Nanowrinkled Carbon Aerogels Embedded with FeN x Sites as Effective Oxygen Electrodes for Rechargeable Zinc-Air Battery.

Rational design of single-metal atom sites in carbon substrates by a flexible strategy is highly desired for the preparation of high-performance catalysts for metal-air batteries. In this study, biomass hydrogel reactors are utilized as structural templates to prepare carbon aerogels embedded with single iron atoms by controlled pyrolysis. The tortuous and interlaced hydrogel chains lead to the formation of abundant nanowrinkles in the porous carbon aerogels, and single iron atoms are dispersed and stabilized within the defective carbon skeletons. X-ray absorption spectroscopy measurements indicate that the iron centers are mostly involved in the coordination structure of FeN4, with a minor fraction (ca. 1/5) in the form of FeN3C. First-principles calculations show that the FeN x sites in the Stone-Wales configurations induced by the nanowrinkles of the hierarchically porous carbon aerogels show a much lower free energy than the normal counterparts. The resulting iron and nitrogen-codoped carbon aerogels exhibit excellent and reversible oxygen electrocatalytic activity, and can be used as bifunctional cathode catalysts in rechargeable Zn-air batteries, with a performance even better than that based on commercial Pt/C and RuO2 catalysts. Results from this study highlight the significance of structural distortions of the metal sites in carbon matrices in the design and engineering of highly active single-atom catalysts.

1. Introduction

Climate change and environmental pollution have motivated the development of sustainable, clean energy technologies, of which rechargeable metal-air batteries have drawn tremendous attention owing to their high energy density and minimal impacts on the environment [1-4]. The overall efficiency of the charge-discharge process of metal-air batteries is determined by two major reactions, oxygen reduction reaction (ORR) and oxygen evolution reaction (OER). Although platinum group metal (PGM) materials, such as Pt/C, RuO2, and Ir/C, possess excellent catalytic activity for either ORR or OER, none of these noble metal catalysts displays a satisfactory performance for both [5], and their scarcity and high costs greatly hinder their practical applications [6, 7]. Therefore, development of bifunctional catalysts with a low cost and high activity is of both fundamental and technological significance, but remains a great challenge.

Recent studies have demonstrated that PGM-free nanocomposites based on carbon materials, such as heteroatom- (including nonmetal and metal atoms) doped porous carbon, are promising bifunctional oxygen catalysts [8-10]. In fact, transition metal-doped carbon catalysts have been widely investigated due to the unique chemical properties caused by their adjustable 3D electronic orbitals. In particular, transition metal-based single-atom catalysts display overwhelming superiority as compared to their nanoparticle and nanocluster counterparts [11-15]. For instance, single-site dispersion of FeN species in a two-dimensional nitrogen-doped porous carbon layer has been found to exhibit a remarkable catalytic activity towards both ORR and OER in alkaline media [1], where a range of catalytic active sites has been proposed, such as CoN2C2, FeN3C, FeN4, and FeN4O [16-18]. However, the effects of structural distortion induced by the single-metal sites on the catalytic activity have long been ignored, although such structural defects are common in pyrolytic carbon.

Herein, biomass hydrogels (i.e., chitosan, gelatin, and agar), which have long been known for their diverse applications and economic advantages [19, 20], were prepared and used as unique precursors, templates, and reactors to produce three-dimensional, nanowrinkled carbon aerogels embedded with FeN single sites [21-23]. Due to the abundant functional groups on the hydrogel chains, defective single-metal sites were dispersed and stabilized within the nanowrinkled, porous carbon aerogels. First-principles calculations showed that the FeN sites in the Stone-Wales configurations induced by the carbon nanowrinkles displayed a much lower free energy for oxygen electrocatalysis than the normal counterparts. Electrochemical measurements exhibited apparent and reversible oxygen electrocatalytic performance towards both ORR and OER. When the nanowrinkled carbon aerogels were used as the air-cathode of a zinc-air battery, the battery displayed a higher open-circuit voltage and higher energy density, as well as better cycling stability than that with commercial Pt/C-RuO2 catalysts.

