Oxygen vacancies-rich molybdenum tungsten oxide nanowires as a highly active nitrogen fixation electrocatalyst
Jincheng Zhang , Mengjie Sun , Jiali Ren , Rui Zhang , Min Ma , Qingzhong Xue , Jian Tian
Ammonia (NH3), as an industrial raw material can produce explosives, plastics, synthetic fibers, resins, and many other industrial compounds [1-3]. NH3 can also be used as a promising carrier for carbon-free energy storage and conversion [4]. However, the inherent properties of strong N≡N bonds make N2 extremely difficult to synthesize NH3 from the hydrogenation of N2 in the atmosphere [5-8], So far, industrial NH3 production still relies mainly on the traditional Haber-Bosch method. However, this process needs a large amount of energy and produces a large amount of carbon dioxide [9-11]. Therefore, developing green and sustainable routes to achieve efficient fixation of N2 under milder conditions is urgent.
Electrochemical nitrogen reduction reaction (eNRR) is carried out using renewable electricity, which is considered a promising option for artificially fixing N2 [12,13]. However, NRR must be driven by an effective electrocatalyst. Although the noble metal-based catalysts (Ru, Pd, Ag Au [14-17], etc.) exhibit good NRR activity, their low abundance and high cost hinder large-scale application [18,19]. Therefore, researchers have given considerable attention to Earth-abundant alternatives.
Transition metal oxides (TMOs) are considered important NRR catalysts due to their structural adjustability, rich redox properties, and earth abundance [20,21]. Therefore, developing high-performance TMOs catalysts through specific adjustments has great potential. Among these TMOs, tungsten trioxide (WO3) has attracted extensive research interest due to its wide range of applications, high electrochemical stability, and abundant supply [22]. In addition, Density functional theory (DFT) calculation has also shown that W has the potential to activate the nitrogen-nitrogen triple bond, which can lead to the nitrogen reduction reaction [23]. However, due to the lack of active sites and limited electrical conductivity of WO3, its NRR catalytic effect is poor [24]. As an important element in biological nitrogenase, Mo is used to catalytically fix N2 in biological systems [25]. Studies on some Mo-based catalysts (MoS2 [26], MoO3 [25], Mo2N [27], and even Mo mono-atomic catalysts [28]) have shown that Mo is the active center of immobilized N2. It helps stabilize -N2Hy intermediates (-N2H, -N2H2, etc.). and desorb -NH2 species during NRR [29,30]. Therefore, by adding MoO3, introducing Mo elements and adjusting the local atomic structure to generate more active sites, the electrochemical NRR activity is enhanced. In addition, oxygen vacancies (OVs), constructed on TMOs, can act as electron trapping centers for π-backdonation of N2 molecules, improving the electrochemical activity and electrical conductivity, promoting energetics and kinetics of NRR reaction [31].
Herein, we report the synthesis of artificial NH3 under environmental conditions using 1D OVs-rich molybdenum-tungsten bimetallic oxide nanowires (CTAB-D-W4MoO3 NWs) as active and selective electrocatalysts. CTAB-D-W4MoO3 NWs realize a large NH3 yield of 60.77 µg h-1 mg-1cat. at −0.70 V vs. RHE and a high faradaic efficiency (FE) of 56.42% at −0.60 V, much superior to molybdenum-tungsten bimetallic oxide nanowires deficient in OVs (CTAB-W4MoO3 NWs: 20.26 µg h-1 mg-1cat. and 17.1%). In addition, CTAB-D-W4MoO3 NWs exhibit excellent cyclic stability and long-term stability.
W4MoO3 nanowires rich in O-vacancies (CTAB-D-W4MoO3 NWs) were obtained using ammonium molybdate tetrahydrate as Mo source and sodium tungstate dihydrate as W source. Fig. 1a shows the synthesis strategy of CTAB-D-W4MoO3 NWs. Firstly, pure W4MoO3 NWs were prepared, and the surfactant cetyltrimethylammonium bromide was added to avoid product aggregation. Then, W4MoO3 NWs with abundant O vacancies were acquired via annealing under an H2 atmosphere. The local atomic structure is changed and abundant oxygen defects are formed in W4MoO3 owing to the reducibility of H2. These defect sites provide more active sites for N2 adsorption, and enhance the electron transport rate of the reaction. The improvement of catalyst nitrogen fixation performance is precisely owing to the large specific surface area and the catalyst defects. In this article, the catalyst samples before and after annealing are named CTAB-W4MoO3 and CTAB-D-W4MoO3 [24].
