Ambient Electrosynthesis of Urea with Nitrate and Carbon Dioxide over Iron‐Based Dual‐Sites

Abstract

The development of efficient electrocatalysts to generate key *NH₂ and *CO intermediates is crucial for ambient urea electrosynthesis with nitrate (NO₃ ⁻) and carbon dioxide (CO₂). Here we report a liquid‐phase laser irradiation method to fabricate symbiotic graphitic carbon encapsulated amorphous iron and iron oxide nanoparticles on carbon nanotubes (Fe(a)@C‐Fe₃O₄/CNTs). Fe(a)@C‐Fe₃O₄/CNTs exhibits superior electrocatalytic activity toward urea synthesis using NO₃ ⁻ and CO₂, affording a urea yield of 1341.3±112.6 μg h⁻¹ mg_cat ⁻¹ and a faradic efficiency of 16.5±6.1 % at ambient conditions. Both experimental and theoretical results indicate that the formed Fe(a)@C and Fe₃O₄ on CNTs provide dual active sites for the adsorption and activation of NO₃ ⁻ and CO₂, thus generating key *NH₂ and *CO intermediates with lower energy barriers for urea formation. This work would be helpful for design and development of high‐efficiency dual‐site electrocatalysts for ambient urea synthesis.

Iron‐based electrocatalyst was fabricated by a facile liquid‐phase laser irradiation technique, exhibiting superior electrocatalytic activity toward urea synthesis with CO₂ and NO₃ ⁻. The superior performance is ascribed to the formed Fe(a)@C and Fe₃O₄ on CNTs as the dual active sites for facilitating the generation of *NH₂ and *CO intermediates with lower energy barriers for urea formation.

Introduction

It is well known that more than 40 % of the world‘s food production depends on fertilizer. ^[1] As an important nitrogen source of fertilizers, the global annual output of urea (CO(NH₂)₂) reaches more than 100 million tons. ^[2] Besides of agricultural fertilizer application, urea is also an important feedstock for pharmaceutical and chemical production.[ ³ , ⁴ ] To date, the dominant industrial route to produce urea is the Bosch‐Meiser process via the reaction of carbon dioxide (CO₂) and ammonia (NH₃) under extremely severe conditions (e.g., 150–200 °C, 150–250 bar). ^[5] This process results in high energy consumption and CO₂ emission, consuming about 80 % of the global NH₃ produced by the energy‐intensive Haber‐Bosch process.[ ⁶ , ⁷ , ⁸ ] Therefore, it is imperative to explore energy‐saving and economical routes for urea synthesis under mild conditions.

As an abundant environmental pollutant, nitrate (NO₃ ⁻) can be efficiently treated by the electrochemical reduction method.[ ⁹ , ¹⁰ ] Inspired by this, the electrochemical nitrate reduction under ambient conditions has been considered as a promising alternative to the Haber‐Bosch process for ammonia synthesis.[ ¹¹ , ¹² , ¹³ , ¹⁴ ] Recent studies have demonstrated that urea can be efficiently synthesized by the electrochemical reduction of CO₂ and NO₃ ⁻ at ambient conditions.[ ¹⁵ , ¹⁶ , ¹⁷ ] Amal et al. reported a Cu−N−C single‐atom catalyst can produce urea from NO₃ ⁻ and CO₂ with a faradic efficiency (FE) of 28 %. ^[15] Yu et al. demonstrated that In(OH)₃ with a specific structure can electrochemically couple CO₂ and NO₃ ⁻ to synthesize urea with a yield rate of 533.1 μg h⁻¹ mg_cat ⁻¹ and a FE of 53.4 %. ^[17] Despite some achievements, the electrosynthesis of urea with CO₂ and NO₃ ⁻ under ambient conditions is still far away from practical application, facing with huge challenge. In the process of urea synthesis with CO₂ and NO₃ ⁻, the key step is to generate active *NH₂ and *CO intermediates with lower energy barriers,[ ¹⁸ , ¹⁹ ] which needs an efficient electrocatalyst enabling adsorption and activation of NO₃ ⁻ and CO₂ to accomplish the C−N coupling.[ ²⁰ , ²¹ , ²² , ²³ ]

