Olivine-type compounds: method for their preparation and use in cathode materials for sodium-ion batteries

Abstract

This disclosure provides for Olivine-type compounds, their preparation and use in cathode materials for sodium-ion batteries. The olivine-type compounds of the invention are obtained by a direct synthesis embodying a hydrothermal method.

Claims

1. A compound of general formula Na.sub.hLi.sub.iM′.sub.jM″.sub.kPO.sub.4, wherein: h is 1, 0<i≤1, 0≤j≤1 and 0≤k≤1, and wherein M′ and M″ are each independently a metal selected from the group consisting of: Mg, Ca, Mn, Fe, Co, Ni and Cu, and optionally i is not (1-h); and/or (0<j≤1 and 0<k≤1); or of general formula NaMn.sub.1−xMg.sub.xPO.sub.4, wherein: x≈0.1; or of general formula Na.sub.1−xMg.sub.xPO.sub.4, wherein: 0<x≤1.

2. A compound according to claim 1, which is: of general formula NaLi.sub.1−xM′.sub.xPO.sub.4, wherein: 0≤x<1; or of general formula NaLi.sub.1−x−yM′.sub.xM″.sub.yPO.sub.4, wherein: 0≤x<1 and 0≤y<1 and x+y<1; and wherein M′ and M″ are each independently a metal selected from the group consisting of: Mg, Ca, Mn, Fe, Co, Ni and Cu.

3. A compound according to claim 1, which is a sodium phospho-olivine compound.

4. A compound according to claim 1, which is: of general formula NaFe.sub.1−x−yMn.sub.xLi.sub.yPO.sub.4, wherein: 0≤x<1 and 0≤y<1 and x+y<1.

5. A cathode material for sodium-ion batteries, comprising: a compound as defined in claim 1, optionally with a carbon material.

6. A sodium-ion battery, comprising a cathode material as defined in claim 5.

7. The compound according to claim 1, wherein 0<j≤1 and 0<k≤1.

8. A compound of general formula Na.sub.1−xLi.sub.xFe.sub.1−yMn.sub.yPO.sub.4 or Na.sub.1−xLi.sub.xFe.sub.yMn.sub.1−yPO.sub.4, wherein: 0<x<1 and 0<y<1.

9. A compound according to claim 8, having the formula Na.sub.0.9Li.sub.0.1Fe.sub.0.22Mn.sub.0.78PO.sub.4 or Na.sub.0.9Li.sub.0.1Fe.sub.0.25Mn.sub.0.75PO.sub.4.

10. A compound of general formula Na.sub.hLi.sub.iM′.sub.jM″.sub.kPO.sub.4, wherein: 0<h≤1, 0<i≤1, 0≤j≤1 and 0≤k≤1, and wherein M′ and M″ are each independently a metal selected from the group consisting of: Mg, Ca, Mn, Fe, Co, Ni and Cu, wherein i is not (1-h).

11. A compound of general formula NaMn.sub.1−xMg.sub.xPO.sub.4, wherein: x≈0.1.

12. A compound of general formula Na.sub.1−xMg.sub.xMnPO.sub.4, wherein: 0<x≤1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) In the appended drawings:

(2) FIG. 1: Schematic illustration of voltage versus capacity of intercalation materials for sodium ion batteries, wherein theoretical capacities and achieved ones are shown with blue bars and green ovals, respectively.

(3) FIG. 2A: Schematic view of olivine-NaFePO.sub.4; FIG. 2B: Schematic view of maricite-NaFePO.sub.4.

(4) FIG. 3: Rietveld refinement of the XRD of Na.sub.0.9Li.sub.0.1MnPO.sub.4, with Y.sub.obs (red line), Y.sub.calc (black line), Y.sub.obs-Y.sub.calc (blue line) and Bragg position (vertical blue lines).

(5) FIG. 4A and FIG. 4B: SEM images of Na.sub.0.9Li.sub.0.1MnPO.sub.4.

(6) FIG. 5: Rietveld refinement of neutron powder diffraction of Na.sub.0.966Li.sub.0.034MnPO.sub.4. Y.sub.obs (red line), Y.sub.calc (black line), Y.sub.obs-Y.sub.calc (blue line) and Bragg position (vertical blue lines), wherein λ=1.544402.

(7) FIG. 6A and FIG. 6B: Schemes of a crystal structure for Na.sub.0.966Li.sub.0.034PO.sub.4 along [010] and [001] directions, where sodium and lithium are located in the same position (blue-orange balls).

(8) FIG. 7A and FIG. 7B: SEM images of Na.sub.0.966Li.sub.0.034MnPO.sub.4.

(9) FIG. 8A and FIG. 8B: Rietveld refinements of XRD of Na.sub.0.9Li.sub.0.1Fe.sub.0.22Mn.sub.0.78PO.sub.4. FIG. 8A shows peaks that correspond to maricite; and FIG. 8B shows refinements of both olivine and maricite structure, with Y.sub.obs (red line), Y.sub.calc (black line), Y.sub.obs-Y.sub.calc (blue line) and Bragg position (vertical blue lines).

(10) FIG. 9A and FIG. 9B: SEM images of Na.sub.0.9Li.sub.0.1Fe.sub.0.22Mn.sub.0.78PO.sub.4.

(11) FIG. 10: Rietveld refinement of Na.sub.0.9Li.sub.0.1Fe.sub.0.25Mn.sub.0.75PO.sub.4.

(12) FIG. 11A and FIG. 11B: Schemes of a crystal structure for Na.sub.0.9Li.sub.0.1Fe.sub.0.25Mn.sub.0.75PO.sub.4 along [010] and [001] directions, wherein sodium and lithium are located in the same position (blue-orange balls), and Mn and Fe share the same position (green-red).

(13) FIG. 12: SEM image of Na.sub.0.9Li.sub.0.1Fe.sub.0.25Mn.sub.0.75PO.sub.4.

(14) FIG. 13: Lebail profile matching of NaMnPO.sub.4 olivine phase.

(15) FIG. 14: Rietveld refinement matching of NaMnPO.sub.4 maricite phase.

(16) FIG. 15: XRD diffraction pattern matching of Na.sub.1−xMg.sub.xMnPO.sub.4.

