Chaoqiao
Yang
a,
Ya
Zhao
a,
Jiaxin
Fan
*b,
Lin
Li
b,
Jinxia
Zhou
a,
Keliang
Wang
b,
Fenglian
Lu
b and
Hongmei
Sun
b
aCollege of Environmental and Chemical Engineering, Dalian University, Dalian, 116622, Liaoning, China
bSchool of Chemistry and Materials Engineering, Liupanshui Normal University, Liupanshui, 553004, Guizhou, China. E-mail: fmf_fjx@163.com
First published on 18th June 2024
Due to their open structure, numerous reactive sites, and eco-friendly characteristics, Prussian blue-based electrode materials find widespread application in aqueous zinc-ion batteries. However, the Prussian blue materials face challenges such as low specific surface area and high defect rate, resulting in slow ion transport and poor cycling performance. This study focuses on synthesizing ferric ferricyanide (FeFe(CN)6) material with reduced vacancies and a higher specific surface area. The synthesis process incorporates citric acid as a chemical assistant, resulting in the as-synthesized C-FeFe(CN)6 featuring a significantly larger specific surface area of 406.625 m2 g−1, compared to FeFe(CN)6 without citric acid (28.143 m2 g−1). Moreover, C-FeFe(CN)6 has a lower vacancy rate of 20% compared to FeFe(CN)6 at 31%. The addition of citric acid effectively slows down the nucleation rate of C-FeFe(CN)6 crystals. When utilized as cathode materials in aqueous zinc–sodium hybrid batteries, C-FeFe(CN)6 demonstrates highly reversible Na+ storage behavior, exhibiting a remarkable reversible capacity of 162.5 mA h g−1 at 0.1C and a good cycling performance with 88% capacity retention over 500 cycles (at a high rate of 45C). These superior performance characteristics make C-FeFe(CN)6 a promising candidate for extensive energy storage applications.
The cathode materials play a crucial role in zinc-ion batteries.11 These materials primarily include manganese-based oxides,12 vanadium-based oxides,13–15 Prussian blue analogues,16 organic compounds17 and other cathode materials.18,19 Prussian blue (PB) has a general chemical formula of AxFe[Fe(CN)6]1−ynH2O, where A represents various possible interstitial cations introduced during synthesis or through electrochemical insertion.20 The cyanide ion's strong ligand field results in the low spin state of the C-coordinated Fe cation in the Fe(CN)6 group, while the N-coordinated Fe ion adopts a high spin state.21 Therefore, high-spin iron and low-spin iron have different redox potentials, which may also depend on water content and ion embedding position.22–24
So far, a considerable amount of work has been done on minimizing the [Fe(CN)]6 defects. Researchers have found that slow-release chelating agents such as citric acid can significantly delay the crystallization of PBA, thus inhibiting the formation of defects.25–28 In this regard, Yang25et al. proposed a simple citrate-assisted kinetic control crystallization method, successfully synthesizing Na1.85Co[Fe(CN)6]0.99·2.5H2O with low defects and high crystallinity. Through this method, the vacancy defect formation rate was reduced from 12% to 1%, while significantly increasing the redox sites of Na+, resulting in a high reversible capacity of 130 mA h g−1 and 90% capacity retention after 800 cycles. Furthermore, Liu27et al. synthesized NaxFeFe(CN)6 with different sodium contents by adding varying masses of sodium citrate. Among them, Na1.70FeFe(CN)6 exhibited a capacity of up to 120.7 mA h g−1 at a current density of 200 mA g−1, and even at a current density of 1200 mA g−1, the capacity remained at 73.6 mA h g−1. In addition, Chen28et al. successfully synthesized s-PBMN samples with high crystallinity and regular morphology by controlling the nucleation rate with sodium citrate. Using this method, the vacancy defect content was reduced from 11% to 2%, and the material exhibited a high specific capacity of 120 mA h g−1. Even after 400 cycles at a current density of 100 mA g−1, the capacity remained at 110 mA h g−1. However, so far, the effectiveness of the aforementioned strategies for eliminating PBA vacancy defects is still relatively low, posing a major challenge. Overall, the above-mentioned studies mainly explore the application of sodium citrate in slowing down the coordination rates of [Fe(CN)6]4− and M2+ (M represents transition metal ion). Hence, in this study, we utilized a dropwise mixed reaction solution and simultaneously introduced citric acid to jointly regulate the reaction rate, aiming to investigate the impact of citric acid on the coordination of [Fe(CN)6]3− and Fe3+. The ferric ferricyanide synthesized via the coprecipitation method presents certain issues such as high vacancy defect, small specific surface area, and large grain size. Therefore, this paper introduces citric acid as a chelating agent to decelerate the nucleation rate of the sample, leading to the production of ferric ferricyanide with fewer defects and a higher specific surface area, consequently enhancing the electrochemical performance of the battery. Additionally, the exploration of the selectivity of Na+ and Zn2+ ions embedding in the cathode C-FeFe(CN)6 structure is rarely documented for Zn–Na hybrid systems.
