1 Introduction

Ammonia (NH3) is an important large-scale industrial product in the fertilizer industry, which has attracted widespread attention as one of the most promising low-carbon energy carriers with low liquefaction pressure and high hydrogen density [1,2,3,4,5,6,7,8]. Industrially, NH3 synthesis is still dominated by the long-standing Harber–Bosch process under harsh conditions, which consumes approximately 1.4% of annual energy consumption and approximately 3% of global carbon dioxide (CO2) emissions [9,10,11,12]. Electrocatalytic N2 reduction reaction (NRR) is generally considered an energy-efficient and sustainable process for synthesizing NH3 at ambient conditions [13, 14]. However, the NH3 selectivity (SNH3) and yield rate (RNH3) are greatly hindered by the high dissociation energy of N≡N tripe bond (941 kJ mol−1), poor solubility of N2 in the electrolyte and the competitive hydrogen evolution reaction (HER) [15, 16]. Compared with NRR, the electrocatalytic nitrate reduction reaction (NitRR) to NH3 is not limited by the low solubility of N2 in aqueous environment, and its thermodynamics is more advantageous because the dissociation energy of the N = O bond (204 kJ mol−1) is much lower than the N≡N tripe bond (941 kJ mol−1) [17, 18]. It is worth noting that the nitrate (NO3−) widely exists in industrial and agricultural wastewater, posing a real potential threat to human health and natural balance [19,20,21,22,23]. Therefore, converting nitrate into valuable and recyclable NH3 is a frontier field that requires in-depth research. However, in the NitRR process, the competitive HER and the complex eight-electron reduction processes hinder the faradaic efficiency (FE) and selectivity of NH3 [24,25,26,27]. Thus, there is an urgent need for efficient NitRR catalysts with both activity and selectivity simultaneously.

Previous studies have reported that due to the appropriate energy and symmetric 3d orbitals of Mn, the Mn–O site has the ability to catalyze NRR, which is beneficial for the adsorption and activation of N2 molecules [28, 29]. In addition, benefiting from the minimum metal sizes, atomically dispersed single-atom catalysts (SACs) are expected to have great potential to improve and may therefore convert NO3− into NH3 at an acceptable overpotential. Meanwhile, the uniform active sites within atomically dispersed metal catalysts ensure high selectivity, thus ensuring a satisfactory FE [30,31,32,33].

Inspired by the above breakthroughs and our previous research on NitRR and SACs [34,35,36,37], we have rationally modulated oxygen (O)-coordinated single-atom Mn catalyst (Mn–O–C) with Mn–(O–C2)4 as a novel NitRR active coordination configuration. The bacterial cellulose (BC) with rich oxygen functional groups is innovatively utilized as Mn2+ impregnation regulator and the precursor to simultaneously derive the carbon support (CBC) and anchor Mn single atoms to CBC via Mn–O coordination bonds during a facile carbothermal reduction process. The resultant Mn–O–C achieves a high FE of 89.0 ± 3.8% at − 0.5 V (vs. RHE) and a desirable RNH3 of 1476.9 ± 62.6 μg h−1 cm−2 at − 0.7 V (vs. RHE) under ambient conditions. Our density functional theory (DFT) calculations reveal that introduction of Mn–(O–C2)4 sites renders NO3− chemisorption, activates the hydrogenation of adsorbed NO3− and suppresses HER, ultimately enhancing the selectivity of NitRR.

2 Experimental Section

2.1 Preparation of Mn NPs/CBC and Mn–O-C

BC pellicle was frozen by liquid nitrogen and freeze-dried in a bulk tray dryer at a sublimating temperature of − 75 °C and a pressure of 0.01 mbar for 48 h. The pre-treated BC was used as the adsorbent to controllably impregnate Mn2+. The BC (1.0 g) was immersed in a 400-mL solution containing 16 mmol of MnSO4·4H2O (concentration of Mn2+: 40 mmol L−1) for 6.0 h. The BC with adsorbed Mn2+ was adequately washed with deionized water, freeze-dried and subjected to the pyrolytic treatment at 360 °C for 2.0 h and then 700 °C for 3.0 h under Ar atmosphere to carbothermally reduce the adsorbed Mn2+ on BC to metallic Mn NPs and simultaneously carbonize BC into graphitic carbon (CBC). The resultant Mn NPs/CBC was adequately washed with deionized water and ethanol and dried at 60 °C under vacuum for 6.0 h, followed by a refluxing acid-etching process using 2.0 M H2SO4 at 120 °C for 6.0 h to remove metallic Mn NPs. The resultant Mn–O–C was adequately washed with distilled water and ethanol and dried under vacuum for 12 h.

