Sodium ion batteries are an emerging candidate to replace lithium ion batteries in large-scale electrical energy storage systems due to the abundance and widespread distribution of sodium. Despite the growing interest, the development of high-performance sodium cathode materials remains a challenge. In particular, polyanionic compounds are considered as a strong cathode candidate owing to their better cycling stability, a flatter voltage profile, and stronger thermal stability compared to other cathode materials. Here, we report the rational design of a biomimetic bone-inspired polyanionic Na3V2(PO4)3-reduced graphene oxide composite (BI-NVP) cathode that achieves ultrahigh rate charging and ultralong cycling life in a sodium ion battery. At a charging rate of 1 C, BI-NVP delivers 97% of its theoretical capacity and is able to retain a voltage plateau even at the ultra-high rate of 200 C. It also shows long cycling life with capacity retention of 91% after 10 000 cycles at 50 C. The sodium ion battery cells with a BI-NVP cathode and Na metal anode were able to deliver a maximum specific energy of 350 W h kg−1 and maximum specific power of 154 kW kg−1. In situ and postmortem analyses of cycled BI-NVP (including by Raman and XRD spectra) HRTEM, and STEM-EELS, indicate highly reversible dilation–contraction, negligible electrode pulverization, and a stable NVP-reduced graphene oxide layer interface. The results presented here provide a rational and biomimetic material design for the electrode architecture for ultrahigh power and ultralong cyclability of the sodium ion battery full cells when paired with a sodium metal anode.
The rapid development of cost-efficient large-scale electrical energy storage systems (ESSs) has led to remarkable progress on the exploitation of renewable energy resources, such as solar energy, wind energy, and tidal energy.1–4 Owing to the abundance and widespread distribution of sodium, sodium ion batteries (SIBs) are a strong candidate to replace lithium ion batteries (LIBs) for large-scale EESs.5–11 Nevertheless, the development of high-performance sodium cathode materials remains challenging.12–15 Extensive efforts have thus been devoted to investigating cathode materials, especially sodium-layered oxides12–14,16–18 and polyanionic compounds.19–23 Despite the high specific capacity of layered compounds, they often lack structural stability in the highly charged state and require a low discharge cutoff voltage to achieve enough capacity.14,17 In a sharp contrast to oxides, most polyanionic compounds have a three-dimensional (3D) robust framework due to strong covalent bonding of the oxygen atom in the polyanion polyhedra. This provides for better cycling stability, flatter voltage profile, stronger thermal stability, and higher safety. Owing to the inductive effect of the polyanion groups, a higher operating voltage can be achieved.21,24
Among the family of polyanion compounds, Na superionic conductor (NASICON)-type Na3V2(PO4)3 (NVP) is of particular interest.19,20 The crystal structure of NVP is composed of [VO6] octahedra and [PO4] tetrahedra interlinked via common apical oxygens to form [V2 (PO4)3] “lantern” units; Na ions occupy Na1 (sixfold coordination) and Na2 (eightfold coordination) sites.25 Such a stable crystal structure shows a small volume expansion (8.26%) during cycling.26 This NASICON-type framework provides 3D open diffusion pathways for Na+ transport, resulting in rapid Na+ diffusivities exceeding 10−11 cm2 s−1,25 which are many orders of magnitude greater than those of other SIB cathodes (e.g., around 10−14 cm2 s−1 for NaMn3O5 and ≈10−17 cm2 s−1 for NaFePO4).27,28 Accompanying the V3+/V4+ redox reaction, Na ions can be extracted/intercalated reversibly from/into 3D framework over a wide range of Na compositions (up to two Na per unit formula of NVP) at 3.4 V vs Na/Na+. A reasonable theoretical capacity of 117 mA h g−1 can be achieved, with specific energy as high as 401 W h kg−1.24,25,29 A NVP-based battery is a strong candidate for stable, high-energy-density EESs,30,31 along with Prussian blue,32 olivine NaFePO433 and mixed polyanionic sodium host materials.34–36
