Nitrogenous fertilizer, as the largest consumed chemical fertilizer, has increased food production to support ~ 40% of the world’s population.1 In this regard, the Haber-Bosch (H-B) method for nitrogen fixation has been considered the greatest invention in the 20th century.2 In conventional agriculture, solid nitrogenous fertilizer is produced in centralized chemical plants and then transported to where it is needed. Both processes are highly dependent on fossil fuels with the emission of huge amounts of greenhouse gas. Moreover, the utilization rate of the applied solid nitrogenous fertilizer is less than 50%,3 inevitably leading to diffusion of unabsorbed nitrogenous fertilizer into underground water. Very recently, the introduction of liquid fertilizer into precision irrigation systems was proven to significantly improve fertilizer use efficiency.4 Thus, the development of new routes for distributed production and in-place utilization of liquid nitrogenous fertilizer driven by renewable energy is of great significance.
Since plants can simultaneously absorb ammonium- and nitrate-nitrogen fertilizer, ammonium nitrate (NH4NO3) solution can serve as a good candidate carbon-free liquid nitrogenous fertilizer.5 Herein, we exhibit a two-step relay strategy to sustainably produce NH4NO3 solution by using air and water as raw materials (Fig. 1). In the first step, air is converted into NOx by plasma technology. In the second step, the produced NOx is dissolved in water and then converted to NH4NO3 solution by electrocatalytic reduction. Both steps can be directly driven by volatile green energy such as wind and solar power. The NH4NO3 solution produced by this distributed two-step relay strategy can be easily integrated into the present precision irrigation system with controllable fertilizer concentrations for different crops. We summarized the significant progress in plasma-driven air-to-NOx conversion and electrocatalytic NOx− reactions.6,7 Surprisingly, the two-step relay strategy has gained an economic advantage over conventional manufacturing for solid NH4NO3 fertilizer based on techno-economic analysis (TEA). This work offers a method for sustainable production and in-place utilization of liquid nitrogenous fertilizer with high use efficiency, assisting the development of smart agriculture.
STEP 1: PLASMA-DRIVEN AIR-TO-NOx CONVERSION
Nitrogen fixation to form nitrogen-containing oxides (NOx/NOx−) can be driven by light irradiation, electricity and plasma.7,8 In 1989, O. A. Ileperuma et al. 9 showed that TiO2-coated ZnO dispersed in an aqueous solution could convert nitrogen to nitrate under ultraviolet radiation. Recently, photocatalytic air conversion for NOx synthesis was developed by our group.10 In addition, a direct electrochemical nitrogen oxidation strategy to form NO3− was mentioned by Chen et al.11 and experimentally proven by our group12 in 2019. Although great efforts have been devoted to this area, the efficiency of NOx/NOx− synthesis via photo/electrocatalytic nitrogen oxidation is still very low because of the difficulty in activating the strong N ≡ N bond (941 kJ mol− 1).13 In contrast, plasma-driven nitrogen fixation has been considered an attractive technology due to its high efficiency in nitrogen activation.14 The plasma process, as one of the oldest phenomena occurring on Earth (e.g., lightning), is generated by the ionization of gases. The first industrial application of atmospheric nitrogen fixation was known as the Birkeland-Eyde (B-E) plasma process.15 Specifically, plasma arcs were generated in the thermal-plasma furnace, and then the air passed the furnace to combust and produce NOx. It was reported that only ~ 3% of input energy was utilized in the B-E process for NOx production. Although several improvement methods, such as waste heat utilization from process gases and the operation of furnaces at high pressure, have been developed, the energy efficiency is still unsatisfactory. Moreover, the NOx products tended to dissociate when not supplied to thermal quenching.15
Instead, non-thermal plasma is taking the dominant status due to its low theoretical energy consumption for nitrogen fixation (approximately 0.2 MJ mol− 1-N) and easy operation at low temperature.14 In a typical non-thermal plasma reactor, the bulk temperature remains ambient, while the temperature of internal electrons can exceed several thousands of K, which allows for selective activation of the N ≡ N bond.16,17 Based on different discharge patterns, non-thermal plasma can be divided into dielectric barrier discharge (DBD) plasma, glow discharge plasma, microwave (MW) plasma, gliding arc discharge plasma and so on. Figure 2a and Supplementary Table S1 summarize the progress on non-thermal plasma-driven NOx synthesis with respect to plasma types, operation parameters and energy costs. For example, DBD plasma coupled with catalysts was able to achieve nitrogen fixation with an energy cost of 18 MJ mol− 1-N.18 Pei et al.19 reduced the specific energy cost for NOx formation to ~ 2.8 MJ mol− 1-N by adopting direct current glow discharge at atmospheric pressure. The value was further reduced to 2.0 MJ mol− 1-N for an electrode-free MW plasma because electrode-free ignition could effectively reduce energy loss to the walls.20 Notably, Patil et al.21 employed gliding arc discharge plasma to prepare NOx with 1% concentration at an energy cost of 1.43 MJ mol− 1-N. It allowed the generation of higher NOx concentrations by increasing the frequency, pulse width, and amplitude.
