Scalable synthesis of a foam-like FeS₂ nanostructure by a solution combustion–sulfurization process for high-capacity sodium-ion batteries

Abstract

Pyrite-type FeS₂ is regarded as a promising anode material for sodium ion batteries. The synthesis of FeS₂ in large quantities accompanied by an improved cycling stability, as well as retaining high theoretical capacity, is highly desirable for its commercialization. Herein, we present a scalable and simple strategy to prepare a foam-like FeS₂ (F-FeS₂) nanostructure by combining solution combustion synthesis and solid-state sulfurization. The obtained F-FeS₂ product is highly uniform and built from interconnected FeS₂ nanoparticles (∼50 nm). The interconnected feature, small particle sizes and porous structure endow the product with high electrical conductivity, good ion diffusion kinetics, and high inhibition capacity of volume expansion. As a result, high capacity (823 mA h g⁻¹ at 0.1 A g⁻¹, very close to the theoretical capacity of FeS₂, 894 mA h g⁻¹), good rate capability (581 mA h g⁻¹ at 5.0 A g⁻¹) and cyclability (754 mA h g⁻¹ at 0.2 A g⁻¹ with 97% retention after 80 cycles) can be achieved. The sodium storage mechanism has been proved to be a combination of intercalation and conversion reactions based on in situ XRD. Furthermore, high pseudocapacitive contribution (i.e. ∼87.5% at 5.0 mV s⁻¹) accounts for the outstanding electrochemical performance of F-FeS₂ at high rates.

---

Introduction

Over the past decades, lithium-ion batteries (LIBs) have dominated the commercial market of energy storage.^1–7 However, the limited resources and high price of lithium hinder their wide application.^8–10 Hence, sodium-ion batteries (SIBs) become one of the most promising alternatives to LIBs considering the similar working mechanism and abundant resources.^11–13 As the size of the Na⁺ ion (1.02 Å) is larger than that of the Li⁺ ion (0.69 Å), it is crucial to develop viable Na-host materials for commercializing the SIB technology.

Pyrite-type FeS₂ has been widely regarded as one of the promising anode materials for SIBs owing to its high theoretical capacity (894 mA h g⁻¹), abundance, cost-effectiveness and eco-friendly nature.^14–16 However, FeS₂-based electrodes suffer from volume expansion up to 280% during the cycling process.¹⁷ Much effort has been devoted to optimize the electrochemical performance of this material. For example, some researchers tuned the cut-off voltage to avoid large volume change,^18–20 and others composited FeS₂ with carbon to buffer the expansion.^21–23 Although improved cyclability has been achieved through reported methods, their low yield and partial capacity sacrifice have posed a great challenge towards commercialization.^24,25 Thus, exploring a facile and cost effective strategy for the large-scale production of FeS₂ while retaining high theoretical capacity is highly desirable.

Herein, we report a scalable and simple strategy to prepare a FeS₂ nanostructure by combining solution combustion synthesis and solid-state reaction. The first step released large amounts of gaseous products and heat, yielding a porous Fe₂O₃ precursor by a solution combustion procedure. Then, the precursor was sulfurized to synthesize a foam-like FeS₂ (F-FeS₂) nanostructure. The uniform F-FeS₂ exhibits high capacity, good rate capability and cyclability. It is found that the unique foam-like structure of F-FeS₂ accounts for its good electrical conductivity, ion diffusion kinetics, and tolerance ability of volume expansion. Furthermore, in situ X-ray diffraction (XRD) has been conducted to study the sodium storage mechanism. This facile strategy could open opportunities for preparing other foam-like metal sulfide materials in considerable amount and with impressive performance in electrochemical fields.

Experimental

Materials

Iron(III) nitrate nonahydrate (Fe(NO₃)₃·9H₂O, ≥98.5%, Sinopharm Chemical Reagent), glycine (C₂H₅NO₂, 99.5–100.5%, Macklin), sublimed sulfur (S, ≥99.5%, Sinopharm Chemical Reagent), sodium hydroxide (NaOH, ≥97%, Macklin), and iron(III) chloride hexahydrate (FeCl₃·6H₂O, ≥99%, Macklin) were directly used as received without any purification.