Unfortunately, NVP has poor intrinsic electronic conductivity (1.63 × 10−6 S cm−1),37 analogous to its lithium-counterpart Li3V2(PO4)3.38 In its pristine form, NVP's theoretical capacity is not fully accessible even at intermediate charge rates, e.g., 67 mA h g−1 at 0.5 C.39 To overcome this critical issue, researchers have mainly focused on electrically conducting coatings based on carbon25,29,40,41 and on nanosizing the NVP particles.38,39 For example, Li et al.42 synthesized a NVP/C with a thin carbon coating layer derived from glucose, which showed the capacities of 98 mA h g−1 at 0.1 C and 76 mA h g−1 at 30 C. Liu et al.43 reduced the size of NVP (20–30 nm), which was encapsulated in a 1D electrospun NVP/C nanofiber, exhibiting the capacities of 103 mA h g−1 at 0.1 C and 58 mA h g−1 at 5 C. While achieving superior gravimetric performance, many nanostructuring methods have limitations due to the resultant low electrode density.44–47
3D hierarchical porous micro/nano-structures, consisting of low-dimensional building blocks, have attracted significant research interest.48–58 Such micro/nano-structures can combine the superior ion diffusion characteristics of the nano-scale building blocks, and synergistically gain additional benefits from a microscale secondary architecture.59,60 For example, a 3D Si/C micro/nano-composite was synthesized via quasi-industrial-scale production, delivering a stable capacity of ∼400 mA h g−1 at 2100 mA g−1.61 Hierarchical MoS2 microboxes constructed by ultrathin nanosheets were synthesized by a template-assisted strategy, exhibiting a stable capacity ∼700 mA h g−1 at 1000 mA g−1.62 Such approaches appear quite useful for the NVP system as well, potentially allowing for a combination of fast rate capability, extended cyclability, and high electrode density.
Herein, we rationally design and fabricate a biomimetic, bone-inspired, NVP, reduced graphene oxide composite, termed BI-NVP. Mammalian bone, consisting of the inner “Spongy bone” and the hard “Compact bone,” is an excellent structural composite [Fig. 1(a)], allowing for strength combined with flexibility. The inner “Spongy bone” is the soft inner part of the bone and is structurally stabilized by the hard “Compact bone” surrounding it. The Spongy bone possesses a surface area ten times higher than the Compact bone. This creates the classic soft-hard composite effect, allowing the bone to flex under stress and yet structurally support the load of the body. Mimicking such an effective biological architecture, i.e., creating a biomimetic NVP system, was our motivation. As will be demonstrated, following such a methodology creates a uniquely cyclable and fast charging NVP-based SIB cathode.
II. RESULTS AND DISCUSSION
A. Structure and chemistry of bioinspired BI-NVP
As shown in Fig. 1, the design of BI-NVP is inspired by the unique porous structure of mammalian bone. The outer dense rGO nanolayers serve as the compact bone, improving the electrical conductivity for a fast charge transfer, adding secondary Na ion conduction paths, and providing mechanical robustness. The inner NVP granules are interconnected with the rGO, forming a porous composite structure, as shown in the scanning electron microscopy (SEM) image [Fig. 1(b)]. The visible mesopores become filled with electrolytes and provide channels for a rapid sodium ion transport. Owing to the tight binding between intimately embedded NVP and nanoscale rGO shell, the structure of composite is stabilized to withstand the sodiation/desodiation volume changes. An example of the NVP-rGO interface is shown in an associated high-resolution transmission electron microscopy (HRTEM) image [Fig. 1(b)].
This BI-NVP composite is fabricated through mixing, freeze-drying, and sintering. The details of the synthesis process are provided in Sec. IV. The sodium-vanadium-phosphate complex ions in the precursor solution interact with the oxygen functional groups on the surface of the graphene oxide (GO) during the mixing. Subsequently, the solvent is removed through freeze-drying while maintaining the original 3D structure and internal porosity. Complex ions begin to crystallize into NVP nanocrystals at 400 °C and further grow into granules. Finally, the oxygen functional groups on the GO surface interact with the NVP to condense and form strong covalent bonds at 750 °C.
