Research ArticlesChemistry

Three-dimensional holey-graphene/niobia composite architectures for ultrahigh-rate energy storage

See allHide authors and affiliations

Science  12 May 2017:
Vol. 356, Issue 6338, pp. 599-604
DOI: 10.1126/science.aam5852

As with donuts, the holes matter

Improving the density of stored charge and increasing the speed at which it can move through a material are usually opposing objectives. Sun et al. developed a Nb2O5/holey graphene framework composite with tailored porosity. The three-dimensional, hierarchically porous holey graphene acted as a conductive scaffold to support Nb2O5. A high mass loading and improved power capability were reached by tailoring the porosity in the holey graphene backbone with higher charge transport in the composite architecture. The interconnected graphene network provided excellent electron transport, and the hierarchical porous structure in the graphene sheets facilitated rapid ion transport and mitigated diffusion limitations.

Science, this issue p. 599

Abstract

Nanostructured materials have shown extraordinary promise for electrochemical energy storage but are usually limited to electrodes with rather low mass loading (~1 milligram per square centimeter) because of the increasing ion diffusion limitations in thicker electrodes. We report the design of a three-dimensional (3D) holey-graphene/niobia (Nb2O5) composite for ultrahigh-rate energy storage at practical levels of mass loading (>10 milligrams per square centimeter). The highly interconnected graphene network in the 3D architecture provides excellent electron transport properties, and its hierarchical porous structure facilitates rapid ion transport. By systematically tailoring the porosity in the holey graphene backbone, charge transport in the composite architecture is optimized to deliver high areal capacity and high-rate capability at high mass loading, which represents a critical step forward toward practical applications.

Batteries and supercapacitors represent two complementary electrochemical energy storage (EES) technologies (14), with the batteries offering high energy density but low power density and supercapacitors providing high power density with low energy density. Although lithium (Li)–ion batteries currently dominate the market for powering consumer electronic devices and are making in-roads into transportation and grid storage, there is a growing technological demand for more rapid energy storage (high power) without compromising energy density. Thus, there is considerable interest in creating materials that combine the high energy density of battery materials with the short charging times and long cycle life of supercapacitors (514).

The combination of high energy density and high power density requires materials that can store a large number of charges (such as Li ions) and an electrode architecture that can rapidly deliver sufficient charges (electrons and ions) in a given charge/discharge duration. This behavior can occur in nanostructured materials that include very thin films and materials reduced to nanoscale dimensions—that is, materials with low mass loading (<1 mg cm−2) (8, 10, 1517). However, such nanoscale materials cannot be readily scaled to electrodes that have practical levels of mass loading (~10 mg cm−2) because of increasing ion diffusion limitations in thicker electrodes. Moreover, the highly promising electrochemical properties achieved in low-mass loaded electrodes rapidly diminish in practical devices when other passive components such as current collectors and separators (~10 mg cm−2) are included (18). As a result, the scaled areal capacity or current density rarely exceeds those of today’s Li-ion batteries (~3 mAh cm−2, 4 mA cm−2) (19, 20).

In general, sustaining the same gravimetric capacity and current density in higher mass-loaded electrodes (for example, 10 versus 1 mg cm−2) requires proportionally higher ion and electron currents across a longer charge transport distance. Recent studies focusing on the question of mass loading in Li-ion batteries showed that the ionic current was especially critical and defined a penetration depth to describe the utilization of the active electrochemical material for a given electrode thickness (19). Therefore, in a thick electrode, the mass transport limit of ions is particularly important because insufficient charge transport could severely degrade the capacity within a fixed charge/discharge window because of considerably higher overpotential (21). These considerations suggest that the mass loading required to translate the high performance achieved by many nanoscale materials into practical devices is a fundamental challenge in electrode design rather than a matter of scaling.