The combination of aqueous electrolytes with Na+ and Zn2+ ions renders the fabrication of rechargeable battery without using the toxic organic electrolyte and highly activated metals (e.g., Li, Na) in ambient air.29 Zinc–sodium hybrid battery can help solve the problems of low energy density and poor cycle performance of PBAs.30 Through introducing an extra charge carrier into the aqueous zinc salt electrolyte, like alkali cation Li+/Na+/K+, the overall energy storage performance can be improved on basis of the appropriate cathode materials and co-insertion of the alkali cation and zinc ion.31 Zinc salt is a strong acid and weak base salt with low ionization degree. Adding sodium salt can improve the conductivity of electrolyte.
In previous studies, we have investigated the electrochemical performance of FeFe(CN)6 synthesized without the addition of citric acid.32 The focus of this article will be on studying the effect of citric acid on the formation of C-FeFe(CN)6 crystals and exploring its storage performance for Zn2+ and Na+ in a Zn–Na hybrid battery. We prepared C-FeFe(CN)6 using a controlled crystallization method and characterized it using XRD, BET, TEM, and other characterization tests. The electrochemical performance of the material was tested using cyclic voltammetry (CV) and galvanostatic charge–discharge (GCD) methods. The findings suggest that the inclusion of citric acid enhanced the specific surface area of ferric ferricyanide materials, mitigated the vacancy defects of Fe(CN)6, improved the overall crystal structure, consequently showcasing outstanding electrochemical performance.
Fig. 1 Schematic representation of the formation mechanism of C-FeFe(CN)6 in a controlled crystallization reaction. |
Fig. 2a illustrates the XRD patterns of FeFe(CN)6 and C-FeFe(CN)6, both displaying a distinct correlation with the face-centered cubic structure of PBAs, indicating high crystallinity (JCPDS no. 73-0687). In Fig. 2b, the infrared spectra of the two samples are presented. The peaks observed at 3441 (3427) cm−1 and 1613 (1626) cm−1 can be ascribed to the stretching and bending of O–H bonds, respectively. Additionally, a prominent peak at 2086 (2089) cm−1 corresponds to the stretching vibration of the CN ligand (Fe3+–CN–Fe3+). Furthermore, the peak at 502 cm−1 corresponds to the in-plane deformation of Fe–C bonds. The thermogravimetric (TG) results in Fig. 2c indicate that the weight loss of C-FeFe(CN)6 in the temperature range of 0–250 °C is 14.89%, equivalent to 2.19 water molecules per C-FeFe(CN)6. Moreover, in comparison to FeFe(CN)6, the water content of the C-FeFe(CN)6 material is reduced, indicating that the addition of citric acid slows down the crystallization rate and inhibits the formation of crystalline water. This results in stronger thermal stability and a more stable crystal structure. Fig. 2d displays the Raman spectra of the two materials, with a single peak at 2147 (2152) cm−1 attributed to the stretching vibration of Fe3+–CN–Fe3+.
