of 81%−98% in a membrane electrode assembly (MEA). This technique holds a promise in achieving the collaborative management of carbon mitigation and wastes recovery.

Graphical abstract

Keywords

Waste protein / Heavy metal wastewater / Ni single atom / Electrospinning / Gas diffusion electrode / CO ₂ reduction

Highlight

	● The utilization of waste proteins and heavy metals for CO2 to CO conversion.
	● Proteins atomically disperse Ni onto a gas diffusion electrode (GDE).
	● Porous structure of GDE is capable of CO2 diffusion and storage.
	● N element is crucial to synthesize efficient Ni single atom catalysts (Ni SACs).
	● GDE is composed of nanofibers with uniformly dispersed Ni SACs.

1 Introduction

The pursuit of carbon neutrality in numerous countries has facilitated extensive research on carbon mitigation strategies in energy-intensive industries (Yuan et al., 2024). However, CO₂ emissions from organic wastes and wastewater have been understudied (Jones, 2013; Lu et al., 2018; Voosen, 2020; Reichstein et al., 2021; Xu et al., 2021; Liu et al., 2024). Studies have shown that organic waste and wastewater, particularly from agriculture and food industries, account for over 7% of global CO₂-equivalent emissions (Gustavsson et al., 2011; Mohareb et al., 2018; Yang et al., 2024). The electrochemical CO₂ reduction reaction (ECO₂RR) has recently exhibited significant potential for mitigating CO₂ emissions and producing valuable carbon-based chemicals (Sharifian et al., 2021). Therefore, directly converting waste-derived CO₂ into value-added products by ECO₂RR holds great promises for sustainable development. Beside the organic matters, heavy metals (HMs) may exist in the wastes, which is toxic but also exhibits catalyzing nature for chemical reactions (Gu et al., 2024). Here, we suppose whether soybean peptide (SP) processing wastewater and HMs derived from electroplating wastewater as the precursors of ECO₂RR catalysts could be feasible for simultaneous CO₂ mitigation and wastes treatment. We propose employing SP to fix HMs Ni in the wastewater, then the formed mixture was used to fabricate a gas diffusion electrode (GDE) with in situ growth of Ni single atoms (SAs) catalyst for low-cost CO₂-CO electrochemical conversion.

Developing efficient and low-cost electrocatalytic materials for converting CO₂ to CO rather than liquid products remains a challenge (Xia et al., 2019; Fan et al., 2020; Peng et al., 2021; Wang et al. 2024; Tan et al., 2025). Current research found that earth-abundant Ni single atom catalysts (SACs) can maximize the efficiency of catalytic sites to facilitate the conversion of CO₂ to CO (Jiang et al., 2018; Zheng et al., 2019; Li et al., 2020; Yang et al., 2020). Ni SACs are typically synthesized using Ni-contained salt and expensive carbon matrix, such as metal organic frameworks (MOFs) (Li et al., 2020; Yang et al., 2020; Sun et al., 2025), carbon nanotubes (CNTs) (Liang et al., 2021), and graphene (Jiang et al., 2018), etc. therefore, utilizing affordable carbon matrices is necessary for facilitating the large-scale production of Ni SACs. Previous studies have indicated the importance of N-element doping to enhance the Ni SACs performance, since the formed Ni-N_x during carbonization was proved to resist metals leaching (Liang et al., 2021) and improve the detachment of absorbed CO (CO*) due to the decreased binding energy (Liu et al., 2021; Qin et al. 2025). Consequently, we proposed the use of proteins, a widely available N-enriched matter, to synthesize SACs. In addition, proteins can effectively bind HMs on a uniform and ultrasmall scale through interactions such as chelation between functional groups and Ni atoms, electrostatic attraction, and van der Waals forces (Blundell and Jenkins, 1977; Ge et al., 2012). Moreover, the total amount of low-cost protein is considerable given the tremendous protein-enriched wastes produced annually, such as agriculture and food wastes. For example, Food and Agriculture Organization of the United Nations (FAO) statistics show that more than 3 billion tons of various food are wasted annually, corresponding to more than 0.24 billion tons of waste protein produced based on the protein contents in 7 main kinds of foods (Table S3, Table S) (Gustavsson et al., 2011). Therefore, there is a great potential to utilize the waste protein as carbon matrix to synthesize highly efficient catalysts for CO₂ conversion.