2. Results and Discussion

2.1. Synthesis and Characterization

In this study, flexible biomass hydrogels were synthesized in a facile process and employed as 3D templates to prepare carbon aerogels embedded with single-metal atoms (Figures 1(a) and ). In order to achieve atomic dispersion, the hydrogel networks were modified by two strategies to minimize metal aggregation. The first is “headstream fixation,” which means immobilization of metal atoms into the hydrogel reactor by complexation agents (e.g., phenanthroline (PM)); and the other, “roadblocks,” is based on rigid templates, such as SiO2 nanoparticles and Zn atoms. Experimentally, a variety of hydrogel/hydrosol networks, i.e., CSSi-Zn/FePM, CSSi/FePM, CSSi/Fe, and CSSi, were prepared by using chitosan (CS) as the structural scaffold, along with a select combination of other precursors, such as SiO2 nanoparticles (Si), FePM, and Zn salt (details in Materials and Methods). The morphological details were first investigated by scanning electron microscopy (SEM) measurements. From , freeze-dried CSSi/FePM, CSSi/Fe, and CSsi hydrosols can be seen to consist of uneven microcavities. However, as shown in Figure 1(b), the microcavities of the CSSi-Zn/FePM hydrogel shows a more uniform size of ca. 50 μm, forming a 3D, continuous framework composed of intertwining CS-Zn chains. This suggests that Zn ions can induce the hydrogelation of CS hydrosol to form much more uniform 3D intertwining networks, which further facilitates the generation of nanowrinkles [24]. Circular dichroism (CD) and UV-vis absorption measurements were then carried out to monitor the structural evolution from CSSi sol to CSSi-Zn/FePM hydrogel. As depicted in , the incorporation of both FePM and Zn2+ ions into chitosan led to marked conformational changes of the CS chains. In UV-vis absorption measurements (), two new peaks can be seen to emerge at ca. 226 nm and 268 nm, due to the strong complexation interaction between Fe and PM [25].

Figure 1

Synthesis and morphological characterization of the NCAC-Zn/Fe carbon aerogel. (a) Schematic representation of the synthesis of the NCAC-Zn/Fe carbon aerogel. (b) SEM image of the freeze-dried CSSi-Zn/FePM hydrogel. Inset is a digital photo of the sample. (c) Bright-field STEM image of the NCAC-Zn/Fe aerogel. Red circles indicate single Fe atoms. Inset is a digital photo of the sample. (d) Dark-field STEM image of the NCAC-Zn/Fe aerogel. Red circles indicate single Fe atoms. Inset is a TEM image of the NCAC-Zn/Fe aerogel, and the scale bar is 30 nm. (e) Elemental maps of NCAC-Zn/Fe aerogels. The scale bar is 10 nm. AFM images of NCAC-Zn/Fe aerogels: (f) adhesion force image and (g) current flow image.

The freeze-dried CSSi-Zn/FePM hydrogel was then used as a 3D reactor to synthesize metal-doped carbon aerogels by controlled pyrolysis, which was then subject to HF etching to remove the SiO2 templates (Figure 1(a)), producing NCAC-Zn/Fe (Figure 1(c) inset). Control samples of NCAC/Fe, CAC/Fe, and CAC were prepared in a similar fashion (details in Materials and Methods). From the transmission electron microscopy (TEM) images in , one can see that the NCAC-Zn/Fe sample displays a highly porous, nanowrinkled structure with rich mesopores of ca. 10 nm in diameter. In both bright-field (Figure 1(c)) and dark-field (Figure 1(d)) scanning transmission electron microscopy (STEM) images, one can clearly see the formation of single-metal atoms (red circles) embedded into the porous carbon matrix. The corresponding elemental maps clearly show that the C, N, and Fe elements are homogeneously distributed across the aerogel (Figure 1(e)). These results indicate successful construction of N-doped carbon aerogels embedded with isolated Fe atoms by using biomass hydrogels as the reactors.

