1 Introduction

Solid oxide fuel cell (SOFC) and Protonic ceramic fuel cell (PCFC) have gained huge attention in the recent years for the direct use of chemical energy of fuels to generate electricity with high efficiency and low emission[1, 2]. According to the classification of transported carriers, ceramic fuel cell (CFC) includes SOFC, which is mainly oxygen ion conduction, and PCFC, which is mainly proton conduction. Conventional SOFC, realised by yttria stabilized zirconia (YSZ) that requires a high working temperature (800–1000 °C) to achieve sufficient ionic conductivity (> 0.1 S/cm) that is demanded for the device operation and it poses a huge challenge for its commercial application. Due to high temperature requirement, it increases the complexity of conventional technology, and ultimately raises the capital cost of the device [3, 4]. Goodenough introduced the concept of materials which could operate at sufficiently low temperatures below 700 °C [5]. In the field of low-temperature SOFC, well-studied doped zirconia electrolytes materials are replaced by gadolinium doped ceria (GDC) as well as samarium doped ceria (SDC) [6, 7]. Compared to the conventional YSZ based fuel cells, the ceria-based electrolytes show lower activation energy and therefore, can work at 400–700 °C [8, 9].

Meanwhile, as an alternative electrolyte of oxygen ion conductor, proton conductor ceramic materials have emerged as potential candidates. In this aspect, perovskite ceramic electrolytes with a high proton conductivity exhibit promising results in hydrogen or water vapor atmosphere. Generally, it is considered that Grotthuss and Vehicle mechanisms are two main mechanisms of proton conductivity [10]. The former is due to the jumps of excess hydrogen through the bonds from one stationary lattice oxygen to the neighbouring lattice oxygen; the latter is due to the attachment of protons with the oxygen ions and then move in the form of OH− ions via translational diffusion mechanism through appropriate vacancies in the perovskites [11]. Protons move easily through the lattice, which results in proton conduction being activated at lower activation energy level compare to that of attachment with oxygen ions and the composed conduction [7]. Here, BaZr0.8Y0.2O3 (BZY) and BaCe0.7Zr0.1Y0.2O3-δ (BCZY) electrolytes are some typical examples for proton conduction [12, 13].Compared to conventional SOFC, PCFC is composed of proton-conducting ceramic oxides and has greater potential for low-temperature operation [14]. Table 1 gives an overview of the electrochemical performance of PCFCs. As shown in Table 1, doped BaCeO3 and doped BaZrO3 electrolytes, which have been extensively investigated in recent years, exhibit a potential to operate at low temperatures (300–550 °C).

Table 1 An overview of the maximum power density (MPD) of PCFCs at low temperature in the recent five years
Full size table

Usually, in ceria-based electrolytes, performance enhancement is achieved through doping of other elements. As an example, samarium doped ceria (SDC) where Sm3+ replacing Ce4+ in CeO2, generates extra vacancies for oxygen and thereafter, oxygen ion conductivity increases through SDC as an electrolyte. In fact, the mechanism of changing overall conductivity of the electrolytes depend on the variation of different doping ratios. Sm, Gd, Y, La, Nd, and Pr are + 3 valence ions, and are useful to generate more oxygen vacancies due to smaller enthalpy value of association and also the ionic compensation among Ce4+ and + 3 valence ions [21]. Especially, in La and Pr co-doped ceria, we have come to know that larger radii of La3+ and Pr3+ ions lead towards the high lattice strain, that increases the defect formation in response [22]. Meanwhile, excess Pr3+ ions limit the carrier mobility and increase the band gap value, and the high band gap limits the formation of oxygen vacancies [23]. However, the doping of La3+ ions into the main ceria lattice results turning it into a donor level, which supports decrease in the band gap between valence and conduction bands and therefore, the phenomena counteracts to increase a possibility of oxygen vacancy formation [24].

Besides, this study is aimed to understand the conduction mechanism through La/Pr co-doped ceria (LCP) electrolyte for high-performance low-temperature CFC. In addition, LCP sample is sandwiched between two identical layers of Ni0.8Co0.15Al0.05LiO2-δ (NCAL) coated Ni foam (NCAL-Ni) in order to construct a mechanically robust symmetric cell. Finally, the device is tested in fuel cell conditions; the cell has achieved a superior power density with a noticeably high ionic conductivity at sufficiently low temperature, i.e. @ 520 °C. Furthermore, the cell was examined for the gain mechanism of the NCAL-Ni/LCP/NCAL-Ni cell and proton conduction was found. This work demonstrates LCP electrolyte with novel proton conduction for developments of CFC, giving a new path for CFC development which could be practically possible at affordable temperatures.