Synthesis of the porous Fe₂O₃ precursor

The Fe₂O₃ precursor was prepared by a simple solution combustion synthesis.²⁶ 15 mmol Fe(NO₃)₃·9H₂O was first dissolved in 50 mL water. Then 15 mmol C₂H₅NO₂ was added to the as-prepared solution under stirring at 80 °C. Subsequently, the mixture solution was continuously stirred at the same temperature for 2 h. The resultant solution was transferred into a preheated evaporating dish. After removing the water, the mixture became a homogeneous colloid and self-ignited rapidly. With a great deal of smoke spurting out, the foam-like Fe₂O₃ precursor was prepared. This synthesis process (Video S1†) and the digital photo of the foam-like Fe₂O₃ precursor (Fig. S1†) can be found in the ESI.†

Synthesis of foam-like FeS₂ (F-FeS₂)

In a typical synthesis, 20 mg of the as-obtained Fe₂O₃ precursor and 25 mg of sublimed sulfur were ground together gently and loaded into a quartz tube evacuated with a mechanical pump and sealed by using a Partulab MRVS-1002 system. The sealed tubes were placed in a muffle furnace and annealed at 450 °C for 3 h. The heating rate was 10 °C min⁻¹ before the temperature reached 400 °C. After that, the heating rate was changed to 5 °C min⁻¹. The macroporous FeS₂ was obtained after cooling to ambient temperature. The sulfurized quantity of the Fe₂O₃ precursor can be easily extended to be more than 500 mg in one synthesis step, by using quartz tubes with a larger length and diameter.

Synthesis of routine FeS₂

For comparison, the routine FeS₂ was also prepared via a conventional precipitation method. Firstly, 150 mmol NaOH was dissolved in 100 mL water. Subsequently, 50 mmol FeCl₃ was dissolved in 50 mL water and added dropwise into the above solution under stirring. The obtained precipitate was washed three times with water and then annealed at 600 °C under air for 5 h to form Fe₂O₃. The as-prepared Fe₂O₃ was sulfurized with the same method mentioned above.

Materials characterization

The phase of the as-obtained products was confirmed by X-ray diffraction (XRD) on a Bruker D8 Advance diffractometer (40 kV, 40 mA, Cu Kα radiation, λ = 0.15418 nm), and the measurement settings were 0.02° and 0.25 seconds per step. X-ray photoelectron spectra (XPS) were recorded using a Thermo ESCALAB 250Xi system with monochromatic Al Kα radiation as the excitation source. Scanning electron microscopy (SEM, JEOL JSM-7001F) and transmission electron microscopy (TEM, Tecnai G2 F30 S-TWIN) were used to characterize the morphology. Energy dispersive X-ray spectroscopy (EDX) elemental mapping and selected area electron diffraction (SAED) were conducted on a Tecnai G2 F30 S-TWIN system. Nitrogen adsorption–desorption isotherms were recorded at 77 K on a Micromeritics TriStar II surface area and porosity instrument. Inductively coupled plasma (ICP) determined elemental constituents were measured on a PerkinElmer RFIC system.

Electrochemical measurements

The electrode slurry was made up of 70 wt% active material, 10 wt% acetylene black, 10 wt% SWCNTs and 10 wt% polyvinylidene fluoride (PVDF) binder homogeneously mixed in 1-methyl-2-pyrrolidinone (NMP) solvent. After stirring for 12 h, the obtained slurry was pasted onto copper foil, and heated under air at 80 °C for 30 min, and then transferred into a vacuum oven for another 4 h of drying at 120 °C to remove NMP. The electrode mass loading is ∼1 mg cm⁻². Electrochemical measurements were carried out via CR2032 coin-type half cells using Na metal as the counter/reference electrode, Whatman glass microfiber filter GF/D membrane as the separator, and the solution of 1 M NaClO₄ in ethylene carbonate (EC)/dimethyl carbonate (DMC) (1 : 1 vol% with 5 vol% fluoroethylene carbonate (FEC)) as the electrolyte (∼0.4 mL per cell). All coin-type cells were assembled in an argon-filled glove box (H₂O ≤ 0.1 ppm, O₂ ≤ 0.1 ppm). Cyclic voltammetry (CV) tests were conducted at various scan rates (0.1–5.0 mV s⁻¹) in a potential window of 0.01–3 V. The electrochemical impedance spectra (EIS) were recorded over a frequency range of 0.01 to 10⁵ Hz with an amplitude of 5 mV. Both CV and EIS tests were conducted using a Biologic VMP3 system. Galvanostatic charge–discharge (GCD) tests were carried out on a LANHE battery tester from 0.01–3 V.