Iron (Fe) based catalysts are well‐known to be employed in the processes of industrial NH₃ production and natural nitrogen‐fixation. ^[24] Recently, Fe‐based materials, such as FeOOH, ^[25] Fe₃O₄, ^[26] Fe₂O₃,[ ²⁷ , ²⁸ , ²⁹ ] single‐atom Fe,[ ³⁰ , ³¹ ] and metallic Fe nanoparticles, ^[32] have been developed for ammonia synthesis and CO₂ reduction at ambient conditions,[ ³³ , ³⁴ ] exhibiting high electrocatalytic activities. According to these reported studies, it can be imagined that Fe‐based catalysts with tunable active components could be promising for electrocatalytic synthesis of urea with CO₂ and NO₃ ⁻. Specially, the introduction of defects into the crystal structure has been proven to be an effective strategy for regulating the electronic structure of catalyst, thus improving the electrocatalytic activity. ^[35] Arrestingly, amorphous material is solid formed by disordered aggregation of mass points, and considered to be the most sterically defective form, often exhibiting unexpected catalytic performance.[ ³⁶ , ³⁷ , ³⁸ ] These strategies could be utilized to design and develop Fe‐based catalysts with multiple active sites for effective adsorption and activation of CO₂ and NO₃ ⁻, thus generating key *NH₂ and *CO active intermediates with lower energy barriers for high‐efficiency electrosynthesis of urea at ambient conditions.

Herein, we report a liquid‐phase laser irradiation route to fabricate symbiotic carbon encapsulated amorphous iron (Fe(a)@C) and iron oxide nanoparticles (Fe₃O₄ NPs) on carbon nanotubes (denoted as Fe(a)@C‐Fe₃O₄/CNTs). The as‐fabricated Fe(a)@C‐Fe₃O₄/CNTs contains two Fe‐based active components, namely, Fe@C NPs with the particle sizes of 10–20 nm and Fe₃O₄ NPs with the particle sizes of 1–5 nm. The presence of two different structural units in Fe(a)@C‐Fe₃O₄/CNTs makes it possible to synergistically electrocatalytic activate CO₂ and NO₃ ⁻ to realize the C−N coupling for urea synthesis. As expected, Fe(a)@C‐Fe₃O₄/CNTs exhibits superior activity toward the electrocatalytic coupling of CO₂ and NO₃ ⁻ for urea synthesis, affording a urea yield of 1341.3±112.6 μg h⁻¹ mg_cat ⁻¹ and a faradic efficiency (FE) of 16.5±6.1 % at −0.65 V (vs. RHE) in 0.1 M KNO₃ electrolyte. Both experimental and theoretical results unveil that Fe(a)@C is mainly responsible for the electrocatalytic reduction of NO₃ ⁻ to form *NH₂ intermediates, while Fe₃O₄ is more beneficial for the electrocatalytic reduction of CO₂ to form *CO intermediates. The synergistically catalytic effect contributes the excellent electrocatalytic performance of urea synthesis at ambient conditions.

Results and Discussion

Figure 1a shows the fabrication process of the catalyst. Typically, carbon nanotubes (CNTs) and FeCl₂ were firstly dispersed in isopropanol, followed by stirring for 12 h to adequately adsorb Fe²⁺, and then irradiating with nanosecond laser for 1 h to obtain graphitic carbon encapsulated amorphous Fe and Fe₃O₄ nanoparticles (NPs) on CNTs (see Experimental Section for details). Figure 1b, c show low‐ and high‐magnification TEM images of the sample, displaying obvious nanotubes structures with various sized Fe‐based NPs loaded on the surface of CNTs. These larger NPs possess the sizes ranging from 10–20 nm (Figure 1c), and high‐resolution TEM (HRTEM) image indicates that the NPs show the specific structures of graphitic carbon encapsulated amorphous Fe (Figure 1d, Figure S1). Besides of these large‐sized Fe(a)@C NPs, ultrafine NPs can be also observed on the surface of CNTs. These ultrafine NPs show particle sizes ranging from 1–5 nm, exhibiting the interplanar spacings of 0.250 and 0.294 nm, assignable to (113) and (022) planes of cubic phase Fe₃O₄, respectively (Figure 1e, Figure S2).