2.2 Electrochemical Measurements

The electrochemical measurements were carried out in a customized H-type glass cell separated by Nafion 211 membrane at room temperature. A CHI 760E electrochemical workstation (CH Instrumental Corporation, Shanghai, China) was used to record the electrochemical response. The Mn–O–C sample on carbon paper, saturated Ag/AgCl electrode and platinum mesh was used as the working electrode, reference electrode and counter electrode, respectively. Before use, the Nafion 211 membrane was successively treated at 80 °C in H2O2 (5.0 wt%) and 0.1 M H2SO4 aqueous solutions and thoroughly rinsed with the deionized water. The working electrode was prepared as follows: 2.5 mg of the targeted electrocatalyst was firstly dispersed in 950 µL of absolute ethanol and 50 µL of Nafion solution (5.0 wt%) under sonication for 30 min to form a homogeneous ink. One hundred microliters of ink was loaded onto a carbon paper (1.0 × 1.0 cm2) and dried under ambient conditions for 40 min before use. The surface area of carbon paper was 0.25 mg cm−2 as the working electrode. The as-fabricated electrode was treated in Ar-saturated 0.1 M K2SO4 + 1000 ppm N-KNO3 solution to activate before being used. It should be noted that the reported NH3 yield in this work is the NH3 product collected only from cathodic compartment. In this work, all measured potentials versus Ag/AgCl were converted to the potentials versus RHE (ERHE) according to the following equation:

$$ E_{{{\text{RHE}}}} = E_{{{\text{Ag}}/{\text{AgCl}}}} + 0.059{\text{pH}} + E_{{{\text{Ag}}/{\text{AgCl}}}}^{ \circ } $$
(1)

where EAg/AgCl is the equilibrium potential under standard conditions and EoAg/AgCl = 0.1967 V versus RHE at 25 °C.

The commercial gas diffusion electrode (GDE) consisted of a working electrode, Nafion 211 membrane and platinum foil anode. The effective catalytic area was 1.0 cm2 with a catalyst loading of 0.25 mg cm−2. Using Ar-saturated 0.1 M K2SO4 + 1000 ppm N-KNO3 as both the flowing cathode and anode electrolyte, the synthesized NH3 can be transported out at a flexible rate. The electrocatalytic nitrate reduction reaction was tested at the constant current density of 50, 100 and 150 mA cm−2. Each experiment was run in triplicate, and the average values with error bars are presented.

The yielded ammonia and the content of nitrite in the electrolyte were measured by colorimetric methods [34, 36, 37].

2.3 In situ Raman and FT-IR Spectroscopy Measurements

For the in situ Raman spectroscopy tests, the samples were recorded on a RXN1-785 Raman spectrometer (Analytik Jena AG, excited wavelength of 785 nm) connected with the CHI 660E electrochemical workstation. The in situ attenuated total reflection surface-enhanced infrared adsorption spectroscopy (ATR-SEIRAS) was performed on a FT-IR spectrometer (Nicolet iS50, Thermo Scientific) equipped with an MCT-A detector with silicon as the prismatic window. Fist, Mn–O–C ink (pure ethanol as a dispersant) was carefully dropped on the surface of gold film, which was chemically deposited on the surface of the silicon prismatic before each experiment. Then, the deposited silicon prismatic served as the working electrode. The platinum mesh and Ag/AgCl electrode containing saturated KCl solution were used as the counter and reference electrodes, respectively. The 0.1 M K2SO4 + 1000 ppm N-KNO3 solution was employed as the electrolyte. Each infrared absorption spectrum was acquired by averaging 128 scans at a resolution of 4.0 cm−1. The background spectrum of the catalyst electrode was acquired at an open-circuit voltage before each systemic measurement, and the measured potential ranges of the electrocatalytic NitRR were − 0.2 to − 0.7 V versus RHE with an interval of 0.1 V.