In order to examine the mechanical and electrical properties of the obtained porous carbon, Fast Force Mapping (FFM) measurements were then carried out. The data presented in Figures 1(f)–1(g) and exhibit a ca. 10 nm variation in the mechanical and electrical properties of the porous carbon, confirming the formation of nanowrinkled carbon. Domains dictated by round features in topography () are outlined by prominent changes in max force () and an increase in the adhesion force (Figure 1(f)). Notably, the adhesion force, which represents the bulk modulus or stiffness of the sample, indicates that these round regions are stiffer in the center and softer around the edges. Typically, sp2-hybridized carbon exhibits hydrophobic characteristics, whereas defective carbons are more hydrophilic [26, 27]. With an AFM tip that consists of a hydrophilic silicon oxide layer, a high adhesion force corresponds to a hydrophilic domain. This implies that the metal centers are most likely situated within the high adhesion force areas. Interestingly, from Figures 1(f)to 1(g), one can see that the soft nodes correspond to high electrical conductance. Taken together, these results suggest that the metal sites are mostly located in the high adhesion and high conductivity areas of the porous carbon aerogel [28, 29]. Both features are conducive to oxygen electrocatalysis.

The porosity of the obtained samples was then quantitatively evaluated by N2 adsorption-desorption measurements. The carbon aerogels obtained above all show a Type IV isotherm (Figures 2(a) and ), which suggests the formation of a complex porous network containing a myriad of mesopores with an average size of ca. 10 nm, in line with the diameter of the SiO2 nanoparticle [30]. From the isotherms, the specific surface area of NCAC-Zn/Fe was estimated to be 609 m2 g−1, with a microporous surface area of 111 m2 g−1, which is the highest among the series, a condition favorable for the formation of abundant active sites ( and ). The corresponding X-ray powder diffraction (XRD) patterns are shown in , where only one broad peak at ca. 25° can be observed, due to the (002) diffraction of graphitic carbon [31]. This carbon diffraction became gradually sharpened from CSC to NCAC-Zn/Fe, indicating an increasing degree of graphitization. Importantly, the fact that no other diffraction features were observed suggests the absence of metal (oxide) nanoparticles. In Raman measurements, the I/I ratio of NCAC-Zn/Fe was estimated to be ca. 1.15, much higher than those of the control samples (Figure 2(b)), signifying the generation of rich defects which may be conducive to the formation of metal active sites [31].

Figure 2

Structural characterization of the NCAC-Zn/Fe carbon aerogel. (a) N2 adsorption-desorption isotherm of NCAC-Zn/Fe. Inset is the corresponding pore size distribution. (b) Raman spectra of CAC, CAC/Fe, NCAC/Fe, and NCAC-Zn/Fe. (c) XPS spectra of the N 1s electrons in CAC/Fe and NCAC-Zn/Fe. (d) K-edge XANES of NCAC-Zn/Fe, FePc, and Fe foil. (e) K-edge EXAFS of NCAC-Zn/Fe, FePc, and Fe foil. (f) The corresponding EXAFS fitting curves for the NCALR/Fe sample.

The elemental compositions of the obtained carbon samples were then quantitatively assessed by inductively coupled plasma atomic emission spectroscopy (ICP-OES) and energy-dispersive X-ray spectroscopy (EDS) measurements. Results from ICP-OES analysis showed that the Fe content in the carbon aerogel was about 0.22 wt% for CAC/Fe, 0.61 wt% for NCAC/Fe, and 0.72 wt% for NCAC-Zn/Fe, in good accordance with the EDS results (, Tables ). The increased metal content suggests the important roles of PM (chelation) and Zn2+ ions (porogen and gel initiator) into fixing Fe centers in the carbon matrix. Further analysis was conducted with X-ray photoelectron spectroscopy (XPS) measurements. First of all, no Fe−O peak can be resolved in the high-resolution scan of the O 1s electrons (), suggesting that Fe atoms are most likely coordinated to other atomic sites such as N and C; and from the XPS spectra of the N 1s electrons of the series of samples (Figures 2(c), ), one can see that the successive introduction of PM and Zn2+ into the precursors increased the N doping from 2.25 at% to 3.60 at% in the carbon matrix (Tables ), and the pyridinic N fraction was the highest in the NCAC/Fe (0.52 at%) and NCAC-Zn/Fe (0.51 at%) samples. In addition, as compared to CSC/Fe, the much stronger Fe-N peak (0.44 at% vs. 0.04 at%) in the NCAC-Zn/Fe sample suggests the generation of more abundant FeN moieties in the carbon aerogels.