Figure 1.

a) Schematic illustration of the synthetic process of Fe(a)@C‐Fe₃O₄/CNTs. b) Low‐ and c) high‐magnification TEM images of Fe(a)@C‐Fe₃O₄/CNTs. HRTEM images of Fe(a)@C (d) (from the red dashed box in c) and Fe₃O₄ NPs (e) (from the yellow dashed box in c). f), g) HAADF‐STEM images and corresponding elemental mapping images of Fe₃O₄ and Fe(a)@C.

Simultaneously obtained two Fe‐based components on CNTs can be ascribed to high‐energy nanosecond laser irradiation of CNTs with adsorbed Fe²⁺ under the help of isopropanol, through partial oxidation and carbon thermal reduction of Fe²⁺ to form Fe₃O₄ and Fe(a)@C, respectively. To further confirm the components, we selected different regions of the sample to obtain the element mapping images. The high‐angle annular dark‐field (HAADF) image in Figure 1f shows merely ultrafine Fe₃O₄ NPs, displaying uniform distribution of Fe, O and C elements in this region, while the HAADF image (Figure 1g) of large‐sized Fe(a)@C NPs indicates that the iron signal is very prominent but the oxygen signal is not obvious at the iron position, implying the formation of Fe(a)@C. The above characterizations indicate that large‐sized Fe(a)@C and ultrafine Fe₃O₄ NPs have been concurrently formed on CNTs. These two Fe‐based components in Fe(a)@C‐Fe₃O₄/CNTs could be advantageous for adsorption and activation of CO₂ and NO₃ ⁻, thus high electrocatalytic activity toward urea synthesis. For comparison, we also employed the impregnation‐pyrolysis method to fabricate Fe‐based sample, only cubic‐phase Fe₃O₄ and FeO NPs can be obtained on CNTs (Figure S3), different to the laser irradiation fabricated sample. In addition, through changing the reaction medium composition during laser irradiation, we also obtained the samples with different molar ratios of Fe₃O₄ to Fe(a)@C, such as Fe(a)@C‐Fe₃O₄/CNTs‐V2 (V_isopropanol/V_water=2 : 1), Fe(a)@C‐Fe₃O₄/CNTs‐V1 (V_isopropanol/V_water=1 : 2), and Fe₃O₄/CNTs‐W (Figure S4), to investigate the Fe‐based component influence on the urea synthesis. The ICP results show that the molar ratio of Fe₃O₄ to amorphous Fe is 0.43 : 1 for Fe(a)@C‐Fe₃O₄/CNTs, 0.47 : 1 for Fe(a)@C‐Fe₃O₄/CNTs‐V2, and 0.71 : 1 for Fe(a)@C‐Fe₃O₄/CNTs‐V1, indicating that the content of Fe₃O₄ is increased in catalyst with increasing water content during laser irradiation, consistent with the TEM results (Figure S4).