Results and discussion

The foam-like FeS₂ was synthesized as described in the synthesis section. For comparison, the routine FeS₂ prepared by a conventional precipitation method was also employed. Fig. 1a shows the XRD patterns of the obtained samples, and the related Rietveld refined patterns are shown in Fig. 1b and c. Both the patterns can be well indexed to pure pyrite FeS₂ (JCPDS no. 24-0076, space group 205). Besides, the lattice parameters evaluated by Rietveld refinement are a = b = c = 5.406 Å for F-FeS₂ and a = b = c = 5.407 Å for the routine FeS₂, which fit well with the JCPDS data (card no. 24-0076, a = b = c = 5.418 Å). The chemical compositions were further determined by XPS (Fig. 1d–g). For F-FeS₂, two peaks located at 720.8 and 708.0 eV can be assigned to Fe 2p_1/2 and Fe 2p_3/2, respectively (Fig. 1d), which are accompanied by the associated satellite peaks.²⁷ Two peaks around 164.5 and 163.3 eV can be attributed to the S 2p_1/2 and S 2p_3/2 species, respectively (Fig. 1e). The XPS spectra of FeS₂ show similar results to those of F-FeS₂. The Fe 2p_1/2 and Fe 2p_3/2 peaks are positioned at 720.7 and 707.9 eV, and the S 2p_1/2 and S 2p_3/2 ones are located at 164.4 and 163.2 eV, respectively (Fig. 1f and g).^28–30 The results from both XRD and XPS confirm that F-FeS₂ and FeS₂ have no difference in phase and composition.

Fig. 1 The phase and composition characterization of the as-prepared FeS₂. (a) XRD patterns of F-FeS₂ and FeS₂. Rietveld refined patterns of (b) F-FeS₂ and (c) FeS₂. XPS spectra of Fe 2p and S 2p of (d–e) F-FeS₂ and (f–g) FeS₂.

The morphology of F-FeS₂ was studied by SEM and TEM, and the surface area was determined using the BET method. The as-prepared F-FeS₂ has foam-like interconnected 3D porous morphology, as shown in Fig. 2a and Fig. S2 (ESI†). Such a porous structure is mainly formed due to the release of a large amount of gaseous products during liquid combustion synthesis.^31,32 Interestingly, the skeleton of the foam structure is composed of FeS₂ nanoparticles with diameters ranging from 25 to 120 nm (Fig. 2b), and most particles stick closely to each other, building up a continuous frame. The EDX elemental mapping results (Fig. 2c) show uniform distribution of Fe and S, confirming that the porous Fe₂O₃ was totally sulfurized into FeS₂. The SAED pattern shown in Fig. 2d indicates the polycrystalline nature of F-FeS₂. Moreover, the as-observed diffraction rings can be well indexed to the (111), (200), (210), (211), (220), (311) planes, respectively, agreeing well with the XRD result. In contrast, FeS₂ shows the morphology of aggregated particles with an average size of 230 nm (Fig. 2e). Meanwhile, the BET result in Fig. 2f indicates that F-FeS₂ offers a larger specific surface area (9.8 m² g⁻¹) than FeS₂ (5.3 m² g⁻¹). Obviously, the larger specific surface area of F-FeS₂ is attributed to the porous structure and the smaller particle sizes.

Fig. 2 (a) SEM and (b) TEM images of F-FeS₂. (c) EDX elemental mapping of Fe (green) and S (brown) in F-FeS₂. (d) SAED image of F-FeS₂. (e) SEM image of FeS₂. (f) Nitrogen adsorption–desorption isotherms of F-FeS₂ and FeS₂.

