Taking advantage of Li-evaporation in LiCoO2 as cathode for proton-conducting solid oxide fuel cells

LiCoO2, a widely used electrode material for Li-ion batteries, was found to be suitable as a cathode material for proton-conducting solid oxide fuel cells (H-SOFCs). Although the evaporation of Li in LiCoO2 was detrimental to the Li-ion battery performance, the Li-evaporation was found to be beneficial for the H-SOFCs. The partial evaporation of Li in the LiCoO2 material preparation procedure led to the in-situ formation of the LiCoO2+Co3O4 composite. Compared to the cell using the pure phase LiCoO2 cathode that only generated moderate fuel cell performance, the H-SOFCs using the LiCoO2+Co3O4 cathode showed a high fuel cell performance of 1160 mW·cm−2 at 700 °C, suggesting that the formation of Co3O4 was critical for enhancing the performance of the LiCoO2 cathode. The first-principles calculation gave insights into the performance improvements, indicating that the in-situ formation of Co3O4 due to the Li-evaporation in LiCoO2 could dramatically decrease the formation energy of oxygen vacancies that is essential for the high cathode performance. The evaporation of Li in LiCoO2, which is regarded as a drawback for the Li-ion batteries, is demonstrated to be advantageous for the H-SOFCs, offering new selections of cathode candidates for the H-SOFCs.


Introduction 
The current energy and environmental problems require the development of sustainable technologies and devices [1][2][3][4][5], and solid oxide fuel cells (SOFCs) that can convert chemical energies into electricity receive considerable attention [6]. The traditional SOFCs have to work at high temperatures (above 800 ℃), reducing the lifetime of the fuel cells [7,8]. Therefore, the development of the SOFCs working at lower temperatures (below 700 ℃) is highly desirable, and proton-common points. Many of these electrode materials are ceramic oxides and contain transition metals as their major compositions. Therefore, it would be reasonable to expect that some electrode materials for the Li-ion batteries could show decent performance for the H-SOFCs as well. More importantly, many of the electrode materials for the Li-ion batteries have been commercialized, and the application of the Li-ion battery electrode materials for the H-SOFCs could advance the development of the H-SOFCs.
In fact, some lithiated oxide cathodes have been used for the SOFCs, but only moderate performance has been achieved [20]. Very recently, LiCo 0.6 X 0.4 O 2 (X = Mn, Sr, and Zn) materials were applied as the cathodes for the H-SOFCs, and the optimal polarization resistance (R p ) is 0.55 Ω·cm 2 at 700 ℃ [21], which is inferior to the newly developed high-performing cathodes. These results imply that the traditional lithiated oxide cathodes might not be a good choice as the cathode candidate for the H-SOFCs. However, all of these have focused on using pure phase lithiated oxide cathodes and evaluated their suitability as the cathodes for the H-SOFCs. The lithiated oxide cathodes, such as the classical LiCoO 2 , contain a large amount of Li element, and the Li element tends to evaporate during the synthesis procedure due to the high-temperature calcination. Therefore, an excess amount of Li should be added during the preparation process to compensate for the Li-evaporation and obtain the pure phase LiCoO 2 [22,23]. It is understandable that the Li loss could be detrimental to the performance of the Li-ion batteries, and the Li-evaporation has to be compensated [24]. However, the working mechanism of the H-SOFCs is different from that of the Li-ion batteries, which does not require the mobility of the Li-ions. It should be noted that the performance of the lithiated oxide cathodes for the H-SOFCs is reported to be only moderate [20,21], even though efforts are devoted to preparing the pure phase lithiated oxide cathodes. One may wonder about the performance of the lithiated oxide cathode for the H-SOFCs if the evaporated Li is not compensated, as no mobility of Li is needed in the H-SOFCs. Therefore, the classical Li-ion battery electrode LiCoO 2 was used in this study as the cathode for the H-SOFCs. The phase compositions for LiCoO 2 with and without the Li compensation were studied, and their performance for the H-SOFCs was investigated and compared, aiming to provide a new cathode system for the H-SOFCs.