2 Experimental

2.1 Materials Preparation

The original material used in this work is industrial-grade rare-earth LaCePr-carbonate purchased from Inner Mongolia Baotou Rare-Earth Hi-Tech Co Ltd, China. The LaCePr-carbonate is an industrial-grade mineral mainly composed of La2(CO3)3 (24.9 wt%), Ce2(CO3)3 (74.7 wt%), and Pr2(CO3)3 (0.4 wt%), which is used as the raw material to produce industrial-grade rare-earth oxides. The electrode material NCAL is purchased from Tianjin Bamo Science &Technology Joint Stock Ltd., China. The LaCePr-carbonate is heat-treated at 800 °C for 4 h to get LCP for the subsequent experiments. Previous study has established that the LCP with the chemical formula La0.33Ce0.62Pr0.05O2-δ [25,26,27,28].

BCZY (BaCe0.7Zr0.1Y0.2O3-δ) is synthesized using sol–gel method. Ba(NO3)2 (> 99%, purchased from Aladdin Bio-Chem Technology Ltd., China), Zr(NO3)4·5H2O (99.99%, purchased from Macklin Biochemical Co., Ltd., China), Ce(NO3)2·6H2O, Y(NO3)3·6H2O (purchased from A.R., Aladdin Bio-Chem Technology Ltd., China) were dissolved in ethylenediamine tetra acetic acid (C10H16N2O8, EDTA, purchased from A.R., Sinopharm Chemical Reagent Co., Ltd, China) solution according to the required stoichiometric ratio. Citric acid monohydrate (C6H8O7·H2O, purchased from A.R., Chron Chemicals Ltd., China) and C10H16N2O8 were added to the combined solution with a molar ratio of cation: citric acid: EDTA = 1: 1.5: 1. Then, ammonia solution (NH3·H2O, purchased from A.R., Sinopharm Chemical Reagent Co., Ltd., China) was used to adjust the solution with pH value to approximately 9, until the solution became transparent via a visual inspection. The resultant solution was continuously stirred at 120 °C for 4 h to evaporate the water to get grey colored gel after the auto-combustion. Finally, the gel was calcined at 1000 ℃ in air for 5 h to obtain final BZCY product.

2.2 Material Characterization

X-ray diffraction (XRD) of the powder samples was performed by an advanced X-ray diffractometer (Bruker D8 ADVANCE, Bruker Company, Germany) with Cu Kα radiation (λ = 1.54060 Å). The morphology and detailed microstructure of the sample were studied using scanning electron microscopy (SEM; SU8010, Hitachi Limited, Japan) and high resolution transmission electron microscopy (TEM; Talos F200x, FEI Co., USA); also, an energy dispersive spectrometer (EDS) coupled with SEM focusing on O, Ce, La, and Pr elements was used to study the elemental composition. Valence state changes and different bonding mechanisms were detected by X-ray photoelectron spectroscopy (XPS; Escalab Xi + , Thermo Fischer Scientific, USA) in C 1 s, O 1 s, La 3d, Ce 3d and Pr 3d orbits. In addition, Fourier transform infrared spectroscopy (FTIR; Nicolet Is10, Thermo Fischer Scientific, USA) was performed to characterize the surface vibrational properties and the functional of the samples.

2.3 Fuel Cells Fabrication and Testing

The LCP fuel cells were fabricated via dry-pressing method. As show in Fig. S1, 0.2 g electrolyte powder was sandwiched between two pieces of NCAL-Ni electrodes by a unidirectional hydraulic press applying 10 MPa of compression to obtain a single cell with a configuration of NCAL-Ni/LCP/NCAL-Ni, which possessed an effective working area of 0.64 cm2. Fig. S2 shows a cross-sectional SEM image of this cell, with a cell thickness of 2 mm and an electrolyte layer thickness of 0.5 mm. H2 with flow rate of 175 mL/min and pumped air with flow rate of 150 mL/min were utilized as fuel and oxidant, respectively. The I-V (current–voltage) and I-P (current-power) characteristics of the cell were measured at various temperatures using an IT8511 electronic load (ITECH Electrical Co., Ltd., China), while the data was recorded by an IT7000 software. An electrochemical workstation (CHI660E, ITECH Electrical Co., Ltd., China) was used to perform the electrochemical impedance spectroscopy (EIS), which was carried out at open circuit voltage from 1 MHz to 0.1 Hz with an amplitude of 5 mV.

