The commercial high temperature (300-500℃) and pressure (150-200 atm) energy-intensive Haber-Bosch process, a century-old craft for ammonia production, is large scale in global energy consumption (~1% of the annual natural gas) and carbon dioxide emission (at 1.87 ton
CO2/ton
NH3) [
1–
4]. Thus formed ammonia is a fundamental chemical in fertilizers and a hydrogen-rich fuel for the next generation [
5]. An alternative, using renewable energy for NH
3 synthesis under ambient conditions, is electrochemical nitrogen reduction reaction (NRR) [
6–
8]. However, the NRR suffers from low Faradaic efficiency (FE) and NH
3 yield due to the limited solubility, high bond energy (N≡N, 941 kJ/mol), and multi-step reductions involving electron/hydrogen [
9,
10]. The wide use of nitrogen fertilizer caused ground water nitrate pollution. As one of the most widespread nitrogen-containing water-soluble contaminants at varying concentrations (0.001-2 mol/L) in industrial waste streams, nitrate (NO
3-) has the lower bond energy of N=O (204 kJ/mol), which decreases the reaction energy barrier and increases the NH
3 yield [
11,
12].
The NO
3--to-NH
3 conversion is involved in nitrate reduction, as an 8e
- transfer process (NO
3- + 6H
2O + 8e
- → NH
3 + 9OH
- in alkaline electrolyte, 0.69 V
versus the reversible hydrogen electrode (RHE), and a slow process also involving 9 protons [
13–
16]. The byproducts (NO
2, N
2O, N
2) and competitive hydrogen evolution reaction (HER) affect Faradaic efficiency (FE) and NH
3 yield [
17,
18]. Therefore, developing catalysts with improved activity, selectivity, and durability remains academically interesting, technically challenging, and industrially demanding.
Single-atom catalysts (SACs) anchor isolated active metal sites by chemically coordinating/bonding them with surrounding atoms from the supports [
19]. In atomic dispersion, SACs maximize the use of metals, providing even superior activity than their nanoparticle (NP) counterparts [
20–
22]. Recently, iron-based SACs have been widely investigated for NO
3-RR, due to their abundant reserves, low cost, and non-toxic property. Notably, the Haber-Bosch catalysts are also Fe-based [
23–
26]. According to Wu
et al. [
27], the impressive Fe SACs (max FE
NH3 = 75%, Yield
NH3 = 0.46 mmol h
-1 cm
-2) isolated Fe-N
4 active sites to prevent the N-N coupling and decrease the thermodynamic barriers compared with Fe NP, Co SAC, and Ni SAC from density functional theory (DFT) calculations. Zhang
et al. [
28] used O atoms to tune the d-center of Fe, making the FeN
2O
2 structure spontaneously trigger the transformation of
*NOH to
*N, which is the rate-determining reaction based on FeN
4. However, the typical M-C-N type (M = Fe, Co, Cu, Ni …) SACs are limited by linear scaling relations due to the multi-intermediate of the NO
3-RR [
24]. Therefore, tuning the coordination environment is important to boost the performance [
29]. Inspired by the composition of the nitrate reductase (containing Fe-S and Mo-S cofactor) [
30,
31], Murphy
et al. [
32] fabricated a bimetallic FeMo-N-C catalyst and demonstrated that the NO
3- groups were firstly reduced to NO
2- by the Mo sites, then the NO
2- was converted to NH
3 on the Fe sites at 100% FE. Nevertheless, using binders to fix the powder catalysts to the supports, may cover some of the active centers and hinder the efficiency. Molybdenum disulfides (MoS
2), a traditional two-dimensional transition-metal dichalcogenides, can grow firmly on carbon-based conductive substrates, present excellent properties in HER [
33], oxygen reduction reaction (ORR) [
34] and NRR [
35]. Notably, by regulating the defects or doping metals, the effect can be further enhanced [
36,
37]. Hence, MoS
2-based catalysts grown on carbon substrate are potential binder-free materials for electrochemical NO
3-RR. In 2022, Li
et al. [
38] supported single iron atoms on MoS
2 nanosheets, showing NO
3-RR activity. The formed NH
3 yield (431.8 ± 38.6 µg h
-1 cm
-2) was relatively low, but this can be improved by changing the synthesis conditions to adjust the active sites, their coordination environments and loading of Fe SACs.