We report on the design and mass loading properties of an electrode architecture using a three-dimensional holey graphene framework (3D-HGF) as the conductive scaffold for electrochemical active material [for example, orthorhombic niobia (Nb2O5)]. The highly interconnected graphene network in the 3D-HGF provides excellent electron transport properties, and its hierarchical porous structure, with large-sized pores (macropores) in the 3D network and tunable micro- to meso-pores in the graphene sheets, facilitates rapid ion transport and mitigates diffusion limitations throughout the entire electrode architecture. This morphology produces interpenetrating electron transport and ion transport paths that enable high capacity at high charge/discharge rates at practical levels of mass loading (22, 23).

We used Nb2O5 as a model system for evaluating the effect of mass loading for the 3D-HGF scaffold. Although the Li insertion properties of Nb2O5 have been known for many years, only recently was it observed that the orthorhombic form of Nb2O5 (T-Nb2O5) is able to retain high levels of charge storage at high rate (for example, 110 mAh g−1 at 60C) (24). The charge storage properties of T-Nb2O5 are not controlled by semi-infinite diffusion as usually occurs in battery materials. Instead, surface-controlled kinetics occurring in the bulk of the material lead to its unusually high-rate capability (24, 25). However, because of the limited electronic conductivity of T-Nb2O5, the high rate capability can only be realized in thin-film electrodes or at relatively low mass loading (generally <2.0 mg cm−2) despite considerable efforts in nanostructure design and the use of carbon-based hybrid architectures (24, 2629). In the thicker electrodes, the overall rate capability is limited by insufficient delivery of ions to the electrode material surface (29). The susceptibility of the high rate performance of T-Nb2O5 to mass loading makes it an interesting model system for investigating the design of electrode architectures.

Synthesis and characterization of the hierarchically porous composites

We prepared the free-standing Nb2O5/HGF composites using a two-step process in which graphene oxide (GO) or holey graphene oxide (HGO) is combined with Nb2O5 (Fig. 1 and supplementary text) (30). The porous composite electrodes are designed to fill with electrolyte and thus enhance the ion transport kinetics (31, 32). We prepared the composites from HGO sheets with varying pore sizes, which were prepared by etching GO sheets with H2O2 for 0, 0.5, 1.0, and 2.0 hours. The oxidative-etching process initiates from the chemically more active oxygenic defect sites and propagates in the basal plane of GO to form increasingly larger pores with increasing etching time (23, 33, 34), as demonstrated through high-resolution transmission electron microscopy (HR-TEM) studies (Fig. 2, A to D). The two-step synthesis produce free-standing 3D porous composites with graphene framework (GF) or holey graphene framework (HGF) as the conductive scaffolds for Nb2O5 nanoparticles (Fig. 2E and fig. S1), which are identified as Nb2O5/GF for the material with a nonholey GO as the graphene source, and Nb2O5/HGF-0.5, Nb2O5/HGF-1.0, and Nb2O5/HGF-2.0 for those materials using HGO as the graphene sources (obtained by etching in H2O2 for 0.5, 1.0, and 2.0 hours), respectively.

Fig. 1 Illustration of the two-step process flow to prepare 3D hierarchically porous composite architecture.

The Nb2O5 is uniformly decorated on the first portion of GO (~4 wt % of the composite) in step one and then mixed with the second portion of GO/HGO (~11 wt % of the composite) in solution, followed by a reduction process to produce the monolithic free-standing composite. HGOs with tailored nanopores were prepared by etching in H2O2 for 0, 0.5, 1.0, to 2.0 hours and used to prepare Nb2O5/GF, Nb2O5/HGF-0.5, Nb2O5/HGF-1.0, and Nb2O5/HGF-2.0, with various degrees of porosity in the graphene sheets. The samples were annealed at 600°C in argon at the end of each step in order to produce the orthorhombic Nb2O5 (T-Nb2O5) and further deoxygenate the RGO sheets so as to improve their electron transport properties. The amount of T-Nb2O5 is controlled to be ~85 wt % in the final composites.

Fig. 2 Material characterization of T-Nb2O5/HGF composites.

