Abstract

Increasing the content of anion vacancies may yield significant improvement in the overall electrochemical energy-storing performance of perovskite materials, where the vacancy sites act as highly favorable ion diffusion paths. However, a detailed study on energy storage mechanism at binary cation sites under the anion deficiency should be further explored in supercapacitor electrode materials. In this study, a simple hydrothermal method and hydrogen gas exposure processes were used to generate oxygen vacancies in the crystal of BiFeO3 (BiFeO3-X) to enhance the overall electrochemical properties. At a current density of 1 A g−1, the BiFeO3-X supercapacitor electrode exhibits a large specific capacitance (461.9 F g−1, 923.8 mF cm-2, and 145.3 mAh g-1) and a high cycling stability (94.4%) after 2,000 cycles. Electrochemical analysis reveals that the oxygen vacancy sites can further increase the electrochemical activity of Bi sites, which is mostly suppressed in the pure crystal lattice, resulting in synergistic energy storage behavior of binary cation sites.

1. Introduction

Compared with other energy-storing devices, electrochemical energy storage (EES) devices are considered more promising candidates for the next-generation energy storage applications with high power density and excellent cycling stability [1, 2]. EES devices can store large amounts of charges via two different mechanisms: (1) ion attachment and capture at the surface of electrodes and (2) fast and reversible electrochemical Faradaic redox reactions on the lattice of materials [3, 4]. Therefore, the overall energy-storing performance can be changed by selecting suitable active electrode materials, designing favorable ion-electrode interfaces (surface area, porous structure, and ion diffusion pathways), and maximizing the surface Faradaic redox reactions. Transition metal-based cation materials are typically used for EES device electrodes; in these materials, the transition metal and electrolyte ion can provide additional charges through Faradaic redox reactions on the lattice of the transition metal sites [57]. This is especially true of transition metal oxides that can provide high theoretical capacitances, owing to their multivalent oxidation states, thereby inducing rich electrochemical Faradaic redox reaction sites [810].

Binary transition metal-based perovskite materials, the most extensively researched ABO3-structured materials (where A and B are transition metal cations), are considered promising candidates for electrochemical energy-storing electrodes [11, 12]. These materials generally present a higher electrical conductivity and richer oxidation redox states than the single-component transition metal compounds [13, 14]. Both transition metal atoms can serve as active sites for the Faradaic redox reactions. Moreover, the ion diffusion and ionic kinetics can be strongly influenced by the selection of cations on the A and B sites [15, 16]. Within the ABO3 structure, the large inner-crystal space of the perovskite materials can yield quite favorable ion diffusivity, enabling a large power-operation capability. These properties are strongly related to the physical ion paths on the ABO3 structure, and by controlling the lattice structure of the perovskite materials, the overall energy storage performance can be significantly modified and optimized [17, 18]. Moreover, the newly suggested energy-storing mechanisms can be utilized by changing the crystal lattice space of ABO3 [1921]. Anion vacancy engineering represents an effective method of controlling the lattice structure. The anion vacancies can increase the electrical conductivity of materials, act as favorable ion diffusion paths, and increase the number of ion adsorption and active sites on the transition metal materials. However, improvements and a detailed analysis of perovskite materials are required [22].

In this study, we synthesized a BiFeO3-X electrode using a simple hydrothermal method and an annealing process. Through exposure to hydrogen gas (H2), anion oxides can be converted to H-included byproducts and can generate anion vacancy sites in perovskite structures. The BiFeO3-X electrode, with induced vacancy sites, exhibited a promising electrochemical performance. Compared with a BiFeO3 electrode, the BiFeO3-X electrode exhibited a 1.5 times higher capacitance of 461.9 F g−1 (923.8 mF cm-2, 145.3 mAh g-1) at a current density of 1 A g−1 and a high cycling stability of 94.4% after 2,000 charge–discharge cycles. The cathodic peak current densities and electrochemical impedance spectroscopy (EIS) curve of BiFeO3-X revealed an enhanced surface-controlled activity, a low charge transfer resistance (), and a steep Warburg impedance region, suggesting that the BiFeO3-X electrode possesses favorable electrochemical properties. A significant enhancement in capacitance relies strongly on the increased electrochemical activity of the Bi sites. According to the cyclic voltammetry (CV) curve, the redox reactions and charge generation contribution from the Bi sites are significantly enhanced, whereas the same activity is suppressed within the pure lattice of BiFeO3-X. Therefore, by inducing anion vacancies, the inherent electrochemical activity of transition metal sites can be significantly increased, thereby representing a promising material strategy for future energy storage applications.