Figure S5 shows the X‐ray diffraction (XRD) patterns of the as‐synthesized Fe@C‐Fe₃O₄/CNTs. As shown, the diffraction peak at around 26.4° is attributed to the (002) plane of graphitic carbon, while other diffraction peaks correspond well to cubic phase Fe₃O₄ (Card No. 01‐076‐1849). However, the Fe₃O₄ NPs on CNTs show poor crystalline structure, mainly owing to the rapid nucleation process of liquid‐phase laser irradiation treatment. ^[39] Furthermore, the elemental composition and chemical state of the sample were determined by the X‐ray photoelectron spectroscopy (XPS) technique. The surface survey XPS spectrum (Figure 2a) reveals the peaks of Fe, O and C elements, consistent with the element mapping results. In Figure 2b, the C 1s XPS spectrum can be divided into two peaks at 284.8 and 285.9 eV, assigned to C=C and C−O, respectively. ^[40] The O 1s XPS spectrum (Figure 2c) can be divided into two peaks of C−O (532.6 eV) and Fe−O (530.5 eV). ^[41] Figure 2d shows the Fe 2p XPS spectrum, and it can be fitted by ten peaks. The peaks located at 733.5, 728.3, 725.7and 723.9 eV are corresponding to Fe 2P_3/2 orbitals of Fe³⁺ (satellite), Fe²⁺ (satellite), Fe³⁺ and Fe²⁺, respectively. ^[42] The peaks at 718.8, 714.7, 712.8 and 711.0 eV is attributed to Fe 2P_1/2 orbitals of Fe³⁺ (satellite), Fe²⁺ (satellite), Fe³⁺ and Fe²⁺, respectively. ^[43] The peaks at 720.5 and 707.9 eV are corresponding to Fe⁰. ^[44] The XPS results combined with the HRTEM and XRD measurements further confirm the co‐existence of Fe(a)@C and Fe₃O₄ NPs on CNTs. Figure S6 shows the Raman spectra of pristine carbon nanotubes (P‐CNTs), carbon nanotubes after laser irradiation treatment (L‐CNTs), and Fe(a)@C‐Fe₃O₄/CNTs. The results indicate that the laser irradiation can induce carbon graphitization, and the introduction of Fe further promotes the process,[ ⁴⁵ , ⁴⁶ ] favourable for electrocatalysis. The BET surface area of Fe(a)@C‐Fe₃O₄/CNTs is 70.63 m⁻² g⁻¹ with dominant mesoporous and microporous structures, beneficial for the exposure of catalytic active sites and mass transport during electrocatalysis (Figure S7).

Figure 2.

a) XPS survey spectrum, b) C 1s, c) O 1s and d) Fe 2p XPS spectra of Fe(a)@C‐Fe₃O₄/CNTs.

The electrochemical measurements of Fe(a)@C‐Fe₃O₄/CNTs were performed for evaluating the urea synthesis performance. In all experiments, 0.1 M KNO₃ solution was used as the electrolyte (see Experimental Section for details). Figure 3a shows the linear sweep voltammetry (LSV) curves of Fe(a)@C‐Fe₃O₄/CNTs in Ar‐ and CO₂‐saturated electrolytes. Obviously, when CO₂ is introduced to the system, Fe(a)@C‐Fe₃O₄/CNTs exhibits larger current densities at the given potentials from −0.4 to −0.9 V (vs. RHE) compared to those in Ar‐saturated electrolyte, possibly suggesting superior electrocatalytic activity toward urea synthesis with CO₂ and NO₃ ⁻. Subsequently, the potentiostatic electrolysis was performed in CO₂‐saturated 0.1 M KNO₃ at different potentials from −0.45 to −0.75 V (vs. RHE). The chronoamperometry curves are shown in Figure 3b. It is noted that the produced urea in this work is analyzed by the diacetyl monoxime colorimetric method (Figure S8). ^[23] After 2 h of electrocatalysis, the UV/Vis spectra of the electrolytes obtained at different potentials are shown in Figure S9, exhibiting the strongest absorbance peak at −0.65 V (vs. RHE). Figure 3c shows the dependence of urea yield and corresponding faradic efficiency (FE) at different potentials. It can be seen that with increasing applied potential, the urea yield and FE are increased, reaching the largest yield of 1341.3±112.6 μg h⁻¹ mg_cat ⁻¹ with a FE of 16.5±6.1 % at −0.65 V (vs. RHE). Both the urea yield and FE obtained by Fe(a)@C‐Fe₃O₄/CNTs are comparable to the recently reported urea electrosynthesis catalysts (Table S1). When the potential is beyond −0.65 V (vs. RHE), the urea yield and FE are decreased, primarily due to the possibly competitive hydrogen evolution reaction (HER), CO₂ reduction reaction (CO₂RR), and nitrate⁻ reduction reaction (NtrRR). ^[15] In this work, all possible products during electrocatalytic urea synthesis were detected (Figure S10–12). The results demonstrate that H₂, CO, NH₃, NO₂ ⁻ and urea are detectable at all investigated potentials (Figure S12). Moreover, at more negative potential (e.g., −0.75 V, vs. RHE), the FEs for H₂ and NH₃ are obviously increased, which result in the decreased urea synthesis efficiency, consistent with the aforementioned results. The stability test was subsequently conducted in CO₂‐saturated 0.1 M KNO₃ for 10 h of electrocatalysis. As shown in Figure 3d, the change of the chronoamperometry curve is slight, indicating good stability of the catalyst. This can be further confirmed by the electron microscopy images, XRD and XPS results of Fe(a)@C‐Fe₃O₄/CNTs after 10 h of electrocatalysis (Figure S13–15). The cycling tests were carried out at −0.65 V (vs. RHE) for 5 cycles with 2 h of electrocatalysis for each cycle, displaying good repeatability of Fe(a)@C‐Fe₃O₄/CNTs (Figures S16).