3 Results and Discussion

3.1 Structural, Morphological and Compositional Analysis

Figure 1 shows XRD pattern of LCP that displays the same fluorite structure as that of CeO2 (PDF No. 34–0394). The 2-theta value of the LCP peak is slightly shifted towards a lower angle as compared to that of pure CeO2. This can be considered as the doping effect of La3+ (1.16 Å) and Pr3+ (1.12 Å) with larger ionic radii instead of Ce4+ (0.97 Å), resulting in lattice expansion [29].

Fig. 1
figure 1

XRD patterns of LCP electrolyte.

Full size image

Figure 2(a) shows the SEM image of LCP, it can be observed that the industrial-grade rare-earth LCP material after direct sintering at 800 °C has an irregular particle structure with some agglomeration. The particle size distribution of LCP is measured between 100 and 300 nm. From the EDS images in Fig. 2(b–e), the main elements of LCP material are observed as uniformly distributed along the overall material structure.

Fig. 2
figure 2

a SEM image of LCP electrolyte, b O, c Ce, d La and e Pr EDS images for elemental mapping of the LCP electrolyte.

Full size image

Figure 3(a) and (b) show the TEM images of LCP with the lattice spacing of 0.271 nm and 0.329 nm in the (111) and (200) planes, respectively, which is larger than that of CeO2. Comparing the TEM images (a) before the testing and (b) after the testing, it can be observed that there is an amorphous layer (dotted area) on the surface of the LCP particles after the testing. The developement of this amorphous layer after electrochemical testing has been easily detected in nanostructured fuel cells with NCAL electrode, due to the reductive NCAL materials in fuel gaes envirmoment together with the initial porous structure of electrolyte. The composition and detailed mechansim of this amorphous layer would be revealed by the following XPS and FTIR analysis.

Fig. 3
figure 3

TEM images of LCP electrolyte a before the testing and b after the testing.

Full size image

3.2 Electrochemical Performance of the Device and Conduction Mechanism

At different temperatures, Fig. 4(a) displays the I-V and I-P characteristics of the LCP electrolyte fuel cell. The LCP device has a MPD of 757 mW/cm2 at 520 °C, and its open circuit voltage (OCV) is 1.071 V. Observed OCV increases with a decrease in temperature, agreeing with the Nernst equation. Furthermore, this figure depicts the MPDs of 619, 552, 292, 153 and 30 mW/cm2 at 500, 480, 460, 440 and 420 °C. The sudden decrease of MPD observed when the temperature was reduced to 420 °C can be explained by phase transition in the electrolyte. In the fuel cell test, the NCAL anode was reduced in hydrogen atmosphere forming LiOH/Li2CO3 mixture, which in molten state diffused into the electrolyte [30, 31]. In the preparation of NCAL-Ni electrode, NCAL powder was mixed with ethanol and terpineol solvent to get slurry and pasted on the Ni foam. Therefore, the ethanol and terpineol are believed to be the main source of the carbon in the amorphous layer. Besides, CO2 in the air is also a possible source. In molten/partially molten states, LiOH/Li2CO3 mixture seals both the porous anode and the electrolyte, resulting in a high OCV that is saturated to the theoretical value [32]. Correspondingly, when the temperature decreases to 420 °C, the MPD of the cell deteriorated dramatically. In the study of NCAL/GDC/NCAL fuel cell, 419 °C was considered as the melting/solidification critical point of LiOH/Li2CO3 mixture [33]. As the LiOH/Li2CO3 mixture solidifies in the LCP device at around 419 °C, the conductivity of the electrolyte dropped abruptly disrupting a production of power output. In addition, the solidification point of LiOH is 471 °C, and the sudden power density degradation at 460 °C may be caused by the partial solidification of LiOH in the LiOH/Li2CO3 mixture.

Fig. 4
figure 4

a I-V and I-P characteristics of the LCP electrolyte fuel cell; b EIS of LCP electrolyte fuel cell tested under H2/air conditions; c Conductivity plots and d Arrhenius plots of the LCP electrolyte fuel cell.

Full size image

Figure 4(b) exhibits the EIS results for the LCP fuel cell in H2/air atmosphere from 520 °C down to 420 °C. The raw data were fitted to the equivalent circuit by Zview. In the EIS, the x-axis intercept at high frequencies indicates the ohmic impedance, and similarly, the x-axis intercepts at medium and low frequencies indicate the charge transfer impedance and mass transfer impedance, respectively [34]. The Nyquist plot can be modeled using an equivalent circuit, where R0 represents the bulk conduction resistance, R1 represents the grain boundary conduction resistance, R2 represents the electrode polarization resistance, and CEP is its corresponding constant element phase. Table 2 shows the results of the fitting calculation.

The ionic conductivity can be calculated by the following equation:

$${\sigma }_{i}=\frac{L}{R\times S}$$
(1)

where L denotes the width of the electrolyte, S denotes the active area of the cell, and R denotes the resistance associated with the electrolyte. Here, R can be represented in two ways: 1. the value of R0 + R1 in the EIS results; 2. the resistance of the I-V curve in the ohmic polarization region [34].

As shown in Fig. 4(c), the ionic conductivity of LCP is 0.111-0.032 S/cm at 520–420 °C, which is significantly higher than that of conventional oxygen ion conductors. In this case, the result obtained by simulating the ohmic resistance in the EIS is the total conductivity influenced by ions and electrons. In contrast, the resistance of the linear part of the ohmic polarization region in the polarization curve corresponds to the ionic conductivity. The ionic conductivity (δi) results from the EIS are not significantly different from the polarization curve (I-V curve) obtained, and it can be judged that the electron conductivity in the LCP fuel cell is negligible Table 2.

Table 2 The fitting parameters obtained from Zview with an equivalent circuit. The R and constant phase element (CPE) have the units of Ω cm2 and F cm−2, respectively
Full size table

The activation energy can be calculated by the following Arrhenius equation [35].

$$\mathrm{ln}({\sigma }_{i}T)=ln(A)-\frac{{E}_{a}}{\mathrm{R}}\times \frac{1}{T}$$
(2 )

Here, A is a pre-exponential factor, T is temperature in kelvin, R is the molar gas constant, and Ea is the activation energy of ion migration. As shown in Fig. 4(d), Ea obtained by EIS is 0.642 eV. Because of its high ionic conductivity and low activation energy, the LCP electrolyte cannot be a conventional oxygen ion conductor as they usually exhibit their activation energy greater than 1.0 eV, whilst the activation energy for surface conduction mechanism can be as low as 0.2 eV [7, 36]. In order to elucidate the conduction mechanism of LCP electrolyte and the reason for its high performance, the investigation carried on further.

To investigate the type of carrier conduction in LCP material, BCZY was used as the blocking layer. BCZY exhibits proton conduction at low temperatures [13, 37]. This method is often used to determine the proton conductivity of electrolyte materials [28, 38]. Figure 5(a) shows the I-V and I-P characteristics of cell that is constructed using two 0.05 g BCZY blocking layers at each side of LCP electrolyte, as shown in Fig. 5(b). The cell has MPD of 732 mW/cm2 at 520 °C with an OCV of 1.062 V. The comparison of MDP at 520 °C from this cell with that (757 mW/cm2) of a cell without the blocking layer in it indicates a marginal difference, proving that proton conduction dominates through a LCP electrolyte layer for low-temperature CFC.

Fig. 5
figure 5

a I-V and I-P characteristics of the NCAL-Ni/BCZY/LCP/BCZY/NCAL-Ni fuel cell measured at 520 °C and b the corresponding schematic of the cell structure.

Full size image

In fuel cells, hydrogen bonding is considered as the pathway for proton conduction [39, 40]. In order to verify the proton conduction mechanism through a LCP electrolyte, FTIR (Fourier transform infrared spectroscopy) characterization was used to investigate any effect in the material before and after the cell’s performance tests, as shown in Fig. 6. The peak at 3435 cm−1 is the O–H bond in LiOH and H2O, and the peak at 2922 cm−1 is caused by the stretching vibration of O–H [41, 42]. These indicate that there are many OH groups in the LCP electrolyte and that protons may be transferred in the electrolyte in the form of O–H. Also, the peak at 2345 cm−1 is caused by the O = C = O antisymmetric stretching vibration, the peaks at 1499 cm−1 and 1439 cm−1 are attributed to the antisymmetric stretching vibration of CO32−, and the peak at 864 cm−1 is caused by the out-plane bending vibration of CO32− [43, 44]. This may be an effect that is related to the infiltration of LiOH/Li2CO3 mixture in the NCAL electrode. The faint peaks at the same positions of 1499 cm−1 and 1439 cm−1 in the cell before the test may be caused by some residues of the LaCePr-carbonate.

Fig. 6
figure 6

FTIR spectra of the LCP cell before and after the test.

