b Research Institute of Processing(RIIP), SINOPEC, Beijing 100083, China
Natural gas, which consists mainly of methane, is an important and cheap natural resource for the production of platform chemicals and as an energy source. Since 2010, several countries have made breakthroughs in the field of shale gas exploiting, thus a large amount of difficult-to-exploit methane can be extracted [1, 2]. Therefore, the utilization of methane has attracted even more attentions and interests. Oxidative coupling of methane (OCM) is regarded as an effective way to utilize methane with added-value, since it can transform methane into C2 products, among which ethylene is a crucial platform organic compound for chemical industry. Up to now, several kinds of catalysts have been tried for this reaction [3-6]. It is commonly accepted that Mn/Na2WO4/SiO2 is still the most promising catalyst for industrial application, on which 27% C2 product yield can be obtained at 850 ℃ [7, 8]. Whereas, the industrially implement of this vital resource reaction is still limited by the one-way C2 yield below 30%, which is regarded as the minimum requirement to industrialize this reaction economically and technically [9, 10]. Therefore, it is still of great necessity to develop more efficient catalysts for this reaction.
A2B2O7 pyrochlore compounds generally own high thermal stability, intrinsic 8a oxygen vacancies and certain surface alkalinity, which fit to the active site requirements for a good OCM catalyst. Therefore, since late 1980s, some A2B2O7 pyrochlores have been tested for this reaction preliminarily, and found to display certain reactivity [11-14]. But to date, only limited literatures are documented and people are still in short of deep understanding on the structure and reactivity relationship for the application of pyrochlore compounds for OCM reaction. Our former study has proved that La2Ce2O7 displays very good reaction performance for OCM . In particular, compared with the stateof-the-art Mn/Na2WO4/SiO2 catalyst, it possesses much better low temperature reaction performance, though its highest C2 product yield is still lower than that of Mn/Na2WO4/SiO2. For instance, even at 650 ℃, 15.3% C2 yield can be achieved on La2Ce2O7, but over Mn/ Na2WO4/SiO2, the reaction is just started with a C2 yield below 1% [16, 17]. Furthermore, La2Ce2O7 demonstrates very potent resistance to sulfur and lead poisoning. It is revealed that the coexistence of both active electrophilic oxygen and alkaline sites and the synergistic interaction between them are the major reasons to control the reaction. With the target to develop better catalysts and gain deeper understanding on structure and reactivity relationship, Ln2Zr2O7 pyrochlore compounds (Ln = La, Pr, Sm and Y) with a fixed Zr4+ B site and varied rare earth A sites have been investigated for OCM reaction as a continued work. Indeed, it is found that the reactivity of the catalysts changes with the change of the A site cation. La2Zr2O7 displays the best reaction performance among all the catalysts, on which 15.1% C2 yield can be obtained even at 750 ℃. Moreover, at low temperature region between 650–700 ℃, it also exhibits much better reaction performance than MnNa2WO4/SiO2. These finding could give people some new insights on how to develop OCM catalysts that can be operated at relatively lower temperature region. In addition, the bulk and surface structure changes of Ln2Zr2O7 have been characterized by different means, and the relationship between the reactivity and structure has been elucidated.
The Ln2Zr2O7 pyrochlore compounds have been prepared by a co-precipitation method. The detailed activity evaluation and characterization experiment procedures of the catalysts are described in the supporting information. For comparison purpose, a ZrO2 sample was also prepared by precipitation method, and a Mn-Na2WO4/SiO2 catalyst was prepared with a traditional impregnation method according to the references .