The crystal structures of CTAB-D-W4MoO3 and CTAB-W4MoO3 were tested by X-ray diffraction (XRD) (Fig. 1b). The precursor (CTAB-W4MoO3) is mainly composed of MoO3 (PDF #47–1320) and WO3 (PDF #33–1387). After annealing, CTAB-D-W4MoO3 is mainly composed of MoO2 (PDF #32–0671) and WO2 (PDF #32–1393). Although masked by stronger dioxide peaks, peaks of MoO3 and WO3 are still present, suggesting the formation of oxygen vacancies following calcination. In contrast, the XRD pattern of CTAB-H-W4MoO3 is similar to that of CTAB-W4MoO3 except for the difference in peak intensity, which further confirms the existence of oxygen vacancies in the catalyst (Fig. S1 in Supporting information). For Raman spectra (Fig. 1c), it can be seen that the characteristic peaks of CTAB-W4MoO3 and CTAB-D-W4MoO3 are similar. The bands at 712 and 821 cm-1 correspond to the stretching mode of M-O-M (M represents metal). The bands at 326 and 263 cm-1 are ascribed to the O-M-O bending mode, suggesting the presence of metal oxides in the catalysts [32]. Yet, the relatively wide peak of CTAB-D-W4MoO3 indicates the possibility of O vacancies in W4MoO3 [24].
The survey X-ray photoelectron spectroscopy (XPS) spectrum of CTAB-D-W4MoO3 (Fig. S2 in Supporting information) shows the presence of W, Mo, and O elements in the sample. Fig. S3a (Supporting information) shows the W 4f spectrum of CTAB-D-W4MoO3 and CTAB-W4MoO3 nanowires. CTAB-D-W4MoO3 not only has peaks of W6+ (35.80 and 38.12 eV, 40.57 eV-satellite peak), but also peaks of W5+ (35.17 and 37.31 eV) and W4+ (32.76 and 34.88 eV) [33]. This is because W6+ undergoes the reduction of oxide during heat treatment in H2 atmosphere, forming O vacancies in the catalysts [34]. Similarly, in the Mo 3d spectrum (Fig. S3b in Supporting information), the peaks at 231.93 and 235.37 eV are ascribed to Mo6+ 3d5/2 and 3d3/2, respectively, while the peaks at 229.15 and 232.59 eV correspond to Mo4+ 3d5/2 and 3d3/2, respectively [35]. The O 1s spectrum in Fig. S3c (Supporting information) further indicates the formation of oxygen vacancies. The main peak at 530.28 eV is attributed to the oxygen metallic bond of O-W or O-Mo [36]. Importantly, another peak is clearly observed at 531.82 eV, corresponding to oxygen vacancies [36]. Furthermore, the peak at 533.26 eV is ascribed to the C—O bond [24]. As shown in Fig. S3d (Supporting information), the electron paramagnetic resonance (EPR) signal intensity of CTAB-D-W4MoO3 NWs is significantly higher than CTAB-W4MoO3 NWs, suggesting the presence of vacancies in CTAB-D-W4MoO3 NWs. For CTAB-D-W4MoO3 NWs, the noticeable signal with a g value of 2.002 is caused by oxygen vacancies. Combined with XPS analysis, the existence of oxygen vacancies in the prepared CTAB-D-W4MoO3 NWs can be confirmed.
As shown in Fig. S4 (Supporting information), the precursor of the hydrothermal product CTAB-W4MoO3 presents the nanowire morphology. After annealing, CTAB-D-W4MoO3 preserves the nanowire structure (Fig. 2a). The transmission electron microscope (TEM) image further confirms this observation (Fig. S5 in Supporting information), and the diameter of the nanowire is about 20 nm (Fig. 2b). In addition, the 0.18 nm lattice fringe corresponds to the (101) crystal plane of WO3-x or MoO3-x (Fig. 2c). Moreover, the energy dispersive spectroscopy (EDS) elemental mapping images show that Mo, W, and O are uniformly distributed across the nanowires (Fig. 2d). The contents of Mo and W in catalyst were measured by inductively coupled plasma mass spectrometry (ICP-MS) (Table S1 in Supporting information).