Fig. 2 Physical properties of C-FeFe(CN)6 and FeFe(CN)6 samples. (a) XRD patterns; (b) FT-IR spectra; (c) TG curves; (d) Raman spectra. |
Fig. 3a displays the XPS full spectra of two samples, revealing the presence of C, N, O and Fe elements in both materials. Moreover, the detailed Fe 2p spectrum (Fig. 3b) shows two distinctive peak groups for Fe 2p3/2 at 708.58 eV and Fe 2p1/2 at 721.48 eV. The Fe 2p3/2 peak can be deconvoluted into two distinct peaks corresponding to C–Fe3+ and N–Fe3+, while the Fe 2p1/2 peak can be deconvoluted into three separate peaks corresponding to C–Fe3+, N–Fe3+, and a satellite signal peak.33 Additionally, Fig. 3c depicts the N2 adsorption–desorption isotherms of C-FeFe(CN)6 and FeFe(CN)6, with the adsorption capacity of C-FeFe(CN)6 notably surpassing that of FeFe(CN)6. The specific surface area of FeFe(CN)6 is measured at 28.143 m2 g−1, whereas that of C-FeFe(CN)6 reaches as high as 406.625 m2 g−1. C-FeFe(CN)6 predominantly demonstrates mesoporous characteristics with pore sizes ranging from 2 to 10 nm (Fig. 3d), and its average pore diameter is 2.472 nm, approximately half that of FeFe(CN)6 (5.056 nm) (Table S1, ESI†). The increased specific surface area of FeFe(CN)6, along with reduced pore size attributed to the addition of citric acid, significantly boosts its electrochemical performance.
Fig. 3 (a) The full XPS spectra; (b) high resolution Fe 2p spectrum of C-FeFe(CN)6; (c) nitrogen adsorption–desorption isotherms; (d) pore sizes distribution. |
Based on the results obtained from ICP-OES, elemental analysis, and TG (Table S2, ESI†), the precise chemical compositions of FeFe(CN)6 and C-FeFe(CN)6 can be expressed as Fe[Fe(CN)6]0.69·□0.31·2.80H2O and K0.18Fe[Fe(CN)6]0.80·□0.20·2.19H2O, respectively, in which □ denotes Fe(CN)6 vacancy (defect). It is evident that C-FeFe(CN)6 has a lower defect of 20% than 31% of FeFe(CN)6, which might be helpful to improve the electrochemical performances.34 The decrease of Fe(CN)6 vacancy could lead to the decrease of bound water in material lattice. The scanning electron microscope (SEM) image depicted in Fig. 4 reveals that the C-FeFe(CN)6 crystals are uniformly dispersed and arranged in a regular pattern. It is apparent that the FeFe(CN)6 crystal grains are relatively large, exhibiting a rough surface, and forming large agglomerated clusters. The particle size distribution of FeFe(CN)6 ranges from 300 to 800 nm, with an average particle size of 535 nm. On the other hand, the particle size distribution of C-FeFe(CN)6 ranges from 120 to 220 nm, with an average particle size of 168 nm. Notably, the presence of citric acid is shown to play a pivotal role in the formation of nanocrystals. Its strong chelating ability enables complexation with Fe3+ ions, thereby retarding the precipitation rate of Fe3+ with Fe(CN)63−, subsequently slowing down crystal nucleation and growth. This mechanism leads to the development of smaller particle-sized C-FeFe(CN)6 crystals.
Fig. 4 SEM images and crystallite size and distribution of FeFe(CN)6 (a)–(c) and C-FeFe(CN)6 (d)–(f) at different magnification. |
To further examine the morphology, structural characteristics, and surface elemental composition of the C-FeFe(CN)6 material, TEM and EDS techniques were employed for sample characterization. As illustrated in Fig. 5, energy dispersive X-ray spectroscopy (EDS) mapping confirmed the homogeneous distribution of K, Fe, C, N, and O elements in C-FeFe(CN)6. The EDS diagram of FeFe(CN)6 confirms the uniform distribution of Fe, C, N, and O elements in the sample (Fig. S5, ESI†). The findings suggest that reducing the nucleation rate contributed to more thorough growth of the C-FeFe(CN)6 crystal, leading to a smoother surface and reduced resistance to ion insertion. This indicates that C-FeFe(CN)6 demonstrates excellent performance over extended cycling periods.
