Structural water engaged disordered vanadium oxide nanosheets for high capacity aqueous potassium-ion storage

Aqueous electrochemical energy storage devices using potassium-ions as charge carriers are attractive due to their superior safety, lower cost and excellent transport properties compared to other alkali ions. However, the accommodation of potassium-ions with satisfactory capacity and cyclability is difficult because the large ionic radius of potassium-ions causes structural distortion and instabilities even in layered electrodes. Here we report that water induces structural rearrangements of the vanadium-oxygen octahedra and enhances stability of the highly disordered potassium-intercalated vanadium oxide nanosheets. The vanadium oxide nanosheets engaged by structural water achieves high capacity (183 mAh g−1 in half-cells at a scan rate of 5 mV s−1, corresponding to 0.89 charge per vanadium) and excellent cyclability (62.5 mAh g−1 in full cells after 5,000 cycles at 10 C). The promotional effects of structural water on the disordered vanadium oxide nanosheets will contribute to the exploration of disordered structures from earth-abundant elements for electrochemical energy storage.

Supplementary Figure S7. The average gravimetric electrode capacities of the fully hydrated disordered KVO in comparison with the partially hydrated and crystalline KVO materials through the 5,000 charge/discharge cycles at a current density of 20 A/g. Figure S8. The neutron PDF analysis of partially hydrated, fully hydrated KVO, as well as commercial V2O5. (a) The comparison of coherence in the Neutron PDFs of the partially hydrated (not soaked in water for two weeks) and fully hydrated (soaked in water for two weeks) disordered KVO nanosheets with Commercial V2O5. Oxygen correlations from fitted local structure models for the (b) fully hydrated sample and (c) partially hydrated sample. Comparison of the fully hydrated and partially hydrated sample interactions from fitted structural models (d) OV-OV (e) OV-OW (f) OW-OW. (OV is the oxygen atoms from V-O bilayers and Ow is the oxygen atoms from structural water molecules.) Figure S9. The comparison of coherence in the X-ray and neutron PDFs of the fully hydrated (soaked in water for two weeks) disordered KVO nanosheets. Figure S10. A simplified model was used to construct stacking patterns containing a single bilayer. The local structure was fit to 6 Å using a fully isotropic, fully occupied model for the vanadium oxide bilayers in which the water (represented by oxygen) and potassium occupy the same atomic position within the layers at the appropriate occupancies to give the correct chemical composition. The stacking pattern is shown in both the (a) ac and (b) bc plane. (c) The resulting X-ray calculated PDFs and difference curves are also shown. Figure S11. A simplified model was used to construct stacking patterns containing two bilayers. The local structure was fit to 6 Å using a fully isotropic, fully occupied model for the vanadium oxide bilayers in which the water (represented by oxygen) and potassium occupy the same atomic position within the layers at the appropriate occupancies to give the correct chemical composition. The vanadium oxide bilayers were shifted using least squares refinement of the X-ray PDF data of the fully hydrated KVO. The stacking pattern is shown in both the (a) ac and (b) bc plane, the gray shadows are of the non-shifted model (as shown in Figure S10) which are there to highlight the shifting of the bilayers. (c) The resulting X-ray calculated PDFs and difference curves are also shown. Figure S12. A simplified model was used to construct stacking patterns containing three bilayers. The local structure was fit to 6 Å using a fully isotropic, fully occupied model for the vanadium oxide bilayers in which the water (represented by oxygen) and potassium occupy the same atomic position within the layers at the appropriate occupancies to give the correct chemical composition. The vanadium oxide bilayers were shifted using least squares refinement of the X-ray PDF data of the fully hydrated KVO. The stacking pattern is shown in both the (a) ac and (b) bc plane, the gray shadows are of the non-shifted model (as shown in Figure S10) which are there to highlight the shifting of the bilayers. (c) The resulting X-ray calculated PDFs and difference curves are also shown. Figure S13. A simplified model was used to construct stacking patterns containing four bilayers. The local structure was fit to 6 Å using a fully isotropic, fully occupied model for the vanadium oxide bilayers in which the water (represented by oxygen) and potassium occupy the same atomic position within the layers at the appropriate occupancies to give the correct chemical composition. The vanadium oxide bilayers were shifted using least squares refinement of the X-ray PDF data of the fully hydrated KVO. The stacking pattern is shown in both the (a) ac and (b) bc plane, the gray shadows are of the non-shifted model (as shown in Figure S10) which are there to highlight the shifting of the bilayers. (c) The resulting