2. Materials and Methods

2.1. Materials and Preparation

Precursors with 2 mM of Bi(NO3)3·5H2O (Sigma-Aldrich), 2 mM of Fe(NO3)3·9H2O (Sigma-Aldrich), and 6 mM of urea (CH4N2O, Sigma-Aldrich) were dissolved in 20 mL of ethylene glycol (C2H6O2, Sigma-Aldrich) to obtain a homogeneous solution through magnetic stirring. This solution was then subjected to solvothermal synthesis within a Teflon-lined autoclave at 140°C for 12 h. The obtained product was filtered and dried overnight. Afterward, BiFeO3 and BiFeO3-X were prepared through an annealing process at 600°C for 2 h in air and 4% H2 atmosphere, respectively. Figure 1 shows a schematic of the BiFeO3-X synthesis process.

2.2. Characterization

The morphology of BiFeO3-X was analyzed via field-emission scanning electron microscopy (FESEM; JEOL JSM-7800F Prime) performed at an acceleration voltage of 20 kV. Moreover, X-ray diffraction (XRD; Rigaku Ultima IV) conducted with a Cu Kα1 source was used to determine the crystallinity of the prepared samples. The crystal structure was further analyzed by means of spherical aberration-corrected transmission electron microscopy (Cs-TEM; JEOL JEM-ARM200F) performed at an acceleration voltage of 100 kV. In addition, energy-dispersive X-ray spectroscopy (EDS) coupled with Cs-TEM was used to determine the elemental distribution. To estimate the oxygen content, thermogravimetric curves were obtained at temperatures ranging from room temperature to 600°C (ramping rate: 1°C min−1) via thermogravimetric analysis (TGA; Pyris TGA N-1000, SCINCO). The valence states of the prepared powders were determined by means of X-ray photoelectron spectroscopy (XPS; ULVAC-PHI VersaProbe II) in the Bi 4f, Fe 2p, and O 1s regions. The surface area and pore size were analyzed using a PSA system (UPA-150, ASAP2010, AutoporeIV) that employed the Barrett-Joyner-Halenda (BJH) method.

2.3. Electrochemical Measurements

For the supercapacitor application, BiFeO3 and BiFeO3-X were loaded on a nickel foam electrode using a typical slurry process. Prior to the loading, the nickel electrode was cleaned with 1 M HCl, ethanol, and deionized water and then dried overnight to prevent surface reoxidation. A typical slurry process involved mechanically mixing 7 mg of the active material, 2 mg of carbon black (Super P, Alfa Aesar), and 1 mg of poly(vinylidene fluoride) (PVDF, Sigma-Aldrich) with 0.2 mL of N-methyl-2-pyrrolidone (NMP, Sigma-Aldrich) until a homogeneous dark-colored slurry was obtained. This slurry was drop-casted on the nickel foam electrode and dried at room temperature for 48 h. A potentiostat (Autolab PGSTAT302N) with a three-electrode system, composed of a platinum sheet (counter electrode, 1 cm2), Ag/AgCl (reference electrode), and the fabricated electrode (working electrode, 1 cm2), was used to measure the electrochemical properties of BiFeO3 (2.1 mg cm−2) and BiFeO3-X (2.0 mg cm−2). The specific capacitance (F cm-2 or F g-1 or mAh g-1) was calculated as follows: where (A) is the discharge current, is the discharge time, (cm2) is the surface area of electrode, (g) is the mass of the active materials, and is the potential range (−1.2 to 0 V versus RHE).