The GCD curves of the C-FeFe(CN)6 electrode in different cycles (2nd, 10th, 20th, 50th, and 100th) at a current density of 15C are presented in Fig. 6c. It can be observed that the battery maintains similar charge–discharge curves from the second cycle to the 100th cycle, indicating excellent cycle performance. Fig. S1b (ESI†) demonstrates the charge–discharge curves of FeFe(CN)6 electrode at the 2nd, 10th, 20th, 50th and 100th cycles at 15C. The FeFe(CN)6 platform shows a slight decrease, while the C-FeFe(CN)6 platform maintains good discharge performance. Furthermore, as depicted in Fig. 6d, long-term cycling performance is assessed at a high current density of 45C over 1000 cycles. The reversible discharge specific capacity decreased from 83.8 to 55.1 mA h g−1, which corresponds to a capacity retention rate of about 65.8%. The factors leading to the capacity attenuation may include the following aspects: incomplete removal of embedded ions during the cycle, the decay of C-FeFe(CN)6 structure, etc. However, for FeFe(CN)6 sample synthesized without citric acid, the capacity significantly decreases after 1000 cycles, retaining only 46.7% of its initial capacity. This is because when Na+ ions are transported in the cathode materials with relatively large grain size, the longer diffusion length and lower material utilization efficiency of Na+ ions lead to sluggish kinetics, which is not conducive to ion transport. This results in poor electrochemical performance of the battery.35–37 The addition of citric acid in the synthesis process slows down the crystal nucleation rate, reduces the crystalline water content, and decreases crystal defects. As a result, it allows for obtaining C-FeFe(CN)6 materials with smaller grain size and higher specific surface area. The results demonstrate that, compared to the original materials, it enhances cycling stability and enables a highly reversible Na+ storage reaction.
Self-discharge rate is an important property for batteries in practical application. To study the self-discharge of Zn//C-FeFe(CN)6, the battery was fully charged at 1C current density and then stood for 24 h before discharging. Fig. S2a and b (ESI†) show the voltage-time and voltage-specific capacity curves of the battery without rest, in which the discharge specific capacity is 117.6 mA h g−1 (Fig. S2b, ESI†). After standing for 24 h, the voltage decays to 1.31 V (Fig. S2c and d, ESI†). It is worth noting that the battery still maintains a high capacity of 103.6 mA h g−1. Therefore, the self-discharge rate of C-FeFe(CN)6 is 0.50% h−1, which is much lower than that of FeFe(CN)6 under the same conditions (Fig. S3, ESI,† self-discharge rate: 1.32% h−1). Due to the mesoporous structure of C-FeFe(CN)6, it is more suitable for storing Na+ ions and can better maintain the stability of the Na+ embedding structure. Zn//C-FeFe(CN)6 battery has a low self-discharge rate, indicating that it has excellent electrochemical performance.
