The effect of NiO on the conductivity of BaZr_0.5Ce_0.3Y_0.2O_3−δ based electrolytes

Abstract

The effects of NiO on the sintering behaviors, morphologies and conductivities of BaZr_0.5Ce_0.3Y_0.2O_3−δ (BZCY532) based electrolytes were systematically investigated. 1 wt% NiO powder was added by different methods during the sample preparation: (i) added during ball-milling before a powder mixture calcination (named BZCY(Ni)532), (ii) no NiO addition in the whole preparation procedure (named BZCY532) and (iii) added after a powder mixture calcination (named BZCY532(Ni)). The conductivities of these three kinds of dense BZCY532 ceramics were investigated in dry air, wet N₂ and wet H₂ atmospheres, respectively. Moreover, the electronic contributions to the total conductivities were also identified in a broad oxygen partial pressure range. According to the achieved results, it can be concluded that the dense BZCY(Ni)532 ceramics showed the best enhanced oxygen and proton conductivities, followed by the BZCY532(Ni) and BZCY532 ceramics. Furthermore, the BZCY(Ni)532 and BZCY532 ceramics showed a tiny electronic conductivity, when the testing temperatures were lower than 800 °C. However, the BZCY532(Ni) ceramics revealed an obvious electronic conduction when they were tested at temperatures of 600–800 °C. Therefore, it is preferable to add the NiO during powder preparation, which can lower the sintering temperature and also increase the conductivity of BZCY532-based electrolytes.

---

Introduction

Barium zirconate and Barium cerate solid solution materials (BaZr_xCe_yY_zO_3−δ, x + y + z = 1) represent one type of promising proton-conducting materials to be used for Intermediate Temperature Solid Oxide Fuel Cells (ITSOFCs).^1–11 Despite these materials having a relatively higher conductivity than yttria-stabilized zirconia (YSZ) based electrolytes, their sintering temperatures are always high up to a temperature of 1550 °C or even higher. This is also accompanied by a long sintering time in order to get densified ceramics.^12,13 However, these severe sintering parameters, especially the high sintering temperature and the long sintering time, will unavoidably lead to several detrimental effects in the final ITSOFC applications. For example, the conductivities of these materials will be reduced due to a barium (Ba) loss, when the sintering temperature is higher than 1500 °C.^12,14–16 Furthermore, the porous structure of the electrodes will be collapsed and their catalytic performance will be reduced, if the electrodes and the electrolyte are co-fired.^17,18 This is detrimental for the cell structure design and for practical mass production. Therefore, various efforts have been done during the recent years to reduce the sintering temperatures and to increase their relative densities as well as the more important proton conductivity property. Among these studies, the addition of sintering aids and the introduction of wet-chemical synthesis methods to prepare ultra-fine powder represent two typical investigation routes.^19,20

Definitely, the sintering temperature can be reduced after the ultra-fine powder has been synthesized by the wet-chemical synthesis method. This procedure always results in a reduced sintering temperature and an improved conductivity. However, still some disadvantages exist. The biggest disadvantage of most wet-chemical synthesis methods is their complex synthesis procedures and the corresponding long synthesis time compared to the widely used solid-state reaction methods. In terms of this, the addition of sintering aids is considered as a much simpler way to realize the reduction of the sintering temperature and to increase the relative density. In previous studies, various transitional metal oxides were used as sintering aids in order to improve the sintering of BaZr_xCe_yY_zO_3−δ based ceramics.^{17,18,21–27} These results indicate that NiO and ZnO are effective sintering aids to improve the sintering of BaZrO₃–BaCeO₃ solid solution materials. Moreover, it was identified that a ZnO addition will not introduce an unfavorable electronic conduction.²¹ However, there exists few research works on the effect of NiO on the total conductivity and especially on electronic conduction. Furthermore, some controversial results with respect to the conductivity property have been reported after NiO addition.^28,29 In 2010, Tong et al. successfully prepared dense Y-doped BaZrO₃ and BaCeO₃ proton conductors through cost-effective solid-state reactive sintering (SSRC) method and using NiO as sintering aid.^30,31 In their studies, the total conductivities of prepared BaZr_0.8Y_0.2O_3−δ conductors are high up to 2.2 × 10⁻² S cm⁻¹ and 3.3 × 10⁻² S cm⁻¹ when they were tested at 600 °C under dry- and wet-argon atmospheres respectively. This high conductivity showed very high commercial potential. After that, the SSRS sintering mechanism was systematically studies in their following research.^27,32 However, they did not carry out a systematic study of the NiO effect on their total conductivity, especially the unfavorable electronic conductivity in severe reducing atmosphere. But, this is an urgent issue should be answered before the materials to be used in commercial products. Combined all of these mentioned above, it is necessary to conduct a systematic study to investigate the effect of NiO additions on the conductivities of BaZr_xCe_yY_zO_3−δ based ceramics and to explore the internal reaction mechanism between NiO and the main electrolyte components.

In this work, the BaZr_0.5Ce_0.3Y_0.2O_3−δ (BZCY532) compound, one of the most BaZr_xCe_yY_zO_3−δ proton-conducting materials that can be used for ITSOFCs, was selected for this investigation. 1 wt% NiO powder was used and it was added by different ways. More specifically, (i) the NiO was added before the powder mixture calcination (added during powder preparation, named BZCY(Ni)532), (ii) after the powder mixture calcination (after powder mixture calcination and then added into the calcined powder, named BZCY532(Ni)) and (iii) without addition (named BZCY532). The effects of NiO on the sintering behaviors, morphologies and conductivities as well as the potential electronic conduction were systematically investigated.