Herein, we tried to increase the single atom iron loading by an in-situ synthesis technique under hydrothermal condition and report the successful preparation of high-loading single atom Fe doped MoS2 grown on carbon fiber cloth (CFC) for NO3-RR, an ultra-high ammonia yield of 28.59 mg h-1 cm-2 at 360 mA/cm2 partial current density, with the corresponding NH3-FE being 96.65% at -0.73 V vs. RHE under alkaline conditions. The results outperform the MoS2/CFC or CFC. The durability in activity and selectivity was excellent, demonstrated by ten times reuse, with one hour in each cycle. The experiments and DFT calculations unveil the synergistic tandem effect of atomic Fe and the surrounding edge Mo sites, indicating that unsaturated Mo is the primary active site for NO3- reduction, and atomic Fe is the main active hydrogen donor. Besides, atomic Fe tunes the edge Mo sites' coordination environment and electronic structure to inhibit HER and facilitate NH3 desorption.
The Fe
1@MoS
2/CFC and MoS
2/CFC catalysts were fabricated by following the same procedures as reported by the authors previously [
39]. The preparation details are provided in Supporting information. High-resolution transmission electron microscope (HRTEM), energy-dispersive X-ray spectroscopy (EDS), X-ray diffraction (XRD), X-ray photoelectron spectra (XPS), and X-ray fine structure spectroscopy (XAFS) characterized the microstructure and elemental composition of the Fe
1@MoS
2 on CFC. HRTEM displays that the Fe atoms distribute uniformly in nanoflower-like MoS
2 with 1T and 2H phases (Fig. S1 in Supporting information). The XRD shows that the peaks at 10.9°, 34.7°, and 57.2° belong to the (002), (100), and (110) of MoS
2, while the 25.1° and 43.0° come from the CFC (Fig. S2 in Supporting information) [
40–
42]. The XPS spectra (Fig. S3 in Supporting information) illustrate that the atomic Fe existed in Fe-S form, suggesting the positive oxidation states of Fe species in Fe
1@MoS
2/CFC, consistent with the XAFS analysis [
38]. The content of Fe was 16.45 wt% in Fe
1@MoS
2, tested by inductively coupled plasma optical emission spectrometer (ICP-OES).
The electronic structure and coordination of Fe species in Fe
1@MoS
2/CFC were obtained by the Fe K-edge X-ray absorption near-edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) measurements.
Fig. 1a demonstrates that the absorption edge of the sample is located between FeO and Fe
2O
3, illustrating the valence state of Fe is between 2
+ and 3
+ [
43,
44]. The Fourier transform of EXAFS spectra (FT-EXAFS) (
Fig. 1b) exhibits the Fe-S interaction in Fe
1@MoS
2/CFC. No Fe-Fe bands appear at 2.2 Å, suggesting the Fe species are atomically isolated by the surrounding S [
45–
48]. The coordination number (CN) and geometric configuration were explored by the EXAFS fitting spectra.
Figs. 1c and
d show the match between the experiments and fitting curves in
R- and
K-spaces. The Fe species in Fe
1@MoS
2/CFC have three coordinating interactions at 2.01 Å (with a CN of 5.30), at 2.29 Å (with a CN of 2.40), and 3.05 Å (with a CN of 1.20), corresponding to Fe-C, Fe-S, and Fe-S-Fe, respectively (Table S1 in Supporting information). The wavelet transforms of EXAFS (
Fig. 1e) was employed to observe the coordination environment of Fe species in Fe
1@MoS
2/CFC. The maximum intensity appears at ~5.1 Å
-1, corresponding to Fe-C and Fe-S. No Fe-metal exists, confirming the atomic-level distribution of Fe species [
49]. LSV (linear sweep voltammetry) was conducted to investigate the NO
3-RR activity of Fe
1@MoS
2/CFC in Ar-saturated electrolytes with 1 mol/L KOH. The scan rate for LSV was 5 mV/s. The largest current density was obtained when the concentration of NO
3- was 0.6 mol/L, and it was enhanced as the magnetic stirring rate up to 500 rpm (Figs. S4 and S5 in Supporting information). The impacts of NO
3- concentration (0.1, 0.3, 0.6, and 1 mol/L) were further tested. The NH
3 yield rate and FE increased with the NO
3- concentration from 0.1 mol/L to 0.6 mol/L, while slightly decreasing in 1 mol/L (Fig. S6 in Supporting information), corresponding to the trend observed in Fig. S4. Therefore, we chose 0.6 mol/L and 500 rpm for NO
3-RR in subsequent tests.
