Yuyao
Ji‡
a,
Mingyu
Yang‡
ab,
Wendong
Cheng
a,
Chengbo
Li
b and
Xingquan
Liu
*a
aUniversity of Electronic Science and Technology of China, Chengdu 610054, China. E-mail: lxquan@uestc.edu.cn
bCollege of Chemistry and Materials Science, Sichuan Normal University, Chengdu 610068, Sichuan, China
First published on 22nd September 2022
At present, most industrial NH3 production comes from the Haber–Bosch process, which causes a series of serious environmental pollution problems. Electrochemical N2 reduction is regarded as a green pathway to deal with this problem. Recently, CeO2 has attracted much attention due to its high thermal stability. Metal doping with smaller ion radius is an effective strategy to regulate oxygen vacancies, increase the defect concentration and enhance the catalyst activity. Herein, we developed Zn-doped CeO2 nanospheres (Zn–CeO2) for the electrochemical NRR. In 0.1 M Na2SO4, Zn–CeO2 achieves a large NH3 yield of 29.01 μg h−1 mgcat.−1 and a high faradaic efficiency of 10.3% at −0.20 V vs. reversible hydrogen electrode, and it also shows good structure stability. The density functional theory (DFT) calculations revealed the reaction mechanism of NRR activity regulated by the doping metal.
As an important oxide catalyst, cerium oxide material has been a research hotspot in the field of ternary catalysts, environmental catalysis and catalyst supports for decades.10,11 However, the reported NRR activity of CeO2 related catalysts is still not high, which is mainly caused by their low electrical conductivity and poor N2 adsorption and activation.9 Recent reports have highlighted that oxygen vacancies (OV) can receive foreign electrons to enhance N2 adsorption and activation, thereby improving the NRR performance of the catalysts.12–15 Moreover, a large number of applications of CeO2 in the field of electrocatalysis mainly benefit from the existence of OV.16 Furthermore, a recent report by Liu et al.17 showed that OVs of CeO2 can be significantly enhanced via doping with transition elements; meanwhile, Liu et al.18 also proved that ZnO is an efficient NRR electrocatalyst. The above studies further motivated us to use elements with smaller ionic radii as dopants to improve the NRR performance of pure CeO2.
In this work, we report Zn-doped CeO2 nanospheres to modulate oxygen vacancies for promoting the NRR performance. As observed, in 0.1 M Na2SO4, the NH3 yield rate of Zn–CeO2 (29.01 μg h−1 mg cat.−1) is much better than that of undoped CeO2, with a high faradaic efficiency (FE) of 10.3%. It also shows good stability. Density functional theory (DFT) calculations reveal the mechanism that doped pairs are the active sites.
The XRD patterns of CeO2 and Zn–CeO2 nanospheres are shown in Fig. 1a, and all the XRD peaks positioned in each sample correspond to (111), (200), (220), (311), (222), (400), (331), and (420) planes. It is worth noting that no other secondary or impurity peaks are observed in Fig. 1a, indicating that Zn has been successfully doped with CeO2. Scanning electron microscopy (SEM) reveals the CeO2 nanospheres before and after Zn doping, as shown in Fig. 1b and c. The high-resolution TEM (HRTEM) image reveals the crystallographic fringes of 0.308 nm and can be well assigned to the (111) lattice plane of CeO2 (Fig. 1e). Furthermore, the corresponding elemental mapping images (Fig. 1f and g) of Zn–CeO2 also demonstrate that Zn is successfully doped in CeO2. From the above characterization results, it can be seen that we successfully synthesized a Zn–CeO2 nanosphere catalyst.
In order to further determine the element valence in the composite material, the XPS spectrum is necessary, and the results show the presence of Ce, O, and Zn elements in the Zn–CeO2 materials. Fig. 2a shows the survey scan of Ce 3d, Zn 2p and O 1s. The XPS spectra of Zn 2p3/2 and Zn 2p1/2 correspond to the binding energies of 1021.6 eV and 1043.8 eV (Fig. 2b). The spin-orbital splitting of 23.1 eV, between the peaks, confirmed that Zn exists as pure metal on the CeO2 matrix rather than its oxide form.19Fig. 2c shows the Ce 3d spectrum, and the peaks located in the range of 881–902 eV correspond to Ce 3d5/2; meanwhile, the peaks in the range of 901–921 eV correspond to Ce 3d3/2.20–22 The above results can be attributed Ce3+ and Ce4+.23 And they are match well with the previous reports.24,25Fig. 2d shows the XPS spectrum of O 1s. The peak seen at low binding energy (529.9 eV) corresponds to the oxygen atom in the CeO2 lattice,26 and the other peak at high binding energy (532.4 eV) corresponds to chemisorbed oxygen of the surface hydroxyl group.27 From the above results, it can be concluded that Ce element exists in the +3 or +4 oxidation state in the composite. For example, in the original cerium oxide, Zn exists in the 0 oxidation state, and oxygen exists in the −2 oxidation state and in the 0 oxidation state. Fig. S1 (ESI†) shows the Raman spectra of Zn–CeO2 and pure CeO2. Notably, the intensity in Zn–CeO2 is less than pure CeO2, implying that more oxygen vacancies are present in Zn–CeO2. Meanwhile, Fig. S2 (ESI†) shows the room temperature electron spin resonance (ESR) spectra of pure CeO2 and Zn–CeO2. The latter shows a definite oxygen vacancy signal at g = 2.018, indicating the formation of a large number of oxygen vacancies after the Ce3+ center.