The morphology and microstructure of the BI-NVP were analyzed by SEM and TEM. As shown in Figs. 2(a) and 2(b) and Fig. S1 in the supplementary material, the rGO sheets cover the surface of the NVP and are interconnected with adjacent NVP granules to form mesoporous networks. During the formation of the interconnected porous structure, the rGO sheets prevent the aggregation of NVP granules into large lumps and serve as the electrically conducting networks between the NVP granules to obtain an efficient charge transfer. Since rGO is known to be Na ion active as well, it also serves as fast ion conduction paths during sodiation and desodiation.63 The structure is composed of numerous NVP granules with a size distribution in the range of 20–50 nm [Fig. 2(c)], which are tightly wrapped with compact rGO sheets. The interconnected porous structure was constructed by aggregated granules on a micrometer scale. Figure 2(d) shows the HRTEM image and fast Fourier transform (FFT) pattern highlighting the atomic structure of the composite. From Fig. 2(d) it may be observed coating thickness of the rGO layers is in the range of 5–10 nm. As shown in Fig. S2 in the supplementary material, per the Energy Dispersive X-ray Spectroscopy (EDXS) elemental maps, the carbon from the rGO uniformly covers the NVP structure. According to the thermogravimetric analysis (TGA) in Fig. S3 in the supplementary material, the BI-NVP composite comprises 95.0 and 5.0 wt. % of NVP and rGO, respectively.
Figure 2(e) shows X-ray diffraction (XRD) patterns of the NVP and of BI-NVP. All the diffraction peaks of the samples are assigned to the rhombohedral (space group: , lattice parameter: a = b = 8.426 Å and c = 21.815 Å) structure of NVP (JCPDS # 053–0018).64 Figure 2(f) shows the Raman spectra of BI-NVP with characteristic modes of NVP being in the range of 250–1100 cm−1. In addition, two characteristic bands of carbonaceous materials are observed at 1346 and 1582 cm−1. The intensity ratio of the D and G bands (ID/IG) of the BI-NVP is ∼1.10, which is typical for the highly disordered rGO materials still containing substantial oxygen. The near-surface chemistry of BI-NVP before and after the 750 °C heat treatment was analyzed by using X-ray photoelectron spectroscopy (XPS). As illustrated in the survey spectrum in Fig. S4 in the supplementary material, the BI-NVP composite exhibits five elemental signals of the Na 1s, O 1s, V 2p, C 1s, and P 2p peaks. The C 1s, O 1s, and V 2p scan spectra are deconvoluted by a Gaussian-Lorentzian function. The bonding configurations, corresponding binding energies, and relative areas of BI-NVP are summarized (see Table S1 in the supplementary material). In the XPS spectra of BI-NVP before heat treatment, the V-O/P-O and V3+ peaks are derived from NVP, while C-C, C-O, C O, and O C-O peaks originate from rGO (see Fig. S5 in the supplementary material). After the heat treatment, as shown in the C 1s spectrum, the relative area of the C-O peak is increased from 13.7% to 41.9%, while those of the O C-O and C O peaks are decreased [Fig. 2(g)]. In the O 1s spectrum, the H2O residual peak is not observed, the relative area of the O C peak is decreased from 32.6% to 20.8%, and the O-C and V-O peaks are shifted from 531.4 to 530.7 eV without significant change of relative area [Fig. 2(h)]. These results demonstrate that the C O bonds on the rGO surface donate electrons to V-O/P-O dangling bonds of the NVP. The increased electron density around the V-O/P-O bonds is delocalized, forming strong bindings between the rGO and NVP such as (C-O-V/C-O-P) during the heat treatment.65–67 These findings are associated with the intimate contact between the NVP and rGO at the heterointerface, as shown in the HRTEM image. The chemical bonding is further confirmed by the V 2p spectrum where the minor V2+ peak of the NVP is observed, which was reduced from V3+ by receiving electrons from the C O double bonds at the surface of the rGO [Fig. 2(i)].68
The coating of the NVP by the rGO layers also changes the textural properties of the BI-NVP, as confirmed by nitrogen adsorption/desorption isotherms shown in Fig. S6 in the supplementary material. Both isotherms are type-IV isotherms. However, a hysteresis indicating mesopores is observed only in the isotherm of the BI-NVP. The BET surface area and pore volume of the BI-NVP are estimated to be 32 m2 g−1 and 0.112 cm3 g−1, respectively, approximately 3 times larger than those of the NVP (see Table S2 in the supplementary material). The pore size distribution obtained by the Non-Local Density Functional Theory (NLDFT) analysis shows that the pore width of BI-NVP spanned between the mesopore and macropore regimes with the average pore width of 62.4 nm, while the pore with a baseline NVP was only located in the macropore regime with the average pore width centered on 432.8 nm.