3. Results and Discussion

3.1. Preparation and Characterization of BiFeO3-X

BiFeO3 and BiFeO3-X were commonly obtained through a simple hydrothermal process and a subsequent annealing process. To obtain an oxygen deficient BiFeO3 (BiFeO3-X), 4 vol% of H2/Ar gas was purged into a furnace to create a reductive atmosphere, resulting in the formation of oxygen vacancy sites during the annealing process (see Figure 1(a)). The physical structures of BiFeO3 and BiFeO3-X were characterized via FESEM and Cs-TEM. The FESEM (Figure 1(b)) and Cs-TEM (Figure 1(c)) images of BiFeO3-X confirmed a nanoporous structure with a pore size of ~100 nm, resulting from the removal of anion oxygen species during the annealing process. Moreover, highly entangled grains and pore channels were observed, as shown in Figure 1(c). In contrast, the Cs-TEM image of BiFeO3 reveals a solid and nonporous structure, resulting in a smaller surface area than that of BiFeO3-X (Supplementary Materials, Figure S1a). The porosity and surface area of the samples were further confirmed through BJH analysis. Based on IUPAC classification, BiFeO3 and BiFeO3-X displayed type IV (a) curves with H1 type hysteresis loop in the N2 adsorption-desorption isotherms ascribed to mesoporous structure. However, the closed area of hysteresis loop of BiFeO3-X was clearly larger than BiFeO3 due to anion vacancies [23]. The pore volume of BiFeO3-X (surface area: 133.0511 m2 g−1) is larger than that of BiFeO3 (surface area: 59.3248 m2 g−1) (Supplementary Materials, Figures S2a and S2b). The elemental distribution of the samples was determined through EDS mapping analysis, and a homogenous distribution of Bi, Fe, and O was confirmed in both BiFeO3-X (Figure 1(d)) and BiFeO3 (Supplementary Materials, Figure S1b). Furthermore, the weight/atomic percentage of the samples were obtained (Supplementary Materials, Figure S3). From the structural analysis, compared with the weight/atomic percentage of oxygen species occurring in BiFeO3, a lower weight/atomic percentage occurred in BiFeO3-X, indicating the presence of anion vacancy sites.

The crystal and chemical structures of BiFeO3-X were verified through XRD and XPS analyses. The induction of oxygen vacancy sites was confirmed by the color change of the material (the inset image of Figure 2(a)). Moreover, the crystal structures of the BiFeO3 and BiFeO3-X samples were similar to that of the perovskite BiFeO3 (JCPDS 86-1518). However, the intensity of the BiFeO3-X sample was less than that of BiFeO3, owing to the lattice distortion resulting from the induced oxygen vacancy sites. The crystal structures of BiFeO3 and BiFeO3-X were further characterized by employing high-resolution Cs-TEM. Consistent with the XRD results, sharp (102) and (104) planes of the perovskite BiFeO3 crystal (lattice spaces: 3.96 and 2.81 Å, respectively) were detected in both the BiFeO3 (Supplementary Materials, Figure S1c) and BiFeO3-X (Figure 2(b)) samples. The presence of oxygen vacancy sites was determined via TGA (Figure 2(c)). Unlike the BiFeO3 samples, the BiFeO3-X samples changed color, consistent with oxygen vacancy refilling steps during the TGA measurements at 520°C (Supplementary Materials, Figure S4). The color differences among BiFeO3, BiFeO3-X, and the thermogravimetry-analyzed BiFeO3-X resulted from the characteristic -band electrons and the surrounding oxygen vacancy concentrations [1921].

The valence states of BiFeO3 and BiFeO3-X were determined via XPS analysis (Figure 2 and Supplementary Materials, Figure S5). In the Bi 4f region (Figure 2(d)), Bi3+ (159.0 and 164.5 eV), although dominant in BiFeO3, occurred at a significantly lower concentration than elemental bismuth (Bi0: 158.0 and 163.0 eV) in BiFeO3-X, owing to the oxygen deficiency [2426]. Similarly, the elemental iron (Fe0) peak occurred at binding energies of approximately 709.0 and 721.9 eV in BiFeO3-X. The Fe2+ (710.8 and 723.7 eV) and Fe3+ (713 and 726.6 eV) peaks were observed in the Fe 2p spectra of both BiFeO3 and BiFeO3-X (Figure 2(e)) [27, 28]. In the O 1s region of BiFeO3 (Figure 2(f)), lattice oxygen species () and O–H bonding, associated with the original perovskite structure, were observed at binding energies of approximately 529.4 and 531 eV, respectively. Compared with these species, the atomic oxygen adsorbate species () associated with the BiFeO3-X sample occurred at a lower binding energy (528.3 eV), reflecting the insufficient coverage of oxygen species on the surface of the material [29, 30]. All the binding energies in this work were calibrated with the C 1s reference (284.5 eV). Thus, the SEM, TEM, BJH, EDS, XRD, and XPS results reveal various characteristics of the nanoporous BiFeO3-X samples obtained through a simple hydrothermal process and subsequent annealing.

