Fast Ionic Diffusion-Enabled Nanoflake Electrode by Spontaneous Electrochemical Pre-Intercalation for High-Performance Supercapacitor

Abstract

Layered intercalation compounds Na_xMnO₂ (x = 0.7 and 0.91) nanoflakes have been prepared directly through wet electrochemical process with Na⁺ ions intercalated into MnO₂ interlayers spontaneously. The as-prepared Na_xMnO₂ nanoflake based supercapacitors exhibit faster ionic diffusion with enhanced redox peaks, tenfold-higher energy densities up to 110 Wh·kg⁻¹ and higher capacitances over 1000 F·g⁻¹ in aqueous sodium system compared with traditional MnO₂ supercapacitors. Due to the free-standing electrode structure and suitable crystal structure, Na_xMnO₂ nanoflake electrodes also maintain outstanding electrochemical stability with capacitance retention up to 99.9% after 1000 cycles. Besides, pre-intercalation effect is further studied to explain this enhanced electrochemical performance. This study indicates that the suitable pre-intercalation is effective to improve the diffusion of electrolyte cations and other electrochemical performance for layered oxides and suggests that the as-obtained nanoflakes are promising materials to achieve the hybridization of both high energy and power density for advanced supercapacitors.

Introduction

Supercapacitors, with high power density and long operating life, are expected to provide potential access to bridge the gap between batteries and conventional capacitors^{1,2,3,4,5,6,7,8,9,10}. However, one of the key challenges for supercapacitors to achieve wide application is increasing their energy density^1,3.

As one of the most promising pseudocapacitive electrode materials, MnO₂ has a relatively high energy density because of its high reversible redox reaction. For example, it can exhibit a redox reaction in sodium electrolyte system based on the following charge storage mechanism (Equations (1))^3,11:

Here are three rate-controlling steps involved in this reaction during the Mn III/IV oxidation state change: (i) Adsorption of electrochemically active ions on the surface of electrode; (ii) Electronic transfer drived by current collector; (iii) Na⁺ (or H⁺) ions diffusion and intercalation/deintercalation.

For rate-controlling step (i), large active surface area is important to enhance the ionic adsorption. For rate-controlling step (ii), we can electrodeposit MnO₂ on the nickel foam to obtain the continuous electronic transfer path. While for rate-controlling step (iii), it is important to improve the long-distance ionic diffusion ability from electrolyte to the solid phase, which is very difficult for traditional MnO₂^12,13. To solve this problem, researchers reduce the diffusion resistance in the solid phase by decreasing the size and shortening the diffusion paths^13,14,15,16, facilitated the ionic diffusion between MnO₂ and electrolyte by constructing an interconnected porous system^{10,17,18,19,20,21,22}, or even loading very small quantity of active mass to achieve high capacitance. However, those strategies can improve the diffusion and intercalation/deintercalation only on the surface or the interface, the low ionic diffusion inside the solid phase is still unsolved fundamentally.

Recently, sodium-intercalation layered compounds Na_xMO₂ (M = Mn^23,24,25, Co^26,27, V²⁸, etc) were synthesized to deliver high electrochemical activity of Na⁺ ions due to their interesting crystal structure. Among them, Na_0.7MnO₂ and Na_0.91MnO₂ are electrochemically attractive due to high sodium cationic distribution and layered intercalation structure²³. However, all those reported Na_xMnO₂ were synthesized by high-temperature solid reactions and that will greatly increase the products' practical energy consumption, complexity and crystal size, which is not suitable for supercapacitors.

In our previous work, we have successfully intercalated metallic ions to improve electrochemical performance for energy storage by secondary hydrothermal process²⁹. Here, to improve the diffusion performance of Na⁺ ions in MnO₂ solid phase, we used a one-step electrodeposition technique, based on the atomic-level control of material composition³⁰, to rationally obtain electrochemically active layered Na_xMnO₂ (x ≥ 0.7) phases (Figure 1a and b). Na⁺ was intercalated spontaneously in MnO₂ crystal interlayers without any rigorous condition such as thermal treatment or complex solid reactions. The as-obtained nanoflake electrode exhibited good Na⁺ diffusion behaviors and excellent electrochemical performance in 1 M Na₂SO₄ solution even when the loading mass was as large as 2.25 mg·cm⁻². Such good results demonstrate that our rational, facile, mild and controllable method to produce Na_xMnO₂ naomaterials is promising for fabrication of high-level energy devices for large-scale practical application^31,32,33.

