Understanding and Regulating C–N Coupling Pathways for Electrocatalytic Urea Synthesis From CO2 and NO3 −/NO2 − Authors: Xingbao Chen, Zhuo Li, Lebin Cai, Paul Kilmartin, Bao Yu Xia, Ziyun Wang Published: 2026 ABSTRACT ABSTRACTElectrocatalytic co‐reduction of CO2 and NO3 −/NO2 − to synthesize urea is a promising alternative to the traditional urea production process. However, the complexity of the reaction pathway hinders clear identification of C–N coupling mechanisms, thereby limiting the ability to guide catalyst design. Therefore, to design catalysts with improved performance, it is necessary to deepen the understanding of the urea synthesis process, particularly the C–N coupling step. This review summarises and discusses the current mechanistic insights into C–N coupling. Several representative C–N coupling pathways reported to date are reviewed and assessed in detail. Intermediate‐regulation strategies on both the carbon and nitrogen sides for promoting C–N coupling are discussed. In addition, commonly employed characterization techniques for elucidating the reaction mechanism are reviewed. Finally, this review provides perspectives on further exploring C–N coupling mechanisms.Keywords: C–N coupling mechanism, Intermediate regulation, urea electrosynthesis Introduction Urea (CO(NH2)2) serves as a crucial artificial nitrogen fertilizer for global crop production [1] and is a key chemical feedstock in the synthesis of plastics, resins, and pharmaceuticals [2, 3]. Present industrial synthesis of urea mainly employs the Bosch‐Meiser method, in which NH3 and CO2 react at high temperature (150–200°C) and pressure (15–25 MPa) [4]. In addition, NH3, as one of the reactants, should be synthesized by the Haber‐Bosch method, which also requires harsh conditions, i.e., 400–600°C and 200–350 atm [5]. Therefore, the recent industrial two‐step process synthesis of urea consumes approximately 1%–2% of the global annual energy and generates a substantial amount of CO2 [6]. By contrast, electrochemical co‐reduction of CO2 and N‐containing (N2, NO, NO2 −, and NO3 −) sources to urea offers an environmentally friendly solution to reduce energy consumption [7, 8, 9, 10]. Compared to N2, which has a highly stable triple bond (N≡N, ∼941 kJ mol−1) [11], the bond energies of the N═O bonds in NO2 − and NO3 − are significantly lower. Moreover, these species are abundant in industrial wastewater, and their infiltration into groundwater causes various diseases, such as methemoglobinemia [12]. Therefore, NO2 − and NO3 − are regarded as ideal nitrogen sources for co‐reduction with CO2 in electrocatalytic urea synthesis [13, 14, 15]. Benefiting from the inherent advantages of electrochemical co‐reduction of CO2 and NO3 −/NO2 − to urea, this field has attracted growing attention [9, 16, 17, 18]. Many strategies have been applied in catalyst design, for example, oxides [19], alloying [20], heteroatom doping [10, 21], vacancy engineering [22], facet regulation [23], and single‐atom engineering [24]. However, the understanding of the C–N coupling mechanism remains ambiguous owing to the intricate urea synthetic pathway, which hinders the rational design of catalysts with satisfactory performance [25]. The urea synthetic pathway involves 16 proton‐coupled electron transfer steps and two C–N coupling steps. For carbon species, gaseous CO2 and the other three possible intermediates may participate in the coupling process: *CO2, *COOH, and *CO. Furthermore, nitrogen species are more diverse and complex, with up to 13 intermediates theoretically existing between *NO3 and *NH2, all of which could potentially serve as coupling partners during C–N bond formation: i)*NO3 → *NO2OH → *NO2 → *NOOH → *NO → *NOH → *N → *NH → *NH2; ii) *NO3 → *NO2OH* → *NO2 → *NOOH → *NO → *HNO → *NHOH → *NH2OH → *NH2; iii) *NO3 → *NO2OH → *NO2 → *NOOH → *NO → *HNO → *ONH2 → *NH2OH → *NH2 [26]. Therefore, it is essential to understand which intermediates dominate the C–N coupling process and how they interact with specific catalytic surfaces. In this review, we summarize recent developments within mechanistic studies in the electrocatalytic co‐reduction of CO2 and NO3 −/NO2 − to urea, focusing on three major aspects: (i) mechanistic insights into C–N coupling derived from theoretical and experimental studies; (ii) intermediate regulation strategies toward efficient C–N coupling; and (iii) advances in in situ characterization techniques that have enhanced our understanding of intermediate species and coupling mechanism. By systematically reviewing recent progress in these areas, we aim to provide a comprehensive understanding of the field and offer guidance for future research directions. Finally, the current challenges and perspectives on uncovering C–N coupling mechanisms are discussed. Representative C–N Coupling Pathways in Electrocatalytic Urea Synthesis Current studies indicate that the first C–N coupling step is critical for CO2 and NO3 −/NO2 − co‐reduction to urea [27, 28, 29]. Based on recent studies, the proposed mechanisms of first C–N coupling step can be broadly categorized into five representative types: (i) *CO2 + *NO2 → *CO2NO2; (ii) *CO + *NO2 → *CONO2; (iii) *CO + *NO → *OCNO; (iv) *CO + *NH2 → *CONH2; (v) direct coupling with gaseous CO2 (Figure 1 and Table 1). The following sections elaborate on each proposed C–N coupling pathway, highlighting the adsorption behaviors on specific catalyst types, key coupling intermediates, and supporting experimental or theoretical evidence. C–N coupling mechanisms of CO2 and NO3 − co‐reduction to urea. (a) *CO2 + *NO2 → *CO2NO2, (b) *CO + *NO2 → *CONO2, (c) *CO + *NO → *OCNO, (d) *CO + *NH2 → *CONH2, (e) direct coupling with gaseous CO2. TABLE 1Recent catalysts for NO3 −/NO2 − and CO2 reduction to urea.C–N couplingC and N sourceCatalystFE (%)Yield rateReferences*CO2NO2 CO2 + NO3 − In(OH)3‐S53.4533.1 µg h−1 mgcat. −1 [31]CO2 + NO3 − MoOx/C27.71431.5 µg h−1 mgcat. −1 [30]CO2 + NO3 − VO‐S‐IO‐660.6910.4 µg h−1 mgcat. −1 [53]CO2 + NO3 − FeN4/B2CuN2@NC71.92072.5 µg h−1 mgcat. −1 [54]CO2 + NO3 − PdAuCuIrCo22.5752.43 mmol h−1 g−1 [55]CO2 + NO3 − 3D Cu/Zn7516 µmol h−1 cm−2 [7]CO2 + NO3 − U‐Zn31.839.3 mmol h−1 g−1 [32]*CONO2 CO2 + NO3 − CuWO4 70.198.5 µg h−1 mgcat. −1 [33]CO2 + NO3 − Cu‐Bi0.1 89.4586.87 µg h−1 mgcat. −1 [56]CO2 + NO3 − IL@Cu‐ZIF‐855.3140 µmol h−1 cm−2 [57]CO2 + NO3 − CuSiOX 79.011606.1 µg h−1 mgcat. −1 [34]*OCNOCO2 + NO3 − Cu1–CeO2 —52.84 mmol h−1 gcat. −1 [35]CO2 + NO3 − Cu‐PMOF25.525.5 mmol h−1 mgcat. −1 [36]CO2 + NO3 − O–BiM/CuOX 23.52180.3 µg h−1 mgcat. −1 [58]CO2 + NO3 − VO‐CeO2‐750—943.6 mg h−1 g−1 [38]CO2 + NO3 − Cu1Au8@CeO2 45.2813.6 µg h−1 mgcat. −1 [59]CO2 + NO3 − N‐C‐1000—610.6 mg h−1 g−1 [37]CO2 + NO3 − Zn–Fe2O3/OV 62.47.48 mg h−1 mgcat. −1 [60]*CONH2 CO2 + NO3 − Ni@CeO2‐X 70.12175.47 µg h−1 mg−1 [39]CO2 + NO3 − SrCo0.39Ru0.61O3−δ 34.11522 µg h−1 mg−1 [61]CO2 + NO3 − CuZn40304.8 mmol h−1 gcat. −1 [62]CO2 + NO3 − Fe(a)@C‐Fe3O4/CNTs16.51341.3 µg h−1 mgcat. −1 [49]CO2 + NO3 − Pd4Cu1/FeNi(OH)2 66.4436.9 mmol h−1 gcat. −1 [63]CO2 + NO3 − CuAC‐CuSA@NC—42.4 mmol h−1 gcat. −1 [64]CO2 + NO3 − Cu‐N4−x‐Cx 28—[24]CO2 + NO3 − Mo‐PCN‐222(Co)33.90844.11 mg h−1 g−1 [65]CO2 + NO3 − Cu‐GS‐800284.3 nmol s−1 cm−2 [24]CO2 + NO3 − FeP0.9S2.9‐x 15.41160.9 µg h−1 mgcat. −1 [66]CO2 + NO3 − Cu1.0/ZnO0.5 37.43.2 µmol h−1 cm−2 [67]CO2 + NO3 − Feacs 39.83643.65 mM h−1 gFe −1 [68]CO2 + NO3 − Ga/Y‐CNP22.141.9 mmol h−1 g−1 [69]CO2 + NO2 − Co–NiOx@GDY64.3913.2 µg h−1 mgcat −1 [41]CO2 + NO2 − Te–Pd NCs/C12.2—[10]CO2 + NO2 − AuCu SANFs24.73,889.6 µg h−1 mgcat. −1 [40]Gaseous CO2 couplingCO2 + NO3 − B‐Cu80.1101.2 µmol h−1 cm−2 [21]CO2 + NO3 − MIL‐125‐NH2‐derived TiO2 48.8843.37 mmol h−1 g−1 [4] Note: Urea yield rates are reported in the original units from the cited literature. Recent catalysts for NO3 −/NO2 − and CO2 reduction to urea. Note: Urea yield rates are reported in the original units from the cited literature. *CO2 + *NO2 → *CO2NO2 Formation of the *CO2NO2 intermediate occurs at an early stage of the co‐reduction, typically after the initial two‐electron reduction of nitrate to surface‐bound *NO2. Subsequently, adsorbed *CO2 on the catalyst surface couples with *NO2 to yield *CO2NO2. Since the coupling intermediates require relatively fewer hydrogenation steps, the mechanism tends to reduce by‐product formation and deliver higher urea selectivity at comparatively lower overpotentials [30]. Through operando synchrotron‐radiation Fourier transform infrared (SR‐FTIR) spectroscopy measurements and density functional theory (DFT) calculations, Lv et al. initially reported that the surface‐bound *CO2 can directly couple with absorbed *NO2 on In(OH)3‐S catalyst, particularly on the (100) facet [31]. As a first step, *NO2 adsorbs on coordinatively unsaturated indium sites, where electron transfer to the adsorbate enhances its reactivity toward C–N coupling. Simultaneously, electron transfer from the indium sites to *CO2 facilitates the C–N coupling step between *CO2 and *NO2 (Figure 2a). Notably, DFT calculations show that the energy barrier for *NO2 hydrogenation to *HNO2 (0.62 eV) is higher than that for coupling with *CO2 (0.35 eV), demonstrating that the C–N coupling step