(17) FIG. 16: LeBail profile matching for Na.sub.1−xLi.sub.xFe.sub.yMn.sub.1−yPO.sub.4 (x≈0.1, y≈0.5), with Y.sub.obs (red line), Y.sub.calc (black line), Y.sub.obs-Y.sub.calc (blue line) and Bragg position (vertical blue lines-maricite and vertical red lines-olivine).

(18) FIG. 17: LeBail profile matching for Na.sub.1−xLi.sub.xFe.sub.yMn.sub.1−yPO.sub.4 (x≈0.1, y≈0.75), with Y.sub.obs (red line), Y.sub.calc (black line), Y.sub.obs-Y.sub.calc (blue line) and Bragg position (vertical blue lines-maricite and vertical red lines-olivine).

(19) FIG. 18: Rietveld refinement of neutron diffraction for maricite Na.sub.1−xLi.sub.xFePO.sub.4, with Y.sub.obs(red line), Y.sub.calc (black line), Y.sub.obs-Y.sub.calc (blue line) and Bragg position (vertical blue lines), wherein λ=1.544402.

(20) FIG. 19: SEM image of a sample of Na.sub.1−xLi.sub.xFePO.sub.4.

(21) FIG. 20: X-Ray diffraction patterns of NaFe.sub.0.5Mn.sub.0.5PO.sub.4 and NaFe.sub.0.75Mn.sub.0.25PO.sub.4 obtained after hydrothermal method in O.sub.2 atmosphere.

(22) FIG. 21: SEM image of mixture active material:carbon (80:20) after ball milling.

(23) FIG. 22A: Galvanostatic charge and discharge curves of Na.sub.1−xLi.sub.xMnPO.sub.4; FIG. 22B: Capacity retention of Na.sub.1−xLi.sub.xMnPO.sub.4 (from Example 1).

DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

(24) The present invention relates to novel olivine-type compounds which may be used in cathode materials for sodium-ion batteries. The compounds of the invention are sodium phospho-olivines. They are prepared by a direct synthesis which embodies a hydrothermal method. The invention is illustrated in further details by the non-limiting examples outlined below.

(25) Generally, direct synthesis of olivine Na.sub.1−xLi.sub.xFe.sub.1−yMn.sub.yPO.sub.4 was carried out by hydrothermal method. Sodium hydroxide (NaOH, Sigma Aldrich, 97%), iron sulfate (FeSO.sub.4.7H.sub.2O, Sigma Aldrich, 99%), manganese sulfate (MnSO.sub.4.H.sub.2O, Sigma Aldrich, 99%), lithium hydroxide (LiOH.H.sub.2O, Sigma Aldrich, 98%) and orthophosphoric acid (H.sub.3PO.sub.4, Alfa Aesar, 85%) were used as a precursors according to the following reaction:
3.NaOH+(1−x).FeSO.sub.4.7H.sub.2O+x.MnSO.sub.4.H.sub.2O+y.LiOH.H.sub.2O+H.sub.3PO.sub.4.fwdarw.Na.sub.1−xLi.sub.xFe.sub.1−yMn.sub.yPO.sub.4

(26) The following Table 1 outlines some detailed examples performed:

(27) TABLE-US-00001 TABLE 1 Moles of reactants used for each hydrothermal synthesis Compound NaOH FeSO.sub.4•7H.sub.2O MnSO.sub.4•H.sub.2O LiOH•H.sub.2O MgSO.sub.4 H.sub.3PO.sub.4 Ex. 1 Na.sub.0.9Li.sub.0.1MnPO.sub.4 0.36 — 0.108 0.012 — 0.12 Ex. 2 Na.sub.0.9Li.sub.0.1MnPO.sub.4 0.36 — 0.089 0.012 — 0.12 Ex. 3 Na.sub.0.9Li.sub.0.9Fe.sub.0.22Mn.sub.0.78PO.sub.4 0.36 0.03 0.078 0.012 — 0.12 Ex. 4 Na.sub.0.9Li.sub.0.1Fe.sub.0.22Mn.sub.0.78PO.sub.4 0.036 0.03 0.089 0.0012 — 0.012 Ex. 5b NaMnPO.sub.4 0.036 0.012 0.012 Ex. 5a NaMnPO.sub.4 0.36 — 0.12 0.12 Ex. 6 NaMnMgPO.sub.4 0.036 — 0.0108 — 0.0012 0.012 Ex. 7 Na.sub.0.9Li.sub.0.1Fe.sub.0.50Mn.sub.0.50PO.sub.4 0.036 0.0054 0.0054 0.0012 — 0.012 Ex. 8 Na.sub.0.9Li.sub.0.1Fe.sub.0.75Mn.sub.0.25PO.sub.4 0.036 0.0078 0.003 0.0012 — 0.012 Ex. 9 Na.sub.0.9Li.sub.0.1FePO.sub.4 0.36 0.108 — 0.012 — 0.12 Ex. 10a Na.sub.0Fe.sub.0.5Mn.sub.0.5PO.sub.4 0.036 0.006 0.006 — — 0.012 Ex. 10b Na.sub.0Fe.sub.0.75Mn.sub.0.25PO.sub.4 0.036 0.009 0.003 0.012

Example 1: Preparation of Na.SUB.1−x.Li.SUB.x.MnPO.SUB.4 .(x≈0.1, See Table 1)

(28) Excess NaOH (0.36 moles) was dissolved in 100 mL of deionized water. 0.108 moles of MnSO.sub.4.H.sub.2O and 0.012 moles of LiOH.H.sub.2O were then dissolved in 90 mL and 10 mL of deionized water, respectively. LiOH solution was added to MnSO.sub.4 solution and mixed under stirring until a homogeneous solution was obtained. 0.12 moles of H.sub.3PO.sub.4 solution were added in a graduated cylinder, which was then filled to 100 mL with deionized water. This last solution was added to the Mn—Li and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Mn—Li—P and was also stirred until a homogeneous dense solution was obtained. The pH of the final solution (total volume 300 mL) was 10. A glass liner was filled with the final solution/mixture and the autoclave was sealed under N.sub.2 atmosphere. Powders crystallized in the olivine structure were grown in the autoclave under stirring at 200° C. for 4 hours (P≈2 MPa).

(29) It should be noted that the deionized water was previously bubbled under N.sub.2 for 4-5 hours. Also, the solutions, the mixtures and the autoclave sealing were carried out in a glove box under N.sub.2 atmosphere. The obtained solution was the filtered and the product was washed with water (3-4 times) and ethanol (1 time). Subsequently, it was dried at 60-75° C. overnight.

