1 Introduction
As global energy demands continue to rise, coupled with growing concerns over the depletion of fossil fuel reserves, the development of cost-effective, sustainable, and environmentally friendly energy technologies has become a critical area of focus. Rechargeable zinc-air batteries (RZABs) have emerged as a promising solution owing to their exceptional properties, including high energy density, the widespread availability of zinc, and environmental compatibility. These characteristics position RZABs as a viable candidate for addressing the dual challenges of energy scarcity and environmental pollution, thus offering a renewable, efficient, and eco-friendly alternative for future energy storage systems [
1–
4]. Additionally, compared to fuel cells, RZABs offer improved safety, exceptional flexibility for wearable devices, and the ability to be recharged, making them an ideal energy source for portable electronic devices [
5,
6]. However, the sluggish reaction kinetics and inferior structural durability of the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) remain significant challenges for the large-scale utilization of RZABs [
7–
9]. To enhance the practical viability of ZABs, the rational design and development of high-performance catalysts that facilitate rapid electron and mass transfer, reduce reaction energy barriers, and improve energy efficiency is essential. Currently, the most advanced ORR and OER electrocatalysts are primarily composed of noble metals, such as Pt/C and RuO
2, and their varied derivatives [
10]. Nonetheless, their inherent limitations, such as poor stability and high costs, substantially hinder their practical applications in the energy sector [
11,
12]. Thus, the development of inexpensive, efficient, and durable catalysts is crucial for the widespread commercialization of RZABs.
Multifunctional metal-nitrogen-carbon single-atom catalysts (M-N-C SACs) have emerged as highly promising candidates for both ORR and OER, owing to their outstanding atomic utilization, excellent catalytic selectivity, tunable electronic structures, and high electrical conductivity [
13–
17]. In this regard, metal phthalocyanines (MPcs), representative macrocyclic M-N-C-based molecules, have gained recognition as effective molecular catalysts for ORR due to their straightforward synthesis and remarkable catalytic performance [
18–
21]. However, these molecular catalysts require highly specific carbon substrates, such as single-/multi-walled carbon nanotubes, 2D graphene, and porous carbons, to achieve satisfactory electrocatalytic performance [
22,
23]. It is important to note that most carbon-based heterogeneous catalysts exhibit limited stability during charging–discharging cycles and under high-voltage conditions in practical applications. The corrosion of carbon nanomaterials plays a key role in activity degradation and reduced lifespan during electrocatalytic oxidation, resulting in the aggregation, dissolution, and detachment of the embedded metal active centers [
24]. Therefore, the design and preparation of suitable carbon substrates to anchor FePc molecules for achieving efficient and stable bifunctional ORR and OER performance remains a significant challenge.
Recently, metal-organic frameworks (MOFs) have attracted significant attention in electrocatalysis due to their adaptable and customizable nanostructures, exceptionally high surface area, and abundant porosity [
25–
27]. Among various MOFs, zeolitic imidazolate frameworks (ZIFs) are particularly notable as a subclass of crystalline inorganic-organic hybrids with high nitrogen content. Through a straightforward pyrolysis process, ZIFs can be easily converted into hierarchically porous carbon nanomaterials with uniformly doped metal-nitrogen sites [
28,
29]. Additionally, Co-based ZIFs are particularly effective in deriving active species (such as Co nanoparticles (NPs) [
30], Co
3O
4 [
31], CoP [
32], CoS
x [
33]) for OER. Inspired by these advantages, a bifunctional CoNC@FePc was successfully synthesized using a straightforward pyrolysis method, based on the Co-ZIF precursor. In this process, FePc molecules were loaded onto Co NPs within an N-doped carbon substrate via a convenient solution-based approach (Fig.1). Thanks to the strong molecular interaction between FePc molecules and Co species, the resulting catalyst exhibited superior bifunctional oxygen catalytic performance, achieving a positive half-wave potential of 0.87 V for ORR and a minimal overpotential for OER. Moreover, the assembled ZAB using CoNC@FePc as the air cathode exhibited an impressive open-circuit potential of 1.53 V and a peak power density of 150.2 mW/cm
2. It also demonstrated excellent electrochemical and structural stability, maintaining consistent performance over a charge–discharge cycling of approximately 100 h.