Experimental

Powder preparation

An improved solid-state reaction method, which included a water-based milling followed by a freeze drying process, was applied for the powder preparation.^33–35 Take BZCY(Ni)532 for example, stoichiometric amounts of raw materials BaCO₃, ZrO₂, CeO₂, Y₂O₃ and NiO (≥99.9% purity, Alfa Aesar GmbH & Co KG, Germany) were ball-milled for 5 h. After that, the powder mixture slurry was dried by freeze drying. Thereafter, the dried powder mixture was calcined at a temperature of 1300 °C and dwelled for 10 h. Finally, the prepared BZCY(Ni)532 powder was ball-milled and dried again before a powder compaction. The powder synthesis procedure for the BZCY532 compound is same as was used for the BZCY(Ni)532 compound, but without the addition of NiO during the process. As for BZCY532(Ni) powder, pure BZCY532 powder was firstly prepared and then the NiO powder was added during the second ball-milling procedure.

Green body preparation and sintering

After these three kinds of powder preparation, all powders were firstly compacted using a uniaxial pressure of 20 MPa to make sure that the green body became an integrated bulk pellet. Following that, the pellets were sealed in a plastic bag and then compacted at a high pressure of 250 MPa by using the cold-isostatic press and using a holding time of 10 min. Finally, the green bodies of the BZCY(Ni)532, BZCY532 and BZCY532(Ni) pellets were obtained.

The prepared green bodies were placed in boat-shaped alumina crucibles, which were surrounded by the powder with the same composition with respect to the buried pellets. After that, these green bodies were firstly heated to a temperature of 600 °C with a heating rate of 1 °C min⁻¹ and held for 2 h, to remove the inside organic binders and moisture. Following that, the temperature was increased up to higher sintering temperatures of 1300–1600 °C with a heating rate of 5 °C min⁻¹. The sintering time was kept constant of 10 h for the sintering of the BZCY(Ni)532 and BZCY532(Ni) pellets. Furthermore, a constant sintering time of 24 h was applied during the sintering of the BZCY532 pellets.

Characterizations

The X-Ray Diffraction (XRD, Philips X'pert X-ray diffractometer equipped with a graphite monochromatized Cu Kα radiation source λ = 1.540598 Å, PANalytical B.V., Netherland) was carried out to determine the phase purity and the crystal structure. Scanning Electron Microscopy (SEM, JSM-7000F, JEOL Ltd, Japan) was used for the morphology determination of the sintered samples. Moreover, a corresponding energy-dispersive X-ray spectroscopy (EDS or EDX, Oxford Instruments combined with an INCA software, Oxford Instruments, United Kingdom) analysis was done for the element analysis of the sintered ceramics. An electronic balance (Sartorius BSA224S equipped with an YDK01-C accessory, Sartorius AG, Germany) was used to measure the density. In addition, the density measurements were carried out at room temperature and using deionized water as the testing medium. Also, the density calculations were done automatically after the water density calibration.

Prior to the conductivity tests, the pellet surfaces were polished. Also, it was made sure that the up and down surfaces were parallel. Following that, an Au paste was coated on both the upper and lower surfaces. Then, the Au coated pellets were heated to a temperature of 850 °C using a heating rate of 1 °C min⁻¹ and held for 30 min. Their conductivities at different testing atmospheres and at different oxygen partial pressures were measured by an electrochemical testing system (MinisTest6000S platform, Toyo Corporation, Japan). More specifically, the conductivities in the atmospheres of dry air, wet N₂ and H₂ were tested at temperatures of 800–200 °C. A frequencies range of 0.1–100 kHz and an excitation voltage of 100 mV were employed. The setting of different oxygen partial pressures was controlled and realized automatically by using the gas mixtures of N₂–O₂ and N₂–H₂O–H₂ to control the oxygen partial pressure in a range of 1 to 10⁻² atm and 10⁻³ to 10⁻²⁴ atm, respectively.

Results and discussion

XRD studies on prepared powders and sintered pellets

Initially, the prepared BZCY(Ni)532 and BZCY532 powders were characterized by XRD, in order to check whether phase changes occur for the prepared powders and the sintered pellets. As there are organic binders in the prepared BZCY532(Ni) powder, it was not characterized by XRD. As shown in Fig. 1(a), the prepared BZCY(Ni)532 and BZCY532 powders were well-crystallized, had a cubic perovskite structure, and belonged to the Pm3̄m space group. However, small amounts of impurities in the BZCY(Ni)532 powder could be detected in the XRD patterns. These presented impurities are connected to the addition of NiO. The component of these impurities most probably belong to the BaY₂NiO₅ (BYN) compound. This was confirmed by Tong et al.'s previous study of BaZr_0.8Y_0.2O_3−δ (BZCY802)^31,32 and BaCe_0.8Y_0.2O_3−δ (BZCY082)³⁰ as well as from our previous studies of BZCY532,^34,36 when using the solid-state reactive sintering method to prepare barium zirconate-cerate based proton conductors and using an addition of NiO as a sintering aid. The XRD pattern of the BYN compound shows a typical peak at the 2θ diffraction angle of 32.5 ± 0.2 (Fig. 1(b)). Therefore, the XRD patterns in a 2θ angle range of 32.0–32.8 was magnified to determine this phase change when the sintering temperatures were changed, see Fig. 2. The BYN phase was also detected for the sintered BZCY(Ni)532 pellets after they had been sintered at a temperature of 1300 °C and dwelled for 10 h (Fig. 2(a)). Even if the sintering temperature was increased up to 1400 °C and 1500 °C, the BYN phase still existed in the sintered BZCY(Ni)532 pellets (Fig. 2(b) and (c)). However, this BYN phase disappeared when the sintering temperature was increased up to 1600 °C (Fig. 2(d)). This behavior was both consistent with Tong et al.'s previous results and our previous results.^{30–32,34,36} This small amount of BYN compound should be decomposed at the temperature of 1600 °C and merged into the main BZCY532 lattice with a Ni substitution at the location of B-site (Zr or Ce).

