Diglyme molecular solvated sodium ion complexes enable the superfast co-intercalation/de-intercalation into graphite interlayers, providing unprecedented prospects for the application of low-dimensional graphitic carbon as fast-charge sodium ion battery anode materials. A thorough understanding of this novel co-intercalation process and resulting solid-electrolyte interphase (SEI) is essential for improving the electrochemical performance of co-intercalation-based high-capacity energy storage systems. This work presents the real-space operando observation of SEI formation and Na-diglyme co-intercalation in the few-layer graphene (FLG) anode as a relevant model of a graphitic anode. The micrometer-sized FLG grid on a nickel current collector was fabricated as a model sample, allowing direct comparative studies using complementary techniques. A reversible sodium-diglyme co-intercalation into the graphene grid was confirmed by Raman spectroscopy, the nanomechanical properties of electrolyte decomposition products on graphene anode and Ni current collector surfaces were studied by ultrasonic force microscopy, and the chemical components of the SEI were confirmed by x-ray photoelectron spectroscopy mapping. We observed a mechanically soft SEI layer formed on the carbon anode surface compared with the electrode current collector surface within the low voltage region (<0.3 V vs Na+/Na), this SEI layer does not affect the reversible Na-diglyme co-intercalations into FLG. At the same time, the SEI layer formed on the Ni current collector mainly contains stiff and thin inorganic species and is electrochemically stable at low voltage regions. Our results clarify the SEI formation behavior on the FLG anode surface in the diglyme electrolyte, providing experimental evidence for the fundamental understanding of Na-diglyme co-intercalation.

Reversible rapid intercalation of charge carrier ions into cathode/anode materials is the key to achieving high-rate performance and long-cycle stability in “rocking-chair” batteries. In sodium ion batteries (SIBs), the conventional intercalation process (naked alkali metal-ions intercalation into graphitic carbon interlayers) was perceived as being thermodynamically unfavorable1 for Na-ions in carbonate-based electrolytes (only a tiny amount of Na-ions can be reversibly intercalated into graphite ≈NaC36 (Ref. 2) with a poor storage capacity of ∼30 mA h g−1). This poor Na+ storage ability in graphitic carbon anode was attributed to a weak cation–π interaction.3 Interestingly, it has been reported that by replacing the conventional carbonate electrolyte with an ether solvent, stable and reversible co-intercalation of the ether-solvated Na-ions can be achieved, leading to an increase in the graphite sodium storage capacity to more than 110 mAh/g.4–6 This novel solvent co-intercalation energy storage strategy has shown excellent kinetics and reversibility.6–8 Although great advances have been achieved in the Na+ solvent co-intercalation energy storage system,4 there is a key question that still remains elusive: How can such large-sized [Na-solvent]+ complexes swiftly shuttle through the electrode–electrolyte interface and rapidly diffuse inside the graphite interlayers?

Several assumptions have been proposed. For example, the negligible formation or absence of solid-electrolyte interphase (SEI) layers at the graphite surface promotes the co-intercalation,9 the minimized interaction between the solvated Na-ions and graphite layers also contributes to the rapid diffusion of the intercalant in the graphite host lattice.10 However, these assumptions are still under debate, and direct experimental observations of the co-intercalation and SEI formation in real-space are mostly missing. A thorough understanding of the formations and properties of SEI in ether electrolyte systems remains elusive compared to the SEI formed on graphite anode surfaces in carbonate-based electrolytes.11 In contrast to the reports of negligible SEI or SEI-free surfaces, the first comprehensive study of surface layer dynamics on t-GICs (ternary graphite intercalation compounds), using soft x-ray photoelectron spectroscopy,12 confirmed that the SEI on the electrodes after several cycles is about 3–8 nm thick and mainly composed of salt (NaFSI) decomposition products and hydrocarbons.13 The authors suggested this SEI does not block the co-intercalation completely but impairs its efficiency. In another salt system (NaOTf + diglyme),9 the electrolyte decomposition was suggested to be a non-recurring reaction with graphite surface groups, and the decomposition products are highly soluble or volatile. As a result, no solid decompositions can be observed by TEM in the graphite anodes after two charge/discharge cycles. Most recently, Liu et al. proposed that only a small proportion of the initial capacity loss of graphite anodes is due to the electrolyte reduction decomposition of PF6- in Na solvation sheath, while the main contributor to the initial capacity loss is identified as the PVDF binder.14 Overall, while the understanding of the interplay between the sodium co-intercalation process and the marginal (or absent) SEI formation is required for in-depth studies, the direct observation of SEI remains a major challenge due to the nanoscale thickness and the air sensitivity of SEI.

It is clear that operando scanning probe microscope (SPM) can accomplish the real-space observation of SEI formation15 as well as its nanomechanical property evolutions on the battery electrode surface under liquid electrolyte environments and was demonstrated to play a very important role in the discovery of SEI layer in LIBs.16 The previously reported operando SPM observation of SEI formation in the ether-based electrolyte on graphite surface was carried out in G3 solvent by Seidl et al. using an electrochemical scanning tunneling microscope (EC-STM).12 Although they concluded that there is no obvious solid decomposition based on these operando EC-STM/Atomic Force Microscope (AFM) images, further interpretations of these STM results were limited by the image quality and insufficient morphology information. Moreover, serious delamination and electrode expansion were found in the typical HOPG model anode by Zhang et al. using electrochemical atomic force microscopy (EC-AFM).17 This delamination and large volume expansion of HOPG caused by the sodium-diglyme co-intercalation are detrimental to the interpretations of surface structural evolutions using EC-AFM.

To address these shortcomings, we fabricated a few-layer graphene (FLG) grid electrode as a close-to-reality model and used the EC-AFM with nanomechanical probing via ultrasonic force microscopy (UFM) approach to observe the surface dynamics and evaluate the nanomorphology and nanomechanical properties simultaneously during the sodium-diglyme complex co-intercalation. Based on the systematic characterization and comparison of nanomechanical and chemical species of SEI layers formed on carbon and metallic current collector surfaces, the voltage-related electrolyte decomposition behavior was proposed to describe the interface chemistry and co-intercalation dynamics of the graphitic carbon as sodium ion battery anode material.