ECO₂RR is generally conducted in bulk electrolyte solution using a single chamber or H-type electrolysis cell (Figs. S1(a) and (b)), but low solubility (0.033 mol/L) and low diffusion (0.0016 mm²/s) of CO₂ molecules in a bulk solution well confine ECO₂RR kinetics to a low level (Verma et al., 2016; Weekes et al., 2018; Nguyen and Dinh, 2020). Despite using HCO₃^–, more soluble and more diffusional in solution, as the feedstock to electro-conversion toward CO and formate have been reported (Li et al., 2019; Sun et al. 2024), the energy penalty and catalysts’ cost were strikingly higher than the counterpart system of CO₂ as the feedstock. Feedstock for ECO₂RR remains preference toward CO₂ molecules. The resultant mass transfer limitation in bulk solution inflicts on current uplift to a commercially viable operation. Delivering humidified CO₂ gaseous, instead, for ECO₂RR serves as a solution against mass transfer limitation. Hike in (0.041 mol/L) and diffusion (16 mm²/s) drives current density up to several hundred mA/cm², making a step toward large-scale application of ECO₂RR. (Weekes et al., 2018). Adaptative to humidified CO₂ gaseous feeding necessitates conversion from a traditional H-type or single chamber electrolysis cell to a membrane electrode assembly (MEA) (Figs. S1(c) and S1(d)). MEA is a contacted electrolysis cell where cathode, ion exchange and anode are clung, leaving zero gap within them. The linchpin in MEA is the gas diffusion electrode (GDE), providing a thin interface for gaseous CO₂ pass through while preventing bulk solution penetration. Traditionally, powdered catalysts are slathered onto a gas diffusion layer (GDL) to manufacture a GDE, yet catalysts cannot resist to be stripped away via continuous electrolyte flourish, inflicting on GDE duration.

In this study, we originally used wasted protein to capture Ni in the real wastewater and directly converted the organic-HMs mixture into a Ni-SACs distributed gas diffusion electrode (GDE) through electrospinning coupled with carbonization. This tactic in situ synthesized Ni SACs onto the surface of nanofibers of GDE, avoiding tedious processes of fabricating GDL and catalyst separately to achieve an overwhelming stability and selectivity toward CO. The satisfactory CO selectivity remained no matter conducting ECO₂RR in a single-chamber or in a MEA. Despite the progress in the development of biomass-derived catalysts has been made, co-management of organic wastes, HMs and CO₂ emission have rarely been reported. Using herein static simultaneously reclaimed protein and HMs from wastes or wastewater, and driven overdue atmospheric CO₂ into non-greenhouse while value-added gas, pushing toward a closed carbon loop or even negative emission in wastes treatment as Fig.1(a) showed.

2 Materials and methods

2.1 Source of the wastes

The wastewater produced from a soybean peptide (SP) processing factory in Shenzhen, China was selected as the source of waste protein to synthesize the GDE. The SP wastewater had a total soluble protein concentration of around 10000 mg/L (as measured by bicinchoninic acid assay) (Table S4). Heavy metals (HMs) contained wastewater was obtained from an electroplating factory in Meizhou, China with Ni (around 183 mg/L) as the dominant metal (Table S5).

2.2 Preparation of waste-derived GDE

The electrospinning technique was used to manufacture GDE with ultrafine nanofibers and ultrahigh porosity for gas diffusion and storage (Yang et al., 2019) (Fig. S2). SP processing wastewater was mixed with HMs-containing wastewater at a protein to Ni mass ratio of 50:1. A 12 wt% polyacrylonitrile (PAN) in solution was used as the flexible support material. These two solutions were simultaneously electrospun into fiber using 18−20 kV voltage applied to the syringe needle, while the collector was charged with −0.05 to −4 kV voltage. The syringe feed rate for both the PAN and SP solutions was set at 1 mL/h. After 45−50 h of operation, the synthesized membrane underwent peroxidation in a muffle furnace at 260 °C for 2 h and then cooled to room temperature. The following carbonization was conducted at 1000 °C for 2 h under argon atmosphere protection. The synthesized GDE was labeled as SP_GDE. For a comparison, we synthesized GDE using only PAN solution and Ni addition according to the same procedures. The resulting material was labeled as NP_GDE. To enhance the surface hydrophobicity of SP_GDE, 5 wt% Nafion 117 solution was drop-cast on SP_GDE (Yang et al., 2020).