Fig. 2a shows the effect of atomic Fe on NO
3-RR and comparison of the electrochemical activity of Fe
1@MoS
2/CFC, MoS
2/CFC, and CFC. The Fe
1@MoS
2/CFC exhibited an onset potential for nitrate reduction at around 0.06 V, more positive than the MoS
2/CFC (-0.03V), and displayed the largest current density among the three samples. Moreover, in
Fig. 2b, without NO
3-, no activity is observed until -0.25 V, which is the onset of the HER, indicating the suppression of the competitive reaction [
50]. According to previous studies, the current peak near 0 V is assigned to the reduction of
*NO
3 (adsorbed NO
3-) to
*NO
2, and the peak at around -0.1 V is mainly attributed to the conversion of
*NO
2 to
*NH
3 [
51,
52].
Figs. 2a and
b demonstrate that the atomic Fe reduced the overpotential of the NO
3-RR and increased the selectivity of the catalyst for NH
3.
The NO
3-RR performances of the Fe
1@MoS
2/CFC at different potentials are shown in
Fig. 2c. The NH
3 yield rate increased with the potential from -0.33 V to -0.73 V and reached 28.59 mg h
-1 cm
-2, but decreased at more negative potentials due to the competitive HER reaction. Additionally, the Fe
1@MoS
2/CFC showed high NH
3-FE above 93% at -0.33 V to -0.83 V, and the maximum NH
3-FE (99%) at -0.43 V. Notably, the NH
3 yield at 18.07 mg h
-1 cm
-2 (Fe
1@MoS
2/CFC) and 90.02% FE in 0.1 mol/L NO
3-, indicate that it operates effectively with a relatively low NO
3- concentration (Fig. S6).
The NO
3-RR activity of the Fe
1@MoS
2/CFC, MoS
2/CFC, and CFC was compared at -0.73 V.
Fig. 2d indicates that the NH
3 yield rate and FE of Fe
1@MoS
2/CFC are 2.64 and 1.16 times higher than MoS
2/CFC (10.85 mg h
-1 cm
-2, 83.61%). If pure CFC is used, only 4.48 mg h
-1 cm
-2 and 73.27% were obtained. Moreover, the NH
3 partial current density delivered by the Fe
1@MoS
2/CFC, MoS
2/CFC, and CFC were 360 mA/cm
2, 135 mA/cm
2, and 65 mA/cm
2, respectively (Fig. S7 in Supporting information).
Fig. 2d and Fig. S7 demonstrate that the atomic Fe promoted activity and selectivity to the NO
3-RR. The content of Fe in the Fe
1@MoS
2/CFC increased with the corresponding ion concentrations in the precursor as expected (Table S2 in Supporting information). Besides, the yield rate and FE of NH
3 improved with the Fe ion precursor concentration from 0.005 mol/L to 0.04 mol/L, due to the more abundant active sites (Fig. S8 in Supporting information). With the absence of nitrate ions or external potential, only very little ammonia was produced (< 0.2 mg h
-1 cm
-2), confirming that the NH
3 was mainly synthesized
via the electrocatalytic NO
3-RR process (Fig. S9 in Supporting information). The durability was tested by repeating ten times of experiments each operating in 1 h.
Fig. 2e shows that Fe
1@MoS
2/CFC has insignificant decreases in NH
3 yield (-9.6%) and FE (-4.7%). The NO
3-RR performance is compared with the recent reports in
Fig. 2f (more details can be found in Table S3 in Supporting information). The Fe
1@MoS
2/CFC has superior NH
3 yield and FE, due to the binder-free electrode being advantageous for its good stability, low-cost, and numerous exposed active sites.
In-situ Raman spectra of Fe
1@MoS
2/CFC and MoS
2/CFC at various applied potentials during the NO
3-RR are displayed in
Figs. 3a and
b, respectively. The peak at 1049 cm
-1 is associated with symmetric stretching of NO
3-, suggesting the
*NO
3 on the surface of the electrodes [
53]. The Raman bands at around 1300 cm
-1 and 1600 cm
-1 are attributed to the D and G bands of the carbon substrates [
54]. The weak signal that appears at 819 cm
-1 is ascribed to Mo
2-O vibration because of the partial oxidation of Mo species under the laser, consistent with the characteristic peak of the electrodes (Figs. S10 and S11 in Supporting information) [
55]. In
Fig. 3a, the peak at 670 cm
-1 demonstrated that the partial atomic Fe in Fe
1@MoS
2/CFC exists as the ferric iron species in alkaline solutions [
56]. The peak intensity decreased with the applied reduction potential at 0.2 V to 0 V, indicating that the ferric iron species is reducing to a relatively low chemical state by external electricity. However, when the potential shifts to more negative, the characteristic peak increases again, suggesting that the low chemical state Fe is oxidized to Fe
3+ again, indicating that atomic Fe participated in the reaction and Fe
1@MoS
2/CFC had dynamic stability [
57].