Figure 1

Phase characterization and schematic Na⁺ intercalation mechanism of pre-intercalated MnO₂ nanoflake electrodes.

(a) Typical XRD of Na_xMnO₂/MnO₂ with electrodeposition time: 20 min (yellow line); 100 min (red line); 200 min (blue line). Symbols *, ⋆ respectively represent Na_0.7MnO₂, Na_0.91MnO₂ and ◊, represent MnO₂. (b) Proposed mechanism for intercalation of Na⁺ ions into MnO₂ interlayers.

Results

According to XRD patterns (Figure 1a), the phase of electrodeposited products changes with deposition time. The Na_0.7MnO₂ (JCPDS No. 00-027-0752) is deposited in the initial 20 min. Then Na_0.91MnO₂ (JCPDS No. 00-038-0965) and MnO₂ (JCPDS No. 00-011-0055) are produced with more Na⁺ intercalated into the layer structure during 20–100 min. With time extended to 200 min, the Na_0.7MnO₂ phase changes to Na_0.91MnO₂. The Atomic Adsorption Spectrum (AAS) was measured to demonstrate the ratio of Na and Mn (Supplementary information, Table 1) and the analysis is identical with the XRD results. Energy dispersive X-ray spectrometry (EDS) mapping (Supplementary information, Figure S1a) further confirms that Na, Mn and O atoms are homogeneously distributed in all of the nanoflakes. Besides, XRD of hydrothermally synthesized α-MnO₂ (JCPDS No. 00-044-0141) controlled nanowires is also given for comparison with Na_xMnO₂.

Typical scanning electron microscopic (SEM) images (Figure 2a) show that excessive agglomerations are prevented effectively. A loose and interconnected nanoflake structure with large surface areas is obtained and these nanoflakes were closely attached to nickel surface (electron path), forming good electrochemical connection. Significantly, these ultrathin nanoflakes with large surface area are beneficial for the adsorption of the electroactive ions and contact with electrolyte. As a result, diffusion and intercalation/deintercalation of Na⁺ ions have been facilitated by the nanoflake morphology. The high-resolution transmission electron microscopic (HRTEM) investigation (Figure 2b, Supplementary Figure S1b) reveals polycrystalline structure of the nanoflake. By comparing interplanar distances measured from HRTEM with the theoretical value, the existence of Na_0.91MnO₂ and Na_0.7MnO₂ is confirmed, consistent with the results from XRD crystal analysis.

Figure 2

Morphology characterization of pre-intercalated MnO₂ nanoflake electrodes.

(a, b) SEM images of the electrodeposited uniform nanoflakes on the nickel skeleton with average thickness of 15–20 nm. (c) HRTEM image of Na_xMnO₂/MnO₂ nanoflakes after 200 min electrodeposition.

Cyclic Voltammetry tests (CVs) were performed to identify the charge storage mechanism in 1 M Na₂SO₄. According to Figure 3a, obviously a pair of redox peaks appear in the CV curve of Na_xMnO₂/MnO₂ electrode while no redox peaks are found in the CV curve of α-MnO₂ sample. It is interesting that according to many reported work, no such peaks related to redox reaction of MnO₂ have been ever observed with respect to α-MnO₂^34,35,36, β-MnO₂³⁶, γ-MnO₂³⁶, δ-MnO₂³⁶, ε-MnO₂^10,34, nanocrystalline MnO₂²⁰, MnO₂ composite electrodes^17,18 and so on when the same Na₂SO₄ is used as electrolyte. The occurrence of redox peaks implies that the charge storage mechanism is mainly based on redox reaction. The potential difference between the anodic and cathodic peak is about 0.06 V, representing a highly reversible intercalation/deintercalation during the whole electrochemical process.