The NRR activity of CTAB-D-W4MoO3 was further evaluated in an H-type electrolytic cell containing 0.1 mol/L Na2SO4. Considering that there may be a few NOx impurities in the electrolyte, we first detect NO3- and NO2- in the electrolyte before and after introducing N2 for 30 min. The corresponding UV visible spectrum is shown in Fig. S6 (Supporting information). The presence of NO3- and NO2- is not found, suggesting that the experimental results are not affected by impurity contamination. Watt/Chrisp and indophenols blue method were used to gage the amount of possible by-product N2H4 and produced NH3, and corresponding calibration curves were shown in Fig. S7 (Supporting information). Firstly, CTAB-D-W4MoO3 samples were tested in N2 and Ar saturated 0.1 mol/L Na2SO4. From Fig. S8 (Supporting information), it can be seen that the current density of CTAB-D-W4MoO3 sample in N2-saturated 0.1 mol/L Na2SO4 shows a significant difference from that in Ar-saturated electrolyte, mainly due to the presence of nitrogen reduction reaction. Subsequently, a further systematic study was conducted on the electrocatalytic NRR performance of CTAB-D-W4MoO3 samples in 0.1 mol/L Na2SO4 within −0.6 V to −0.8 V vs. RHE. Fig. 3a shows the average ammonia production rate and Faraday efficiency of CTAB-D-W4MoO3 samples running chronoamperometry testing for 2 h, calculated according to formulas. Fig. 3b shows the UV–vis absorption spectra of the electrolyte at each potential after staining with indophenol blue indicator for 2 h, suggesting that the electrocatalytic NRR process can occur within −0.6 V to −0.8 V. For the CTAB-D-W4MoO3 catalyst, it can be observed that as the potential gradually becomes negative, the NH3 production rate presents a trend of first increasing and then decreasing, achieving the maximum NH3 production rate at −0.70 V, reaching 60.77 µg h-1 mg-1cat.. The Faraday efficiency (FE) gradually decreases with the decrease of potential, with the highest FE at −0.60 V of 56.42%. This phenomenon is attributed that under a relatively negative applied potential, protons occupy more active sites, so that competitive hydrogen evolution reaction occupies the dominant position, which hinders the adsorption of N molecules on the surface of the catalyst.
Due to the peak ammonia production rate of CTAB-D-W4MoO3 catalyst at −0.70 V, the durability of this catalyst material was further studied at this potential. The chemical stability and durability of CTAB-D-W4MoO3 catalyst were evaluated by cyclic and long-term electrolytic tests respectively. Five cycle tests were conducted at −0.70 V, and there was no obvious fluctuation in ammonia production rate and Faraday efficiency (Fig. 3c), which proved that CTAB-D-W4MoO3 catalyst had excellent electrochemical stability. Meanwhile, a long-term electrocatalytic nitrogen reduction process was conducted at −0.70 V for 18 h, during which the current density did not significantly decrease (Fig. 3d), further demonstrating the good durability of the catalyst. The SEM and EDS mapping images of the CTAB-D-W4MoO3 catalyst after long-term testing showed that the morphology of the catalyst is relatively intact and there is no obvious change (Figs. S9 and S10 in Supporting information). In addition, according to the time-varying curves of current density at different potentials, it can be seen that the current density keeps on stable within −0.60 V to −0.80 V, indicating the good stability of the catalyst (Fig. S11 in Supporting information).