is kinetically favored (Figure 2b). As a result, In(OH)3‐S exhibits outstanding electrocatalytic performance, achieving a Faradaic efficiency of 53.4% for urea, as well as nitrogen selectivity of 82.9% and carbon selectivity approaching 100%. (a) Corresponding atomic configurations for each step of urea production on In(OH)3. (Adapted with permission from Ref. [31], copyright 2021 Springer Nature). (b) The proposed mechanism of urea synthesis through DFT on In(OH)3‐S. (Adapted with permission from Ref. [31], copyright 2021 Springer Nature). (c) Comparison of Gibbs free energy changes associated with urea formation on MoOx/C and MoO3. (Adapted with permission from Ref. [30], copyright 2023 Wiley‐VCH). (d) Online DEMS spectra of U‐Zn. (Adapted with permission from [32], copyright 2025, Royal Society of Chemistry). (e) In situ Raman spectra of CuWO4 in 0.1 M KNO3 under CO2 flow at various potentials and under Ar at open circuit. (Adapted with permission from Ref. [33], copyright 2023 Springer Nature). (f) Online DEMS spectra of CuSiOX in 0.1 M KNO3 and 0.1 M KHCO3 under CO2 at various potentials. (Adapted with permission from Ref. [34], copyright 2024 Wiley‐VCH). Similarly, Sun et al., reported that *CO2 directly couples with *NO2 to form *CO2NO2 on MoOx/C catalysts, thereby initiating urea formation [30]. Mechanistic insights into the urea formation pathway were unveiled by DFT calculations through identifying the most energetically favorable reaction route. In detail, the process of urea formation begins with the reduction of NO3 − to *NO2. Subsequently, *NO2 faces two competitive reaction pathways: coupling with *CO2 or hydrogenation to form *NO. Notably, *NO2 preferentially undergoes C–N coupling with *CO2 to form a *CO2NO2 intermediate, as this step requires a substantially lower energy barrier compared with its further reduction to *NO (Figure 2c). The lower C–N coupling energy barrier leads to a Faradaic efficiency of NO3 − reduction to NH3 as high as 98.14% in the absence of CO2. However, when CO2 is introduced, the Faradaic efficiency of NH3 decreases to 14.59%, and achieves a urea Faradaic efficiency of 27.7%. Wang et al., further demonstrated the *NO2 and *CO2 coupling mechanism on uncoordinated Zn nanosheets through online differential electrochemical mass spectrometry (DEMS) [32]. The signals of *NO2, *CO2NO2, *COOHNH2, and *CO(NH2)2 were observed, but the signals of *COOH and *CO were absent (Figure 2d). The result suggests that the coupling step occurs between *CO2 and *NO2. In summary, the C–N coupling process can directly occur between adsorbed *CO2 and *NO2, without the need for a deep hydrogenation process with these intermediates. In this mechanism, the C–N bond formation at the early stage of nitrate reduction allows *NO2 to participate in the coupling reaction before it further converts into more deeply reduced intermediates such as *NO or *NHx. This coupling mode effectively inhibits the competition of the continuous hydrogenation pathway in terms of kinetics, thereby reducing the tendency of the formation of ammonia and other by‐products, and facilitating the selection of the reaction to form urea. Therefore, the direct coupling of *CO2 and *NO2 provides an available reaction pathway for urea production. *CO + *NO2→*CONO2 *CO and *NO2 as coupling intermediates were proposed by Zhao et al. CO2 undergoes two hydrogenation steps to yield *CO, which then acts as the carbon‐containing intermediate [33]. Simultaneously, NO3 − is reduced to *NO2 through proton‐coupled electron transfer steps. These two surface‐adsorbed species, *CO and *NO2, then couple to initiate the C─N bond formation process. In situ Raman results revealed the peaks of *CO and *NO2 at 1979 cm−1 and 1428 cm−1, respectively. Notably, the peak intensity reached its maximum at −0.2 V vs. RHE, which coincided with the potential at which the highest urea yield and FEurea were observed (Figure 2e). This correlation provided evidence that *CO and *NO2 are likely the key intermediates involved in the C–N coupling process. Additionally, a series of carbon‐ and nitrogen‐containing reactants was tested to validate the proposed mechanism. The results revealed that when any downstream intermediates beyond *CO and *NO2 were introduced into the reaction system, urea formation was not detected, further supporting the role of *CO and *NO2 as key coupling intermediates on CuWO4. Qiu et al. also proposed the *CO + *NO2 coupling mechanism on a CuSiOX catalyst, which was validated through in situ attenuated total reflection Fourier transform infrared spectroscopy (ATR‐FTIR) spectroscopy, DEMS, and DFT calculations [34]. As shown in Figure 2f, the intermediate of *CONH (m/z = 43), *CONHO (m/z = 59), and *CONO2 (m/z = 74) were observed. Additionally, the *CONO2 intermediate was detected by in situ ATR‐FTIR spectroscopy, with a characteristic adsorption peak observed at 1695 cm−1. Among all experimentally detected intermediates, *CONO2 appears at the earliest stage of the hydrogenation pathway, suggesting its potential role as the initial product of C–N coupling. Generally, nitrate reduction (NO3RR) or nitrite reduction (NO2RR) initiates at more positive potentials than CO2 reduction (CO2RR), making it difficult to simultaneously sustain high surface coverages of *NO2 and *CO on most catalysts. To exploit the *NO2 and *CO coupling pathway, catalysts must be rationally engineered to (i) lower the overpotential for converting CO2 to *CO and/or (ii) raise the activation barrier for the further reduction of *NO2, thereby sustaining concurrent high coverages of both intermediates. *CO + *NO → *OCNO Another proposed C–N coupling mechanism involves the reaction between *CO and *NO to form *OCNO, which has been identified as a key intermediate in several recent studies [35, 36, 37]. Tan et al. introduced the coupling mechanism on Cu‐PMOF (Figure 3a) and proved that via in situ Attenuated Total Reflection Surface‐Enhanced Infrared Absorption Spectroscopy (ATR‐SEIRAS) [36]. The peaks at 1774, 2291, and 2156 cm−1 were attributed to the stretching vibrations of *NO, *CO, and *OCNO, respectively, demonstrating the *CO + *NO → *OCNO pathway (Figure 3b). Online DEMS was employed to identify the key intermediates involved in the urea synthesis process. Cu‐PMOF exhibited signals of at m/z = 30, 58, and 88, which correspond to *NO, *ONCO, and *ONCONO intermediates, respectively. They suggested that *NO and *CO were absorbed on the same Cu site and formed a *OCNO intermediate with no apparent structural changes. Therefore, this coupling process is kinetically favorable, as indicated by a negative Gibbs free energy change (ΔG < 0). The hydrogenation of *CO2 to *CO requires the highest energy barrier in the overall reaction (ΔG = 0.21 eV), so this step is identified as the rate‐determining step. Wei et al. uncovered a distinct *CO + *NO coupling pathway on VO‐enriched CeO2, in which *NO inserts into oxygen vacancies and couples directly with *CO to form *OCNO rather than conventional protonation routes (Figure 3c) [38]. The kinetic barriers for protonation and coupling steps were calculated through DFT, demonstrating that the energy barrier of *CO and *NO coupling reaction on VO‐enriched CeO2 only needs 0.27 eV, lower than *NO protonation (1.75 eV) and *CO coupling with *NH (0.85 eV)or *NH2 (0.85 eV). Meanwhile, Wang and his colleagues used operando SR‐FTIR to observe the coupling of *NO and *CO to form the key *OCNO intermediate, indicating that the first C–N bond could occur on the metal‐free catalyst (N‐C‐1000) [37]. These studies provide evidence that the *CO + *NO coupling pathway, involving the formation of *OCNO as a key intermediate, can occur on a variety of catalyst surfaces, including both metal‐based and metal‐free systems. (a) Schematic illustration of the urea synthetic routes for Cu‐PMOF. (Adapted with permission from Ref. [36], copyright 2025 Wiley‐VCH). (b) In situ ATR‐SEIRAS spectra at various applied potentials for Cu‐PMOF in 0.05 M NO3 −. (Adapted with permission from Ref. [36], copyright 2025 Wiley‐VCH). (c) Schematic diagram of *CO and *NO coupling mechanism on VO‐CeO2. (Adapted with permission from Ref. [38], copyright 2022 American Chemical Society). (d) In situ FTIR spectra collected in CO2‐saturated electrolyte with and without NO3 − at 0.05 V vs. RHE. (Adapted with permission from Ref. [21], copyright 2025 American Chemical Society). (e) cNO3‐ and pCO2‐dependent urea formation rate on B–Cu and Cu catalysts at 0.05 V vs. RHE. (Adapted with permission from Ref. [21], copyright 2025 American Chemical Society). *CO + *NH2 → *CONH2 The coupling pathway between *CO and *NH2 is another frequently proposed reaction route. This mechanism suggests that the C–N bond formation occurs at a relatively late stage of the overall reaction. Specifically, CO2 undergoes two proton‐coupled electron transfer steps to form *CO, while NO3 − is reduced through seven electron transfer steps to yield *NH2. Subsequently, *CO couples with *NH2 to form the *CONH2 intermediate, which then undergoes a second coupling step with another *NH2 species to finally generate urea. Liang et al. confirmed the *CONH2 mechanism