The reaction performance of the Ln2Zr2O7 catalysts has been tested and shown in Fig. 1. Pure ZrO2 displays very poor reaction performance for OCM, as evidenced by the very low methane conversion, C2 product selectivity and yield in the whole tested temperature region. In comparison, all the Ln2Zr2O7 catalysts exhibit significantly improved reaction performance, indicating the formation of pyrochlore structure generates the active sites required for OCM reaction. With the increasing of the reaction temperature from 650 ℃ to 750 ℃, while the methane conversion improves, the C2 product selectivity is nearly constant, proving that the C2 yield increase is contributed mainly by the methane conversion change. In addition, the results imply that the C2 product selectivity on the catalysts is not sensitive to the reaction temperatures. As shown in Fig. 1C, the C2 product yields achieved over the catalysts follow the sequence of La2Zr2O7 > Pr2Zr2O7 > Sm2Zr2O7 ≈ Y2Zr2O7 > ZrO2. Among all the catalysts, La2Zr2O7 shows the best reaction performance, on which the highest C2 yield around 15.1% can be achieved even at 750 ℃. For clarification, the coupling product distribution on the catalysts collected at 750 ℃ is presented in Table S1 (Supporting information). It is revealed that over all the catalysts, the percentage of C2H6 is higher than that of C2H4. However, La2Zr2O7, the best catalyst in this study, owns the highest C2H4 selectivity in all the catalysts, testifying that this catalyst also has more abundant active sites favourable for C2H6 dehydrogenation.
|Fig. 1. Reaction performance of the Ln2Zr2O7 catalysts. (A) CH4 conversion, (B) C2 product selectivity, (C) C2 product yield, (D) Stability test at 750 ℃ over La2Zr2O7 catalyst.|
For an OCM catalyst, the stability is an important parameter to estimate its application potential, since this reaction generally requires high temperature to proceed. Where upon, the long-term stability of La2Zr2O7, the best catalyst, has been evaluated at 750 ℃ for 80 h. As demonstrated in Fig. 1D, the methane conversion, C2 selectivity and yield remains constant during the time span without any decrease, testifying this catalyst possesses superior stability for the high temperature OCM process.
Currently, Mn-Na2WO4/SiO2 is still the most promising OCM catalyst from the point of view of industrial application, on which the highest one way C2 product yield close to 27% can be obtained at 850 ℃ . However, this catalyst has very low activity below 700 ℃. To reduce production cost, it is always desirable to find catalysts that can be operated at relatively lower temperatures region. As compared in Fig. S1 (Supporting information), La2Zr2O7, the best catalyst in this study, displays much better reaction performance than Mn-Na2WO4/SiO2 under the same condition in the temperature region of 650~700 ℃. For instance, even at 650 ℃, 5.4% C2 yield has already been achieved over La2Zr2O7, but the reaction over Mn-Na2WO4/SiO2 is just started with an extremely low C2 yield of 0.2%. At 700 ℃, the C2 yield over La2Zr2O7 increases to 9.8%, which is still much higher than that over Mn-Na2WO4/SiO2.However, starting from 750 ℃, the C2 yield over La2Zr2O7 becomes about 1% lower than that over Mn-Na2WO4/SiO2. In brief, Ln2Zr2O7 pyrochlores possess surface active sites that can match the requirements of OCM reaction. With suitable element combination, La2Zr2O7 and Pr2Zr2O7 catalysts having outstanding low temperature reaction performance and high stability has been attained. This implies that with further optimization, pyrochlore type of OCM catalysts that can function at lower temperature could be eventually obtained.
The phase compositions of the catalysts have been analysed by XRD, with the patterns shown in Fig. 2. Pure ZrO2 shows the typical diffraction feature of monoclinic phase structure, as also testified by the lattice parameters in Table S2 (Supporting information). In contrast, for all the Ln2Zr2O7 catalysts, cubic pyrochlore is the only detectable crystalline phase, no any impurity diffraction belonging to the individual oxides can be observed, indicating that pure pyrochlore phase has been successfully synthesized. The intensive peaks observed indicate that all the smaples are well crystallized. With the decreasing of rA/rB ratio from La2Zr2O7 to Y2Zr2O7, the peak position of the Ln2Zr2O7 pyrochlores shifts gradually to higher 2θ angles, as dispalyed in the enlarged Fig. 2B, and the side length becomes smaller, indicating the presence of cell shrinking effect caused by the decrease of the radii of the A site cations. It was reported formerly that with the decreasing of the rA/rB ratio, the crystalline structure of the samples could transform from ordered pyrochlore (La2Zr2O7) to less ordered pyrochlore (Pr2Zr2O7 and Sm2Zr2O7) and eventually to defective cubic fluorite phase (Y2Zr2O7) . As a result, the surface oxygen property related to the 8a intrinsic oxygen vacancies could be changed, which subsequently influence the OCM reaction performance . The XPS and O2-TPD results indeed testify the change of the oxygen property, which will be discussed in more detail in the following sections.
The texture properties of the Ln2Zr2O7 catalysts have also been analysed by N2 adsorption-desorption, with the isotherms and pore size distribution profiles displayed in Fig. S2 (Supporting information). Pure ZrO2 depicts a typical type Ⅳ isotherm with an H1-type hysteresis loop, and its pore size distribution profile shows a wide peak, which corresponds to an average pore size of 21 nm (Table S2), implying the presence of inter-particle pores. In contrast, all the Ln2Zr2O7 catalysts possess also a typical type Ⅳ isotherm but with an H2-type hysteresis loop, indicating that they have the same texture structure with each other but which is different from that of the individual ZrO2. However, the quantification results in Table S2 manifest that both Sm2Zr2O7 and Y2Zr2O7 have much higher pore volumes than that of La2Zr2O7 and Pr2Zr2O7. In addition, the surface areas of the former two catalysts are higher than the latter two. Theoretically, the higher pore volumes and surface areas are generally favourable for the reactants diffusion and contacting with the surface active sites, which might promote the reaction performance of the catalysts. Whereas, as depicted in Fig. 1, the reaction performance Sm2Zr2O7 and Y2Zr2O7 is much worse than La2Zr2O7 and Pr2Zr2O7, which indicates that the texture structure change is not the crucial factor to affect the OCM reaction performance on the catalysts.
Raman technique was adopted to investigate the structure change of Ln2Zr2O7 compounds, with the profiles manifested in Fig. S3 (Supporting information). It is particularly noted here that the XRD patterns in Fig. 2 have indicated that all the Ln2Zr2O7 samples are well crystallized due to high temperature calcination. Therefore, it is rational to propose that the change of Raman spectra can be ascribed to the phase change instead of crystalinity change of the samples. It is noticed that all the Ln2Zr2O7 catalysts show complete different spectra from pure ZrO2, which confirms the structure difference observed by XRD. A well-ordered cubic A2B2O7 pyrochlore generally has six typical Raman peaks at 299, 395, 500, 520, 601 and 750 cm-1 [21, 22]. The Raman peaks at 520 cm-1 is assigned to A1g mode related to the O－B－O banding vibrations. The most intense Raman band at about 299 cm-1 is assigned to Eg mode. Moreover, the Raman band at around 395, 500, 601 and 750 cm-1 can be assigned to the F2g mode. It is well known that Raman spectroscopy can provide explicit information to distinguish between a pyrochlore, biphasic mixture and a defect-fluorite material . The presence of all the six Raman bands in the profile of La2Zr2O7 testifies the formation of ordered pyrochlore phase in its bulk. In contrast, on the profiles of Pr2Zr2O7 and Sm2Zr2O7, only the strongest F2g peak is still detected with the disappearance of other bands, testifying that the pyrochlore phase is still present but become less ordered . On the profile of Y2Zr2O7, all the Raman bands assigning to pyrochlore phase disappear completely, indicating the complete transformation of the crystalline phase. As discussed above for the XRD results, it is believed that a defective cubic fluorite phase could have eventually been formed in this sample.
H2-TPR was used to investigate the redox properties of the Ln2Zr2O7 catalysts, with the profiles shown in Fig. S4A (Supporting information). For all of the Ln2Zr2O7 catalysts with different A site cations, a small reduction peak is observed in the temperature region of 300–500 ℃, as better indicated by the enlarged profiles in Fig. S4B (Supporting information). It is noted that pure ZrO2 shows also a minute reduction peak at ~345 ℃. Theoretically, if all the lattice oxygen of a Ln2Zr2O7 compound can be reduced, the O/ (Ln + Zr) atomic ratio should be 1.75. However, the quantified results in Table S3 (Supporting information) prove that on all the catalysts with varied A sites, the ratios are much lower than 1.75, proving the peak can be assigned to the reduction of a very small amount of surface deficient oxygen species. In addition, the oxygen amount follows the order of La2Zr2O7 > Pr2Zr2O7 > Sm2Zr2O7 = Y2Zr2O7, which is nearly in accordance to the reaction performance of the catalysts. Therefore, it is reasonable to believe that the amount of surface active oxygen sites could be important for the reaction performance.
To further understand the oxygen property of the catalysts, O2- TPD experiments were also conducted for all the Ln2Zr2O7 catalysts, with the profiles shown in Fig. S5 (Supporting information). Two groups of desorption peaks are observed for all the Ln2Zr2O7 samples and the pure ZrO2. The low temperature peak at ~100 ℃, which is labbled as α peak, is assigned to the desorption of loosely bonded surface oxygen species . The broad desorption peak above 200 ℃, which is labelled as β peak, is obviously composed of two overlapped peaks. Therfore, it was deconvoluetd into two peaks at ~300 and ~410 ℃, respectively. The peak at ~ 300 ℃ may correspond to the desorption of surface electrophilic oxygen species formed by the inducing of the inherent 8a oxygen vacancies in the lattice , and the peak at ~410 ℃ could be assigned to the desorption of surface mobile lattice oxygen . Since the β group oxygen sites are close to the OCM reaction temperature range, we hence believe that they should contribute mainly to the reaction. As quantified in Table S4 (Supporting information), the amount of β group oxygen sites follows the sequence of La2Zr2O7 > Pr2Zr2O7 > Sm2Zr2O7 > Y2Zr2O7, which is in line with the H2-TPR results. It is worth noting here that La2Zr2O7 possesses the largest amount of surface active oxygen species, thus having the highest OCM performance among all the catalysts. This testifies again that the amount of surface active oxygen species is an important factor to decide the OCM reaction performance.
To further investigate the impact of A site change on the surface oxygen property of the Ln2Zr2O7 catalysts, XPS experiments were also performed, with the spectra shown in Fig. 3. For all the Ln2Zr2O7 catalysts, including the comparison pure ZrO2, two overlapped O 1s peaks are obviously observed. Fig. S6 (Supporting information) proves that carbonate is formed on the surface of all the catalysts, as indicated by the XPS peak at ~288.9 eV. Therefore, part of the surface oxygen could be contributed by surface carbonates, which is generally inert for any oxidation reaction . Based on this, the O 1s spectra of the samples are deconvoluetd into three peaks according to the typical binding energies. The major peak at ~529.0 eV can be assigned to surface lattice O2-, the peak at ~531.0 eV can be ascribed to oxygen species related to surface carbonates, and the peak at ~532.8 eV can be attributed to surface superoxide O2- cations . Former studies have substantiated that superoxide O2- sites are important for the activation of methane molecules and the selective formation of coupling products during OCM reaction [29, 30]. Therefore, the surface superoxide O2- site percentages of the catalysts are quantified and listed in Table S5. Apparently, the percentages of the Ln2Zr2O7 catalysts follow the sequence of La2Zr2O7 > Pr2Zr2O7 > Sm2Zr2O7 > Y2Zr2O7, which is well consistent with the H2-TPR and O2-TPD results.
|Fig. 3. Analyzing surface oxygen property of the Ln2Zr2O7 catalysts by XPS. (A) O 1s of La2Zr2O7, (B) O 1s of Pr2Zr2O7, (C) O 1s of Sm2Zr2O7, (D) O 1s of Y2Zr2O7, (E) O 1s of ZrO2, (F) the relationship between surface active oxygen amount and C2 product yields.|
For better understanding, the surface superoxide O2- percentages of the catalysts and C2 product yields achieved at 650, 700 and 750 ℃ are plotted in Fig. 3F. It is apparent that with the decreasing of the superoxide O2- percentages in the order of La2Zr2O7 > Pr2Zr2O7 > Sm2Zr2O7 > Y2Zr2O7, the C2 product yields degrade with the same order. This strongly demonstrates again that the amount of surface active electrophilic oxygen species is a major factor to decide the OCM reaction performance of the La2Zr2O7 catalysts.
It is particularly noted here that although ZrO2 also possesses a small amount of surface active oxygen species, our reaction results in Fig. 1 have testified that it displays very poor reaction performance for OCM, with the formation of negligible amount of C2 product due to the short of surface alkaline sites. Many previous studies have testified that both surface active electrophilic oxygen species and alkaline sites are indispensible for OCM reaction. The surface alkaline sites are active for methane molecules adsorption and activation, and the surface electrophilic oxygen sites are responsible for the formation of coupling products, although sometimes they are also involved into methane activation . Our former work has also testified that the concerted interaction between surface alkaline sites and active oxygen species indeed controls the OCM reaction performance of a catalyst .
According to this discussion, the amount of surface alkaline sites could correlate to the OCM reaction performance of a catalyst intimately. Therefore, to probe the effect of the differed A sites on the surface alkalinity of the Ln2Zr2O7 catalysts, they are subjected to CO2-TPD analysis. As manifested by Fig. 4, pure ZrO2 displays no observable CO2 desorption in the tested temperature region, indicating the absence of alkaline sites on its surface, which explains its very low OCM activity with the formation of negligible amount of C2 products. In contrast, all Ln2Zr2O7 catalysts depict two CO2 desorption peaks. The first peak emerges in the temperature region of 50–200 ℃, which is assigned to weak surface alkaline sites. The second peak emerges in the temperature range of 200–500 ℃, which is assigned to alkaline sites with moderate strength. It is worth noting here that the alkaline sites with moderate strength are particularly beneficial to the selective formation of C2 coupling products [15, 31-33]. Therefore, the CO2 desorption amount is quantified in Table S6 (Supporting information) for all the catalysts, which indicates that the amount of the moderate alkaline sites decreases in the order of La2Zr2O7 > Pr2Zr2O7 > Y2Zr2O7 > Sm2Zr2O7. In brief, the substitution of the A sites of Ln2Zr2O7 compounds affects not only its surface active oxygen amount, but also its surface alkaline sites quantity, which could eventually influencing the reaction performance of the catalysts.
For clarification, the amount of moderate surface alkaline sites is plotted against the C2 product yields achieved at 650, 700 and 750 ℃ in Fig. 4B. It is typically observed that the C2 product yields on the catalysts increase with the increasing of the moderate alkaline sites amount. Whereas, though Y2Zr2O7 possesses a larger amount of surface moderate alkaline sites than Sm2Zr2O7, it displays similar reaction performance to the latter. This proves again that surface alkalinity is not the only factor deciding the reaction performance. Taking into account of the surface oxygen property explored by H2-TPR, O2-TPD and XPS, it is rational to propose that both the surface oxygen species and alkaline sites are critical for OCM reaction, and their concerted interaction determines the reaction performance. By tuning the combination of A and B sites of A2B2O7 compounds, catalysts with suitable amount of both active sites could be achieved, thus developing catalysts having industrialization potential for OCM.
In conclusion, with the objective to develop catalysts having application potential for OCM, especially at relatively lower temperature region. A series of Ln2Zr2O7 compounds with a fixed Zr4+ cation B sites but with varied rare earth A sites (Ln = La, Pr, Sm and Y) have been prepared by a co-precipitation method in this study. The surface and bulk properties of the catalysts have been characterized by different means, and correlated with their reaction performance.
(1) XRD and Raman results have proved that pure Ln2Zr2O7 compounds have been successfully prepared for all the catalysts. With the decreasing of the rA/rB ratio, the crystalline structure of the compounds transform from an ordered pyrochlore (La2Zr2O7) to a less ordered pyrochlore (Pr2Zr2O7 and Sm2Zr2O7) and eventually to a defective cubic fluorite phase (Y2Zr2O7).
(2) As testified by H2-TPR, O2-TPD and XPS, the amount of surface active superoxide O2- species follows the order of La2Zr2O7 > Pr2Zr2O7 > Sm2Zr2O7 > Y2Zr2O7, which is well consistent with the reaction performance of the catalysts, and indicates that the abundance of surface active oxygen sites is a critical factor influencing the reaction performance. CO2-TPD results have demonstrated a better catalyst generally possesses a larger amount of surface moderate alkaline sites, which is another factor to affect the reaction performance.
(3) It is concluded that the concerted interaction between the two types of surface active sites controls the reaction performance of the Ln2Zr2O7 catalysts. La2Zr2O7 possesses the largest amount of both kinds of surface sites, hence displaying the optimal reaction performance among all the catalysts. In comparison with the state-of-the art Mn/Na2WO4/SiO2, La2Zr2O7 exhibits much better reaction performance at a low temperature region (< 750 ℃). Therefore, the information included in this study could give people some new insights on how to develop improved OCM catalysts that can be operated at relatively lower temperatures.Acknowledgments
This work is supported by the Natural Science Foundation of China (Nos. 21567016, 21566022, 21666020), the Natural Science Foundation of Jiangxi Province (Nos. 20181ACB20005, 20171BAB213013 and 20181BAB203017), the Key Laboratory Foundation of Jiangxi Province for Environment and Energy Catalysis (No. 20181BCD40004), the Education Department of Jiangxi Province (Nos. GJJ150016, GJJ150085 and KJLD14005), the China Postdoctoral Science Foundation (No. 2018M631294), Innovation Fund Designated for Undergraduate Students of Nanchang University of China (No. 201802369) and the Graduate Student Creativity Funding of Nanchang University (No. 201802062), which are greatly acknowledged by the authors.Appendix A. Supplementary data
Supplementary material related to this article canbefound, in the online version, at doi:https://doi.org/10.1016/j.cclet.2019.03.031.
H. Ding, Y. Xu, C. Luo, et al., Energy Convers. Manage. 171 (2018) 12-19. DOI:10.1016/j.enconman.2018.05.088
E.C. Umeozor, S.M. Jordaan, I.D. Gates, Energy 152 (2018) 594-600. DOI:10.1016/j.energy.2018.03.151
S. Parishan, P. Littlewood, A. Arinchtein, V. Fleischer, R. Schomaecker, Catal. Today 311 (2018) 40-47. DOI:10.1016/j.cattod.2017.08.019
Y.A. Ivanova, E.F. Sutormina, N.A. Rudina, A.V. Nartova, L.A. Isupova, Catal. Commun. 117 (2018) 43-48. DOI:10.1016/j.catcom.2018.08.001
M. Haneda, Y. Katsuragawa, Y. Nakamura, Towata A., Front. Chem. 6 (2018) 581. DOI:10.3389/fchem.2018.00581
Z. Cheng, D.S. Baser, S.G. Nadgouda, et al., ACS Energy Lett. 3 (2018) 1730-1736. DOI:10.1021/acsenergylett.8b00851
S. Gu, H.S. Oh, J.W. Choi, et al., Appl. Catal. A:Gen. 562 (2018) 114-119. DOI:10.1016/j.apcata.2018.05.031
J. Wang, L. Chou, B. Zhang, et al., J. Mol. Catal. A:Chem. 245 (2006) 272-277. DOI:10.1016/j.molcata.2005.09.038
M. Albrecht, U. Rodemerck, D. Linke, E.V. Kondratenko, React. Chem. Eng. 3 (2018) 151-154. DOI:10.1039/C7RE00208D
K. Takahashi, I. Miyazato, S. Nishimura, J. Ohyama, et al., ChemCatChem 10 (2018) 3223-3228. DOI:10.1002/cctc.v10.15
A.C. Roger, C. Petit, L. Hilaire, et al., Catal. Today 21 (1994) 341-347. DOI:10.1016/0920-5861(94)80155-X
C. Petit, J.L. Rehspringer, A. Kaddouri, et al., Catal. Today 13 (1992) 409-416. DOI:10.1016/0920-5861(92)80166-K
C.A. Mims, A.J. Jacobson, R.B. Hall, et al., J. Catal. 153 (1995) 197-207. DOI:10.1006/jcat.1995.1122
A.C. Roger, C. Petit, A. Kiennemann, J. Catal. 167 (1997) 447-459. DOI:10.1006/jcat.1997.1601
J. Xu, L. Peng, X. Fang, et al., Appl. Catal. A:Gen. 552 (2018) 117-128. DOI:10.1016/j.apcata.2018.01.004
C. Uzunoglu, A. Leba, R. Yildirim, Appl. Catal. A:Gen. 547 (2017) 22-29. DOI:10.1016/j.apcata.2017.08.020
W. Riedel, L. Thum, J. Moeser, et al., J. Phys. Chem. C 122 (2018) 22605-22614. DOI:10.1021/acs.jpcc.8b07386
R.E. Owen, O'Byrne J.P., D. Mattia, et al., Chem. Commun. (Camb.) 49 (2013) 11683-11685. DOI:10.1039/c3cc46791k
C. Karakaya, H. Zhu, C. Loebick, J.G. Weissman, R.J. Kee, Catal. Today 312 (2018) 10-22. DOI:10.1016/j.cattod.2018.02.023
X. Zhang, X. Fang, X. Feng, et al., Catal. Sci. Technol. 7 (2017) 2729-2743. DOI:10.1039/C7CY00004A
Jafar, S.N. Achary, P. Niliesh, et al., J. Nucl. Phys. Mater. Sci. Radiat. Appl. 475 (2016) 192-199.
M.L. Sanjuán, C. Guglieri, Díaz-Moreno S., et al., Phys. Rev. B 84 (2011) 104207. DOI:10.1103/PhysRevB.84.104207
J. Tian, H. Peng, X. Xu, et al., Catal. Sci. Technol. 5 (2015) 2270-2281. DOI:10.1039/C4CY01553C
X. Zhang, L. Peng, X. Fang, et al., Int. J. Hydrogen. Energ. 43 (2018) 8298-8312. DOI:10.1016/j.ijhydene.2018.03.086
F. Yu, X. Xu, H. Peng, et al., Appl. Catal. A:Gen. 507 (2015) 109-118. DOI:10.1016/j.apcata.2015.09.023
Z. Wang, H. Zhu, L. Ai, et al., J. Colloid Interface Sci. 478 (2016) 209-216. DOI:10.1016/j.jcis.2016.06.010
X. Xu, L. Li, J. Huang, et al., ACS Catal. 8 (2018) 8033-8045. DOI:10.1021/acscatal.8b01692
J. Xu, Y. Zhang, Y. Liu, et al., Eur. J. Inorg. Chem. 2019 (2019) 183-194. DOI:10.1002/ejic.201801250
Iwamoto, M. Lunsford, et al., J. Phys. Chem. 84 (1980) 3079-3084. DOI:10.1021/j100460a022
V. Ferreira, P. Tavares, J. Figueiredo, et al., Ind. Eng. Chem. Res. 51 (2012) 10535-10541. DOI:10.1021/ie3001953
L. Peng, J. Xu, X. Fang, et al., Eur. J. Inorg. Chem. (2018) 1787-1799.
Y. Hou, W. Han, W. Xia, et al., ACS Catal. 5 (2015) 1663-1674. DOI:10.1021/cs501733r
P. Huang, Y. Zhao, J. Zhang, Y. Zhu, et al., Nanoscale 5 (2013) 10844-10848. DOI:10.1039/c3nr03617k