The structural configuration of the FeN functional moiety was then examined by X-ray absorption spectroscopy (XAS) measurements. From Figure 2(d), one can see that the Fe K-edge X-ray absorption near-edge spectrum (XANES) of NCAC-Zn/Fe is very similar to that of FePc but markedly different from that of an Fe foil, suggesting a comparable oxidation state (+2) of the Fe centers in NCAC-Zn/Fe and FePc. In the extended X-ray absorption fine structure (EXAFS) spectrum of the Fe foil, the Fe-Fe peak is well-defined at 2.21 Å (Figure 2(e)); however, this peak is absent in NCAC-Zn/Fe, consistent with the atomic dispersion of Fe in the NCAC-Zn/Fe sample (Figure 1). In fact, both the NCAC-Zn/Fe and FePc samples display only a single major peak at 1.41 Å, which can be assigned to the Fe-N bond. Furthermore, the first shell of NCAC-Zn/Fe is well fitted with 3.8 N and 0.2 C with the same bond length of 1.94 Å (Figures 2(f) and ). Taken together, these results suggest that the Fe centers in NCAC-Zn/Fe were mostly involved in the coordination structure of FeN4, with a minor fraction (ca. 1/5) in the form of FeN3C.

2.2. Theoretical Investigation of the ORR Using the NCAC-Zn/Fe Catalyst

First-principles calculations were then carried out to shed light on the contributions of FeN4 and FeN3C moieties to the electrocatalytic activity. It is likely that the interlaced 3D structure of the hydrogel networks and the tortuous CS chains can lead to the formation of abundant wrinkles in the obtained porous carbon aerogels. Therefore, the electrocatalytic activities of the Stone-Wales- (SW-) defect FeN4 (FeN4 SW) and FeN3C (FeN3C SW) moieties, which can be formed by the nanowrinkles of carbon matrices, are examined by theoretical calculations. Figures 3(a)–3(b) and show the side view and top view of the atomic models of the four kinds of Fe-N centers. From the atomic models, one can see that the normal FeN4 and normal FeN3C moieties exhibit a planar structure in the carbon matrices, while FeN4 SW and FeN3C SW show a distorted nonplanar structure. The simulated scanning tunneling microscopic (STM) images of the FeN4 and FeN4 SW moieties are presented in Figures 3(c)–3(d) and . As compared to normal FeN4, the SW defects cause significant redistribution of electron densities of FeN4 and adjacent carbon atoms. Figure 3(e) displays the total density of states (DOS) of normal and SW Fe-N centers. According to Figure 3(f), for the FeN4 SW on a graphene sheet, the Fe atom makes the largest contributions to the DOS near the Fermi level (red peak), which is similar to that (black line) of normal FeN4. Apparently, the marked state of FeN4 SW (highlighted by arrows in Figure 3(e)) is much closer to the Fermi level than that of normal FeN4, indicating a higher probability of donating electrons and reducing oxygen.

Figure 3

Atomic model and DOS of metal sites; Gibbs free energy diagrams of ORR. (a, c) Side/top view and simulated STM image (at a bias of -1.0 V) of normal FeN4 doped graphene sheets. (b, d) Side/top view and simulated STM image (at a bias of -1.0 V) of Stone-Wales FeN4- (FeN4 SW-) doped graphene sheets. (e) Density of state (DOS) of normal FeN4 and FeN4 SW-doped graphene sheets. (f) DOS of FeN4 SW and Fe 3d. (g) Free energy diagrams of ORR processes on normal FeN4, normal FeN3, FeN4 SW, and FeN3 SW at the applied potential of +0.9 V vs. RHE.

To evaluate the ORR activity of these Fe-N metal centers, the reaction free energy is calculated at the applied potential of +0.9 V vs. RHE and plotted in Figure 3(g). One can see that the first two electron-transfer steps are exothermic and the last two endothemic, with the rate determining step (RDS) most likely the fourth electron-transfer step of water formation and desorption. In comparison with normal FeN4 and FeN3, both FeN4 SW and FeN3 SW show much lower endothermic energies (0.179 eV and 0.228 eV), implying a lower reaction overpotential. These results suggest that the nanowrinkles can enhance the electrocatalytic activity of Fe-N centers on the carbon matrices by forming SW defects, as manifested below in electrochemical tests.