We used Ar/O2 plasma etching to create the micro-size patterns in the FLG grown on a nickel (Ni) current collector as the model electrode. The plasma etching rate was enhanced by putting a magnet underneath the Ni TEM grid (see supplementary material, Fig. S1). TEM grids with two different mesh numbers, 2000 and 200, were used as the masks for patterning the grid structure for AFM and x-ray photoelectron spectroscopy (XPS) measurements, respectively. The as-prepared patterns were imaged by scanning electron microscopy (SEM) (JSM-7800F, JEOL, Japan) with an energy-dispersive spectroscopy (EDS) imaging module. The sample cross section was prepared by beam-exit cross section polishing.18,19 The graphitic characteristic of the FLG model electrode was confirmed by a confocal Raman microscope (LabRAM HR Evolution, Horiba) with an excitation wavelength of 532 nm. The chemical components of SEI were studied by XPS using Kratos Analytical AXIS Supra spectrometer with a monochromatic Al Kα 1486.7 eV x-ray source, operating at 15 kV and 15 mA, equipped with an electron gun for charge neutralization. All the spectra were referenced to C1s at 285.0 eV and analyzed using CASAXPS (Casa Software Ltd, UK). Reference 600+ potentiostat (Gamry, USA) was used for cyclic voltammetry (CV) measurements. CV data were obtained at a scanning rate of 0.5 mV s−1 between open circuit potential (OCP) to 0.01 V vs Na+/Na, using patterned graphene on Ni electrode as a working electrode, sodium metal as reference/counter electrode, and 1M NaOTf in diglyme as the electrolyte. Our electrolyte does not contain any additives that could influence SEI formation. The SEI formation on the model electrode surface during the first CV cycle was studied by electrochemical atomic force microscopy with ultrasonic excitation mode (EC-UFM).20,21 The detailed description of the EC-UFM setup and physical principles can be found in supplementary material Note 1. Operando electrochemical EC-AFM/UFM was performed on the Bruker Multimode (Bruker, USA) inside the glovebox (MBRAUN, Germany) with oxygen and moisture content below 0.5 ppm. The AFM tips used were Bruker ScanAsyst-fluid+ with a force constant of about 0.4 N/m. After plasma etching, the patterned sample is horizontally mounted on the custom-made AFM electrochemical cell as shown in Fig. 3(a). Before electrochemical measurements, open circuit potential (OCP) was recorded for 300 s after the injection of electrolyte, then 200 nN force was applied to the AFM tip to clean the electrode surface by scanning in a 30 × 15 μm2 area before commencing operando EC-AFM measurements.

We used the magnetic-field-enhanced plasma etching method [Fig. 1(a)] to prepare the grid structure on the FLG grown on top of nickel (Ni). The Ni is used as the current collector for the charge/discharge of FLG anodes. The pattern and periodic size of the grid electrode are selected by changing the mesh number of TEM grid masks. After the argon plasma etch step, the initial graphene at the square hole areas was removed by the plasma, leaving the underneath Ni layer exposed as shown in Figs. 1(b) and 1(c). The average etch depth is controlled to around 26 nm. This guarantees the graphene layer to be completely removed, leaving the Ni current collector layer exposed as confirmed by the SEM image in Fig. 1(c). The exposed Ni current collector shows higher electronic conductivity as confirmed by the conductivity mapping by AFM (supplementary material, Fig. S2). Figure 1(d) shows the deflection channel of the cross section AFM image of the model sample, in which the different layers are observed (see also cross section SEM and AFM height/mechanical images in supplementary material, Figs. S3 and S4). The beam-exit cross-sectional polishing (BEXP) due to the oblique polishing angle magnifies the thin layer structure on the section by five times,19 enabling precise determination of the thickness of multilayer structures. In the oblique section image, the thickness of the few-layer section is between 4 and 11 nm, which corresponds to a real thickness of around 0.8–2.2 nm, estimating the FLG to be consisting of 2–7 graphene layers. This agrees well with the Raman spectrum in Fig. 1(e), in which a sharp strong 2D-band peak is observed. The G/2D ratio is about 0.6 in the model sample, indicating the graphene has a layer number between single and quintuple layers.22  Figure 1(f) shows the surface effective modulus measured by UFM, and the surface topography of the scan area is shown in Fig. 1(g). The different contrast in the effective modulus image indicates the different nanomechanical properties of the current collector (Ni hole) and graphene anode surface as sketched in the inset of Fig. 1(g). This capability of differentiating nanomechanical properties on the sample surface can be found in supplementary material Note 1, and this UFM approach will be used to study the various SEI component formations during the electrochemical cycles as discussed below.

FIG. 1.

(a) Sketch of the model sample (FLG grid electrode) prepared by magnetic-field-enhanced plasma etching method. (b) AFM topography and (c) SEM image of the as-prepared model sample surface, the inset is the nickel elemental energy-dispersive x-ray spectroscopy (EDX) mapping. (d) Beam-exit cross section polishing (BEXP)19 oblique low-angle cross section AFM image of the sample. (e) The Raman spectra of FLG. (f) The effective modulus image measured by UFM and the (g) the AFM topography image at the same scanning area. The inset shows the cross section schematic diagram of a Ni hole in the FLG model electrode.

FIG. 1.

(a) Sketch of the model sample (FLG grid electrode) prepared by magnetic-field-enhanced plasma etching method. (b) AFM topography and (c) SEM image of the as-prepared model sample surface, the inset is the nickel elemental energy-dispersive x-ray spectroscopy (EDX) mapping. (d) Beam-exit cross section polishing (BEXP)19 oblique low-angle cross section AFM image of the sample. (e) The Raman spectra of FLG. (f) The effective modulus image measured by UFM and the (g) the AFM topography image at the same scanning area. The inset shows the cross section schematic diagram of a Ni hole in the FLG model electrode.

Close modal

The [Na-diglyme]+ co-intercalation into the FLG electrode was elucidated by employing the nondestructive operando electrochemical Raman spectroscopy [Fig. 2(a)]. Figures 2(d) and 2(e) show the operando Raman spectra of FLG during the first CV cycle (the full spectra, from 1000 to 3000 cm−1, can be found in supplementary material, Fig. S5). As shown in the schematic diagram in Fig. 2(b), the D band derives from the disordered structures induced by sp3 hybridization,23 such as defect structures in crystals, as well as the A1g breathing vibration mode of sp2 carbon atoms.23 Therefore, the insertion of sodium ions into the defect sites will suppress the vibration mode of the D band. The G band corresponds to the E2g in-plane bond-stretching vibration of sp2 carbon atom pairs [Fig. 2(c)]. This stretching vibration is highly sensitive to ion doping, and the co-intercalation of ions/molecules in graphene causes a significant change in the G band.24 The Raman spectra for graphite intercalation compounds (GICs) with stage n > 2 are known to exhibit a doublet G band, containing a Guc peak at around 1580 cm−1 and a Gc peak at around 1601 cm−1.25 The Guc peak corresponds to the in-plane vibration within the graphitic carbon layers, while the Gc peak can be attributed to the vibration of carbon layers with intercalated solvated sodium ions. Furthermore, the 2D band is the harmonic of the D band and does not require defects for activation.24 

FIG. 2.

