Human activities discharge plenty of nitrogen-containing substances into surface water in the form of NO₂^–/NO₃^–, posing severe risks to public health. Traditional catalytic conversion of harmful NO₂^–/NO₃^– into harmless nitrogen molecules fails to realize the product upgrading and the maximum use of energy and resources. Recent studies indicate the electrocatalytic coupling of CO₂ and NO₂^–/NO₃^– into urea is an ideal alternative, which simultaneously realize the utilization of CO₂, control of water pollution and acquirement of high value-added products.

In the early stages, Furuya et al. [32,33] used Ni-phthalocyanine (Ni-Pc) electrocatalysts for simultaneous reduction of CO₂ and NO₂^– to produce urea, an optimal current efficiency of 40% was obtained at −1.5 V vs. standard hydrogen electrode (SHE), but the complex reaction pathways of CO₂ reduction reaction (CO₂RR) allowed very few active species to combine with N-containing intermediates to complete the C–N coupling reaction, and the competitive hydrogen evolution reaction (HER) also limits the reaction efficiency. Subsequent studies have shown that some transition metal oxides and p-block-elements based electrocatalysts can be candidates for this reaction because their inherent weak H adsorption abilities can restrain the HER [22,34]. Especially, inspired by the defect engineering strategy, a series of robust defective electrocatalysts have been developed to simultaneously activate N- and C-containing intermediates and reduce the HER for more effective urea synthesis.

Cao et al. [35] designed O vacancy (V_O)-riched Cu-doped anatase TiO₂ nanotubes catalysts (Cu–TiO₂) to improve the electrocatalytic co-reduction of CO₂ and NO₂^–. A little red shift of (101) plane was observed in Cu–TiO₂, indicating Cu doping reduced the TiO₂ lattice constant (Fig. 1A). The remarkably increase in Ti³⁺/Ti⁴⁺ ratio indicated Cu doping contributed to the occurrence of V_O on Cu–TiO₂ (Fig. 1B). The V_O facilitated the adsorption and activation of NO₂^– and cooperated with high CO₂RR active Cu sites to realize efficient urea synthesis, and a Faraday efficiency (FE) of 43.1% was obtained at a low overpotential of −0.4 V vs. reversible hydrogen electrode (RHE) (Fig. 1C). The potentials mentioned in this work are relative to RHE, unless otherwise stated. Meng et al. [36] found V_O–ZnO porous nanosheets display an excellent performance in electrocatalytic synthesis of urea from CO₂ and NO₂^–. The experimental results indicated CO₂ and NO₂^– filled the surface vacancy for accelerating urea synthesis, and the V_O–ZnO displays a ~3- and 2-fold enhancement of the FE_urea and urea yield rate (r_urea), respectively, compared with that of pure ZnO at an applied potential of −0.79 V (Figs. 1D and E). Lv et al. [37] reported the intriguing electrocatalytic urea synthesis process on a defect engineered indium oxyhydroxide catalyst with V_O (V_O–InOOH) using CO₂ and NO₃^– as reactants. V_O–InOOH exhibited much better performance compared with pristine InOOH in terms of both urea yield (592.5 vs. 378.4 µg h^–1 mg_cat.^–1) and FE_urea (51.0% vs. 26.3%), indicating the presence of V_O resulted in the improvement of C–N coupling efficiency. Wei et al. [38] used the V_O-riched CeO₂ (V_O–CeO₂) as the electrocatalyst for urea synthesis with CO₂ and NO₃^– as reactants. Experimental results indicated the urea production was tightly associated with the V_O concentration, and CeO₂ treated with 750 ℃ (V_O–CeO₂-750), the one had the highest V_O contents (Fig. 1F), exhibited the highest urea yield of 943.6 mg h⁻¹ g⁻¹ at −1.66 V, which was almost three times as much as that of pristine CeO₂ (Fig. 1G). Beside to metal-based electrocatalysts, a novel metal-free F-riched carbon nanotubes (F-NCT) electrocatalyst was developed for urea synthesis with CO₂ and NO₃^– as reactants by Liu et al. [39]. The F-doping introduced lots of defects in CNT according to the Raman spectra (Fig. 1H), and a high r_urea of 6.36 mmol h^–1 g^–1 with a corresponding FE_urea of 18.0% was achieved for F-CNT catalyst mediated urea synthesis at –0.65 V (Fig. 1I). The r_urea was 3.8 times as high as that obtained from CNT catalyzed urea synthesis, proving the defects introduced by F doping facilitate the C–N coupling reaction to form urea. Recently, Li et al. [40] integrated facet and defect engineering strategy to develop a highly efficient V_O-rich indium hydroxide nanocube catalyst with dominant facets (denoted as V_O-S-IO-6) for urea synthesis from the electrocatalytic co-reductive coupling of NO₃^– and CO₂. The In 3d orbital of V_O-S-IO-6 shifted to lower binding energy compared with M-IO and S-IO, implying the increase of the local electron density of In caused by delocalized electrons after oxygen escaping (Fig. 1J). Linear cyclic scanning (LSV) measurements proved V_O-S-IO-6 had the best performance toward electrocatalytic co-reduction of NO₃^– and CO₂ (Fig. 1K), and a high r_urea of 910.4 µg h^–1 mg^–1 and FE_urea of 60.6% was achieved at −0.6 V (Fig. 1L), indicating the introduction of V_O contributed to the C–N coupling toward urea synthesis.