A half-cell configuration was assembled to evaluate the sodium storage behavior of F-FeS₂. Fig. 3a shows the initial five CV curves of the F-FeS₂ electrode recorded between 0.01 and 3 V at a scan rate of 0.1 mV s⁻¹. As can be seen from the cathodic scan, two groups of peaks can be observed. A sharp cathodic peak located at 1.07 V is observed in the first cycle, which disappears in the subsequent cycles. This indicates an irreversible reaction process. It is substituted with a new reduction peak at 2.12 V. Another group of peaks, belonging to the discharging process, can be detected below 0.76 V in all cycles. The associated in situ XRD patterns in Fig. 3b show that the intensity of the F-FeS₂ peaks gradually decreases during the discharging process and no peak shift can be observed. No other crystalline compound is detected until the Na₂S was formed during the discharging process. These observations indicate that the phase of F-FeS₂ has changed around 1 V and the iron sulphides converted into Na₂S below 0.7 V. For the charging process, two groups of peaks, associated with Na⁺ extraction, are located around 1.32 and 2.54 V as seen in Fig. 3a. Each group contains at least two obvious peaks, indicating the corresponding multi-stepped reactions.³³ Along the same process, no other diffraction peak change occurs except for the decomposition of Na₂S from the lowest potential. This decomposition accounts for the first group of reaction peaks in the charging process, whereas another group of peaks appear to correspond to a different desodiation process. Previous studies have reported that the formation of Na_xFeS₂ can be observed during the first cycle discharging process, as well as a charging/discharging potential above 0.8 V in all subsequent scans.^14,18 Besides, the lattice fringe of NaFeS₂ can be observed by ex situ HRTEM for the samples charged/discharged to the mentioned potentials (Fig. S3, ESI†) in our study. Moreover, our CV plots present a similar profile as in the previous report, which was associated with an intercalation reaction.³⁴ It is reasonable to believe that the intercalation reaction occurs in this electrochemical process. Hence, we propose the following reaction mechanism:

FeS₂ + xNa⁺ + xe⁻ → Na_xFeS₂ (x < 2) (1)

Na_x−yFeS₂ + yNa⁺ + ye⁻ ⇋ Na_xFeS₂ (y < x < 2) (2)

Na_xFeS₂ + (4 − x)Na⁺ + (4 − x)e⁻ ⇋ 2Na₂S + Fe (3)

Fig. 3 (a) Initial five CV cycles of F-FeS₂ for SIBs at 0.1 mV s⁻¹. (b) In situ XRD patterns of F-FeS₂ for SIBs from OC voltage discharging to 0.01 V, then recharging to 3 V at a rate of 0.2 A g⁻¹; the black symbols represent the diffraction peaks of Be and BeO, the red arrow represents the initial peaks of F-FeS₂ and the star mark represents the strongest Na₂S peaks. (c) Initial five GCD curves of F-FeS₂ for SIBs at 0.1 A g⁻¹.

Eqn (1) is associated with a sharp peak located at 1.07 V in the first cycle, and it is substituted by eqn (2) in the following cycles. The cathodic peak at 2.12 V and the anodic peaks around 2.54 V originate from this reaction. The irreversible phase change of FeS₂ matches well with the previous report.²¹Eqn (3) describes the reduction below 0.76 V and the oxidation around 1.32 V in all cycles. In particular, we observed the peak shift of Na₂S near 23.6° as seen from the magnified in situ XRD patterns in Fig. S4 (ESI†). The Na₂S diffraction peak associated with 0.01 V matches exactly with that of pure Na₂S. It is interesting to find that these peaks tend to appear at lower angles before discharging to 0.01 V, and keep shifting to larger angles after it. This may be attributed to the interphase generated in the conversion process, whose interplanar spacing shrinks from 3.79 to 3.74 Å during the process. In addition, the typical diffraction peaks of Be and BeO can also be seen in Fig. 3b, which are attributed to the Be window used in the in situ XRD test.

Fig. 3c presents the corresponding initial five GCD profiles of the F-FeS₂ anode at 0.1 A g⁻¹. The first cycle delivers a discharge specific capacity of 1341 mA h g⁻¹ and a charge specific capacity of 840 mA h g⁻¹. The initial capacity loss is due to the irreversible phase transformation reaction and the formation of a solid–electrolyte interface (SEI) layer. The SEI generating reaction also accounts for the capacity decrease between the first scan curve and the subsequent curves. This matches well with the intensity decay in the reduction peaks below 0.76 V and the oxidation peaks around 1.32 V shown in Fig. 3a. The discharge plateaus are observed around 2.1 V and 0.5 V, while the charge plateaus locate around 1.3 V and 2.5 V. This result also matches well with the redox peaks in the CV curves. Both CV curves and GCD profiles overlap well from the second cycle onward, demonstrating the good reversibility of the reaction.

The rate capabilities and cycling stabilities of the F-FeS₂ and FeS₂ electrodes are presented in Fig. 4a and b. The F-FeS₂ anode exhibits specific capacities of 823, 771, 721, 693, 662, 630, 608, 582, and 861 mA h g⁻¹ at current densities of 0.1, 0.2, 0.5, 1.0, 2.0, 3.0, 4.0, 5.0, and 0.1 A g⁻¹, respectively. These values are much higher than those of FeS₂ (Fig. 4a). Besides, the rate performance of F-FeS₂ demonstrates the highest specific capacities among previous reports (Fig. S5, ESI†). For cyclability, F-FeS₂ shows a specific capacity of 755 mA h g⁻¹ at 0.2 A g⁻¹ after 80 cycles, with 97% capacity retention. Meanwhile, the 20^th, 50^th, and 80^th GCD curves during cycling can be found in Fig. S6 (ESI†), and the similar profiles reveal the stability of our F-FeS₂. The coulombic efficiency below 99% after tens of cycles might be attributed to the side reactions between the ester-based electrolyte and sulfur anionic groups.^6,18,25,35 In comparison, FeS₂ only delivers 194 mA h g⁻¹ at 0.2 A g⁻¹ after 80 cycles, with 33% capacity retention. The slight increase of the discharge capacity of F-FeS₂ during the initial fifty cycles can be ascribed to the activation process, which has been observed in other transition metal sulfides.^36–38 In contrast, the specific capacity of FeS₂ decreases rapidly in the first few cycles. Considering the same phase and composition of F-FeS₂ and FeS₂, it is reasonable to believe that the foam-like structure enables the improved electrochemical performance of F-FeS₂. In addition, we also investigated the sodium storage performance of SWCNTs, and it displays a 5^th-cycle specific capacity of 91.8 mA h g⁻¹ at 0.1 A g⁻¹ (Fig. S7, ESI†). Taking the mass percentage (10% of the loading mass) into account, the capacity contribution from the SWCNT additive was negligible.

Fig. 4 (a) Rate capabilities of F-FeS₂ and FeS₂ for SIBs from 0.1–5.0 A g⁻¹. (b) Cycling stabilities of F-FeS₂ and FeS₂ at 0.2 A g⁻¹. (c) EIS of F-FeS₂ and FeS₂ half cells.

To reveal the reaction kinetics for such good Na-storage performance of F-FeS₂, both samples were subjected to EIS. Both samples show similar Nyquist plots obtained at open circuit voltage. A depressed semicircle at high frequency followed by a diffusion drift at low frequency can be clearly observed (Fig. 4c). Obviously, F-FeS₂ has a smaller depressed semicircle diameter than FeS₂ at high frequency. This suggests that the F-FeS₂ electrode has better charge transfer performance than FeS₂, which might be attributed to the interconnected feature of F-FeS₂. The low frequency line is associated with the ion diffusion of electrodes.¹⁵ The line slope of FeS₂ is smaller than that of F-FeS₂, indicating the faster transport of ions in the F-FeS₂ electrode cells. The porous structure and relatively small particle sizes offer a larger surface area. This allows more contact sites with the electrolyte, thus shortening the ion transporting distance.^15,20 The EIS spectra have also been fitted to an equivalent circuit, and the calculated resistances and Warburg coefficients are agreeable with the analysis above (Fig. S8, ESI†). All of these account for the high discharge capacity at each scan rate. Besides, the pores of the foam structure of FeS₂ offer more space for volume expansion during the sodium storing process. This guarantees the structural stability of F-FeS₂, resulting in a high capacity F-FeS₂ electrode with good cyclability.

To better understand the good high-rate performance of F-FeS₂, the reaction kinetics was further analyzed by distinguishing the surface-controlled pseudocapacitive capacity and diffusion-controlled capacity. Fig. 5a shows the CV curves of F-FeS₂ for various scan rates ranging from 0.1–5 mV s⁻¹. It can be found that the peaks of all the CV curves show a similar shape, although broadened peaks can be observed at increased scan rates. The preserved peaks even at high scan rates indicate small polarization during the reaction process.³⁹ The contribution of pseudocapacitive effects is qualitatively revealed by the b value from the equation log(i) = b log(v) + log(a) where both a and b are constants, and i and v represent the current and scan rate.^40–42 The series of peak currents marked in Fig. 5a were chosen to determine the b value by fitting lines based on this equation. The calculated b value of F-FeS₂ is 0.93. As the value of b is very close to 1, it suggests that there is a large contribution from the pseudocapacitive storage in the F-FeS₂ electrode.^43–45 To quantify the pseudocapacitive contribution, the equation i = k₁v + k₂v^0.5 was used according to previous reports, where i and v represent the current and scan rate, respectively, and k is a constant related to a fixed potential.^46–48 By fitting the value of k₁ at different voltage stages, we calculated a series of pseudocapacitive contributions to the current. As a result, the direct view of pseudocapacitive storage contribution (red part) for F-FeS₂ at 1.0 mV s⁻¹ is shown in Fig. 5c, and it is determined to be ∼82.5%. This indicates the surface pseudocapacitance-dominated process in the F-FeS₂ electrode. The overall pseudocapacitive contributions at different scan rates were obtained using the same method, as shown in Fig. 5d. As the scan rates increase, the pseudocapacitive contribution also increases gradually. A maximum value of ∼87.5% can be obtained at 5.0 mV s⁻¹. It suggests that the pseudocapacitive behavior dominates the whole reaction process, especially at high scan rates. It is well known that the surface-controlled pseudocapacitive process is much faster than the diffusion-controlled process.¹⁹ Such high pseudocapacitive contribution may be ascribed to the unique foam-like structure with relatively large surface area and small particle sizes, resulting in good rate capability of the F-FeS₂ electrode.