B. Electrochemical performances of SIB based on BI-NVP
The electrochemical behavior of BI-NVP and baseline NVP was investigated using a range of techniques. As shown by the cyclic voltammetry (CV) curves in Figs. 3(a)–3(b) and Fig. S7 in the supplementary material, both electrodes exhibit redox peaks corresponding to the changes in oxidation states of the V3+/V4+ redox couples around 3.4 V vs Na/Na+. The current is measured at scan rates of 1–100 mV s−1. The weight normalized redox peak currents of the BI-NVP are considerably higher than that of NVP, indicating a higher overall utilization of the theoretical capacity of 117 mA h g−1. At high scan rates, the redox peaks of the BI-NVP are well preserved, while those of the NVP are distorted and resistive. The peak-to-peak separation, power-law coefficient, diffusion coefficient, charge transfer resistance, and activation energy of both electrodes are evaluated to further investigate the redox kinetics. The BI-NVP electrode exhibits a peak-to-peak separation of 331 mV at 50 mV s−1, lower than that of the 547 mV for NVP. This difference in peak-to-peak separation of two electrodes is enlarged at the higher scan rates, supporting the argument for facile kinetics due to the BI architecture.
The b values, which represent the slope of the fitted power law and are derived from the anodic and cathodic peaks as functions of the scan rate (i = aνb), are 0.55 and 0.62 for BI-NVP, and 0.47 and 0.44 for NVP (R2 > 0.99). These results are shown in Fig. 3(c). These results indicate that both of electrodes are under diffusion-controlled redox kinetics.69–72 The Randles-Sevcik equation [Eq. (1)] can be used to calculate the Na-ion diffusion coefficient, by plotting the peak current Ip vs the square root of the scan rate ν1/2. These results are shown in Fig. S8 in the supplementary material.
where Ip is the peak current (A), n is the electron transfer number of the reaction (n = 1), A is the surface area of the electrode (cm2), DNa is the Na-ion diffusion coefficient, C0 is the concentration of Na ions in the cathode (3.47 × 10−3 mol cm−3 for NVP), and ν is the scan rate (V s−1).71,73,74 The DNa of the BI-NVP electrode, obtained by using the slope of ν1/2 vs Ip, is 3.90 × 10−10 cm2 s−1, more than an order of magnitude higher than the 1.17 × 10−11 cm2 s−1 of NVP (R2 > 0.99). This indicates that the combination of the rGO shell, the nanosized NVP particulates and the mesopores creates a more favorable environment for Na ion diffusion.