(a) Schematic diagram of the operando electrochemical Raman cell for the FLG during the first cyclic voltammetry (CV) cycle (0.1 mV/s). Schematic diagram of (b) A1g and (c) E2g Raman band in graphitic carbon. (d) G band and (e) 2D band of the measured operando Raman spectra.

FIG. 2.

(a) Schematic diagram of the operando electrochemical Raman cell for the FLG during the first cyclic voltammetry (CV) cycle (0.1 mV/s). Schematic diagram of (b) A1g and (c) E2g Raman band in graphitic carbon. (d) G band and (e) 2D band of the measured operando Raman spectra.

Close modal

From the full Raman spectra in Fig. S5, one can observe that during the cathodic scan from 2.5 to 0.9 V, the intensity of the D band decreases and undergoes a red shift. This change indicates that the [Na-diglyme]+ complex is adsorbing and partially intercalating into the defects of graphene during the cathodic scan in this voltage range. Further decreasing the electrode voltage from 0.5 to 0.01 V, the intensity of the D band increases, accompanied by a blue shift, while the intensity of the Gc peak disappears suddenly with the voltage dropping below 0.5 V. This suggests that within this voltage window, [Na-diglyme]+ complexes leave the defect sites, and a large number of [Na-diglyme]+ moieties exist between the graphene layers, evidenced by the significantly enhanced G band (supplementary material, Fig. S6).

The detailed evolution of the G peak is shown in Fig. 2(d). Under an excitation laser with a wavelength of 523 nm (∼1.58 eV), we observed the emergence of a Gc peak at the beginning of intercalation (at ∼1.7 V vs Na+/Na), indicating the onset of the staging process. As the intercalation reaction proceeds, the intensity of the Gc peak increases dramatically while the Guc peak red shifts and disappears [see typical Raman spectrum in Fig. 2(c)], coinciding with the time when the discharge voltage reaches around 0.9 V vs Na+/Na. The disappearance of the Guc layer indicates the absence of non-intercalated carbon layers, which occurs when the reaction reaches a stage-1 compound [one layer of graphene in between each two intercalant layers, as shown in Fig. 2(d)]. At this point (< 0.3 V), a rapid enhancement of the Gc peak intensity is observed, reaching ten times the initial Guc peak intensity. This dramatic change in the Gc peak intensity can be attributed to the Pauli blocking of destructive interference Raman pathways26 that takes place when the Fermi level approaches half of the excitation laser energy. The narrowing of the Gc peak can be simply attributed to the increasing structural order of [Na-diglyme]+ within the carbon interlayers, which has also been ascribed to increased phonon lifetime in charged graphene.7 

The intensity of the 2D band remains unchanged from 2.5 to 0.9 V [Fig. 2(e)], but after around 0.9 V, it undergoes a red shift and a decrease in intensity. This indicates that after 0.9 V, the carrier density in graphene increases significantly due to the massive insertion of sodium ions, which increases the inner strain and electronic doping of the graphene confirming the conclusions from the G band dynamics. Hence, we believe that the intercalation of solvated Na-ions in FLG follows the Daumas–Herold model, wherein the interlayer malleability of graphene layers accommodates the presence of the intercalating agent.25 At the end of the cathodic scan, the 2D band became nearly invisible till stage-2 GIC started to form stage-1 GIC (when all the graphene layers were charged).25 

During the first anodic scan (de-intercalation of sodium ions), the D band exhibits a blue shift with a reduced intensity, the Guc peak reemerges, and the 2D band experiences a blue shift, all signifying the departure of sodium ions from the FLG anode. Moreover, we also found that the reversibility of co-intercalated species was not affected after the first CV cycle. As shown in Fig. S7, although the D peak becomes weak in the second and third CV cycles, the shifting of the G peak, as well as the 2D peak, is still reversible and consistent with the trends observed during the first cycle, which is consistent with the previously reported reversibility of Na-diglyme intercalation into FLG.27 In summary, operando Raman spectra confirm that the intercalation of [Na-diglyme]+ into the FLG electrode appears to be highly reversible.

Operando EC-AFM and UFM were introduced to further observe the ion insertion and SEI formation in the FLG electrode. As shown in Fig. 3(a), in our operando AFM cell, a piezo-transducer was mounted underneath the electrochemical cell to generate the ultrasonic vibration with a sub-nanometer vibration amplitude at about 4 MHz frequency with low-frequency (1.2 kHz) amplitude modulation. As the vibration frequency is significantly higher than the frequencies the AFM cantilever can respond to, the ultrasonic vibration results in the effective indentation on the sub-nanometer scale of the tip into the sample. Due to the nonlinearity of tip–surface force interaction, that is detected as an average additional “ultrasonic” force at the kHz modulation frequency that carries the nanomechanical properties information.21 The low-frequency AFM signal was extracted by the lock-in amplifier and used to construct the nanomechanical properties on the sample surface (see supplementary material, Note 1). In UFM, the vertical modulation of tip–sample surface interaction, with a modulation amplitude of a few angstroms to nanometers, can reduce the average force of the AFM tip applied on the fragile SEI layer. This is similar to the conventional tapping mode and peak-force tapping mode,28 but, with a simplified signal feedback system that can effectively prevent the scratching off of the SEI layer from the electrode surface during the operando characterizations.

FIG. 3.

(a) Schematic diagram of EC-AFM cell with ultrasonic extrication transducer. (b) Cyclic voltammetry curves of graphene model electrode in the EC-cell. (c)–(e) 3D topography images and height distribution functions of electrode surface at voltage ranges of 1.59–1.33, 0.64–0.44, and 0.39–0.13 V during the CV scan measured by operando EC-AFM (the white arrow indicates AFM image capturing directions).

FIG. 3.

(a) Schematic diagram of EC-AFM cell with ultrasonic extrication transducer. (b) Cyclic voltammetry curves of graphene model electrode in the EC-cell. (c)–(e) 3D topography images and height distribution functions of electrode surface at voltage ranges of 1.59–1.33, 0.64–0.44, and 0.39–0.13 V during the CV scan measured by operando EC-AFM (the white arrow indicates AFM image capturing directions).

