In situ investigation of ruthenium doped lanthanum nickel titanium double perovskite and its exsolution behaviour

Exsolution, an innovative method for fabricating perovskite-based oxides decorated with metal nanoparticles, has garnered significant interest in the fields of catalyst fabrication and electrochemical devices. Although dopant exsolution from single perovskite structures has been extensively studied, the exsolution behaviour of double perovskite structures remains insufficiently understood. In this study, we synthesized B-site double perovskite Ru-doped lanthanum nickel titanates with a 7.5 at% A-site deficiency, and systematically investigated the exsolution process that formed nickel metal nanoparticles on the material surface, across a broad reduction temperature range of 350–1000 °C. Both Ex situ and in situ characterization revealed that small, uniform Ni nanoparticles exsolved at low temperatures, whereas the exsolution of ruthenium required higher reduction temperatures beyond 1000 °C. Within the reduction temperature range of 350–500 °C, a notable finding is the reconstruction of exsolved nanoparticles, implying that Ni particles exist in a thermodynamically metastable state. Electrochemical impedance spectroscopy (EIS) showed a decreased area specific resistance (ASR) during the progress of exsolution. The increase in current density of a full solid oxide cell (SOC) in electrolysis mode and the doubling of peak power density in fuel cell mode attributed to the exsolution of Ni nanoparticles highlight the potential application of metal exsolution in electrode materials for SOCs.


Figures
Electronic Supplementary Material (ESI) for Nanoscale Advances.This journal is © The Royal Society of Chemistry 2024            Three different reducing atmospheres were utilized in this work: wet 5%H2/Ar, dry 5%H2/Ar and high vacuum in microscope.The high vacuum condition in transmission electron microscope was considered as 1.3×10 -10 bar.The partial pressure of oxygen and hydrogen content were evaluated as 2.73×10 -11 bar and 7.8×10 -18 bar.
For the hydrogen gas, ESS quadrupole mass spectroscopy 4 was also used to confirm the difference between wet and dry 5%H2/Ar: Table S.5 mole fraction of oxygen, hydrogen and water calculated based the ionic current measured by quadrupole mass spectroscopy.
From Table S.5, the oxygen and hydrogen mole fraction of wet and dry 5%H2/Ar are similar to each other.Hence, it's comparable to correlate the in-situ observation of mass loss in dry 5%H2/Ar and the ex-situ observation after reduction in wet 5%H2/Ar.
Table S. 6 Fitted parameters of the components in the equivalent circuits of the spectrum measured after reduction at 800°C in 5%H2/Ar.

Equations
For double perovskite with two elemental cations occupying B-site positions, the ordering of B-cations was assessed through the degree of ordering, S 5 : where gB represents the correct occupancy of Ni or Ti-Ru cations at either the (0, ½, 0) or (½, 0, 0) positions in this work.
For thermogravimetric analysis, Amaya-Dueñas et al. 6 characterized the Ni exsolution from La(Sr)Cr0.85Ni0.15O3−δwith TGA in 5%H2/Ar and calculated the oxygen deficiency, δ.The exsolution and decomposition processes can be assumed as a subsequent reaction after oxygen loss.Consequently, according to the method described, the loss of oxygen can be derived using Equation S2 below: where ML1.85NTR is the molecular weight of L1.85NTR.Δm is the measured mass loss from TG curve.AO is the atomic mass of oxygen and lO is the loss of oxygen.Based on the Rietveld refinement result, if we assume the molecular formula of L1.85NTR as La1.888NiTi0.9Ru0.1O5.832, the lO of the sample reduced at 800°C is approximately 0.25 mol per mol of L1.85NTR.

Figure S. 2
Figure S.2 XRD pattern and the Rietveld refinement profiles of L1.85NTR.The factors of agreement are Rp = 4.95, Rwp = 6.41,Rexp = 4.11 and χ 2 = 2.43.Two inset plots are the zoomed pattens of unresolved peaks.The green arrows mark the Cu Kβ reflections.Two subtle features were located at 33.2° and 62.5°, remaining unresolved.However, given the main diffraction signals strongly correspond to lanthanum nickel titanate double perovskite, the product was considered as majority single phase and adopted in the following study.

Figure S. 4
Figure S. 4 XRD patterns of L1.85NTR sample reduced at 800°, with 2 theta positions ranging from 28.2° to 30.9° highlighting the main peaks of La2O3 (a), and from 43.5° to 48.5° highlighting the main peaks of Ni metal phase (b).

Figure S. 5
Figure S.5 SEM images of L1.85NTR sample reduced at 800°C for the particle analysis.

Figure S. 7
Figure S. 7 TG curve of stoichiometric sample L2NTR and corresponding temperature profile as a function of time in 5%H2 atmosphere.The mass loss behaviour exhibited a two-step mass loss,

Figure S. 8
Figure S.8 Schematic illustration of the thermal cycle applied to the samples.

Figure
Figure S.11 (a) Schematic illustration of the testing rig for the EIS measurement; (b,c) SEM images of the microstructure of the L1.85NTR/LSGM/L1.85NTRsymmetric cell.

4
Rietveld refined lattice parameter and atomic parameters of double perovskite phase in the reduced L1.85NR sample based on reference structure ICSD186433 3 .

Table S .
1 Goldschmidt tolerance factor and Le Bail refined lattice parameters of L2NT and L2NTR.Table S.2 ICP results of stoichiometric sample L2NT and L2NTR and A-site deficient sample L1.85NTR.
B-site positions were assumed fully occupied by nickel, titanium, and ruthenium cations.Oxygen concentration is calculated based on experimental results of metal cations.Table S.3 Rietveld refined lattice parameters and atomic parameters of L1.85NTR