on Ni‐CeO2‐X via in situ ATR‐FTIR substitution experiments with various C‐ and N‐sources [39]. They utilized CO2, CO, HCHO, and CH3OH as carbon sources and NO3 − as the nitrogen source. Urea formation was only observed when CO2 or CO was employed. These results demonstrate that *CO can act as a reactive intermediate in the C–N coupling process. When CO2 was used as the carbon source and KNO2, NH2OH, HCONH2, NH3, and NH4Cl were separately used as nitrogen sources, urea was detected only in systems with KNO2, NH2OH, or HCONH2. The detection of urea when using HCONH2 as the nitrogen source suggests that *CONH2 may serve as a key intermediate in urea synthesis. This intermediate is likely formed through the coupling of *CO and *NH2, indicating that *NH2 is a crucial N‐containing species involved in the C–N coupling process. Moreover, numerous studies on NO2 − reduction systems have identified *CONH2 as the initial C–N coupling intermediate in urea synthesis [10, 22, 40, 41]. For example, Feng et al. reported the *NH2 and *CO coupling mechanism in NO2 − and CO2 co‐reduction system on Te‐doped Pd nanocrystals [10]. The coupling mechanism of *CO and *NH2 relies on the deep reduction process of CO2 and NO3 −/NO2 −. Therefore, under this mechanism, the C–N coupling usually occurs in a more negative reaction potential range. In this mechanism, *CO and *NH2 represent the last active intermediates that the carbon‐containing and nitrogen‐containing intermediates, respectively, which can participate in coupling for generating urea. Further reduction of either intermediate diverts the reaction pathway toward competing side products such as CO and NH3. Therefore, in the *CO and *NH2 coupling mechanism, promoting the effective generation of both while inhibiting their excessive reduction is critical for achieving high‐selectivity urea synthesis. Gaseous CO2 Coupling Mechanism Aqueous amine solutions have been widely employed in industrial CO2 capture, particularly for treating low‐concentration CO2 streams such as postcombustion flue gas (10%–15% CO2), owing to the strong affinity between CO2 molecules and amino functional groups [42, 43]. During the NO3 − and NO2 − reduction processes, a variety of surface‐adsorbed ammonia‐related intermediates are commonly generated. Therefore, the ammonia intermediates formed during the electrochemical reduction of NO3 − or NO2 − may serve as active sites for C–N coupling through chemisorptive interaction with CO2. Hu et al. proposed a coupling mechanism in which *NO2 − captures CO2 through the Eley–Rideal (ER) mechanism to facilitate C–N bond formation [21]. In situ FTIR was used to uncover the C–N coupling mechanism. At 0.05 V vs. RHE, no *OCO signal was detected when only CO2 was present, suggesting that CO2 alone could not be activated on Cu. By contrast, in the CO2 + NO3 − system, both the C─N bond signal associated with urea and the *OCO signal were observed, supporting that C–N coupling proceeds via direct participation of gaseous CO2 (Figure 3d). They further investigated the C–N coupling mode by adjusting the concentration of NO3 − and the partial pressure of CO2. When the concentration of NO3 − increased from 0.01 to 0.2 M, the rate of urea production initially rose significantly, but gradually plateaued at approximately 0.06 M, indicating that the surface coverage of *NO2 had reached saturation. In contrast, as the partial pressure of CO2 increased from 0.1 to 1 atm, the rate of urea production continued to increase without showing a saturation trend (Figure 3e). This kinetic behaviour indicates that CO2 does not participate in the reaction through surface adsorption of intermediates but directly couples with the adsorbed *NO2 in gaseous form, which is consistent with an ER mechanism. In further research, the same ER mechanism was also observed on Fe2O3 and Cd‐Fe2O3 [44]. These studies collectively suggest that various nitrogen‐containing intermediates can interact with CO2, enabling alternative pathways for C─N bond formation. The gaseous CO2 coupling pathway offers a novel strategy to enhance urea selectivity. This route does not require prior activation of CO2 to *CO2, *COOH, or *CO, steps that typically demand more negative potentials. Employing this mechanism lowers the operating overpotential for urea electrosynthesis, thereby suppressing competing reactions. In addition, this mechanism provides an easy‐to‐handle and effective strategy to enhance the kinetic favorability of C–N coupling by increasing the local CO2 concentration at the catalyst surface. Thermodynamic and Kinetic Analysis for Different C–N Coupling Pathways As discussed above, different C–N coupling pathways are often manifested on different catalyst surfaces. This suggests that no single coupling pathway is universally favored. Instead, the dominant pathway depends on the thermodynamic and kinetic characteristics of the catalytic system. Furthermore, operating conditions, particularly the applied potential and electrolyte pH, could alter thermodynamic and kinetic characteristics. Therefore, it is important to consider the catalytic conditions in which these mechanisms are more likely to occur, as well as the fundamental thermodynamic and kinetic differences between pathways. For the *CO2 + *NO2 → *CO2NO2 pathway, C–N coupling occurs before extensive reduction of the carbon and nitrogen intermediates, which tends to suppress the formation of by‐products. However, the feasibility of this pathway is still constrained by the behavior of both C‐ and N‐intermediates. From a thermodynamic perspective, the adsorption of CO2 to *CO2 is often considered the rate‐determining step for CO2RR [45, 46]. As a result, on active CO2RR catalysts, after the formation of *CO2, further hydrogenation to *COOH or *OCHO is favorable. This suggests that preventing the further hydrogenation of CO2, thereby enabling its participation in C–N coupling at an early stage of reduction, is a consideration for this mechanism. In the NO3RR process, the rate‐determining step varies across catalysts. The hydrogenation of *NO2 on catalysts is generally regarded as one of the rate‐determining steps during NO3RR [47, 48]. Therefore, from an energetic perspective, the relatively high activation energy of *NO2 hydrogenation hinders its further reduction, making the coupling with *CO2 more competitive. From a kinetic perspective, this can also lead to the accumulation of *NO2 on catalyst surfaces, facilitating C–N coupling with *CO2. In contrast, on those catalysts where the hydrogenation of *NO2 is not the rate‐determining step, the hydrogenation of *NO2 is relatively easier, which kinetically suppresses the *NO2 + *CO2 coupling pathway. Overall, the early coupling pathway of *CO2 + *NO2 relies on the stability of *NO2 and the hindered process of *CO2 hydrogenation. Therefore, this mechanism is more common in systems where the adsorption of *CO2 is weak, or CO2RR activity is limited, and the energy barrier for *NO2 hydrogenation is relatively high. Different from the early coupling pathway, C–N coupling occurs at a later reduction stage, generally between *CO and *NO2 or *NO. From the energetic standpoint, C–N coupling between *CO and *NO2/*NO generally does not involve significant energy barriers. Therefore, this pathway has been proposed as feasible in many theoretical computational studies [27]. However, the actual occurrence is often limited by kinetic factors, namely whether C‐ and N‐intermediates can coexist within the same potential window. Specifically, the generation of *CO from CO2RR typically requires a more negative potential, while the hydrogenation of *NO2 and *NO usually occurs over a more positive potential range [33]. If the active potential ranges of these processes are mismatched, it becomes difficult to establish sufficient surface coverage of both *CO and *NO2/*NO simultaneously, which limits C–N coupling through this pathway. Therefore, these coupling pathways are more favorable on catalysts that combine a low overpotential for CO2 to CO conversion and relatively high energy barriers for *NO2/*NO hydrogenation. The *NH2 + *CO coupling pathway is a frequently discussed coupling route because it typically exhibits a relatively low energy barrier [10, 49]. However, the hydrogenation of *NH2 to *NH3 often also proceeds with a relatively low energy barrier [50, 51]. Therefore, from a kinetic perspective, this pathway is mainly constrained by the stability of *NH2. When hydrogenation of *NH2 towards *NH3 occurs rapidly, it is difficult to maintain an adequate surface coverage of *NH2, thereby limiting its coupling with *CO. This coupling pathway can become dominant only when the hydrogenation of *NH2 is sufficiently suppressed, allowing *NH2 to remain on the surface and maintain sufficient coverage for C–N bond formation. Consequently, this mechanism is more likely to occur in catalytic systems that can slow the hydrogenation of *NH2. Unlike the Langmuir‐Hinshelwood (LH) coupling pathways discussed above, which require the participation of two adsorbed intermediates, the ER mechanism allows gaseous CO2 to directly couple with surface nitrogen intermediates. Therefore, this