Our electrochemical tests are carried out in U-shaped electrolyzers separated by membranes. The optimum catalyst load is 0.1 mg cm−2. For more accurate determination of ammonia concentration, as shown in Fig. S3a (ESI†), the time-dependent current density curves of Zn–CeO2 were obtained. Fig. S3b (ESI†) presents the UV-Vis absorption. After electrolysis, the obtained NH3 in the solution phase was spectrophotometrically determined by the indophenol blue method,28 and another possible by-product (N2H4) was detected by the method of Watt and Chrisp.29 The corresponding calibration curves are shown in Fig. S4 (ESI†), respectively. It is worth noting that we did not detect the byproduct hydrazine in the reaction solution (Fig. S5, ESI†). The ammonia production rate and Faraday efficiency of the catalyst at different voltages are shown in Fig. 3a, and the ammonia production rate (29.01 μg h−1 mgcat.−1) and Faraday efficiency (10.3%) of the catalyst were the highest at −0.2 V, outperforming most reported NRR electrocatalysts listed in Table S1 (ESI†). To better demonstrate the NRR performance of the Zn–CeO2, we compare the electrochemical activity between the material and the precursor by the amount of ammonia produced (Fig. 3b). The results showed that the composites Zn–CeO2/CP showed the best NRR activity, and its NRR performance is much higher than that of CeO2/CP and bare CP. Stability is another indicator of electrochemical performance; the ammonia production rate and Faraday efficiency were tested after six cycles at −0.6 V (Fig. 3c). And UV-vis absorption spectra (Fig. S6, ESI†) confirm the high stability of Zn–CeO2/CP. After the cycling test, we found that the ammonia production rate and Faraday efficiency of the catalyst did not change basically. In addition, the current density of the catalyst did not fluctuate after 24 hours of electrolysis (Fig. 3d). This further shows that the catalyst has good electrochemical stability. The amount of NH3 measured in the Ar-saturated electrolyte at each potential was very small and insignificant compared to the N2-saturated electrolyte (Fig. S7, ESI†), indicating that the NH3 product is mainly generated by the supply of N2 electrocatalyzed by Zn–CeO2.
Fig. 3 (a) NH3 yields and FEs at each given potential. (b) NH3 yields at −0.20 V. (c) Stability test of Zn–CeO2 for 6 cycles at −0.20 V. (d) The curve for Zn–CeO2 at −0.20 V. |
In order to explore the mechanism by which Zn, as a dopant, regulates defect concentration to enhance NRR activity from a microscopic perspective, the reaction mechanism of the NRR on the CeO2(111) surface was observed by DFT calculation. There are several well-established reaction pathways for the NRR depending on the specific adsorption modes of N2 molecules.30,31 To evaluate the potential of Zn–CeO2 as an electrocatalyst for nitrogen reduction, typical reaction paths through the distal mechanism were optimized and the corresponding free energy profiles/structures are summarized in Fig. 4a. Importantly, the N2 adsorption energy on the Zn–CeO2(111) surface with VO is higher than that on pure MnO2 (−0.24 vs. −0.58 eV). These results are attributed to the stronger electronic interaction between VO, N2 and Ce atoms. However, the barrier of NNH* protonation is greatly reduced after Zn doping. The second NH3* formation is the Zn–CeO2 limiting step with a critical energy barrier of 0.24 eV (NH2* + H+ + e− → NH3*). As shown in Fig. S8 (ESI†), the results show that the Zn doped surface has more vacant orbitals, which can enhance the adsorption energy of N2. The density of states (DOS) indicates that two orbitals hybridize after N2 adsorption by adding H atoms to the adsorbent, and N2 hydrogenation is performed by adding H atoms to the adsorbent based on a distal or alternate mechanism (Fig. S9, ESI†). The top panel of Fig. S10 (ESI†) shows the end-to-end adsorption configuration (a) with VO and (b) without VO for the N2 molecule on the CeO2(111) surface. For the former, the N2 molecule occupies VO, and one of its terminal N atoms interact directly with VO. In general, our theoretical calculations point out that Zn atoms can significantly improve the NRR performance, which is in good agreement with the experimental electrochemical results.
In conclusion, Zn has been shown to be an effective dopant to regulate the CeO2 defect concentration to enhance the NRR performance. In 0.1 M Na2SO4, Zn–CeO2 attains the largest NH3 yield of 29.01 μg h−1 mgcat.−1 and highest FE of 10.3% at −0.20 V. Moreover, it has good electrochemical stability, and its catalytic activity is basically unchanged after 24 hours of electrolysis. DFT calculation shows that doping Zn element with small ion radius in CeO2(111) can regulate and increase the concentration of oxygen vacancies, thus promoting the adsorption and activation of N2. Moreover, Ce3+ formed by oxygen vacancy defects is more likely to capture electrons, thus improving the NRR activity of the catalyst. This work not only provides an attractive scheme for the construction of the defect concentration on the catalyst surface but also opens new opportunities to explore cerium-based catalysts for N2 fixation applications.
Footnotes |
† Electronic supplementary information (ESI) available: Experimental section and supplementary figures. See DOI: https://doi.org/10.1039/d2ya00241h |
‡ Yuyao Ji and Mingyu Yang contribute equally to this manuscript. |
This journal is © The Royal Society of Chemistry 2022 |