Close modal

The first cathodic CV scan of the graphene grid electrode during the EC-AFM measurement can be found in Fig. 3(b). As shown in Fig. 3(b), a large cathodic peak is found at around 0.5–1.0 V vs Na+/Na during the CV cycles, corresponding to the massive Na-diglyme co-intercalation at this voltage range in full agreement with the Raman spectra in Fig. 2. Apart from this peak, some reductive current peaks in the voltage range of 1.0–2.5 V vs Na+/Na were also observed, which may be attributed to the sodium adsorption and initial electrolyte decompositions in the FLG at the defect sites. Therefore, the operando EC-AFM/UFM was performed during the first cathodic scan at the voltage range around 2.0–0.01 V vs Na+/Na to understand the initial SEI formation and ion intercalation. During this CV scan, the EC-AFM recorded the 3D surface topographic [Figs. 3(c)–3(e) and Fig. S8] and nanomechanical evolution [Figs. 4(c) and 4(h)] simultaneously. During the cathodic scan from 1.81 to 0.69 V, no significant changes can be found on the electrode surface. However, the height distribution function in Fig. S8 reveals that the height difference between the carbon material “mesh” and metallic current collector square “holes” increased from 26 to 29 nm in this voltage range. This is attributed to the carbon layer expansion around the voltage region of 0.94–0.69 V vs Na/Na+ [Fig. 3(f)], which is consistent with the CV curve in Fig. 3(b) showing that the large cathodic peak starts to appear around this voltage range. Interestingly, as shown in Fig. 3(d), at the voltage range 0.64–0.44 V, the initial average graphene thickness increased by ∼4–5 nm (Fig. S8) due to the Na-diglyme co-intercalation. Considering the initial average thickness of the graphene anode of about 1.5 nm, the large expansion ratio of total thickness should correspond to the stage-1 co-intercalation state of graphitic carbon, in which the initial interlayer spacing (∼0.335 nm) expands to about 1.13 nm after each intercalated layer is occupied by the sodium-diglyme complexes.6–8 This is fully consistent with the Raman spectra in Fig. 2 where a strong enhancement of the G peak intensity was observed in this voltage range. As shown in Fig. 3(e), we further observed many larger SEI particles' formation on the electrode surface (see also deflection mappings in Fig. S8 and the cross section profiles in Fig. S9 in the supplementary material), which affects the accurate determination of the graphene layer expansion. During de-intercalation, we observed serious delamination of the binder-free FLG anode due to the lack of strain applied to the electrode (Fig. S10). This happened when the FLG working electrode current changes from reduction to oxidation upon the anodic scan as shown in the EC-AFM measurement during a full CV cycle in supplementary material Fig. S10(a).

FIG. 4.

One-dimensional (1D) ultrasonic force–distance curves (a) and normal force–distance curves (b) measured on the Ni hole, exposed FLG and SEI particles (see the measured points in FIG. S11) on the postmortem FLG electrode surface. (c)–(h) Two-dimensional (2D) topography images and the corresponding nanomechanical images of electrode surface during the first cathodic scan at different voltage regions measured by UFM.

FIG. 4.

One-dimensional (1D) ultrasonic force–distance curves (a) and normal force–distance curves (b) measured on the Ni hole, exposed FLG and SEI particles (see the measured points in FIG. S11) on the postmortem FLG electrode surface. (c)–(h) Two-dimensional (2D) topography images and the corresponding nanomechanical images of electrode surface during the first cathodic scan at different voltage regions measured by UFM.

Close modal

The dynamic SEI formation process can be further studied by combining 1D nanomechanical force spectroscopy and 2D topographical mapping as shown in Fig. 4. The nanomechanical properties of the FLG and Ni current collector were probed by 1D ultrasonic force spectroscopy (UFS)21,29 [Figs. 4(a) and 4(b)] in the postmortem FLG electrode. UFS detects the ultrasonic deflection as a function of the normal force applied by the AFM tip, allowing us to compare the local effective stiffness of the sample surface (see supplementary material Note 1). During the tip approach to the sample surface, we recorded the UFS response [Fig. 4(a)] and normal force response [Fig. 4(b)] simultaneously. At zero indentation, the tip suddenly jumps to the sample surface due to the attraction force, leading to the artificial UFS transient increase. As the tip further indents into the sample, a drastic increase in UFS signal to a maximum value was observed in the Ni hole region, indicating the highest effective stiffness of the Ni hole region compared to FLG and SEI particles. This suggests the surface passivation layer formed on the Ni hole region has a high mechanical stiffness. It is also interesting to note that the SEI particles formed on FLG show negligible UFS response during the tip indentation, indicating a softer nature of these SEI particles. This is also confirmed by the notably smaller slope value of SEI particles in the force–distance curves [Fig. 4(b)] compared to the Ni hole and FLG surface.

2D nanomechanical mapping was also recorded during the sodium intercalation [Figs. 4(c)–4(h)]. Initially, at the stationary OCP of around 1.9 V, a relatively high scanning force (∼200 nN) was applied to the AFM tip to clean the non-Faradaic adsorption/precipitation species on the electrode surface. Between 1.8 and 1.6 V, the electrode surface was clean without any adsorption/decomposition contaminations according to the morphology image in Fig. 4(c). At the same scan area, as shown in the UFM nanomechanical mapping in Fig. 4(c′), the nanomechanical image shows a stiffer signal (purple contrast) on the Ni holes compared to the FLG grid (blue contrast), which is consistent with the previous report that the out-of-plane Young's modulus of FLG (∼25 GPa)30 is smaller than that of the Ni metal (∼170 GPa). Interestingly, with the decrease in electrode voltage from about 1.6 to 0.9 V, the cathodic current starts to increase [Fig. 3(b)], while no visible SEI formation was observed on the topography images [Figs. 4(d) and 4(e)]. However, from the nanomechanical images [Figs. 4(d′) and 4(e′)] at these voltage regions, one can find that the stiffness of Ni holes becomes smaller than the stiffness of the FLG grid. This indicates that a “thin” SEI layer was formed on top of the Ni metal, which reduced the mechanical stiffness during these voltage regions. Between 0.64 and 0.44 V, the graphene grid electrode surface was fully covered by the SEI particles [Fig. 4(g)], and these SEI particles had similar mechanical strength as the SEI formed on the Ni “holes” as supported by the nanomechanical image in Fig. 4(g′). At the lowest voltage region (0.39–0.13 V), dramatic electrolyte decomposition results in the large SEI particles formed on the graphene grid electrode surface as shown in Fig. 4(h). Importantly, from Fig. 4(h′), one can find that the relative stiffness of these large SEI particles is much smaller than the ones formed at high voltage regions, indicating that the later formed SEI at lower voltage regions may be composed of a larger fraction of soft organic SEI, following different decomposition paths compared to the SEI formed in the higher voltage regions.