3 Results and Discussion

3.1 Synthesis and Characterizations of Mn–O–C

Similar to the synthetic method we previously reported [34,35,36], bacterial cellulose (BC) with rich O-containing functional groups and nanofiber network structures (Fig. S1) was used as the adsorption regulator to controllably adsorb Mn2+, followed by freeze-drying, high-temperature pyrolysis and acid washing treatment, to achieve oxygen-coordinated single Mn sites with certain loading supported on BC-converted graphitic carbon (Mn–O–C) (Fig. 1a). As revealed by the transmission electron microscopy (TEM), the as-prepared Mn–O–C still maintains the initial fiber-like aggregation morphology after carbonization fixation and acid-etching process (Fig. 1b). Further observations by high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) indicate that no metal nanoparticles had formed (Fig. 1c). The dominant diffraction peaks corresponding to carbon in the X-ray diffraction (XRD) pattern together with the broad D and G bonds in the Raman spectrum (Fig. S2) are consistent with a partially graphitized carbon structures, which is confirmed by the selected area electron diffraction (SAED) image (Fig. 1d). The enlarged aberration-corrected HAADF-STEM images of Mn–O–C confirm that the Mn atoms were atomically dispersed on the CBC support (Fig. 1e, f). Homogeneously dispersed single Mn sites can clearly be observed according to the different intensity profiles (Fig. S3). In addition, the elemental mappings reveal the homogenous distribution of C, O and Mn over the whole CBC support (Fig. 1g). The actual loading of Mn in Mn–O–C sample was measured to be 1.01% by the inductively coupled plasma atomic emission spectrometer (ICP-AES) analysis. The N2 physisorption isotherm and pore size distribution demonstrate its high Brunauer–Emmett–Teller (BET) specific surface area of 496.8 m2 g−1 and micro- and mesoporous structure (Fig. S4), which are beneficial to exposure of the isolated Mn sites and the mass transport of electrolytes during electrolysis [34,35,36].

Fig. 1
figure 1

a Schematic for diagram illustrating synthetic procedure of Mn–O–C. b TEM image, c HAADF-STEM image, and d SAED pattern of Mn–O–C. e, f Enlarged aberration-corrected HAADF-STEM images of Mn–O–C. g Elemental mapping of Mn–O–C

Full size image

3.2 Confirmation of Mn–(O–C2)4 Single-atom Site

X-ray photoelectron spectroscopy (XPS) was then used to characterize the Mn–O–C catalyst surface and composition (Fig. S5). As shown in Fig. S5a, the survey XPS spectra show the presence of Mn, O and C elements in the Mn–O–C sample. The corresponding elements contents show that the atomic percentages (at%) of Mn, O and C are 0.23%, 10.3% and 89.47%, respectively (Table S1). The high-resolution C 1s and O 1s XPS spectra (Fig. S5b, c) indicate the existence of rich oxygen functional groups and the formation of Mn–O bonds in Mn–O–C [34,35,36], suggesting that Mn single atoms could be anchored to the graphitic carbon substrate via Mn–O coordination bonds. The high-resolution Mn 2p XPS spectrum (Fig. 2a) shows the peaks located at 641.3, 642.8 and 645.5 eV, corresponding to the Mn2+, Mn3+ and specific shakeup satellite peak [38, 39].

Fig. 2
figure 2

a High-resolution XPS spectra of Mn 2p of Mn–O–C. b Mn K edge XANES spectra and c k3-weighted FT-EXAFS spectra of Mn–O–C and references. d, e Mn K edge EXAFS fitting curves of Mn–O–C at R space and k space. f The proposed Mn–(O–C2)4 coordination configuration in Mn–O–C (C, O and Mn atoms are in color of gray, red and orange, respectively). g–j Mn K edge WT-EXAFS of Mn–O–C and references