Fig. 1 (a) Represents the XRD patterns of the prepared BZCY(Ni)532 and BZCY532 powders, and (b) shows the magnified parts in a 2θ angle range of 32.0–32.8.

Fig. 2 XRD patterns of sintered BZCY(Ni)532, BZCY532 and BZCY532(Ni) pellets when using the following sintering temperatures: (a) 1300 °C, (b) 1400 °C, (c) 1500 °C and (d) 1600 °C.

It seems that the phase change regularity at the 2θ angle range of 32.0–32.8 for the sintered BZCY532(Ni) pellets is almost same as sintered BZCY(Ni)532 pellets. As the NiO powder was added after that a powder mixture calcination had been made, it should be more difficult to form BYN compound during the BZCY532(Ni) green body sintering compared to the BZCY(Ni)532 sintering. Therefore, the existence form of Ni element in the sintered BZCY532(Ni) pellets might be a phase mixture of NiO and BYN when they were sintered at temperatures of 1300–1500 °C. Even if no peaks can be detected within the 2θ angle range of 32.0–32.8 when the BZCY532(Ni) pellets were sintered at a temperatures of 1600 °C, tiny NiO particles still existed and they can't be detected and reflected from the obtained XRD patterns due to the limitations of XRD detection. This can be confirmed in the following EDS and conductivity measurements. Specifically, the sintered BZCY532(Ni) pellets have an obvious electronic conduction in the temperature range of 600–800 °C. This phenomena should be induced by the remnant NiO, even if the total NiO addition is only 1 wt% and even if only part of it was merged into the main BZCY532 lattice.

Relative density and microstructure analysis of the sintered pellets

Usually, a higher sintering temperature will lead to a higher relative density when performing a ceramic sintering. However, this rule is only true for the pure BZCY532 ceramics in this study, but not for the BZCY(Ni)532 and BZCY532(Ni) ceramics (Fig. 3). More specifically, the relative densities of the sintered BZCY532 pellets show a linear increase with increasing sintering temperatures. Due to their intrinsic refractory property, a sintering temperature of 1600 °C is necessary to obtain dense BZCY532 ceramics. Even so, the relative densities of the sintered BZCY532 pellets with a sintering temperature of 1600 °C is still lower than the relative densities of sintered BZCY(Ni)532 and BZCY532(Ni) pellets at the same sintering temperature. In regards to sintered BZCY(Ni)532 and BZCY532(Ni) pellets, their relative densities show similar changing curves when the sintering temperatures were changed. Moreover, the sintered BZCY(Ni)532 pellets always show a little higher relative density than the sintered BZCY532(Ni) pellets'. To be more specific, the relative densities of the sintered BZCY(Ni)532 and BZCY532(Ni) pellets increased when the sintering temperature increased from 1300 °C to 1400 °C. However, their relative densities did not continue to increase when the sintering temperatures were even higher. In contrast, a reversed change tendency was occurred when the sintering temperatures of 1500 °C and 1600 °C were applied. According to the furnace limitation in the lab, no higher sintering temperatures were tested in this study.

Fig. 3 The relative densities of the sintered BZCY(Ni)532, BZCY532 and BZCY532(Ni) pellets when different sintering temperatures were applied.

In order to better explain this abnormal phenomenon during the relative density measurement, the cross-sectional morphologies of these sintered pellets were taken and the result are shown in Fig. 4. It is clear that some pores existed in the sintered BZCY(Ni)532 pellets when using a sintering temperature of 1300 °C (Fig. 4(a1)). When the sintering temperature was increased to 1400 °C, almost no pores existed (Fig. 4(a2)). It is interesting to note that there are more pores in the sintered BZCY(Ni)532 pellets when they are sintered at temperatures of 1500 °C and 1600 °C. Meanwhile, the grain sizes of the sintered BZCY(Ni)532 pellets will grow up to a size of 1–5 μm (the mean grain size is equal or greater than 3 μm) at 1500 °C and to a size of 2–20 μm (the mean grain size is equal or greater than 10 μm) at 1600 °C compared to 0.5–2 μm (the mean grain size is around 1.5 μm and is more homogenous) at 1400 °C, see Fig. 4(a2)–(a4) and Table 1. The above obtained decreased relative densities at temperatures of 1500 °C and 1600 °C for sintered BZCY(Ni)532 pellets are probably caused by these newly generated pores.

Fig. 4 Cross-sectional morphologies of sintered (a1–a4) BZCY(Ni)532, (b1–b4) BZCY532 and (c1–c4) BZCY532(Ni) pellets. Note that all the scales in the SEM image was unified as 1 μm with a magnification of 5000 times.