CTAB-W4MoO3 nanowires without oxygen vacancies and CTAB-H-W4MoO3 nanowires under N2 calcination were prepared as comparison samples, respectively, and the NRR properties were measured. The ammonia production rate and FE of CTAB-W4MoO3 (20.26 µg h-1 mg-1cat. and 17.1%) and CTAB-H-W4MoO3 (32.85 µg h-1 mg-1cat. and 18.78%) (Fig. 3e) were both significantly lower than those of CTAB-D-W4MoO3 NWs (60.77 µg h-1 mg-1cat. and 32.25%), and superior to most typical catalysts (Table S2 in Supporting information). Furthermore, In the electrochemical impedance spectroscopy (EIS) test results (Fig. S12 in Supporting information), the impedance radius of CTAB-D-W4MoO3 catalyst containing oxygen defects is significantly smaller than that of CTAB-W4MoO3, indicating that CTAB-D-W4MoO3 has a faster surface electron migration rate and can realize electron transfer. To further confirm the origin of NH3, the 15N isotopic test was conducted on CTAB-D-W4MoO3. As shown in Fig. 3f, a triple peak corresponding to 14N2 and a double peak at 15N2 appeared, confirming that the ammonia originated from the supplied nitrogen. The excellent performance of CTAB-D-W4MoO3 catalyst is further demonstrated by the use of two ammonium detection methods (Nessler's reagent method and Indophenol blue method, Fig. S13 in Supporting information).
In comparison of the absorbance of the electrolyte before and after the reaction, no by-product N2H4 is found, further indicating that CTAB-D-W4MoO3 catalyst has excellent selectivity (Fig. S14 in Supporting information). To exclude contamination of the electrolyte itself, we first detect NH4+ in the electrolyte before and after introducing N2 for 30 min. The corresponding UV–vis spectrum is shown in Fig. S15a (Supporting information). The presence of NH4+ is not found, suggesting that the experimental results are not affected by impurity contamination. Fig. S15b (Supporting information) shows the UV–vis absorption spectrum of the 2 h electrolysis process at −0.70 V vs. RHE without catalyst loading on the working electrode. It can be seen that there was no NH3 generation during this process, indicating that the CTAB-D-W4MoO3 catalyst catalyzed the generation of ammonia. In addition, to check that all NH3 measured in the cathode chamber after the experiment was generated by the electrocatalytic nitrogen reduction process, electrocatalytic nitrogen reduction tests were conducted in Ar-saturated (−0.70 V) and N2-saturated (at an open circuit potential) electrolyte. Figs. S15c and d (Supporting information) show the corresponding UV–vis absorption spectra under two test conditions, indicating that NH3 was not detected under either condition. This result indicates that all previously detected NH3 comes from the NRR process. All the blank control experiments mentioned above have confirmed that the generated NH3 comes from the electrocatalytic NRR process carried out by introducing nitrogen gas, rather than interference from the environment, reactor, or reactants.
DFT calculations are used to examine possible NRR pathways on CTAB-D-W4MoO3 (Fig. 4a). It can be seen that after the first step of hydrogenation, the *N-NH intermediate can be further hydrogenated through distal and alternative pathways, respectively. According to the Gibbs free energy diagram, the potential‐determining step (PDS) of both pathways is *N-NH → *NH—NH. Due to the tendency of the alternative pathway towards lower energy (0.19/0.47 eV), the second hydrogenation step is achieved by the alternative pathway. The optimized geometries and the corresponding free energy changes of each step for CTAB-D-W4MoO3 and CTAB-W4MoO3 are shown in Fig. 4b. For the first step of hydrogenation (*N-N→*N-NH), the process on CTAB-D-W4MoO3 is a downhill pathway with ΔG=−0.15 eV, significantly lower than the process on CTAB-W4MoO3 (ΔG=1.90 eV), thus facilitating the further progress of the NRR process. Therefore, we conclude that the introduction of O vacancies in CTAB-D-W4MoO3 can reduce the energy barrier formed by the intermediate of *N-NH, facilitate the activation and further hydrogenation of *N-N, promote the NRR process, and improve the NRR performance.