3.2. Electrochemical Measurements

The electrochemical properties of BiFeO3 and BiFeO3-X were measured using a three-electrode configuration composed of a platinum counter electrode, an Ag/AgCl reference electrode, and the samples loaded on a nickel foam as a working electrode. Figure 3(a) and Figure S6 present the CV curves of BiFeO3 and BiFeO3-X obtained for potential ranging from −1.2 to 0.0 V with respect to mass and surface area, respectively. The integrated CV area for BiFeO3-X is much larger than that of BiFeO3, indicating a higher electrochemical capacitance. The BiFeO3-X sample exhibited an enhanced electrochemical performance, as evidenced by the more closely gathered redox peaks, indicating a higher electrochemical reversibility compared with that of BiFeO3. Interestingly, the strong redox peaks at −0.4 and −0.9 V were detected only for BiFeO3-X. The redox peak shapes of BiFeO3-X were retained up to a scan rate of 50 mV s−1, reflecting the fine electrochemical behavior of the electrode (Figure 3(b)). The charge-storage mechanism of BiFeO3 and BiFeO3-X was determined from the relationship between the logarithm of the anodic peak current and the logarithm of the scan rate derived from the CV curves plotted against the scan rates. The peak current density, , and sweep rate, , obey the following power-law relation: where and are adjustable values. An exponent value of indicates primarily linear diffusion kinetics of the electrolyte ions, whereas indicates the dominance of surface redox reactions. Consider the gradient shown in Figure 3(c). Although values close to 0.5 (rather than 1.0) were obtained for both BiFeO3 (0.52) and BiFeO3-X (0.67), implying a diffusion-controlled electrochemical reaction, the oxygen deficiency enhanced the surface-controlled Faradaic reactions of BiFeO3-X. The surface redox reactions and diffusion-controlled process of BiFeO3-X were evaluated from the following relation: where and denote the surface redox reaction and diffusion-controlled process, respectively [31, 32]. In BiFeO3-X, the redox reaction contributed 63.4% of the total capacity associated with a scan rate of 5 mV s−1 (Supplementary Materials, Figure S7a). When the scan rate was increased (from 5 to 50 mV s−1), the reaction yielded the dominant storage kinetics (93.2%) (Supplementary Materials, Figure S7).

The galvanic charge–discharge curves obtained at a current density of 2 A g−1 confirmed that both the BiFeO3 and BiFeO3-X samples exhibited diffusion and surface Faradaic redox behaviors. Moreover, owing to its porous nanostructure and high electrochemical activity (Figure 3(d)), the charge–discharge time of the BiFeO3-X electrode was longer than that of the BiFeO3 electrode. As a result, BiFeO3-X showed superior specific capacitance of 461.9 F g-1 (923.8 mF cm-2, 145.3 mAh g-1) to BiFeO3 316.0 F g−1 (300.9 mF cm-2, 88.9 mAh g-1) at the current density of 1 A g-1 and the other current densities (2, 5, 10, and 20 A g−1) (Figure 3(e)). Approximately 62.6% of the capacitance was retained after the charging rate was increased from 1.0 to 20.0 A g−1, demonstrating the high conductivity and good diffusion kinetics of BiFeO3-X. The supercapacitor properties of BiFeO3-X were evaluated by comparing various parameters (see Table S1, Supplementary Materials). At a high current density of 1 A g−1 (or at ), the specific capacitance of the BiFeO3-X electrode was larger than that of the other electrodes, indicating its excellent supercapacitor behavior.

Further, EIS analysis was conducted at an open-circuit potential with frequency ranging from 100 kHz to 10 Hz for BiFeO3 and BiFeO3-X (Figure 3(f)). The fitted Nyquist plot using equivalent circuit revealed that the solution resistance () and charge transfer resistance () of BiFeO3-X (0.21 and 2.20 Ω, respectively) are smaller than those of BiFeO3 (0.26 and 3.86 Ω, respectively). The semicircles in the high frequency region and straight lines in the low frequency region indicate the and the Warburg impedance () associated with ion-diffusion processes. In addition, compared with BiFeO3, the BiFeO3-X electrode exhibited faster surface and charge transfer kinetics, thereby demonstrating superior electrochemical kinetics and stability.