Table 1 Mean grain size of BZCY532 based ceramics (unit: μm)

	1300 °C	1400 °C	1500 °C	1600 °C
BZCY(Ni)532	—	≥1.5	≥3	≥10
BZCY532	—	—	—	1.5
BZCY532(Ni)	—	1.5	3	5

---

For sintered BZCY532 pellets, the cross-sectional morphologies show a porous structure, when sintering temperatures of 1300–1500 °C were applied, see Fig. 4(b1)–(b3). It is undeniable that the pores become less and less when an increased sintering temperature was employed. Finally, the BZCY532 pellets turned into dense ceramics after a higher sintering temperature of 1600 °C performed (Fig. 4(b4)). Moreover, the grain sizes of the sintered BZCY532 pellets, which were sintered at a temperature of 1600 °C, are in a range of 0.5–3 μm (the mean grain size is around 1.5 μm).

Compared to sintered BZCY(Ni)532 pellets, the morphologies of the sintered BZCY532(Ni) pellets showed similar results (Fig. 4(c1)–(c4)). The main difference between them is the grain size. More specifically, the grain sizes of sintered BZCY532(Ni) pellets are 0.5–2.5 μm (the mean grain size is around 1.5 μm) at 1400 °C, 1–4 μm (the mean grain size is around 3 μm) at 1500 °C and 1–10 μm (the mean grain size is around 5 μm) at 1600 °C. Therefore, the biggest difference between the sintered BZCY(Ni)532 and BZCY532(Ni) pellets' morphologies was reflected in the grain size when using a 1600 °C sintering temperature (Table 1).

Besides the cross-sectional morphology determinations, the elemental content analysis for dense BZCY(Ni)532 and BZCY532(Ni) ceramics, which were prepared at a temperature of 1600 °C, were also carried out by EDS. For this test, the Ni contents in the grain and grain boundary were paid more attention to identify the existence form of Ni element. In addition, its detrimental electronic conduction properties will be discussed in the following sections. Moreover, different locations and areas were measured in order to make sure the achieved results were accurate. The obtained results were shown in Fig. 5 and summarized in Tables S1 and S2.† From these results, it can be seen that the Ni content in the grain and grain boundary as well as the triangle region (area 25 in Fig. 5(b)) is almost same for dense BZCY(Ni)532 ceramics, as shown in Fig. 5(a) and (b) and Table S1.† Instead, the Ni content has a higher value in the grain boundary compared to its content in the grain for dense BZCY532(Ni) ceramics, as shown in Fig. 5(c) and (d) and Table S2.†

Fig. 5 (a) and (b) Represent selected areas and points as well as the typical achieved spectrum of dense BZCY(Ni)532 ceramics used for EDS characterization with respect to grain and grain boundary, respectively. (c) and (d) Represent selected areas and points of dense BZCY532(Ni) ceramics used for EDS characterization with respect to grain and grain boundary, respectively. Note that the ceramics used for EDS test were prepared at a temperature of 1600 °C.

The decomposition of the BYN phase will accelerate the main BZCY532 phase sintering and it will speed up the grain growth, especially when the sintering temperature was higher than 1450 °C.³² Due to the very small grain growth for the sintered BZCY532(Ni) pellets, the amount of BYN phase in the BZCY532(Ni) pellets should be limited and the existence form of Ni should mainly be in a NiO state (Fig. 5 and Table S2†). Moreover, the lower XRD diffraction intensities for the 2θ angle range of 32.0–32.8 can further confirm this point. According to these results, it can preliminary be concluded that an addition of NiO during a powder preparation is more efficient compared to the addition after a powder calcination, with respect to a reduced sintering temperature and a faster grain growth.

Conductivities at different testing atmospheres and oxygen partial pressures

In order to further investigate the effect of NiO on the total conductivity and to eliminate the potential effect of the BYN phase, these three kinds of pellets that were sintered at a temperature of 1600 °C were selected for the conductivity investigation. They were measured in dry air, wet N₂ and wet H₂ atmospheres, respectively. Fig. 6(a)–(c) shows the typical impedance spectra of BZCY(Ni)532, BZCY532 and BZCY532(Ni) ceramics, when they were tested at a temperature of 600 °C.

Fig. 6 Typical impedance spectra of sintered (a) BZCY(Ni)532, (b) BZCY532 and (c) BZCY532(Ni) ceramics, which were tested at a temperature of 600 °C.

Compared to BZCY532 ceramics, the BZCY(Ni)532 and BZCY532(Ni) ceramics always show a reduced resistance regardless if they were tested in dry air, wet N₂ or wet H₂ atmospheres. Also, a Zview 3.1 software was employed for impedance spectra fitting. In addition, an equivalent electric circuit of (R₁CPE₁)(R₂CPE₂)R₃ was applied for a detailed fitting. In this electric circuit, R₁, R₂ and R₃ represent the bulk, grain-boundary and electrode resistances, respectively. Moreover, CPE represents the constant phase element. After the completion of the impedance spectra fitting, their conductivities were calculated using the following equation:

σ = L/(SR) (1)

where σ is the conductivity, L is the thickness of the electrolyte, S is the tested electrolyte area and R is the corresponding resistance of the electrolyte. The total conductivities of the sintered BZCY(Ni)532, BZCY532 and BZCY532(Ni) ceramics, which were tested in dry air, wet N₂ and wet H₂ atmospheres, are shown in Fig. 7(a)–(c), respectively. Furthermore, these data are shown and compared using the same testing atmospheres to better illustrate the effect of NiO on their corresponding conductivities, see Fig. 7(d)–(f). Moreover, their conductivities at a temperature of 600 °C is listed in Table 2, to enable a better comparison with previously reported data.