Materials and method
LiCoO 2 was prepared by using Li 2 CO 3 and Co(NO 3 ) 3 as starting materials [25]. Li 2 CO 3 was dissolved in dilute nitric acid. According to Refs. [22,24], 4 mol% Li-excess was needed to compensate for the Li-evaporation during the synthesis to form the pure phase LiCoO 2 . In contrast, the stoichiometric ratio of Li 2 CO 3 and Co(CO 3 ) 3 was used in this study, making the molar ratio of Li∶Co keep at 1∶1 for the starting materials. In other words, no Li-excess was applied to compensate for the Li-evaporation during the synthesis procedure. The as-prepared powders were calcined at 850 ℃ for 3 h, followed by the phase examination using X-ray diffraction (XRD). The XRD was performed at a scanning rate of 3 (°)·min -1 with Cu Kα radiation. The morphologies of the powders were characterized by scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HR-TEM). The stability of the powders was examined by treating the powders in both CO 2 and H 2 O-containing atmospheres at a high temperature. The CO 2 and H 2 O concentrations were 20% and 30%, respectively. Then, the XRD was used to analyze the phase of the powders before and after the treatments.
To evaluate the efficiency of the LiCoO 2 cathodes for the H-SOFCs, the cathode slurry was deposited onto the electrolyte, and then being co-fired at 800 ℃ for 10 min. The electrolyte material used in this study was BaCe 0.7 Zr 0.1 Y 0.2 O 3−δ (BCZY), and the anode was NiO-BCZY. The details for the preparations of the electrolyte powders and half-cells can be found in Ref. [26]. The single cells were tested by using humidified H 2 (3% H 2 O) as the fuel. IV and electrochemical impedance spectroscopy (EIS) measurements were carried out by an electrochemical workstation (Squidstat Plus, Admiral Instrument) under open circuit voltage (OCV) condition. The frequency range was from 1 MHz to 0.1 Hz with an amplitude of 5 mV. The morphologies of the tested cells were observed by the SEM.

Results and discussion
Figure 1(a) shows the XRD pattern for LiCoO 2 without Li-excess after being fired at 850 ℃ for 3 h. The LiCoO 2 powders with 4 mol% Li-excess were also prepared via the same procedure, and its XRD is shown in Fig. 1(a) as well. A pure phase is obtained with the LiCoO 2 powders with 4 mol% Li-excess. No other Li-related compounds, such as Li 2 O or Li 2 CO 3 , can be detected. This result suggests that the excess Li compensates for the Li-evaporation during the calcination procedure, agreeing with Ref. [24] that 4 mol% Li-excess is usually used to offset the Li-evaporation during the synthesis process, making LiCoO 2 achieve a pure phase. In contrast, if one looks at the XRD pattern of the LiCoO 2 powders without Li-excess, the main peaks correspond to the LiCoO 2 phase. However, some extra peaks corresponding to Co 3 O 4 can be found, suggesting that the evaporation of Li in LiCoO 2 could lead to Co 3 O 4 as the second phase if the Li-evaporation is not compensated by excess Li sources. Figure 1(b) shows the SEM image of the LiCoO 2 powders without Li-excess, and one can see that large sheet-like powders are LiCoO 2 , and the small particles are Co 3 O 4 . The TEM image of the powders (Fig. 1(c)) shows a similar result that LiCoO 2 and Co 3 O 4 form the composite. The HR-TEM image shown in Fig. 1 4 . The C element detected as a carbon tape was used in the SEM observations, and the Au element detected as Au sputtering was used to increase the conductivity of the samples. Therefore, it can be concluded that the LiCoO 2 powders without Li-excess finally becomes the LiCoO 2 +Co 3 O 4 composite powders. Figure 1(e) shows the scheme of the LiCoO 2 powder preparation. After the sol-gel procedure, the precursor was calcined at 850 ℃ for 3 h. When there is excess Li, the Li-evaporation can be compensated with the production of the pure phase LiCoO 2 . In contrast, when the stoichiometric ratio of Li and Co is used, no Li-excess is used, and thus no compensation of the Li-evaporation happens for the final product, leading to the formation of the LiCoO 2 +Co 3 O 4 composite. The compensation of Li to obtain the pure phase LiCoO 2 is understandable for the Li-ion batteries as the Li ions are required to move during the charge and discharge processes for the Li-ion batteries, and the deficiency in the Li content could lead to the reduced capacity of the batteries [32]. However, the Li-evaporation and the formation of Co 3 O 4 might promote the cathode performance in the fuel cells as Co 3 O 4 is reported to facilitate the oxygen reduction reaction (ORR) at the cathode for the SOFCs and improve the cathode performance [33]. Therefore, it is reasonable to expect that the LiCoO 2 +Co 3 O 4 composite could give a good performance for the H-SOFCs. Figure 2(a) shows the IV and power density curves of a single H-SOFCs using the LiCoO 2 +Co 3 O 4 composite as the cathode. The cell's peak power densities (PPDs) are 1160, 896, and 603 mW·cm −2 at 700, 650, and 600 , respectively. The PPDs tested at ℃ a temperature higher than 700 ℃ are not reported here, as obvious oxygen-ion conductions appear for protonconducting oxides above 700 ℃ [4].
