Enhanced NH3 Synthesis from Air in a Plasma Tandem-Electrocatalysis System Using Plasma-Engraved N-Doped Defective MoS2

We have developed a sustainable method to produce NH3 directly from air using a plasma tandem-electrocatalysis system that operates via the N2–NOx–NH3 pathway. To efficiently reduce NO2– to NH3, we propose a novel electrocatalyst consisting of defective N-doped molybdenum sulfide nanosheets on vertical graphene arrays (N-MoS2/VGs). We used a plasma engraving process to form the metallic 1T phase, N doping, and S vacancies in the electrocatalyst simultaneously. Our system exhibited a remarkable NH3 production rate of 7.3 mg h–1 cm–2 at −0.53 V vs RHE, which is almost 100 times higher than the state-of-the-art electrochemical nitrogen reduction reaction and more than double that of other hybrid systems. Moreover, a low energy consumption of only 2.4 MJ molNH3–1 was achieved in this study. Density functional theory calculations revealed that S vacancies and doped N atoms play a dominant role in the selective reduction of NO2– to NH3. This study opens up new avenues for efficient NH3 production using cascade systems.


5.
The remaining Faradaic efficiency is most likely due to the competing HER reaction, since the entire process involved only eNO2 -RR and HER, and no liquid byproducts (e.g., N2H4) were detected (see Section S-Ⅳ-3 in SI). Nevertheless, the H2 produced was not measured due to its very low concentration caused by a high flow rate (9 SLM). But we will measure H2 concentration in our future work by using GC, as suggested by the reviewer. Figure S1. Schematic of the plasma tandem-electroreduction system for sustainable ammonia production.

S-I-2: Electrochemical evaluation.
All the electrochemical tests were conducted using two H-type cells (GAOSSUNION) connected in series with well-sealed rubber rings and stainless-steel tubes and the chambers of each cell were separated by a proton exchange membrane (DuPont, Nafion 117). All data were collected with a typical three-electrode system connected to a CHI-660E electrochemical workstation (CHI Instrument, Inc.). The as-prepared electrocatalysts, Ag/AgCl, and platinum foil were used as the working, reference, and counter electrodes, respectively. The potentials were converted into the RHE according to the equation: ERHE = EAg/AgCl + 0.207 V + 0.0591  pH. A solution with 0.1M KOH was used as the electrolyte and NOx absorbent. The two H-cell had volumes of 200 ml and 80 ml, respectively, and were purged with pure argon for 20 minutes before testing. The cyclic voltammetry (CV) curves were performed at a scan rate of 5 mV s -1 and electrochemical impedance spectroscopy (EIS) was carried out in a frequency range of 1 MHz to 10 mHz with an AC amplitude of 10 mV. Catalysts with a geometric area of 1  1 cm 2 were used for potentiostatic tests at various potentials (-0.13 ~ -0.53 V vs RHE) for 1 h at a stirring rate of 280 rpm. The electrochemically active surface area (ECSA) was calculated from the following equations: ECSA = Cdl/Cs, where Cs is the specific capacitance of an atomically smooth planar surface and is considered to be 40 F cm -2 for a 0.1M KOH medium.

Ammonia quantification by typical indophenol blue method.
The quantification of ammonia production was conducted by the typical indophenol blue method. 0.5 mL of the reacted solution was extracted from the cathode chamber, and then diluted to 4 mL for the following detection. Afterwards, 0.32 mL of 1 wt% sodium nitroferricyanide (Na2[Fe(CN)5NO] 2H2O) aqueous solution was added into diluted samples, followed by addition of 2.4 mL of 0.32 M NaOH solution containing 10.4 M sodium salicylate (C6H4(OH)COONa), and 0.8 mL of 0.3 M NaClO and 0.75 M NaOH mixture. The indophenol blue absorbance was measured using an Ultraviolet-visible (UV-vis) spectrophotometer (MAPADA, UV-1800) after 2 hours of incubation.
For reference, the calibration was built with standard ammonia chloride solution (0.5 -5 g ml -1 ).

Hydrazine quantification by Watt and Chrisp method.
The ammonia production was quantified using a Watt and Chrisp method. The coloring agent was prepared by mixing 5.99 g of para-(dimethylamino)benzaldehyde (p-C9H11NO) with 30 mL of hydrochloric acid (HCl) and 300 mL of ethanol (C2H5OH). Then, 9 mL of 1.0 M HCl was added to 1 mL reacted solution, followed by the addition of 5 mL of coloring agent. The absorbance was measured using the aforementioned UV-vis spectrophotometer after 30 min of incubation. For reference, the calibration was built with standard hydrazine solution (1 -5 g ml -1 ).