Nanomechanical property measurements indicate the SEI formed on the Ni and FLG surfaces are distinctly different. We therefore used XPS mapping to further reveal the chemical components of the SEI formed on the electrode surface. We prepared the model electrode with a larger grid size (mesh number 200, with a period of the grid ∼127 μm) to meet the resolution of XPS elemental mapping. The sample underwent a cathodic scan to 0.01 V vs Na/Na+ reference electrode and then disassembled for the XPS measurement via inert gas transfer. The XPS mapping image contrast derives from the intensity of the Na and F elements at the binding energy positions corresponding to the NaF (F1s at ∼686 eV and Na1s at ∼1072 eV). As shown in Figs. 5(a) and 5(b), the higher sodium and fluorine signals are observed on the Ni square areas compared with the carbon grid area, indicating the abundant NaF, a widely reported inorganic SEI component,31 formed on the Ni current collector. This is consistent with nanomechanical property measurements in Fig. 4 showing a stiff SEI layer (compared with the soft organic SEI formed on graphite at low voltage region) was formed on the Ni hole region. The F1s spectrum of the mapping area in Fig. 5(f) contains a larger metallic fluoride peak and a smaller organic-fluoride peak, which further confirms the dominant fluoride species is NaF. We also noticed that the carbon and oxygen mappings do not show a visible contrast between the graphene and the current collector region. This could be because of the effect of the adsorption of the organic DMC solvent that was used to wash the disassembled sample.

FIG. 5.

The XPS elemental mapping and corresponding spectra of (a) sodium, (b) fluorine, (c) carbon, and (d) oxygen on the sample surface after the first cathodic CV scan. (The mesh number of the TEM grid used for preparing the XPS elemental mapping sample is 200, and the scale bar size is 60 μm). XPS spectra of graphene anode surface after the first cathodic scan. (e) Full survey spectrum, (f) C1s, (g) F1s, (h) O1s, and (i) Na1s elemental spectra, respectively.

FIG. 5.

The XPS elemental mapping and corresponding spectra of (a) sodium, (b) fluorine, (c) carbon, and (d) oxygen on the sample surface after the first cathodic CV scan. (The mesh number of the TEM grid used for preparing the XPS elemental mapping sample is 200, and the scale bar size is 60 μm). XPS spectra of graphene anode surface after the first cathodic scan. (e) Full survey spectrum, (f) C1s, (g) F1s, (h) O1s, and (i) Na1s elemental spectra, respectively.

Close modal

From the full spectrum survey [Figs. 5(e)–5(i)] on the FLG electrode surface, one can find that fluorine, sodium, oxygen, and carbon are the four main elements on the electrode surface. The sulfur signal intensity is negligible, indicating that either the SEI contains only a small amount of sulfide or the sulfide SEI species are more soluble in the electrolyte. The high-resolution spectra of the four main elements are shown in Figs. 5(f)–5(i). As shown in the C1s spectrum [Fig. 5(f)], the lowest binding energy peak (C–Si or C–Ni) derives from the silicon carbide or the carbon–nickel alloy in the substrate. In addition, apart from the carbon–hydrogen/oxygen species that can be attributed to the organic SEI species,32 an additional C-Fx bond can be found at ∼293.5 eV.33 Since there is no PVDF binder in our electrode, the existence of carbonate fluoride indicates the decomposition of the salt anions that are the only fluoride source in this system. This anion decomposition derived SEI34 is also confirmed by the organic-F peak (∼689.2 eV) in the F1s spectrum [Fig. 5(g)]. Additionally, a metal-F peak at a lower binding energy position (∼685.8 eV) can be attributed to NaF, which is a typical inorganic SEI phase that has been reported in both ester and ether-based electrolyte,31 regardless of the types of the salt used. Interestingly, from the O1s and Na1s spectra in Figs. 5(h) and 5(i), one can find that almost no NaxOy-related peaks33,35 are detected. This is consistent with the previous reports that the inorganic SEI phase formed in the ester-based electrolyte mainly consists of NaF, rather than sodium oxide or sodium carbonate.36 Moreover, the site-specific XPS measurements were further performed to confirm the lateral heterogeneity of SEI components as shown in Fig. S12. From Fig. S12, one can find that the evident difference between the SEI formed on Ni and FLG is the amount of NaF. Ni surface shows a greater proportionality in the SEI of NaF compared to FLG, while the organic and metallic fluoride SEI seems to equally exist on the FLG surface. All the indexed chemical species are summarized in supplementary material, Table S1.

Based on the operando Raman, EC-AFM/UFM, and ex situ XPS measurements, we can now summarize our understanding of sodium-diglyme co-intercalation and SEI formation on the FLG model sample. As shown in the sectional view of the model electrode in Fig. 6, the initial model electrode contains an etched hole with a depth of around 26 nm. The thin and uniform SEI species formed in the voltage region 1.59–1.33 V. Although this SEI formation does not change the surface topography drastically [Figs. 4(d) and 4(e)], it was observed in our UFM nanomechanical measurements in Figs. 4(d′) and 4(e′). Importantly, as sketched in Fig. 6, we found the soft organic SEI formation at low voltage regions (0.39–0.13 V) does not affect the reversibility of sodium-diglyme co-intercalation. Meanwhile, the SEI component formed on the Ni current collector surface stays constant at lower voltage regions. The XPS mapping in Figs. 5(a)–5(d) and nanomechanical mapping in Fig. 4 also confirm that the SEI formed on the Ni current collector is mainly NaF. The mechanical stability of the SEIs formed on FLG and Ni surfaces was compared in the supplementary material Fig. S12 using nano-indentation. The results also support that the NaF-rich SEI formed on the Ni surface has a high yield strength.

FIG. 6.

Sketch of SEI formation and Na-diglyme co-intercalation in the model electrode during the first cathodic scan. The voltage regions 1.59–1.33 and 0.39–0.13 V correspond to the same regions determined by the operando EC-AFM/UFM in Fig. 3.

FIG. 6.

Sketch of SEI formation and Na-diglyme co-intercalation in the model electrode during the first cathodic scan. The voltage regions 1.59–1.33 and 0.39–0.13 V correspond to the same regions determined by the operando EC-AFM/UFM in Fig. 3.