show the cross-sectional view of the tested cell. A tri-layer structure can be observed, and the electrolyte layer contacts well with the cathode and anode. The fuel cell performance is much higher than those of the H-SOFCs using the traditional cathodes and comparable to or even higher than those of many H-SOFCs using recently developed cathodes, as indicated in Table 1 [34][35][36][37][38][39][40][41][42]. This result suggests that the LiCoO 2 +Co 3 O 4 composite is a high-performance cathode for the H-SOFCs.
The cell performance between the pure phase LiCoO 2 cathode and the LiCoO 2 +Co 3 O 4 cathode was further compared. Figure 3(a) shows the performance of the cell with the LiCoO 2 cathode and the LiCoO 2 +Co 3 O 4 cathode tested at 700 ℃. The PPD for the cell using the LiCoO 2 cathode is 879 mW·cm -2 , and the PPD value increases to 1160 mW·cm -2 for the cell using the LiCoO 2 +Co 3 O 4 cathode under the same testing condition. The EIS analysis confirms that the in-situ formation of Co 3 O 4 in the LiCoO 2 cathode is beneficial for the reduction of the R p . One can see from Fig. 3(b) that the ohmic resistance (R ohmic ) is similar, being 0.148 and 0.141 Ω·cm 2 at 700 ℃ for the cells using the LiCoO 2 cathode and LiCoO 2 +Co 3 O 4 cathode, respectively. In contrast, an obvious difference in the R p can be observed. The R p for the pure phase LiCoO 2 cathode is 0.123 Ω·cm 2 , and the R p value decreases to 0.043 Ω·cm 2 for the cell using the LiCoO 2 +Co 3 O 4 cathode under the same testing condition. The shrinkage in the R p is about 65% with the use of the LiCoO 2 +Co 3 O 4 cathode instead of the pure phase LiCoO 2 cathode. The R p for the LiCoO 2 +Co 3 O 4 cathode is even smaller than those for many recently reported cathodes for the H-SOFCs, as indicated in Table 1. It should be noted that only LiCoO 2 or LiCoO 2 +Co 3 O 4 was used as the cathode, and the LiCoO 2 -based material was not mixed with the BCZY to form the composite cathode. The close R ohmic for both cells is expected because both cells are identically prepared except for the different cathodes used. In addition, the same Ni-BCZY anode is used for both cells, so the considerable gap in the R p should come from the cathode. The performance enhancement for the LiCoO 2 + Co 3 O 4 cathode is not only observed at 700 ℃ but also  for LiCoO 2 is calculated to be 3.76 eV, which is relatively large compared with those for other cathode materials for the SOFCs. This result indicates that the V O formation in the pure phase LiCoO 2 has to overcome a high energy barrier. This result could explain that the reported lithiated oxide cathodes only show moderate performance for the SOFCs. When the LiCoO 2 material is prepared without an excess of Li, the final product is the LiCoO 2 +Co 3 O 4 composite, and Fig. 4(a) shows the DFT-calculated configuration of the LiCoO 2 /Co 3 O 4 supercell. Both the www.springer.com/journal/40145 at the LiCoO 2 side has the lowest energy, which means that the V O is more favourable to form at the LiCoO 2 side. The O V E at the LiCoO 2 /Co 3 O 4 interface is calculated to be 0.46 eV, which is about one order of magnitude lower than that at the pure phase LiCoO 2 , indicating that the V O formation is thermodynamically more favourable with the appearance of Co 3 O 4 . The oxygen species of LiCoO 2 and LiCoO 2 + Co 3 O 4 are investigated by X-ray photoelectron spectroscopy (XPS), and the results are shown in Figs. 4(b) and 4(c), respectively. The ratio between the adsorbed oxygen and the lattice oxygen reflects the V O content [50,51], and the ratio is 0.69 and 1.02 for LiCoO 2 and LiCoO 2 + Co 3 O 4 , respectively, suggesting that the V O content in LiCoO 2 +Co 3 O 4 is higher than that in the pure phase LiCoO 2 that agrees with the DFT calculation results. The iodometric titration method was further used to measure the oxygen vacancy content in LiCoO 2 and LiCoO 2 +Co 3 O 4 . It is found that the oxygen vacancy concentration in the pure phase LiCoO 2 is 0.13, and the value increases to 0.22 for the LiCoO 2 +Co 3 O 4 composite, suggesting an increase in the oxygen vacancies for LiCoO 2 +Co 3 O 4 compared with those for the pure phase LiCoO 2 . The result is consistent with the DFT calculations and XPS analysis.
