• La-based perovskites mixed with carbon black were used as cathode catalysts in Li–O2 batteries.

Lithium-ion batteries (LIBs) are widely utilized in laptops, smartphones, power banks, renewable energy systems, and electric vehicles worldwide (Lu et al., 2013; Li et al., 2018; Dunn et al., 2021; Muruganantham et al., 2022; Zhu et al., 2022; Anirudha et al., 2023). However, the density of energy storage in LIBs is still insufficient to fulfill the increasing energy requirements for advanced electric transportation (Wei et al., 2021; Luo et al., 2022). The recently developed metal–air batteries, including Zn–oxygen, Na–oxygen, and Li–oxygen (Li–O₂) batteries, have attracted considerable interest because of their advantages such as low cost, high flexibility, and high theoretical energy density (Hardin et al., 2013; Rahman et al., 2013; Kang et al., 2022; Li et al., 2022; Salado and Lizundia, 2022; Yang et al., 2022). In addition, in comparison with LIBs, Li–O₂ batteries offer much higher gravimetric energy density, which could reach the theoretical value of 11,680 Wh kg⁻¹ because oxygen electrodes can directly use oxygen from the surrounding environment while discharging; therefore, oxygen does not need to be stored within the battery (Girishkumar et al., 2010). However, Li–O₂ batteries have been facing several serious challenges, including high overvoltage, poor rate capacity, and short cycle life, which are mainly caused by the sluggish dynamics of the air electrode during Li₂O₂ formation (2Li⁺ + 2e⁻ + O₂ → Li₂O₂, oxygen reduction reaction (ORR)) and Li₂O₂ decomposition [Li₂O₂ → 2Li⁺ + 2e⁻ + O₂, oxygen evolution reaction (OER)] (Thapa et al., 2010; Thapa and Ishihara, 2011; Pan et al., 2019; Cui et al., 2021; Yin et al., 2021; Zhan et al., 2021). Furthermore, battery performance is also affected by the components present in ambient air, resulting in significantly lower practical specific energy (Zhang et al., 2018).

Recently, Li–O₂/CO₂ batteries have attracted considerable attention because they capture and utilize carbon (Takechi et al., 2011; Lim et al., 2013; Yin et al., 2017; Zou et al., 2019; Chen et al., 2020; Savunthari et al., 2021; Iputera et al., 2022; Wang et al., 2023; Wu et al., 2023). In addition, during discharge, Li⁺ ions react with O₂ and CO₂ to produce Li₂O₂ and Li₂CO_3. These products can improve the battery capacity by 289% compared to batteries operating on 100% O₂ (Takechi et al., 2011). Therefore, the generation of Li₂CO₃ is beneficial for battery capacity. At the same time, Li₂O₂ and Li₂CO₃ that precipitate on the air electrode surface during discharge are difficult to completely decompose during charge (Gallant et al., 2012). These discharge products (Li₂O₂ and Li₂CO₃) block the pores of the air electrode and thus cause performance degradation because they hinder air supply and liquid electrolyte diffusion (Liu et al., 2017; Zhao et al., 2018). Hence, it is necessary to promote Li₂CO₃ generation on oxygen electrodes during discharge and decomposition of discharge products during charge.

Because reactions on the air electrode significantly affect battery performance, many studies have focused on enhancing the electrocatalytic activity of the oxygen electrode. Therefore, catalyst addition to oxygen electrodes in Li–O₂ batteries is necessary to improve the electrochemical activities of ORR during discharge and OER during charging (Thapa et al., 2010; Thapa and Ishihara, 2011; Pan et al., 2019; Yin et al., 2021). Currently, Pt-based catalysts are considered excellent catalysts for the ORR and OER (Wu and Yang, 2013; Liu et al., 2019), and IrO₂ is known to be the best catalyst for the OER (Lee et al., 2012). Although these materials show excellent performance, their high cost limits their application in Li–O₂ batteries. Recently, some studies have demonstrated that perovskite catalysts are beneficial for oxygen reduction during discharge and the decomposition of discharge products during charge (Xu et al., 2013; Du et al., 2014; Jin et al., 2014; Zhang et al., 2014; Ma et al., 2020; Du et al., 2021). Oxygen vacancies in the perovskite LaCoO₃ have been reported to enhance the bifunctional catalytic activity (ORR and OER) because of the valence electron transformation of the Co ions (Du et al., 2021). In addition, the battery discharge capacity and long-term cycling stability were also remarkably increased. Zhang et al. used porous LaNiO₃ as a catalyst for an air electrode and improved the discharge capacity from 2,545 to 3,407 mAh g⁻¹ (Zhang et al., 2014). However, it is unclear how the catalysts affect the discharge products and battery capacity.