Figure 3(d) shows the electrochemical impedance spectroscopy (EIS) Nyquist plot of BI-NVP and NVP taken at open circuit voltage at room temperature. Each Nyquist plot is composed of a high-frequency semicircle, a Warburg region, followed by a steep sloping line in the low-frequency region. The Warburg region is related to Na-ion diffusion coefficient. Figure S9 in the supplementary material displays the linear relationship of the real part of the impedance with the inverse square root of the angular frequency. The Warburg factor is determined from the slope σ and is used to calculate the DNa using Eqs. (2) and (3).75
where Z′ is the real part resistance, ω is the angular frequency, R is the gas constant, T is the absolute temperature, A is the surface area of the electrode, n is the electron transfer number of the reaction, F is the Faraday constant, and C is the concentration of Na ion in the cathode. The DNa value of NVP and BI-NVP cathode was 1.25 × 10−12 cm2 s−1 and 8.14 × 10−11 cm2 s−1, respectively. Similar to the result calculated by the Randles-Sevcik equation, BI-NVP has a considerably higher DNa value than that of the NVP.
Furthermore, EIS was conducted at various temperatures in the range of 40 °C–90 °C at the charge state potential of 3.4 V vs Na/Na+. These results are shown in Figs. 3(e) and S10. The charge transfer resistance RCT is estimated by extrapolating the corresponding semicircle at each temperature (see Table S3 in the supplementary material). At all temperatures, the RCT values of the BI-NVP are lower than those of the NVP. Figure 3(f) shows the reciprocal temperature (1000/T) vs ln(RCT−1) (R2 > 0.99), which obeys the Arrhenius equation [Eq. (4)].
where A is the frequency factor, Ea is the activation energy for interfacial charge transfer, R is the universal gas constant, and T is the absolute temperature (K).76 The BI-NVP cell exhibits Ea of 3.96 kJ mol−1 vs 6.77 kJ mol−1 for NVP.
The electrochemical performance of BI-NVP and baseline NVP are investigated through galvanostatic charge-discharge (GCD) measurements. The GCD profiles at various C-rates are shown in Fig. 4(a). The plateau region is observed around 3.4 V vs Na/Na+ in good agreement with the CV results. For BI-NVP, the 1 C-rate capacity of 114 mA h g−1 is very close to the theoretical capacity of 117 mA h g−1. When the C rate is increased to 100 C, the BI-NVP exhibits a capacity retention of 59%, delivering a discharge specific capacity of 68 mA h g−1. These results are shown in Fig. 4(b). In this C-rate range, the Coulombic efficiencies (CEs) are approaching 100%, which indicates good reversibility. Even at an ultra-high rate of 200 C, the voltage plateau is retained and a capacity of 49 mA h g−1 is achieved. The BI architecture also contributes to excellent cyclic stability, allowing for the necessary expansion-contraction without the electrode pulverizing. Cycling stability tests were conducted at 1 and 50 C, as shown in Figs. 4(c) and 4(d). At the 50 C-rate, the capacity retention of the BI-NVP with respect to the first cycle is 91% after 10 000 cycles. A CE approaching 100% is achieved during the cycling. At 1 C-rate, the capacity retention is 80% after 10 000 cycles, while that of NVP is 12% after 2000 cycles. It notes that the inherently low electronic conductivity and volume variation of NVP during an ion insertion/desertion causes the sluggish charge storage kinetics and the micro-stress generated during a long cycling, resulting in structural demolition.19,77 On the other hand, the outer dense layer of mechanically robust conductive rGO, the strong bonding between rGO and NVP, and the hierarchical structure of BI-NVP could effectively delocalize electrochemical and mechanical stress, inhibiting permanent damage for improved cyclability. The cycling performance of BI-NVP is among the most favorable reported, as demonstrated in Fig. S11 and Table S4 in the supplementary material.39,68,78–93
We calculated the specific energy and power of a SIB based on a BI-NVP cathode pairing with a Na metal anode (see Supplementary Note 1). As shown in the Ragone plot in Fig. 4(e), this SIB shows highly favorable specific energy–power characteristics, including a relatively flat energy profile, which is needed for fast and ultrafast charging applications. The specific energy profile remains relatively flat even for <1-min charge times. The maximum specific power and specific energy are 154 kW kg−1 and 350 W h kg−1 (218 kW L−1 and 495 W h L−1), respectively. For example, when the cell is fully charged within 3.6 s, a specific energy–specific power of 93 W h kg−1 and 86 kW kg−1 are achieved. The fast charging and high-power capabilities are comparable to hybrid supercapacitors and approaches that of electric double layer capacitor (EDLC) based systems.30 Meanwhile the specific energy is superior to most SIBs, primarily due to the use of a Na metal rather than a hard carbon anode. Figure S12 and Table S5 in the supplementary material display the Ragone chart characteristics of state-of-the-art SIBs from literatures, including systems based on a Na metal anode coupled to NFPP/HC, Prussian blue@C, NMFPP, NFPP@rGO, and NFP cathodes.32–36 It may be observed that BI-NVP is among the most favorable in terms of energy and power.