To unveil the role of atomic Fe and NO
3-RR mechanism, the Gibbs free energy of water dissociation and NO
3--to-NH
3 pathway with deoxidation (
*NO
3 →
*NO
2 →
*NO →
*N) and hydrogenation (
*N →
*NH →
*NH
2 →
*NH
3) were calculated using DFT, under a potential of 0 V and pH 13.5 (
Figs. 3c-
f) [
58,
59]. The adsorption configurations of different intermediates on the Fe/Mo site of Fe
1@MoS
2 and MoS
2 are provided (Figs. S12-S17 in Supporting information).
Figs. 3d and
e show that the adsorption energy of
*NO
3 (Δ
GNO3) on the Mo site is more robust than Fe, demonstrating that unsaturated Mo on the edge of Fe
1@MoS
2 is thermodynamically more favorable to NO
3-RR. Meanwhile, the Δ
GNO3 (-1.804 eV) on Mo site in Fe
1@MoS
2 is higher than Δ
GH2O (-0.32 eV), indicating that NO
3- is more easily adsorbed on Mo site than H
2O and explaining the high selectivity of Fe
1@MoS
2 for NO
3-RR (
Fig. 3c) [
60,
61]. Similarly, we found water splitting on atomic Fe priority by comparing the Δ
GNO3 (-0.668 eV) and Δ
GH2O (-0.782 eV). Moreover,
Fig. 3c shows that the Mo site has an extraordinary adsorption ability for active hydrogen (
*H), promoting the HO-H cleaving on the surrounding Fe sites [
62]. Furthermore, for the competitive HER, the Mo site in Fe
1@MoS
2 has the most stable ΔG
H to suppress the H
2 formation [
63].
EPR spectroscopy probes that the Fe
1@MoS
2/CFC exhibits the strongest DMPO-H signals with high
*H formation capacity in the electrolyte without NO
3- (Fig. S18 in Supporting information). In the presence of NO
3-, the DMPO-
*H signals surrounding Fe
1@MoS
2/CFC are hardly detected, indicating that the
*H is consumed quickly by NO
3-RR. However, for MoS
2/CFC, the DMPO-
*H still exists, representing a sluggish reaction rate (Fig. S19 in Supporting information) [
64].
From the overall NO
3-RR steps displayed in
Figs. 3d-
f, the Mo site in Fe
1@MoS
2 has the most vigorous
*NO
3 adsorption energy and lowest NH
3 desorption energy barrier, the rate-determining step (RDS), compared with the pure MoS
2 [
65,
66]. In addition, the rapid electron transfer efficiency and fast reaction kinetics of Fe
1@MoS
2/CFC were identified by their smaller Tafel slope of 321 mV/dec than those of MoS
2/CFC (341 mV/dec) and CFC (497 mV/dec) (Fig. S20 in Supporting information) [
67].
Therefore, the crucial function of atomic Fe has been summarized: (1) enhancing the H
2O adsorption and supplying sufficient active hydrogen for the Mo site to accelerate the NO
3-RR [
67,
68]; (2) regulating the coordination environment and electronic structure of the surrounding Mo atom, thereby enhancing the adsorption capacity of the
*NO
3 and partial nitrogen-containing intermediates, making NH
3 desorption easier, and promoting the reactivation of the active sites.
The in-situ grown catalysts, the high-load atomic Fe in MoS2 on CFC, show superior electrocatalytic NO3-RR activity for ammonia synthesis under ambient conditions. Using Fe1@MoS2/CFC, at a rate of 28.59 mg h-1 cm-2, NH3 is formed, 96.65% FENH3 at -0.73 V. The electrode has good durability with insignificant degradation after ten repeated times. The synergistic tandem effect of atomic Fe and surrounding Mo edge sites boosted the NO3-RR performance in activity and selectivity because of the adequate supply of active hydrogen, enhanced electron transfer, increased nitrate adsorption, and decreased free energy of ammonia desorption, making it promising in future applications in ammonia synthesis.
PDF
Full
66
37
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn
