Xiaochen Li and
Hongwei Gao*
Key Laboratory of Plant Resources and Chemistry in Arid Regions, Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Urumqi 830011, China. E-mail: gaohongw369@ms.xjb.ac.cn; Fax: +86-991-3858319; Tel: +86-991-3858319
First published on 27th March 2018
Lanthanum-based perovskite-type oxides represented by LaBO3 (B = Co, Fe, Mn) have been thought to present strong limitations for practical application although they are active for catalytic removal of NO. Cerium (Ce) substitution has been extensively studied to modify the properties of perovskites. It is noted that a new phase of ceria (CeO2) can be separated from perovskites when the doping ratio exceeds the solution limit (x > S). This review outlines the relationship between the existence of CeO2 phase and catalytic activity. CeO2 dispersing on the lattice surface or small particles are beneficial for catalytic activity, but larger particles are adverse. Ce-doped LaBO3 perovskites exhibiting the best activity must contain additional CeO2 phases. In addition, CeO2-supported LaBO3 perovskite catalysts are discussed.
From the thermodynamic view, the process (NO = 1/2N2 + 1/2O2; ΔG0f = −86 kJ mol−1)5 is favored, but it is kinetically hindered because of its high activation energy (364 kJ mol−1).6 To solve the afore-mentioned problem, a good deal of catalytic methods and catalysts have been studied. Currently, NO removal methods that we have learned can be roughly classified as two major technologies: NO decomposition and NO reduction. The former can be regarded as a “green” process since no additional toxic reactant, such as ammonia, is required and the NOx gases are decomposed into N2 and O2 as harmless components. Oppositely, the latter (e.g. NOx storage and reduction (NSR), selective catalytic reduction (SCR), etc.) usually requires the addition of a supplemental reductant which causes further problems, such as higher equipment cost and secondary pollution. At present, three principal types of catalysts active for NO removal are zeolites,7–9 noble metals10,11 and composite mixed oxides.12–16 However, the first, zeolites, are subject to hydrothermal deactivation and are only active at high temperature while the second, noble metals, are costly and easily agglomerated during the reaction. In contrast to the above two catalysts, composite mixed oxides, especially perovskite-type catalyst with the general formula of ABO3, have considered as a group of promising catalysts for NO removal as they are lower price, mixed valence states of the transition metals and higher stability.17–19
Lanthanum-based perovskites (LaBO3), some of which exhibit excellent performance for NO removal,20–22 have been extensively investigated. As far as we know, unmodified LaBO3 perovskites present strong limitations (e.g. comparatively small surface area and low catalytic activity at low temperatures)21,22 and therefore reliable technologies need to be developed. Among these technologies tackling aforementioned problems, we find that doping Ce into perovskite (e.g. La0.6Ce0.4CoO3) or CeO2 as a “support” (e.g. LaCoO3/CeO2) are effective. In addition, it have been reported that the solubility of Ce in LaBO3 perovskites is low and CeO2 appears in the mixed-oxides when Ce addition exceeds its solubility limit (S). The valence alternation ability of Ce3+/Ce4+ makes CeO2 a good oxygen reservoir which stores and releases oxygen under oxidizing and reducing conditions, respectively.23,24 In this review, CeO2 is focused because of its significant role in improving catalytic activity of LaBO3 perovskites catalysts. The roles played CeO2 by will be introduced by the following two topics: (i) CeO2 as separated phase (ii) CeO2 as the support.
Fig. 1 Crystal structure of the LaBO3 perovskite oxide.4 |
Ce often appears as a good promoter in perovskite lattice because of its two different valence states Ce4+ or Ce3+.4,34,35 The substitution of La3+ cation in LaBO3 with higher valence Ce4+ will affect indirectly the electronic state of the B-site ion. As mentioned above, Sr2+ are used to partially substitute the A-site ions in order to increase the basicity and to produce oxide anion vacancies by charge compensation. To get a better and clearer understanding about Ce doping, we take LaCoO3 as an example to explain the process of defect formation. Three reactions have been considered for Ce doping,36,37 the first two of which are for doping stoichiometric LaCoO3. The substitution of La3+ cation in LaCoO3 with Ce4+ leads to creation of La3+ vacancies (eqn (1)) or reduction of Co3+ to Co2+(eqn (2)).
(1) |
(2) |
However, as prepared LaCoO3 calcined at ≥900 °C is typically slightly reductively non-stoichiometric.38 Therefore, the third reaction that Ce is doped into the LaCoO3 lattice containing oxygen lattice vacancies was considered. This means that those oxygen vacancies were refilled.
(3) |
These defects caused by Ce doping lead to the fact that perovskite-type oxides show promising structural, surface and catalytic properties for catalyst application. In addition, supported lanthanum-based perovskite-type catalysts, e.g. LaBO3/CeO2, were also extensively studied.
For a low Ce content (which is below its solubility in perovskite structure), Ce can be incorporated into the LaBO3 lattice to form a solid solution. The valence change of B is complicated due to the addition of Ce. The substitution of La3+ in LaBO3 with Ce3+ result in no change of valence state of B ions (Bn+). The higher valence Ce4+ substitution lead to the fact that electroneutrality of the lattice is not maintained. As such, cation vacancies are formed and valence state of B ions is changed. No interstitial oxygen ions, however, are detected because perovskite lattice cannot accommodate them.39,40 For high Ce content (which exceed their solubility limit in perovskite structure), CeO2 are formed and segregated from perovskite oxides. The extra phase, CeO2, is either highly dispersed on the surface of Lal−xCexBO3 for comparatively low Ce content or found as larger particles for higher Ce content. The presence of CeO2 result in appearance of Ce vacancy sites or migration of (excess) B ions outside the perovskite lattice to yield ByOx or both.41 In this case, CeO2 are believed to play a key role in enhancement of catalytic performance.
Fig. 2 XRD patterns of the La1−xCexCoO3 perovskites (x = 0, 0.05, 0.10 and 0.15).46 |
La1−xCexCoO3 | ||||||
---|---|---|---|---|---|---|
x = 0 | x = 0.05 | x = 0.1 | x = 0.2 | x = 0.3 | x = 0.4 | |
SSA (m2 g−1) | 8.9 | 11.4 | 12.6 | 11.0 | 9.8 | 10.8 |
In addition to above-mentioned physical structures, the effect of Ce substitution on redox properties have also been discussed in this paper. Firstly, it has been reported that partial substitution with Ce can affect the surface characteristics which are always more important than lattice. The surface characteristics obtained from X-ray photoelectron spectroscopy (XPS) has been observed in the reports of Erdenee et al.43 and Wen et al.45 In their work, no change of La valence state (La3+) was detected when La was substituted with Ce. But they all agreed that the trivalent Co ion is partly changed. As has been reported by Wen et al.,45 when the doping ratio was below 0.1, tetravalent Ce ions were fully dissolved into the perovskite structure so that electronic unbalance was generated. As a result, a part of Co3+ in the B-site became Co2+ or La3+ vacancies were created as a charge compensation mechanism.36,37 When x was larger than 0.1, the average oxidation state of the Co cation increased because the existence of CeO2 might lead to deficiency of A-cation. Of course, they also showed XPS spectra of O 1s. According to ref. 43, 45 and 47, there are three different forms of oxygen including lattice oxygen (Olatt), adsorbed oxygen (Oads), and surface adsorbed water species. Wen et al.45 found that the La0.8Ce0.2CoO3 achieved the highest percentage of adsorption oxygen which participated in oxidation reactions. It is generally accepted that among Ce doped LaCoO3 samples, La0.8Ce0.2CoO3 shows the best activity.43,45 However, we cannot directly draw the conclusion that the improvement of catalytic activities are due to high amount of adsorbed oxygen, owing to the fact that although the perovskite LaCoO3 also gets high amount of adsorbed oxygen (Table 2), its activity is low. Additionally, it was proved that Ce ions were present in the trivalent and tetravalent form.43,46,47 The relative concentration of Ce3+, defined as Ce3+/(Ce3+ + Ce4+), was listed in Table 3. It varied from 19.2% to 21.0%, indicating that Ce4+ was dominant in all the samples. Secondly, a result of temperature programmed reduction of H2 (H2-TPR) experiment investigating effect of Ce addition on the redox ability of LaCoO3 is shown in Fig. 3. The literature46 also displayed the H2-TPR profiles of La1−xCexCoO3, in which two obvious reduction peaks were observed, which could be attributed to the reduction of Co3+ to Co2+ and Co2+ to Co0, respectively. As was seen from Fig. 3, two reduction peaks respectively located at about 400 °C and 600 °C. When a certain amount of Ce was introduced into the LaCoO3 perovskite structure, the two peaks moved forward to lower temperature direction correspondingly, indicating that Ce4+ insertion increased the catalyst reducibility. In other word, Ce-doped catalysts showed a better reducibility as the reduction temperatures of these catalysts were lower than pure LaCoO3. And the sample with La1−xCexCoO3 (x = 0.2) leads to the highest decrease in the reduction peak temperatures. This may be because (1) it creates an easier reducibility of the Co3+ into Co2+ and, (2) Ce4+ increases the number of cation vacancies within the lattice.
La1−xCexCoO3 | ||||||
---|---|---|---|---|---|---|
x = 0 | x = 0.05 | x = 0.1 | x = 0.2 | x = 0.3 | x = 0.4 | |
% | 46.6 | 42.2 | 43.5 | 48.4 | 39.8 | 38.3 |
La1−xCexCoO3 | |||||
---|---|---|---|---|---|
x = 0 | x = 0.05 | x = 0.1 | x = 0.3 | x = 0.5 | |
Ce3+/(Ce3+ + Ce4+) (%) | — | 20.6 | 21.0 | 20.3 | 19.2 |
Fig. 3 H2-TPR profiles of the La1−xCexCoO3 (0 ≤ x ≤ 0.4) catalysts.43 |
According to Wen's paper,45 the La1−xCexCoO3 samples synthesized through a citrate show the highest conversion (80%) at about 300 °C in a flow of NO + O2 + He when x was 0.2. However, the catalytic activities decreased at higher x values. They thought that these factors including the presence of CeO2 and the decrease of adsorbed oxygen amount caused the decrease of activity. Combined with his results of XRD, when x > 0.1, the peaks of CeO2 were observed. This means that small CeO2 appeared in the mixed-oxide. Thus, it is concluded that the existence of small CeO2 nano-particles is beneficial to the improvement of catalytic activities, but catalytic activities decreased because larger particles are formed at higher Ce content.
Giannakas et al.50 investigated and compared the catalytic performance of LaFeO3, La0.85Sr0.15FeO3, La0.8Sr0.1Ce0.1FeO3 and La0.8Ce0.2FeO3 which were prepared via a reverse micelles microemulsion route for NO reduction by CO, finding that the full sequence of catalytic activity of tested solids is La0.8Ce0.2FeO3 > La0.8Sr0.1Ce0.1FeO3 > La0.85Sr0.15FeO3 ≥ LaFeO3. They pointed that (1) the above sequence is in accordance with the sequence of increment of ssa of the solids and, (2) La0.8Ce0.2FeO3 and La0.8Sr0.1Ce0.1FeO3 show much higher catalytic activity than the solids containing no Ce, which may be caused by the presence of CeO2 phase in these two solids. Qin et al.51 also reported that catalytic reaction of La1−xCexFeO3 for NO reduction with CO and illustrated the influence of Ce substitution on SO2 resistance. It was found that although LaFeO3 exhibited an excellent catalytic performance without SO2 (100% NO conversion at 500 °C), catalytic reaction decreased drastically when SO2 gas was added to the CO + NO system(conversion rate decreased to 40–50% after approximately 100 min). Doping a certain amount of Ce into the LaFeO3 perovskite structure could obviously improve the SO2 resistance. For example, La0.6Ce0.4FeO3 sample maintained a conversion rate of 80% during the latter 270 min after addition of SO2 gas into the CO + NO reaction system. The reason was suggested to be the addition of Ce4+ which could prevent the adverse effect of SO2 on the LaFeO3 perovskite by absorbing SO2 and forming Ce(SO4)2. Thus, the presence of Ce4+ is the key factor to the improvement of SO2 resistance in LaFeO3 perovskite catalysts.
Recently, attapulgite (ATP) was considered as support for La1−xCexFeO3 catalysts, not only because of its inexpensive but also for its adsorption of many pollutants. A recent research52 on La1−xCexFeO3/ATP was conducted for testing its photocatalytic reduction of NO at low temperature. Following investigation on photo-SCR activity and stability of La1−xCexFeO3/ATP samples (Fig. 4) proves that the substitution of Ce is helpful for the improvement of catalytic performance. It is noteworthy that the conversion rate of NO reaches close to 80% when Ce doping amount is 0.3 at room temperature. Combined with XRD results, CeO2 is precipitated from the solid solution when the Ce doping ratio is 0.3. This implies that the existence of CeO2 is no harm for NOx reduction.
Fig. 4 Photo-SCR denitrification activity and stability of La1−xCexFeO3/ATP (x = 0.1–0.5).52 |
Besides the NO removal, La1−xCexFeO3 was also considered as promising catalyst in some reactions. For example, Xiang et al.49 reported that the temperature of CH4 conversion at 90% was as low as 510 °C when x = 0.3. According to Ma's report,53 in the case of x = 0.5, the conversion of CO is about 68% at 530 °C.
For catalysts, the effect of the substitution of La by Ce in these catalysts is more complex than that in La1−xCexMnO3, since the Ce cation not only changes the reducibility of Mn cation (Mn3+ ↔ Mn4+), but also may affect the oxidation state of the B′ cation. In the case of , the redox couple can be Ce4+/Ce3+, Mn4+/Mn3+and B′n+/B′(n−1)+ since oxidation state of B′ cation may be variable. According to He's paper,57 the reaction (Ce3+ + Cu2+ → Ce4+ + Cu+) taken place during the preparation process of the La0.8Ce0.2Mn0.6Cu0.4O3 catalysts. The presence of Cu+ and its redox equilibrium (Cu+ ↔ Cu2+) would facilitate the NO adsorption and reduction by CO. Similarly, Tarjomannejad et al.58 studied the LaMn1−xFexO3 and La0.8Ce0.2Mn0.3Fe0.7O3 perovskites catalysts, finding that the substitution of La3+ by Ce4+ result in decrease of Mn4+/Mn3+ and Fe4+/Fe3+ratios. Of course, the substituted samples can also lead to the formation of oxygen vacancies and the increase of Oads/Olatt ratio. An integration of all these factors resulted in the enhancement of catalytic activity (90% NO conversion at 323 °C). In addition, if the B′ cation exists in a permanent valence state, Ce4+/Ce3+and Mn4+/Mn3+ are the redox couples. The role of Ce is just to adjust the valence state of Mn cation or/and the ratio of Oads/Olatt. However, the catalytic performance of samples have no relation to whether the valence state is varied or not. Sometimes the B′ cation with a varied valence state can promote the catalytic reaction,57,58 sometimes inhibit. In the literature,59 the substitution of Mn by Co in La0.8Ce0.2MnO3 catalysts results in decreasing catalyst activity.
Besides, it has been reported that preparation methods interfere with the generation of CeO2 phase. In the paper,57 a considerable amount of CeO2 was found by XRD in the La0.8Ce0.2Cu0.4Mn0.6O3 and La0.8Ce0.2Ag0.4Mn0.6O3 samples synthesized by sol–gel method, but not in samples synthesized by the reverse microemulsion method.
Another factor for catalytic performance of supported samples is preparation method. H2-TPR experiment has been done to investigate the structure and reducibility of the catalysts prepared by two methods.66 The result shown in Fig. 5 indicated that preparation method had relation to the amount of H2 consumed, and it was suggested that LaMnO3/CeO2 prepared by the dry impregnation (DI) method showed higher activity than that by the precipitation–deposition (PD) method. In order to find the reason, the catalysts were characterized by XRD, BET and EXAFS. In the result of XRD, the peaks corresponding to the perovskite were detected for LaMnO3/CeO2-PD catalysts but hardly detected for the LaMnO3/CeO2-DI catalysts, indicating that LaMnO3 perovskite phases were highly dispersed on the CeO2 support. The BET surface areas of LaMnO3/CeO2-DI (46 m2 g−1) was lower than that of LaMnO3/CeO2-PD (68 m2 g−1). The formation of a perovskite oxide phase for the LaMnO3/CeO2 catalysts was confirmed by the EXAFS studies. Combined with the H2-TPR study results, they propose that the higher activity may be ascribed to the strong interaction between the perovskite phase and the CeO2 support although this catalyst has smaller surface area. Oppositely, Alifanti et al.67 concluded that there was no interaction between the support and the elements of the perovskite.
Fig. 5 H2-TPR profiles of LaMnO3/CeO2 catalysts and LaMnO3.66 |
In addition, the morphology of CeO2 has also a great influence on the catalytic activity, which has been studied by Wang et al.68 It was found that the supported catalysts were more active than the unsupported catalyst (La0.8Ce0.2MnO3) and La0.8Ce0.2MnO3/CeO2 nanopolyhedra exhibited the higher catalytic activity than the other two morphologies of CeO2 supported samples. The results described above are possible due to higher specific surface area and larger number of oxygen vacancies.
In summary, the modification method of perovskite oxides is effective for improving their catalytic activity, even though the promotion mechanism is still not clearly explained. Such a supported perovskite-type catalysts can have wide application prospects for the possibility of increasing its activity by simple preparation methods.
For LaBO3/CeO2 catalysts, promotion mechanism is still not clearly explained. However, comparing with CeO2 support or pure LaBO3 perovskite catalyst, the catalytic activities of the majority of supported catalysts are enhanced. The simple preparation methods provide a feasible scheme for improvement of catalytic activity of LaBO3 perovskites.
This journal is © The Royal Society of Chemistry 2018 |