Fig. 5 (a) CV curves of F-FeS₂ for scan rates from 0.1 to 5 mV s⁻¹. (b) Functional relationship of peak current (i) vs. scan rate (v). (c) Capacity contribution of F-FeS₂ at 1 mV s⁻¹. (d) Capacitive contribution comparison at different scan rates.

Besides, the good electrochemical performance of F-FeS₂, which benefits from the foam-like structure, can also be explained from the structural mechanistic aspect. As shown in Fig. S9 (ESI†), Na, S and Fe elements distribute evenly throughout the 3^rd-cycle discharged material, indicating a thorough chemical conversion process. This means that our porous foam-like structure consisting of small FeS₂ nanoparticles is suitable for kinetically reversible reactions.³⁴ In addition, FeS₂ tends to react with alkali metals via a radial movement of a reaction front triggered by the contact of two materials.^49,50 Thus, the sufficient contact area and ion transport pathway offered by the porous structure are beneficial for sodium storage. Besides, Fe ions, based on their short diffusion length,⁵¹ tend to nucleate uniformly alongside Na₂S in the resultant particle during the conversion reaction process.^34,49,52 As a result, the conversion products only showed a slight increase in the diameters of the original particles, retaining the initial shape. Meanwhile, the neighboring pores in our porous foam-like structure can accommodate the associated volume expansion, maintaining the initial contact of particles. Hence, the interconnected foam-like structure can be preserved, even undergoing volume expansion after the discharge/charge process (Fig. S10, ESI†).

Conclusions

In summary, the facile and scalable synthesis of foam-like F-FeS₂ with high purity is successfully realized through the combination of solution combustion synthesis and solid-state sulfurization. It is observed that the interconnected nanoparticles make up the skeleton of porous F-FeS₂. For the sodium storage in F-FeS₂, CV analysis and in situ XRD techniques were used to prove that the reversible reaction is based on the combination of insertion and conversion reaction mechanisms. Benefiting from the unique structure, the as-prepared F-FeS₂ manifests impressive sodium storage properties with high specific capacity (823 mA h g⁻¹ at 0.1 A g⁻¹), good rate capability (581 mA h g⁻¹ at 5 A g⁻¹) and cyclability (754 mA h g⁻¹ at 0.2 A g⁻¹ with 97% retention after 80 cycles). Additionally, the resistance analysis explains its high capacity, and the proven pseudocapacitance contribution clarifies its good rate capability. This work holds great promise for commercializing high capacity FeS₂-based SIBs, offering scientific insights into the rational design of structures to enhance their electrochemical performance.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This investigation was supported by the Natural Science Foundation of China (No. 51472122, 51772152), the Fundamental Research Funds for the Central Universities (No. 30915011201, NJ20160030), the “333 project” of Jiangsu and the PAPD of Jiangsu.