pathway can occur in catalyst systems without CO2RR activity. Under this mechanism, the *CO2 adsorption step can be bypassed, thereby avoiding side reactions arising from the further reduction of carbon species [44]. Meanwhile, in a flow cell system, CO2 can maintain a relatively high local concentration at the interface, providing favorable kinetic conditions for direct coupling between gaseous CO2 and surface nitrogen species. However, the direct coupling of inactivated linear CO2 molecules with nitrogen species might require a higher reaction energy barrier. Compared to the coupling pathways involving *CO2, the thermodynamic feasibility of the ER process may be lower. Therefore, lowering the reaction barrier to gaseous CO2 coupling by regulating the catalytic interface or local environment might be a key challenge for promoting this C–N pathway. Through an in‐depth analysis and summary of current C–N coupling mechanisms, it is evident that the C─N bond formation pathway largely depends on the adsorption behavior of key intermediates and their associated reaction energy barriers on the catalyst surface. Different catalyst surfaces support distinct C–N coupling routes, which significantly influence the selectivity and energy efficiency of urea formation. Therefore, before focusing on enhancing catalyst performance, it is essential to first identify and understand the feasible C–N coupling pathways specific to a given catalytic system. This strategic approach is a critical prerequisite for advancing research in CO2 and NO3 −/NO2 − co‐reduction to urea. Intermediate Regulation Strategies for Efficient C–N Coupling Efficient electrocatalytic urea synthesis from CO2 and nitrate fundamentally relies on the formation of C–N bonds between carbon‐ and nitrogen‐derived intermediates at the electrode interface. Numerous catalyst design strategies have been reported to promote C–N coupling [7]. Currently, researchers mainly focus on enhancing the selectivity of electrochemical urea synthesis through refined material design. However, from a mechanistic perspective, regardless of how complex the catalyst structure is, the fundamental improvement in performance still stems from the regulation of the generation, stability, and transformation behavior of key intermediates during the reaction process. In other words, material design is not the goal; rather, it is through altering the interface electronic structure and adsorption properties to effectively guide the reaction pathways of intermediates, thereby ultimately influencing the C–N coupling efficiency and urea selectivity. Therefore, by concentrating the research strategy for enhancing urea selectivity on the regulation of the reaction behavior of key intermediates, it will help provide more mechanism‐oriented guidance for subsequent studies [52]. Accordingly, this chapter examines regulation strategies for C–N coupling from an intermediate‐centric perspective. We first summarise carbon‐side regulation approaches that control the generation, configuration, and availability of coupling‐active carbon intermediates. We then discuss nitrogen‐side regulation strategies that retain key nitrogen species and suppress their over‐hydrogenation toward ammonia. By organizing diverse catalyst design methods through the perspective of intermediate behavior, this chapter provides a unified framework for understanding how coordinated regulation of carbon and nitrogen reaction manifolds enables efficient urea electrosynthesis. Regulation of Carbon‐Derived Intermediates As for carbon‐containing species, surface‐bound *CO2 and *CO are widely regarded as the two most relevant carbon intermediates involved in C–N coupling during electrocatalytic urea synthesis. Notably, the optimal strategy for regulating carbon‐side activation is not universal but strongly depends on the specific C–N coupling pathway involved. Adsorption Configuration Regulation For the *CO2 + *NO2 → *CO2NO2 pathway, modulating the adsorption configuration of CO2 has been demonstrated to facilitate the C–N coupling process, as certain adsorption geometries are more favorable for effective bond formation with *NO2 [53, 54]. Zhou et al. reported a *CO2‐side activation strategy based on deliberate regulation of CO2 adsorption geometry (Figure 4a) [54]. By constructing dual‐metal hetero‐single‐atom sites with a FeN4/B2CuN2 coordination environment, the authors induced a strong antibonding interaction between CO2 and the Cu–B moiety, which selectively stabilizes a single O‐monodentate (*OCO) adsorption mode rather than conventional C‐linked or bridging O,O‐linked configurations. (a) The scheme of the single oxygen adsorption pathway for C–N coupling. (Adapted with permission from Ref. [54], copyright 2025 Wiley‐VCH). (b) In situ ATR‐FTIR spectra for FeN4/B2CuN2@NC. (Adapted with permission from Ref. [54], copyright 2025 Wiley‐VCH). (c) Free‐energy profiles of the CO2/NO3 − co‐reduction to urea on FeN4/CuN4@NC (blue) and FeN4/B2CuN2@NC (purple), together with the corresponding optimized reaction geometries. (Adapted with permission from Ref. [54], copyright 2025 Wiley‐VCH). (d) Bioinspired alternating Cu–W bimetallic site design in CuWO4 for urea electrosynthesis. (Adapted with permission from Ref. [33], copyright 2023 Springer Nature). (e) Faradaic efficiency and yield rates of urea for CuWO4 at different potentials. (Adapted with permission from Ref. [33], copyright 2023 Springer Nature). (f) CO yield rate and Faradaic efficiency of CuNi(OH)2 under CO2RR. (Adapted with permission from Ref. [72], copyright 2024 Wiley‐VCH). (g) Urea yield rates and Faradaic efficiencies obtained on Ag NPs, CuNi(OH)2 NSs, their physical mixture, and the Ag–CuNi(OH)2 composite. (Adapted with permission from Ref. [72], copyright 2024 Wiley‐VCH). (h) Differential charge density of Cu1Pd(111). (Adapted with permission from Ref. [33], copyright 2023 Springer Nature). Operando ATR‐FTIR spectroscopy, combined with isotope labelling experiments, revealed that this adsorption geometry enables the direct coupling between *OCO and *NO2 to form the *OCONO2 intermediate, effectively suppressing competing *CO2 and *COOH formation. Firstly, they employed liquid chromatography‐mass spectrometry/mass spectrometry (LC–MS/MS) to detect urea using 13CO2 or 12CO. They found that only m/z = 62 (13CO(NH2)2) was observed for FeN4/B2CuN2@NC, certifying that CO2 is the C‐containing coupling source rather than CO. Moreover, in situ ATR‐FTIR showed a peak at 1232 cm−1 for FeN4/B2CuN2@NC, which was assigned to *OCOH, and no *CO2 nor *COOH peaks were detected (Figure 4b), indicating the *OCO adsorption mode. KNO2, NH2OH, HCONH2, NH3, and NH4Cl were substituted for KNO3 to identify the nitrogen‐containing intermediates involved in the C–N coupling process. Only when KNO2 was used as a nitrogen source could urea be observed, which illustrates that *NO2 is the intermediate involved in C–N coupling. The results demonstrated that the C–N coupling mechanism on FeN4/B2CuN2@NC is *OCO + *NO2 → *CO2NO2. DFT calculations showed that the energy barrier of the first C–N coupling step *OCO and *NO2 was 0.19 eV, while that of *CO2 and *NO2 coupling was 1.02 eV (Figure 4c). As a result, compared with catalyst FeN4/CuN4@NC, which relies on *CO2/CO and *NO2 coupling, catalyst FeN4/B2CuN2@NC exhibited a higher Faradaic efficiency and urea yield, reaching 71.9% and 2072.5 µg h−1 mgcat. −1, respectively. These comprehensive experimental and theoretical results collectively validate that the *OCO + *NO2 pathway, enabled by single‐O adsorption, is energetically and kinetically more favorable for C–N coupling. Similarly, Li et al. reported that CO2 activation occurs via the formation of a specific In–O–C–O–In adsorption configuration on exposed (100) facets, which facilitates electron transfer to the surface‐bound CO2 intermediate [53]. This coordination mode provides a favorable structural environment for direct C–N coupling between *NO2 and *CO2 intermediates, thereby promoting efficient urea formation. The underlying reasons for promoting C–N coupling by regulating the adsorption configuration are as follows: the first is a difference in steric hindrance. When the adsorption configuration is the C‐linked *CO2, the carbon atom directly adsorbs on the catalyst surface and is spatially confined, leading to significant steric hindrance to coupling with nitrogenous intermediates. In contrast, in the O‐linked *OCO configuration, the carbon atom protrudes outward and is highly exposed, significantly reducing the steric hindrance and providing highly favorable geometric conditions for C–N coupling. The second reason may be that the C‐linked adsorption configuration allows direct electron transfer from the catalyst surface to the carbon atom, increasing its local electron density and reducing its electrophilicity. In contrast, in the O‐linked *OCO configuration, influenced by the electron‐withdrawing inductive effect of the metal‐oxygen interaction, the electron cloud shifts toward the oxygen atom [70]. This enhances the electrophilicity of the carbon atom, thereby thermodynamically promoting C–N coupling with the nucleophilic nitrogen