(A to D) TEM images of graphene sheets with tailored pores obtained by etching in H2O2 for 0, 0.5, 1.0, and 2.0 hours, respectively. (E) Cross-sectional scanning electron microscopy image of Nb2O5/HGF composite shows 3D hierarchically porous structure. (Inset) A free-standing monolithic composite used to make the electrode. (F) XRD patterns of the as-synthesized Nb2O5/G powders before and after 600°C annealing, and free-standing Nb2O5/HGF composite. (G) TEM image of graphene sheets with uniformly decorated T-Nb2O5 nanoparticles. (H) HR-TEM image of T-Nb2O5 nanoparticles. (I) Raman spectra of Nb2O5/G powder after thermal annealing, and free-standing Nb2O5/GF and Nb2O5/HGF electrodes. The D and G bands are characteristic of RGO; The Raman bands at 120, 230, 310, and 690 cm−1 further confirm the orthorhombic phase of T-Nb2O5. (J) Comparison of DFT pore size distributions. The prominent pore size shifts from micropores (~1.5 nm) to mesopores (2.7 nm) for the composite prepared from HGO with increasing etching time.

The as-synthesized Nb2O5 on GO in the first step is amorphous and can be effectively converted into the orthorhombic phase (JCPDS 30-873) upon annealing at 600°C in argon, as indicated by the x-ray diffraction (XRD) studies (Fig. 2F and fig. S2). The crystallite size of T-Nb2O5 is ~15 nm as determined by the Scherrer equation based on XRD peak width (35). The TEM image (Fig. 2G) further confirms that T-Nb2O5 nanoparticles are homogeneously decorated on the graphene sheets, with a size of around 10 to 15 nm, which is consistent with that determined from the XRD studies. The HR-TEM image shows a lattice spacing of 0.39 nm (Fig. 2H), corresponding to the (001) plane of the orthorhombic phase. Raman spectroscopy studies show the expected D and G bands of reduced GO (RGO) in the composites, and the characteristic Raman bands at 120, 230, 310, and 690 cm−1 for the orthorhombic phase T-Nb2O5 (Fig. 2I). In addition, N2 adsorption/desorption isotherms (fig. S3) show various cumulative volumes of pores, and the density functional theory (DFT) analysis indicates that the prominent pore sizes in the 3D Nb2O5/HGF composites nearly double from ~1.5 to 2.7 nm with increasing etching time (Fig. 2J), which is consistent with HR-TEM studies (Fig. 2, A to D). The specific surface area of the composites also increases from 63 to ~83 m2 g−1 with the increasing etching time from 0 to 2.0 hours (table S1).

The two-step synthesis approach effectively produces mechanically strong 3D porous composites with high mass loading of Nb2O5 nanoparticles and sufficient electrical conductivity for high power performance. In contrast to other synthesis methods (figs. S4 and S5), the two-step approach decouples the active material loading step from the formation of the 3D architecture, thus offering a general strategy to incorporate a wide range of active materials into the 3D architecture without affecting the overall structure.