Figure 3.

a) LSV curves of Fe(a)@C‐Fe₃O₄/CNTs measured in Ar‐ and CO₂‐saturated 0.1 M KNO₃ electrolytes. b) The chronoamperometry curves of Fe(a)@C‐Fe₃O₄/CNTs measured in CO₂‐saturated 0.1 M KNO₃ with different potentials. c) The dependence of urea yield and FE of Fe@C‐Fe₃O₄/CNTs on potential. d) The stability test of Fe(a)@C‐Fe₃O₄/CNTs for urea synthesis.

In order to confirm the source of urea produced in this work, we subsequently conducted several control experiments as follows: i) Open‐circuit potential was employed to Fe(a)@C‐Fe₃O₄/CNTs in CO₂‐saturated 0.1 M KNO₃ for 2 h of electrocatalysis; ii) Fe(a)@C‐Fe₃O₄/CNTs was employed in CO₂‐saturated 0.1 M K₂SO₄ for 2 h of electrocatalysis at −0.65 V (vs. RHE); iii) Carbon paper was employed as the working electrode at −0.65 V (vs. RHE) in CO₂‐saturated 0.1 M KNO₃ for 2 h of electrocatalysis; iv) Fe(a)@C‐Fe₃O₄/CNTs was employed in Ar‐saturated 0.1 M KNO₃ for 2 h of electrocatalysis at −0.65 V (vs. RHE). For all cases, no urea product can be detected (Figure S17), demonstrating that the urea produced in this work is indeed from the Fe@C‐Fe₃O₄/CNTs catalyzed coupling reaction of CO₂ and NO₃ ⁻. This conclusion can be further validated by the isotope‐labelling experiments, which were conducted in CO₂‐saturated 0.1 M K¹⁵NO₃ and K¹⁴NO₃ electrolytes. The ¹H NMR spectra identified the formation of ¹⁵NH₂CO¹⁵NH₂ and ¹⁴NH₂CO¹⁴NH₂ products according to the chemical shift of doublet coupling and single coupling, respectively. ^[23] As shown in Figure 4a, the ¹H NMR spectrum of the electrolyte with K¹⁵NO₃ shows typical two peaks of ¹⁵NH₂CO¹⁵NH₂ at 5.50 and 5.73 ppm, while the ¹H NMR spectrum of the electrolyte with K¹⁴NO₃ only exhibits one typical peak of ¹⁴NH₂CO¹⁴NH₂ at 5.62 ppm. Furthermore, after feeding with ¹³CO₂ gas, the ¹³C NMR spectrum shows one typical peak for ¹³C‐labelled urea (Figure 4b). The ¹H NMR results confirmed that the produced urea is originated from the electrocatalytic coupling reaction of CO₂ and NO₃ ⁻ on Fe@C‐Fe₃O₄/CNTs. In this work, the by‐products of NH₃ and NO₂ ⁻ during electrocatalytic urea synthesis are detectable, which may affect the accuracy of urea determination using the diacetyl monoxime method. ^[47] For this, we also employed the ¹H NMR to quantitatively measure the produced urea and simultaneously verify the reliability of the diacetyl monoxime method. The results demonstrate that the ¹H NMR measured ¹⁵NH₂CO¹⁵NH₂ concentration is essentially consistent with that detected by the colorimetric method, confirming the accuracy of the used quantitative method (Figure S18).