Close modal

The previous report suggests a marginal formation of SEI layers at the graphite surface can promote the revisability of sodium-diglyme co-intercalation.9 Our operando AFM/UFM measurements unambiguously confirm the SEI formation on the FLG surface is non-negligible, but this SEI does not affect the reversibility of sodium-diglyme co-intercalation and de-co-intercalation according to the operando Raman studies in Fig. 2. This is different from the previous study, which suggested that the ether-based electrolyte provides an SEI-free surface for the co-intercalation of solvated complexes.37 The difference may be due to the better solubility of sodium SEI, which was removed/dissolved during the electrode wash before the postmortem analysis, further emphasizing the importance of operando/in situ measurements that can maintain the integrity of the fragile SEI components during the characterizations. Moreover, via the nanomechanical property measurement, we can confirm that the electrolyte decomposition products at relatively high voltage regions (>0.5 V vs Na+/Na) and low voltage regions (<0.5 V vs Na+/Na) should be the stiff inorganic (NaF) and soft organic SEI components, respectively. However, how the sodium-diglyme complexes shuttle through this SEI layer and what is the underlying mechanism remain open questions. On the methodology aspects, our plasma etched model electrode can serve as a relevant and efficient platform for further studies of these fundamental interfacial electrochemical processes of sodium-diglyme co-intercalation systems. On the battery performance aspects, improving the FLG–electrolyte interface compatibility and Na+ transportation capability are two key points that can be studied in future. For example, it has been reported that the Li-based SEI components, formed by the alloy transform of crystalline SnS nanosheets on graphene surface into amorphous Li2S and Sn with good ion conductivity and interfacial stability, can facilitate Na+ transportations.38 Therefore, we suggested that surface modification with Na ionic conductors may be able to enhance the Na+ transportation and further enhance the cycle stability of the FLG electrode.

In this work, we used plasma etching to prepare the FLG on the Ni model electrode to study the sodium-diglyme complex co-intercalation and SEI formation in the ether-based electrolyte. Operando Raman characterization confirmed that the sodium-diglyme complexes can be reversibly co-intercalated and de-co-intercalated in the FLG model anode. During the co-intercalation, we observed the SEI formed during the first CV cycles, containing the electrochemical reduction of electrolyte on the Ni current collector (inorganic component, NaF), as well as abundant soft organic SEI component on the carbon active material surface. Importantly, this soft organic SEI layer starts to appear during the sodium-diglyme co-intercalation and violently grows at low voltage regions. We, therefore, believe that inhibiting soft, unstable organic SEI formation at low voltage regions, or artificially modifying the carbon electrode surface by forming a more stable SEI species, such as NaF, would further increase the cycle stability of sodium ion batteries using ether-based electrolytes.

See the supplementary material for more details on sample preparations and ultrasonic force microscopy.

The authors wish to acknowledge the financial support by the Faraday Institution (Grant No. FIRG018), EU Graphene Flagship Core 3 project, EPSRC Project No. EP/V00767X/1, the National Natural Science Foundation of China (Nos. 61574037, 11874113, 11344008, and 11204038), the Science Foundation of the Fujian Province (No. 2023J01521), and the Foreign Science and Technology Cooperation Project of Fuzhou Science and Technology Bureau (No. 2021-Y-086). We are also grateful to Bruker UK, Leica Instruments, LMA Ltd, and Gamry Instruments for the in-depth application support of the relevant instrumentation. The authors also acknowledge NEXGENNA consortium for the new methodology development.

The authors have no conflicts to disclose.

Yue Chen and Shaohua Zhang contributed equally to this work.

Yue Chen: Conceptualization (lead); Data curation (equal); Formal analysis (equal); Funding acquisition (lead); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Supervision (equal); Validation (equal); Writing – original draft (lead). Shaohua Zhang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Resources (equal); Visualization (equal); Writing – original draft (equal). Weijian Zhang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Resources (equal); Validation (equal). Alessio Quadrelli: Conceptualization (equal); Data curation (equal); Investigation (equal); Methodology (equal); Resources (equal); Writing – review & editing (equal). Samuel Jarvis: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Jing Chen: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Validation (equal). Hongyi Lu: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Resources (equal). Nagarathinam Mangayarkarasi: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Methodology (equal); Resources (equal); Validation (equal); Visualization (equal); Writing – review & editing (equal). Yubiao Niu: Conceptualization (equal); Data curation (equal); Funding acquisition (equal); Methodology (equal); Resources (equal); Supervision (equal); Validation (equal); Visualization (equal). Jianming Tao: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Resources (equal); Validation (equal); Visualization (equal). Long Zhang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Validation (equal); Visualization (equal); Writing – review & editing (equal). Jiaxin Li: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Resources (equal); Writing – review & editing (equal). Yingbin Lin: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Resources (equal); Visualization (equal); Writing – review & editing (equal). Zhigao Huang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Supervision (equal); Writing – review & editing (equal). Oleg Kolosov: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding authors upon reasonable request.