2.3. Electrocatalytic Activity towards ORR

The electrocatalytic activity of the nanowrinkled carbon aerogels obtained above was then investigated in 0.1 M KOH. First, electrical impedance spectroscopy (EIS) analysis was carried out to investigate the electron-transfer kinetics. For the NCAC-Zn/Fe catalyst, the small diameter at high frequency and the steep tail at low frequency suggest excellent channels for both mass transfer and charge transfer. Such a low impedance is anticipated to facilitate ORR electrocatalysis (). Figures 4(a) and 4(b) show the ORR polarization curves and H2O2 yields of the carbon aerogels, in comparison to commercial Pt/C (20 wt%). As a metal-free catalyst, the CAC sample shows a rather apparent electrocatalytic activity with an onset potential (Eonset) of +0.94 V vs. RHE and a half-wave potential (E1/2) of +0.79 V, much more positive than those of other carbon catalysts reported in recent literature [15, 32, 33]. This suggests that biomass alone may be exploited as a carbon source to fabricate metal-free ORR electrocatalysts. Notably, doping of the FePM complex into the CSSi leads to a marked enhancement of the catalytic performance with Eonset = +1.10 V and E1/2 = +0.90 V (NCAC-Zn/Fe), which is even better than those of commercial Pt/C (+0.99 V and+ 0.83 V) [34]. Likewise, the NCAC-Zn/Fe single-atom catalyst shows the lowest average H2O2 yield (1.45%) within the potential range of +0.2 V to +0.9 V, signifying a high-efficiency 4e− reduction pathway (Figures ). From the Koutecky-Levich plots, the kinetic current density (J) at +0.85 V was estimated to be 9.12 mA cm−2, about 3.3 times that of Pt/C (2.80 mA cm−2, Figure 4(c)). Both NCAC-Zn/Fe and Pt/C show a low Tafel slope (85 vs. 87 mV dec−1) in the high potential range, illustrating an efficient kinetic process of ORR on these two catalysts (Figure 4(d)). Besides, in contrast with Pt/C, the NCAC-Zn/Fe exhibits remarkable durability and methanol tolerance (). In addition, the E1/2 and diffusion-limited current density of NCAC-Zn/Fe were comparable to those of Pt/C in acidic media (0.1 M HClO4), suggesting the high ORR activity of the single iron atom catalysts even at low pH (Figure 4(e)).

Figure 4

ORR performance. (a) ORR polarization curves of CAC, CAC/Fe, NCAC/Fe, and NCAC-Zn/Fe, as well as the Pt/C at 1600 rpm in 0.1 M KOH at the potential sweep rate of 5 mV s−1. (b) H2O2 yield of CAC, CAC/Fe, NCAC/Fe, NCAC-Zn/Fe, and Pt/C. (c) Eonset, E1/2, and J (at +0.85 V) of the NCAC-Zn/Fe carbon aerogels and Pt/C catalyst. (d) Tafel plots of NCAC-Zn/Fe and Pt/C. (e) ORR performance in acidic media. ORR polarization curves of the NCAC-Zn/Fe and the Pt/C at 1600 rpm on RDE in 0.1 M HClO4 at the potential sweep rate of 5 mV s−1. (f) ORR activity of aerogels derived from agar and gelatin hydrogels. ORR polarization curves of NCAA-Zn/Fe and NCAG-Zn/Fe as well as Pt/C at 1600 rpm on RDE in 0.1 M KOH at the potential sweep rate of 5 mV s−1.

To distinguish the contributions of the Fe center and adjacent nonmetal atoms to the electrocatalytic activity, electrochemical measurements were then carried out with the addition of SCN− as the poisoning species. One can see that upon the addition of 10 mM SCN− into 0.1 M KOH, the E1/2 of NCAC-Zn/Fe exhibited a negative shift of 20 mV (). The relatively mild performance deterioration suggests that in addition to the Fe sites, adjacent nonmetal atoms also play a critical role in driving the catalytic reaction. This is actually in good agreement with the formation of FeN SW moieties in the carbon skeletons, where structural distortion leads to the activation of adjacent C atoms (Figures 3 and ) [8].

Notably, other biomass hydrogels, such as gelatin and agar, can also be used as templates to fabricate nanowrinkled carbon aerogels embedded with single-metal atoms in a similar fashion. The resulting catalysts, NCAG-Zn/Fe and NCAA-Zn/Fe, both displayed excellent catalytic activities towards ORR in alkaline media, with an E1/2 of +0.92 and +0.89 V and an Eonset of +1.12 and 1.10 V, respectively (Figure 4(f) and inset). These results highlight the universality of the synthetic strategy in the preparation of high-performance ORR electrocatalysts ().