Figure 4.

a) ¹H NMR spectra obtained by using K¹⁵NO₃ and K¹⁴NO₃ as the reactants with CO₂ and standard ¹⁵N‐labelled‐urea, ¹⁴N‐labelled‐urea samples. b) ¹³C NMR spectra of 0.1 M KNO₃ electrolyte at −0.65 V (vs. RHE) for 2 h with feeding ¹³CO₂ and standard ¹³C‐labelled urea. In situ FTIR spectroscopy measurements under various potentials for Fe(a)@C‐Fe₃O₄/CNTs during electrocatalytic coupling of nitrate and carbon dioxide. c) Three‐dimensional in situ FTIR spectra in the range of 1000–3600 cm⁻¹. d) Infrared signals in the range of 1100–1800 cm⁻¹.

To visually verify the C−N coupling mechanism, the in situ Fourier transform infrared (FTIR) measurements (Figure S19) were performed on Fe(a)@C‐Fe₃O₄/CNTs with the signals collected from 1000 to 3600 cm⁻¹, and the applied potential was set at open‐circuit, −0.45, −0.55, −0.65 and −0.75 V (vs. RHE), respectively (Figure 4c). For the given potentials, from 1100 to 1800 cm⁻¹ (Figure 4d), the bending and rocking frequencies of −NH₂ in urea were found at 1637 and 1172 cm⁻¹,respectively. ^[17] There is also a wagging mode of −NH₂ at 1304 cm⁻¹. ^[17] In addition, a stretching peak appears at 1665 cm⁻¹, corresponding to C=O band. The band at 1455 cm⁻¹ is attributed to the C−N stretching of free urea.[ ⁴⁸ , ⁴⁹ ] Not only that, the stretching mode of C−N can also be observed at 1415 cm⁻¹, indicating that a part of urea could be coordinated with Fe²⁺ through oxygen atom on C=O group and N atom on N−H group to form Fe²⁺‐urea complex. ^[50] The in situ FTIR measurements unambiguously demonstrate that the C−N coupling is successfully achieved through the Fe(a)@C‐Fe₃O₄/CNTs catalyzed coupling reaction of CO₂ and NO₃ ⁻. In addition, the in situ Raman measurements also reveal the formation of C−N stretching mode (symmetric A₁) appeared at 1000 cm⁻¹, ^[51] indicating that C−N coupling is indeed occurred (Figure S20, 21).

In this work, we also evaluated the urea synthesis performance of the electrocatalysts with different molar ratios of Fe₃O₄ to Fe(a)@C. As shown in Figure S22, the Fe(a)@C‐Fe₃O₄/CNTs exhibits more superior electrocatalytic activity toward urea synthesis than Fe(a)@C‐Fe₃O₄/CNTs‐V2 and Fe(a)@C‐Fe₃O₄/CNTs‐V1, indicating that a suitable ratio of Fe₃O₄ to Fe(a)@C in catalyst is very important for its high urea synthesis performance. To further understand the role of Fe‐based components and the catalytic mechanism of Fe(a)@C‐Fe₃O₄/CNTs for C−N coupling with CO₂ and NO₃ ⁻, we also performed a series of comparative experiments using Fe(a)@C‐Fe₃O₄/CNTs catalysts after acid etching (Fe(a)@C‐Fe₃O₄/CNTs‐AE), calcination at 300 °C in N₂ (Fe(a)@C‐Fe₃O₄/CNTs‐CN), and calcination at 300 °C in air (Fe(a)@C‐Fe₃O₄/CNTs‐CA). As shown in Figure S23a–c, after acid etching, ultrafine Fe₃O₄ NPs on CNTs are almost completely disappeared, while Fe(a)@C NPs are well maintained (Figure S23d), owing to the coating role of carbon layer. The elemental mapping images (Figure S23e) exhibit a strong Fe signal but almost ignorable O signal at Fe site. The high‐resolution Fe 2p XPS spectra (Figure S24) show that Fe⁰ is the dominant form of iron in Fe(a)@C‐Fe₃O₄/CNTs‐AE, and the oxidized iron components are very little, meaning almost complete removal of Fe₃O₄ nanoparticles on CNTs. The XRD patterns show the diffraction peaks of only graphitic carbon without any iron oxides in Fe(a)@C‐Fe₃O₄/CNTs‐AE (Figure S25), supportable for the above results. To investigate the influence of the catalyst's crystalline degree on urea synthesis, Fe(a)@C‐Fe₃O₄/CNTs‐CN was fabricated, exhibiting nanoparticles morphology with different sizes on CNTs (Figure S26a). After thermal treatment in N₂, the carbon coated amorphous Fe NPs also exhibit the crystalline structure at a certain extent with a lattice spacing of 0.203 nm, corresponding to the (011) plane of Fe (Figure S26b, c). The elemental mapping and XPS results further confirm the co‐existence of crystalline structure Fe and Fe₃O₄ (Figure S26d, Figure S27). For the calcined sample in air (Fe(a)@C‐Fe₃O₄/CNTs‐CA), amorphous Fe⁰ is almost disappeared, and transformed to iron oxides (Figure S28, 29).