Figure 3

Cyclic voltammograms of Na_xMnO₂/MnO₂ electrodes in 1M Na₂SO₄.

(a) Cyclic voltammograms of Na_xMnO₂/MnO₂ (electrodeposition time: 200 min) and pure α-MnO₂ at scan rate of 1 mV·s⁻¹; inset: comparison of discharge process between Na_xMnO₂/MnO₂ and α-MnO₂ at current density of 1 mA·cm⁻². (b) Cyclic voltammograms of Na_xMnO₂/MnO₂ (electrodeposition time: 200 min) at different scan rates from 1 mV·s⁻¹ to 500 mV·s⁻¹; inset: the capacitance calculated by CV curve at corresponding scan rate, potential window: 0–0.8 V (vs. SCE).

According to Figure 3b, with the increase of the scan rate, the redox current gets larger, the anodic peak shifts towards positive potential and the cathodic peak shifts towards negative potential. And the inset shows that, the specific capacitance of the electrode increase as the scan rate decrease from 500 to 1 mV·s⁻¹. As observed in both Figure 3a inset and Figure 4a inset b, it is interesting that the slopes of potential-time curves change with inflexion points arising between 0.4 and 0.6 V during charge and discharge process, representing the occurrence of redox reaction, which is greatly consistent with the appearance of redox peaks in CV.

Figure 4

Areal and gravimetric capacitance of pre-intercalated MnO₂ with different deposition time.

(a) Areal capacitance for different deposition time (20 min, 100 min, 200 min) at different current densities ranged from 0.17 to 30 mA·cm⁻², current densities from left to right: 0.17, 0.25, 0.5, 1, 3, 5, 7, 10, 13, 15, 20, 25, 30 mA·cm⁻²; inset a, galvanostatic discharge profiles of each electrode at 0.17 mA·cm⁻²; inset b, galvanostatic charge and discharge profiles of electrode after electrodeposition of 200 min at current densities from 1 to 20 mA·cm⁻². (b) Gravimetric capacitance normalized by given loading mass of different deposition time; inset: capacities normalized to mAh·g⁻¹ at current densities 0.17, 0.25 and 0.5 mA·cm⁻².

According to the galvanostatic charge–discharge curves of electrodes in 1 M Na₂SO₄ at room temperature, average capacitance during the whole discharge process has been calculated from Equation (2):

Where i (A) is the current density applied in charge/discharge, t (s) is the time elapsed in the discharge cycle, S (cm²) is the geometric area of the active electrode and V (V) is the voltage interval of the discharge.

The areal capacitances are estimated from discharge curves at different current densities from 0.17 to 30 mA·cm⁻² (Figure 4a). The capacitance contributed by bare nickel foam can be negligible (Supplementary Figure S4). All the capacitance curves show that the areal capacitances increase as the areal current densities decrease. At the lowest current densities 0.17 mA·cm⁻², the obtained areal capacitance for 200, 100, 20 min reach up to maximum 2.08, 1.46, 0.25 F·cm⁻², respectively, which are comparable to as-reported high capacitance^17,37,38. Even at the largest current density 30 mA·cm⁻² which produces the lowest areal capacitance, the areal capacitance of 10–30 mF·cm⁻² in our work is still much higher than that of reported MnO₂ film^14,17,18,20. When compressed at 10 MPa, the nickel foam-supported Na_xMnO₂/MnO₂ electrode is paper-like with a thickness of merely 50 μm (Supplementary Figure S5 inset), much thinner than other supporters such as textiles¹⁷, sponge¹⁰, ITO glasses¹⁸, PET²⁰, bringing a higher volumetric capacitances over 100 F·cm⁻³ (Supplementary Figure S5) based on volume of the total electrode.