Tuning electrochemical properties by porosity

In order to ensure a proper comparison for the different Nb2O5/HGF composite electrodes, the amount of T-Nb2O5 is controlled to be around 85 weight % (wt %) for all composites [fig. S6, thermogravimetric analysis (TGA)]. The porosity is controlled at 0.60 ± 0.03, with a tap density of 1.54 ± 0.09 g cm−3 for all the composite electrodes after compression for coin cell assembly (supplementary text). To probe the effect of structural features on the charge transport kinetics, we first conducted electrochemical impedance spectroscopy (EIS) measurements on composites by using a symmetric cell with two identical electrodes, which can provide more accurate impedance spectra of the electrode-electrolyte interface than that of typical EIS measurement of asymmetric cell with a counter electrode (such as a lithium metal electrode) (36, 37). Before lithiation, the Nyquist plots (Fig. 3A and fig. S7) of a symmetric cell at a practical mass loading of 11 mg cm−2 describe a nonfaradaic process with a state of charge (SOC) at 0%. These plots exhibit a 45° slope in the frequency region between ~5 and 100 Hz and quasi-vertical lines at lower frequency (<1 Hz). These features indicate a nonfaradaic process that can be further validated analytically by an equivalent circuit by using a transmission line model (TLM) for porous electrodes (fig. S8A) (36, 37). The projection of the 45° slope to the real axis reflects the ionic resistance for the electrolyte-filled pores inside the 3D electrode structures (Fig. 3A and fig. S7), which is a key parameter for characterizing the rate capability of a porous HGF electrode during the charge/discharge process (37). This projection is defined as Rion/3, as derived from a TLM for cylindrical pores (the complete derivation is available in the supplementary text) (36, 37). The determination of Rion/3 for each composite electrode in Fig. 3A is shown in fig. S7, A to D. In our simulations, we followed the approach of Ogihara et al., using the finite length Warburg element open circuit terminus (Wo) for the TLM at 0% SOC (36). We added another circuit element (fig. S7E) to account for an additional resistance at high frequency because the EIS showed a depressed semicircle in the frequency region >100 Hz. This contribution (Rhigh) is believed to arise from an interfacial charge transfer resistance of the graphene component in the composites (38, 39). Another parameter derived from the TLM is the electrolyte resistance, Rsol. The simulations (fig. S7) enable us to obtain values for Rion, Rhigh, and Rsol (table S2) for the various electrodes shown in Fig. 3A.

Fig. 3 Evolution of kinetic properties and electrochemical characteristics with porosity.

(A) Comparison of Nyquist plots obtained from potentiostatic EIS of a symmetric cell using two identical electrodes (11 mg cm−2) at a SOC 0%. The projection of the 45° slope in the high-frequency region is used to determine the ionic resistance for the electrolyte-filled porous architectures in a nonfaradaic process (fig. S4, A to D). The open symbols and solid lines represent the experimental and simulation results, respectively. (B) The ionic resistance (Rion) as a function of the porosity of the electrode materials. With increasing porosity, the ionic resistance is reduced substantially. (C) Galvanostatic charge-discharge curves of electrodes with tunable nanoporosity at a rate of 10C in the voltage window 1.1 to 3.0 V (versus Li/Li+). (D) Comparison of specific capacities (normalized by the total mass of the electrodes) at various rates (1 to 100C) for composite electrodes with tunable nanoporosity.

The gradual changes in projection length values for the different electrodes show a decrease in ionic resistance (Rion) from 27.1 to 8.8 ohm cm2, with increasing pore size in the HGF scaffold (Fig. 3B and table S2). Additionally, the variation in the imaginary part of the capacitance with frequency indicates that the optimized porous architecture (Nb2O5/HGF-2.0) exhibits a much shorter time constant (fig. S9) (40). These studies demonstrate that the ion transport kinetics can be improved by tailoring the pore size in the holey graphene sheets that form the 3D graphene scaffold. Here, the in-plane pores in the holey graphene sheet function as ion transport shortcuts in the hierarchical porous structure to facilitate rapid ion transport throughout the entire 3D electrode and greatly improve ion access to the surface of the T-Nb2O5. As expected, neither Rhigh nor Rsol shows significant variation with electrode porosity (table S2).

We have carried out a series of galvanostatic studies to examine the effect of C-rate and composite architecture at a mass loading of 6 mg cm−2. The results indicate that the Nb2O5/HGF-2.0 electrode consistently exhibits higher capacity for a given C-rate (Fig. 3, C and D, and figs. S10 and S11). The continuous decrease in voltage with increasing capacity is observed in all electrodes, suggesting a pseudocapacitive behavior. Another apparent characteristic of Nb2O5/HGF electrodes is that there is a relatively high cut-off voltage of 1.1 V, which is consistent with previous studies (41).

