Abstract

Non-Faradaic electrochemical modification of catalytic activity (NEMCA) with electric field applications in solid oxide cells (SOCs) is thought to be induced by spillover effects of lattice oxygen from the bulk, although the detailed mechanism has not still been clear. In SOCs, important phenomena such as fuel decomposition, charge transfer, etc. occur at the triple phase boundary (TPB) as a highly active site that consists of catalyst, electrolyte, and gas phases. NEMCA is expected to be also induced strongly by the surface mechanism on TPB, and understanding surface reactions on TPB is essential for improving catalyst and cell performances. However, the reliable TPB model has not been still uniquely defined to discuss the property of TPB although various studies have been reported. Therefore, in this study, we have focused on the TPB model comprising Ni catalyst cluster; YSZ electrolyte; and gas phase, and aimed to identify a reliable TPB model for theoretical studies by using first-principles calculations as an initial step. In concrete, we identified firstly the stable structures of YSZ surface models by using DFT calculations taking into account oxygen vacancy positions, yttrium atom arrangements, yttria concentration, and other factors. Thereafter, we discussed a reliable Ni/YSZ interface model based on the most stable YSZ model proposed above results by evaluating the Ni structure, interface stability, and so on.In this study, DFT calculations with a plane-wave basis set were implemented using CASTEP, and GGA-PBE exchange-correlation functional was used. The plane-wave cutoff energy was set as 489.8 eV, the OTFG-ultrasoft was used as the pseudopotentials, and the spin-polarization is considered because YSZ is a ferromagnetic substance. In YSZ surface models, yttria concentrations are set to 4.35 mol% and 9.1 mol% which shows the maximum ion conductivity of ZrO2. The three-layer YSZ (111) slabs with 15 Å vacuum layer with 2×2 and 2×4 unit cells were used for repeated slab models. The DFT+U method is used to obtain the correct electric structure of metal oxides with partially filled d or f-orbital shells, and k-points were set to 4×2×1. In Ni/YSZ interface models, we considered Ni cluster and Ni belt type models based on the most stable 2×4 YSZ surface model (9.1 mol%) proposed in this study. We also considered both cases of (111) and (100) facets for the contact interface. In the case of Ni/YSZ interface models, the DFT+U method was not considered to improve the calculation convergence and k-points were set to 4×2×1. A schematic diagram of the Ni/YSZ interface model is shown in the attached Figure.In the case of the 2×2 YSZ model with the yttria concentration of 9.1 mol%, the YSZ model is stabilized when oxygen vacancy is on the second O atom layer and the second neighbor to Y atoms, indicating that improvement in the geometry instability of ZrO2 for 8-coordination is more important than keeping local electron neutrality. We have also found that oxygen vacancy positions are more sensitive to the YSZ surface stability than Y atom arrangements. In the case of the 2×4 YSZ model with the yttria concentration of 4.35 mol%, the YSZ model where there are Y atoms on the first layer is stabilized. The crystal structure achieves a more stable structure by varying bond lengths when Y3+ with the larger ionic radius is replaced with Zr4+. Therefore, the structure is easier to stabilize when the Y atom exists on the surface than in the bulk due to the higher degree of freedom of the Y atom. The most stable structure of the 2×4 YSZ model with the yttria concentration of 9.1 mol% given based on the above results is 0.18 eV more stable than the previously reported structure [1]. This is because the number of Y atoms on the first layer with the second neighbor from oxygen vacancy is larger than the previously reported structure. We evaluated then the structural stability of the Ni/YSZ interface model based on the above YSZ model. As a result, we have found that adhesion energy between Ni and YSZ is independent of the relative position of Ni atoms. In addition, the larger the number of Ni atoms is, the more stable the structure is. This is because the electronic property of Ni atoms approaches the metal as increasing the number of Ni atoms. Other results and a detailed discussion will be reported in our meeting publications (ECS Transactions) and the presentation.[1] M. Shishkin and T. Ziegler, Phys. Chem. Chem. Phys., 16, 1798-1808 (2014). Figure 1

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call