N₂ is the most abundant and cheapest N source on the earth, and the report on electrocatalytic N₂ reduction to NH₃ inspired researchers to synthesize urea in N₂-integrated electrocatalytic CO₂ reduction in spite of the inertness and high bonding energy of the N≡N bonds [41,42]. Compared with the NO₂^–/NO₃^–-integrated electrocatalytic CO₂ reduction for urea synthesis, the N₂-integrated electrocatalytic CO₂ reduction one not only lowers the cost of raw material but also circumvents the use of energy-intensive fossil-fuel [10–12].

Kayan et al. [43] firstly used aqueous solution as the proton source to explore the co-reduction of N₂ and CO₂ for the production of urea on a polypyrrole-coated Pt electrode under a mixed atmosphere of 30 bar N₂ and 30 bar CO₂, and an optimized urea formation rate of 31.8 µg h^–1 cm^–2 and of FE_urea of 6.9% at −0.325 V vs. NHE were achieved. The replace of ambient working condition to high pressure one favors to save energy and minimize the requirement of synthesis equipment. Inspired by the fact that metal alloying and defect engineering can promote the electronic interactions and provide more active sites for electrocatalytic reaction [44–47]. Wang’s group [48] developed the electrocatalyst of PdCu alloy nanoparticles on V_O-riched defective TiO₂ nanosheets (Pd₁Cu₁/TiO₂–V_O) for the efficient electrocatalytic synthesis of urea via the coupling of N₂ with CO₂ under ambient conditions. The interaction between TiO₂–V_O-400 (TiO₂ with treatment at 400 ℃) with the most abundant defect sites and PdCu alloy was verified by X-ray photoelectron spectroscopy (XPS) analysis (Figs. 2A and B). The temperature programmed desorption (TPD) spectra illustrated that the proposed electrocatalyst displayed a stronger chemical adsorption ability for CO₂ and N₂ due to the strong electronic interaction between the bimetal alloys and the TiO₂-V_O-400 (Fig. 2C). Operando synchrotron radiation−Fourier transform infrared spectroscopy (SR–FTIR) analysis proved the formation of urea (Figs. 2D and E), and a desirable r_urea of 3.36 mmol g^–1 h^–1 and FE of 8.92% at –0.4 V were obtained (Figs. 2F and G). Also, higher r_urea and FE_urea of Pd₁Cu₁/TiO₂-V_O-400 than that of Pd₁Cu₁/TiO₂-V_O were identified. On the basis of this work, Pan et al. [49] further found Pd₁Cu₁/TiO₂-V_O had better performance than Pd₁/TiO₂-V_O on electrocatalytic co-reduction of N₂ and CO₂ for urea synthesis, indicating the important role of dual-atom Pd₁Cu₁ site. Wu et al. [50] reported the defective bimetallic Cu–Bi catalyst for urea electrosynthesis. HR-TEM images showed the bimetallic Cu–Bi had ordered regions consisting of missing-atom defects (Fig. 2H), and the EPR analysis also proved the presence of higher concentration of unsaturated Bi sites with unpaired electrons (Fig. 2I). Significantly, defective Cu–Bi catalyst offered a maximum urea concentration of 0.45 ± 0.06 mg/L and the largest faradaic efficiency of 8.7% ± 1.7% at −0.4 V (Fig. 2J), and meanwhile, it exhibited much higher active in producing urea compared with monometallic (Cu and Bi) and bimetallic (Cu–Bi mixture and intact Cu–Bi) catalysts (Fig. 2K). This work provides insights into intriguing relationships between defective structure and catalytic activity, shedding light on the rational design of defective electrocatalysts for producing urea from N₂ and CO₂ molecules.