In this study, La_1-xSr_xCoO_3-δ (x = 0.1, 0.3, and 0.5) and La_0.9Sr_0.1YbO_3-δ were used as catalysts for oxygen electrode to determine their effects on Li–O₂ battery capacity. Additionally, the performances of two types of carbon black as cathode substrates were tested. Finally, the discharge products were analyzed to explain the differences in battery performance for different cathode materials. The findings of this study are expected to facilitate the development of catalysts for air electrodes in Li–O₂ batteries.

The phase purity of L9SC, L7SC, L5SC, and L9SYb powders was verified by analyzing their crystal structures using XRD (Figure 1). The diffraction peaks of L9SC, L7SC, L5SC, and L9SYb phases match well with the corresponding reference patterns thus confirming that each as-prepared material is composed of a single major phase. In the case of La_1-xSr_xCoO_3-δ, the formation of impurity phases with increasing Sr doping is not observed, as shown in Figure 1. Therefore, the as-prepared materials have sufficiently high purities and crystallinities for use as catalysts.

Figure 2 shows the charge–discharge curves for Li–O₂ in the voltage range of 4.3–2.4 V at a constant current of 0.1 mA cm⁻² at room temperature. First, only carbon black (Vulcan XC500 and Super P) electrodes without catalysts were tested in one-cycle charge-discharge tests (Figure 2A). The charge–discharge capacity results are listed in Table 1. These results indicate that the performance of Super P is superior to that of Vulcan XC500. In particular, the discharge capacity of Super P is almost twice that of Vulcan XC500. Super P, with a large specific surface area, could provide more reaction sites for Li⁺ ions from the anode (Wang et al., 2018). Subsequently, as shown in Figure 2B, L5SC was used as a catalyst and mixed with Vulcan XC500 and Super P, and the obtained cathode was tested. The results show that the addition of L5SC improved both the charge and discharge capacities. In the case of Vulcan XC500, the charge capacity increased from 926 to 2,247 mAh g⁻¹, and the discharge capacity increased from 1,394 to 2,108 mAh g⁻¹. This indicates that L5SC can promote the OER and ORR in Li−O₂ batteries. Furthermore, in the case of Super P, the charge and discharge capacities were significantly enhanced from 955 to 5,500 mAh g⁻¹ and from 2,253 to 6,032 mAh g⁻¹, respectively. Super P has a large specific surface area (62 m²/g) for mixing with L5SC; therefore, the addition of L5SC produced a greater effect on OER and ORR of the Li−O₂ battery, resulting in excellent Li−O₂ battery performance.

The capacities of cathodes made of Vulcan XC500 mixed with different catalysts (L5SC, L7SC, and L9SC) were measured using one-cycle charge–discharge tests (Figure 3; Table 2). According to the results, the capacities show the same tendencies during charge and discharge. The L5SC/XC500 electrode shows the best performance, and the L7SC/XC500 electrode shows a higher capacity than the L9SC/XC500 electrode. These results suggest that catalyst activity could be promoted by increasing the number of oxygen vacancies. It has been previously shown that the catalytic activity of Sr-doped LaCoO₃ in the OER could increase with the number of oxygen vacancies (Mefford et al., 2016; Lu et al., 2019). Furthermore, the number of oxygen vacancies in perovskites has also been reported to be related to catalytic performance for ORR (Gayen et al., 2020; Ji et al., 2020). Therefore, it was considered that the charge and discharge capacities increased because the oxygen vacancies in Sr-doped LaCoO₃ promoted the OER and ORR of the Li–O₂ batteries.

The capacities of cathodes made of Super P mixed with different catalysts (L5SC, L7SC, L9SC, and L9SYb) were measured using one-cycle charge–discharge tests (Figure 4; Table 3). According to the results, the capacities show the same tendencies during charge and discharge. The L5SC/Super P demonstrated the highest capacity in this study. This is because L5SC has the highest catalytic activity in the OER and ORR, and Super P has a larger specific surface area than Vulcan XC500. In addition, compared with Sr-doped LaCoO₃, L9SYb shows relatively low catalyst activity in OER and ORR. In fact, the OER and ORR performances of L9SYb were previously investigated in only a few studies because of its low catalytic activity. Therefore, the effects of the catalysts can be easily elucidated by comparing the differences between the performances of L5SC and L9Yb.