In conclusion, the well-defined W4MoO3 nanowires rich in OVs (CTAB-D-W4MoO3 NWs) are constructed and prepared with the assistance of surfactant CTAB. CTAB-D-W4MoO3 NWs exhibit a high activity and selectivity with an NH3 yield of 60.77 µg h-1 mg-1cat. at −0.70 V and a FE of 56.42% at −0.60 V, outperforming CTAB-W4MoO3 NWs (NH3 yield: 20.26 µg h-1 mg-1cat.; FE: 17.1%). Meanwhile, CTAB-D-W4MoO3 NWs also show excellent electrochemical stability. The outstanding NRR performance of CTAB-D-W4MoO3 NWs is primarily ascribed to the customized electronic structure, rich active sites, high conductivity and large specific surface area. DFT calculations confirm that the introduction of O vacancies in CTAB-D-W4MoO3 reduces the energy barrier formed by the intermediate of *N-NH, which facilitates the activation and further hydrogenation of *N-N, promotes the NRR process, and improves the NRR performance. This study provides a rational design of high-performance electrocatalysts.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this pap
Jincheng Zhang: Writing – review & editing. Mengjie Sun: Writing – original draft. Jiali Ren: Writing – original draft. Rui Zhang: Writing – original draft. Min Ma: Validation. Qingzhong Xue: Writing – review & editing. Jian Tian: Writing – review & editing.
This work is supported by the National Natural Science Foundation of China (No. 51872173) and Natural Science Foundation of Shandong Province (No. ZR2022JQ21).
Supplementary material associated with this article can be found, in the online version, at doi:
T. Wu, M.M. Melander, K. Honkala, ACS Catal. 12 (2022) 2505–2512. doi: 10.1021/acscatal.1c05820
Z. Wang, J. You, J. Li, et al., ChemistrySelect 8 (2023) e202301199. doi: 10.1002/slct.202301199
M. Sun, C. Ma, M. Ma, et al., Mater. Today Phys. 30 (2023) 100945. doi: 10.1016/j.mtphys.2022.100945
Y. Zhao, J. You, L. Wang, et al., Int. J. Hydrogen Energy 46 (2021) 39146–39182. doi: 10.1016/j.ijhydene.2021.09.137
J.J. Li, J.H. You, Z.W. Wang, et al., J. Mater. Sci. 58 (2023) 10274–10287. doi: 10.1007/s10853-023-08683-1
X. Chen, S. Zhang, X. Qian, et al., Appl. Cataly. B: Environ. 310 (2022) 121277. doi: 10.1016/j.apcatb.2022.121277
Q. Wang, X. Xue, Y. Lei, et al., Small 16 (2020) 2001571. doi: 10.1002/smll.202001571
D. Chen, J. Lan, F. Xie, et al., Chem. Eng. J. 475 (2023) 146137. doi: 10.1016/j.cej.2023.146137
X. Li, J. You, J. Li, et al., ChemCatChem 16 (2024) e202301108. doi: 10.1002/cctc.202301108
X. Qian, C. Ma, U.B. Shahid, et al., ACS Catal. 12 (2022) 6385–6393. doi: 10.1021/acscatal.2c01172
X. Wang, M. Luo, J. Lan, et al., Adv. Mater. 33 (2021) 2007733. doi: 10.1002/adma.202007733
G. Liu, R. Yao, J. You, et al., Mater. Today Commun. 39 (2024) 109052. doi: 10.1016/j.mtcomm.2024.109052
X. Qian, Y. Wei, M. Sun, et al., Chin. J. Catal. 43 (2022) 1937–1944. doi: 10.1016/S1872-2067(21)64020-2
W. Li, C. Zhang, M. Han, et al., ACS Appl. Mater. Interfaces 11 (2019) 44186–44195. doi: 10.1021/acsami.9b14770
H. Huang, L. Xia, X. Shi, et al., Chem. Commun. 54 (2018) 11427–11430. doi: 10.1039/c8cc06365f