(30) X-Ray diffraction measurement was used for the structural characterization. FIG. 3 shows the Rietveld refinement for the obtained compound, where impurities (maricite phase or unknown) are highlighted (*). After refinement, the unit cell parameters obtained were a=10.5391(6) Å, b=6.3307(4) Å and c=4.9906(3) Å, which are slightly lower than those reported for the olivine NaMnPO.sub.4 with Pnma space group (see Table 2 below). This slight difference could be attributed to the small amount of Li in the structure. Electronic microscopy was used to study the degree of homogeneity and the particle size as well. As it can be seen from FIG. 4A and FIG. 4B, two different morphologies can be observed, one of which is attributed to the olivine iron manganese phosphate (big crystals) and the other one to impurities (small crystals). Compositional studies of the final product were carried out by EDAX analysis. These measurements demonstrated that the ratio for Na:Mn:P was close to 1:1:1 (Li ions cannot be detected by EDAX measurements, Table 3 below).

(31) In this example, direct synthesis of olivine Na.sub.1−xLi.sub.xMnPO.sub.4 has been demonstrated, with the presence of some unknown impurities and impurities attributable to the maricite phase.

(32) TABLE-US-00002 TABLE 2 Refined unit cell parameters of olivine Na.sub.1−xLi.sub.xFe.sub.1−yMn.sub.yPO.sub.4 and comparison with reported values for NaFePO.sub.4 and NaMnPO.sub.4 after a Rietveld refinement a b c Ex. 1 Na.sub.0.9Li.sub.0.1MnPO.sub.4 10.5391(6) 6.3307(4) 4.9906(3) Ex. 2 Na.sub.0.9Li.sub.0.1MnPO.sub.4 10.555(1) 6.3383(9) 4.9965(7) Ex. 3 Na.sub.0.9Li.sub.0.1Fe.sub.0.22MN.sub.0.78PO.sub.4 10.516(1) 6.3060(7) 4.9795(7) Ex. 4 Na.sub.0.9Li.sub.0.1Fe.sub.0.25MN.sub.0.75PO.sub.4 10.5249(9) 6.3067(5) 4.9745(5) Ex. 5a NaMnPO.sub.4 10.5446(5) 6.3294(3) 4.9920(2) Database NaFePO.sub.4 [21] 10.4063(6) 6.2187(3) 4.9469(3) Database NaMnPO.sub.4 [29] 10.5578(3) 6.3359(2) 4.9966(2)

(33) TABLE-US-00003 TABLE 3 Summary of the composition of every compound determined by EDAX % At Compound Na Fe Mn P O Ex. Na.sub.0.9Li.sub.0.1MnPO.sub.4 17.06 — 11.33 15.91 55.70 1 Ex. Na.sub.0.9Li.sub.0.1MnPO.sub.4 16.30 — 15.13 18.53 50.04 2 Ex. Na.sub.0.9Li.sub.0.1Fe.sub.0.22Mn.sub.0.78PO.sub.4 16.00 3.62 9.86 16.72 53.80 3 Ex. Na.sub.0.9Li.sub.0.1Fe.sub.0.25Mn.sub.0.75PO.sub.4 14.65 4.75 12.64 17.52 50.42 5b Ex. Na.sub.0.9Li.sub.0.1FePO.sub.4 17.90 13.40 — 17.20 51.50 4

Example 2: Preparation of Na.SUB.1−x.Li.SUB.x.MnPO.SUB.4 .(x≈0.1, See Table 1)

(34) In this case a lower amount of MnSO.sub.4.H.sub.2O was used with respect to Example 1 (see Table 1). Excess NaOH (0.36 moles) was dissolved in 100 mL of deionized water. Then 0.089 moles of MnSO.sub.4.H.sub.2O and 0.012 moles of LiOH.H.sub.2O were dissolved in 90 mL and 10 mL of deionized water, respectively. LiOH solution was added to MnSO.sub.4 solution and mixed under stirring until a homogeneous solution was obtained. 0.12 moles of solution of H.sub.3PO.sub.4 were added in a graduated cylinder, which was then filled to 100 mL with deionized water. This last solution was added to the Mn—Li solution and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Mn—Li—P solution and was also stirred until a homogeneous dense solution was obtained. The pH of the final solution (total volume 300 mL) was 10. A glass liner was filled with the final solution/mixture and the autoclave was sealed under N.sub.2 atmosphere. Powders crystallized in the olivine structure were grown in the autoclave under stirring at 200° C. for 4 hours (P≈2 MPa).

(35) It should be noted that the deionized water was previously bubbled under N.sub.2 for 4-5 hours, and that the solutions, the mixtures and the autoclave sealing were carried out in a glove box under N.sub.2 atmosphere. However, a test was done using no bubbled water and sealing the autoclave under air, which also resulted in olivine phase.

(36) After that the obtained solution was filtered and the product was washed with water (3-4 times) and ethanol (1 time). Then, it was dried at 60-75° C. overnight.

(37) In this synthesis, a high degree of purity has been achieved. For this reason, the structural characterization was carried out using neutron powder diffraction. The aim of the use of neutron diffraction was to refine the occupancy and the site position of lithium ions. For the Rietveld refinement the olivine NaMnPO.sub.4 structure was used as a starting structural model. FIG. 5 reveals that the compound crystallized with the characteristic orthorhombic olivine phase (Pnma). The unit cell parameters determined were a=10.555(1) Å, b=6.3383(7) Å and c=4.9965(7) Å (Table 2). Thanks to the neutron diffraction, it was possible to refine the occupancies and to determine in which site lithium was located, and also the final composition of the compound. Table 4 below summarizes the atomic positions and the occupancies for each atom in the structure. Taking into account the results obtained in the Rietveld refinement, it was found that the Li is located in the sodium site with a final composition of Na.sub.0.966Li.sub.0.034MnPO.sub.4. The polyhedral representations of the final crystal structure along [010] and [001] directions are shown in FIG. 6A and FIG. 6B.