The 3D porous composites with tailored nanopores exhibit enhanced specific capacity compared with the macroporous electrode without tailored nanopores (Nb2O5/GF), and this difference widens with increasing C-rate (Fig. 3D). For example, at 100C, the Nb2O5/HGF-2.0 electrode delivers a specific capacity of 75 mAh g−1, which is more than two times that of the Nb2O5/GF electrode (35 mAh g−1). Sweep voltammetry experiments are consistent with these results because surface-controlled kinetics dominates these electrodes (figs. S12 and S13 and supplementary text). Hence, the tailored porosity is essential for supporting the high rate capability of T-Nb2O5. Last, because the Nb2O5/HGF electrodes contain no additional conductive additive or binder, their gravimetric capacities are considerably higher than other Nb2O5 electrodes when normalized by the total mass of the electrode material (table S3); Moreover, the loading of the Nb2O5/HGF electrodes is 6 and 11 mg cm−2, whereas that of the other electrodes is in the range of 1 to 2 mg cm−2.

Effects of mass loading on electrode performance

Three different levels of mass loading were investigated, corresponding to a typical loading amount for research studies (1 mg cm−2), one that is representative of practical levels of loading (11 mg cm−2) and an intermediate level of loading (6 mg cm−2). The studies also considered different C-rates and different electrode architectures. For the latter, the electrochemical behavior of one of the optimized 3D porous HGF composites with tailored nanopores (Nb2O5/HGF-2.0) was compared with two other composites: 3D porous Nb2O5/GF composite with no in-plane nanopores and Nb2O5/G composite with a randomly stacked graphene network (the different synthesis processes are described in the supplementary materials, materials and methods, and the structural comparison is provided in fig. S14).

For the Nb2O5/G control electrodes, it is apparent that the mass loading can dramatically alter the galvanostatic charge-discharge characteristics (Fig. 4A). The voltage-capacity curves exhibit an increasingly steeper slope and larger voltage drop with increasing mass loading. This response results from an increasingly larger internal resistance that in turn leads to higher overpotentials and lower capacities (21). For example, at 10C the capacity of the Nb2O5/G control electrode decreases from 85 mAh g−1 at mass loading of 1 mg cm−2 to 25 mAh g−1 at 11 mg cm−2 (Fig. 4, A and D). Such rapidly degrading performance with increasing mass loading highlights the challenges in delivering sufficient ion current densities to retain the same gravimetric energy storage performance in thicker electrodes.

Fig. 4 Effects of mass loading on electrochemical characteristics.

(A and B) Galvanostatic charge-discharge curves for (A) the Nb2O5/G control electrode and (B) the Nb2O5/HGF-2.0 electrode at a rate of 10C for the mass loadings of 1, 6, and 11 mg cm−2. (C) Comparison of the rate performance between 1C and 100C for Nb2O5/HGF-2.0 (open) and Nb2O5/G (solid) electrodes under different mass loadings (1, 6, and 11 mg cm−2). (D and E) Retention of specific capacity at (D) 10C and (E) 50C as a function of mass loading for three different electrodes. The properties of Nb2O5/HGF-2.0 and Nb2O5/GF electrodes were normalized by the total mass of the electrode materials (free of conductive additives and binders); the control Nb2O5/G electrodes were normalized by the total mass of the electrode materials (including binders and conductive additives).

The charge-discharge curves of the optimized Nb2O5/HGF-2.0 electrode show a relatively small voltage drop and capacity loss, with increasing mass loading (Fig. 4B). This indicates that the 3D hierarchically porous HGF architecture leads to a much lower internal resistance. As a result, the Nb2O5/HGF-2.0 electrode shows much less capacity degradation induced by mass loading at various C-rates (Fig. 4C), and a capacity of 139 mAh g−1 is maintained with 11 mg cm−2 loading at a high rate of 10C. This value is only 7% less than the 1 mg cm−2 mass loading for the same electrode architecture (Fig. 4D). Even at the high rate of 50C, the Nb2O5/HGF-2.0 electrode with a mass loading of 11 mg cm−2 retains a capacity of ~74 mAh g−1, a decrease of only 28% from that of the 1 mg cm−2 electrode (Fig. 4E). In contrast, increased mass loading for Nb2O5/G control electrodes (Fig. 4C) leads to a significant decrease in energy storage properties with essentially diminished capacity at high C-rate (only 7 mAh g−1 at 50C). The Nb2O5/HGF-2.0 electrode also exhibits stable cycling performance (fig. S15). After 10,000 cycles at 10C, the capacity is ~125 mAh g−1 (90% retention), and Coulombic efficiency is above 99.9%, demonstrating the robust porous architecture of the 3D composites.

