Article pubs.acs.org/IC
Enabling the Electrochemical Activity in Sodium Iron Metaphosphate [NaFe(PO3)3] Sodium Battery Insertion Material: Structural and Electrochemical Insights Ritambhara Gond,† Sher Singh Meena,‡ S. M. Yusuf,‡ Vivekanand Shukla,§ Naresh K. Jena,§ Rajeev Ahuja,§ Shigeto Okada,⊥ and Prabeer Barpanda*,† †
Faraday Materials Laboratory, Materials Research Center, Indian Institute of Science, C.V. Raman Avenue, Bangalore 560012, India Solid State Physics Division, Bhabha Atomic Research Centre, Mumbai 400085, India § Condensed Matter Theory Group, Materials Theory Division, Department of Physics and Astronomy, Uppsala University, Box 516, SE-751 20, Uppsala, Sweden ⊥ Institute for Materials Chemistry and Engineering (IMCE), Kyushu University, 6-1 Kasuga-Koen, Kasuga, Fukuoka 816-8580, Japan ‡
S Supporting Information *
ABSTRACT: Sodium-ion batteries are widely pursued as an economic alternative to lithium-ion battery technology, where Fe- and Mn-based compounds are particularly attractive owing to their elemental abundance. Pursuing phosphate-based polyanionic chemistry, recently solid-state prepared NaFe(PO3)3 metaphosphate was unveiled as a novel potential sodium insertion material, although it was found to be electrochemically inactive. In the current work, employing energy-savvy solution combustion synthesis, NaFe2+(PO3)3 was produced from low-cost Fe3+ precursors. Owing to the formation of nanoscale carbon-coated product, electrochemical activity was enabled in NaFe(PO3)3 for the first time. In congruence with the first principles density functional theory (DFT) calculations, an Fe3+/Fe2+ redox activity centered at 2.8 V (vs Na/Na+) was observed. Further, the solid-solution metaphosphate family Na(Fe1−xMnx)(PO3)3 (x = 0−1) was prepared for the first time. Their structure and distribution of transition metals (TM = Fe/Mn) was analyzed with synchrotron diffraction, X-ray photoelectron spectroscopy, and Mössbauer spectroscopy. Synergizing experimental and computational tools, NaFe(PO3)3 metaphosphate is presented as an electrochemically active sodium insertion host material.
■
INTRODUCTION
already created a splash among existing classical oxide cathode materials. Among the PO4-based (SIB) polyanionic insertion materials, special attention is geared toward Fe-containing ones owing to their elemental abundance.5 Among all Na−Fe−P−O quaternary systems, simple phosphates [(PO4)3−], pyrophosphates [(P2O7)4−], and mixed phosphates [(PO4)2(P2O7)10−] have been recently reported as economically viable sodium insertion hosts. Robust electrochemical activity involving Fe3+/Fe2+ redox potentials in the 2.5−3.5 V window has been observed in maricite NaFePO4, Na2FeP2O7, and Na4Fe3(PO4)2P2O7.6−11 These polyanionic cathodes typically demonstrate chemical/ thermal stability and superior capacity retention to layered oxides, due to their robust and stable frameworks with tightly bound polyanions. In addition, the voltage-composition profiles are generally simpler than those of layered oxides, with distinct single-phase or two-phase behavior having a well-defined phase boundary.12 These Na−Fe−P−O quaternary cathodes involve
Owing to their high volumetric and gravimetric energy density, Li-ion batteries (LIBs) have unparalleled ubiquity in current mobile technologies encompassing myriads of portable electronics to (plug-in) hybrid electric vehicles. However, when the ball rolls to the large-scale production court such as (stationary) grid storage, cost instead of energy density becomes the overriding factor, thereby paving way for the of illation to Na-ion battery (SIB) technology. Sodium, being the fifth most abundant element with a uniform geographic distribution, can lead to an economic battery technology visà-vis lithium batteries.1−3 Although Li- and Na-based batteries have similar underlying operating principles, owing to the wide structural diversity between Li- and Na-based insertion host materials, at times it is difficult to design SIB cathode materials by simply mimicking LIB cathode materials. Therefore, significant efforts have been geared toward the discovery of novel insertion materials and reaction mechanisms to realize viable cathodes for advanced SIB technology.4 In search of SIB cathode materials, the phosphate or diphosphate class of cathode materials containing Li/Na and transition metals has © 2017 American Chemical Society
Received: March 4, 2017 Published: May 2, 2017 5918
DOI: 10.1021/acs.inorgchem.7b00561 Inorg. Chem. 2017, 56, 5918−5929
Article
Inorganic Chemistry
NaH 2PO4 ·H 2O + 1−x Fe3 +(NO3)3 ·9H 2O + x Mn(CH3COO)2 ·4H 2O
a minimal volume change (2−4%) upon cycling, making them suitable for long-term operation. This value compares favorably with those of other Na-ion battery materials, such as NaFePO4 and NaFeSO4F, which exhibit huge volume change of 15% and 14.5%, respectively.6 Pursuing the rich Na−Fe−P−O chemistry, recently metaphosphate NaFe(PO3)3 has been synthesized by a classical solid-state route by Lin et al. as a potential sodium insertion material.13 Involving (PO3)− chains, it has a complex structure having large tunnels for Na migration. Nevertheless, large Na− Na hopping distances and poor electronic conductivity led to electrochemical inactivity in this metaphosphate despite having an open framework.13 To solve this issue, we have investigated this novel NaFe(PO3)3 metaphosphate. Using soft chemical combustion synthesis involving economic Fe3+ precursors14,15 and careful cathode optimization, we have observed the Fe3+/ Fe2+ electrochemical activity in NaFe(PO3)3 for the first time. The crystal structure and electrochemical properties of isostructural NaMn(PO3)3 have been investigated synergizing experimental and computational tools. This work describes the structural aspects of a novel Na(Fe1−xMnx)(PO3)3 (x = 0, 0.25, 0.50, 0.75, 1) solid-solution family by different characterization tools. First-principles density functional theory (DFT) calculation along with galvanostatic measurement has been used to probe the redox voltage, sodium diffusion path, and possible hindrance to electrochemical activity in NaFe(PO3)3 metaphosphate.