intermediate. Promoting the Generation of *CO Intermediates In addition to *CO2, *CO has also been identified as a viable carbon intermediate capable of undergoing coupling with nitrogen‐derived species during electrocatalytic urea synthesis [36, 71]. Therefore, by promoting the reduction of CO2 to produce *CO, the possibility of the coupling reaction between *CO and nitrogen intermediates can be effectively enhanced [10]. This strategy has been exemplified in several catalyst systems. For instance, CuWO4 has been reported to enhance the C–N coupling process by promoting the generation of *CO [33]. In this system, Cu sites are responsible for facilitating CO2 to *CO conversion, while adjacent W‐based oxide motifs preferentially stabilize nitrogen‐derived intermediates such as *NO2 (Figure 4d). Researchers compared the urea production behaviors of WO3 and CuWO4 catalysts and found that the urea yield on WO3 reached its maximum at −0.8 V vs. RHE (61.7 ± 1.2 µg h−1 mg−1). However, after introducing Cu sites that could promote the generation of *CO, the peak urea yield of CuWO4 (99.5 ± 3.0 µg h−1 mg−1) significantly shifted to −0.2 V vs. RHE, and it was significantly higher than that of WO3 (Figure 4e). This result indicates that by enhancing the generation of *CO, the C–N coupling can be promoted. Ye et al. achieved a significant improvement in urea selectivity by introducing Ag sites that could promote the generation of *CO into the CuNi(OH)2 system [72]. The CO2 reduction test on CuNi(OH)2 only yields trace amounts of CO, indicating a low CO2 reduction activity (Figure 4f). Consequently, CuNi(OH)2 exhibited a lower urea yield in the co‐reduction system. In contrast, after loading Ag nanoparticles with CO2 reduction activity to *CO on CuNi(OH)2, Ag‐CuNi(OH)2 showed a significantly increased urea yield (Figure 4g). This result further highlights the significant role of the *CO generation ability in regulating the C–N coupling process. In the system where *CO serves as the primary coupling intermediate, promoting the reduction of CO2 to *CO is highly beneficial for increasing the surface *CO coverage. From a kinetic perspective, the rate of C–N coupling is typically positively correlated with the *CO coverage. Regulation of *CO Adsorption Strength Beyond improvements in *CO generation, regulating the adsorption strength of *CO on catalyst surfaces is another important strategy for enhancing C–N coupling [68]. Specifically, overly weak *CO adsorption results in rapid desorption, while excessively strong binding favors further hydrogenation to hydrocarbons or oxygenates, thereby limiting the participation of carbon species in urea formation [24, 39, 61, 73]. Therefore, an optimal *CO adsorption strength is required to balance surface stabilization and chemical reactivity. For instance, Zhao et al. effectively modulated the adsorption intensity of the *CO intermediate by regulating the exposed crystal planes of the Cu2O catalyst [73]. DFT calculations indicated that the binding of *CO to the Cu(110) surface was weak, leading to the desorption of *CO and weakening the effective coupling between *CO and *NH species. On the contrary, on the Cu(100) surface, the C–N coupling process is more likely to occur. This is mainly attributed to the fact that the Cu(100) surface has a stronger adsorption ability for *CO, thereby providing favorable conditions for *CO to participate in the subsequent C–N bond formation. The strategy of promoting C–N coupling by regulating the adsorption strength of *CO was also demonstrated in the work of Xu et al. [63]. On the Pd4Cu1/FeNi(OH)2 catalyst, the Pd sites serve to stabilize the *CO intermediate and prevent its premature desorption, thus providing more favorable reaction conditions for C–N coupling between *CO and the nitrogen source intermediate. DFT calculations revealed that on the Cu1Pd(111) single‐atom alloy surface, the free energy change for *CO desorption is increased relative to pure Pd surfaces, indicating that *CO is more strongly retained at the catalyst surface. Differential charge density analysis on the Cu1Pd(111) facet (Figure 4h) shows a clear electron delocalization from the Cu atom to the surrounding Pd atoms, which can be attributed to the higher electronegativity of Pd. This trend is further supported by Bader charge analysis, indicating that the Cu atom transfers approximately 0.21 e− to neighboring Pd atoms on the Cu1Pd(111) facet, while the pristine Pd(111) facet maintains an almost neutral charge distribution. Therefore, electrophilic *CO preferentially stabilizes on the electron‐rich Pd sites. As a result of the enhanced *CO adsorption capacity, the C–N coupling efficiency at the Pd4Cu1/FeNi(OH)2 site significantly increased, achieving a urea Faraday efficiency of up to 76.2% at −0.5 V vs. RHE. Additionally, a very small amount of CH4 was detected within the potential range of −0.1 to −0.3 V, which is also related to the enhanced *CO adsorption strength. Spatial and Interfacial Regulation of *CO for C–N Coupling Even when *CO is sufficiently generated and stabilized, effective C–N coupling critically depends on whether *CO is spatially positioned in close proximity to nitrogen‐derived intermediates. In conventional dual‐site catalysts, the diffusion length of surface‐bound *CO is typically very short, and *CO migration across heterogeneous interfaces is energetically unfavorable [74]. As a result, *CO and nitrogen‐derived intermediates remain spatially separated, which severely limits effective C–N coupling and hampers urea formation [75]. Therefore, engineering spatial and interfacial environments that confine or localise *CO near coupling‐active nitrogen sites represents another promising strategy for promoting C–N coupling toward urea synthesis [76, 77]. To overcome the spatial separation between *CO produced at CO2RR‐active sites and nitrogen‐derived intermediates, Li et al. introduced a *CO spillover strategy that enables effective migration of *CO toward coupling‐active sites, thereby facilitating urea synthesis (Figure 5a) [77]. In this system, palladium–gold alloys (Pd2Au1) and ruthenium dioxide (RuO2) serve as CO2RR‐active and NO3RR‐active components, respectively, enabling efficient generation of *CO and *NH2 intermediates. The Pd2Au1/RuO2 interface exhibits a relatively lower work function than Pd/RuO2 (Figure 5b,c). The decreased work function suppresses excessive interfacial charge accumulation, resulting in a lower energetic barrier for *CO migration. Accordingly, *CO generated on Pd2Au1 sites can more readily spillover across the interface toward RuO2, where nitrogen‐derived intermediates are preferentially stabilized. This facilitated *CO spillover enables effective spatial coupling between carbon‐ and nitrogen‐derived intermediates, ultimately promoting urea formation. (a) Schematic illustration of remote *CO spillover‐enhanced tandem urea electrosynthesis on metal/metal oxide catalysts. (Adapted with permission from [77], copyright 2025 Wiley‐VCH). Interfacial electronic configuration analysis of (b) Pd/RuO2 (ΔΦ = 0.44 eV) and (c) Pd2Au1/RuO2 (ΔΦ = 0.05 eV). (Adapted with permission from [77], copyright 2025 Wiley‐VCH). (d) High‐Angle Annular Dark‐Field Scanning Transmission Electron Microscopy (HAADF‐STEM) image of B‐FeNi‐DASC. (Adapted with permission from [76], copyright 2022 Springer Nature). (e) The product distributions of CO2RR, NO3RR and urea synthesis on Ni‐SAC, Fe‐SAC, I‐FeNi‐DASC, and B‐FeNi‐DASC at −1.4 V vs. RHE. (Adapted with permission from [76], copyright 2022 Springer Nature). Different from the aforementioned *CO spillover approach, Zhao et al. achieved the coupling of CO2RR and NO3RR functionalities by constructing a Ni–Fe diatomic pair (B‐FeNi‐DASC), in which the carbon‐ and nitrogen‐reduction sites are intrinsically linked at the atomic scale (Figure 5d) [76]. This design integrates the CO2RR and NO3RR active sites and deliberately regulates their spatial proximity, allowing the coupling of the C intermediate and the N intermediate. Specifically, Ni–SACs exhibit high activity toward the reduction of CO2 to CO, while Fe–SACs show favorable performance in converting NO3 − to NH3. However, when these two catalytic sites exist independently, the catalytic ability is inefficient in promoting the necessary C–N coupling for the formation of urea. Although the isolated diatomic Fe–Ni electrocatalyst (I‐FeNi‐DASC) is active for CO2RR and NO3RR individually, its contribution to urea electrosynthesis remains minimal. In contrast, the B‐FeNi‐DASC demonstrates markedly improved performance in urea electrosynthesis, indicating that Fe–Ni pairs play a vital role in enhancing the reaction kinetics of the C–N coupling process and overall reaction efficiency (Figure 5e). Regulation of Nitrogen‐Containing Intermediates Carbon‐side regulation focuses on the generation, stabilization, and spatial management of coupling‐active carbon intermediates; efficient urea electrosynthesis also critically depends on the behavior of nitrogen‐derived species [78, 79]. As discussed in Section 2, several nitrogen intermediates, including *NO2, *NO, *NH2, have been proposed to participate in C–N coupling through different reaction pathways. In practice, these intermediates tend to desorb readily or undergo rapid hydrogenation, which limits their effective interaction with carbon intermediates [80]. Accordingly, design strategies are required to address the instability of nitrogen intermediates during urea electrosynthesis. Generally, the intermediate can be prevented from desorbing or from being further reduced to NH3 by regulating its adsorption strength and inhibiting excessive hydrogenation. In the following sections, we have summarised several representative design strategies. Adsorption Strength Regulation of Nitrogen Intermediates Enhancing the surface retention of coupling‐active nitrogen intermediates, such as *NO2, is an effective nitrogen‐side regulation strategy for enabling efficient C–N coupling. During electrochemical reduction, *NO2 tends to desorb from catalyst surfaces [81, 82], resulting in an insufficient population available for subsequent C─N bond formation. Insufficient stabilization of *NO2 often limits its effective coupling with carbon species. Adsorption strength regulation focuses on tuning the interaction between nitrogen intermediates and catalytic surfaces to suppress their premature loss and maintain sufficient surface availability for coupling with carbon species. Recent studies have confirmed that this strategy effectively promotes C–N coupling [56, 57]. For instance, Yin et al. enhanced the adsorption capacity of *CO and *NO2 by modifying Cu‐ZIF‐8 with ionic liquids [57]. Compared with the original Cu‐ZIF‐8 (with *CO: −0.56 eV, *NO2: −0.25 eV), the IL@Cu‐ZIF‐8 modified by ionic liquids exhibits stronger adsorption capabilities for *CO and *NO2, with its adsorption energies reduced to −0.70 eV and −1.34 eV, respectively (Figure 6a). This enhanced adsorption stabilizes reactive intermediates and further facilitates subsequent C–N coupling. As a result, IL@Cu‐ZIF‐8 achieves a urea yield up to 140 µmol h−1 cm−2 (∼42 000 mg h−1 gcat −1) with a Faradaic efficiency of 55.3% at −0.5 V vs. RHE (Figure 6b,c). Moreover, Zhang et al. demonstrated that Bi‐doped Cu catalyst exhibits enhanced *NO2 adsorption relative to pristine Cu, which increases the surface population of *NO2 and consequently improves urea selectivity [56]. These results collectively demonstrate that reinforcing *NO2 adsorption is an effective nitrogen‐side strategy to prevent premature intermediate loss and promote C–N coupling. (a) The adsorption energies of *CO and *NO2 on IL@Cu‐ZIF‐8 and Cu‐ZIF‐8. (Adapted with permission from Ref. [57], copyright 2026 Springer Nature). (b) Faradaic efficiency of product distribution with applied potential for IL@Cu‐ZIF‐8. (Adapted with permission from Ref. [57], copyright 2026 Springer Nature). (c) Urea production rate with applied potential. (Adapted with permission from Ref. [57], copyright 2026 Springer Nature). Free energy profiles comparing the coupling of *NO, *N, *NH, and *NH2 with *CO, and the corresponding protonation steps, on VO‐enriched (d) and VO‐deficient CeO2 (e). (Adapted with permission from Ref. [38], copyright 2022 American Chemical Society). (f) Faradaic efficiencies of urea and NH4 + during CO2 and NO3 −/NO2 − co‐reduction at different reactant concentrations. (Adapted with permission from Ref. [66], copyright 2025 American Chemical Society). Suppression of Over‐Hydrogenation of Nitrogen Intermediates In addition to adsorption regulation, efficient urea electrosynthesis requires suppressing the over‐hydrogenation of nitrogen intermediates during nitrate reduction, as excessive hydrogenation readily drives these species toward ammonia formation at the expense of C–N coupling [83]. Once adsorbed, nitrogen intermediates such as *NO2, *NO, and *NH2 can readily undergo sequential proton‐coupled electron‐transfer steps, thermodynamically favoring deep hydrogenation to NH3 under reductive conditions. This competitive pathway substantially consumes nitrogen flux and reduces the availability of coupling‐active intermediates for C−N bond formation. Therefore, it is necessary to regulate the reaction process reasonably to inhibit the excessive hydrogenation of nitrogen intermediates, thereby preventing their deep reduction to NH3. Previous studies have shown that limiting the continuous hydrogenation of nitrogen intermediates is essential for preventing their diversion toward downstream intermediates and preserving their involvement in C–N coupling pathways [84]. Wei et al. suppress the further hydrogenation of N‐containing intermediates through the oxygen vacancies (VO) in CeO2, thereby promoting its coupling with *CO rather than hydrogenation toward downstream intermediates [38]. For vacancy‐deficient CeO2, the calculated energy barrier for *N–*CO coupling is lowest in all possible C–N coupling steps. However, *N is more prone to hydrogenation, which redirects the reaction toward downstream intermediates and ultimately leads to low urea selectivity (Figure 6d). After the introduction of VO, the activation energy barrier for the hydrogenation of nitrogen intermediates significantly increases. At this point, the coupling of *NO and *CO becomes the most likely C–N coupling step, and its reaction energy barrier (0.27 eV) is lower than that of the hydrogenation process of *NO (Figure 6e). In addition, to verify the promotion of suppressing the hydrogenation process on the formation of urea through C–N coupling, Liu et al. conducted comparative experiments on the FeP0.9S2.9‐x catalyst using NO3 − and NO2 − as nitrogen sources. The results showed that when NO2 −, which is more prone to the hydrogenation reaction, was used as the nitrogen source, the Faradaic efficiency of urea decreased significantly, while the selectivity of ammonia increased markedly (Figure 6f) [66]. DFT calculations further showed that the free‐energy barriers for *NH2 hydrogenation and *NH2–CO coupling are comparable, indicating that suppressing *NH2 over‐hydrogenation is essential for maintaining an effective C–N coupling pathway. Collectively, these studies demonstrate that suppressing the over‐hydrogenation of nitrogen intermediates is a general nitrogen‐side regulation strategy for preserving coupling‐active species and promoting efficient C–N bond formation during urea electrosynthesis. In this section, regulation strategies for electrocatalytic urea synthesis were summarised from both the carbon and nitrogen perspectives. Carbon‐side regulation primarily aims to control the generation, stabilization, and availability of coupling‐active carbon intermediates, whereas nitrogen‐side regulation focuses on retaining key nitrogen intermediates and suppressing their over‐hydrogenation toward ammonia. While these strategies are discussed separately for clarity, increasing evidence suggests that high urea selectivity often arises from the simultaneous implementation of multiple regulation directions, in which coordinated control over intermediate generation, retention, and reaction pathways across both manifolds is achieved. In Situ Characterization Techniques Uncovering C–N Coupling Mechanisms We have summarised a variety of C–N coupling pathways and intermediate regulation strategies that have been proposed. Most mechanistic insights into electrocatalytic urea synthesis are inferred from in/ex situ analysis and theoretical calculations [85]. Therefore, in situ characterization is essential for elucidating the detailed reaction mechanism. Techniques such as in situ Fourier‐transform infrared spectroscopy, Raman spectroscopy, and online DEMS enable real‐time monitoring of reaction intermediates [46], providing critical insights into the dynamic catalytic process and C–N coupling mechanism (Figure 7a). In this section, we introduce these characterization techniques and summarise representative studies in which they have been applied to probe C–N coupling mechanisms. (a) Schematic illustration of in situ characterization for probing reaction intermediates. (b) In situ infrared spectroscopy of m‐Cu2O. (Adapted with permission from Ref. [87], copyright 2023 Chinese Chemical Society). (c) In situ Raman spectroscopy of Pd4Cu1–Ni(OH)2. (Adapted with permission from Ref. [63], copyright 2023 Springer Nature). (d) Schematic illustration of the DEMS setup. (f) Online DEMS analysis of CuSn/CS−1. (Adapted with permission from Ref. [79], copyright 2025 Springer Nature). In Situ Fourier‐Transform Infrared Spectroscopy In situ FTIR spectroscopy is a widely used technique in electrocatalysis, which enables real‐time monitoring of surface‐absorbed intermediates under reaction conditions [86]. During the co‐reduction of CO2 and NO3 −, in situ FTIR spectroscopy has been employed to detect C‐containing, N‐containing, and C–N coupling intermediates. By monitoring intermediates along with their appearance, disappearance, intensity enhancement, and spectral shifts, in situ FTIR provides supportive evidence for elucidating the reaction pathway. For example, Qiu et al. detected two peaks at ∼3170 and ∼1890 cm−1 on multihole‐Cu2O (m‐Cu2O) through in situ FTIR spectra, which were assigned to the stretching vibrations of *NH2 and *NO, respectively (Figure 7b) [87]. Additionally, peaks