(38) TABLE-US-00004 TABLE 4 Final atomic coordinates, and occupancy of olivine Na.sub.0.966Li.sub.0.034MnPO.sub.4 Atomic position atom x y z B.sub.iso Occ. Na 0.00000 0.00000 0.0000 1.092(7) 0.483(1) Li 0.00000 0.00000 0.0000 1.092(7) 0.017(1) Mn 0.2841(1) 0.75000 0.9922(3) 0.543(3) 0.5000 P 0.1065(1) 0.75000 0.4385(3) 0.568(2) 0.5000 O1 0.1132(1) 0.75000 0.7450(2) 0.734(1) 0.5000 O2 0.4676(1) 0.75000 0.1610(2) 0.734(1) 0.5000 O3 0.17524(7) 0.9433(1) 0.3233(1) 0.734(1) 1.0000

(39) Moreover, the high degree of homogeneity can be seen from scanning electron microscopy images (FIG. 7A and FIG. 7B), where it shows that the particles have a rhombohedral-shape from 5 μm to 20 μm in length.

(40) In this example, direct synthesis of olivine Na.sub.0.966Li.sub.0.034MnPO.sub.4 has been demonstrated. In addition, the addition of the non-stoichiometric amount of MnSO.sub.4.H.sub.2O leads to a more pure compound than the use of stoichiometric amount of MnSO.sub.4.H.sub.2O. Finally, neutron diffraction allowed for determining the occupancy and the site in the crystal structure for the lithium ion. It is confirmed that Li and Na are located in the same crystallographic position and that LiOH.H.sub.2O is used in excess. However, the possibility of the presence of anitisites could not be ruled out.

Example 3: Preparation of a Na.SUB.1−x.Li.SUB.x.Fe.SUB.y.Mn.SUB.1−y.PO.SUB.4 .(x≈0.1, y≈0.22, See Table 1)

(41) Excess NaOH (0.36 moles) was dissolved in 100 mL of deionized water. Then, 0.03 moles of FeSO.sub.4.7H.sub.2O, .sub.0.78 moles of MnSO.sub.4.H.sub.2O, and 0.012 moles of LiOH.H.sub.2O were dissolved in 25 mL, 65 mL, and 10 mL of deionized water, respectively. First, FeSO.sub.4 solution was added to MnSO.sub.4 and was stirred until a homogeneous solution was obtained. Then, LiOH solution was added to Fe—Mn solution and mixed under stirring until a homogeneous solution was obtained. 0.12 moles of H.sub.3PO.sub.4 solution were added in a graduated cylinder, which was filled to 100 mL with deionized water. This last solution was added to the Fe—Mn—Li solution and it was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the solution of Fe—Mn—Li—P, which was also stirred until a homogeneous dense solution was obtained. The pH of the final solution (total volume 300 mL) was 10. A glass liner was filled with the final solution/mixture and the autoclave was sealed under N.sub.2 atmosphere. Powders crystallized in the olivine structure were grown in the autoclave under stirring at 200° C. for 4 hours (P≈2 MPa).

(42) It should be noted that the deionized water was previously bubbled under N.sub.2 for 4-5 hours. In addition, the solutions, the mixtures and the autoclave sealing were carried out in a glove box under N.sub.2 atmosphere. Afterwards, the obtained solution was filtered and the product was washed with water (3-4 times) and ethanol (1 time). Then, it was dried at 60-75° C. overnight.

(43) The structural characterization was carried out using X-Ray diffraction. The Rietveld refinement revealed that the compound crystallized with the orthorhombic olivine phase (Pnma), however, there are some peaks that correspond to maricite phase as impurities (*) (FIG. 8A). A second Rietveld refinement was carried out using both olivine and maricite structure which determined that the unit cell parameters for the olivine phase are a=10.516(1) Å, b=6.3060(7) Å and c=4.9795(7) Å (FIG. 8B and Table 2). Moreover, Rietveld refinement allows us to do a phase quantification, which determined that 75% corresponds to olivine and 25% to maricite.

(44) The scanning electronic images show the presence of two different morphologies; one of them corresponds to the olivine compound (rectangular-shaped) and the other one to the maricite as impurities (FIG. 9A and FIG. 9B). Moreover, EDAX measurements determined that the ratio for Na:Fe:Mn:P was close to 1:0.25:0.75:1. Li ions cannot be detected by EDAX detector, as in the previous example.

(45) Neutron diffraction has been performed to be able to determine the atomic coordinates, the atomic occupancy, and to elucidate the amount of lithium as well as its site in the crystal structure. Due to the presence of impurities, obtaining a stable refinement for determining the final composition presented some challenges. Nevertheless, it has been used to confirm that the structure of Na.sub.0.9Li.sub.0.1Fe.sub.0.22Mn.sub.0.78PO.sub.4 corresponds to the olivine phase.

(46) In this example, the direct synthesis of olivine Na.sub.0.9Li.sub.0.1Fe.sub.0.22Mn.sub.0.78PO.sub.4 has been demonstrated, with the presence of maricite phase as an impurity.

Example 4: Preparation of a Na.SUB.1−x.Li.SUB.x.Fe.SUB.y.Mn.SUB.1−y.PO.SUB.4 .(x≈0.1, y≈0.25, See Table 1)

(47) Excess NaOH (0.036 moles) was dissolved in 10 mL of deionized water. 0.003 moles of FeSO.sub.4.7H.sub.2O, 0.0078 moles of MnSO.sub.4.H.sub.2O, and 0.0012 moles of LiOH.H.sub.2O were then dissolved in 2.5 mL, 6.5 mL, and 10 mL of deionized water, respectively. First, FeSO.sub.4 solution was added to MnSO.sub.4 and was stirred until a homogeneous solution was obtained. Then, LiOH solution was added to Fe—Mn solution and mixed under stirring until a homogeneous solution was obtained. After that, 0.012 moles of H.sub.3PO.sub.4 solution were added in graduated cylinder, which was then filled to 10 mL with deionized water. This last solution was added to the Fe—Mn—Li solution and it was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the solution of Fe—Mn—Li—P, which was also stirred until a homogeneous dense solution was obtained. The pH of the final solution (total volume 30 mL) was 10 or higher. Powders crystallized in the olivine structure were grown in the autoclave at 200° C. for 4 hours. In order to avoid any transformation from olivine to maricite, a fast cooling (quenching) of the autoclave was carried out.