Fig. 7 (a)–(c) Represent the total conductivities in dry air, wet N₂ and wet H₂ atmospheres of the sintered BZCY(Ni)532, BZCY532 and BZCY532(Ni) ceramics, respectively. Also, (d)–(f) show a conductivity comparison for the different ceramics when they were tested in dry air, wet N₂ and wet H₂ atmospheres, respectively.

Table 2 Conductivity comparison including parts of the widely recognized literature data for BaCeO₃–BaZrO₃ based materials, where NiO additions have been used

Composition	Synthesis	NiO addition	Sintering/grain size	R.D.^a [%]	Conductivity [mS cm^−1]	Electronic conduction	Ref.
a R.D.: relative density.b CS: conventional sintering.c n.r.: not reported.d SSRS: solid-state reactive sintering with 1 wt% NiO addition.
BaCe0.9Gd0.1O3−δ (BZCG091)	EDTA complexing	No	CS^b, 1400 °C for 4 h/1.2 μm	100	18.4 (wet air) 16.5 (wet H2–Ar)	Yes	26
BaCe0.9Gd0.1O3−δ (BZCG091)	EDTA complexing	2 mol% after calcination	CS, 1200 °C for 10 h/0.5 μm	90	8 (wet air) 6 (wet H2–Ar)	Yes	26
BaCe0.9Y0.1O3−δ (BZCY091)	Oxalate co-precipitation	4 mol% after calcination	CS, 1250 °C for 10 h/≈5 μm	95	7 (wet H2, 500 °C) n.r.^c (600 °C)	n.r.	28
BaCe0.9Gd0.1O3−δ (BZCG091)	Solid-state reaction	No	CS, 1600 °C for 3 h/1–3 μm	86	≈8 (wet air) ≈5.6 (wet H2)	n.r.	29
BaCe0.9Gd0.1O3−δ (BZCG091)	Solid-state reaction	1 mol% before calcination	CS, 1450 °C for 3 h/5–20 μm	97	≈10 (wet air) ≈9.5 (wet H2)	n.r.	29
BaZr0.5Ce0.3Y0.2O3−δ (BZCY532)		1 wt% before calcination	SSRS^d, 1600 °C for 5 h/2–3 μm	98	2.1 (total, wet air)	n.r.	36
BaZr0.5Ce0.3Y0.2O3−δ (BZCY532)	Sol–gel	No	CS, 1550 °C for 8 h/1–5 μm	n.r.	4.4 (wet H2)	n.r.	42
BaZr0.1Ce0.7Y0.2O3−δ (BZCY172)	Solid-state reaction	No	CS, 1550 °C for 10 h/n.r.	n.r.	15 (wet H2/air)	n.r.	43
BaZr0.4Ce0.5Y0.2O3−δ (BZCY451)	Co-precipitation	No	CS, 1650 °C for 4 h/n.r.	≥95	8 (wet air, 700 °C)	n.r.	44
BaZr0.7Ce0.2Y0.1O3−δ (BZCY721)	Solid-state reaction	No	CS, 1700 °C for 6 h/0.5–5 μm	90–95	1.89 (wet H2)	n.r.	45
BaZr0.3Ce0.6Y0.1O3−δ (BZCY361)	Solid-state reaction	No	CS, 1700 °C for 6 h/0.5–5 μm	90–95	2.33 (wet H2)	n.r.	45
BaZr0.6Ce0.3Y0.1O3−δ (BZCY631)	Solid-state reaction	1 wt% before calcination	SSRS, 1500 °C for 4 h/3.4 μm	98.4	4.02 (wet H2/N2)	n.r.	46
BaZr0.5Ce0.4Y0.1O3−δ (BZCY541)	Solid-state reaction	1 wt% before calcination	SSRS, 1500 °C for 4 h/3.3 μm	98.2	6.08 (wet H2/N2)	n.r.	46
BaZr0.5Ce0.3Y0.2O3−δ (BZCY(Ni)532)	Solid-state reaction	1 wt% before calcination	CS, 1600 °C for 10 h/2–20 μm (mean value ≥ 10 μm)	97.0	3.4 (dry air) 13.8 (wet N2) 44.6 (wet H2)	Tiny (≥800 °C) No (≤700 °C)	This work
BaZr0.5Ce0.3Y0.2O3−δ (BZCY532)	Solid-state reaction	No	CS, 1600 °C for 24 h/0.5–3 μm (mean value: 1.5 μm)	96.4	0.5 (dry air) 1.1 (wet N2) 1.5 (wet H2)	No	This work
BaZr0.5Ce0.3Y0.2O3−δ (BZCY532(Ni))	Solid-state reaction	1 wt% after calcination	CS, 1600 °C for 10 h/1–10 μm (mean value: 5 μm)	96.4	0.9 (dry air) 1.7 (wet N2) 4.9 (wet H2)	Yes	This work

---

As this work is focused on NiO additions, the data in this table is also focused on the reported conductivity data using NiO additions. Also, their corresponding addition methods as well as the electronic conduction data are summarized. From this table, it can be seen that the conductivities achieved in this study for the BZCY(Ni)532, BZCY532 and BZCY532(Ni) ceramics were also as high as or even higher than previously reported results, which ranged from 10⁻³ S cm⁻¹ to 10⁻² S cm⁻¹ in dry air, wet Ar, wet N₂, wet H₂, or some other atmospheres. These results further confirm that dense BZCY532 based ceramics can be used as electrolytes for ITSOFC applications.