Full size image

For further surface analysis of LCP, the valence state changes of the material was investigated using XPS. As shown in Fig. 7(a), the main characteristic peaks of O, Ce, La and Pr can be found in a wide range of XPS spectra from simple LCP samples. In contrast, in the tested LCP cell, the characteristic peaks of Ce, La, and Pr are no longer detectable, while the faint characteristic peaks of Li elements can be detected. In Fig. 7(b), the XPS spectra of C 1 s before the test and after the test can be fitted into two peaks. Position 1 (298.21 eV) and 1’ (298.77 eV) can be assigned to the CO32−, position 2 (284.83 eV) and 2’ (284.76 eV) can be assigned to the hydrocarbon [33]. In Fig. 7(c), the peaks of adsorbed oxygen, associated with OH−, CO32− and oxygen vacancies, with binding energy around position 1 (531.35 eV) and 1’ (531.55 eV) appears in the O 1 s XPS spectrum, while position 2 (528.54 eV) is attributed to lattice oxygen [33, 45]. It is noticeable that the peaks indicating OH− and CO32− are significantly enhanced after the test, while the peaks of lattice oxygen are almost difficult to be detected. This indicates that during the fuel cell testing, the NCAL anode was reduced under hydrogen environment and formed LiOH/Li2CO3 mixture in the molten state [32]. The LiOH/Li2CO3 mixture diffuse into the electrolyte and is filled between the interfaces of the LCP particles forming an amorphous layer, as shown in Fig. 3(b). This amorphous layer has been reported in previous nanostructured LST, BCY and GDC fuel cells using the same dry-pressing method similar in LCP too [32, 33, 42, 46]. The authors speculate that this amorphous layer forms a high-speed pathway for ion conduction on the surface of LCP particles, which helps to enhance the proton conductivity of LCP electrolyte and hence improve the power output of the cell.

Fig. 7
figure 7

XPS spectra of (a) the wide survey of all elements; b C 1s, c O 1s, d La 3d e Ce 3d and f Pr 3d of LCP.

Full size image

Recently, proton conduction is reported at the interface of nanostructured oxides and carbonates [47]. Ricca et al. suggested that LiKCO3 is coated on the surface of YSZ, forming a high-speed pathway for ion transport in the two-phase interface region [48]. Wang et al. investigated the SDC/Na2CO3 nanocomposite electrolyte and concluded that the interface due to the introduction of carbonate enhances the proton conduction and used a “Swing Model”. In their model, protons form meta-stable hydrogen bonds with oxygen ions from both CO32− group and the surface of SDC as a conduction pathway [49]. Similarly, in the LCP electrolyte, the LiOH/Li2CO3 mixture coated on the surface of LCP particles establishes a high-speed pathway for proton transfer. The bending and stretching vibrations of CO32− demonstrate the process of hydrogen bond formation and breaking between proton and CO32− groups. Along the direction of hydrogen concentration, the H+-CO32− bond becomes one of the main pathways for proton conduction. Moreover, O–H stretching vibration was also observed, which may be caused by hydrogen bond between protons and oxygen atoms on the LCP surface. The concentration difference cell method was used to investigate the LCP ion conduction mechanism without the influence of LiOH/Li2CO3 mixture. The concentration cell method is often used to determine the carrier type and a number of ion transfer within fuel cells [50]. For instance, LiAl0.5Co0.5O2, BCY and other materials demonstrated the conduction mechanism using same method [46, 51]. Hydrogen concentration cells (HCC) were fabricated using platinum as the symmetric electrode and oxygen concentration cells (OCC) were fabricated using silver as the symmetric electrode. The voltage of HCC was recorded under the conditions of 5%H2 + 95%Ar and pure H2 on one and other side of the cell [52]. The voltage of the OCC was recorded under the conditions of 2%O2 + 98%Ar and 21%O2 + 79%N2 on one and other side of the cell. The theoretical voltage value of concentration cell was calculated by Eq. (3), where p1 and p2 are the partial pressures of fuel on each side of the cell, respectively, and n is the number of transferred electrons. The transfer number of protons and O2− can be calculated by Eq. (4), based on the measured voltage Eobs and the theoretical voltage ENernst.

$${E}_{Nernst}=\frac{RT}{nF}ln\left(\frac{p1}{p2}\right)$$
(3 )
$${t}_{ion}=\frac{{E}_{obs}}{{E}_{Nernst}}$$
(4 )

The proton transfer number of HCC shown in Fig. 8(a) is 42.8%-63.5% at 400–550 °C. As shown in Fig. 8(b), the oxygen ion transfer number of OCC is 16.9%-27.1%. This result indicates that proton conduction dominates in pure LCP with the presence of oxygen ion conduction. In the results of Fig. 6, the O–H stretching vibration peak at 2922 cm−1 in the NCAL-Ni/LCP/NCAL-Ni cell indicates that the proton conduction may be through the formation of O–H bonds between protons and oxygen atoms on the LCP surface. It is also one of the main pathways of proton conduction in the similar Grotthuss model. Figure 7(d–f) show that in LCP before the test, Ce element is mainly present at + 4 valence, La element is mainly present at + 3 valence, while Pr element is present at both + 3 and + 4 valence. In the fluorite structure of ceria-based materials, Ce4+ is doped by La3+ and Pr3+/Pr4+ in high proportion, forming many oxygen vacancies. These oxygen vacancies provide a conduction pathway for O2−. However, the low oxygen ion transfer number is due to the effect of low temperature below 550 °C on the one hand, and on the other hand, the doping of rare-earth cations with large radii such as La3+ and Pr3+ make the formed oxygen vacancies unstable as revealed by the study of Nakayama et al. using the first principle [53]. In fuel cell conditions, both H+ and O2− can be sourced externally from hydrogen and oxygen. In contrast, in HCC or OCC tests, different gas atmospheres will result in different ion conduction mechanisms with the fuel cell atmosphere [54].

Fig. 8
figure 8

The observed OCV, theoretical OCV, and transfer number from (a) hydrogen concentration cell and (b) oxygen concentration cell.

Full size image

According to the combined analysis, the authors have proposed the conduction mechanism in NCAL-Ni/LCP/NCAL-Ni fuel cell, as shown in Fig. 9. During the cell operation, the LiOH/Li2CO3 mixture in molten or partially molten state is automatically coated on the surface of the LCP particles forming a core–shell structure [55]. The proton has a smaller radius and lighter mass and is believed to form a hydrogen bond with the neighboring lattice O atom [56]. In the two-phase interfacial region, proton conduction takes place mainly in two ways: 1. H–O bonds are formed among protons and oxygen atoms of CO3 group; 2. H–O bonds are formed among protons and oxygen atoms on the LCP surface. These H–O bonds are continuously forming and breaking, allowing proton conduction along this high-speed pathway in the direction of hydrogen concentration. On one hand, the rotational motion of the CO3 group enables the proton to pass across the CO3 groups through the H–O bond. On the other hand, proton passes from one oxygen atom to another one on the surface of the LCP particles via the H–O bond. The stretching and bending vibrations of CO32− and the stretching vibrations of O–H demonstrate these processes. Expectedly, the protons may form both H–O bonds to the CO3 group and the oxygen atom on LCP surface. In this case, the rotating CO3 group may promote the transfer of protons across the oxygen atoms on the LCP surface. For the LCP electrolyte with low oxygen ion conduction, when NCAL is used as the electrode, the LiOH/Li2CO3 mixture formed by the reduction of the anode greatly promotes the proton conduction, making the contribution of O2− in the NCAL-Ni/LCP/NCAL-Ni fuel cell almost negligible. In addition, the molten constituent covers the surface of LCP particles, which may hinder the internal electron conduction through the bulk electrolyte and can help to avoid internal short circuit [31].

Fig. 9
figure 9

Conduction mechanism in NCAL-Ni/LCP/NCAL-Ni fuel cell.

Full size image

4 Conclusions

LCP based low-temperature CFC using NCAL as the electrodes, has achieved a peak power density more than 750 mW/cm2 at 520 °C. Adding BCZY blocking layer helps to demonstrate that the NCAL-Ni/LCP/NCAL-Ni cell to become dominated with proton conduction. In the molten state LiOH/Li2CO3 mixture on the surface of LCP particles, protons form H–O bonds with oxygen atoms in CO3 groups and meanwhile form H–O bonds with oxygen atoms on the surface of LCP, establishing a high-speed pathway for proton transfer. The high-speed pathway at the interface between the LCP particles and the LiOH/Li2CO3 mixture has a great influence on the enhancement of the ionic conductivity. It was demonstrated via HCC/OCC tests that the LCP electrolyte is a mixed conductor with proton and O2− when there is no LiOH/Li2CO3 mixture produced by the NCAL electrode. LiOH/Li2CO3 mixture promoted the proton conduction through an LCP electrolyte layer to a promising level that oxygen ion conduction was almost negligible through the NCAL-Ni/LCP/NCAL-Ni device.