(48) After that, the obtained solution was filtered and the product was washed with water (3-4 times) and ethanol (1 time). Then, it was dried at 60-75° C. overnight. It should be noted that the use of bubbled or non-bubbled deionized water does not affect to the final product. In both cases olivine phase has been obtained. Also, in this example, solutions, mixtures and the autoclave sealing were carried out under O.sub.2 atmosphere.

(49) It has been observed that when kept under non-O.sub.2 atmosphere (more than 1 hour purge under N.sub.2 in a glovebox) the reaction is not successful. It seems that O.sub.2 is necessary to obtain the olivine phase. In addition, fast cooling (quenching) has been used to avoid the phase transformation from olivine to maricite.

(50) This methodology has also been used in Example 1 and Example 2, which, in both cases, results in pure olivine phase (not shown in this disclosure).

(51) The structural characterization was carried out using X-Ray diffraction and the Rietveld refinement confirmed that the compound crystallized with the orthorhombic olivine phase (Pnma). The use of a quenching during the cooling process helps avoid the presence of maricite impurities. However, there is a small peak (*) which is attributable to an unknown impurity (FIG. 10). In the Rietveld refinement, the occupancy of Na:Fe:Mn:P was fixed according to the ICP measurement (1:0.25:0.75:1); the unit cell parameters were determined to be a=10.5249 (9) Å, b=6.3067 (5) Å and c=4.9745 (5) Å (Table 2). The polyhedral representations of the final crystal structure along [010] and [001] directions are shown in FIG. 11A and FIG. 11B. The scanning electronic images show the presence of rectangular-shaped single particles with a width of 1.5-2 μm and from 6 to 12 μm long (FIG. 12).

(52) In this example, direct synthesis of olivine Na.sub.0.9Li.sub.0.1Fe.sub.0.25Mn.sub.0.75PO.sub.4 has been demonstrated, with an absence of maricite phase as impurity thanks to fast cooling (quenching). This reaction was carried out in O.sub.2 and using no-bubbled deionized water.

Examples 5a and 5b: Preparation of NaMnPO.SUB.4 .(x≈0.1, See Table 1)

Example 5a: Reaction Under O.SUB.2 .Atmosphere

(53) Excess NaOH (0.036 moles) was dissolved in 10 mL of deionized water. Then 0.012 moles of MnSO.sub.4.H.sub.2O were dissolved in 10 mL of deionized water. 0.012 moles of solution of H.sub.3PO.sub.4 were added in a graduated cylinder, which was then filled to 10 mL with deionized water. This last solution was added to the Mn solution and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Mn—P solution and was also stirred until a homogeneous dense solution was obtained (total volume 30 mL). A Teflon liner was filled with the final solution/mixture and the autoclave was sealed in O.sub.2 atmosphere. Powders crystallized in the olivine structure were grown in the autoclave under stirring at 200° C. for 4 hours.

(54) In this embodiment of the invention, the deionized water was not bubbled previously. Further, the solutions, the mixtures, and the autoclave sealing were carried out in air atmosphere. Afterwards, the obtained solution was filtered and the product was washed with water (3-4 times) and ethanol (1 time). Then, it was dried at 60-75° C. overnight.

(55) X-Ray diffraction was used for the phase identification and the structural characterization. FIG. 13 corresponds to a LeBail profile matching which reveals that the compound crystallized with the characteristic orthorhombic olivine phase (Pnma). The unit cell parameters were determined to be a=10.5446(5) Å, b=6.3294(3) Å and c=4.9920(2) Å (Table 2). In this case, the impurities are not attributed to maricite phase.

Example 5b: Reaction Under N.SUB.2 .Atmosphere

(56) Excess NaOH (0.36 moles) was dissolved in 100 mL of deionized water. Then 0.108 moles of MnSO.sub.4.H.sub.2O were dissolved in 100 mL. Then, 0.12 moles of H.sub.3PO.sub.4 solution were added in a graduated cylinder, which was then filled to 100 mL with deionized water. This last solution was added to the Mn-solution and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Mn—P and was also stirred until a homogeneous dense solution was obtained (total volume 300 mL). A glass liner was filled with the final solution/mixture and the autoclave was sealed under N.sub.2 atmosphere. Powders crystallized in the maricite structure were grown in the autoclave under stirring at 200° C. for 4 hours (P≈2 MPa).

(57) In this embodiment of the invention, the deionized water was previously bubbled under N.sub.2 for 4-5 hours. In addition, the solutions, the mixtures, and the autoclave sealing were carried out in a glove box under N.sub.2 atmosphere. After that, the obtained solution was filtered and the product was washed with water (3-4 times) and ethanol (1 time). Then, it was dried at 60-75° C. overnight.

(58) X-Ray diffraction was used for the phase identification and structural characterization. FIG. 14 corresponds to a Rietveld profile matching which reveals that the compound crystallized with the characteristic orthorhombic maricite phase (Pnma). However, there are some peaks that correspond to olivine phase as impurities (*). The determined unit cell parameters were a=9.0990(4) Å, b=6.9010(2) Å and c=5.1176(2) Å for the maricite phase and a=10.540(1) Å, b=6.3314(6) Å and c=4.9926(6) Å for the olivine phase. Moreover, Rietveld refinement allows for phase quantification, in which it was determined that 90% corresponds to maricite and 10% to olivine.

(59) If the reaction is carried under air atmosphere, an olivine phase is obtained, whereas maricite phase is obtained under N.sub.2 and using bubbled deionized water.

Example 6: Preparation of NaMn.SUB.1−x.Mg.SUB.x.PO.SUB.4 .(x≈0.1, See Table 1)

(60) Excess NaOH (0.036 moles) was dissolved in 10 mL of deionized water. 0.0108 moles of MnSO.sub.4.H.sub.2O and 0.0012 moles of MgSO.sub.4 were then dissolved in 9 mL and 1 mL of deionized water, respectively. First, MgSO.sub.4 solution was added to MnSO.sub.4 and was stirred until a homogeneous solution was obtained. 0.012 moles of H.sub.3PO.sub.4 solution were subsequently added to a graduated cylinder, which was then filled to 10 mL with deionized water. This last solution was added to the Mn—Mg solution and it was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the solution of Mn—Mg—P, which was also stirred until a homogeneous dense solution was obtained. The pH of the final solution (total volume 30 mL) was higher than 10. Powders crystallized in the olivine structure were grown in the autoclave at 200° C. for 4 hours. In order to avoid any transformation from olivine to maricite, a fast cooling (quenching) of the autoclave was carried out.