1.
K.
Nobuhara
,
H.
Nakayama
,
M.
Nose
,
S.
Nakanishi
, and
H.
Iba
, “
First-principles study of alkali metal-graphite intercalation compounds
,”
J. Power Sources
243
,
585
(
2013
).
2.
M. M.
Doeff
,
Y.
Ma
,
S. J.
Visco
, and
L. C.
De Jonghe
, “
Electrochemical insertion of sodium into carbon
,”
J. Electrochem. Soc.
140
(
12
),
L169
(
1993
).
3.
Y.
Liu
,
B. V.
Merinov
, and
W. A.
Goddard
IIII, “
Origin of low sodium capacity in graphite and generally weak substrate binding of Na and Mg among alkali and alkaline earth metals
,”
Proc. Natl. Acad. Sci. U. S. A.
113
(
14
),
3735
(
2016
).
4.
M. L.
Divya
,
Y.-S.
Lee
, and
V.
Aravindan
, “
Solvent Co-intercalation: an emerging mechanism in Li-, Na-, and K-ion capacitors
,”
ACS Energy Lett.
6
(
12
),
4228
(
2021
).
5.
J.
Park
,
Z. L.
Xu
, and
K.
Kang
, “
Solvated ion intercalation in graphite: Sodium and beyond
,”
Front. Chem.
8
,
432
(
2020
);
[PubMed]
B.
Jache
and
P.
Adelhelm
, “
Use of graphite as a highly reversible electrode with superior cycle life for sodium-ion batteries by making use of co-intercalation phenomena
,”
Angew. Chem.
53
(
38
),
10169
(
2014
).
6.
H.
Kim
,
J.
Hong
,
Y.-U.
Park
,
J.
Kim
,
I.
Hwang
, and
K.
Kang
, “
Sodium storage behavior in natural graphite using ether-based electrolyte systems
,”
Adv. Funct. Mater.
25
(
4
),
534
(
2015
).
7.
A. P.
Cohn
,
K.
Share
,
R.
Carter
,
L.
Oakes
, and
C. L.
Pint
, “
Ultrafast solvent-assisted sodium ion intercalation into highly crystalline few-layered graphene
,”
Nano Lett.
16
(
1
),
543
(
2016
).
8.
Z.
Zhu
,
F.
Cheng
,
Z.
Hu
,
Z.
Niu
, and
J.
Chen
, “
Highly stable and ultrafast electrode reaction of graphite for sodium ion batteries
,”
J. Power Sources
293
,
626
(
2015
).
9.
M.
Goktas
,
C.
Bolli
,
E. J.
Berg
,
P.
Novák
,
K.
Pollok
,
F.
Langenhorst
,
M. v.
Roeder
,
O.
Lenchuk
,
D.
Mollenhauer
, and
P.
Adelhelm
, “
Graphite as cointercalation electrode for sodium-ion batteries: Electrode dynamics and the missing solid electrolyte interphase (SEI)
,”
Adv. Energy Mater.
8
(
16
),
1702724
(
2018
).
10.
S. C.
Jung
,
Y.-J.
Kang
, and
Y.-K.
Han
, “
Origin of excellent rate and cycle performance of Na+-solvent cointercalated graphite vs. poor performance of Li+-solvent case
,”
Nano Energy
34
,
456
(
2017
).
11.
V. A.
Agubra
and
J. W.
Fergus
, “
The formation and stability of the solid electrolyte interface on the graphite anode
,”
J. Power Sources
268
,
153
(
2014
);
K.
Xu
, “
Electrolytes and interphases in Li-ion batteries and beyond
,”
Chem. Rev.
114
(
23
),
11503
(
2014
);
[PubMed]
S. J.
An
,
J.
Li
,
C.
Daniel
,
D.
Mohanty
,
S.
Nagpure
, and
D. L.
Wood
, “
The state of understanding of the lithium-ion-battery graphite solid electrolyte interphase (SEI) and its relationship to formation cycling
,”
Carbon
105
,
52
(
2016
).
12.
L.
Seidl
,
N.
Bucher
,
E.
Chu
,
S.
Hartung
,
S.
Martens
,
O.
Schneider
, and
U.
Stimming
, “
Intercalation of solvated Na-ions into graphite
,”
Energy Environ. Sci.
10
(
7
),
1631
(
2017
).
13.
J.
Maibach
,
F.
Jeschull
,
D.
Brandell
,
K.
Edstrom
, and
M.
Valvo
, “
Surface layer evolution on graphite during electrochemical sodium-tetraglyme Co-intercalation
,”
ACS Appl. Mater. Interfaces
9
(
14
),
12373
(
2017
).
14.
M.
Liu
,
L.
Xing
,
K.
Xu
,
H.
Zhou
,
J.
Lan
,
C.
Wang
, and
W.
Li
, “
Deciphering the paradox between the Co-intercalation of sodium-solvent into graphite and its irreversible capacity
,”
Energy Storage Mater.
26
,
32
(
2020
).
15.
Y.
Chen
,
W.
Wu
,
S.
Gonzalez-Munoz
,
L.
Forcieri
,
C.
Wells
,
S. P.
Jarvis
,
F.
Wu
,
R.
Young
,
A.
Dey
,
M.
Isaacs
,
M.
Nagarathinam
,
R. G.
Palgrave
,
N.
Tapia-Ruiz
, and
O. V.
Kolosov
, “
Nanoarchitecture factors of solid electrolyte interphase formation via 3D nano-rheology microscopy and surface force-distance spectroscopy
,”
Nat. Commun.
14
(
1
),
1321
(
2023
).
16.
A. M.
Tripathi
,
W. N.
Su
, and
B. J.
Hwang
, “
In situ analytical techniques for battery interface analysis
,”
Chem. Soc. Rev.
47
(
3
),
736
(
2018
).
17.
X.-R.
Zhang
,
J.-Y.
Yang
,
Z.-Y.
Ren
,
K.-Y.
Xie
,
Q.
Ye
,
F.
Xu
, and
X.-R.
Liu
, “
In-situ observation of electrolyte-dependent interfacial change of the graphite anode in sodium-ion batteries by atomic force microscopy
,”
New Carbon Mater.
37
(
2
),
371
(
2022
).
18.
H.
Pan
,
Y.
Chen
,
W.
Pang
,
H.
Sun
,
J.
Li
,
Y.
Lin
,
O.
Kolosov
, and
Z.
Huang
, “
Complementary sample preparation strategies (PVD/BEXP) combining with multifunctional SPM for the characterizations of battery interfacial properties
,”
MethodsX
8
,
101250
(
2021
).
19.
A. J.
Robson
,
I.
Grishin
,
R. J.
Young
,
A. M.
Sanchez
,
O. V.
Kolosov
, and
M.
Hayne
, “
High-accuracy analysis of nanoscale semiconductor layers using beam-exit Ar-ion polishing and scanning probe microscopy
,”
ACS Appl. Mater. Interfaces
5
(
8
),
3241
(
2013
).
20.
A. P.
McGuigan
,
B. D.
Huey
,
G. A. D.
Briggs
,
O. V.
Kolosov
,
Y.
Tsukahara
, and
M.
Yanaka
, “
Measurement of debonding in cracked nanocomposite films by ultrasonic force microscopy
,”
Appl. Phys. Lett.
80
(
7
),
1180
(
2002
);
F.
Dinelli
,
H. E.
Assender
,
N.
Takeda
,
G. A. D.
Briggs
, and
O. V.
Kolosov
, “
Elastic mapping of heterogeneous nanostructures with ultrasonic force microscopy (UFM)