Full size image

To further determine the valence state of Mn, the X-ray absorption spectroscopy near-edge structure (XANES) spectra were employed. As shown in Fig. 2b, the XANES spectra show the adsorption peak position for Mn–O–C is situated between those for the MnO and MnO2, further revealing its typical electronic structure of Mnδ+ (2 < δ < 4), which is in agreement with the above XPS analysis [38, 39]. The extended X-ray absorption fine structure (EXAFS) curves (Fig. 2c) show the characteristic peak of the Mn–Mn bond at approximately 2.30 Å for the Mn foil and Mn–O bonds at approximately 1.40–1.70 Å for MnO and MnO2. The predominant peak at approximately 1.65 Å ascribed to the Mn–O scattering pathway can be observed (Fig. 2c), and there was no peak corresponding to the Mn–Mn scattering pathway, indicating the existence of the single Mn atomic structure. Based on the EXAFS spectra fitting, the coordination numbers of Mn–O are given by CNMn-O = 4.0 in Mn–O–C sample (Table S2). The simulated EXAFS spectra are well fitted with measured curves (Fig. 2d, e), verifying the rationalization of optimized structure models, confirming the formation of Mn–(O–C2)4 sites in the Mn–O–C sample (Fig. 2f). Meanwhile, the wavelet transforms (WT) EXAFS oscillation of Mn foil, MnO, MnO2 and Mn–O–C was further analyzed to study the atomic configuration. As shown in Fig. 2g-j, the WT contour plots of the Mn–O–C sample display the intensity maximum at approximately 7.0 Å−1, corresponding to the Mn–O coordination bond. Moreover, the Mn–Mn signals at approximately 8.0 Å−1 are absent as compared to the MnO, MnO2 and Mn foil references. The above results clearly prove the formation of precisely regulated Mn–(O–C2)4 sites, which may endow the Mn–O–C catalyst with high activity and selectivity in catalytic NitRR [34,35,36].

3.3 NitRR Performance Evaluation

The NitRR performance of Mn–O–C was evaluated in a typical H-type electrolytic cell under ambient conditions in Ar-saturated 0.1 M K2SO4 + 1000 ppm N-KNO3. We first performed the linear sweep voltammetry (LSV) in Ar-saturated 0.1 M K2SO4 electrolyte with and without KNO3 to study the NitRR catalytic activity of Mn–O–C (Fig. 3a). The obviously enhanced current density under the same potential confirms that nitrate can be effectively reduced by Mn–O–C. It should be noted that before electrocatalytic NitRR by a potentiostatic method, Ar gas was purged to the electrolyte for at least 20 minutes. Figure 3b shows the dependence of NH3 yield rate (RNH3) and FE on the applied potentials for 2.0 h electrolysis from − 0.2 to − 0.7 V (vs. RHE). The calculated RNH3 and FEs based on three repeated experiments are derived from the recorded chronoamperometric curves (Fig. S6a) under different potentials with the yielded NH3 being determined by the indophenol blue method (Figs. S6b and S7). The well-established colorimetric methods (Fig. S8) were used to quantify the unreacted NO3− in the electrolytes [40, 41]. When applying the more negative potentials from − 0.2 to − 0.7 V (vs. RHE), NH3 yield rate can gradually increase. The Mn–O–C electrocatalyst achieves a high RNH3 of 1476.9 ± 62.6 μg h−1 cm−2 at − 0.7 V (vs. RHE) and a FE of 89.0 ± 3.8% at − 0.5 V (vs. RHE) under ambient conditions, which is comparable to most of recently reported NitRR electrocatalysts (Table S3). Note that the decreased FE here (from − 0.6 V (vs. RHE)) is due to the enhanced competition of the HER at more negative applied potentials [42, 43]. Although the electrodynamic potential of NO3− to NO2− is higher than that of NO3− to NH3 [40,41,42,43], NO2− is an essential intermediate of NitRR and is also quantified by colorimetric method (Fig. S9) [40,41,42,43]. As shown in Fig. S10, NO2− is almost detectable for the Mn–O–C in all potential ranges, while the FE of NO2− is much less than NH3, which indicates that NH3 is the main product of NitRR. The selectivity of NH3 (SNH3) of Mn–O–C at different applied potentials was further studied, the RNH3 shows the same trend with the increase of potential, and the highest SNH3 was 86.1 ± 0.3% (Fig. 3c).

Fig. 3
figure 3

a LSV curves of the Mn–O–C in Ar-saturated 0.1 M K2SO4 and 0.1 M K2SO4 + 1000 ppm N-KNO3 electrolyte. b RNH3 and FE at each given potential. c SNH3 of Mn–O–C obtained at different potentials. d 1H NMR spectra of 14NH4+ and 15NH4+ standards, and the resultant samples from Mn–O–C-catalyzed NitRR using 14NO3− and 15NO3− as nitrogen source, respectively. e Comparison of RNH3 from Mn–O–C using 15NO3− as nitrogen source at different potentials. f Durability test on the applied potential for Mn–O–C. g Cycling stability test of Mn–O–C at − 0.5 V (vs. RHE) for 10 cycles with 2.0-h NitRR period per cycle. h Corresponding RNH3 and FE of each testing cycle. i RNH3 and FE of CBC, Mn NPs/CBC and Mn–O–C at − 0.5 V (vs. RHE) for 2.0 h NitRR