(61) FIG. 15 illustrates the X-Ray diffraction pattern of the compound obtained by hydrothermal method. The pattern has been compared with the reported one in Example 2 (Na.sub.1−xLi.sub.xMnPO.sub.4). The XRD reveals that the compound crystallized with the characteristic orthorhombic olivine phase (Pnma). Nonetheless, there are some peaks attributed to unknown impurities (*). The Lebail profile matching is ongoing in order to determine the unit cell parameters. However, when comparing both compositions, the peaks of Na.sub.1−xMg.sub.xMnPO.sub.4 are shifted at higher 2θ leading to a decrease of the unit cell parameters. Taking into account the higher ionic radii of Mg than Li, peaks shifted at lower 2θ were expected.

(62) In this example, the method was performed under O.sub.2 atmosphere.

Example 7: Preparation of a Na.SUB.1−x.Li.SUB.x.Fe.SUB.y.Mn.SUB.1−y.PO.SUB.4 .(x≈0.1, y≈0.5, See Table 1)

(63) Excess NaOH (0.036 moles) was dissolved in 10 mL of deionized water. Then 0.0054 moles of FeSO.sub.4.7H.sub.2O, 0.0054 moles of MnSO.sub.4.H.sub.2O, and 0.012 moles of LiOH.H.sub.2O were dissolved in 4.5 mL, 4.5 mL and 1 mL of deionized water, respectively. First, Mn-solution was added to FeSO.sub.4 solution and was mixed. Then LiOH solution was added to the Fe—Mn solution and mixed under stirring until a homogeneous solution was obtained. After that 0.012 moles of solution of H.sub.3PO.sub.4 were added in a graduated cylinder, which was then filled to 10 mL with deionized water. This last solution was added to the Fe—Mn—Li solution and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Fe—Mn—Li—P solution and it was also stirred until a homogeneous dense solution was obtained. The pH of the final solution (total volume 30 mL) was higher than 10. A Teflon liner was filled with the final solution/mixture and the autoclave was sealed under N.sub.2 atmosphere. Powders crystallized in the maricite structure were grown in the autoclave under stirring at 200° C. for 4 hours.

(64) In this embodiment of the invention, the deionized water was previously bubbled under N.sub.2 for 4-5 hours. The solutions, the mixtures, and the autoclave sealing were carried out in a glove box under N.sub.2 atmosphere. The obtained solution was then filtered and the product was washed with water (3-4 times) and ethanol (1 time). Then, it was dried at 60-75° C. overnight.

(65) X-Ray diffraction was used for the structural characterization and phase identification. FIG. 16 corresponds to a XRD LeBail profile matching of the compound obtained after hydrothermal synthesis, which reveals that the compound crystallized with the characteristic orthorhombic maricite phase (vertical blue lines). However, there are some peaks that correspond to olivine phase as impurities (FIG. 16, vertical red lines). The determined unit cell parameters were a=9.0563(5) Å, b=6.8781(5) Å and c=5.0861(3) Å for the maricite phase and a=10.522(2) Å, b=6.321(2) Å and c=4.9827(8) Å for the olivine phase (estimated composition NaFe.sub.0.25Mn.sub.0.75PO.sub.4). Due to the low amount of the olivine phase (5% or less), the phase quantification is very difficult and cannot be accurate enough.

Example 8: Preparation of a Na.SUB.1−x.Li.SUB.x.Fe.SUB.y.Mn.SUB.1−y.PO.SUB.4 .(x≈0.1, y≈0.75, See Table 1)

(66) Excess NaOH (0.036 moles) was dissolved in 10 mL of deionized water. 0.0078 moles of FeSO.sub.4.7H.sub.2O, 0.003 moles of MnSO.sub.4.H.sub.2O, and 0.012 moles of LiOH.H.sub.2O were then dissolved in 6.5 mL, 2.5 mL, and 1 mL of deionized water, respectively. First, Mn-solution was added to FeSO.sub.4 solution and was mixed. LiOH solution was then added to Fe—Mn solution and mixed under stirring until a homogeneous solution was obtained. After that, 0.012 moles of solution of H.sub.3PO.sub.4 were added in a graduated cylinder, which was then filled to 10 mL with deionized water. This last solution was added to the Fe—Mn—Li solution and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Fe—Mn—Li—P solution and it was also stirred until getting a homogeneous dense solution. The pH of the final solution (total volume 30 mL) was higher than 10. A Teflon liner was filled with the final solution/mixture and the autoclave was sealed under N.sub.2 atmosphere. Powders crystallized in the maricite structure were grown in the autoclave under stirring at 200° C. for 4 hours.

(67) In this embodiment of the invention, the deionized water was previously bubbled under N.sub.2 for 4-5 hours. In addition, the solutions, the mixtures, and the autoclave sealing were carried out in a glove box under N.sub.2 atmosphere. After that, the obtained solution was filtered and the product was washed with water (3-4 times) and ethanol (1 time). Then, it was dried at 60-75° C. overnight.

(68) X-Ray diffraction was used for the structural characterization and phase identification. FIG. 17 corresponds to a XRD LeBail profile matching of the compound obtained after hydrothermal synthesis which reveals that the compound crystallized with the characteristic orthorhombic maricite phase (FIG. 17, vertical blue lines). However, there are some peaks that correspond to olivine phase as impurities (FIG. 17, vertical red lines). The determined unit cell parameters were a=9.0267(5) Å, b=6.8692(5) Å and c=5.0678(3) Å for the maricite phase and a=10.518(1) Å, b=6.309(1) Å and c=4.9838(6) Å for the olivine phase (estimated composition NaFe.sub.0.25Mn.sub.0.75PO.sub.4). Due to the low amount of the olivine phase (5% or less), the phase quantification presented some challenges.