References

1. Y. Fang, Y. Lv, F. Gong, A. A. Elzatahry, G. Zheng and D. Zhao, Adv. Mater., 2016, 28, 9385–9390.
2. S. Xia, J. Ni, S. V. Savilov and L. Li, Nano Energy, 2018, 45, 407–412.
3. Y. Zhang, Y. Wang, L. Wang, C.-M. Lo, Y. Zhao, Y. Jiao, G. Zheng and H. Peng, J. Mater. Chem. A, 2016, 4, 9002–9008.
4. X. Li, Y. Ma, G. Cao and Y. Qu, J. Mater. Chem. A, 2016, 4, 12487–12496.
5. R. Dangol, Z. Dai, A. Chaturvedi, Y. Zheng, Y. Zhang, K. N. Dinh, B. Li, Y. Zong and Q. Yan, Nanoscale, 2018, 10, 4890–4896.
6. W. Yao, J. Chen, L. Zhan, Y. Wang and S. Yang, ACS Appl. Mater. Interfaces, 2017, 9, 39371–39379.
7. C. Wang, S. Wang, Y.-B. He, L. Tang, C. Han, C. Yang, M. Wagemaker, B. Li, Q.-H. Yang, J.-K. Kim and F. Kang, Chem. Mater., 2015, 27, 5647–5656.
8. J. Ni and L. Li, Adv. Funct. Mater., 2018, 28, 1704880.
9. T. Wu, C. Zhang, H. Hou, P. Ge, G. Zou, W. Xu, S. Li, Z. Huang, T. Guo, M. Jing and X. Ji, Adv. Funct. Mater., 2018, 28, 1705744.
10. D. Wu, G. Zhang, D. Lu, L. Ma, Z. Xu, X. Xi, R. Liu, P. Liu and Y. Su, J. Mater. Chem. A, 2018, 6, 13613–13618.
11. X. Zhu, D. Liu, D. Zheng, G. Wang, X. Huang, J. Harris, D. Qu and D. Qu, J. Mater. Chem. A, 2018, 6, 13294–13301.
12. Z. Xu, K. Yao, Z. Li, L. Fu, H. Fu, J. Li, L. Cao and J. Huang, J. Mater. Chem. A, 2018, 6, 10535–10542.
13. P. F. Wang, H. Xin, T. T. Zuo, Q. Li, X. Yang, Y. X. Yin, X. Gao, X. Yu and Y. G. Guo, Angew. Chem., Int. Ed., 2018, 57, 8178–8183.
14. W. Chen, S. Qi, L. Guan, C. Liu, S. Cui, C. Shen and L. Mi, J. Mater. Chem. A, 2017, 5, 5332–5341.
15. B. H. Hou, Y. Y. Wang, J. Z. Guo, Q. L. Ning, X. T. Xi, W. L. Pang, A. M. Cao, X. Wang, J. P. Zhang and X. L. Wu, Nanoscale, 2018, 10, 9218–9225.
16. D. Li, Y. Sun, S. Chen, J. Yao, Y. Zhang, Y. Xia and D. Yang, ACS Appl. Mater. Interfaces, 2018, 10, 17175–17182.
17. M. Walter, T. Zund and M. V. Kovalenko, Nanoscale, 2015, 7, 9158–9163.
18. Z. Hu, Z. Zhu, F. Cheng, K. Zhang, J. Wang, C. Chen and J. Chen, Energy Environ. Sci., 2015, 8, 1309–1316.
19. W. Zhao, C. Guo and C. M. Li, J. Mater. Chem. A, 2017, 5, 19195–19202.
20. Y. Chen, X. Hu, B. Evanko, X. Sun, X. Li, T. Hou, S. Cai, C. Zheng, W. Hu and G. D. Stucky, Nano Energy, 2018, 46, 117–127.
21. Z. Liu, T. Lu, T. Song, X.-Y. Yu, X. W. Lou and U. Paik, Energy Environ. Sci., 2017, 10, 1576–1580.
22. M. J. Choi, J. Kim, J. K. Yoo, S. Yim, J. Jeon and Y. S. Jung, Small, 2018, 14, 1702816.
23. F. Bu, P. Xiao, J. Chen, M. F. Aly Aboud, I. Shakir and Y. Xu, J. Mater. Chem. A, 2018, 6, 6414–6421.
24. L. Li, S. Peng, N. Bucher, H.-Y. Chen, N. Shen, A. Nagasubramanian, E. Eldho, S. Hartung, S. Ramakrishna and M. Srinivasan, Nano Energy, 2017, 37, 81–89.
25. K. Zhang, M. Park, L. Zhou, G. H. Lee, J. Shin, Z. Hu, S. L. Chou, J. Chen and Y. M. Kang, Angew. Chem., Int. Ed., 2016, 55, 12822–12826.
26. K. Deshpande, A. Mukasyan and A. Varma, Chem. Mater., 2004, 16, 4896–4904.