W. Xu, G. Fan, J. Chen, et al., Angew. Chem. Int. Ed. 59 (2020) 3511–3516. doi: 10.1002/anie.201914335
J. Li, W. Wang, W. Chen, et al., Nano Res. 11 (2018) 4774–4785. doi: 10.1007/s12274-018-2062-4
Q. Zhang, J. You, X. Zhang, et al., Energy Fuels 38 (2024) 6659–6678. doi: 10.1021/acs.energyfuels.3c05202
Q. Wang, Y. Lei, D. Wang, et al., Energy Environ. Sci. 12 (2019) 1730–1750. doi: 10.1039/c8ee03781g
Z.W. Wang, J.H. You, J.J. Li, et al., Catal. Sci. Technol. 13 (2023) 274–296. doi: 10.1039/d2cy01300b
Y. Zhao, J.H. You, Z.Y. Wang, et al., Inter. J. Hydrogen Energy 70 (2024) 599–605. doi: 10.1016/j.ijhydene.2024.05.186
L. Luo, B. Wang, J. Wang, et al., Phys. Chem. Chem. Phys. 23 (2021) 16658–16663. doi: 10.1039/d1cp01874d
T. He, S.K. Matta, A. Du, Phys. Chem. Chem. Phys. 21 (2019) 1546–1551. doi: 10.1039/c8cp06978f
X. Tang, J. Liu, K. Zhan, et al., Chem. Asian J. 15 (2020) 2984–2991. doi: 10.1002/asia.202000822
J. Han, X. Ji, X. Ren, et al., J. Mater. Chem. A 6 (2018) 12974–12977. doi: 10.1039/c8ta03974g
L. Zhang, X. Ji, X. Ren, et al., Adv. Mater. 30 (2018) 1800191. doi: 10.1002/adma.201800191
X. Ren, G. Cui, L. Chen, et al., Chem. Commun. 54 (2018) 8474–8477. doi: 10.1039/c8cc03627f
L.L. Han, X.J. Liu, J.P. Chen, et al., Angew. Chem. Int. Ed. 58 (2019) 2321–2325. doi: 10.1002/anie.201811728
H. Cheng, L.X. Ding, G.-F. Chen, et al., Adv. Mater. 30 (2018) 1803694. doi: 10.1002/adma.201803694
X. Zhang, Y. Guo, J. Tian, et al., Appl. Catal. B: Environ. 232 (2018) 355–364. doi: 10.1016/j.apcatb.2018.03.074
G. Zhang, Q. Ji, K. Zhang, et al., Nano Energy 59 (2019) 10–16. doi: 10.1016/j.nanoen.2019.02.028
X. Zhong, Y. Sun, X. Chen, et al., Adv. Funct. Mater. 26 (2016) 5778–5786. doi: 10.1002/adfm.201601732
F. Shen, W. Jiang, G. Qian, et al., J. Power Sources 458 (2020) 228014. doi: 10.1016/j.jpowsour.2020.228014
J. Wang, K. Chang, Z. Sun, et al., Appl. Catal. B: Environ. 251 (2019) 162–167. doi: 10.1016/j.apcatb.2019.03.065
Y. Sun, X. Hu, W. Luo, et al., ACS Nano 5 (2011) 7100–7107. doi: 10.1021/nn201802c
Z. Guo, Q. Ma, Z. Xuan, et al., RSC Adv. 6 (2016) 16730–16735. doi: 10.1039/C5RA26689K
Figure 1 (a) The schematic synthesis diagram of W4MoO3 nanowires deficient in oxygen vacancies (CTAB-W4MoO3) and W4MoO3 nanowires rich in O-vacancies (CTAB-D-W4MoO3 NWs). (b) XRD patterns and (c) Raman spectra of CTAB-W4MoO3 and CTAB-D-W4MoO3.
Figure 2 (a) Scanning electron microscope (SEM), (b) TEM, (c) high resolution transmission electron microscope (HRTEM), and (d) corresponding EDS elemental mapping images of CTAB-D-W4MoO3.
Figure 3 (a) NH3 yields and FEs of CTAB-D-W4MoO3 at a series of potentials. (b) UV–vis absorption spectra of 0.1 mol/L Na2SO4 after 2 h electrolysis on CTAB-D-W4MoO3. (c) NH3 yields and FEs of CTAB-D-W4MoO3 at −0.70 V during recycling tests. (d) Time-dependent current density curves of CTAB-D-W4MoO3 for NRR at −0.70 V. (e) NH3 yields and FEs of CTAB-D-W4MoO3 CTAB-H-W4MoO3 and CTAB-W4MoO3. (f) 1H NMR spectra of the electrolyte after NRR using 15N2 and 14N2 as the nitrogen source, respectively.
扫一扫看文章
扫一扫关注我们
DownLoad:
下载:
下载: