Exsolution trends and co-segregation aspects of self-grown catalyst nanoparticles in perovskites

In perovskites, exsolution of transition metals has been proposed as a smart catalyst design for energy applications. Although there exist transition metals with superior catalytic activity, they are limited by their ability to exsolve under a reducing environment. When a doping element is present in the perovskite, it is often observed that the surface segregation of the doping element is changed by oxygen vacancies. However, the mechanism of co-segregation of doping element with oxygen vacancies is still an open question. Here we report trends in the exsolution of transition metal (Mn, Co, Ni and Fe) on the PrBaMn2O5+δ layered perovskite oxide related to the co-segregation energy. Transmission electron microscopic observations show that easily reducible cations (Mn, Co and Ni) are exsolved from the perovskite depending on the transition metal-perovskite reducibility. In addition, using density functional calculations we reveal that co-segregation of B-site dopant and oxygen vacancies plays a central role in the exsolution.

cations, serving as the driving force for steering fast nanoparticle growth (Nano Lett., 2016, 16 (8), pp 5303-5309). This previous work is not mentioned in the manuscript. 3. The occurrence of metal or oxide nanoparticle as consequence of the exsolution is essentially demonstrated in this work by using X-ray diffraction and transmission electron microscopy with EDS analysis. TEM analysis provides a local information involving just a few nanoparticles. However, there is no proper analysis of the surface modifications on a statistical basis after the exolution in terms of composition and oxidation states. On the other hand, a quantitative determination of the surface atomic composition and oxidation states may be provided by X-ray photoelectron spectroscopy analysis. The catalytic activity is essentially related the surface characteristics since reactions occur at the interface; thus, I strongly recommend a surface analysis e.g. by XPS for this work. 4. The aim of using catalysts made of perovskite with exsolved transition metal on the surface in SOFCs is to provide an enhanced activity, being this promoted by the metal nanoparticles on the surface, while keeping good stability towards redox and thermal cycles. The performance of hydrogen-fed SOFCs for the perovskites containing exsolved transition metals and oxides on the surface was investigated but no attempt was made to demonstrate the stability under redox and thermal cycles. In this regard, no obvious advantage over perovskites decorated with transition metals is observed. 5. Moreover, no appropriate durability studies have been made to show the stability of the nanoparticles on the perovskite surface after prolonged operation. 6. Another important aspect is dealing with the use of perovskites anodes for the direct utilisation of hydrocarbons in SOFCs. It would be important to investigate if the synthesised materials can operate in dry methane, as a model organic fuel, and a comparison with relevant literature reports should be made to show if there is any relevant progress beyond the state of the art. In particular, perovskites containing exsolved transition metals have shown already promising capability as an active and stable electrode for SOFCs in various fuels (Nano Lett., 2016, 16 (8), pp 5303-5309) but the comparison should be extended to other perovskites decorated with metal particles or stabilised by thermal treatments for application as anodes in SOFC fed with dry hydrocarbons. 7. In general, the exsolution of transition metals from perovskite backbones in reducing atmosphere has been widely studied in the literature. The authors have mentioned some previous works in this field but relevant literature dealing with anode-based perovskites for SOFCs, including stabilisation procedures and decoration of perovskites with transition metal nanoparticle for application as SOFC anodes is not covered.

Reviewer #3 (Remarks to the Author):
This report shows good performance and nice nanoparticle production. The decomposition of these layered phases is novel although the nature of the exsolution could be better described The term periodic seems to mean following the periodic table rather than something much more exciting that I was anticipating. This periodicity is not that surprising and could be more or less predicted from standard free energy considerations, perhaps iron is a little out of step. I am not too sure how much the dft adds to this. It would be better if the exsolution was from a single phase rather than from a hexagonal /cubic perovskite mix.Which phase yields the metals. The comparison lacks some more competitive papers, more care could have been taken.
General Comments: This manuscript discusses the "periodicity" in the exsolution of transition metal nanoparticle on the layered perovskite oxide and the co-segregation energy. While the first concept is elusive, the second one has been proposed by others. While I believe that the study provides significant new insight on the exsolution process, there are some critical issues need to be considered. I list them here: Comment 1. At line 77, the authors claim there isn't any secondary phase, however the hexagonal phase is indeed a secondary phase.