at ∼1630 cm−1 and in the 1165–1316 cm−1 range corresponded to the stretching modes of C═O and C−O, respectively, indicating the presence of carbonaceous intermediates. Notably, the peak at approximately 2112 cm−1 is attributed to the vibrational mode of *ONCO that is likely formed through the coupling of *CO and *NO. These results demonstrate that m‐Cu2O is capable of simultaneously activating and reducing CO2 and NO3 −, while also facilitating the coupling between *CO and *NO intermediates. In Situ Raman Spectroscopy In situ Raman spectroscopy is an effective technique for probing both the structural evolution of catalysts and the formation of reaction intermediates during electrocatalytic processes [88]. In situ Raman spectroscopy enables the investigation of catalyst structural changes under different applied potentials or over time at a fixed potential, as well as the adsorption behavior of reaction intermediates on the catalyst surface [89]. In the field of electrocatalytic CO2 and NO3 − co‐reduction, in situ Raman spectroscopy has been rarely employed to monitor catalyst reconstruction during the reaction. Therefore, this section focuses exclusively on the application of in situ Raman spectroscopy for identifying surface‐bound intermediates and uncovering mechanistic insights into the C–N coupling process. Using in situ Raman spectroscopy at −0.5 V vs. RHE from 0 to 45 min, Ye et al., confirmed the C–N coupling mechanism on Pd4Cu1‐Ni(OH)2 catalyst [63]. The appearance of two vibrational peaks at 1216 and 1296 cm−1 can be attributed to the wagging modes of *NO2 and *NH2 intermediates, respectively. In addition, the peaks at 590, 1597, and 1683 cm−1 that appeared with high intensity at 45 min are attributed to the OCN bending mode, the δNH2 vibration of formamide (HCONH2), and the H‐bonded formamide signal, respectively (Figure 7c). Combined with the characteristic peak around 1000 cm−1, which is attributed to the C─N bond in urea, it can be inferred that *CONH2 serves as a key intermediate in the urea formation pathway. Additionally, Zhao et al. employed in situ Raman spectroscopy to understand the C–N coupling mechanism on the CuWO3 catalyst [33]. Two characteristic vibrational peaks at 1428 and 1979 cm−1 were observed, which were assigned to surface‐adsorbed *NO2 and *CO intermediates, respectively. The intensities of these peaks reached their maximum at −0.2 V vs. RHE, indicating a high coverage of both species. The maximum urea Faradaic efficiency is observed at this potential, implying that the simultaneous high coverage of *NO2 and *CO facilitates their coupling and promotes C─N bond formation. These results underscore the value of in situ Raman spectroscopy for elucidating the C–N coupling mechanism during electrochemical urea synthesis. However, the practical application of in situ Raman spectroscopy in probing surface intermediates remains limited due to the typically low surface coverage of key intermediates and their intrinsically weak Raman activity. Online Differential Electrochemical Mass Spectrometry Differential electrochemical mass spectrometry (DEMS) is a powerful in situ analytical technique that enables the real‐time detection and quantification of volatile or gaseous intermediates and products generated during electrochemical reactions. By coupling an electrochemical cell directly to a mass spectrometer via a gas‐permeable membrane, DEMS enables continuous monitoring of reaction species under working conditions without disturbing the reaction environment (Figure 7d). This technique has been extensively employed in CO2 reduction and NO3 − reduction to elucidate reaction pathways by identifying key intermediates and products such as *CO, *HCO, *NO2, *NO, *NH2OH, and *NH3. However, intermediates formed during C–N coupling are often difficult to detect due to their low volatility, short lifetimes, and low coverage. Therefore, while DEMS provides valuable insights into individual half‐reactions and precursor species, it is typically used as a supplemental tool to probe C–N coupling mechanisms. For instance, Wu et al., proposed a C–N coupling mechanism on CuSn/CS−1 [79]. Through online DEMS analysis, the CONH2 signal was observed under applied potential, with no CONO2, CONH, or CONHO signals. The results suggest that *CO and *NH2 are dominant intermediates in the C–N coupling process (Figure 7e). However, because many crucial reaction intermediates cannot be detected by DEMS, it remains an auxiliary technology to elucidate the C–N coupling mechanism. Conclusion and Perspectives Coupling CO2 and NO3 − electroreduction to synthesize urea offers a sustainable route toward carbon neutrality and efficient nitrogen removal from wastewater. This approach offers a promising strategy for producing green chemicals and promoting environmental remediation. This review focuses on recent advances in the electrocatalytic co‐reduction of CO2 and NO3 −, emphasizing C–N coupling mechanisms, the regulation of reaction intermediates, and in situ characterization approaches. Firstly, five representative C–N coupling mechanisms were reviewed, including: (i) the reaction between *CO2 and *NO2 forming *CO2NO2, (ii) *CO with *NO2 forming *CONO2, (iii) *CO with *NO yielding *OCNO, (iv) *CO coupling with *NH2 to form *CONH2, and (v) direct coupling between gaseous CO2 and N intermediates. In addition, several intermediate‐regulation strategies that promote C–N coupling were discussed. Finally, in situ electrochemical characterization techniques employed to elucidate the C–N coupling mechanisms were summarized. However, the current understanding of the C–N coupling mechanism is still limited, mainly because of the limitations of current characterization methods and theoretical calculation accuracy. In this regard, the following potential perspectives may contribute to a deeper understanding of the C–N coupling mechanism, guide rational catalyst design and ultimately enhance urea production performance. Developing More Advanced in Situ/Operando Characterization Techniques Currently, in situ characterization methods cannot directly elucidate the C–N coupling mechanism [90]. Understanding this process relies predominantly on detected intermediate signals, integrated with theoretical calculations and relevant experimental observations for indirect interpretation. This limitation is mainly because the existing in situ characterization techniques are unable to directly observe the initial steps of the formation of the C─N bond. For example, the observation of *CONH2 in in situ FTIR does not definitively prove that the C–N coupling occurred between *CO and *NH2, as the coupling might have taken place at an earlier stage, and *CONH2 could merely be a downstream intermediate formed after subsequent hydrogenation steps. To better understand the C–N coupling mechanism, it is necessary to develop characterization techniques that can more effectively monitor intermediates during the co‐reduction of CO2 and NO3 −. Improving temporal resolution and enhancing signal sensitivity are two promising development directions for existing in situ characterizations. Under electrochemical operating conditions, many C–N coupling intermediates are highly transient and have low surface coverage, making them difficult to detect by conventional steady‐state in situ spectroscopy [91]. Accordingly, improving temporal resolution through the integration of time‐resolved or rapid‐scan spectroscopic techniques therefore represents a critical step toward capturing these short‐lived intermediates before they evolve into more stable products. Meanwhile, strategies that enhance spectroscopic sensitivity are required to access intermediates with intrinsically weak signals. In this regard, enhanced Raman surface‐enhancement approaches provide a representative example. The amplified vibrational signals of adsorbed reaction intermediates enable the detection of previously inaccessible species, particularly those formed at the early stages of C–N coupling. In addition to the improvements to the existing characterizations, the development of new in situ and operando characterizations is equally important. Several emerging approaches, for example, in situ soft X‐ray spectroscopy and electrochemical scanning probe microscopies, offer new opportunities for probing C–N coupling processes. By resolving electronic structure, surface bonding, and reaction kinetics, these advanced techniques may ultimately enable direct observation of C─N bond formation. Furthermore, the inability to directly observe the C–N coupling process might also be due to the fact that in situ characterizations cannot fully reflect the actual operating conditions. Although many in situ characterizations are conducted in simulated reaction environments, they differ from the electrode polarisation state and local environmental conditions (local pH and ion concentration) in the actual reaction process. The intermediate signal obtained under non‐realistic conditions may not correspond to the key intermediates in the actual reaction pathway. Therefore, even if certain C–N coupling species are detected under in situ conditions, they may not