Figure 4b shows that, based on loading mass of 2.25, 1.17, 0.23 mg·cm⁻² for 200, 100, 20 min, the obtained high gravimetric capacitance of 903, 1237, 1058 F·g⁻¹ at 0.17 mA·cm⁻² can be correspondingly normalized to 205, 280, 240 mAh·g⁻¹ (Figure 4b inset). The electrode with 100 min deposition exhibits the larger gravimetric capacitance than that with 200 min deposition. According to current understanding, with the increase of active mass, the gravimetric capacitance decreases, due to large internal resistance (Supplementary information, Table 3) and difficult ionic diffusion.

Considering that the highest areal capacitance is obtained when the deposition time is 200 min, we further investigated its cycling performance, as shown in Figure 5. Typically such electrode exhibits both high areal capacitance of 350 mF·cm⁻² at 5 mA·cm⁻² (2.2 A·g⁻¹) and ultra-high reversibility with excellent capacitance retention of 99.9% after 1,000 cycles and 95.2% after 5000 cycles (Supplementary Figure S6), which represents long life in practical applications because of good structural and electrochemical stability of Na_xMnO₂/MnO₂ electrodes.

Figure 5

Cycling performance of Na_xMnO₂/MnO₂ electrode at current density of 5 mA·cm⁻² (2.2 A·g⁻¹).

Inset: (a) the capacity retention during 1000 cycles and (b) discharge profiles with respect to initial cycle (black line), 500^th cycle (red line), 1000^th cycle (blue line).

We calculated the half cell (3-electrode) energy density and power density of as-electrodeposited electrodes according to the Equations (3) and (4) as follows:

C (F·cm⁻²) represents the areal capacitance; V (V) is the voltage window; R (Ω·cm⁻²) is the areal internal resistance (Supplementary information, Table 3). According to the Ragone Plot in Figure 6, ultrahigh power densities (based on active material) are obtained from 64.6 to 836.9 KW·kg⁻¹ at different current densities, which are at least tenfold higher than those of conventional MnO₂ supercapacitors, resulting from a perfect ionic and electronic mobility during the electrochemical process. Besides, high energy density up to 110 Wh·kg⁻¹ (half cell, based on active material; or 55 Wh·kg⁻¹ for full cell), is also comparable to the reported values²⁰.

Figure 6

The Ragone Plot of electrodeposited electrode.

Energy density and power density with different deposition time (20, 100, 200 min) at different currents from 0.17 to 20 mA·cm⁻². The area for reported MnO₂ supercapacitors and commercial supercapacitors are plotted and modified by reference to Lang's supporting information²⁰.

Discussion

As reported, the layered intercalation compounds Na_xMnO₂ were conventionally prepared by high-temperature solid reaction^23,24,25,34. Here, the sodium pre-intercalation Na_xMnO₂ (x = 0.7 or 0.91) was firstly synthesized by electrodeposition method, a kind of wet electrochemical methods. The probable formation mechanism of Na_xMnO₂ is shown in Figure 1b. During the electrodeposition process, the MnO₂, sharing [MnO₆] octahedra edge with each other, is firstly deposited on the Ni skeleton with a series of successive electrochemical redox reaction, then the nanoflakes formed. In particular, under the large concentration of sodium ions (Na⁺:Mn²⁺ = 100:1), the as-synthesized layered MnO₂ simultaneously enables sodium ions to intercalate into its interlayers, according to the electrostatic balance and energy minimization theory. The final sodium distributive layered Na_xMnO₂ (x = 0.7) is obtained. With longer deposition time, more sodium ions are allowed to enter into the interlayers of MnO₂. Consequently, higher sodium content is obtained in the final electrodeposited product of Na_xMnO₂ (x = 0.91), confirmed by XRD results. The finding of hybrid MnO₂ phase in the final product is consistent with previous reports of electrodeposition^{10,17,19,20,21,39,40,41}, which indicates simultaneous formation of MnO₂ with spontaneous intercalation of Na⁺ ions in MnO₂ interlayers. The XRD results further confirm that the intercalation of Na⁺ ions into MnO₂ interlayers is time dependent with phase transition from Na_0.7MnO₂ to Na_0.91MnO₂.

