Lattice Expansion and Crystallite Size Analyses of NiO-BaCe_0.54Zr_0.36Y_0.1O_3-δ Anode Composite for Proton Ceramic Fuel Cells Application

Abstract

This study reports on the structure analyses of NiO-BCZY (BCZY = BaCe_0.54Zr_0.36Y_0.1O_3-δ) anode composite materials with the ratio of 50:50 for proton ceramic fuel cells (PCFCs) application. A product of sintered NiO-BCZY was developed to understand the structural properties of the anode materials. The objectives of this work were (a) to investigate the lattice expansion of the anode by using a high-temperature XRD (HT-XRD) from 400–700 °C; and (b) to calculate the crystallite size of the sample by using Scherrer’s and Williamson Hall’s methods. The results obtained from the HT-XRD revealed that the diffraction peaks of NiO and BCZY are matched with the cubic phase perovskite structure. For example at T = 400 °C, the lattice parameter of NiO is a = 4.2004 Å and BCZY is a = 4.3331 Å. The observation also showed that the lattice expansion increased with the temperature. Furthermore, analyses of the Scherrer and Williamson Hall methods, respectively, showed that the crystallite size is strongly correlated with the lattice expansion, which proved that the crystallite size increased as the operating temperature increased. The increment of crystallite size over the operating temperature contributed to the increment of conductivity values of the single cell.

1. Introduction

Fuel cells are one of the great sustainable energies which use electrochemical devices that convert chemical energy directly into electrical energy at high efficiency due to the lack of the Carnot constraint of the standard energy conversion chain. One type of fuel cell is a solid oxide fuel cell (SOFC) which comprises oxygen ion (O²⁻)-SOFC and hydrogen ion (H⁺)-SOFC. The H⁺-SOFC, also known as PCFC (proton ceramic fuel cell), is one of the best solutions to operate at intermediate temperatures in the range of 500–800 °C [1,2,3]. In this operating temperature regime, the components of PCFC, namely, electrode (anode and cathode) and electrolyte materials, are affected by the thermal heat. The changes in size and shape of the materials due to the heat treatment and heating profiles become significant problems, particularly on the anode side which is frequently made of metal and ceramic (cermet) [4,5,6].

It is important to accurately estimate the structure (for example the crystallite size) of anode cermet as a function of temperature because it influences the characteristics of polycrystalline materials. One of the widely used polycrystalline materials as an anode composite for PCFC is NiO-BaCe_0.54Zr_0.36Y_0.1O_3-δ (BCZY) [7,8,9,10]. Nickel is used in the composite anode because it has high catalytic properties for hydrogen oxidation [11]. Furthermore, the use of the composite anode enhances thermal compatibility, reduces interface resistance and lengthens the triple-phase boundary (TPB) [12,13]. In terms of NiO:BCZY composition, it has been reported that more than 40% nickel content can effectively improve electrochemical performance [14,15]. Several studies have discovered that some problems occur, such as fracture formation caused by thermal mismatch between anode layers [16,17]. Hence, data on anode crystallite size are very useful and significant to ensure that the electrolyte component has close TEC under the desired working conditions [18,19,20,21].

Most studies have showed a linear relationship between the size of the materials’ crystallites and the annealing temperature [22,23]. On the other hand, correlations between the crystallite size of NiO-BCZY anode composite at working temperatures of PCFC are still small in number. Thus, in this work, Scherrer’s and Williamson Hall’s methods were adopted to calculate the crystallite size of NiO-BCZY that has undergone operating temperatures from 400–700 °C. Williamson Hall’s method (W-H method) is more relevant and accurate than Scherrer’s method as the lattice strain, lattice stress, and also energy density is taken into account in the calculation [24,25].

In terms of electrical performance of the anode material, research done by Rhidwan et al., showed that crystallite size was inversely proportional to the grain resistance [26]. At 750 °C, the crystallite size of electrolyte bismuth-based materials showed that an almost linear relationship to conductivity can be observed [27]. Thus, the effect of lattice expansion and crystallite size trends on the electrochemical performance, such as conductivity and power density of NiO-BCZY, were also identified in this study.

2. Materials and Methods

Firstly, to prepare the anode composite powder, the raw material of barium nitrate (Ba(NO₃)₂), cerium (III) nitrate hexahydrate (Ce(NO₃)₃.6H₂O), zirconyl (IV) nitrate hydrate (Zr(NO₃)₂O.xH₂O), yttrium (III) nitrate hexahydrate (Y(NO₃)₃.5H₂O) were dissolved in deionized water. Nickel (II) nitrate (Ni(NO₃)₂ and citric acid (CA) were added to the solution, and its pH was adjusted to 7 by using ammonia hydroxide. Next, the mixture was heated at 120 °C overnight and dried at 325 °C for about 2 h. The dried powder was ground and calcined at 1100 °C for 10 h and then pressed at the pressure of 5 MPa for about 1 min to produce a circular pellet. The obtained pellet was sintered at 1400 °C for 6 h.

The sintered pellet was ground using a mechanical grinder to a powder form before being subjected to the high-temperature X-ray diffraction (HT-XRD) from 400 to 700 °C. The Rietveld refinement technique was utilized by using commercial software of Highscore Plus to calculate the lattice expansion at operating temperatures. After refinement, the material structures were observed through visual for electronic and structural analysis (VESTA). Furthermore, the isolated and selected high-intensity peaks that correlate to a larger diffraction angle were chosen to determine the crystallite size using Scherrer’s method and several Williamson Hall plot models. These two methods give an extensive range of crystallite sizes.

For Scherrer’s method, the crystallite size of the sample (D) was calculated using Equation (1):

$$D=\frac{k\lambda }{{\beta }_{hkl}\cos {\theta }_{hkl}}$$

(1)

where Scherrer’s constant (k) = 0.9, wavelength (λ) = 0.154056 nm for Cu-K_α radiation, θ_hkl is Brag diffraction angle and β_hkl is the broadening of the hkl diffraction peak measured at half of its maximum intensity in radians.

On the other hand, the Williamson Hall’s method is utilized to predict a more accurate calculation of crystallite size by using three models: uniform deformation model (UDM); uniform stress model (USDM); and energy density model (UDEDM) using Equation (2), Equation (3) and Equation (4), respectively.

$${\beta }_{hkl}=\left(\frac{k\lambda }{D}\right)+4\epsilon \sin \theta$$

(2)

$${\beta }_{hkl}\cos \theta =\left(\frac{k\lambda }{D}\right)+4\sigma \frac{\sin \theta }{E_{hkl}}$$

(3)

$${\beta }_{hkl}\cos \theta =\left(\frac{k\lambda }{D}\right)+4\sin \theta {\left(\frac{2u}{E_{hkl}}\right)}^{\frac{1}{2}}$$

(4)

where E_hkl is the young modulus for the cubic crystal.

To further characterize the electrochemical properties of NiO-BCZY anode composite, the impedance measurement and I-V characteristics were carried out. The anode composite was fabricated as an anode substrate of PCFC single cell. The substrate was coupled with cathode and electrolyte materials with configuration of anode (substrate)|electrolyte (thin film)|cathode (thin film). The in-house developed of BCZY and LSCF-BCZY (LSCF = La_0.6Sr_0.4Co_0.2Fe_0.8O_3-δ)-BCZY were employed as electrolyte and composite cathode, respectively [28,29]. The fabricated single cell of NiO-BCZY|BCZY|LSCF-BCZY following the previously reported procedure [30] was placed at the sample holder of custom-made conductivity station for the conductivity and power density measurements. The measurement was done using electrochemical impedance spectroscopy (EIS) ZIVE SP2 Electrochemical Workstation (ZIVELAB WonATech) in the temperature ranging from 500 to 700 °C under hydrogen fuel at anode side and stagnant air at cathode side. Impedance spectrum of the cell was analyzed using ZIVE^® Smart Manager™ software and I-V polarization curve was plotted for power density assessment.

3. Results

3.1. XRD Pattern

Figure 1 shows the final diffraction pattern of high-temperature XRD for the fine NiO-BCZY anode composite from 400–700 °C. A finely ground powder is needed to achieve an excellent signal-to-noise ratio (SNR) and avoid fluctuation intensity. Both factors will reduce the preferred orientation and thus avoid spottiness and inconsistency analyses [31]. After allowing the pattern shift peaks, all main peaks in the XRD pattern matched with the Joint Committee of Powder Diffraction Standards (JCPDS) file number for NiO is 01-078-0423 and BCZY is 01-089-2485. The respective JCPDS number of NiO and BCZY used was matched with most of the recent papers reported [32,33,34]. However, a secondary phase of cerium oxide, CeO (JCPDS no.: 00-0040-0593) was also detected in the spectrum as reported by [35]. The presence of this secondary phase will lead to the non-homogenous formation of NiO-BCZY anode composite as a result of partial decomposition of the pre-prepared BCZY phase [36] due to the incomplete reaction between NiO and the ceramic part.

Figure 2 shows the peaks of BCZY at 2θ = ± 29.2 and NiO at 2θ = ± 43.1 is shifted to the low angle of 2θ (left side) indicating that the lattice parameter increased as the temperature raised to 700 °C. This trend was also reported by Sultan et al. [37] on thermal expansion of semiconductor material. The authors explained a diffuse scattering phenomenon that caused by thermal expansion will reduced the intensity in the Bragg positions. The values of lattice parameter for BCZY and NiO at the temperature of 400 to 700 °C are presented in

Table 1. The Lattice parameter of NiO and BCZY at temperatures of 400 to 700 °C.

Temp. (°C)	Lattice Parameter, a (Å)
	NiO	BCZY
400	4.2004	4.3331
500	4.2064	4.3400
600	4.2115	4.3433
700	4.2152	4.3462

. The increase in value shows that the lattice went through thermal expansion while heated [38,39]. The value for the goodness of fit (GOF) obtained was in the range of 0.82 to 3.48. The low number of GOFs signifies that the XRD data was effectively refined. Figure 3 shows the result of refinement for Rietveld analysis of XRD pattern and the cubic structure of NiO-BCZY anode composite that has been observed using VESTA software.

3.2. Crystallite Size

The result of crystallite size (D) calculated using Scherrer’s method and crystallite size (D), lattice strain (ε), deformation stress (σ), and density of energy (U) calculated using, UDM, USDM, and UDEDM models are summarized in

Table 2. Summary of crystallite size (D), lattice strain (ε), deformation stress (σ) and energy density (U) calculated using a different model for NiO at temperatures of 400 °C to 700 °C.

NiO
Temp, (°C)	Scherrer Method	Williamson-Hall Method
		UDM		USDM			UDEDM
	D (nm)	D (nm)	ε (×10−3)	D (nm)	ε (×10−3)	σ	D (nm)	ε (×10−3)	σ	U (×10−5)
400	48.71	33.37	12.00	33.37	0.19	28.53	33.37	1.22	186.20	11.32
500	48.79	34.98	11.00	34.98	0.77	118.04	34.98	1.07	163.80	8.76
600	49.01	36.48	8.90	35.15	0.18	28.04	35.32	1.19	183.00	10.94
700	49.34	38.01	8.60	35.32	0.18	27.47	34.65	1.17	179.40	10.51

for NiO and

Table 3. Summary of crystallite size (D), lattice strain (ε), deformation stress (σ) and energy density (U) calculated using a different model for BCZY at temperatures of 400 °C to 700 °C.

BaCe0.54Zr0.36Y0.1O3-δ
Temp, (°C)	Scherrer Method	Williamson-Hall Method
		UDM		USDM			UDEDM
	D (nm)	D (nm)	ε (×10−3)	D (nm)	ε (×10−3)	σ	D (nm)	ε (×10−3)	σ	U (×10−5)
400	30.91	109.71	3.09	109.71	3.09	472.97	109.71	3.09	472.90	73.01
500	31.62	132.86	5.07	132.85	5.07	777.38	132.85	5.07	777.30	197.20
600	34.31	135.34	4.88	135.34	4.88	746.83	135.34	4.88	746.80	182.00
700	41.82	169.18	4.18	169.18	4.18	639.60	169.18	4.18	639.50	133.50

for BCZY. The crystallite size of NiO is larger in Scherrer’s method as compared to the W-H method which is in contrast with the crystallite size of BCZY. Both methods are similar in terms of dependency on the diffraction angle, θ but can be distinguished as explained in Equations (1)–(4) (Scherrer’s is 1/cos θ dependent and W-H is tan θ dependent). Thus, due to different θ positions of respective NiO and BCZY, it will affect the crystallite size and strain broadening data [40]. In addition, the same reason was also explained by Ilyas et al. [41] with the additional factor that contributes to the size and strain such as the widening of the diffraction peak, β. Theoretically, the broadening of the diffraction peak, β is inversely proportional to crystallite size, D [42]. Overall, the average crystallite size showed positive increments as temperature increased due to crystal lattice dilation [37].

In addition, the lattice strain is specified as a lattice expansion or contraction due to changes in crystallite size that come from the modification of atomic arrangement [43]. Since strain can be obtained through a slope of the linear fit to the data, a positive slope of lattice strain is attributed to the lattice expansion [44] as shown in Figure 4. Ideal data for the strain were supposed to be decreased as temperature increased as lattice expansion was taken into account; however, due to poor techniques of fitting data, inconsistent results were achieved. The same inconsistent observation was also reported by Yusoff et al. [45].

3.3. Electrochemical Measurement

A study on electrical anode performance of NiO-BCZY pellet has been reported elsewhere [46]. Hence, as a continuity work, a single cell was fabricated with a configuration anode|electrolyte|cathode:NiO-BCZY|BCZY|LSCF-BCZY to intensively evaluate the effect of lattice expansion and crystallite size of anode materials on the electrochemical performance. Figure 5 shows a typical Nyquist plot of NiO-BCZY|BCZY|LSCF-BCZY single cell was measured at 700 °C with an inductance ‘tail’ below the x-axis region and two well-defined arcs. The impedance spectrum was fitted using the equivalent circuit L_s-R_o-(R₁Q₁)-(R₂Q₂) where L_s and R_o correspond to the inductance and ohmic resistance, respectively. R₁ and R₂ indicate the resistance from arc-1 and arc-2 where R₁ + R₂ is the polarization resistance, R_p of the cell.

Table 4. The area specific resistance and total conductivities of single cell at 500–700 °C.

Temperature (°C)	Ro (Ωcm2)	Rp (Ωcm2)	Total Conductivity (×10−3 Scm−1)
700	34.19	7.90	1.90
600	44.37	8.30	1.52
500	52.45	10.26	1.28

shows the R_o, R_p and conductivity values of the cell at 500–700 °C. The conductivity increases with increasing temperature demonstrates the involved reactions are temperature dependent and indicates that the corresponding electrochemical reactions are thermally activated processes [47]. The total conductivity of the cell is comparable to a study done by Senari et al. [48] in a similar operation atmosphere.

As studied by Hossain et al. [49] the conductivity mechanism was reported as a result of the contribution of ions transportation. For example, the proton transport in crystal lattice of perovskite strongly affect the proton conductivity that occur between the lattices through hopping and reorientation mechanism [50]. Thus, perovskite materials that have large lattice volume tend to form high proton conductivity and vice versa [51]. Greater lattice size promotes more energy to the lattice vibration, which speeds up proton transport and results in better proton conductivity [52].

Shown in Figure 6 is the current- voltage (I-V) and current-power (I-P) plots of the single cell measured when the open circuit voltage (OCV) was stabilized at 700 °C. As compared to our previous study [53], the maximum power density of present findings at 500–700 °C (

Table 5. Power density value of single cell at 500–700 °C.

Temperature (°C)	Power Density (mW/cm−2)
700	0.48
600	0.30
500	0.18

) have improved from nW/cm² to mW/cm² without the need of pore former and material modification. However, the power density of this in-house single cell is noticeably lower than the state-of-art in PCFC technology, particularly NiO-BCZY-based anode composite [9,10,14,46,49]. One of the reasons is due to the ohmic resistance of electrolyte that dominates the cell resistance and influence the OCV of the cell (~0.21 V at 700 °C). The small OCV value indicates insufficiently dense electrolyte membrane to prevent gas mixture from anode to cathode, which might explain the overall performance of the cell. Hence, promising fabrication techniques should be taken into consideration for future development [54,55]. On the other hand, it is best to limit or reduce the extreme nickel diffusion from Ni-based cermet anode into barium cerate electrolyte through altering fabrication procedures, for instance, as this tends to result in poor cell performance by lowering the ionic conductivity [56]. Since each research study compares different cell designs, it might be difficult to thoroughly understand the factors that affect the output performance of fuel cells, including fabrication technique, current collector, and gas flow.

4. Conclusions

The increase in lattice parameter and unit cell volume indicates that the NiO-BCZY anode composite undergoes lattice expansion at operating temperatures from 400 to 700 °C. The results from Scherrer’s method and Williamson-Hall’s method showed that the crystallite size increases as the temperature increases. The average value obtained for crystallite size from different models shows that the lattice strain and deformation stress affected the crystallite size. It can be concluded that the Williamson-Hall method gives a more accurate value compared to Scherrer’s method. The electrochemical results of the single cell NiO-BCZY|BCZY|LSCF-BCZY exhibited almost linear correlation with the obtained crystallite size data at operating temperatures.