Example 9 (Comparative): Preparation of Na.SUB.1−x.Li.SUB.x.FePO.SUB.4 .(x≈0.1, See Table 1)

(69) Excess NaOH (0.36 moles) was dissolved in 100 mL of deionized water. Then 0.108 moles of FeSO.sub.4.7H.sub.2O and 0.012 moles of LiOH.H.sub.2O were dissolved in 90 mL and 10 mL of deionized water, respectively. LiOH solution was added to FeSO.sub.4 solution and mixed under stirring until a homogeneous solution was obtained. 0.12 moles of solution of H.sub.3PO.sub.4 were added in graduated cylinder, which was then filled to 100 mL with deionized water. This last solution was added to the Fe—Li solution and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Fe—Li—P solution and it was also stirred until a homogeneous dense solution was obtained. The pH of the final solution (total volume 300 mL) was 9-10. A glass liner was filled with the final solution/mixture and the autoclave was sealed under N.sub.2 atmosphere. Powders crystallized in the maricite structure were grown in the autoclave under stirring at 200° C. for 4 hours (P≈2 MPa).

(70) In this example, the deionized water was previously bubbled under N.sub.2 for 4-5 hours. In addition, the solutions, the mixtures, and the autoclave sealing were carried out in a glove box under N.sub.2 atmosphere. After that, the obtained solution was filtered and the product was washed with water (3-4 times) and ethanol (1 time). Then, it was dried at 60-75° C. overnight.

(71) Due to the high degree of purity of the sample, neutron diffraction was used for the structural characterization. FIG. 18 shows the Rietveld refinement for the obtained compound. In this case the compound crystallizes in the maricite Pnma structure with a=8.9953(9) Å, b=6.8610(7) Å and c=5.0455(5) Å as unit cell parameters, which are the values reported for maricite NaFePO.sub.4 [24]. Three different structural models were used in the Rietveld refinement: (1) Li is located in the Na site, (2) Li is located in the Fe site and (3) Li is located in both Na and Fe sites. Due to the low amount of Li in the structure, distinguishing among the structural models presented some challenges. A time of flight—secondary ion mass spectroscopy analysis was carried out at IREQ facilities, which showed that the amount of Li in this structure was 0.2%. Moreover, scanning electron microscopy images showed a high degree of homogeneity, where the particles had a platelet-shape making up the aggregates (FIG. 19).

(72) Despite the good Rietveld refinement of the neutron diffraction, it is difficult to determine the Li site due to its lower amount in the structure (assuming that Li ions are located in the Na site as in the previous examples).

(73) In this comparative example, direct synthesis of olivine Na.sub.1−xLi.sub.xFePO.sub.4 has not been demonstrated, as maricite Na.sub.1−xLi.sub.xFePO.sub.4 has been obtained instead.

Examples 10a and 10b (Comparative): Preparation of a NaFe.SUB.y.Mn.SUB.1−y.PO.SUB.4 .(y≈0.5 or y≈0.75, See Table 1)

Example 10a: NaFe.SUB.0.5.Mn.SUB.0.5.PO.SUB.4

(74) Excess NaOH (0.036 moles) was dissolved in 10 mL of deionized water. Then 0.006 moles of FeSO.sub.4.7H.sub.2O, 0.006 moles of MnSO.sub.4.H.sub.2O were dissolved in 5 mL and 5 mL of deionized water, respectively. Firstly, Mn-solution was added to FeSO.sub.4 solution and it was mixed until homogeneous solution was obtained. After that 0.012 moles of solution of H.sub.3PO.sub.4 were added to a graduated cylinder, which was then filled to 10 mL with deionized water. This last solution was added to the Fe—Mn solution and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Fe—Mn—P solution and it was also stirred until getting a homogeneous dense solution. The pH of the final solution (total volume 30 mL) was higher than 10. A Teflon liner was filled with the final solution/mixture and the autoclave was sealed under O.sub.2 atmosphere. Powders crystallized grown in the autoclave under stirring at 200° C. for 4 hours. After that a fast cooling (quenching) of the autoclave was carried out.

Example 10b: NaFe.SUB.0.75.Mn.SUB.0.25.PO.SUB.4

(75) Excess NaOH (0.036 moles) was dissolved in 10 mL of deionized water. Then 0.009 moles of FeSO.sub.4.7H.sub.2O and 0.003 moles of MnSO.sub.4.H.sub.2O were dissolved in 7.5 mL and 2.5 mL of deionized water, respectively. First, Mn-solution was added to FeSO.sub.4 solution and it was mixed until a homogeneous solution was obtained. After that, 0.012 moles of solution of H.sub.3PO.sub.4 were added to a graduated cylinder, which was then filled to 10 mL with deionized water. This last solution was added to the Fe—Mn solution and was stirred until a homogeneous solution was obtained. Finally, NaOH solution was added to the Fe—Mn—P solution and it was also stirred until getting a homogeneous dense solution. The pH of the final solution (total volume 30 mL) was higher than 10. A Teflon liner was filled with the final solution/mixture and the autoclave was sealed under O.sub.2 atmosphere. Powders crystallized grown in the autoclave under stirring at 200° C. for 4 hours. After that a fast cooling (quenching) of the autoclave was carried out.

(76) X-Ray diffraction was used for the structural characterization and phase identification. FIG. 20 shows the X-Ray diffraction patterns for compounds NaFe.sub.0.5Mn.sub.0.5PO.sub.4 (Example 10a, red pattern) and NaFe.sub.0.75Mn.sub.0.25PO.sub.4 (Example 10b, green pattern). Also, the patterns have been compared with the characteristic maricite and olivine NaFePO.sub.4 database. As mentioned above, this synthesis was used to explore the effect of Li-absence and the O.sub.2 atmosphere when the Fe content is higher than 0.25. Both compounds do not crystallize in the characteristic olivine structure. Mainly, NaFe.sub.0.75Mn.sub.0.25PO.sub.4 crystallizes in the maricite phase, however there are some peaks attributed to unknown impurities (*). Whereas the NaFe.sub.0.5Mn.sub.0.25PO.sub.4 (Example 10b, green pattern) is more complex, this compound does not crystallize in the maricite phase or the olivine phase.

(77) This comparative example explored the effect of non Li-doping and O.sub.2 atmosphere. Direct synthesis of olivine NaFe.sub.xMn.sub.1−xPO.sub.4 (x≈0.5 and x≈0.75) was not been achieved. Rather, maricite NaFe.sub.0.75Mn.sub.0.5PO.sub.4 has been obtained instead, as the higher amount of Fe probably induces the maricite phase.