In addition, the distance between Co 3d and O 2p in LiCoO 2 +Co 3 O 4 is reduced compared with that in the pure phase LiCoO 2. Figures 5(a)   As the cathode for the H-SOFCs, another primary concern is the stability of the material, as this parameter is essential for its potential practical applications. The LiCoO 2 +Co 3 O 4 composite was treated in the CO 2containing atmosphere at 600 ℃, and the result is shown in Fig. 6(a). One can see that there is no extra peak under the treatment of 12 h, suggesting that the LiCoO 2 + Co 3 O 4 composite has good chemical stability against CO 2 . The good chemical stability of the LiCoO 2 + Co 3 O 4 composite can be expected that no high-basicity element (such as Ba) is involved in the material, restricting its reaction with acid gases (such as CO 2 ). The LiCoO 2 +Co 3 O 4 composite also presents good chemical stability against steam. The composite powders were treated in a 30% H 2 O-containing air atmosphere at 600 ℃ for 10 h. Figure 6(b) shows the XRD patterns for the LiCoO 2 +Co 3 O 4 composite before and after the treatment. The XRD pattern remains unchanged after the treatment, suggesting that there is no evident reaction between the LiCoO 2 +Co 3 O 4 composite and H 2 O, thus demonstrating its excellent stability against H 2 O at high temperatures. If we enlarge the XRD patterns before and after the treatment, one can see that the XRD peak slightly shifts to a lower angle after the treatment (Fig. S4 in the ESM), suggesting the expansion of the lattice. This phenomenon is also reported in Ref. [53], which is due to the adsorption of water in the lattice of the material. Although the peak shift is not very profound because the powders are treated at 600 ℃, and the hydration degree decreases with the increasing temperatures [54], the peak shift can still be observed. The excellent chemical stability of the LiCoO 2 +Co 3 O 4 composite also leads to good fuel cell stability under the operation condition. Figure 6(c) shows the stability test result for the fuel cell. The cell works in a stable way for more than 500 h, suggesting that the LiCoO 2 +Co 3 O 4 composite provides high performance for the H-SOFCs and integrates the stability of the single-cell well.

Conclusions
In this study, we took advantage of the Li-evaporation in LiCoO 2 to form the LiCoO 2 +Co 3 O 4 composite as a cathode for the H-SOFCs. The Li-evaporation in LiCoO 2 was regarded as a drawback for the Li-ion batteries that required excess Li in the preparation procedure for the compensation. However, LiCoO 2 without the Li compensation turned out to be the LiCoO 2 +Co 3 O 4 composite after the calcination, which was applied as the cathode for the H-SOFCs. The LiCoO 2 +Co 3 O 4 composite generated higher fuel cell performance and lower R p for the H-SOFCs compared with the pure phase LiCoO 2 cathode. The DFT calculations coupled with experiments indicated that the appearance of Co 3 O 4 due to the Li-evaporation in LiCoO 2 was beneficial to the formation of the oxygen vacancies, facilitating the cathode reactions. The improved oxygen vacancy content and the enhanced catalytic activity induced by the formation of the Co 3 O 4 nanoparticles contribute to better cathode performance and larger fuel cell output. In addition, the high electrochemical performance of LiCoO 2 +Co 3 O 4 for the H-SOFCs did not impair its stability. The LiCoO 2 +Co 3 O 4 composite presented excellent stability in both CO 2 and H 2 O. In addition, the good stability of the single-cell using the LiCoO 2 + Co 3 O 4 cathode under the working condition was also demonstrated, suggesting that LiCoO 2 +Co 3 O 4 composite integrated both high performance and good stability. This study bridges the electrode materials between the Li-ion batteries and the H-SOFCs and indicates that the disadvantageous feature in the Li-ion batteries may intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.