27. J. Zhang, C. Du, Z. Dai, W. Chen, Y. Zheng, B. Li, Y. Zong, X. Wang, J. Zhu and Q. Yan, ACS Nano, 2017, 11, 10599–10607.
28. X. Wen, X. Wei, L. Yang and P. K. Shen, J. Mater. Chem. A, 2015, 3, 2090–2096.
29. Y. Liang, P. Bai, J. Zhou, T. Wang, B. Luo and S. Zheng, CrystEngComm, 2016, 18, 6262–6271.
30. W. Ma, X. Liu, X. Lei, Z. Yuan and Y. Ding, Chem. Eng. J., 2018, 334, 725–731.
31. W. Wen and J.-M. Wu, RSC Adv., 2014, 4, 58090–58100.
32. F. T. Li, Q. Wang, J. Ran, Y. J. Hao, X. J. Wang, D. Zhao and S. Z. Qiao, Nanoscale, 2015, 7, 1116–1126.
33. Z. Shadike, Y.-N. Zhou, F. Ding, L. Sang, K.-W. Nam, X.-Q. Yang and Z.-W. Fu, J. Power Sources, 2014, 260, 72–76.
34. A. Douglas, R. Carter, L. Oakes, K. Share, A. P. Cohn and C. L. Pint, ACS Nano, 2015, 9, 11156–11165.
35. Y. Zhou, J. Tian, H. Xu, J. Yang and Y. Qian, Energy Storage Mater., 2017, 6, 149–156.
36. B. H. Hou, Y. Y. Wang, J. Z. Guo, Y. Zhang, Q. L. Ning, Y. Yang, W. H. Li, J. P. Zhang, X. L. Wang and X. L. Wu, ACS Appl. Mater. Interfaces, 2018, 10, 3581–3589.
37. D. Li, D. Yang, X. Yang, Y. Wang, Z. Guo, Y. Xia, S. Sun and S. Guo, Angew. Chem., Int. Ed., 2016, 55, 15925–15928.
38. J. Li, D. Yan, X. Zhang, S. Hou, T. Lu, Y. Yao and L. Pan, J. Mater. Chem. A, 2017, 5, 20428–20438.
39. Z. Lu, N. Wang, Y. Zhang, P. Xue, M. Guo, B. Tang, Z. Bai and S. Dou, Electrochim. Acta, 2018, 260, 755–761.
40. W. Ren, H. Zhang, C. Guan and C. Cheng, Adv. Funct. Mater., 2017, 27, 1702116.
41. R. Sun, S. Liu, Q. Wei, J. Sheng, S. Zhu, Q. An and L. Mai, Small, 2017, 13, 1701744.
42. P. Ge, C. Zhang, H. Hou, B. Wu, L. Zhou, S. Li, T. Wu, J. Hu, L. Mai and X. Ji, Nano Energy, 2018, 48, 617–629.
43. G. Zou, H. Hou, C. W. Foster, C. E. Banks, T. Guo, Y. Jiang, Y. Zhang and X. Ji, Adv. Sci., 2018, 5, 1800241.
44. Y. Long, J. Yang, X. Gao, X. Xu, W. Fan, J. Yang, S. Hou and Y. Qian, ACS Appl. Mater. Interfaces, 2018, 10, 10945–10954.
45. F. Niu, J. Yang, N. Wang, D. Zhang, W. Fan, J. Yang and Y. Qian, Adv. Funct. Mater., 2017, 27, 1700522.
46. C. Zhang, S.-H. Park, S. E. O'Brien, A. Seral-Ascaso, M. Liang, D. Hanlon, D. Krishnan, A. Crossley, N. McEvoy, J. N. Coleman and V. Nicolosi, Nano Energy, 2017, 39, 151–161.
47. D. Yu, Q. Pang, Y. Gao, Y. Wei, C. Wang, G. Chen and F. Du, Energy Storage Mater., 2018, 11, 1–7.
48. P. He, M. Yan, G. Zhang, R. Sun, L. Chen, Q. An and L. Mai, Adv. Energy Mater., 2017, 7, 1601920.
49. M. T. McDowell, Z. Lu, K. J. Koski, J. H. Yu, G. Zheng and Y. Cui, Nano Lett., 2015, 15, 1264–1271.
50. Y. Sun, N. Liu and Y. Cui, Nat. Energy, 2016, 1, 16071.
51. J. H. Chen and W. W. Harvey, Metall. Trans. B, 1975, 6, 331.
52. F. Wang, H. C. Yu, M. H. Chen, L. Wu, N. Pereira, K. Thornton, A. Van der Ven, Y. Zhu, G. G. Amatucci and J. Graetz, Nat. Commun., 2012, 3, 1201.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8nr06675b

---