Implications of high mass loading

Increasing the areal capacity can reduce the relative overhead from the current collectors and separators and is thus critical for achieving higher cell level energy and power density and lowering cost (19). For this reason, the areal capacity provides an important measure for assessing the performance of an EES system (18, 19, 21).

The relation between areal capacity (mAh cm−2) and areal mass loading per (mg cm−2) gives us the specific capacity (mAh g−1). As indicated in Figs. 3 and 4, the specific capacity depends on C-rate, the mass loading, and the electrode architecture; thus, we can determine their influences on areal capacity. The areal capacity versus mass loading for the optimized Nb2O5/HGF-2.0 composite electrode and various control electrodes at a charge-discharge rate of 10C is plotted in Fig. 5A. The Nb2O5/HGF-2.0 electrode has a much higher slope at the 10C as compared with other architectures. The linear response for the Nb2O5/HGF-2.0 composite electrode suggests that no limitation has been reached in mass loading at the rate of 10C (which corresponds to a current density of 22 mA cm−2), whereas the other control electrodes exhibit apparent plateaus or even a decrease in capacity, with increasing mass loading due to the worsening charge transport characteristics in thicker electrodes.

Fig. 5 Performance metrics for electrodes with high mass loading.

(A) Dependence of areal capacity on mass loading at 10C for Nb2O5/HGF-2.0, Nb2O5/GF, and Nb2O5/G electrodes. (B) The correlation of areal capacity with mass loading (1 to 22 mg cm−2) at various C-rates for the Nb2O5/HGF-2.0 electrode. (C) The effect of current density and mass loading on the areal capacity of the Nb2O5/HGF-2.0 electrode. (D) The comparison of areal performance metrics of Nb2O5/HGF-2.0 electrode with various commercial and research anodes, including graphite anodes, high-capacity Si anodes, and high-rate Nb2O5 anodes. (E) Translation of specific capacities at various current densities when the mass of current collectors (~10 mg cm−2) is included for the Nb2O5/HGF-2.0 electrode with mass loadings of 1 and 11 mg cm−2.

A second relationship evident from Figs. 3 and 4 is the dependence of specific capacity on C-rate for the Nb2O5/HGF-2.0 electrode. In Fig. 5B, the areal capacity increases linearly, with mass loading at C-rates of 5 and 10C up until 11 mg cm−2. At higher mass loading levels, deviations from linearity occur as the areal capacity begins to reach a plateau at the 5C rate and experiences a slight decrease in areal capacity at 10C. In the linear region, there is a relatively small difference in slope, which is expected because the specific capacities vary only by <10% for the different levels of mass loading. The plateau-like behavior at 5C has been interpreted as reaching a limiting condition. Specifically, Gallagher et al. proposed that the plateau was an indication that the penetration depth for the ionic current was being reached (19). The transition between the linear response and the plateau suggests that the higher loading becomes progressively less beneficial. The decrease in areal capacity (for example, 10C at loadings above 11 mg cm−2) was interpreted as being the result of concentration gradients being generated. It would seem that under the conditions involved here—thick electrodes and high current density—concentration effects are likely to develop.