2.4. Electrocatalytic Activity towards OER and Zinc-Air Battery Performance

The electrocatalytic activity of the NCAC-Zn/Fe aerogels towards OER was then examined and compared with commercial RuO2 in 1 M KOH with iR correction. From Figure 5(a), one can see that for NCAC-Zn/Fe, an overpotential (η10) of +370 mV was needed to achieve the current density of 10 mA cm−2, a performance comparable to that of commercial RuO2 (η10 = +340 mV). The NCAC-Zn/Fe also exhibits a Tafel slope of 98 mV dec−2, close to that of RuO2 (71 mV dec−2), signifying a favorable OER kinetic (). Thanks to the excellent electrocatalytic performance towards both ORR and OER, the NCAC-Zn/Fe SACs show a low potential difference (ΔE) of only 0.71 V between the OER potential at 10 mA cm−2 (EOER,10) and the ORR potential at 3 mA cm−2 (EORR,3), much smaller than those of bifunctional M−N−C catalysts reported recently in the literature [11, 35, 36].

Figure 5

Zn-air performance tests. (a) Polarization curves for OER and ORR by NCAC-Zn/Fe and Pt/C-RuO2. Inset shows the corresponding ΔE between E and E and the Tafel slope of OER. (b) OCV and (c) power densities of Zn-air batteries assembled by NCAC-Zn/Fe and Pt/C-RuO2. (d) Discharge tests at various current densities (5, 10, 20, and 50 mA cm−2) of NCAC-Zn/Fe and Pt/C-RuO2. (e) Charge-discharge curves of Zn-air batteries assembled by NCAC-Zn/Fe and Pt/C-RuO2 at 10 mA cm−2 for 1100 cycles (400 s per cycle). (f) Comparison of charge-discharge voltage gap for NCAC-Zn/Fe with leading results reported in recent literature [5, 11, 37–40].

With such a remarkable bifunctional performance, the NCAC-Zn/Fe aerogels were tested as the air-cathode for a Zn-air battery, in comparison with those using a commercial Pt/C-RuO2 mixture (mass ratio 1 : 1), along with a Zn plate as the anode. From Figures 5(b)to 5(c), the NCAC-Zn/Fe-Zn-air battery can be seen to show an open-circuit voltage (OCV) of 1.50 V and a maximum power density of 231 mW cm−2, about 6 mV and 20 mW cm−2 higher than those of the Pt/C-RuO2 counterpart. Figure 5(d) shows the corresponding constant current discharge tests at various current densities (5, 10, 20, and 50 mA cm−2) of the two batteries. One can see that the NCAC-Zn/Fe-Zn battery exhibited a much higher discharge voltage within a wide range of current densities (5 to 50 mA cm−2). At the constant current density of 10 mA cm−2, the NCAC-Zn/Fe-Zn battery displayed a stable and optimal potential of 1.36 V for 41 h. By normalizing the energy output to the weight of dissipated Zn, the calculated specific capacity and energy density were estimated to be 780 mAh g−1 and 956 Wh kg−1, respectively, markedly higher than those of Pt/C-RuO2 (). Also, the small charge-discharge voltage gap of the NCAC-Zn/Fe-Zn battery in Figure 5(e) indicates excellent rechargeability (). Impressively, the battery also delivers a stable potential plateau in the charge-discharge test at the constant current density of 10 mA cm−2 during prolonged operation. After 1100 continuous charge-discharge cycles (400 s for each cycle), the NCAC-Zn/Fe-Zn battery still afforded a high round-trip efficiency of 59% and a narrow discharge-recharge voltage gap of 0.79 V, much better than those of Pt/C-RuO2 and other leading oxygen electrocatalysts reported in recent literature (Figure 5(f)) [5, 11, 37–40]. Taken together, these results demonstrate that the NCAC-Zn/Fe aerogels derived from biomass hydrogels can be used as high-performance bifunctional oxygen electrodes for Zn-air batteries, thanks to its high open-circuit voltage, large power density, and superb durability.