To evaluate the high power characteristics of a BI-NVP cathode, the full SIB cell test was conducted pairing with a phosphorous-incorporated nanoporous carbon (P-aCN) anode.94 The P-aCN exhibited the discharge capacities of 113 and 39 mA h g−1 at 1000 and 10 000 mA g−1 in a half-cell, respectively, with 96% and 99% of initial CE (see Fig. S13 in the supplementary material). The full SIB cell was balanced controlling the mass ratio of BI-NVP and P-aCN on a basis of their reversible discharge capacities (see Supplementary Note 2). Consequently, the full SIB cell achieved the maximum discharge capacity of 108 mA h g−1 at 1 C with a CE of 95%, delivering the energy density of 134 W h kg−1 (Fig. S14 in the supplementary material). Furthermore, the capacity retention of the full SIB cell was 85% over 10 000 cycles at a very high rate of 100 C.
C. In situ and postmortem analyses
In situ Raman measurements were performed to analyze the structural change of BI-NVP during the charging and discharging (see Supplementary Note 3). Figure S15 in the supplementary material shows the CV curve of BI-NVP and in situ Raman spectra collected at each potential point. There were negligible changes in the D and G bands of rGO during a charging and discharging process. Unfortunately, the variation in the electronic structure of BI-NVP could not be captured due to the overlapping with Raman peaks of electrolyte solutions. These results imply that when the BI-NVP stores Na+ ion at positive potential, the rGO shell serves to facilitate charge transfer, not contributing to a faradaic process of storing Na+ ion.
Postmortem analyses were conducted to better understand the long-term cyclic stability of the BI-NVP SIB. For these analyses, the BI-NVP SIB was cycled 5000 times at 1 C. It may be observed that the overall morphology of BI-NVP composite is preserved, with minimal cycling-induced pulverization, etc. [Fig. 5(a)]. The HRTEM image of the post-cycled electrode shown in Fig. 5(b) confirms that the interface between the NVP and rGO layer remains intact, without the two separating. The distributions of the elements along the rGO–NVP interface was analyzed by scanning transmission electron microscopy-electron energy-loss spectroscopy (STEM-EELS), as shown in Figs. 5(c) and 5(d). The graphene layer with a thickness of approximately 6 nm uniformly covers the surface of the NVP. This is consistent with the HRTEM results and supports the cycling stability of BI-NVP observed electrochemically. Figure 5(e) shows the XRD pattern of the post-cycled BI-NVP electrode. All diffraction peaks, except for those of the polyvinylidene fluoride binder and aluminum current collector, can be indexed to the NVP phase. As verified by Rietveld refinement, the crystal parameters of BI-NVP after cycles were estimated as a = b = 8.7175 Å and c = 21.7941 Å with reliability values of Rp = 13.2% and Rwp = 13.3% (shown in Fig. S16 and Table S6 in the supplementary material), which is consistent with pristine BI-NVP and previous literatures.95 The minor difference of these parameters before and after cycles indicates the negligible lattice distortion at a fully charged state, which is associated with the structural integrity of biomimetic composite architecture as confirmed by electrochemical and other postmortem analyses.