For the BZCY(Ni)532, BZCY532 and BZCY532(Ni) ceramics, the conductivities changes are very similar, when they were tested in different atmospheres. Take BZCY(Ni)532 ceramics for example (Fig. 7(a)), the conductivity data can simply be divided into three parts: (I) 200–400 °C, (II) 500–700 °C and (III) 800 °C. In the temperature region of 200–400 °C, the total conductivities tested in a dry air atmosphere is higher than those tested in the wet N₂ and H₂ atmospheres. Moreover, the differences between their conductivities that were tested in wet N₂ and H₂ atmospheres are very small. However, this tendency was reversed when they were tested in the temperature region of 500–700 °C. In this temperature region, the conductivity values in a wet H₂ atmosphere is always higher than tested in wet N₂ atmosphere. The lowest conductivities were obtained in dry air atmosphere. This phenomena indicate that proton conductivities exist in wet N₂ and H₂ atmospheres. For BZCY(Ni)532 ceramics, hydroxyl groups will be formed when they are exposed to a hydrogen/water-rich atmosphere. This newly generated hydroxyl groups can provide proton conductivities (eqn (2)). According to Grotthuss mechanism,³⁷ the actual conducting charge carriers are protons (H⁺). These protons will provide a faster transmission compared to an oxygen-ion conduction provided by oxygen vacancies . This can be explained by eqn (3) and it was confirmed by previous data from isotope experiments for similar materials.^38–41

where E_a is the activation energy, A is the pre-exponential factor, T is the Kelvin temperature, k is the Boltzmann constant and m is the mass of effective particle, respectively. However, the total conductivities of BZCY(Ni)532 ceramics in wet N₂ and H₂ atmospheres decreased when the testing temperature continued to increase up to 800 °C. Meanwhile, the conductivities tested in a dry air atmosphere were equal to or even higher than those determined in wet N₂ and H₂ atmospheres. The main reason for this decreased conductivities in wet N₂ and H₂ atmospheres is attributed to the dehydrogenation effect, which occurs at high temperatures (the reaction direction of eqn (2) will be turned into left). This will lead to reduced total conductivities. For BZCY532 and BZCY532(Ni) ceramics (Fig. 7(b) and (c)), this above mentioned phenomena is similar to what was found for the BZCY(Ni)532 ceramics.

As for the comparison in the same testing atmosphere, these three kinds of sintered samples also showed a similar change (Fig. 7(d)–(f)). In consideration of the attractive proton conductivities for the BZCY532 based electrolytes, their total conductivities that were tested in a wet H₂ atmosphere was selected and discussed in detail in this section (Fig. 7(f)). The BZCY(Ni)532 ceramics had the highest total conductivity when they were tested in the temperature range of 200–700 °C. The conductivities were lower for the BZCY532(Ni) ceramics and the lowest for the BZCY532 ceramics. In this temperature region, the total conductivities of BZCY532(Ni) ceramics are always higher than the BZCY532 ceramics, except at a temperature of 200 °C. Here, we suppose that the increased conductivities for the BZCY(Ni)532 and BZCY532(Ni) ceramics are attributed to increased oxygen vacancies, which occur after a NiO addition (eqn (4) and (5)). This means the Zr-site or Ce-site was substituted by a Ni element and this substitution leads to an increased amount of effective proton conducting charge carriers (eqn (2)). The increased conductivities for BZCY(Ni)532 and BZCY532(Ni) ceramics in a dry air atmosphere can further confirm this deduction. This is due to the presence of very small amounts of water vapor inside this testing atmosphere. Therefore, the total conductivity in this atmosphere should be attributed to oxygen vacancies. When the testing temperature is as high as 800 °C and a wet H₂ atmosphere is used, the total conductivities for the BZCY(Ni)532 ceramics also decreased, due to the dehydration effect. Though the total conductivities of BZCY532 and BZCY532(Ni) can still increase a little in a wet H₂ atmosphere, the growth rate of their conductivities is decelerated. This is due to the reduced number of proton charge carriers at a temperature of 800 °C.

From Fig. 7, it can be seen clearly that NiO plays a positive role in improving the conductivity. Though this is a desired result to improve the total conductivity, the potential electronic conductivity should be excluded for their use in practical applications for the ITSOFC electrolytes. Therefore, the total conductivities of these three kinds of ceramics were tested in a broad range of oxygen partial pressures, 1 to 10⁻²⁴ atm. In Fig. 8(a), the total conductivities of the BZCY(Ni)532 ceramics showed a decreased tendency, when they were tested at temperatures of 600 °C and 700 °C and for all oxygen partial pressures. This was induced by the reduced oxygen content in the decreasing oxygen partial pressures, which led to a reduced hole conduction (eqn (6)) and a reduced oxygen ion conduction. However, its total conductivity at 800 °C showed an increase, when the oxygen partial pressure was lower than 10⁻¹⁵ atm. This phenomenon indicate that there might be some electronic conduction inside, when the BZCY(Ni)532 ceramics were exposed to a reducing atmosphere and the testing temperature was higher than 800 °C. Therefore, the above assumed Ni substitution at the location of a Zr-site or a Ce-site is not completely correct. There should be tiny interstitial Ni particles present in the sintered BZCY(Ni)532 ceramics (eqn (7) and (8)). These interstitial Ni particles will lead to an electronic conduction at high temperatures, such as 800 °C used in this work.

Fig. 8 Total conductivities of sintered (a) BZCY(Ni)532, (b) BZCY532 and (c) BZCY532(Ni) ceramics, when they were tested in a broad range of oxygen partial pressures (1 to 10⁻²⁴ atm) and at temperatures of 600–800 °C. (d)–(f) Represent the total conductivity comparison for the sintered BZCY(Ni)532, BZCY532 and BZCY532(Ni) ceramics, when they were tested in a same testing atmosphere when the testing temperatures of 600 °C, 700 °C and 800 °C were applied, respectively.