3. Conclusion

In this study, a facile, scalable strategy was developed for the preparation of nanowrinkled carbon aerogels embedded with FeN active sites by utilizing biomass hydrogels as the precursors and reactors. The resulting nanowrinkled carbon aerogels (NCAC-Zn/Fe) showed an excellent and reversible ORR/OER electrocatalytic performance with a low voltage gap of only 0.71 V for oxygen electrocatalysis. With the obtained carbon aerogels as the (air) cathode catalysts of a Zn-air battery, the battery exhibited a higher open-circuit voltage, greater power density, and superior durability than that based on a mixture of commercial Pt/C-RuO2 catalysts. First-principles calculation showed that FeN sites in Stone-Wales defect formed by the carbon nanowrinkles were most likely responsible for the excellent electrocatalytic activity. Results from the present study suggest that creating structural distortion of metal sites in carbon matrices can be exploited as an effective strategy for the design and engineering of advanced electrocatalysts based on atomically dispersed metal centers.

4. Materials and Methods

4.1. Reagents

Potassium hydroxide (KOH), iron(II) chloride tetrahydrate (FeCl2·4H2O), SiO2 nanoparticles (15 nm), zinc(II) acetate (Zn(OAc)2), potassium thiocyanate (KSCN), gelatin, and agar were purchased from Aladdin Reagents (Shanghai, China). Perchloric acid (HClO4) and ammonium hydroxide (NH4OH) were purchased from Xiya Reagents (Chengdu, China). Chitosan and Nafion solution were purchased from Sigma-Aldrich (USA). Commercial Pt/C (20 wt%) and high-purity zinc plate (99.999%) were obtained from Johnson Matthey. Polytetrafluoroethylene (PTFE, 60 wt%, D-210C) was purchased from Japan DaJin. All other reagents were analytical grade, and ultrapure water (Mill-Q, 18.3 MΩ cm) was used throughout this study.

4.2. Instrumentation

TEM studies were carried out with a T20 FEI Tecnai G2 instrument. Scanning electron microscopy images were obtained with a Hitachi S-4800 field-emission scanning electron microscope. STEM with EDS studies were performed on a JEOL JEM-ARM200CF with aberration-corrected STEM. Topography, maximum force, adhesion force, and current flow were measured using the Fast Force Mapping technique with an Oxford Instruments Asylum Cypher S AFM housed in an Ar-gas-filled glove box. Raman spectra were acquired with a Renishaw inVia Raman microscope. CD spectra were recorded on Jasco J-815 CD spectrometer (Japan). UV-vis spectra were acquired with a Shimadzu UV-2450 Spectrophotometer (Japan). ICP-OES studies were performed on a SPECTROBLUE SOP instrument. XRD and XPS measurements were carried out on a D/MAX2550 X-Ray Power Diffractometer and a Thermo Fisher-VG Scientific ESCALAB 250Xi X-Ray Photoelectron Spectrometer, respectively. N2 adsorption-desorption isotherms were obtained with a Micromeritics ASAP 2020 Surface Area and Porosity Analyzer. Fe K-edge EXAFS measurements were performed at the Quick-EXAFS Beamline of the Taiwan Photon Source in transmission mode, and the results were analyzed by using the FeNC structural model with the Athena program. An RST 5200F electrochemical workstation (Zhengzhou, China) was used to perform the voltammetric measurements. Rotating Disk Electrode (RDE, Pine Research Instrument) tests were carried out at the rotation rates of 400 to 1600 rpm.

4.3. Synthesis of Catalysts

In a typical reaction, 60 mg of chitosan (CS), 40 mg of SiO2 nanoparticles (15 nm), and 2.56 mL of acetic acid (1%) were placed in screw-cap vial under magnetic stirring for 90 min, into which was added 90 μL of NH3H2O (20 wt%) to adjust the solution pH to 7.0. 200 μL of 0.2 M Fe-phenanthrolene (Fe(PM)32+) and 80 μL of 1 M zinc acetate (Zn(OAc)2) were then added into the above solution under stirring. Sonication treatment for 6 min yielded a hydrogel, which was denoted as CSSi-Zn/FePM (“Si,” “Zn,” and “FePM” stand for SiO2 nanoparticles, Zn2+ ions, and Fe(PM)32+, respectively). The CSSi-Zn/FePM hydrogel obtained above was freeze-dried and then heated to 900°C at the heating rate of 5°C min−1 in an Ar atmosphere (containing 3% H2). After heating at 900°C for 3 h, the sample was cooled down to room temperature and subjected to HF etching to remove SiO2 nanoparticles, affording an Fe-N-codoped carbon aerogel, which was referred to as NCAC-Zn/Fe.