Electrochemical Test

(78) First test: preliminary tests were conducted using active material power (Example 1) mixed with Denka carbon with a ratio of 80:20 (active material: carbon) using ball milling. Ball milling was used to decrease the particle size of the prepared material and also to ensure good homogeneity of the mixture. FIG. 21 shows a SEM image after ball milling treatment; a good homogeneity of the sample but also the presence of agglomerates of small particles (mainly, less than 500 nm) can be observed.

(79) Coin cell was assembled using a sodium metal as anode, Watman GF/D borosilicate glass fiber as a separator, and 1M NaPF.sub.6 Ethylene carbonate:Dimethyl carbonate (EC:DMC) as an electrolyte, as well as using a MACCOR battery test system.

(80) Na.sub.1−xLi.sub.xMnPO.sub.4 (theoretical capacity of 145 mA.Math.h/g) was tested versus lithium and sodium metal using galvanostatic charge and discharge curves at C/20 (FIG. 22A and FIG. 22B). While the material is electrochemically active for both ions, both curves exhibit low capacity, far from the theoretical one, as well as a poor cyclability and high polarization.

(81) Second test: a challenge with these new olivine compounds is to find a methodology for the carbon coating process. Temperatures over 400° C. are probably to be avoided, since a phase transition from olivine to maricite may occur. A composite may be prepared by mixing Na.sub.0.966Li.sub.0.034MnPO.sub.4 (Example 2) with C65 using a ball milling. Then the mixture may be thermally treated under Argon at 400° C. for 3 hours. The compound obtained may be tested versus metallic sodium and compared with the results obtained in the first test.

(82) The scope of the claims should not be limited by the preferred embodiments set forth in the examples, but should be given the broadest interpretation consistent with the description as a whole.

(83) The present description refers to a number of documents, the content of which is herein incorporated by reference in their entirety.

REFERENCES

(84) [1] R. A. Huggins. Advanced Batteries: Materials Science Aspects. (Springer, 2009). [2] J. M. Tarascon and M. Armand, Nature 2001, 414, 359. [3] M. Armand and J. M. Tarascon, Nature 2008, 451, 652. [4] M. S. Whittingham, Chem. Rev. 2004, 104, 4271. [5] J. B. Goodenough, J. Power Sources 2007, 174, 996. [6] J. W. Fergus, J. Power Sources 2010, 195, 939. [7] A. Padhi, K. S. Nanjuindaswamy and J. B. Goodenough, J. Electrochem. Soc. 1997, 144, 1188. [8] B. L. Ellis, K. T. Lee and L. F. Nazar, Chem. Mater. 2010, 22, 691. [9] H. Hyand, S. C. Yin and L. F. Nazar, Electrochem. Solid State Lett. 2001, A, A170-A171. [10] K. Zaghib, A. Guerfi, P. Hovington, A. Vijh, M. Trudeau, A. Mauger, J. B. Goodenough and C. M. Julien, Journal of Power Sources 2013, 232, 357-369. [11] J.-M. Tarascon, Nat. Chem. 2010, 2, 510. [12] C. Wadia, P. Albertus and V. J. Srinivasan, Power Sources 2011, 196, 1593. [13] V. Palomares, P. Serras, I. Villaluenga, K. B. Hueso, J. Carretero-Gonzalez and T. Rojo, Energy Environ. Sci. 2012, 5, 5884. [14] M. D. Slater, D. Kim, E. Lee and C. S. Johnson, Adv. Funct. Mater. 2013, 23, 947. [15] S. Komaba, W. Murata, T. Ishikawa, N. Yabuuchi, T. Ozeki, T. Nakayama, A. Ogata, K. Gotoh and K. Fujiwara, Adv. Funct. Mater. 2011, 21, 3859. [16] V. Palomares, M. Casas-Cabanas, E. Castillo-Martinez, M. H. Han and T. Rojo, Energy Environ. Sci. 2013, 6, 2312. [17] S. W. Kim, D. H. Seo, X. H. Ma, G. Ceder and K. Kang, Adv. Energy Mater. 2012, 2, 710. [18] G. H. Newman and L. P. Klemann, J. Electrochem. Soc. 1980, 197, 2097. [19] Shyue Ping Ong, Vincent L. Chevrier, Geoffroy Hautier, Anubhav Jain, Charles Moore, Sangtae Kim, Xiaohua Ma and Gerbrand Ceder, Energy Environ. Sci. 2011, 4, 3680-3688. [20] Y. Le Page and G. Donnay, Can. Miner. 1977, 15, 518-521. [21] P. Moreau, D. Guyomard, J. Gaubicher and F. Boucher, Chemistry of Materials 2010, 22, 4126-4128. [22] A. Sun, F. R. Beck, D. Haynes, J. A. Poston, S. R. Narayanan, P. N. Kumta and A. Manivannan, Mater. Sci. Eng., B 2012, 177, 1729-1733. [23] J. Kim, D.-H. Seo, H. Kim, I. Park, J.-K. Yoo, S.-K. Jung, Y.-U. Park, W. A. Goddard III and K. Kang, Energy Environ. Sci. 2015, 8, 540-545. [24] K. Zaghib, J. Trottier, P. Hovington, F. Brochu, A. Guerfi, A. Mauger, C. M. Julien, J. Power Sources 2011, 196, 9612-9617. [25] Pier Paolo Prosini Cinzia Cento a, Amedeo Masci and Maria Carewska, Solid State Ionics 2014, 263, 1-8. [26] M. Casas-Cabanas, V. V. Roddatis, D. Saurel, P. Kubiak, J. Carretero-Gonzalez, V. Palomares, P. Serras and T. Rojo, J. Mater. Chem. 2012, 22(34), 17421. [27] Montserrat Galceran, Damien Saurel, Begona Acebedo, Vladimir V. Roddatis, Egoitz Martin, Teófilo Rojo and Montse Casas-Cabanas, Phys. Chem. Chem. Phys. 2014, 16, 8837-8842. [28] Violeta Koleva, Tanya Boyadzhieva, Ekaterina Zhecheva, Diana Nihtianova, Svetlana Simova, Georgi Tyuliev and Radostina Stoyanova, CrystEngComm, 2013, 15, 9080-9089. [29] K. T. Lee, T. N. Ramesh, F. Nan, G. Botton and L. F. Nazar, Chem. Mater., 2011, 23, 3593-3600. [30] Chun Li, Xue Miao, Wei Chu, Ping Wu and Dong Ge Tong, J. Mater. Chem. A 2015, 00, 1-6.