Subsequently, the electrocatalytic measurements were performed using Fe(a)@C‐Fe₃O₄/CNTs‐AE, Fe(a)@C‐Fe₃O₄/CNTs‐CN, and Fe(a)@C‐Fe₃O₄/CNTs‐CA for evaluating their urea synthesis performance with CO₂ and NO₃ ⁻. The chronoamperometric measurements of all investigated electrocatalysts were conducted at −0.65 V (vs. RHE) in CO₂‐saturated 0.1 M KNO₃ (Figure S30). The UV/Vis spectra of the electrolytes after electrocatalysis are presented in Figure S31. Figure S32 shows the urea yield and FE of the investigated electrocatalysts. Obviously, Fe(a)@C‐Fe₃O₄/CNTs‐CN and Fe(a)@C‐Fe₃O₄/CNTs‐CA exhibit lower urea synthesis performance compared to Fe(a)@C‐Fe₃O₄/CNTs‐AE, while the urea yield and FE of Fe(a)@C‐Fe₃O₄/CNTs‐AE are only about 47 % and 32 % of those obtained from Fe(a)@C‐Fe₃O₄/CNTs. The results indicate that the increase in the crystallinity of amorphous Fe and only iron oxides in catalysts lead to obviously decreased urea synthesis performance, and the significance of the synergistic effect of Fe‐based dual‐sites on high‐efficiency urea synthesis. The NtrRR and CO₂RR measurements of Fe(a)@C‐Fe₃O₄/CNTs‐CA and Fe(a)@C‐Fe₃O₄/CNTs‐AE demonstrate that Fe(a)@C sites are more beneficial for NtrRR, while Fe₃O₄ sites favors for CO₂RR (Figure S33–35). Their synergistic effect contributes high urea synthesis performance.

The above comparable experiments have demonstrated that Fe(a)@C and Fe₃O₄ in the catalyst are critically important for high‐efficiency electrocatalytic urea synthesis with CO₂ and NO₃ ⁻. The density functional theory (DFT) calculations were subsequently utilized to unveil the catalytic reaction mechanisms of urea synthesis. According to the structure characterizations, we established C+Fe₃O₄, carbon‐coated crystal Fe(011) (Fe(c)@C) and carbon‐coated amorphous Fe (Fe(a)@C) surface structure models (Figure 5a). The detailed setting conditions are given in the experimental section. Firstly, both CO₂ and NO₃ ⁻ adsorption behavior was investigated on the surface of the above models. The structural evolution of the active intermediates at the site is shown in Figure S36–41. For CO₂ adsorption, Fe(a)@C exhibits more negative adsorption free energy of −0.974 eV (ΔG _*CO2) compared with Fe(c)@C (−0.364 eV) and C+Fe₃O₄ (−0.447 eV) (Table S2). While NO₃ ⁻ adsorption free energy (ΔG _*NO3) for Fe(a)@C, Fe(c)@C and C+Fe₃O₄ is −0.751, −0.37, and −0.617 eV, respectively (Table S3). Compared to Fe(c)@C, Fe(a)@C is more favourable for the adsorption of CO₂ and NO₃ ⁻. This is because, according to the calculated partial density of state (PDOS) results of the carbon on Fe surface, p orbital of C on Fe(a)@C at −2–0 eV (the top of the valence band) was higher than that on Fe(c)@C, indicating that after Fe amorphization, the activity of graphitic C is enhanced (Figure S42a). This was further proved by the charge density difference analysis. As shown in Figure S42b, c, the C atom in Fe(a)@C gained obviously electrons from Fe, which improved its chemical activity.

