Reversing the Hydrogenation Pathways of Nitrogen‐Containing Intermediates for the Kinetics‐Matched Urea Electrosynthesis

ABSTRACT

Urea electrosynthesis from carbon dioxide (CO₂) and nitrate (NO₃ ⁻) is a promising sustainable route. However, the kinetic mismatch between key intermediates remains the major challenge for achieving selective C‒N coupling. Herein, indium‐doped titanium dioxide (In‐TiO₂) nanofibers were developed to regulate the hydrogenation pathway for realizing kinetics‐matched urea electrosynthesis. In situ spectroscopic analysis and theoretical calculations reveal that In doping reverses the hydrogenation pathway of nitrogen‐containing intermediates from the Eley‐Rideal (E‐R) to the Langmuir–Hinshelwood (L–H) mechanism. This shift is attributed to the sufficient *H supply guaranteed by the regulated interfacial water structure. Such reversed hydrogenation pathway balances *H utilization between CO₂ and NO₃ ⁻ reduction, enabling well‐matched formation kinetics of key intermediates for efficient C‒N coupling. Owing to the above merits, In‐TiO₂ achieved the remarkable average urea yield rate of 56.5 mmol h⁻¹ g⁻¹ with a Faradaic efficiency of 32.8%. This work provides mechanistic insights into the hydrogenation pathways regulation for efficient urea electrosynthesis.

The kinetics‐matched urea electrosynthesis is realized through reversing the hydrogenation pathway of N‐intermediates from Eley‐Rideal to the Langmuir‐Hinshelwood mechanism on In‐TiO₂. This competitive utilization of *H can suppress the over‐hydrogenation of N‐intermediates, which enables the selective C‒N coupling between *NH and *CO intermediates. The as‐developed catalyst also achieves efficient urea electrosynthesis utilizing flue gas.

1. Introduction

Urea (CO(NH₂)₂) is an essential nitrogen fertilizer in modern agriculture and a key feedstock in the chemical manufacturing [1, 2, 3]. Conventional urea production relies on high‐pressure and high‐temperature processes, which incur substantial fossil energy consumption and severe pollution [4, 5, 6]. In contrast, urea electrosynthesis powered by renewable electricity enables the use of nitrate‐containing wastewater and CO₂ as sustainable feedstocks [7, 8]. This strategy offers significant potential for urea production under mild reaction conditions while delivering notable environmental benefits [9]. However, the electrochemical synthesis of urea is fundamentally hindered by low C‒N coupling selectivity [10].

The C‒N coupling process constitutes a critical step in urea electrosynthesis, which depends on the individual CO₂RR and NO₃RR pathways to furnish carbon‐ and nitrogen‐based intermediates [11, 12]. Most studies have centered on promoting the formation of key intermediates towards C‒N coupling [13, 14, 15], such as *CO, *NO₂, *NO, *NH, and *NH₂ [16, 17, 18, 19]. For instance, Wang et al. revealed that electron‐deficient Cu enhances the adsorption of key intermediates, including *CO and *NO [20]. Zhang et al. reported a Fe(a)@C‐Fe₃O₄/CNTs catalyst, in which the Fe(a)@C and Fe₃O₄ active sites selectively facilitate the formation of *CO and *NH₂ intermediates [21]. Nevertheless, the selective C‒N coupling remains hindered by severe kinetic mismatches in intermediates generation, which may stem from the over‐hydrogenation of nitrogen‐containing intermediates and sluggish CO₂ hydrogenation [22, 23, 24, 25].