Postmortem XRD analysis at each potential was also conducted to monitor the change in the crystalline structure throughout the Na-ion insertion/deinsertion process. Non-cycled samples were employed for this work (see Supplementary Note 4). Figure 5(f) and Fig. S17 show the XRD patterns of BI-NVP and NVP, respectively, at various states of charge/discharge in the range of 3.0 V to 3.9 V, tested at 10 C-rate. The shifts in the (113) Bragg peaks are employed as a tracking marker, as it best captures the dilation associated with Na ion insertion into the Na2 site in the crystal structure. It may be observed that the contraction-dilation with BI-NVP is highly reversible, with the peak shifting to high-low angle according to the charge-discharge. On the contrary, even at the first cycle, the (113) diffraction peak for the baseline NVP is irreversibly shifted to small angles during the charging and discharging, indicating the irreversible expansion of its representative channel. The increase in d-spacing of the NVP was seven times greater than that of the BI-NVP, due to the residual Na ions inside the lattice by the sluggish redox kinetics.
Inspired by the tough and resilient mammalian bone structure, we fabricated a BI-NVP-based sodium metal battery with ultra-high power, high energy, and extended cycling life. The maximum specific power and specific energy were 154 kW kg−1 and 350 W h kg−1, respectively. The BI-NVP cell delivered a high capacity of 114 mA h g−1 at 1 C, almost identical to the theoretical capacity of 117 mA h g−1, and specific capacity of 49 mA h g−1 at 200 C. Tested at 50 C, the capacity retention of the BI-NVP with respect to the first cycle was 91% at cycle 10 000, with a CE approaching 100%. When the cell was fully charged within 3.6 s (at 1000 C), an energy of 93 W h kg−1 was stored, which is among the most favorable high power capabilities reported for a SIB. Compared to the previously reported SIBs, the BI-NVP-based cell exhibited the highest energy in the ultrahigh power regime above 10 kW kg−1. The facile and reversible redox kinetics of the BI-NVP cell were confirmed by the small peak-to-peak separation, high b values and diffusion coefficient, and low activation energy. The cyclic stability of BI-NVP is confirmed by the postmortem and in situ XRD analyses. This study provides a new bioinspired approach for the designs of fast charge–long cycle life sodium storage materials.
IV. EXPERIMENTAL METHODS
The baseline NVP was prepared by the simple sol–gel method. Initially, 0.192 g of citric acid (Sigma-Aldrich, 98%) and 0.378 g of oxalic acid dehydrate (Daejung, 99.5%) were dissolved in 25 ml of de-ionized water to prepare a reducing and chelating agent. Subsequently, 0.117 g of NH4VO3 (Sigma-Aldrich, 99%) was dissolved in the reducing agent (mixture of citric and oxalic acid) at 80 °C. A solution (25 ml) containing 0.06 g of NaOH (Sigma Aldrich, >98%) and 0.173 g of NH4H2PO4 (Sigma Aldrich, >99%) was then prepared and added to the first solution. The obtained NVP precursor solution was stirred at 80 °C until its color turned dark blue. Subsequently, the solution was dried in a conventional oven at 80° C for 72 h until the solvent was completely evaporated. The dried gel was milled in a mortar, transferred to a furnace, and heated at 400 °C for 4 h in an argon atmosphere. The resulting powder was milled and heated again at 750 °C for 12 h in an argon atmosphere to obtain highly crystalline powders. The BI-NVP was prepared by mixing the as-prepared NVP precursor solution and GO solution. The NVP precursor solution obtained by the above procedure was stirred with 10 ml of a 0.5 wt. % GO solution for 12 h. The mixed solution was instantly frozen in liquid nitrogen for 30 min and subsequently vacuum-dried for 72 h. The dried sample was milled in a mortar, transferred to a furnace, and heated according to the above two-step process to obtain a highly crystalline powder and reduce the functional groups on the graphene oxide surface.