A pre-intercalation effect is proposed to explain the excellent electrochemical performance of Na_xMnO₂/MnO₂ in this work. For traditional hydrothermally synthesized α-MnO₂ in Na₂SO₄ electrolyte, it exhibits a pseudo-constant rate over the entire CV tests, due to the difficult and long distance Na⁺ diffusion in the MnO₂ interlayers, responding to rectangular shape without redox peaks in the CV curve. However, the observed redox peaks in Na_xMnO₂ electrode are attributed to a pre-intercalation effect. On one hand, this kind of intercalation of Na⁺ in MnO₂ nanoflackes can be regarded as a pre-intercalation process of electrolyte cations during the electrodeposition process. On the other hand, such pre-intercalation facilitates the subsequent electrochemical ion diffusion, intercalation/deintercalation and transport process, leading to the redox process of Mn between Mn (IV) and Mn (III). The as-observed redox peaks in CV curves at 0.4–0.6 V (vs. SCE) are exactly identical with the theoretical transformation potential between Mn (IV) and Mn (III), according to the Mn Pourbaix diagram (Supplementary Figure S7). Moreover, with the increase of the scan rate up to 500 mV·s⁻¹, the kinetics of Faradic redox reactions and the high rate of electronic and ionic transport are able to successfully occur, which indicates the Na_xMnO₂/MnO₂ ultrathin nanoflakes on Ni electrode allow easy and efficient access of electrons and ions. The sodium ion diffusion coefficient (D_Na⁺) of Na_xMnO₂/MnO₂ was calculated to testify the facilitated Na⁺ diffusion by pre-intercalation, according to the Randles-Sevick equation^42,43. The D_Na⁺ here is up to 3.07 × 10⁻⁶ cm²·S⁻¹ (Supplementary information, Calculation 1), which is three orders of magnitude larger than that of currently reported compounds²³, revealing the fast diffusion of Na⁺ is accelerated by pre-intercalation. Besides, the ultrahigh power density and battery level energy density (Figure 6) also reveal a much better ionic and electronic mobility during the electrochemical process than that of traditional MnO₂. Such results can prove that the as-obtained Na_xMnO₂ phases exhibit high electrochemical activity of Na⁺ to enhance the redox process of Mn (IV)/Mn (III) in Na₂SO₄ electrolyte.

To further identify the pre-intercalation effect, we have compared the electrochemical performance of M_xMnO₂ (M:Na⁺ or Li⁺) in different electrolytic systems (Na₂SO₄ or Li₂SO₄) (Supplementary discussion). As a result, in Na₂SO₄ electrolyte, better electrochemical performance is obtained when Na⁺ instead of Li⁺ intercalated in MnO₂ interlayers (Supplementary Figure S3). Such result reveals that the pre-intercalation effect of electrolyte cations in active mass is greatly related to the subsequent electrochemical performance.

In summary, sodium ions have been spontaneously intercalated into MnO₂ interlayers with two layered Na_0.91MnO₂ and Na_0.7MnO₂ phases obtained by electrodeposition. Such Na_xMnO₂/MnO₂ electrode exhibits excellent supercapacitor performance especially in Na⁺ electrolytic system, delivering both ultrahigh power density and energy density up to 110 Wh·kg⁻¹ (half cell based on active materials). These electrodes also maintain outstanding electrochemical stability with capacitance retention up to 99.9% after 1000 cycles. According to the calculation and electrochemical comparison, the Na⁺ diffusion of Na_xMnO₂/MnO₂ electrode is confirmed to be greatly facilitated. The pre-intercalation effect was proposed to explain the excellent electrochemical behaviors of Na_xMnO₂/MnO₂. This electrochemical pre-intercalation proves to be an effective route to enhance the diffusion of electrolyte cations for layered oxides and the as-obtained Na_xMnO₂/MnO₂ (x = 0.7 or 0.91) nanoflakes are promising materials for supercapacitors to achieve the hybridization of both high energy and power densities for large-scale industrial applications.