Three control samples were prepared in the same fashion except that only one or two of the starting materials (SiO2 nanoparticles, Zn2+ ions, and Fe(PM)32+) were used to prepare the biomass hydrogel precursors (i.e., CSSi, CSSi/Fe, and CSSi/FePM). The corresponding aerogels were denoted as CAC, CAC/Fe, and NCAC/Fe, respectively.

4.4. Electrochemistry

Electrochemical tests were carried out in a three-electrode electrochemical cell with a graphite rod as the counter electrode and an Ag/AgCl (saturated KCl) electrode as the reference electrode. The Ag/AgCl reference electrode was calibrated against a reversible hydrogen electrode (RHE) and all potentials in the present study were referenced to this RHE. To prepare a catalyst ink, 3 mg of the catalysts obtained above was dispersed in a 475 μL mixed solvent of H2O and ethanol (v : v = 1 : 1) and 25 μL of a Nafion solution (5%) under sonication for 1 h to form a homogeneous dispersion (6 mg mL−1). For ORR tests, the catalyst ink was loaded onto a cleaned glassy carbon electrode at the catalyst loading of 250 μg cm−2 for cyclic voltammetry and 400 μg cm−2 for RDE and RRDE measurements in 0.1 M KOH, respectively. For OER tests, the catalyst was loaded onto a carbon paper at the mass loading of 1.0 mg cm−2 in 1.0 M KOH.

In RDE measurements the disk current density (J) is defined by the Koutecky-Levich (K-L) equation: where J is the limiting current density, J is the kinetic current density, ω is the rotation rate, n is electron transfer number, F is the Faraday constant (96,485 C mol−1), C0 is the O2 concentration in the electrolyte solution (1.2 × 10−6 mol cm−3), and ν is the kinematic viscosity of the electrolyte (0.01 cm2 s−1 for 0.1 M KOH). J can be determined from the intercept of the K-L plot (J−1 vs. ω−1/2).

In RRDE measurements, the H2O2 yield and electron transfer number (n) can be calculated by equation (2). where I and I are the ring current and disk current, respectively, and N is the collection efficiency of the ring electrode (0.37).

4.5. Fabrication of Home-Made Zinc-Air Battery

To make a zinc-air battery, a 6 M KOH solution containing 0.2 M zinc acetate was used as the electrolyte. A zinc plate was used as the anode. The air cathode was composed of three layers, i.e., a gas diffusion layer, a Ni foam layer, and a catalyst layer. The Ni foam was subject to sonication treatment in 0.1 M HCl, H2O, and ethanol consecutively, and vacuum dried at 80°C for 3 h. The catalyst layer was prepared by mixing 60 mg NCAC-Zn/Fe catalyst (or a mixture of 30 mg Pt/C and 30 mg RuO2), 10 mg acetylene black (as a conductive agent), and 30 mg PTFE emulsion. The total thickness of the cathode was ca. 0.4 mm after compression with a manual tablet machine and vacuum dried at 80°C for 3 h.

4.6. DFT Calculations

DFT calculations were carried out by Quantum ESPRESSO which is an open-source planewave code [41]. A two-dimensional supercell was built based on an 8 × 8 unit cell (127-129 atoms in total). For avoiding the interactions between periodic images, the vacuum at the z-axis was set at 14 Å. The ultrasoft pseudopotential was adopted [42]. The kinetic and charge density cutoff were set at 40 and 200 Ry, respectively. The 2 × 2 × 1 Monkhorst-Pack K-point grids were sampled for the supercell. The total energy was converged to 10−3 eV for geometric relaxation. The Marzari-Vanderbilt smearing was adopted with a smearing of 0.01 Ry [43]. The electronic energy and force were converged to 10−8 Ry and 10−4 a.u., respectively. The phonon contribution to zero-point energy and entropy was calculated based on the density functional perturbation theory [44, 45]. STM calculations were carried out based on the Tersoff and Hamann approximation [46] as implemented in the open-source Quantum ESPRESSO package [41] at a bias of -1.0 or +1.0 V, as described in the literature [47].