Previous studies generally held the view that the hydrogenation of CO₂RR/NO₃RR intermediates followed the Eley‐Rideal (E‐R) mechanism [26, 27, 28]. In that circumstance, CO₂ and NO₃ ⁻ obtain protons (H⁺) from solvent water to be hydrogenated through the proton‐coupled electron transfer (PCET) process [29, 30]. Recent reports suggested that the hydrogenation of carbon‐ and nitrogen‐containing intermediates could also proceed through the Langmuir–Hinshelwood mechanism [31, 32]. Such hydrogen transfer route relies on the active hydrogen (*H) generated via interfacial water dissociation on the catalyst surface. For example, Lum et al. found that the L–H mechanism was the dominant hydrogen transfer route during CO₂RR process as evidenced by the experiments conducted in mixed H₂O‐D₂O electrolytes [33]. Wu et al. demonstrated that the Mo sites in Cu‐MoS₂/MoO₂ could provide sufficient *H to enable efficient C‒N coupling by accelerating water dissociation [34]. Given the complexity of intermediates hydrogenation mechanism, it is pivotal to elucidate and regulate the hydrogen transfer routes for achieving well‐matched intermediate formation kinetics towards C‒N coupling [35]. In particular, for the largely unregulated NO₃RR process, excessive or frustrated hydrogenation favors the generation of NH₃ or NO₂ ⁻ by‐products, respectively [36]. To circumvent the issue of kinetic mismatch, the hydrogen transfer route in generating nitrogen‐containing intermediates should be rationally regulated, which still lacks exploration.

Herein, the well‐matched kinetics for C‒N coupling was realized over an indium doped titanium (In‐TiO₂) catalyst by reversing the hydrogenation pathway of nitrogen‐containing intermediates, which led to a urea yield rate of 56.5 mmol h⁻¹ g⁻¹ and Faradaic efficiency (FE) of 32.8%. Based on the systematic in situ differential electrochemical mass spectrometry (DEMS) investigations, the hydrogenation process of the nitrogen‐containing intermediates was reversed from E–R to L–H route, while the carbon‐containing intermediates intrinsically proceeded via the L–H mechanism. The balanced utilization of *H under the L–H mechanism could suppress the over‐hydrogenation of *NH intermediates into *NH₂ and NH₃, thereby ensuring the well‐matched kinetics between *NH and *CO generation for selective C‒N coupling. In addition, the hydrogenation of the carbon‐containing intermediate was also accelerated due to the reduced relative Gibbs free energy (ΔG) and sufficient *H supply, which collectively contributed to the improved kinetic match toward urea formation (Figure 1). This work not only deepens the understanding of hydrogenation pathways for key‐intermediates, but also provides a new perspective for modulating reaction kinetics toward electrochemical C‒N coupling.

FIGURE 1.

Schematic illustration of reversed hydrogenation pathway during the urea electrosynthesis process on TiO₂ and In‐TiO₂.

2. Results and Discussion

2.1. Synthesis and Characterization of Catalysts

The TiO₂ and In‐TiO₂ catalysts were synthesized via the electrospinning method followed by calcination (for details, please see methods in Supporting Information). Field‐emission scanning electron microscopy (FESEM) images reveal that both the precursor fibers and the calcined catalyst nanofibers possess continuous nanofibrous morphologies with smooth surfaces (Figures S1 and S2). Transmission electron microscopy (TEM) images of In‐TiO₂ show the average diameter of approximately 100 nm, which is almost the same as that of pristine TiO₂ (Figure S3). As shown in Figure 2a, the x‐ray diffraction (XRD) pattern of TiO₂ exhibits the diffraction peaks at 25.3°, 37.8°, 48.0°, and 55.1°, which can be indexed to the (101), (004), (200), and (211) planes of anatase TiO₂ (PDF# 21‒1272) [37]. The representative In‐TiO₂ catalyst with the optimized doping content of 1.5 at% displays the nearly identical characteristic peaks to those of pristine TiO₂, no obvious change in the anatase structure after indium doping (for In‐TiO₂ with other doping amounts, please see corresponding characterizations in supporting information, Figures S4–S7). The Raman spectra of TiO₂ and In‐TiO₂ further confirm the anatase phase (Figure S8) [38]. Determined by x‐ray photoelectron spectroscopy (XPS), two peaks are located at 444.6 and 452.1 eV (Figure 2b), which can be assigned to the In 3d_5/2 and 3d_3/2 of In³⁺, respectively [39]. A pair of Ti⁴⁺ peaks at 458.6 and 464.4 eV are observed for In‐TiO₂ without any noticeable shift in binding energy relative to pristine TiO₂, indicating the unchanged chemical state of Ti after Indium doping (Figure 2c). High‐resolution TEM (HRTEM) images display the lattice spacing of 0.238 and 0.347 nm, corresponding to the (004) and (101) crystal planes, respectively (Figure 2d) [40]. The abovementioned (004), (101), and (200) crystal planes can also be confirmed by the selected area electron diffraction (SAED) pattern (Figure 2e). High‐angle annular dark‐field scanning transmission electron microscopy (HAADF‐STEM) image, complemented by the elemental mapping images, could illustrate the uniform dispersion of In, Ti, and O elements on an individual In‐TiO₂ nanofiber (Figure 2f). The content of Indium on TiO₂ is determined as approximately 1.5 at% from the associated energy‐dispersive x‐ray spectroscopy (EDS) spectrum, consistent with the nominal doping level (Figure S9, Table S1). The inductively coupled plasma optical emission spectrometer (ICP‐OES) validates the presence of Indium element (Table S2).

FIGURE 2.

Structural and morphological characterization of In‐TiO₂. (a) XRD patterns. (b) In 3d XPS spectrum. (c) Ti 2p XPS spectra. (d) HRTEM image of In‐TiO₂ and (e) corresponding SAED pattern. (f) HAADF‐STEM image and the corresponding EDS elemental mappings of In‐TiO₂.

2.2. Electrocatalytic Performance of Urea Synthesis

The electrochemical performance of urea synthesis was rigorously investigated using an H‐cell equipped with a three‐electrode system under ambient conditions (please see testing and quantification details in experimental sections). As displayed in the linear sweep voltammetry (LSV) curves (Figure S10), the increased current density can be attributed to reactions involving CO₂ when switching the atmosphere [41]. In contrast to pristine TiO₂, the representative In‐TiO₂ (for In‐TiO₂ with other doping amounts, please see corresponding urea electrosynthesis results in Figures S14–S21) exhibits superior urea electrosynthesis activity across the potential range of −0.9 V to −1.3 V versus reversible hydrogen electrode (RHE). In‐TiO₂ achieves a maximum average urea yield rate of 56.5 mmol h⁻¹ g⁻¹ at −1.3 V versus RHE (Figure 3a), which is more than double that of TiO₂ (26.7 mmol h⁻¹ g⁻¹). The maximum urea FE could reach up to 32.8% at −1.0 V versus RHE (Figure 3b), substantially exceeding that of pristine TiO₂ (5.14%). The concentration of urea was further confirmed via ¹H nuclear magnetic resonance (NMR) spectroscopy (Figure S22). The urea yield rates determined by ¹H NMR measurements are in good agreement with that by UV–vis spectroscopy (Figure S22c), independently validating the quantification accuracy. Additionally, the source of urea was investigated by isotope‐labeling ¹H NMR experiments. The single peak at 5.62 ppm for CO(¹⁴NH₂)₂ and doublet peaks at 5.54 and 5.69 ppm for CO(¹⁵NH₂)₂ are observed (Figure S22d) upon using ¹⁴NO₃ ⁻ and ¹⁵NO₃ ⁻ as nitrogen sources, respectively, verifying the origin of nitrogen in urea. Control experiments display no detectable urea in the absence of CO₂, NO₃ ⁻, catalyst, or applied potential (Figure S23), further confirming the urea electrosynthesis process via the co‐reduction of CO₂ and KNO₃. Electrochemical active surface area (ECSA) measurements display comparable values for pristine TiO₂ and In‐TiO₂, suggesting that the enhanced electrocatalytic performance of In‐TiO₂ originates from its intrinsically higher activity instead of differences in ECSA (Figure S24). The Faradaic efficiency distribution of all products on TiO₂ and In‐TiO₂ are shown in Figures 3c and S25–S27. In stark contrast with pristine TiO₂, the increase in the nitrite‐producing FE as well as the concomitant decrease in the ammonia‐producing FE are observed on In‐TiO₂, implying slower hydrogenation of nitrogen‐containing intermediates. The accumulation of CO together with suppressed H₂ evolution indicates promoted CO₂ activation and inhibited hydrogen evolution reaction (HER) activity on In‐TiO₂, respectively. The above changes suggest the shift in hydrogenation preference induced by Indium doping, which favors CO₂ activation and C‒N coupling rather than the hydrogenation of nitrogen‐containing species. The urea FE and yield rate of In‐TiO₂ catalyst remain highly stable over 16 h (Figure 3d). By the analysis of XRD, Raman, TEM, HRTEM, SAED, and HAADF‐STEM performed for In‐TiO₂ after the catalytic reaction, the crystalline phase and nanostructure can be retained, which plays a critical role in guaranteeing the stability in urea electrosynthesis performance (Figure S28–S30 and Table S3). When using simulated flue gas (N₂ ∼82%, CO₂ ∼15%, and O₂ ∼3%) as the feeding gas, In‐TiO₂ still achieves an average urea yield rate of 33.5 mmol h⁻¹ g⁻¹ with the FE of 17.6% (Figures 3e,f). These results verify that In‐TiO₂ favors C‒N coupling even under complex gas compositions. Such direct utilization of flue gas has the potential to reduce the cost associated with additional gas separation systems and enables more sustainable urea production [42, 43].

FIGURE 3.

Urea electrosynthesis performance. (a) Urea yield rates at different potentials in 0.1 M KNO₃ electrolyte with CO₂ feeding gas and (b) corresponding FEs. (c) FEs of different products for TiO₂ and In‐TiO₂ at −1.0 V versus RHE in 0.1 M KNO₃ electrolyte with CO₂ feeding gas. (d) Long‐term durability test for In‐TiO₂. (e) Urea yield rates at different potentials in 0.1 M KNO₃ electrolyte with flue gas and (f) corresponding FEs.

2.3. Hydrogenation Mechanism and C‒N Coupling Pathway

Although the product distribution displays different hydrogenation behaviors between TiO₂ and In‐TiO₂, the underlying hydrogenation pathways governing CO₂RR and NO₃RR process remain unclear. To elucidate this, the electrocatalytic tests were performed with tert‐butyl alcohol (TBA), which could serve as a selective quenching agent of hydrogen radical [44]. In the individual CO₂RR tests (for details, please see Supporting Information), In‐TiO₂ exhibits markedly higher CO yield rate than pristine TiO₂ (Figure S31), confirming the promoted CO₂ activation. Upon the addition of 1 M TBA, the CO yield rates on both TiO₂ and In‐TiO₂ decrease (Figure 4a). This corresponds to the typical L–H pathway, during which the CO₂ hydrogenation relies on the *H [22]. As for the individual NO₃RR tests, the NH₃ formation is suppressed by indium doping (Figure S32). When the TBA concentration increases (from 0 to 1 M), the NH₃ yield rates on TiO₂ remain nearly unchanged (Figure 4b). In sharp contrast, the NH₃ yield rates on In‐TiO₂ decrease gradually under the same conditions. Such results suggest that indium doping converts the dominant NO₃RR hydrogenation mechanism from E–R to L–H. Because both NO₃RR and CO₂RR proceed via the L–H mechanism on In‐TiO₂, the surface *H species are simultaneously consumed. The balanced *H utilization avoids the excessive hydrogenation toward NH₃, which can guarantee the sufficient supply of nitrogen‐containing intermediate (*NH) for selective C‒N coupling.

FIGURE 4.

Mechanism investigations of hydrogenation mechanism and C‒N coupling pathway. (a) CO yield rates of individual CO₂RR with different TBA concentration at −1.0 V versus RHE. (b) NH₃ yield rates of individual NO₃RR with different TBA concentration at −1.0 V versus RHE. (c) In situ DEMS signals from urea electrosynthesis for TiO₂. (d) Schematic illustration of the proposed reaction mechanism for TiO₂ and In‐TiO₂. (e) In situ DEMS signals from urea electrosynthesis for In‐TiO₂.

To identify the possible intermediates and elucidate the C‒N coupling reaction pathway, in situ differential electrochemical mass spectrometry (DEMS) was first performed under conditions where the nitrate and CO₂ coexisted (Figure S33). As shown in Figure 4c, e, the signals of COOH⁺ (m/z = 45) and CO⁺ (m/z = 28) detected on both TiO₂ and In‐TiO₂ confirm that the CO₂RR proceeds via the “CO₂→ *COOH → *CO” route. Unlike following the same CO₂RR pathway, the NO₃RR routes on In‐TiO₂ differs markedly from that on pristine TiO₂. As for pristine TiO₂, the signals at m/z of 46, 47, 30, 31, 14, 15,16, and 17, along with the absence of signal at 32, suggest the “*NO₂ → *NO₂H → *NO → *NOH/*NHO → *N → *NH → *NH₂ → *NH₃” pathway (Figure 4c). When using the In‐TiO₂ catalyst, the signals at m/z of 46, 47, 30, 31, 32, 15, 16, and 17 can be detected, while the signal corresponding to N⁺ (m/z = 14) is absent (Figure 4e). Such stark contrast suggests that the NO₃RR pathway is altered to “*NO₂ → *NO₂H → *NO → *NOH/*NHO → *NHOH → *NH → *NH₂ → *NH₃” after incorporation of indium. Based on the above *H reliance analysis, the reversion of nitrogen‐containing species on In‐TiO₂ might be attributed to the conversion of hydrogen transfer route from E–R to L–H mechanism (Figure 4d). Isotope‐labeled in situ DEMS experiments, together with measurements under Individual NO₃RR conditions, agree well with the above conclusions (Figure S34–S36). In addition, the intensities of in situ DMES signals under individual electrolysis conditions are higher than those observed during co‐electrolysis, with the signals of *NH and *CO showing the most pronounced differences (Figure S37, S38). Suggested by the detected signals of CONH⁺ (m/z = 43) and CONH₂ ⁺ (m/z = 44) in combination with the above results, *NH and *CO act as key intermediates in C‒N coupling to form *CONH, which is subsequently hydrogenated to *CONH₂ rather than undergoing the secondary C‒N coupling step. As reflected in the absence of signals at m/z of 58, 59, 60, 61, 74, 75, or 80 (corresponding to CONO⁺, CONOH⁺, CONHOH⁺, CONH₂OH⁺, CO₂NO⁺/CONO₂ ⁺, CO₂NOH⁺, or CO₂NO₂ ⁺, respectively) the possibility of C‒N coupling at earlier reaction stages can be excluded (Figure S39). The corresponding isotope‐labeled in situ DEMS signals are also collected to validate this.

In situ Raman was further conducted to gain deeper insight into the C‒N coupling mechanism. As shown in Figures 5a, b, the Raman shift located at 152, 402, 520, and 632 cm⁻¹ can be assignable to the E_g, B_1g, A_1g, and E_g active modes for anatase, respectively [38, 45, 46]. Both TiO₂ and In‐TiO₂ exhibit the characteristic bond of Ti‒CO at 378 cm⁻¹, confirming the active site of Ti to generate CO [47]. The peak located at 1003 cm⁻¹ is assigned to the v(C‒N) vibration of urea, which emerges at a lower applied potential and exhibits a higher peak intensity on In‐TiO₂ compared to pristine TiO₂ [48]. This verifies the enhanced C‒N coupling selectivity after the incorporation of In. Another observed peak at 2086 cm⁻¹ corresponds to the C‒O stretching vibration [49]. In situ attenuated total reflection Fourier transform infrared (ATR‐FTIR) spectroscopy was also performed to affirm the formation of C‒N bonds. As displayed in Figure 5c, d, the bands at 3225 and 3360 cm⁻¹ are assign to the stretching vibrational modes of *NH₂ [50]. The bands located at 1636 are attributed to the vibration of N = O [51]. The distinct downward bands at approximately 1240 and 1148 cm⁻¹ are attributed to N‒O antisymmetric stretching vibration of NO₂ ⁻ and NH₄ ⁺, respectively [52, 53]. Upon the introduction of CO₂, two new infrared bands appear at approximately 1299 and 1410 cm⁻¹, which can be assigned to the ‒NH₂ wagging mode of urea and the C‒N stretching vibration, respectively [54, 55]. These observations indicate the formation of C‒N bonds and confirm the existence of urea.

FIGURE 5.

In situ Raman spectra of (a) TiO₂ and (b) In‐TiO₂ at different potentials (−0.7 V_RHE ∼ −1.5 V_RHE) during the co‐reduction of CO₂ and NO₃ ⁻. In situ ATR‐FTIR spectra of In‐TiO₂ during (c) NO₃ ⁻ reduction and d) the co‐reduction of CO₂ and NO₃ ⁻.

Density functional theory (DFT) calculations were employed to elucidate the origin of the enhanced C‒N coupling selectivity on In‐TiO₂. Based on the in situ Raman results, titanium sites are reasonably identified as the active centers. This is further corroborated by the optimized adsorption configurations of key intermediates in Figures S40–S45. All the key intermediates generated from both CO₂RR and NO₃RR procedures prefer to absorb on Ti sites rather than absorb on In sites. As shown in Figure 6a, the lowest energy pathways for individual NO₃RR on TiO₂ and In‐TiO₂ were investigated. The significantly lower ΔG of the PDS (*HNO₃ → *NO₂) on In‐TiO₂ (0.99 eV) compared to that on pristine TiO₂ (2.32 eV) suggests the facilitated hydrogenation of nitrogenous species. A significant thermodynamic divergence occurs after *NO formation. Although the formation of *NO on pristine TiO₂ is exergonic, the excessive adsorption of *NO acts as a thermodynamic sink, inhibiting the subsequent hydrogenation. The hydrogen transfer route of *NO on TiO₂ follows the “NOH pathway” (*NO→ *NOH→ *N→ *NH). The initial hydrogenation steps from *NO to *NOH or *NHO are both endergonic, requiring ΔG of 1.14 eV or 1.29 eV, respectively. In sharp contrast, the evolution of nitrogen‐containing intermediates on In‐TiO₂ is more moderate. The transition from *NO to *NHO is exergonic (ΔG = −0.02 eV), rendering the “NHO pathway” (*NO→ *NHO→ *NHOH→ *NH) more thermodynamically favorable than the “NOH pathway”. Such results further confirm the reversed hydrogenation pathway of the nitrogen‐containing intermediates. As revealed by the temperature‐programmed desorption (TPD) analysis, In‐TiO₂ presents CO₂ and CO desorption peaks at lower temperatures relative to that of pristine TiO₂ (Figure 6b). These results indicate that In doping optimizes the adsorption strength of reactants and alleviates CO poisoning. As depicted in Figure 6c, the PDS for CO₂RR on both catalysts corresponds to the hydrogenation of *CO₂. Compared with pristine TiO₂ (0.93 eV), In‐TiO₂ displays a significantly reduced ΔG (0.44 eV) to facilitate the generation of *CO intermediate. Suggested by the above results, In doping is thermodynamically in favor of promoting the hydrogenation of both carbon‐ and nitrogen‐containing intermediates. However, the decreased NH₃ yield rate is observed on In‐TiO₂ under both co‐reduction and individual NO₃RR conditions (Figures S25 and S32). This implies that the excessive hydrogenation of nitrogen‐containing intermediates is suppressed. The reversed hydrogenation pathway on In‐TiO₂ will balance the utilization of *H, thus enabling the well‐matched C‒N coupling kinetics.

FIGURE 6.

Reaction mechanism. (a) Free energy diagram of NO₃RR on TiO₂ and In‐TiO₂. (b) CO‐ and CO₂‐ TPD of TiO₂ and In‐TiO₂. (c) Free energy diagram of CO₂RR on TiO₂ and In‐TiO₂. (d) Free energy diagram for urea electrosynthesis on TiO₂ and In‐TiO₂. (e) Differential charge density of optimized *CONH adsorption on TiO₂ and In‐TiO₂ (atoms correspond to blue: Ti, red: O, purple: In, brown: C, silver: N, white: H. Yellow stands for accumulation of electrons and cyan stands for depletion of electrons). (f) D‐band center calculated for TiO₂ and In‐TiO₂.

Furthermore, the free energy profiles in Figure 4d suggest the ΔG of 1.12 eV for the initial C‒N bond formation on pristine TiO₂. In sharp contrast, In‐TiO₂ catalyst enables thermodynamically spontaneous C‒N coupling with the ΔG of −0.32 eV. The lower ΔG for subsequent *NHCONH₂ hydrogenation further suggests the effectively decreased overall energy landscape for urea synthesis because of In doping. The enhanced C‒N coupling selectivity can be explained by the stabilized *CONH intermediate, which adopts a more favorable “Ti‒O‒C‒NH‒In” configuration (Figure S43). Analysis of the key *CONH intermediate through charge density difference analysis and Bader charge analysis (Figure 6e,f) reveal greater electron transfer on In‐TiO₂ (0.76 e⁻) than that on pristine TiO₂ (−0.02 e⁻), confirming the promoted electron accumulation in *CONH after In doping [56, 57]. This observation is further supported by the partial densities of states (PDOS) analysis in Figure 4f, where the d‐band center of In‐TiO₂ upshifts further toward the Fermi energy level by 0.29 eV compared to pristine TiO₂ [58]. Such upshift generally results in optimized *CONH adsorption [59], which plays a determinant role in realizing selective C‒N coupling.

2.4. Origin of Reversed Hydrogenation Pathway

As discussed above, In doping not only alters the hydrogenation pathway of nitrogen‐containing intermediates, but also implies a distinct involvement of *H in the overall C‒N coupling process. To further elucidate the origin of active hydrogen on In‐TiO₂, the interfacial water regulation behavior was investigated using in situ Raman spectroscopy (Figure 7a‒c). The broad Raman bands (3000 ∼ 3700 cm⁻¹) assignable to the O‒H stretching mode of interfacial water can be de‐convoluted into three peaks based on the Gaussian fitting principle (Figure 7a, c). The low wavenumber component and the medium wavenumber component are indexed to the four‐coordinated hydrogen‐bonded water (4‐HB‐H₂O) and two‐coordinated hydrogen‐bonded water (2‐HB‐H₂O), respectively [60]. The high wavenumber component is attributed to K⁺ ion hydrated water (K⁺·H₂O) with weak hydrogen‐bond interactions [61]. The lower content of 4‐HB‐H₂O can be observed on In‐TiO₂, while the ratio of 2‐HB‐H₂O and K⁺·H₂O are markedly increased (Figure 7c). The population of water are not sensitive to the applied potentials but the electrode surface. This shift suggests a weakened interfacial H‐bond network, which may promote the accessibility of *H on In‐TiO₂ while inhibiting the long‐range Grotthuss‐type H⁺ transfer [61, 62]. Kinetic isotopic effect (KIE) tests were conducted to further explore the *H generation ability. The KIE values were calculated by comparing the current densities measured in H₂O and D₂O [63]. As shown in Figure S46, In‐TiO₂ delivers lower average KIE value (1.50) compared to that on TiO₂ (2.85), demonstrating the accelerated H₂O dissociation process. Quasi‐in situ electron paramagnetic resonance (EPR) spectroscopy was further employed to monitor the generation of *H, while 5,5‐dimethyl‐1‐pyrroline‐N‐oxide (DMPO) was utilized as the hydrogen radical trapping reagent [64]. Intense nine‐line EPR signals of DMPO‐H (1:1:2:1:2:1:2:1:1) are observed in 0.1 M Na₂SO₄ (Figure 7d). The stronger DMPO‐H signals indicates the promoted *H generation on In‐TiO₂. In addition to the lower ratio of 4‐HB‐H₂O, the enhanced *H supply can also be explained by the decreased ΔG from 0.81 eV on pristine TiO₂ to −0.36 eV on In‐TiO₂ (Figure 7e and S47). As suggested by the regulation of interfacial water and boosted water dissociation to guarantee sufficient *H supply, In‐TiO₂ may allow the hydrogenation of intermediates via the L–H mechanism [65]. Given that In does not directly adsorb nitrogen‐containing species, the reversed hydrogenation pathway can be explained by the increased accessibility of *H stemmed from the regulated interfacial water.

FIGURE 7.

Regulation of interfacial water and identification of active hydrogen. (a) Peak fitting of the interfacial water based on the in situ Raman spectra for TiO₂ and In‐TiO₂. (b) Schematic diagram of in situ Raman techniques. (c) Estimated proportions of 4‐HB‐H₂O, 2‐HB‐H₂O, and K⁺·H₂O from the peak areas for TiO₂ and In‐TiO₂. (d) Quasi in situ EPR spectra of TiO₂ and In‐TiO₂ using DMPO as the trapping reagent. (e) Free energy diagram of HER on TiO₂ and In‐TiO₂.

3. Conclusion

In summary, In‐TiO₂ nanofibers were developed via electrospinning followed calcination for kinetics‐matched urea electrosynthesis. Comprehensive electrochemical measurements, in situ spectroscopic analysis, and theoretical calculations collectively revealed that In doping could regulate the interfacial water structure and promote H₂O dissociation, thereby promoting the accessibility of *H while guaranteeing the sufficient *H supply. Such strategy reversed the hydrogenation pathway of nitrogen‐containing intermediates from the E‐R mechanism (*NO→ *NOH→ *N→ *NH) to the L‐H mechanism (*NO→ *NHO→ *NHOH→ *NH), which enabled the balanced utilization of *H between CO₂ and NO₃ ⁻ reduction. The kinetic mismatch between carbon‐ and nitrogen‐ derived intermediates hydrogenation was effectively relieved for selective C‒N coupling, therefore, as‐developed In‐TiO₂ catalyst achieved an average urea yield rate of 56.5 mmol h⁻¹ g⁻¹ with the FE of 32.8%, and also demonstrated the feasibility of directly utilizing flue gas as feeding gas. This work highlights the vital role of hydrogenation pathway regulation in resolving the kinetic mismatch in electrochemical urea synthesis process, which may provide guidance for the electrochemical synthesis of chemicals relying on the hydrogenation process.

Author Contributions

Liwei Guo: conceptualization, methodology, formal analysis, data curation, validation, visualization, investigation, writing – original draft. Chu Zhang: data curation, formal analysis, validation, investigation. Chunshuang Yan: project administration, funding acquisition, data curation. Qi Long: formal analysis. Shijie Chen: investigation. Chengyun Tang: investigation. Shengji Tian: investigation. Tong Chen: funding acquisition, resources. Zihan Chen: formal analysis. Chunhui Yang: resources. Yumin Qian: software, resources, formal analysis, visualization. Chade Lv: conceptualization, supervision, writing – review and editing, validation, funding acquisition, project administration.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: The authors have cited additional references within the Supporting Information [66, 67, 68, 69, 70, 71, 72, 73].

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.