2 Results and discussion
To investigate the morphology of CoNC@FePc, scanning electron microscopy (SEM) and transmission electron microscopy (TEM) were employed. The SEM images provided detailed insights into the structural features of CoNC@FePc, revealing that its morphology largely retained the features of the Co-ZIF precursor (Figs. S1–S3). Fig.2(a) and Fig.2(b) present the TEM and high-resolution TEM (HRTEM) images, clearly revealing the hexagonal morphology of CoNC. The HRTEM image displayed two distinct fringe lattices of 0.336 and 0.205 nm, attributed to the graphitic C (002) and metallic Co (111) planes, respectively [
34]. Additionally, the aberration-corrected annular dark-field scanning transmission electron microscopy (AC-HAADF-STEM) image (Fig.2(c)) revealed a pronounced concentration of bright spots within the material. These spots were indicative of atomically dispersive FePc molecules and the
in-situ formation of Co species. Larger bright spots corresponded to Co NPs, while the smaller isolated bright spots were consistent with individual FePc molecules. This distinctive imaging pattern confirmed the presence of Co NPs and FePc molecules, both embedded on the porous graphitic N-doped carbon substrate.
In Fig.2(d), the powder X-ray diffraction (PXRD) analysis of CoNC@FePc revealed diffraction patterns similar to those of CoNC. Specifically, two prominent peaks were observed at 44.09° and 51.37°, corresponding to the (111) and (200) crystallographic planes of metallic cobalt (PDF #97-005-3805). Remarkably, no distinct diffraction peaks associated with FePc were detected after it was dispersed and anchored onto the porous CoNC, indicating that FePc species remained well-dispersed without noticeable aggregation. This suggests that the interaction between FePc molecules and the CoNC matrix effectively prevented the formation of larger aggregates. Furthermore, energy-dispersive X-ray spectroscopy (EDS) and corresponding elemental mapping confirmed the uniform distribution of C, N, O, Co, and Fe across the ZIF-derived carbon substrate (Fig.2(e), Fig. S4).
The nitrogen adsorption–desorption isotherms of MOF-derived CoNC exhibited typical type-IV behavior, indicative of the presence of abundant mesoporosity and a substantial surface area of 106.2 m
2/g (Fig.3(a)). However, upon incorporation of molecular catalysts of FePc into the porous CoNC framework, the surface area of the resulting CoNC@FePc composite significantly decreased to 34.2 m
2/g. This reduction suggests that the FePc molecules have effectively infiltrated and occupied the mesopores of the CoNC, as summarized in Table S1. Raman spectra (Fig.3(b)) clearly confirmed the presence of FePc, with two distinct peaks at 1390.83 and 1506.62 cm
−1, characteristic of FePc’s molecular vibrations [
35]. In addition, two more significant peaks at 1331 and 1577 cm
−1 were observed, corresponding to the D band and G band of the carbon material, respectively. The relative intensity ratio (
ID/
IG) of the D to G bands was calculated to be 1.09 for CoNC@FePc and 0.98 for CoNC, suggesting a higher level of disorder and an increased concentration of defects in the CoNC@FePc composite. This enhanced structural disorder in CoNC@FePc, compared to CoNC, reflects the incorporation of FePc and its impact on the carbon matrix [
36].