X-ray calculated PDFs and difference curves are also shown. Figure S14. A simplified model was used to construct stacking patterns containing five bilayers. The local structure was fit to 6 Å using a fully isotropic, fully occupied model for the vanadium oxide bilayers in which the water (represented by oxygen) and potassium occupy the same atomic position within the layers at the appropriate occupancies to give the correct chemical composition. The vanadium oxide bilayers were shifted using least squares refinement of the X-ray PDF data of the fully hydrated KVO. The stacking pattern is shown in both the (a) ac and (b) bc plane, the gray shadows are of the non-shifted model (as shown in Figure S10) which are there to highlight the shifting of the bilayers. (c) The resulting X-ray calculated PDFs and difference curves are also shown. Figure S15. A simplified model was used to construct stacking patterns containing six bilayers. The local structure was fit to 6 Å using a fully isotropic, fully occupied model for the vanadium oxide bilayers in which the water (represented by oxygen) and potassium occupy the same atomic position within the layers at the appropriate occupancies to give the correct chemical composition. The vanadium oxide bilayers were shifted using least squares refinement of the X-ray PDF data of the fully hydrated KVO. The stacking pattern is shown in both the (a) ac and (b) bc plane, the gray shadows are of the non-shifted model (as shown in Figure S10) which are there to highlight the shifting of the bilayers. (c) The resulting X-ray calculated PDFs and difference curves are also shown. Figure S16. Investigation of the difference between inter and intra-bilayer interactions and stacking of bilayers. (a) The isolated inter and intra-bilayer correlations to the fit of X-ray PDF in main text Fig. 1. Although the first inter-bilayer correlations start at 5 Å, it is not until 9 Å where the inter-bilayer correlations become more significant. Nevertheless, the shape of the distinct peaks is still due to the intra-bilayer interactions as the inter-bilayer interactions produce non distinct broad features. (b) The decreasing goodness of fit (RWP) with increasing number of shifted layers in the different stacking models, individual fits shown in Figure S10 to S15. The increasing number of shifted bilayers, all possessing the same local structure for vanadium oxide bilayers, significantly reduced the RWP with only minor shifts suggesting that there is disordered staking of the bilayers in the material rather than a different stacking pattern. (c) Difference curves of the stacking models with the various numbers of independently shifted bilayers. An improvement can clearly be seen in the fit of the data at high-r with increasing number of shifted bilayers, these results lead us to believe that our assertion of turbostratic disorder accounts for the increasingly poor fit at high-r range. Three cycle CV measurement at 1 mV/s in a 3 M KCl electrolyte conducted while collecting the in situ XRD spectra. The in situ XRD spectra of the disordered KVO also contained a brag diffraction peak at 22.94° 2θ, identified as the (020) diffraction plane, that also showed a potential dependence during cycling. The (020) diffraction plane shifts opposite to the (001) layered basal diffraction plane, but changed an order of magnitude smaller than the (001) plane, by only 0.22° 2θ. During the charging process the (020) shifted to higher 2θ angle (23.16° 2θ) and smaller dspacing (1.850 Å), due to the increased oxidation state of V and shorter V-O bond distance. Upon discharging the (020) peak shifts to lower 2θ (22.94° 2θ) and larger d-spacing (1.867 Å) as the V is reduced and the V-O bond distance increases.  Supplementary Table S1. X-ray/Neutron PDF results of the disordered KVO nanosheets from Figure 1d&e.

Supplementary Notes
Calculation of capacitance in half-cell.
The CV measurements were conducted in 1M KCl electrolyte at scan rates from 5 to 200 mV s -1 within a potential range from -0.1 V to 0.9 V (vs. Ag/AgCl). The mass-specific capacitance (CMS) from CV measurement in half-cell was calculated from: Where i (A) is the measured current at certain time of t (s), m (g) is the mass of active material loaded on working electrode, ∆V (V) is potential window, t0 (s) and tF (s) are respective times at the initial potential and the final potential.
Electrochemical full cell calculations.
The cell capacitance Ccell (F), cell mass-specific capacitance CMS (F g -1 ), electrode mass-specific capacitance CMS(electrode) (F g -1 ), discharge cell capacity by mass Cdischarge (mAh g -1 ), discharge electrode capacity by mass Cdischarge(electrode) (mAh g -1 ), specific energy EMS (Wh kg -1 ), specific power PMS (W kg -1 ) and coulombic efficiency (η) and energy efficiency (γ ) were calculated by following equations: Cell capacitance: Where i (A) is the applied constant current, t (s) is discharge time of the cell device, U (V) is potential window, M (g) is the total mass of active materials on both electrodes and m (g) is the mass of active materials on one electrode.