2.3 Electrochemical measurements

To deeply examine the catalytic performance of GDE, we first ground GDEs into powder and then drop-cast the powder onto a carbon paper (Tory TGP-H-060). ECO₂RR was conducted in a single chamber electrolysis cell (Fig. S1(a)) using a 0.5 mol/L KHCO₃ solution as the electrolyte. An Ag/AgCl electrode was used as the reference electrode, and a platinum plate served as the counter electrode. The potential was expressed relative to the reversible hydrogen electrode (vs. RHE) using the following equation:

(1)\mathit{E} \mathrm{v}\mathrm{s}.\mathrm{R}\mathrm{H}\mathrm{E}=\mathit{E}\mathrm{v}\mathrm{s}.\mathrm{A}\mathrm{g}/\mathrm{A}\mathrm{g}\mathrm{C}\mathrm{l}+0.197+0.059\times \mathrm{p}\mathrm{H}

ECO₂RR at high current density was conducted in a MEA that is consist of a GDE, an oxygen evolution electrode (IrO₂-coated titanium mesh, more details were provided in Supplementary Materials), flow plates, and a Nafion 117 membrane (Fig. S1(d)). The CO₂ flow was controlled by a mass flow meter (MC-100SCCM-D-DB9M, Alicat, USA) at 20 or 50 standard cubic centimeter per minute (sccm). The electrolyte flow was controlled by a peristaltic pump (BT100-2J, Dichuang, China) at 50 mL/min. All electrochemical measurements were conducted using a potentiostat (1010E, Gamry, USA).

2.4 Analysis and calculations

Gas products of CO₂RR were detected by a gas chromatograph (GC) (TRACE 1310, Thermo, USA) equipped with a thermal conductivity (TCD). The partial current density of individual products was calculated by the following formula:

(2)j_i=x_i\times v \times z_i\times F\times P_0\times \mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{e}\mathrm{a}/RT,

where x_i represents the volume fraction of a given gas product i, v is the flow rate, z_i is the number of transferred electrons, F is the Faraday constant, P₀ is 101.325 kPa, R is the gas constant, and T is the temperature. The corresponding Faradaic Efficiency (FE) was calculated as follows:

(3)\mathit{F}\mathit{E}=j_i/j_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}\times 100\mathrm{\%},

the turnover frequency (TOFs, h⁻¹) for ECO₂RR products was evaluated based on the electrochemical active surface area (ECSA) normalization. Details in calculation can be found in Supplementary Materials (Jiang et al., 2018):

(4)\mathrm{T}\mathrm{O}\mathrm{F}\mathrm{s}=(j_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}\times t\times {\mathit{F}\mathit{E}}_{\mathrm{C}\mathrm{O}})/(2F\times \mathrm{t}\mathrm{h}\mathrm{e}\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{N}\mathrm{i}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{e}\mathrm{s}).

Energy efficiency can be calculated as the follows:

(5)\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}=(E^0\times {\mathit{F}\mathit{E}}_{\mathrm{C}\mathrm{O}})/{\mathit{E}}_{\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}},

where E⁰ = E_{{\mathrm{O}}_2}−E_CO = 1.23−(−0.11) = 1.34 V, with E_{{\mathrm{O}}_2} and E_CO representing the O₂/H₂O and CO₂/CO equilibrium potentials, respectively.

The potential protein amount that can be extracted from 7 kinds of agriculture and food waste was calculated based on the statistics of Food and Agriculture Organization of the United Nations (FAO) (Tables S2 to S3) (Gustavsson et al., 2011).

2.5 GDE characterizations

The morphology and structure of GDE were characterized via scanning electron microscopy (SEM, TESCAN MIRA4), transmission electron microscopy (TEM, FEI Talos F200X), aberration-corrected scanning transmission electron microscopy (STEM, Titan Cubed Themis G2300) and X-ray diffraction (XRD, RIGAKU Ultima IV). Chemical bonds and compositions were provided by Fourier Transform Infrared Spectrometer (FTIR, Scientific Nicolet iS20) and X-ray photoelectron spectroscopy (XPS, Scientific K-Alpha). The contact angle of GDE surface was observed by a contact angle testing analyzer (KRUSS DSA25). Porous structure, specific surface area and CO₂ adsorption capacity were examined by a BET analyzer (Micrometrics ASAP 2460) using N₂ and CO₂. Surface anti-wettability was detected by a Confocal Laser Scanning Microscope (Nikon AX-SHR). 1 mol/L KOH solution with dilute Rhodamine B served as the color reagent and a 488 nm laser was selected as the excitation source. Heavy metals concentration was detected by an inductively coupled plasma mass spectrometry (NexION 1000, PerkinElmer). Scanning electrochemical microscopy (Princeton Applied Research) was operated by applying GDE with −1.24 V vs. RHE and microelectrode with 0.2 V vs. RHE to consume evolved CO.