To characterize the morphology and structure of NVP and BI-NVP, field-emission scanning electron microscopy (FESEM, LEO SUPRA 55) and ultra-corrected energy filtering transmission electron microscope (UC-EF-TEM, Libra 200 HT Mc Cs, 200 kV) were used. The scanning transmission electron microscopy-electron energy-loss spectroscopy (STEM-EELS) elemental mapping and line profiling was conducted using a Schottky field emission type electron gun (ZrO/W emitter). The carbon content of BI-NVP was measured by thermogravimetric analysis (TGA) at a heating rate of 5° C min−1 from room temperature to 1000 °C in air. X-ray diffraction (XRD) analysis was performed using a D8 Advance (3 kW) with a θ/θ goniometer that was equipped with a Cu Kα radiation generator to confirm the composition and phase purity. The crystalline structure after cyclic test was analyzed using Rietveld refinement using FullProf™ software. The Raman spectra were obtained at room temperature using a Jobin Yvon/HORIBA HR evolution Raman spectrometer equipped with an integral BX 41 confocal microscope. Radiation from an air-cooled frequency-doubled Nd:YAG laser (532 nm) was used as the excitation source. Raman scattering was detected at a 180° geometry using a multichannel air-cooled (–60 °C) charge-coupled device (CCD) camera (1024 × 256 pixels). X-ray photoelectron spectroscopy (XPS) spectra were measured with monochromatic Al Kα (AXIS-NOVA and Ultra DLD, 1486.6 eV). N2 adsorption/desorption isotherms were obtained by the Brunauer-Emmett-Teller (BET) apparatus (BELSORP-max). The BET method was used to calculate the specific surface area of samples. The pore size distributions are exhibited by the NLDFT mode.
C. Electrochemical analysis
The electrochemical properties of the NVP and BI-NVP SIB cell were investigated using a coin-cell system (CR2032). For the fabrication of positive electrodes, active material, carbon black, and polyvinylidene fluoride (PVDF) were dispersed in N-methyle-2-pyrrolidone (NMP) with the mass ratio of 8:1:1. Then the slurry was coated on aluminum foil with the areal mass loading of 1.0 mg cm−2 and the electrode tap density of 1.42 mg cm−3. A sodium metal foil was used as negative electrode. A glass microfiber filter (Whatman) was used as the separator. The electrolyte is 1.0 M NaClO4 dissolved in propylene carbonate (PC) solution. All the coin cells were assembled in an argon-filled glovebox and tested at room temperature. The galvanostatic charge/discharge (GCD) tests were conducted on a battery cycler (Maccor, Series4000). Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) measurements were performed using a potentiostat/EIS system (Bio-Logic, VMP3). All the C-rates and gravimetric capacities were normalized by the mass of active material in a cathode (1 C = 117 mA h g−1). The specific energy and specific power of the SIB cell were normalized based on the active masses of both the BI-NVP cathode and Na metal anode. The mass balance of BI-NVP cathode and Na metal anode is fixed at 10:1 to equalize the capacity of both electrodes (see Supplementary Note 1).
See the supplementary material associated with this article.
All authors contributed equally to this manuscript. All authors reviewed the final manuscript.
H.S.P., K.H.S., S.K.P., P.N., and M.S.C. would like to acknowledge the financial support from the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (Grant No. NRF-2020R1A3B2079803), Republic of Korea. D.M., Y.W., and P.L. (research co-conception, research guidance, and manuscript preparation) were supported by the National Science Foundation, Division of Materials Research, Award No. 1938833. S.M.B. (analysis) at Brookhaven National Laboratory was supported by the Assistant Secretary for Energy Efficiency and Renewable Energy, Vehicle Technology Office of the U.S. Department of Energy through the Advanced Battery Materials Research (BMR) Program under Contract No. DE-SC0012704.
The authors declare no competing financial interest.
The data that support the findings of this study are available from the corresponding author upon reasonable request.