Response to C1:
The crystal chemistry of these perovskite oxides is in fact complex due to the formation of hexagonal BaMnO 3 -type perovskites when synthesis is carried out in air. Raveau et al. [J. Phys.: Condens. Matter, 14, 1297(2002] and Sengodan et al. [Nat. Mater., 14, 205 (2015)] have reported that the R x Ba 1-x MnO 3 (R=La, Pr and Nd) compound is generally a mixture of two main phases namely hexagonal BaMnO 3 and cubic perovskites.
This hexagonal perovskite phase increases with increasing the Ba content. Hence, hexagonal perovskite phase is the characteristic phase of Pr 0.5 Ba 0.5 MnO 3 and is not a secondary phase (impurity phase).

Comment 2.
At line 88, it is easy to understand that MnO, Co, and Ni are more easily exsolved than Fe. However, it is not clear what the authors mean by "periodicity" as described in the following sentence.
Response to C2: Thanks for the comment. In the context of chemistry, periodicity is defined as "trends or recurring variations in element properties". In our manuscript, the periodicity means trend of exsolution among the B-site doping elements (Mn, Co, Ni and Fe). To avoid the misunderstanding, the text has been modified appropriately to "These results clearly show that MnO, Co, and Ni are more easily exsolved to form nanoparticles than Fe, and thus B-site transition metals show some trend to exsolve in the layered perovskite oxide." Comment 3. At line 93 and Figure 2, the SEM image after exsolution shows not only the exsolution of the nanoparticles but also a flatter surface. The small clusters of the host material also vanish after exsolution. This is not in agreement with the authors' comment that the surfaces are similar. Explanations should be provided.

Response to C3:
In the SEM image after exsolution, the formation of flatter surface and removal of the small clusters from the host material are due to the phase transition of simple perovskite to layered perovskite. We did not claim that the surface of simple perovskite (before exsolution) and layered perovskite (after exsolution) are similar. In contrast, we claim  Nonetheless, if necessary, we will further examine the segregation energies on the other promising surfaces for B-metal exsolution based on our XRD data. Response to C6: Thanks for your consideration. The DFT total energy calculations intrinsically consider the columbic repulsion between charged defects and oxygen vacancy in the system. Therefore, by comparing the segregation energy of Pr or Ba compared to Co, we can provide more information for the competitive formation of PrO or BaO in addition to that of BO x . In addition, no BaO formation was detected in experimental results (XRD and TEM). Response to C1: We thank reviewer #2 for generally supporting our paper. Comment 4. The aim of using catalysts made of perovskite with exsolved transition metal on the surface in SOFCs is to provide an enhanced activity, being this promoted by the metal nanoparticles on the surface, while keeping good stability towards redox and thermal cycles.
The performance of hydrogen-fed SOFCs for the perovskites containing exsolved transition metals and oxides on the surface was investigated but no attempt was made to demonstrate the stability under redox and thermal cycles. In this regard, no obvious advantage over perovskites decorated with transition metals is observed. Comment 5. Moreover, no appropriate durability studies have been made to show the stability of the nanoparticles on the perovskite surface after prolonged operation.

Response to C5:
In this work, we focused on phenomenon in terms of periodicity of exsolution. In addition, PrBaMn 1.7 T 0.3 O 5+δ (T= Co and Ni) perovskite anodes decorated with metal nanoparticles show high performance in H 2 fuel compared to other ceramic anodes.
However, we will include the stability test in H 2 and C 3 H 8 at 700 o C for further consideration.  (Nano Lett., 2016, 16 (8), pp 5303-5309) but the comparison should be extended to other perovskites decorated with metal particles or stabilised by thermal treatments for application as anodes in SOFC fed with dry hydrocarbons.