3 Results and discussion

3.1 Selectivity toward CO₂ to CO electrochemical conversion

The GC analysis confirmed that CO and H₂ were the predominant products generated during ECO₂RR using GDEs as catalysts. The absence of liquid products was further verified by ¹H NMR spectroscopy results (Fig. S3). The performance of SP_GDE in ECO₂RR was examined across a potential from −0.45 to −1.1 V vs. RHE (Fig.1(b)). Notably, SP_GDE exhibited excellent selectivity, achieving FE_CO exceeding 80% across a broad potential range from −0.6 to −1.0 V vs. RHE, underscoring the superior CO₂ conversion efficiency of SP_GDE compared with NP_GDE. The maximum FE_CO is 96% at −0.8 V vs. RHE, while corresponding {\mathrm{F}\mathrm{E}}_{{\mathrm{H}}_2} is below 4%. The minimum CO signal detected at −0.45 V vs. RHE by SP_GDE indicates that the onset overpotential is smaller than 340 mV which is notably lower than 490 mV required by NP_GDE. NP_GDE was more liable to carry out hydrogen evolution reaction (HER) rather than CO evolution with approximately 20% FE_CO at its optimal potential of −1.54 V vs. RHE.

3.2 GDE characterizations

Before carbonization, the pristine GDE exhibits an initial size of 200 cm², indicating its potential scalability (Fig. S2c). GDE with a thickness of ~400 µm, comprise interconnected fibers with diameters in the range of several hundred nanometers (Fig.2(a)–2(c) and S4). The porous structure of GDE with a pore size distribution of 1–120 nm indicates the development of macropores, mesopores, and micropores during carbonization. These pores play distinct roles in CO₂ adsorption. Particularly, micropores (smaller than 20 nm) serve as primary adsorption sites for CO₂ molecules (kinetic diameter of 3.4 Å). Additionally, mesopores (20–50 nm) and macropores (larger than 50 nm) function as channels to facilitate gas transport. The CO₂ adsorption isotherm confirms that GDE can store CO₂ with a capacity of 4–7 cm³ of CO₂/g-GDE under atmospheric pressure. Moreover, the BET analysis reveals negligible differences in CO₂ storage capacity between SP_GDE and NP_GDE, indicating that the introduction of SP does not influence the porous structure and CO₂ storage capacity. Hydrophobicity, a crucial factor influencing GDE performance, is significantly improved through the treatment of GDE (SP_DGE and NP_GDE) with a 5 wt% Nafion 117 solution (Yang et al., 2020), resulting in a water contact angle of 135.7° (Fig.2(f) and S5). We further used a confocal laser scanning microscope to assess the penetration depth of a Rhodamine B water solution, highlighting the enhanced water resistance achieved through Nafion treatment. Notably, after Nafion treatment, the penetration depth of the Rhodamine B solution decreases from ~40 to ~20 μm, further confirming the improved hydrophobicity of the SP_GDE surface.

Before the analysis of chemical bonds, we first examined the structure of the Ni-containing SP protein. The XPS N 1s spectrum exhibits three split peaks at 398.3, 399.4, and 400.1 eV, corresponding to NH₂/NH₃ groups in SP proteins (Fig.3(a) and S6). After Ni adsorption onto SP, the intensity of the peak at 399.4 eV significantly decreases, indicating the formation of coordination between N_x and Ni. This coordination is considered a crucial catalytic site for CO₂ to CO conversion (Zheng et al., 2019; Yang et al., 2020; Yao et al. 2024; Qin et al. 2025). The emergence of Ni–S coordination (Fig.3(b)) further indicates the formation of various complex structures between Ni and SP. In these complexes, Ni mainly exists as Ni²⁺ (Fig.3(c)). After Ni adsorption, significant alterations were observed in the FTIR spectra. A series of changes observed around 1400 cm⁻¹ (Fig.3(d)) correspond to the symmetric stretching vibration of COO⁻, indicating interactions between Ni and C-containing groups (Ye et al. 2025). Additionally, the peak around 2940 cm⁻¹ is attributable to the symmetric stretching vibration of dispersed NH₂⁺/NH₃⁺, indicating a crucial complexation form between SP and HMs (Blundell and Jenkins, 1977; Weng and Xu, 2016).