The data in Fig. 5B are rearranged in Fig. 5C to show the effect of both loading and current density on the areal capacity of the Nb2O5/HGF-2.0 electrode. The highest loading (22 mg cm−2 here) leads to an areal capacity reaching 3.9 mAh cm−2. The curve shape for each mass loading level is sigmoidal: The high areal capacities will level off to a maximum value at sufficiently low current density as all the electrode material becomes accessible to electrolyte penetration (19), and all the samples with different mass loadings show a fairly steep decrease in areal capacity when the current density is increased beyond 20 mA cm−2. This response suggests that charge transport within the electrolyte is becoming the limiting factor at such high current density (19). That is, ion transport in the electrolyte-filled porous electrode is responsible for the fall-off in areal capacity, indicating that the capacity of the active electrode is not being fully used.

We have also compared the areal capacity versus current density with state-of-the-art commercial graphite anodes (19, 20, 42) and representative research anodes (such as Nb2O5 and Si) (Fig. 5D) (1517, 24, 26, 27, 4345). In general, in contrast to the rapid decay in areal capacity with increasing current density observed in typical commercial or research devices, our composite electrode exhibits a much more gradual change. This is particularly evident in the linear plot of the areal capacity versus current density (fig. S16). At >5 mA cm−2, the optimized Nb2O5/HGF-2.0 electrode delivers a much higher areal capacity at a given current density. Moreover, the composite electrodes continue to provide energy storage at high current density beyond 20 mA cm−2, at which no other battery material system seems to perform.

The advantage of having electrodes with high mass loading becomes more apparent when the mass of the inactive components (such as current collectors ~10 mg cm−2) are taken into account (18, 21). As demonstrated in Fig. 5E, an electrode with a mass loading of 11 mg cm−2 will experience ~50% decrease in both the specific capacity and current density if the mass of the current collector (~10 mg cm−2) is considered. By comparison, the corresponding performance for an electrode with a mass loading of 1 mg cm−2 will be reduced >90%. Thus, even with a high performance material, when the mass loading is low, the performance of the active material has much less influence on the final device because the mass of the passive components dominates the total electrode mass. This analysis only considers the additional weight from the current collectors. The inclusion of separators and packaging (table S4) could further diminish the performance of the device and further establish the importance of achieving high mass loading for practical devices.

Together, our studies demonstrate that tailored porosity in the 3D conductive scaffold is essential for achieving optimized charge transport and high-rate energy storage at practical levels of mass loading. We showed that the Nb2O5/HGF composite electrode with optimized porosity is able to support a significant increase in mass loading with little decrease in electrochemical performance. At rates as high as 10C, there is little difference in specific capacity for mass loadings ranging from 1 to 11 mg cm−2, the latter being relevant to practical device applications. The achievement of high area capacity with high-rate capability at large mass loading represents an essential step toward practical electrochemical energy storage devices.

Supplementary Materials

www.sciencemag.org/content/356/6338/599/suppl/DC1

Materials and Methods

Supplementary Text

Figs. S1 to S16

Tables S1 to S4

References (4648)

References and Notes

  1. Materials and methods are available as supplementary materials.
Acknowledgments: H.S., H.F., and I.S. extend sincere appreciation to the Deanship of Scientific Research at King Saud University for its funding of this research through grant PEJP-17-01 (material preparation and electrochemical studies). X.D. and M.L. thank the U.S. Department of Energy (DOE) Office of Basic Energy Sciences, Division of Materials Science and Engineering, award DE-SC0008055 (structural characterizations). Y.H. M.D. and Z.Z. appreciate the support from the National Science Foundation through award DMR-1437263 (ion transport studies). B.D. and J.L. greatly appreciate the support by the Office of Naval Research (N00014-16-1-2164) (impedance analysis). L.M., J.L., X.X., and G.H. thank the Chinese Scholar Council scholarship for the financial support (materials preparation of electrochemical studies). L.M. also thanks a postdoctoral fellowship from Hunan University. Holy-graphene/niobia composite are available from the University of California, Los Angeles (UCLA) under a materials transfer agreement with the University. A provisional patent application has been filed by UCLA (UCLA Case 2017-216) that covers the subject described here.
View Abstract

Navigate This Article