N₂-integrated electrocatalytic CO₂ reduction for urea synthesis has huge application prospects, but the exceedingly sluggish kinetics of N₂ reduction limits the overall rate of urea production [11,12,22]. Current works indicate the r_urea and FE_urea are unsatisfactory (Table 1) [2,23,35–40,48–70]. Of note, current works about defective electrocatalysts mediated coupling of CO₂ and N₂ for urea synthesis are scarce, and the development of more robust defective electrocatalysts is bound to push this field forward.

The mechanism of electrocatalytic urea synthesis from NO₂^–/NO₃^–-integrated CO₂ reduction is still controversial, there are mainly four possible explanations in current reported works (Table 2). The ^*CO + ^*NH₂ coupling mechanism observed in NO₂^–/NO₃^–-integrated electrocatalytic CO₂ reduction for urea synthesis was firstly proposed by Shibata and Furuya et al. [33,71,72]. Recently, Liu et al. [39] proved the feasibility of this mechanism through density functional theory (DFT) calculations in their recent F-CNT work, and the formation of ^*CO and ^*NH₂ intermediates was more favorable on F-doped C active sites ("C–F₂" moieties) than on C active sites (Fig. 3A). The experimental results also proved the F-CNT showed better performance toward urea synthesis than CNT, hence we deduced the introduction of V_O caused by F doping contributed to urea production. Of note, this coupling mechanism has long been controversial, and some new ideas have also put forward. For instance, Meng et al. [36] proposed that the V_O–ZnO porous nanosheets mediated electrocatalytic C–N formation came from the coupling of ^*NH₂ and ^*COOH intermediates, which were generated from the reduction of NO₂^– and CO₂, respectively. In situ infrared (IR) spectra measurements indicated the ^*COOH peaks detected at reductive potentials under the CO₂ + NO₂^– conditions were weaker than that detected at the CO₂-free conditions, and a peak attributed to C–N bond was observed, indicative of the consumption of ^*COOH and occurrence of C–N coupling (Fig. 3B). The experimental results indicated V_O promoted urea electrosynthesis, we therefor deduced V_O was conducive to the formation of ^*NH₂ and ^*COOH intermediates. Yu’s group [23,37] proposed the ^*CO₂ + ^*NO₂ coupling mechanism using In-based materials as electrocatalysts. In their V_O–InOOH work [51], DFT calculations indicated the urea synthesis was triggered by the thermodynamically spontaneous reduction of NO₃^– to the ^*NO₂ and the ^*CO₂NH₂ protonation was the potential-limiting step (PLS) for the overall urea formation process (Fig. 3C). Significantly, the V_O contributed to lower the energy barrier for the protonation of the ^*CO₂NH₂ intermediate to accelerate urea synthesis, and Bader charge analysis (Fig. 3D) and the SR-FTIR analysis (Fig. 3E) further proved this idea. In the V_O-S-IO-6 work reported by Li et al. [40] proposed the same C–N coupling mechanism, they illustrated the introduction of V_O significantly tuned the electronic structure around the catalytic sites to accelerate the cleavage of In–O bond in V_O-S-IO-6 adsorbed ^*CO₂NH₂, thereby benefiting the protonation process and boosting the electrocatalytic performance of urea synthesis. Recently, Wang’s group proposed the urea formation via a two-step coupling process. In their V_O–CeO₂ work [38], in situ sum frequency generation (SFG) spectroscopy analysis indicated the core of reaction mechanism was that ^*NO intermediate species adsorbed to V_O sites of the catalyst, and the subsequent coupling process with ^*CO to form ^*OCNO was more favorable than the hydrogenation reaction of ^*NO, whereas the (proton-coupling electron-transfer, PCET) process of ^*NO tended to occur on the V_O-deficient electrocatalyst (Figs. 3F and G). DFT calculations indicate the C–N coupling was thermodynamically and kinetically favorable on Vo-enriched CeO₂, and meanwhile, a novel two-step coupling mechanism, namely, ^*CO + ^*NO → ^*ONCO and ^*ONCO + ^*NO → ^*ONCONO, was proposed (Fig. 3H). In addition, another two-step coupling mechanism of ^*CO + ^*NO → ^*ONCO; (2) ^*ONCO + ^*NO → ^*ONCONO was also proposed [59,60], but the related works did not involve defective electrocatalysts.