After preoxidation and carbonization processes, differences between SP_GDE and NP_GDE are observed. The FTIR spectra of both SP_GDE and NP_GDE exhibit comparable peaks above 900 cm⁻¹, (Fig.3(d)), with C–H stretching vibrations (990–1150 cm⁻¹) and C–N vibrations (1124 or 1219 cm⁻¹). However, significant differences were observed in the lower absorbance region. Specifically, the FTIR spectrum of NP_GDE displays distinct peaks at 594 and 669 cm⁻¹, indicative of metal–O vibrations, suggesting interactions between Ni and oxygen-containing functional groups in NP_GDE (Weng and Xu, 2016). The XRD spectra of both SP_GDE and NP_GDE reveal a graphitic matrix and peaks at 26.2° and 44° (Fig.3(e)), with no detection of other crystal phases. The XPS analysis confirms that the Ni content is below 1 at% (Fig. S7), consistent with previous reports on Ni SACs (Zheng et al., 2019; Yang et al., 2020; Yao et al. 2024; Qin et al. 2025). Additionally, no Ni⁰ (853.5 eV) peak associated with Ni nanoparticles was detected (Fig.3(f)). The N 2p_3/2 spectrum (Fig.3(h)) exhibits five typical peaks corresponding to pyridinic N (398.2 eV), N_x–Ni (399.4 eV), pyrrolic N (400.5 eV), graphitic N (401.3 eV), and oxidized N (403.0 eV) species in SP_GDE. Although both SP_GDE and NP_GDE contain similar Ni species, their N chemical forms differ. NP_GDE contains a smaller proportion of N_x–Ni and a higher pyridinic N content, which may reduce its conductivity of NP_GDE owing to the lower electron cloud density of pyridine (Zheng et al., 2019; Yang et al., 2020; Qin et al. 2025). The S–Ni coordination disappeared after carbonization (Fig.3(b) and Fig.3(h)), indicating that S–Ni is unstable during carbonization and cannot effectively disperse Ni SAs.

The nanofiber surface of SP_GDE exhibits a homogeneous structure with uniform dispersion of Ni and amorphous carbon as the matrix, and no Ni nanoparticles or clusters were detected either via SEM or TEM (Fig.4 and S8, S9). The uniform dispersion of Ni and S elements onto the nanofiber surface suggests that waste SP and Ni mainly attach to the surface of the PAN nanofibers rather than forming nanofibers during electrospinning. This behavior was consistent with our finding that an SP + Ni solution without PAN cannot form threadlike fibers during electrospinning, unlike PAN. The distribution of electrocatalytic sites on the SP_GDE surface (maintained at −0.6 V vs. RHE) was investigated using a scanning electrochemical microscopy system with a Pt ultramicroelectrode tip poised at 0.84 V vs. RHE. During the scanning of the SP_GDE surface, the tip oxidizes the ECO₂RR products (either CO or H₂) to generate a measurable current that visualizes electrochemical activity. The current distribution exhibits a wave-like pattern with a characteristic width of ~400 nm, consistent with the diameter of SP_GDE nanofibers. This indicates that efficient electrocatalytic sites are efficiently distributed on the surface of each nanofiber (Fig.4(f)–Fig.4(h), and S10). In contrast, the epoxy resin used to fix the SP_GDE samples produces negligible current signals, forming a distinct boundary with the SP_GDE material.

The nanofiber surface of NP_GDE is similar to that of SP_GDE (Figs. S11 and S12). However, nanoparticles with crystal cells were detected on the surface of some nanofibers via AC-HAADF-STEM (Fig. S13). These crystal cells exhibit distinguishable brightness compared with the surrounding carbon matrix. The corresponding elemental mapping displays Ni element agglomeration on these crystal cells. The ~0.2 nm fringe pattern indicates that these nanoparticles are pure Ni⁰ nanoparticles. Ni SAs form at the nanofiber surface (Fig. S14). However, no Ni⁰ peaks are observed in the XRD and XPS spectra of NP_GDE (Fig.3(f) and Fig.3(g)) owing to the low loading of Ni nanoparticles on the nanofibers. Generally, the electrospinning technique can homogeneously disperse Ni SAs onto a matrix. Nevertheless, owing to the agglomeration tendency of Ni, a dispersing agent (SP) is required to stabilize Ni in its atomic form.