The high-resolution C 1s XPS spectra (Fig.3(c)) were carefully deconvoluted into four subpeaks, corresponding to different carbon species: O‒C = O at 292.3 eV, C‒O at 289.6 eV, C‒N at 287.1 eV, and C‒C at 284.8 eV. Similarly, the deconvoluted N 1s spectra (Fig.3(d)) revealed four nitrogen species within the CoNC@FePc structure: oxidized nitrogen at 404.1 eV, graphitic nitrogen at 402.1 eV, pyrrolic nitrogen at 400.4 eV, and pyridinic nitrogen at 398.9 eV (Table S2). The high-resolution spectra of Co 2p3/2 and Co 2p1/2 peaks, shown in Fig.3(e), appeared at binding energies of 777.6/792.3, 781.0/796.5, 784.8/800.5, and 788.0/804.0 eV, corresponding to Co(0), Co(III), Co(II), and satellite peaks, respectively. Importantly, the Co 2p3/2 peak shifted significantly by 0.9 eV to higher binding energies, suggesting effective charge redistribution between the Co species and the molecular FePc catalysts. This shift likely reflected the establishment of strong electronic interactions, enhancing charge transfer and potentially influencing the electronic structure and catalytic performance of the CoNC@FePc composite.
The electrochemical performance of CoNC@FePc was systematically evaluated using a rotating ring-disk electrode (RRDE). As illustrated in Fig.4(a) and S7, the linear sweep voltammetry (LSV) curves showed that the optimal CoNC@FePc exhibited a superior half-wave potential (E1/2) of 0.87 V, alongside a high diffusion-limiting current (JL) of 5.42 mA/cm2. These values notably outperform those of Pt/C (E1/2= 0.85 V, JL = 5.24 mA/cm2) and CoNC (E1/2 = 0.79 V, JL = 4.11 mA/cm2), highlighting the enhanced catalytic performance and electrochemical activity of CoNC@FePc. The Tafel slope for CoNC@FePc was calculated to be 38.8 mV/dec, significantly lower than those for CoNC (53.3 mV/dec) and Pt/C (91.1 mV/dec), indicating much faster kinetic efficiency for ORR (Fig.4(b)).
The ORR pathway selectivity, depicted in Fig.4(c), revealed that CoNC@FePc effectively facilitated the preferred 4-electron transfer pathway, minimizing the formation of hydrogen peroxide (H2O2). To assess the electrochemical active surface area (ECSA), the electrochemical double-layer capacitance (Cdl) was measured. CoNC@FePc demonstrated the largest Cdl value of 23.1 cm−2, surpassing both CoNC (20.2 mF/cm2) and Pt/C (14.8 mF/cm2), as shown in Fig.4(d). This indicates that CoNC@FePc possesses a significantly larger electrochemically accessible surface area, contributing to its enhanced catalytic activity.
Furthermore, upon the introduction of a methanol (MeOH) solution, CoNC@FePc exhibited only a minor change in current density (8.7%), in sharp contrast to the more substantial decrease observed for Pt/C (18.1%) (Fig.4(e)). After a 16-h electrochemical evaluation, CoNC@FePc retained 95.4% of its initial current density, a performance significantly superior to Pt/C, which maintained only 88.3% (Fig.4(f)). This substantial retention of current density underscores the exceptional long-term stability of CoNC@FePc.
Regarding the OER performance, CoNC@FePc demonstrated a low overpotential (η10) of 314 mV at a current density of 10 mA/cm2, outperforming both RuO2 (331 mV) and CoNC (377 mV) (Fig.4(g)). The calculated Tafel slope for CoNC@FePc was 73.4 mV/dec, significantly lower than the values for RuO2 (143.8 mV/dec) and CoNC (91.2 mV/dec), suggesting faster reaction kinetics and superior catalytic efficiency for OER (Fig.4(h)). Additionally, the excellent bifunctional catalytic activity of CoNC@FePc was confirmed by its narrow voltage gap (ΔE) of 0.679 V, outperforming the Pt/C+RuO2 pair (0.703 V), CoNC (0.812 V), and other benchmark electrocatalysts (Fig.4(i), Table S3). These results collectively highlight CoNC@FePc as a highly efficient and durable electrocatalyst for both oxygen reduction and OERs.