Figure 5 shows XRD results for L5SC/Super P and L9SYb/Super P before and after one-cycle charge–discharge test. In Figure 5, L5SC and L9SYb are labeled with blue triangle and red cross, respectively, and Li₂CO₃ and Li₂O₂ are labeled with blue circle and orange diamond, respectively. The comparison of Figures 5A, B shows that more Li₂CO₃ was generated on the L5SC/Super P cathode than on the L9SYb/Super P cathode. This indicates that L5SC promoted the reaction between C (from electrolyte solvent or cathode material), O₂, and Li⁺ to generate Li₂CO₃ during discharge. According to the literature, the generation of Li₂CO₃ can increase the capacity of Li/CO₂–O₂ batteries (Takechi et al., 2011; Zou et al., 2019). Yin et al. suggested that two electrons are involved in the generation of Li₂CO₃ (Yin et al., 2017). At the same time, only one electron is involved in the formation of Li₂O₂. Therefore, the generation of Li₂CO₃ could significantly increase the discharge capacity of L5SC/Super P catalysts. We could not determine the amount of Li₂CO₃ produced through XRD measurements; therefore, we conducted CV experiments on Li–O₂ batteries with L5SC/Super P and L9SYb/Super P cathodes (Figure 6).

CV curves for Li–O₂ batteries with L5SC/Super P and L9SYb/Super P were obtained in the voltage range of 2.0–4.5 V at a constant current of 0.1 mA cm⁻² (Figure 6). To observe complete reduction and oxidation peaks in CV curves, they were obtained in the potential range of 4.5–2.0 V. The ORR peak in CV curves corresponds to the generation of Li₂O₂ and Li₂CO₃, while the OER peak corresponds to the evolution of O₂ and CO₂ from Li₂O₂ and Li₂CO₃. In addition, according to previous studies, the theoretical voltage for Li₂O₂ oxidation is 2.90 V for the reaction Li₂O₂ → 2Li⁺ + 2e⁻ + O₂ (Lim et al., 2013). At the same time, the theoretical voltage for the oxidation of Li₂CO₃ is 3.82 V through the reaction Li₂CO₃ → 2Li⁺ + 2e⁻ + 1/2O₂ + CO₂ (Lim et al., 2013). Therefore, Li₂CO₃ is more chemically stable than Li₂O₂. The mentioned voltage levels (2.9 and 3.82 V) are indicated in Figure 6. In Figure 6, for both cathodes, the current density during oxidation increased at 2.9 V and reached a maximum at 4.5 V. In particular, for L5SC/Super P, the current density significantly increases in the voltage range of 3.82–4.5 V. In addition, the CV curve area of the L5SC/Super P battery is larger than that of the L9SYb/Super P battery. This indicates that a greater amount of generators in the L5SC/Super P cathode was oxidized and decomposed to Li ions and gases during oxidation. Above 3.82 V this phenomenon is more pronounced. This suggests that more Li₂CO₃ was produced in the L5SC/Super P cathode. Therefore, the L5SC/Super P cathode demonstrated the best discharge capacity in this study because of promoted Li₂CO₃ generation.

The electrochemical impedance spectroscopy (EIS) results during charge and after discharge are shown in Figure 7. For L5SC/Super P, the ohmic and polarization resistances significantly increase after discharge. This indicates that a greater amount of non-conductive Li₂O₂ and Li₂CO₃ was generated in L5SC/Super P, which blocked electronic conduction and O₂ diffusion. As shown in Figures 7A, B, the polarization resistance of L9SYb/Super P is larger than that of L5SC/Super P, which also indicates that L9SYb has lower catalytic activity than L5SC.

In this study, the charge–discharge performances of L9SC, L7SC, L5SC, and L9SYb catalysts for Li-O₂ batteries were investigated. In addition, the charge–discharge performances of two types of carbon black (XC500 and Super P) were determined. According to the literature, La_1-xSr_xCoO_3-δ is as a superior OER and ORR catalyst compared to La_0.9Sr_0.1YbO_3-δ. Additionally, an increase in the number of oxygen vacancies and an increase in the surface area by blending with Super P carbon have been previously reported to improve the catalytic activity in cathodic reactions in Li–O₂ batteries. In this study, the L5SC/Super P air electrode showed the best charge and discharge capacities corroborating the abovementioned findings from the literature. In addition, the type of generator (Li₂O₂ or Li₂CO₃) could be considered a factor that affects the discharge capacity. The results of this study show that the battery capacity increases with the amount of Li₂CO₃ generated. The L5SC/Super P cathode material promoted the production of Li₂CO₃ and thus showed excellent performance. At the same time, based on XRD and CV results, L9SYb/Super P showed low Li₂CO₃ yields, which also indicated that it did not promote the reaction of Li⁺ ions with oxygen.