3.3 ECO₂RR performance in a MEA

Ni SACs were successfully synthesized on both SP_GDE and NP_GDE surfaces. However, the presence of Ni nanoparticles on the NP_GDE surface affected its product selectivity compared with SP_GDE, as the Ni nanoparticles influenced HER even at low concentrations. Acid soaking is a common method for removing nanoparticles. The selectivity toward CO in NP_GDE increased with prolonged soaking time in 2 mol/L H₂SO₄ (Fig. S15). The FE_CO of SP_GDE slightly decreased with increasing soaking time. Notably, the non-treated SP_GDE exhibited the highest CO selectivity. This indicates that acid soaking is not required for SP_GDE owing to the absence of Ni nanoparticles on SP_GDE and the protective effect of the synthesis strategy on Ni SAs. At lower carbonization temperatures (700–800 °C), the synthesized SP_GDE exhibited favorable CO selectivity (Fig. S16). This suggests that SP_GDE synthesis requires a more moderate energy input compared with ZIF-8-derived Ni SACs, which are typically synthesized at higher temperatures (900–1000 °C) to volatilize residual Zn.

The TOFs (h) of SP_GDE for CO production at −1.1 V vs. RHE was measured as 10305.82 h⁻¹ based on ECSA normalization (Fig.5(a)), which is comparable to most reported values (Jiang et al., 2018; Zheng et al., 2019). TOFs reflect the number of catalytic sites occurring, the target products generated, and the reactants consumed per unit active site under temperature, pressure, reactant ratio, and reaction conditions. Moreover, TOFs represent the catalytic reaction rate. In the ECO₂RR system, TOFs effectively indicate the catalytic activity near the electrode surface within the reaction-diffusion layer, which is closely related to ESCA (Jiang et al., 2018; Zheng et al., 2019). The electrochemical double-layer capacitance, a basic index for ESCA, was determined to be 8.49 and 7.51 mF/cm² for SP_GDE and NP_GDE respectively, (Fig. S17). These values were consistent with many reported studies and even comparable to some graphite-based matrices (2.56–8.20 mF/cm²). In electrochemical reactions, a larger ESCA indicated a greater effective area on the electrode for reactions to occur, leading to higher reaction rates and current densities (Jiang et al., 2018; Zheng et al., 2019). In the overall two-electron ECO₂RR process, the rate-determining step at low overpotentials was the electron–proton transfer from the adsorbed CO₂ (*CO₂^·−) to the *COOH intermediate. The Tafel slope for CO₂ to CO conversion catalyzed by SP_GDE was 171 mV/dec, suggesting that single electron transfer is the rate-limiting step, which may involve the generation of *CO₂^·−, *COOH, or both (Fig.5(b)). This indicates that CO₂ adsorption and activation on the SP_GDE surface occur relatively rapidly (Jiang et al., 2018; Zheng et al., 2019; Liu et al., 2021).

At high current densities, SP_GDE exhibited excellent performance in MEA (Fig. S18). Particularly, the CO₂ to CO conversion rate rapidly increased above a 2.0-V cell voltage and maintained a high FE_CO (> 90%) across a wide range of current densities between 10 and 50 mA/cm² (Fig.5(c)). These values correspond to an energy efficiency of > 50% (Fig.5(d)). As the current further increased to 70 mA/cm², CO selectivity remained stable at 77%. The relatively stable CO selectivity after 8 h of high current operation coupled with nearly unchanged chemical properties of SP_GDE indicated the stability of SP_GDE (Fig.5(e) and S19). Despite some changes in FTIR and XRD spectra (Fig. S19) owing to the unwashed KHCO₃, most chemical bonds exhibit minimal changes after electrolysis, confirming the effective protection of Ni SACs in SP_GDE.

4 Conclusions

A Ni SACs-laden GDE is fabricated using waste SP protein and Ni from wastewater through electrospinning. The resulting material exhibits excellent activity for CO₂ to CO electro-conversion. The SP protein in wastewater effectively pre-anchors Ni ions and enriches nitrogen elements, thereby promoting the formation of N_x–Ni active sites. These sites significantly enhance ECO₂RR performance, achieving a maximum FE_CO of 96%. The electrospinning technique is suitable for producing GDE in large quantities, ensuring uniform and controlled exposure of N_x–Ni sites along the GDE nanofibers. This unique architecture and the complex liquid/gas/solid interface formed on the GDE surface enable high current operation (90% FE_CO at 50 mA/cm²). In addition to facilitating organic waste disposal, the use of GDE-based CO₂ reduction provides additional environmental benefits, such as carbon capture/fixation (478 million tons annually) and HMs recovery (5 million tons annually). This study presents a strategy to address climate change challenges and promote sustainable waste management practices.