Full size image

The 1H nuclear magnetic resonance (NMR) analysis was subsequently carried out to confirm the yielded NH3 exclusively resulted from the Mn–O–C-catalyzed NitRR [40,41,42,43]. The experiments were performed using 14NO3− and 15NO3− as nitrogen source in Ar-saturated 0.1 M K2SO4 electrolyte at − 0.5 V (vs. RHE) over a 2.0-h NitRR period. Figure 3d displays the 1H NMR spectra of the standards and the yielded 14NH4+ and 15NH4+ in NitRR samples, confirming that the yielded NH3 is indeed originated from the Mn–O–C-catalyzed NitRR. The yielded 15NH4+ and concentrations were quantified by both the indophenol blue method (Fig. S7) and 1H NMR analysis (Fig. S11). As shown in Figs. 3e and S12a, the yielded 15NH4+ concentrations determined by 1H NMR method ranged from − 0.4 to − 0.6 V (vs. RHE), very closely approximated to those determined by the indophenol blue method. Such closely approximated NH3 yield rates from 15NO3− confirmed by the two analytical methods infer that the yielded NH3 is indeed originated from the Mn–O–C-catalyzed NitRR.

The control experiments were subsequently conducted to eliminate the environmental interference on the yielded NH3 (Fig. S12b). Only ignorable NH3 concentration can be detected when the experiments were carried out using Ar-saturated 0.1 M K2SO4 with 15NO3− but without applied potential (OCP). The detected NH3 concentration is also ignorable using the Mn–O–C catalyst in Ar-saturated 0.1 M K2SO4 at − 0.5 V (vs. RHE) for 2.0 h (Fig. S12b). The above experimental results eliminate any noticeable environmental interference to the yielded NH3 from NitRR. The durability of Mn–O–C for NitRR was evaluated using Ar-saturated 0.1 M K2SO4 electrolyte at an applied potential of − 0.5 V (vs. RHE) for 20 h (Fig. 3f). It can be seen that the chronoamperometric profile exhibits a slight decrease in the current density during the entire test period and ~ 3.5% loss of FE can be achieved from 2.0 to 20 h, demonstrating high stability of Mn–O–C. The reusability measurement results also indicate that at − 0.5 V (vs. RHE) with a reaction time of 2.0 h for each cycle, no noticeable decay could be observed in the total current density during 10 consecutive cycles (Fig. 3g). Moreover, ignorable change in RNH3 and FE (Fig. 3h) further confirms a superior reusability of Mn–O–C. After 10 NitRR measurement cycles, the used Mn–O–C still exhibits atomically dispersed nature of Mn (Figs. S13–S15), no metallic Mn or Mn-related oxides can be detected (Fig. S16), indicating high structural stability of the O–C-coordinated Mn–O–C. In contrast, the catalytic performance of CBC (Fig. S17), Mn NPs/CBC (Fig. S18) and Mn–O–C was further evaluated in Ar-saturated 0.1 M K2SO4 + 1000 ppm N-KNO3 electrolyte. At − 0.5 V (vs. RHE) for 2.0 h of reaction, no NH3 product can be detected for CBC, while the Mn NPs/CBC can give a RNH3 of 252.8 ± 12.9 μg h−1 cm−2 and a FE of 15.5 ± 0.8%, which was much less than Mn–O–C (RNH3: 966.1 ± 40.9 μg h−1 cm−2 and FE: 89.0 ± 3.8%), implying that the CBC and Mn NPs/CBC are comparatively unfavorable for NitRR (Fig. 3i).