,”
Surf. Interface Anal.
27
(
5–6
),
562
(
1999
);
C. S.
Deng
,
H. E.
Assender
,
F.
Dinelli
,
O. V.
Kolosov
,
G. A. D.
Briggs
,
T.
Miyamoto
, and
Y.
Tsukahara
, “
Nucleation and growth of gas barrier aluminium oxide on surfaces of poly(ethylene terephthalate) and polypropylene: Effects of the polymer surface properties
,”
J. Polym. Sci., Part B
38
(
23
),
3151
(
2000
).
21.
J. L.
Bosse
,
P. D.
Tovee
,
B. D.
Huey
, and
O. V.
Kolosov
, “
Physical mechanisms of megahertz vibrations and nonlinear detection in ultrasonic force and related microscopies
,”
J. Appl. Phys.
115
(
14
),
144304
(
2014
).
22.
D.
Graf
,
F.
Molitor
,
K.
Ensslin
,
C.
Stampfer
,
A.
Jungen
,
C.
Hierold
, and
L.
Wirtz
, “
Spatially resolved Raman spectroscopy of single- and few-layer graphene
,”
Nano Lett.
7
(
2
),
238
(
2007
).
23.
X.
Li
,
J.
Sun
,
W.
Zhao
,
Y.
Lai
,
X.
Yu
, and
Y.
Liu
, “
Intergrowth of graphite-ike crystals in hard carbon for highly reversible Na-ion storage
,”
Adv. Funct. Mater.
32
(
2
),
2106980
(
2021
).
24.
Z.
Xu
,
J.
Wang
,
Z.
Guo
,
F.
Xie
,
H.
Liu
,
H.
Yadegari
,
M.
Tebyetekerwa
,
M. P.
Ryan
,
Y.
Hu
, and
M.
Titirici
, “
The role of hydrothermal carbonization in sustainable sodium-ion battery anodes
,”
Adv. Energy Mater.
12
(
18
),
2200208
(
2022
).
25.
C.
Sole
,
N. E.
Drewett
, and
L. J.
Hardwick
, “
In situ Raman study of lithium-ion intercalation into microcrystalline graphite
,”
Faraday Discuss.
172
,
223
(
2014
).
26.
C.-F.
Chen
,
C.-H.
Park
,
B. W.
Boudouris
,
J.
Horng
,
B.
Geng
,
C.
Girit
,
A.
Zettl
,
M. F.
Crommie
,
R. A.
Segalman
,
S. G.
Louie
, and
F.
Wang
, “
Controlling inelastic light scattering quantum pathways in graphene
,”
Nature
471
(
7340
),
617
(
2011
).
27.
J.
Wang
,
H.
Wang
,
R.
Zhao
,
Y.
Wei
,
F.
Kang
, and
D.
Zhai
, “
Mechanistic insight into ultrafast kinetics of sodium cointercalation in few-layer graphitic carbon
,”
Nano Lett.
22
(
15
),
6359
(
2022
).
28.
Z.
Zhang
,
S.
Said
,
K.
Smith
,
R.
Jervis
,
C. A.
Howard
,
P. R.
Shearing
,
D. J. L.
Brett
, and
T. S.
Miller
, “
Characterizing batteries by in situ electrochemical atomic force microscopy: A critical review
,”
Adv. Energy Mater.
11
(
38
),
2101518
(
2021
).
29.
B. J.
Robinson
and
O. V.
Kolosov
, “
Probing nanoscale graphene-liquid interfacial interactions via ultrasonic force spectroscopy
,”
Nanoscale
6
,
10806
10816
(
2014
).
30.
B.
Hajgató
,
S.
Güryel
,
Y.
Dauphin
,
J.-M.
Blairon
,
H. E.
Miltner
,
G.
Van Lier
,
F.
De Proft
, and
P.
Geerlings
, “
Out-of-plane shear and out-of plane Young's modulus of double-layer graphene
,”
Chem. Phys. Lett.
564
,
37
(
2013
).
31.
L.
Suo
,
O.
Borodin
,
Y.
Wang
,
X.
Rong
,
W.
Sun
,
X.
Fan
,
S.
Xu
,
M. A.
Schroeder
,
A. V.
Cresce
,
F.
Wang
,
C.
Yang
,
Y.‐S.
Hu
,
K.
Xu
, and
C.
Wang
, ““
Water‐in-salt” electrolyte makes aqueous sodium-ion battery safe, green, and long-lasting
,”
Adv. Energy Mater.
7
(
21
),
1701189
(
2017
);
J.
Song
,
B.
Xiao
,
Y.
Lin
,
K.
Xu
, and
X.
Li
, “
Interphases in sodium‐ion batteries
,”
Adv. Energy Mater.
8
(
17
),
1703082
(
2018
);
K.
Li
,
J.
Zhang
,
D.
Lin
,
D. W.
Wang
,
B.
Li
,
W.
Lv
,
S.
Sun
,
Y. B.
He
,
F.
Kang
,
Q. H.
Yang
,
L.
Zhou
, and
T. Y.
Zhang
, “
Evolution of the electrochemical interface in sodium ion batteries with ether electrolytes
,”
Nat. Commun.
10
(
1
),
725
(
2019
).
[PubMed]
32.
Y.
Zhen
,
R.
Sa
,
K.
Zhou
,
L.
Ding
,
Y.
Chen
,
S.
Mathur
, and
Z.
Hong
, “
Breaking the limitation of sodium-ion storage for nanostructured carbon anode by engineering desolvation barrier with neat electrolytes
,”
Nano Energy
74
,
104895
(
2020
).
33.
Q.
Lin
,
J.
Zhang
,
D.
Kong
,
T.
Cao
,
S.-W.
Zhang
,
X.
Chen
,
Y.
Tao
,
W.
Lv
,
F.
Kang
, and
Q.-H.
Yang
, “
Deactivating defects in graphenes with Al2O3 nanoclusters to produce long-life and high-rate sodium-ion batteries
,”
Adv. Energy Mater.
9
(
1
),
1803078
(
2019
).
34.
C.
Yan
,
L.-L.
Jiang
,
Y.-X.
Yao
,
Y.
Lu
,
J.-Q.
Huang
, and
Q.
Zhang
, “
Nucleation and growth mechanism of anion-derived solid electrolyte interphase
,”
Angew. Chem., Int. Ed.
60
,
8521
8525
(
2021
).
35.
X.
Zhang
,
F.
Hao
,
Y.
Cao
,
Y.
Xie
,
S.
Yuan
,
X.
Dong
, and
Y.
Xia
, “
Dendrite‐free and long-cycling sodium metal batteries enabled by sodium-ether cointercalated graphite anode
,”
Adv. Funct. Mater.
31
(
15
),
2009778
(
2021
).
36.
J.
Zhang
,
D.-W.
Wang
,
W.
Lv
,
S.
Zhang
,
Q.
Liang
,
D.
Zheng
,
F.
Kang
, and
Q.-H.
Yang
, “
Achieving superb sodium storage performance on carbon anodes through an ether-derived solid electrolyte interphase
,”
Energy Environ. Sci.
10
(
1
),
370
(
2017
).
37.
B.
Jache
,
J. O.
Binder
,
T.
Abe
, and
P.
Adelhelm
, “
A comparative study on the impact of different glymes and their derivatives as electrolyte solvents for graphite co-intercalation electrodes in lithium-ion and sodium-ion batteries
,”
Phys. Chem. Chem. Phys.
18
(
21
),
14299
(
2016
).
38.
W.
Liu
,
Z.
Chen
,
Z.
Zhang
,
P.
Jiang
,
Y.
Chen
,
E.
Paek
,
Y.
Wang
, and
D.
Mitlin
, “
Lithium-activated SnS–graphene alternating nanolayers enable dendrite-free cycling of thin sodium metal anodes in carbonate electrolyte
,”
Energy Environ. Sci.
14
(
1
),
382
(
2021
).
Published open access through an agreement withJISC Collections