References

[1] Blundell T L, Jenkins J A. (1977). The binding of heavy metal to proteins. Chemical Society Reviews, 6(2): 139–171

[2] Fan L, Xia C, Zhu P, Lu Y, Wang H. (2020). Electrochemical CO₂ reduction to high-concentration pure formic acid solutions in an all-solid-state reactor. Nature Communications, 11(1): 3633

[3] Ge J, Lei J, Zare R N. (2012). Protein-inorganic hybrid nanoflowers. Nature Nanotechnology, 7(7): 428–432

[4] Gu C H, Pan Y, Wei T T, Zhang A Y, Si Y, Liu C, Sun Z H, Chen J J, Yu H Q. (2024). Upcycling waste sewage sludge into superior single-atom Fenton-like catalyst for sustainable water purification. Nature Water, 2(7): 649–662

[5] GustavssonJ, CederbergC, Sonesson U, OtterdijkR, MeybeckA (2011). Global Food Losses and Food Waste-Extent, Causes and Prevention. Rome: Food and Agriculture Organization of the United Nations

[6] Jiang K, Siahrostami S, Zheng T, Hu Y, Hwang S, Stavitski E, Peng Y, Dynes J, Gangisetty M, Su D. . (2018). Isolated Ni single atoms in graphene nanosheets for high-performance CO₂ reduction. Energy & Environmental Science, 11(4): 893–903

[7] Jones N. (2013). Troubling milestone for CO₂. Nature Geoscience, 6(8): 589

[8] Li T, Lees E W, Goldman M, Salvatore D A, Weekes D M, Berlinguette C P. (2019). Electrolytic conversion of bicarbonate into CO in a flow cell. Joule, 3(6): 1487–1497

[9] Li Z, He D, Yan X, Dai S, Younan S, Ke Z, Pan X, Xiao X, Wu H, Gu J. (2020). Size-dependent nickel-based electrocatalysts for selective CO₂ reduction. Angewandte Chemie International Edition, 59(42): 18572–18577

[10] Liang S, Jiang Q, Wang Q, Liu Y. (2021). Revealing the real role of nickel decorated nitrogen-doped carbon catalysts for electrochemical reduction of CO₂ to CO. Advanced Energy Materials, 11(36): 2101477

[11] Liu C, Wu Y, Sun K, Fang J, Huang A, Pan Y, Cheong W C, Zhuang Z, Zhuang Z, Yuan Q. . (2021). Constructing FeN₄/graphitic nitrogen atomic interface for high-efficiency electrochemical CO₂ reduction over a broad potential window. Chem, 7(5): 1297–1307

[12] Liu Y, Li H, Chen S, Zhang L, Li S, Lv H, Gao J, Cui S, Jiang K. (2024). Synergetic pathways of water-energy-carbon in ecologically vulnerable regions aiming for carbon neutrality: a case study of Shaanxi, China. Frontiers of Environmental Science & Engineering, 18: 106

[13] Lu L, Guest J S, Peters C A, Zhu X, Rau G H, Ren Z J. (2018). Wastewater treatment for carbon capture and utilization. Nature Sustainability, 1(12): 750–758

[14] Mohareb E A, Heller M C, Guthrie P M. (2018). Cities’ role in mitigating united states food system greenhouse gas emissions. Environmental Science & Technology, 52(10): 5545–5554

[15] Nguyen T N, Dinh C T. (2020). Gas diffusion electrode design for electrochemical carbon dioxide reduction. Chemical Society Reviews, 49(21): 7488–7504

[16] Peng C, Luo G, Zhang J, Chen M, Wang Z, Sham T K, Zhang L, Li Y, Zheng G. (2021). Double sulfur vacancies by lithium tuning enhance CO₂ electroreduction to n-propanol. Nature Communications, 12(1): 1580

[17] Qin J, Wang Q, Han B, Jin C, Luo C, Sun Y, Dai Z, Wang S, Liu H, Zheng X. . (2025). Heterogeneous Fe–Ni dual-atom catalysts coupled N-vacancy engineering for enhanced activation of peroxymonosulfate. Applied Catalysis B: Environment and Energy, 360: 124538

[18] Reichstein M, Riede F, Frank D. (2021). More floods, fires and cyclones-plan for domino effects on sustainability goals. Nature, 592(7854): 347–349