Methods

Modified electrodeposition for Na_xMnO₂

Currently, solutions for electrodeposited MnO₂ were reported to be consisted of Mn (II) compounds such as MnSO₄ or Mn(CH₃COO)₂ as manganese sources^{17,19,20,21,39,40,41} and 0.1–0.2 M Na₂SO₄ as additives^17,20,41. In our work, 0.02 M Mn(CH₃COO)₂ and 1 M Na₂SO₄ were used as mother solution, which provided a much larger concentration ratio (Na⁺:Mn²⁺ = 100:1), so as to conduct the intercalation of Na⁺ ions in MnO₂ layers by electrodeposition.

Sodium intercalation was carried out by anodically galvanostatic electrodeposition, using a three-electrode configuration with nickel foam (1 × 1 cm) as working electrode, Pt as counter electrode and Ag/AgCl (saturated KCl) as reference electrode. Here nickel foam was used to provide support for electrodeposited Na_xMnO₂ and good electron conductivity for the total electrode. The deposition was carried out by Autolab Potentiostat Galvanostat. As-electrodeposited Na_xMnO₂/MnO₂ nanoflakes were obtained at current density of 350–500 μA·cm⁻² for 20, 100, 200 min with loading mass of 0.23, 1.17, 2.25 mg·cm⁻², respectively Large nickel foams with area of 5–10 cm² were used to identify the average loading mass by microbalance before and after the electrodeposition in a fully dried condition. As-electrodeposited electrode was naturally dried under 25°C to prevent the detachment of active materials from nickel foam. Loading mass with respect to each electrode was identified by over three times. Li_xMn₂O₄ was also prepared for comparison in electrochemistry with Na⁺ pre-intercalated electrode. Method and process were the same for pre-intercalation of Li⁺ ions into MnO₂ layers, with only Na₂SO₄ (1 M) solution replaced by Li₂SO₄ (1 M) solution. Pure α-MnO₂ was synthesized by hydrothermal method with 2 mmol KMnO₄ and 3 mmol MnSO₄ in 80 ml deionized water at 160°C for 8 h.

Characterization

The crystal phase and purity of the product were characterized by X-ray powder diffraction. Before XRD analysis, electrodeposited powder were firstly striped from nickel foam, checked by magnet to prevent the impurity of nickel and then ultrasonicated for 60 min. X-ray diffraction (XRD) data of samples were collected with a D8 Bruker Advance X-ray diffractometer, using Cu Kα radiation (γ = 1.5418 Å) in a 2θ range from 10° to 80° at room temperature. Atomic Adsorption Spectrum was charaterized by GBC (Australia). Scanning electron microscope (SEM) performed on JEOL-7100F was used to characterize the morphology of the synthesized nanomaterials. Transmission electron microscopic (TEM), high-resolution TEM images were recorded by using a JEOL JEM-2010 FEF microscope at an accelerating voltage of 200 kV.

Electrochemical measurement

A three electrodes method consisting of nickel foam as working electrode (typically 1 cm²), Pt wire and Ag/AgCl (saturated KCl) electrodes as counter and reference electrodes were used. Cyclic voltammetry and galvanostatic charge–discharge studies were performed using an Autolab Potentiostat Galvanostat. Na₂SO₄ (1 M) solution was used as the electrolyte for all the electrochemical test. To compare the electrochemical behaviors of Na_xMnO₂/MnO₂ with those of Li_xMn₂O₄/MnO₂ in Na⁺ or Li⁺ electrolytic system and further illustrate our advantages of pre- intercalation, Li₂SO₄ (1 M) solution was also used. All electrodeposited electrodes with free-standing Na_xMnO₂/MnO₂ nanoflakes or Li_xMn₂O₄/MnO₂ were well rinsed by deionized water and then set as working electrode. For the hydrothermally synthesized α-MnO₂, the working electrode was consisted of 60 wt% of the active material, 35.5 wt% of conductivity agent (carbon black, Super-P-Li) and 4.5 wt% of binder (polytetrafluoroethylene).