In contrast to the BZCY(Ni)532 ceramics, the pure BZCY532 ceramics showed a decreased tendency for all oxygen partial pressure and temperature data (Fig. 8(b)). This indicate that the pure BZCY532 ceramic should be a pure oxygen and proton conducting electrolyte. As for the BZCY532(Ni) ceramics (Fig. 8(c)), these ceramics had some obvious electronic conductivities, when they were exposed to a reducing atmosphere and tested at temperatures of 600–800 °C in this study.

Combined with all the results mentioned above, the final existence form of Ni element between the BZCY(Ni)532 and BZCY532(Ni) ceramics should be different as the addition procedure is different. For BZCY(Ni)532 ceramics, NiO was added during the powder mixture ball-milling process and then the powder mixture was co-calcined. Therefore, the Ni element should first form a BYN compound after the powder calcination and thereafter be sintered in a lower temperature region of 1300–1500 °C. When the sintering temperature is as high as 1600 °C, the BYN compound will be decomposed. Then, the majorities of the Ni element will substitute part of the B-site element (Zr or Ce). After the Ni introduction in the compound, more oxygen vacancies will be created. These extra oxygen vacancies will always lead to more effective charge transfer and enhanced oxygen and proton conductivities (eqn (4)–(8) and Fig. 7). This phenomena should also apply for the solid-state reactive sintering method to prepare BZCY532 ceramics in our previous study, which also showed an increased conductivity after NiO addition.³⁶ As for BZCY532(Ni) ceramics, NiO should be more difficult to be melted into the pre-existing lattice of pure BZCY532 compound. Instead, its function should be more equivalent to that of sintering aids.²⁷ Even if parts of NiO was formed as BYN compound and it merged into the main BZCY532 phase (i.e. increased conductivity in dry air atmosphere), the major part of NiO or Ni particles should exist at the location of grain boundaries based on previously reported results.³⁰ As is known, the NiO or Ni compound has a catalytic properties for H₂, especially at high temperatures. This catalytic process will definitely produce an extra electronic conductivity and an enhanced total conductivity during tests in wet N₂ and H₂ atmospheres, but this is not favorable. According to the conductivity data of BZCY532(Ni) ceramics that were tested at various oxygen partial pressure conditions, the above assumption of NiO or Ni aggregated at the location of grain boundary should be more reasonable. Also, the Ni substitution at the location of Zr-site or Ce-site for BZCY532(Ni) ceramics should be limited. This can be further identified by the achieved EDS results in this work (Fig. 5 and Tables S1 and S2†). Therefore, it is preferable to add NiO powder during BZCY532 powder preparation. Also, the NiO addition after BZCY532 powder preparation is detrimental for final ITSOFC applications.

Conclusions

In this study, 1 wt% NiO was added before a BZCY532 powder calcination (BZCY(Ni)532) and after a powder calcination (BZCY532(Ni)). Also, trials were made without NiO addition (BZCY532). The purpose was to investigate the effects of NiO on the sintering behaviors, morphologies and total conductivities of these ceramics. The prepared dense BZCY(Ni)532, BZCY532 and BZCY532(Ni) ceramics were tested in dry air, wet N₂ and wet H₂ atmospheres, respectively. Moreover, their total conductivities at different oxygen partial pressures, 1 to 10⁻²⁴ atm, were also determined. Based on the achieved results in this study, the main conclusions may be summarized as follows:

(1) An addition of NiO powder before a powder calcination or after a powder calcination lead to an improvement of the BZCY532 sintering. Moreover, it is preferable to add the NiO powder before a powder calcination to obtain dense BZCY532 ceramics compared to adding NiO after a calcination.

(2) The addition of NiO before a powder calcination can accelerate the grain growth, while the grain growth improvement is limited when a NiO addition is made after a powder calcination.

(3) Dense BZCY(Ni)532 ceramics showed tiny electronic conduction when the testing temperature is lower than 800 °C. Instead, dense BZCY532(Ni) ceramics revealed an obvious electronic conduction when they were tested at temperatures of 600–800 °C. As for dense BZCY532 ceramics, they were identified as pure oxygen-ion and proton conductors to be used for ITSOFCs. Therefore, the addition of NiO before a powder calcination is preferable compared to after a powder calcination with respect to the unfavorable electronic conduction that is created. The research result of NiO addition in this study may be also applied to other sintering aids and this should be further studied.

Acknowledgements

The Olle Eriksson Foundation Scholarship at KTH and China Scholarship Council (CSC) are acknowledged for the financial support. Zhe Zhao gratefully acknowledges the Program for Professor of Special Appointment (Eastern Scholar) at Shanghai Institutions of Higher Learning.