To further evaluate the electrocatalytic NitRR performance, the RNH3 of Mn–O–C was determined using a commercial gas diffusion electrode (GDE) (Figs. 4a and S19) [44, 45]. Firstly, LSV of Mn–O–C was conducted in Ar-saturated 0.1 M K2SO4 solution with and without NO3−, respectively. As shown in Fig. 4b, the current density enhancement with NO3− over the one without NO3− indicates that Mn–O–C is active for NitRR catalysis in the flow cell. Moreover, the LSV of Mn–O–C tested in the presence of NO3− exhibits a remarkable reduction peak at − 0.4 V (vs. RHE) (Fig. 4b), which may be due to the electrochemical reduction of NO3− [46]. It is worth noting that the significant decrease in current density in the flow cell compared to the H cell (Fig. 3a) may be due to the lack of strong gas (e.g., N2, O2, CO2)–liquid contact significantly enhanced by GDE, and the flow cell can significantly inhibit the competitive HER on the Mn–O–C surface under high current density by continuously cycling the main product ammonia and by-product hydrogen away from the electrode, while also greatly increasing the distance required for nitrate diffusion, lowering the maximum obtainable current densities. The dependence of RNH3 and FE at the constant current density of 50, 100 and 150 mA cm−2 is shown in Fig. 4c. As expected, Mn–O–C shows the highest of RNH3 of 3706.7 ± 552.0 μg h−1 cm−2 at a current density of 100 mA cm−2, twofold lager than on the H cell (1476.9 ± 62.6 μg h−1 cm−2). As an essential intermediate of NitRR, NO2− is also found for Mn–O–C at different current densities (Fig. S20). In order to validate the electrocatalytic NitRR mechanism of Mn–O–C, we utilized in situ ATR-SEIRAS to monitor the evolution of NitRR intermediates [47] and the experimental setup and cell are displayed in Figs. 4d and S21. Figure 4e shows the infrared signals when the in situ electrocatalytic NitRR on Mn–O–C during the negative scan from − 0.2 to − 0.7 V (vs. RHE). At the applied potential, the significantly enhanced infrared peaks at 1352 cm−1 were assigned to N–O asymmetric stretching vibration of NO3− [48, 49], which indicates the activation and consumption of NO3− catalyzed by Mn–O–C. Meanwhile, the infrared peaks at 1225 cm−1 were attributed to N–O antisymmetric stretching vibration of NO2− [48, 50], confirming the formation of by-product NO2− from NitRR, which is also consistent with the electrochemical experimental results. Interestingly, the vibration bands of adsorbed NO in bridge adsorption mode were detected at 1528 cm−1 [48, 51]. In addition, the gradually enhanced infrared peaks at 1457 cm−1, which can be ascribed to N–H bending vibration of NH4+ [50, 52]. Based on the in situ ATR-SEIRAS analysis, we proposed the following pathway for the NitRR on Mn–O–C surface: NO3−  → *NO3 → *NO2 → *NO → *NH2OH → *NH3. We further conducted in situ ATR-SEIRAS measurement at − 0.5 V (vs. RHE) for 1.0-h NitRR. The intensity of the characteristic peaks at 1225, 1352, 1457 and 1528 cm−1 is increased obviously from 10 to 60 min, implying that the NitRR takes place gradually with reaction time under the given electrocatalytic conditions (Figs. 4f and S22). We also performed in situ Raman spectroscopy tests in Ar-saturated 0.1 M K2SO4 solution with NO3− electrolyte (Fig. 4g). As shown in Fig. 4h, when OCP was applied, two main Raman peaks at 989 and 1056 cm−1 were attributed to SO42− and NO3−, respectively. As the applied potential is scanned from OCP to − 0.7 V (vs. RHE), a new signal appears at 1022 cm−1 which is in accordance with NH4+ standard, confirming the formation of NH3 from NitRR. We further investigated the electrochemical NitRR at − 0.5 V (vs. RHE) for 1.0 h by the in situ Raman spectroscopy measurements (Fig. 4i). It was also found that the Raman intensity of the peaks belonging to NH4+ at around 1022 cm−1 gradually enhanced from 10 to 60 min. The above in situ ATR-SEIRAS and Raman spectroscopy measurements confirm the successful realization of electrocatalytic NitRR to NH3 over Mn–O–C in this work.

Fig. 4
figure 4

a Physical photograph of flow cell reactor for electrocatalytic NitRR. b LSV curves of the Mn–O–C in the flow cell with and without NO3−. c RNH3 and FE at the constant current density of 50, 100 and 150 mA cm−2. d Physical photograph of in situ ATR-SEIRAS reactor for electrocatalytic NitRR. e In situ ATR-SEIRAS spectra of electrocatalytic NitRR on Mn–O–C at different potentials. f In situ ATR-SEIRAS spectra of Mn–O–C during electrocatalytic NitRR at − 0.5 V (vs. RHE) for 1.0 h. g The physical photograph of in situ Raman reactor for electrocatalytic NitRR. h In situ Raman spectra of electrocatalytic NitRR on Mn–O–C at different potentials. i In situ Raman spectra of Mn–O–C during electrocatalytic NitRR at − 0.5 V (vs. RHE) for 1.0 h