Figure 5.

a) Structures of C+Fe₃O₄, Fe(c)@C and Fe(a)@C. b) Free energy profiles of CO₂ reduction to *CO on C+Fe₃O₄, Fe(c)@C and Fe(a)@C. c) Free energy profiles of NO₃ ⁻ reduction to *NH₂ on C+Fe₃O₄, Fe(c)@C and Fe(a)@C. d) Free energy profile of C−N coupling on C+Fe₃O₄.

After unveiling the adsorption behavior of NO₃ ⁻ and CO₂, the feasibility of *CO₂ reduction to *CO and *NO₃ reduction to *NH₂ was next disclosed on these three models. The *CO₂ electroreduction to *COOH intermediate through a proton‐coupled electron transfer (PCET) step, which is endothermic by 0.028 eV for Fe(a)@C, but exothermic by −0.139 and −0.555 eV for Fe(c)@C and C+Fe₃O₄, respectively (Figure 5b). Compared with Fe(a)@C, Fe(c)@C and C+Fe₃O₄ models were more conducive to *CO₂ reduction. The *COOH transformed to the *CO through the second PCET step, which was endothermic with a reaction free energy (ΔG) of 0.971, 0.69 eV for Fe(c)@C and Fe(a)@C. But for C+Fe₃O₄, it was a spontaneous step with a downhill energy step and the ΔG was −0.757 eV. Thus, direct conversion of *CO₂ to *CO on C+Fe₃O₄ surface was energetically favorable than other models. Another important step in the production of urea is the formation of *NH₂, which reflects the ease of NO₃ ⁻ reduction. Figure 5c shows the free energy profiles of NO₃ ⁻ reduction to *NH_2, from which it can be found that all the reduction steps of NO₃ ⁻ to *NH₂ for these three models were spontaneous steps. The Fe(a)@C site has the lowest energy, which is more favorable for the generation of *NH₂. Collectively considering, in the process of urea synthesis, C+Fe₃O₄ was more beneficial for CO₂RR process, and Fe(a)@C facilitates NtrRR process, consistent with the experimental results. For urea formation, the most important intermediate is *CONH₂, which was produced by coupling of key *CO and *NH₂ intermediates. ^[52] In this case, the coupling of *NH₂ and *CO should involve the desorption of active intermediates and the selection of coupling sites. The calculations (Table S4) revealed that the desorption energy of *NH₂ (−5.699 eV) is much lower than that of *CO (2.567 eV), indicating that *NH₂ is easily desorbed from Fe(a)@C and then dissociated to C+Fe₃O₄ sites to couple with *CO, thus forming *CONH₂ intermediates. The coupling process (Figure 5d, Figure S43) showed that the transition state (TS) energy barrier is about 0.79 eV, followed by an exothermic process to *CONH₂, which is a spontaneous step with a downhill energy step with the ΔG of −2.63 eV.

Conclusion

In summary, Fe(a)@C‐Fe₃O₄/CNTs with Fe(a)@C and Fe₃O₄ NPs as cooperative catalytic sites was synthesized through a liquid‐phase laser irradiation technique. The Fe(a)@C‐Fe₃O₄/CNTs exhibited outstanding activity toward electrocatalytic coupling of CO₂ and NO₃ ⁻ to produce urea with a urea yield of 1341.3±112.6 μg h⁻¹ mg_cat ⁻¹. The in situ spectroscopy and ¹H NMR measurements confirmed that the yielded urea was originated from the coupling reaction of CO₂ and NO₃ ⁻. The theoretical calculations unveiled the significant role of Fe(a)@C and Fe₃O₄ dual‐sites on adsorption and activation of NO₃ ⁻ and CO₂, thus generating key *NH₂ and *CO intermediates for urea formation.

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Geng J., Ji S., Jin M., Zhang C., Xu M., Wang G., Liang C., Zhang H., Angew. Chem. Int. Ed. 2023, 62, e202210958; Angew. Chem. 2023, 135, e202210958.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.