According to Fig. 7a, both the CV curves of FeFe(CN)6 and C-FeFe(CN)6 show a sharp symmetric Fe2+/Fe3+ redox peak. The peak of FeFe(CN)6 is located at 1.00/1.26 V, while the peak of C-FeFe(CN)6 is located at 1.08/1.28 V. Fig. 7b displays cyclic voltammetry curves of Zn//C-FeFe(CN)6 battery at different scanning speeds. With the increase of scanning rate from 5 to 8 mV s−1, the peak current intensity increases significantly, while the peak potential shifts slightly. In addition, Fig. 7c illustrates the rate capacity of C-FeFe(CN)6 and FeFe(CN)6 electrodes. The specific discharge capacities of C-FeFe(CN)6 electrode at 1, 5, 10, 15, 20, 25 and 30C are 130.6, 115.0, 103.6, 96.3, 91.3, 87.5, 84.4, and 116.4 mA h g−1 (average values), respectively. The specific discharge capacities of FeFe(CN)6 electrode at 1, 5, 10, 15, 20, 25, 30 and 1C are 111.5, 93.9, 85.8, 81.4, 78.1, 75.4, 73.1 and 94.0 mA h g−1, respectively. Under the same charging rate, the specific capacity of the C-FeFe(CN)6 electrode is significantly greater than that of the FeFe(CN)6 electrode. Notably, when the rate is reduced from 30 to 1C, the capacity of C-FeFe(CN)6 electrode recovers to 116.4 mA h g−1, which is 89.15% of the initial value, demonstrating good rate performance. Fig. 7d exhibits the EIS curves of Zn//C-FeFe(CN)6 in the initial, 200th and 1000th cycles, and the arc in the mid-frequency region represents the charge transfer resistance (Rct). The EIS data are fitted by the Zsimpwin software, and the values are presented in Table S4 (ESI†). The initial Rct value is 804.5 Ω. With the cycling, Rct drops to 548.7 and 244.0 Ω after the 200th and 1000th cycles, respectively. Based on the cycling process, it is evident that the Rct value gradually decreases, indicating that the insertion of Na+ ions significantly stabilizes the material structure, thereby mitigating the collapse and damage of the crystal structure during cycling. Conversely, Fig. S4 (ESI†) exhibits the EIS curves of Zn//FeFe(CN)6 at different cycles. The initial Rct value is 500.7 Ω (Table S5, ESI†). As the cycle progresses, the Rct value decreases to 356.7 Ω after the 200th cycle and then increases to 1112 Ω after the 1000th cycle. The results indicate that the introduction of Na+ ions disrupt the crystal structure, resulting in unsatisfactory cycling performance. In Fig. 9d, the XRD spectrum and SEM images of both uncycled and cycled 1000th C-FeFe(CN)6 electrode slices are presented. Following 1000 cycles, the XRD spectrum of the C-FeFe(CN)6 cathode retains its face-centered cubic phase, while the SEM image displays a well-defined nano-cubic morphology, providing further evidence of the remarkable stability exhibited in the repetitive Na+ insertion/extraction reaction within the C-FeFe(CN)6 framework. Fig. S6 (ESI†) displays the XRD and SEM patterns of FeFe(CN)6 electrodes before and after 1000 cycles. Following 1000 cycles, the XRD pattern of the FeFe(CN)6 cathode continues to exhibit the face-centered cubic phase, while the SEM image reveals partial structural damage, suggesting varying degrees of crystal structure impairment due to ion embedding and extraction.
Fig. 8 Characterization of C-FeFe(CN)6 electrode at different states. (a) XRD patterns; (b) the magnifications of (200) peak; (c) the full XPS spectra; (d) high-resolution XPS spectra of Fe 2p. |
The high-resolution XPS spectra of Na 1s and Zn 2p in various charge–discharge states are illustrated in Fig. 9. The findings suggest that the level of Na+ ion insertion surpasses that of Zn2+ ions, establishing the insertion and extraction of Na+ ions as the predominant processes. We roughly calculated the capacity contribution of Na+ and Zn2+ based on their embeddings, which are 63.8% and 34.3%, respectively. Consequently, the electrochemical reaction of the Zn//C-FeFe(CN)6 battery can be succinctly outlined as:
Cathode: C-FeFe(CN)6 + 1.65Na+ + 1.65e− ↔ C–Na1.65FeFe(CN)6 | (1) |
Anode: 0.82Zn ↔ 0.82Zn2+ + 1.65e− | (2) |
Total: C-FeFe(CN)6 + 1.65Na+ + 0.82Zn ↔ C–Na1.65FeFe(CN)6 + 0.82Zn2+ | (3) |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ma00303a |
This journal is © The Royal Society of Chemistry 2024 |