Full size image

3.4 DFT Prediction of NitRR Activity of Mn–(O–C2)4 Site

Based on the aforementioned in situ ATR-SEIRAS results, we performed DFT calculations to understand the electrochemical NitRR mechanism on Mn–O–C catalyst. Figure 5a shows the DFT optimized Mn–(O–C2)4 configuration on graphitic carbon with four Mn–O bond lengths of 1.909, 1.909, 1.908 and 1.909 Å. It is known that adequate NO3− adsorption and activation on catalyst are the essential prerequisite for an efficient NitRR process. Thus, NO3− adsorption on the Mn–(O–C2)4 unit was evaluated via theoretical adsorption energy. As shown in Fig. 5b, the adsorption energy of NO3− adsorbed on Mn–(O–C2)4-coverage graphene is − 3.51 eV, which indicates the strong interaction between NO3− and the Mn-SAs/CBC catalyst. Moreover, the projected density of states (PDOS) of the NO3− bonded to the catalytic sites (Fig. 5c) exhibit a significant hybridization near the Fermi level, and the O 2p orbital shows a vast majority overlap with the Mn 3d orbital according to the charge density difference calculation (Fig. 5d). The result confirms the role of Mn in NitRR adsorption and further activation via electronic donation [31, 53]. The Gibbs free energies diagram further provides more details about the NitRR mechanism on the Mn–(O–C2)4 site. As shown in Figs. 5e and S23, NO3− is first adsorbed to give *NO3 with a dramatic energy decrease up to − 3.34 eV, implying favorable NO3− adsorption. *NO3 spontaneously decomposed into *NO2 with an energy release of − 2.27 eV. Then, by adsorbing a proton to couple with an electron transfer, one N–O bond in *NO2 would be broken, and *NO2 transformed to *NO with a downhill free energy change of − 2.12 eV. Next, three continued protonation steps of *NO generate *NHO, *NH2O and *NH2OH, respectively. The rate-determining step over Mn–(O–C2)4 was the formation of *NO to *NHO (*NO → *NHO) with a free energy increase of 0.75 eV, which was much higher than the energy increase of *NHO to *NH2O (0.17 eV). Subsequently, *NH2OH goes through another N–O cleavage and hydrogenation to form *NH3. Finally, *NH3 desorbs form the Mn–O–C catalyst surface to produce free NH3 via consuming energy of 0.02 eV. Besides, the Gibbs free energies of the HER process on the Mn–(O–C2)4 site were further calculated (Figs. 5f and S24). However, the free energy of H2O adsorbed to form *H2O is uphill and the value is 0.22 eV, which is much higher than NO3− adsorption to form *NO3 (− 3.34 eV). Therefore, the isolated Mn–(O–C2)4 sites are beneficial for active adsorption of NO3−, the favorable formation of *NHO, and suppression of the competition from the HER.

Fig. 5
figure 5

a DFT optimized configurations of Mn–(O–C2)4. b NO3− adsorption on Mn–(O–C2)4. c The computed projected density of states (PDOS) of Mn–(O–C2)4 after NO3− adsorption. The Fermi level was set to be as denoted by the orange dashed line. d The optimized charge density difference of the Mn–(O–C2)4 after adsorbing NO3−. e, f Gibbs free energy diagram of NitRR to NH3 and water dissociation on Mn–(O–C2)4

Full size image

4 Conclusions

In this work, an atomically dispersed and oxygen-coordinated Mn–O–C catalyst has been developed through an impregnation–pyrolysis–etching synthetic approach, which is effective to significantly regulate the density active sites. The high activity of the Mn–O–C catalyst was evidenced by a superior RNH3 of 1476.9 ± 62.6 μg h−1 cm−2 at − 0.7 V (vs. RHE) and a FE of 89.0 ± 3.8% at − 0.5 V (vs. RHE) under ambient conditions. An in situ FT-IR and Raman spectroscopic investigations combined with DFT calculations found that the exclusive existence of Mn–(O–C2)4 sites can effectively suppress the competitive HER and greatly promote the adsorption of reacting intermediates and thus high NitRR performance. The current research results also show that nitrogen pollution management is expected to be realized through the electrocatalytic approach of using single-site Mn catalysts, thus contributing to the future sustainability of fertilizer and renewable fuel recovery in many aspects.