[19] Sharifian R, Wagterveld R M, Digdaya I A, Xiang C, Vermaas D A. (2021). Electrochemical carbon dioxide capture to close the carbon cycle. Energy & Environmental Science, 14(2): 781–814

[20] Sun J, Tu R, Xu Y, Yang H, Yu T, Zhai D, Ci X, Deng W. (2024). Machine learning aided design of single-atom alloy catalysts for methane cracking. Nature Communications, 15(1): 6036

[21] Sun M, Guan W, Chen C, Wu C, Liu X, Meng B, Chen T, Han Y, Wang J, Xi S. . (2025). Mechanistic insight into the synergy between nickel single atoms and nanoparticles on N-doped carbon for electroreduction of CO₂. Journal of Energy Chemistry, 100: 327–336

[22] Tan A, Zhao F, Zhang Y, Li G, Wu C, Liu Z, Li J, Liu J. (2025). Innovative application of transfer learning on small-scale datasets: analysis and optimization of catalyst ink for the low-iridium membrane electrode assemblies of proton exchange membrane water electrolysis. Chemical Engineering Science, 302: 120814

[23] Verma S, Kim B, Jhong H R M, Ma S, Kenis P J A. (2016). A gross-margin model for defining technoeconomic benchmarks in the electroreduction of CO₂. ChemSusChem, 9(15): 1972–1979

[24] Voosen P. (2020). The hunger forecast. Science, 368: 226–229

[25] Wang W, Shao C, Feng Y, Cheng Y. (2024). Effect of single atom loading on the CO₂ reduction activity of pure and defective W₂CO₂ MXene. Molecular Catalysis, 569: 114550

[26] Weekes D M, Salvatore D A, Reyes A, Huang A, Berlinguette C P. (2018). Electrolytic CO₂ Reduction in a flow cell. Accounts of Chemical Research, 51(4): 910–918

[27] WengS, Xu Y (2016). Analysis of Fourier Transform Infrared Spectroscopy (3rd edition). Beijing: Chemical Industry Press (in Chinese)

[28] Xia C, Zhu P, Jiang Q, Pan Y, Liang W, Stavitski E, Alshareef H N, Wang H. (2019). Continuous production of pure liquid fuel solutions via electrocatalytic CO₂ reduction using solid-electrolyte devices. Nature Energy, 4(9): 776–785

[29] Xu X, Sharma P, Shu S, Lin T S, Ciais P, Tubiello F N, Smith P, Campbell N, Jain A K. (2021). Global greenhouse gas emissions from animal-based foods are twice those of plant-based foods. Nature Food, 2(9): 724–732

[30] Yang H, Lin Q, Zhang C, Yu X, Cheng Z, Li G, Hu Q, Ren X, Zhang Q, Liu J. . (2020). Carbon dioxide electroreduction on single-atom nickel decorated carbon membranes with industry compatible current densities. Nature Communications, 11(1): 593

[31] Yang H, Wu Y, Li G, Lin Q, Hu Q, Zhang Q, Liu J, He C. (2019). Scalable production of efficient single-atom copper decorated carbon membranes for CO₂ electroreduction to methanol. Journal of the American Chemical Society, 141(32): 12717–12723

[32] Yang M, Li X, Chao W, Gao X, Wang H, Lu L. (2024). Renewable biosynthesis of isoprene from wastewater through a synthetic biology approach: the role of individual organic compounds. Frontiers of Environmental Science & Engineering, 18: 28

[33] Yao R, Sun K, Zhang K, Wu Y, Du Y, Zhao Q, Liu G, Chen C, Sun Y, Li J. (2024). Stable hydrogen evolution reaction at high current densities via designing the Ni single atoms and Ru nanoparticles linked by carbon bridges. Nature Communications, 15(1): 2218

[34] Ye C, Guo Z, Zhou Y, Shen Y. (2025). Nickel-based dual single atom electrocatalysts for the nitrate reduction reaction. Journal of Colloid and Interface Science, 677: 933–941

[35] Yuan X, Zhang X, Yang Y, Li X, Xing X, Zuo J. (2024). Emission of greenhouse gases from sewer networks: field assessment and isotopic characterization. Frontiers of Environmental Science & Engineering, 18: 119

[36] Zheng T, Jiang K, Ta N, Hu Y, Zeng J, Liu J, Wang H. (2019). Large-scale and highly selective CO₂ electrocatalytic reduction on nickel single-atom catalyst. Joule, 3(1): 265–278

RIGHTS & PERMISSIONS

Higher Education Press 2025