be the C–N intermediates in the actual process. In contrast, operando characterizations can better reflect the information of intermediates under actual reaction conditions. Therefore, developing operando characterization techniques applicable to the electrochemical co‐reduction of CO2 and NO3 −/NO2 − to urea is expected to help better explore the C–N coupling mechanism. The complex solid–liquid interface environment often leads to severe background signal interference and peak overlap in in situ/operando spectra. Therefore, in situ/operando characterization techniques should be combined with isotope labelling experiments (13CO2, 15NO3 −). By comparing and observing the isotope shifts of characteristic peaks of key intermediate species, interference from interface solvent molecules and background ions would be effectively excluded. This avoids incorrect identification of the intermediate species and provides more accurate experimental evidence for the C–N coupling mechanism. It is worth noting that because multiple C–N coupling pathways may coexist in this reaction system, even if all potential C–N coupling intermediates are observed (such as *CO2NO2, *COOHNO2, *CONO2), the species appearing earliest along the pathway cannot be considered as the key intermediate for the dominant C–N coupling. Therefore, establishing a quantitative relationship between spectral peak intensity and surface coverage is helpful for accurately analyzing the actual coverage of each intermediate, and thereby determining which C–N coupling mechanism is most dominant. Improving the Reliability and Relevance of Theoretical Calculations DFT calculations are the main method for studying the C–N coupling mechanisms. It can directly simulate the C–N coupling process and simulate the free energy for every hydrogenation and coupling step. However, current theoretical calculations still face some challenges in understanding the C–N coupling mechanism. Generally, the determination of reaction mechanisms relies on calculating the free energy of reaction pathways on ideal catalyst surfaces [92]. However, ideal models are difficult to accurately reflect realistic electrochemical conditions. In order to enhance the accuracy of theoretical calculations in describing realistic reaction conditions, future efforts can be made to improve physical model construction and kinetics analysis. For the model construction, the theoretical calculations should gradually translate from idealized models to realistic electrochemical reaction environments. Introducing key interface factors, such as electric field effects, solvent structure, and surface structure flexibility during the calculation process, helps to enhance the accuracy of theoretical models in describing the actual reaction conditions. In addition, the analysis should not only focus on adsorption energies and reaction energy barriers obtained from thermodynamic analyses. Instead, kinetic analysis should also be considered. For the C–N coupling reaction, some important intermediates have short lifetimes and low surface coverage. These species may be difficult to couple with other intermediates, although the energy barrier for their coupling is relatively low. So, thermodynamic analysis is often difficult to depict real reactions accurately. Therefore, to determine the actual C–N coupling steps, it is essential to calculate the C–N coupling mechanism through introducing kinetic analysis. Synergistic of Computational Chemistry and Experiment for Mechanism Study Studying the reaction process together with computational calculation and experimental characterizations is promising to further reveal the C–N coupling mechanism. In past decades, although in situ characterization techniques and theoretical calculations have developed rapidly, the integration between theory and experiment remains insufficient. Currently, synergy between theoretical calculations and experiment is still mainly as complementary approaches, rather than a fully integrated framework with reciprocal feedback and constraint. To further integrate computational calculations and experiments, several key descriptors that can be tested experimentally and effectively distinguish different C–N coupling mechanisms should be employed. Such descriptors should enable computational calculations and experiments to constrain each other and enhance their cooperation in analyzing C–N coupling mechanisms. In this context, the physical model is established to generate concrete descriptors that can be directly evaluated based on experimental data, rather than only to identify thermodynamically favorable reaction pathways. Correspondingly, experiments should not only demonstrate the existence of intermediates but also quantify reaction trends and their responses to reaction conditions. By strengthening the connection between computational calculation and experiments, they can clearly and effectively uncover crucial mechanisms in C–N coupling, which further guides the design of urea electrosynthesis catalysts. Author Contributions Ziyun Wang: writing – review and editing, conceptualization, writing – original draft, funding acquisition, resources, supervision, formal analysis, project administration. Xingbao Chen: writing – original draft, validation, visualization, writing – review and editing. Zhuo Li: writing – original draft, visualization. Paul Kilmartin: project administration, supervision, formal analysis, visualization, resources. Lebin Cai: visualization. Bao Yu Xia: writing – original draft, writing – review and editing, supervision. Conflicts of Interest The authors declare no conflicts of interest. Acknowledgements This study is supported by the Marsden Fund Council from Government funding (25‐UOA‐003) and Catalyst: Seeding General Grant (24‐UOA‐048‐CSG), managed by Royal Society Te Apārangi. X.C. acknowledges the China Scholarship Council for a scholarship. Open access publishing facilitated by The University of Auckland, as part of the Wiley ‐ The University of Auckland agreement via the Council of Australasian University Librarians. Biographies Xingbao Chen is currently a Ph.D. student in the School of Chemical Sciences at the University of Auckland. He received his master's degree in Materials Science and Engineering from Wuhan University of Technology. His research interests focus on electrocatalysis, especially CO2 reduction and urea electrosynthesis. His work emphasizes catalyst design and reaction mechanism understanding through carefully controlled experiments and in situ characterization. Zhuo Li is a Ph.D. student in the School of Chemical Sciences at the University of Auckland. His research focuses on aqueous interfacial electrochemistry for advanced energy‐storage systems, with a particular emphasis on aqueous zinc‐ion batteries. He investigates the decoupling of Zn–H2O interactions in bulk electrolytes, integrating first‐principles calculations and finite‐element simulations to elucidate the underlying mechanisms and guide materials design. Lebin Cai is a Ph.D. student under the guidance of Prof. Bao Yu Xia at Huazhong University of Science and Technology (HUST). His current research primarily focuses on the design and construction of advanced materials for water electrolyzers, as well as fundamental research and scaling‐up strategies for electrochemical carbon dioxide capture. Paul Kilmartin is a professor at the University of Auckland, where he has been on the academic staff since completing his Ph.D. in 1997. Paul's scientific training is in electrochemistry, coupled with the use of chromatographic, spectroscopic and surface analysis techniques. Paul's main areas of research are evenly divided between wine chemistry and electrochemistry, with projects on chemical sensors and organic electrosynthesis, including several international collaborations. Bao Yu Xia is currently a full professor in the School of Chemistry and Chemical Engineering at Huazhong University of Science and Technology (HUST), China. His research interests are in the areas of nanostructured functional materials and their application in sustainable energy and environmental technologies. Ziyun Wang is a computational chemist specializing in theoretical catalysis and machine learning. He conducted his doctoral research in theoretical catalysis at Queen's University Belfast and the University of Manchester under the guidance of Prof. Peijun Hu and Prof. Chris Hardacre. Subsequently, he conducted postdoctoral research at Stanford University with Prof. Jens Nørskov and Prof. Thomas Bligaard at Stanford University, and with Prof. Edward Sargent at the University of Toronto. In 2021, Dr. Wang joined the School of Chemical Sciences at the University of Auckland, where he is currently an Associate Professor. Contributor Information Bao Yu Xia, Email: byxia@hust.edu.cn. Ziyun Wang, Email: ziyun.wang@auckland.ac.nz. Data Availability Statement Data sharing not applicable to this article as no datasets were generated or analysed during the current study. References Associated Data Data Availability Statement Data sharing not applicable to this article as no datasets were generated or analysed during the current study.