To evaluate the practical viability of the as-synthesized catalysts, a series of zinc-air batteries (ZABs) were assembled. As shown in Fig.5(a), the open circuit voltage (OCV) of CoNC@FePc-based ZABs was 1.53 V, outperforming the Pt/C-based ZABs (1.47 V). Moreover, the specific discharge capacity and energy density of CoNC@FePc-based ZABs were substantially higher, reaching 846.5 mAh/g Zn, compared to Pt/C, which exhibited 691.2 mAh/g Zn (Fig.5(b)). In Fig.5(c), the assembled CoNC@FePc-based ZABs demonstrated an impressive peak power density of 150.2 mW/cm2, substantially exceeding that of the commercial Pt/C-based ZABs (139.7 mW/cm2). Under identical operational conditions, the optimized CoNC@FePc-based batteries consistently delivered higher output voltages across a broad range of discharge current densities, highlighting their outstanding rate capability and stability at varying loads (Fig.5(d)).
Fig.5(e) illustrated the discharge‒charge cycling profiles of both ZAB types at a constant current density of 10 mA/cm2. Notably, the CoNC@FePc-based ZAB exhibited a remarkably narrow voltage gap of 0.747 V, with charging and discharging voltages of 1.968 and 1.221 V, respectively, over an extended period of 100 h. In contrast, the commercial Pt/C+RuO2-based ZAB demonstrated significantly lower cycling stability, with a limited operational lifespan of only 35 h. This substantial difference highlights the exceptional cycling stability and energy efficiency of CoNC@FePc-based ZABs, emphasizing their potential for high-performance, long-duration energy storage applications.
To gain a deeper insight into the intrinsic catalytic behaviors of FePc molecules on Co nanoparticle substrates for ORR, density functional theory (DFT) calculations were performed. In the computational model depicted in Fig.5(f) and S11, an FePc molecule was directly anchored to the Co (001) lattice plane, constructing the Co@FePc model. The ORR and OER pathways were modeled to involve four consecutive proton-coupled electron transfer steps, each step associated with the formation of reaction intermediates (Fig. S12) [
37]. At an applied potential of 0.00 V, all the reaction steps displayed a favorable downhill free energy profile, indicating that both catalysts exhibit promising catalytic activity (Fig.5(g)).
As shown in Fig.5(h), the rate-determining step (RDS) for Co@FePc was identified as the reduction of *O2 to *OOH, with a theoretical overpotential of 0.33 eV. This value is significantly lower than the RDS for FePc, which is the reduction of *O to *OH, with an overpotential of 0.54 eV. In the context of OER (Fig. S13), the RDS for Co@FePc was the conversion of *OH to *O, exhibiting a markedly lower energy barrier of 0.44 eV compared to FePc, which has an energy barrier of 0.84 eV. These results not only reinforce the superior ORR/OER performance of Co@FePc, but also provide critical insights into its enhanced catalytic efficiency.
3 Conclusions
In conclusion, FePc molecules were successfully anchored onto Co NPs within an N-doped carbon substrate through direct pyrolysis of Co-ZIF precursors. Owing to the electron interactions between Fe sites in FePc and Co NPs, the as-synthesized CoNC@FePc exhibited excellent bifunctional properties for both ORR and OER, achieving an E1/2 potential of 0.87 V in a 0.1 mol/L KOH electrolyte and a low OER overpotential of 314 mV at a current density of 10 mA/cm2. The optimized CoNC@FePc catalyst was further employed in homemade ZABs, demonstrating a peak power density of 150.2 mW/cm2 and exceptional charge and discharge cycling stability for over 100 h, outperforming the Pt/C + RuO2 air cathode. The high activity of CoNC@FePc for bifunctional oxygen reactions was further confirmed by relevant theoretical calculations, where the RDS was identified as the conversion of *OH to *O for Co@FePc. This paper proposes a novel synthetic strategy for developing superior bifunctional oxygen electrocatalysts and offers significant potential for advancements in clean energy storage and conversion technologies.
PDF
(2506KB)