To further understand the charge–discharge mechanism, the BiFeO3-X sample was subjected to CV measurements in 0.1, 1, and 6 M KOH electrolytes (Figure 4(a)). The ambiguous redox peaks (at −0.5 and −0.8 V versus Ag/AgCl) obtained for 0.1 M KOH were resolved into distinct pairs of redox reaction peaks. This indicated that Fe2+/Fe3+ and Bi0/Bi3+ reactions occurred as the concentration of OH ions increased in the electrolyte [2528]. The possible electrochemical reactions of BiFeO3-X during charge-discharge are as follows:

Notably, the intensity of the oxidation peak occurring at approximately −0.4 V versus Ag/AgCl increased considerably, reflecting the activation of Bi as BiOOH (Figure 4(b)). This result indicates that BiFeO3-X possessed dual active sites (Bi, Fe) with a tunable activity against the pH environment of the electrolyte. Moreover, with anion vacancies, the electrochemical activity of both the Bi and Fe sites was synergistically enhanced, and the electrochemical contribution of Bi increased significantly. This corresponded to a significant improvement in the electrochemical activity of BiFeO3-X with the incorporation of anion vacancy sites. Furthermore, as an important performance parameter, the cycling stability of BiFeO3-X was measured at a current density of 2 A g−1 in 1 M KOH (Figure 4(c)). Approximately 94.4% of the initial cycling stability was retained without significant changes in the Coulombic efficiency during 2,000 cycles of the measurement. Moreover, after the electrochemical cycling, the electrode underwent negligible morphological and structural changes, indicative of stable electrochemical properties (Supplementary Materials, Figure S8).

4. Conclusions

We report the improved electrochemical performance of BiFeO3-X materials resulting from anion defect sites induced on their crystal structure. With the anion vacancies, the electrochemical activity of both the Bi and Fe sites was synergistically improved, although the electrochemical contribution of the Bi sites was considerably larger than that of the Fe sites. Therefore, at a current density of 1 A g−1, the specific capacitance of the BiFeO3-X electrode (461.9 F g−1, 923.8 mF cm-2, and 145.3 mAh g-1) was significantly higher than that of BiFeO3 (316.0 F g−1, 300.9 mF cm-2, and 88.9 mAh g-1), and high electrochemical stability was achieved. The realized nanoporous structure and defect sites represent a simple means of enhancing the inherent electrochemical property of BiFeO3 materials. Moreover, the subsequent vacancy engineering method is an effective strategy for maximizing the electrochemical activity of oxide materials for application in high-performance EES devices.

Data Availability

Access to data is restricted according to the policies of institutions and funding organization.

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Authors’ Contributions

Seunghwan Jo was responsible for the conceptualization, methodology, and writing—original draft. Sangyeon Pak was responsible for the data curation, formal analysis, and investigation. Young-Woo Lee was responsible for the methodology, data curation, and formal analysis. SeungNam Cha was responsible for the methodology, formal analysis, and validation. John Hong was responsible for the funding acquisition, supervision, and writing—review and editing. Jung Inn Sohn was responsible for the funding acquisition, supervision, and writing—review and editing.

Acknowledgments

This work was financially supported by the National Research Foundation of Korea (NRF) grants funded by the Korean Government (MSIT) (2019R1A2C1007883 and 2021M3H4A3A02093515) and by the Korea Carbon Industry Promotion Agency funded by the Ministry of Trade, Industry and Energy of the Republic of Korea (G2820220800015).

Supplementary Materials

Cs-TEM and EDS images of BiFeO3, BJH analysis for BiFeO3 and BiFeO3-X, EDS profiles for BiFeO3 and BiFeO3-X, digital photograph of the thermogravimetry-analyzed BiFeO3-X, X-ray photoelectron spectroscope (XPS) full spectra of BiFeO3 and BiFeO3-X, cyclic voltammetry curve BiFeO3-X compared to BiFeO3 and nickel foam substrate, contribution of surface redox reaction and diffusion-controlled process, and postanalysis after electrochemical cycling. (Supplementary Materials)