STEP 2: ELETRO-CATALYTIC NITRATE-TO-AMMONIA CONVERSION
The obtained NOx (NO2/NO) from the plasma process firstly get through aqueous solution. NO is insoluble in water, and NO2 reacts with water to form nitrate and NO. With the presence of O2, the produced NOx finally dissolves in the electrolyte to form nitrate. Thus, the second step in the relay strategy is the electrocatalytic reduction of half nitrate to form NH4NO3 solution. The rational design and optimization of electrocatalysts with maximum efficiency and selectivity for ammonia requires a fundamental understanding of the reaction mechanism.6,22 The process of the electrochemical nitrate reduction reaction (NO3RR) can be summarized as the deoxidation and hydrogenation of nitrogen-containing intermediates, involving the transfer of eight electrons and nine protons. 28,30,31 There are many stable intermediates with various valence states, including NO2, NO, N2O, N2 and NH2OH, which can escape from the catalyst surface as products. Although the thermodynamic equilibrium potential of each intermediate varies considerably, many intermediates may coexist in the actual catalytic process due to the specific kinetic processes and the differences in their adsorption energy.23 As a result, there is selective competition for the production of any target product.24 The reaction pathway of the NO3RR begins with the adsorption of nitrate to the active site and its gradual deoxidation process. This process can be summarized by equations (1)–(3).
*NO3− → *NO3 + e- (1)
*NO3 + H2O + 2e− → *NO2 + 2OH− (2)
*NO2 + H2O + 2e− → *NO + 2OH− (3)
Next, the resulting *NO intermediate is hydrogenated. The mechanism of N-O bond cleavage can be divided into direct breaking followed by hydrogenation and successive hydrogenation followed by dehydration.
Direct breaking followed by hydrogenation model:
*NO + H2O + 2e− → *N + 2OH− (4)
*N + H2O + 2e− → *NH + OH− (5)
*NH + H2O + 2e− → *NH2 + OH− (6)
*NH2 + H2O + e− → *NH3 + OH− (7)
Successive hydrogenation followed by dehydration model:
*NO + H+ + e− → *HNO (8)
*HNO + 2H+ + 2e− → *H2NOH (9)
*H2NOH + 2H+ + 2e− → *NH3 + H2O (10)
In equations (4)–(7), if two *N intermediates couple, a byproduct of N2 will be generated. In equations (8)–(10), *NH2OH might be generated as a byproduct. Different hydrogenation pathways produce different intermediates and byproducts.25 Therefore, it is crucial to design electrocatalysts with high activity and selectivity toward the synthesis of goal products from the NO3RR.23,26 Here, we summarized various types of catalysts for the NO3RR, including metal-free electrocatalysts, transition metal-based electrocatalysts, and noble metal-based electrocatalysts (Fig. 2b and Supplementary Table S2). Metal-free eletrocatalysts are represented by carbon-based materials. Polymeric carbon nitride with a controlled number of nitrogen vacancies exhibited 90% Faradaic efficiency (FE) and 0.03262 mmol g− 1 h− 1 yield for ammonia product at -0.75 V versus reversible hydrogen electrode (RHE).27 Transition metal-based catalysts have long been the focus of research in electrocatalysis due to their low cost, high activity, and potential for large-scale applications. Cobalt phosphide (CoP) nanosheets with a three-dimensional structure exhibited ~ 100% FE and a 9.56 mol h− 1 m− 2 yield of ammonia at a potential of -0.3 V vs. RHE.28 Compared to metallic cobalt nanosheets, the introduction of phosphorus atoms into the lattice effectively stabilized the active phase and optimized the energy barriers of the key steps in the NO3RR. Sargent et al.29 prepared a series of CuNi alloys with various Cu:Ni compositions by electrodeposition, which electrochemically reduced nitrate into ammonia with a FE of 99 ± 1% and a yield of 0.48 mmol h− 1 cm− 2 at -0.15 V vs. RHE. With increasing Ni content in the CuNi alloys, the d-band center of Cu upshifted, and the anti-bonding occupation decreased, which greatly enhanced the intermediate adsorption energy. Noble metal-based catalysts with empty d-orbitals have been considered state-of-the-art electrocatalysts for the NO3RR due to the role of d-orbital electrons in enhancing charge injection of the lowest vacancy molecular orbital of nitrate. Luo et al.30 designed rhodium nanoflowers as NO3RR electrocatalysts, in which the low-coordination Rh atoms facilitated the adsorption of nitrate ions and stabilized the intermediates. As a result, this electrocatalyst reduced nitrate to ammonia with a 95% FE and 0.0149 mmol h− 1 cm− 2 yield at + 0.2 V vs. RHE. Very recently, we proposed a novel three-step relay mechanism to decrease the reaction overpotential for the NO3−RR. A series of RuxCoy catalysts were designed and prepared to perform the three-step relay mechanism, in which Ru15Co85 exhibited the optimal catalytic performance (onset potential: +0.4 V vs. RHE, FE: 96.8%, yield: 3.2 mol gcat−1 h− 1).31 Moreover, great progress has been made in scaling up NO3RR. An electrochemical system with a working volume of 500 L was used to electrochemically reduce nitrate to ammonia by Feng's group.32 The selectivity of NH4+ in this system was over 92.0% with concentrations ranging from 241 mg L− 1 to 2527 mg L− 1, and the productivity of NH4+ did not decay after 2 months of continuous operation. This proves the practical application potential of the NO3RR to produce ammonia.