Nitrate and nitrite (NO3and NO2 -) quantification.
First, the reacted solution was extracted from the cathode chamber and diluted to detectable levels. 1 ml of the sample was measured by ion chromatography (IC, ICS-3000). A series of standard potassium nitrate and potassium nitrite solutions are used to obtain the concentration-intensity curves.

Nuclear magnetic resonance (NMR) analysis.
Ammonia and 15 N isotope-labeling were also quantitatively determined by 1 H nuclear magnetic resonance (NMR, 600 MHz) using maleic acid (C4H4O4) as the internal standard. Firstly, the calibration curve was established as follows: 40 ml 0.1 M KOH (with a standard concentration of 1000, 500, 250, 125 ppm ammonia chloride); Secondary, the pH of the above solution was adjusted by adding dropwise to 2 ~ 3; Afterwards, the maleic acid (with the concentration kept at 100 ppm) was added into the mix solution and the acidulated solution was tested by a 600 MHz liquid Agilent DD2 NMR spectrometer at room temperature; Finally, the calibration curve was achieved using the peak area ratio between NH4 + and C4H4O4 because the NH4 + concentration and the area ratio were positively correlated.
As for the ammonia and 15 N detection, the process was the same as explained above, except that the standard concentration of 1000, 500, 250, 125 ppm ammonia chloride was replaced by the reacted solution.
The electric power converting NOx intermediaries to ammonia is calculated by Eq. (S3), where P2 is power (W), U2 is applied potential (V) and I2 is current (A). In the electrolyser, a certain voltage was applied, and the output current was recorded. We took the average current multiplied by the applied voltage to calculate the power.
S-Ⅱ-3: Calculation of the Faradic efficiency (FE) and ammonia production rate. The FE of electrocatalytic NO2 -, NO3 -, and NO to NH3 conversion were calculated as follows: The ammonia production rate was calculated as follows: where F is the Faraday constant (96,485 C mol -1 ), CNH3 is the measured NH3 concentration, V is the volume of the cathodic electrolyte, Q is the total charge passing the electrode, t is the reduction time, and Scat. is the total area of catalysts. Note: In this study, the FE was calculated based on function (S5).
S-Ⅲ-1: HRTEM images of phase transformation from 2H to 1T.   S-Ⅲ-3: Comparison of the electrochemical properties of various catalysts. The results revealed that the N-MoS2/VGs possess the largest specific surface area (SBET), the largest value of double-layer capacitance (Cdl), the largest electrochemically active surface area (ECSA) and the smallest electrochemical impedance.

N isotope-labeling experiment.
To demonstrate that the produced ammonia originates from the source of N2 gas rather than potential contamination, 15 N2 isotopic experiments were carried out, with the ammonia detected by 1 H NMR to distinguish between 15 NH4 + and 14 NH4 + . The NMR test method was listed in section S-Ⅳ-4: "Nuclear magnetic resonance (NMR) analysis". The 15 NH4 + and 14 NH4 + production rates are almost the same, excluding the influence of the surroundings or the plasma, and demonstrating that the produced ammonia effectively originates from the source of N2 gas rather than potential contamination. Figure S14. (a) NMR spectra of the 14 NH4 + and 15 NH4 + detection and (b) corresponding NH3 production rates comparison.

S-Ⅶ-1: Calculation methods.
All the first-principles calculations were performed by the "Vienna ab initio simulation package" (VASP) [2] with the theoretical basis of density functional theory (DFT). Generalized gradient approximation (GGA) with Perdew, Burke, and Ernzerhof [3] (PBE) was adopted to treat the exchange and correlation effects. The nuclei-electron was described by the projector augmented wave (PAW) pseudopotentials. The van der Waals interactions were included using the DFT-D3 method. We utilized the Gaussian smearing method and set the value of SIGMA as 0.05. The bandgap error of MoS2 in the DOS calculation was corrected by DFT + U (U=2.0eV). The cut-off energy for the plane-wave basis was set at 400 eV. 10 −6 eV and -0.05 eV Å −1 were set as the convergence criteria for total energy and the Hellman-Feynman force. A 3-layer (3×2) MoS2 (100) bulk was built to simulate the supercell model and optimized the lattice parameters for convergence. The bottom layer of the z-axis was fixed to improve computational efficiency. The surface atoms of MoS2 were exposed with a sufficient vacuum gap of 20 Å in the z-direction to simulate the reaction on the catalyst surface, as shown in Figure S15. A 3 × 3 × 1 k-point grid was set in the first Brillouin zone. The adsorption energy (Eads) was defined as: Eads = Eslab + Emol -Etotal (S9) Where Eslab, Emol, and Etotal represent the energy of the isolated slab module, the energy of isolated species, and the energy of the species-adsorbed system.

Plasma type Application Results
Ref.