N₂-integrated electrocatalytic CO₂ reduction reaction involves multistep electrochemical processes and nonelectrochemical processes, and probably share the similar C–N coupling mechanism as the co-reduction of CO₂ and NO₂^–/NO₃^–. Generally, the coupling reaction occurs through the formation of C–N bonds via the thermodynamically spontaneous reaction between ^*N=N^* and CO. In the Pd₁Cu₁/TiO₂-V_O-400 work [48], DFT calculations illustrated that the key urea precursor of tower-like ^*NCON^*, was formed via the reaction between ^*N=N^* and CO in a thermodynamically and kinetically feasible way, then ^*NCON^* was hydrogenated following a distal pathway to form urea, and the third proton-coupled electron transfer process was energy demanding and was the PLS of urea (Fig. 4A). The formation of C–N bond was proved by SR-FTIR analysis (Figs. 2E and F), but the structure of ^*NCON^* intermediates and the role of V_O were not well identified. The same C–N coupling process was proposed by Zhang’s group [64] in the Bi–BiVO₄ work, but ^*NCON^* was hydrogenated in an alternating pathway rather than in a distal pathway, and the PLS was the generation of ^*NHCONH intermediate (Figs. 4B and C). This hypothetical ^*NCON^* pathway for urea formation involved two N atoms bond with one C atom simultaneously, which was obviously different from the pathways in electrocatalytic co-reduction of CO₂ and NO₂^–/NO₃^–. Of note, such a dual C–N bond-forming step circumvented the formation of highly unstable formamide or carbamic acid intermediates. In the defective Cu–Bi catalyst work reported by Wu et al. [50], Raman signal of C–N intermediates on defective Cu–Bi immersed in both N₂ and CO₂ atmosphere was observed (Fig. 4D), and no evidence of co-reduction N₂ and CO₂ was found on contrast catalysts including Cu, Bi, mixed Cu–Bi, and intact Cu–Bi (Fig. 4E), indicating the defective Cu–Bi had the most abundant active sites to produce urea by coupling N₂ and CO₂. DFT calculations indicated the defective Cu–Bi Gibbs free energy decreased in the presence of N₂ and CO₂, while the intact Cu–Bi did not show a significant preference for ^*COOH (^*N₂) and ^*COOH, which suggested that the defective Cu–Bi could offer important advantages over that intact Cu–Bi on coupling C–N intermediates (Fig. 4F). In addition, rate-determining step (RDS) barriers were also used to evaluate the catalytic activity. As shown in Figs. 4G and H, the RDS for the distal and alternating styles of intact Cu–Bi was from ^*NN + ^*CO to ^*NCON^* step, amounting to a barrier of 2.63 eV, and relative lower RDS barriers were posed by comparing the distal (from ^*NHCON to ^*NHHCON, 0.78 eV) and alternating (from ^*NN + ^*CO to ^*NCON^*, 0.76 eV) styles in defective Cu–Bi. This work paved the way to further advances in designing other defective electrocatalysts for N₂ and CO₂ co-reduction operation. Subsequent works also proposed the same coupling steps as the above two works, but further investigation was still desired to verify the C–N coupling process, especially the formation of ^*NCON^* intermediate. Recently, theoretical scientists have carried out some theoretical calculation works to help researchers to better understand the C–N coupling mechanism of N₂-integrated electrocatalytic CO₂ reduction reaction [73–76], but the role of defect was not well explored.

The mechanisms for urea synthesis from N-integrated O₂ reduction are controversial, and more detailed in situ characterization should be identified, especially the formation of key intermediates. Anyway, the defects are certainly proved to act the role of extra active sites to modulate the nearby active sites’ electronic structure, facilitating the formation of vital intermediates and higher reaction efficiency. Further research is bound to develop more robust defect engineering strategy to improve the selectivity, FE and yield of urea synthesis.