References

1. E. Fabbri, A. Magraso and D. Pergolesi, MRS Bull., 2014, 39, 792–797.
2. E. Fabbri, L. Bi, D. Pergolesi and E. Traversa, Adv. Mater., 2012, 24, 195–208.
3. Y. Guo, R. Ran and Z. Shao, Int. J. Hydrogen Energy, 2010, 35, 5611–5620.
4. Y. Guo, Y. Lin, R. Ran and Z. Shao, J. Power Sources, 2009, 193, 400–407.
5. J. Qian, W. Sun, Z. Shi, Z. Tao and W. Liu, Electrochim. Acta, 2015, 151, 497–501.
6. Y. Li, R. Guo, C. Wang, Y. Liu, Z. Shao, J. An and C. Liu, Electrochim. Acta, 2013, 95, 95–101.
7. J. Li, J.-L. Luo, K. T. Chuang and A. R. Sanger, Electrochim. Acta, 2008, 53, 3701–3707.
8. P. Sawant, S. Varma, B. N. Wani and S. R. Bharadwaj, Int. J. Hydrogen Energy, 2012, 37, 3848–3856.
9. B. He, D. Ding, Y. Ling, L. Zhao and J. Cheng, Int. J. Hydrogen Energy, 2014, 39, 19087–19092.
10. Z. Shi, W. Sun and W. Liu, J. Power Sources, 2014, 245, 953–957.
11. J. Lagaeva, D. Medvedev, A. Demin and P. Tsiakaras, J. Power Sources, 2015, 278, 436–444.
12. K. H. Ryu and S. M. Haile, Solid State Ionics, 1999, 125, 355–367.
13. K. Katahira, Y. Kohchi, T. Shimura and H. Iwahara, Solid State Ionics, 2000, 138, 91–98.
14. H. G. Bohn and T. Schober, J. Am. Ceram. Soc., 2000, 83, 768–772.
15. F. Iguchi, T. Yamada, N. Sata, T. Tsurui and H. Yugami, Solid State Ionics, 2006, 177, 2381–2384.
16. P. Babilo, T. Uda and S. M. Haile, J. Mater. Res., 2007, 22, 1322–1330.
17. S. Tao and J. T. S. Irvine, J. Solid State Chem., 2007, 180, 3493–3503.
18. S. Tao and J. T. Irvine, Adv. Mater., 2006, 18, 1581–1584.
19. Z. P. Shao, W. Zhou and Z. H. Zhu, Prog. Mater. Sci., 2012, 57, 804–874.
20. L. Bi and E. Traversa, J. Mater. Res., 2014, 29, 1–15.
21. P. Babilo and S. M. Haile, J. Am. Ceram. Soc., 2005, 88, 2362–2368.
22. S. Duval, P. Holtappels, U. Stimming and T. Graule, Solid State Ionics, 2008, 179, 1112–1115.
23. M. A. Azimova and S. McIntosh, Solid State Ionics, 2009, 180, 160–167.
24. C. Peng, J. Melnik, J. Li, J. Luo, A. R. Sanger and K. T. Chuang, J. Power Sources, 2009, 190, 447–452.
25. C. Peng, J. Melnik, J.-L. Luo, A. R. Sanger and K. T. Chuang, Solid State Ionics, 2010, 181, 1372–1377.
26. M. Amsif, D. Marrero-López, J. Ruiz-Morales, S. Savvin and P. Núñez, J. Power Sources, 2011, 196, 9154–9163.
27. S. Nikodemski, J. Tong and R. O'Hayre, Solid State Ionics, 2013, 253, 201–210.
28. R. Costa, N. Grünbaum, M.-H. Berger, L. Dessemond and A. Thorel, Solid State Ionics, 2009, 180, 891–895.
29. E. Gorbova, V. Maragou, D. Medvedev, A. Demin and P. Tsiakaras, Solid State Ionics, 2008, 179, 887–890.
30. J. H. Tong, D. Clark, L. Bernau, A. Subramaniyan and R. O'Hayre, Solid State Ionics, 2010, 181, 1486–1498.
31. J. H. Tong, D. Clark, M. Hoban and R. O'Hayre, Solid State Ionics, 2010, 181, 496–503.
32. J. H. Tong, D. Clark, L. Bernau, M. Sanders and R. O'Hayre, J. Mater. Chem., 2010, 20, 6333–6341.
33. J. Bu, P. G. Jönsson and Z. Zhao, ECS Trans., 2014, 59, 315–320.
34. J. Bu, P. G. Jönsson and Z. Zhao, Ceram. Int., 2015, 41, 2558–2564.
35. J. Bu, P. G. Jönsson and Z. Zhao, Ceram. Int., 2015, 41, 2611–2615.
36. J. Bu, P. G. Jönsson and Z. Zhao, J. Power Sources, 2014, 272, 786–793.
37. N. Agmon, Chem. Phys. Lett., 1995, 244, 456–462.
38. S. Shin, H. Huang, M. Ishigame and H. Iwahara, Solid State Ionics, 1990, 40, 910–913.
39. K. C. Liang and A. S. Nowick, Solid State Ionics, 1993, 61, 77–81.
40. N. Bonanos, Solid State Ionics, 2001, 145, 265–274.
41. D. A. Stevenson, N. Jiang, R. M. Buchanan and F. E. G. Henn, Solid State Ionics, 1993, 62, 279–285.
42. E. Fabbri, A. Depifanio, E. Dibartolomeo, S. Licoccia and E. Traversa, Solid State Ionics, 2008, 179, 558–564.
43. C. Zuo, S. Zha, M. Liu, M. Hatano and M. Uchiyama, Adv. Mater., 2006, 18, 3318–3320.
44. Z. Zhong, Solid State Ionics, 2007, 178, 213–220.
45. S. Ricote, N. Bonanos and G. Caboche, Solid State Ionics, 2009, 180, 990–997.
46. S. Ricote, N. Bonanos, A. Manerbino and W. G. Coors, Int. J. Hydrogen Energy, 2012, 37, 7954–7961.

---

Footnote

† Electronic supplementary information (ESI) available: EDS data for grain and grain boundary elemental composition. See DOI: 10.1039/c6ra09936j

---