■
+ NH4H 2PO4 + CH4N2O → Na(Fe2 +1 − xMn 2 + x)(PO3)3 + NOx ↑ +NH3 ↑ +CO2 ↑ + H 2O↑
Material Characterization. Powder X-ray diffraction measurements of the as-synthesized Fe−Mn solid solution series were collected at room temperature with a PANalytical X’Pert Pro diffractometer equipped with a Cu Kα target of monochromatic wavelength (λ = 1.5404 Å) operating at 40 kV/30 mA. Typical scans were made in the 2θ range of 10−90° with a scanning step of 0.02626° in Bragg−Brentano geometry. Rietveld refinement was performed using the GSAS program with the EXPGUI front-end17−19 for Na(Fe1−xMnx)(PO3)3 (x = 0, 0.25, 0.50, 0.75, 1). Synchrotron measurements were performed at the BL-18B Indian Beamline at the High Energy Accelerator Research Organization (KEK-Photon Factory, Tsukuba, Japan) using a Mo Kα target of monochromatic wavelength (λ1 = 0.786 Å) to study phase purity and peak shifting in the mixed-metal metaphosphates. The energy of the beamline (E = 15.77 keV) was calibrated with a Si (640b NIST) standard. The crystal structure was illustrated using VESTA software.20 Field emission scanning electron microscopy (FE-SEM, FEI Inspect F 50 at 10 kV) was used to compare the size, morphology, and elemental distribution mapping of the final powder products. The elemental compositions in the metaphosphate end-products were qualitatively analyzed by energy dispersive X-ray analysis (EDX) and were subsequently quantified by using a PerkinElmer Optima 8300 inductively coupled plasma (ICP) atomic emission spectrometer. The selected area diffraction pattern (SAED) of the NaFe(PO3)3 powder was collected by an FEI Tecnai F 30 STwin TEM (200 kV) microscope. FT-IR spectra of powder samples (diluted in KBr pellets) were acquired by an Agilent Cary 600 Series FTIR spectrometer in the wavenumber range of 400−4000 cm−1 (cycle number = 4). The Mössbauer spectra of the samples were recorded at room temperature with a conventional spectrometer operated in constant-acceleration mode in transmission geometry with a 57Co source in the Rh matrix of 1 mCi. The recorded Mössbauer spectra were fitted using the WinNormos site fit program. The calibration of the velocity scale was done using an enriched α-57Fe metal foil. The isomer shift values are relative to the Fe metal foil (δ = 0.0 mm s−1). Chemical State Analysis. In order to identify the chemical states of atoms in as-prepared Na(Fe1−xMnx)(PO3)3 metaphosphate powders, XPS analysis was carried out for NaFe(PO3)3, Na(Fe0.5Mn0.5)(PO3)3, and NaMn(PO3)3 by a Kratos Axis Ultra DLD with an incident monochromated X-ray beam from the Al target. XPS data were collected at an accelerating voltage of 13 kV and emission current of 9 mA. Shift corrections were done by considering carbon as reference at 284.6 eV binding energy for all the samples. Computational Investigation. The DFT-based calculations were performed considering spin-polarization with the Perdew−Burke− Ernzerhof exchange−correlation functional21,22 as implemented in the Vienna Ab-initio Simulation Package (VASP).23 The generalized gradient approximation (GGA) with Hubbard U approximation (GGA+U) was used for the calculations.24 The (U−J) parameters for Fe and Mn were chosen to be 4 and 3.9, respectively, in accordance
EXPERIMENTAL SECTION
Material Synthesis. NaFe(PO3)3 was synthesized by a conventional solid-state (dry) method as well as a solution combustion (wet) method. The conventional solid-state route involved intimate mixing of NaH2PO4·H2O (Merck, ≥98%), FeC2O4·2H2O (Aldrich, ≥99%), and NH4H2PO4 (Merck, ≥99%) in a molar ratio of 1:1:2, respectively. These precursors were mixed by (dry) planetary milling for 3 h at 350 rpm using Cr-hardened stainless steel milling media and container to get a precursor mixture with reduced particle size, facilitating better reactivity. The resulting powder was annealed inside a tubular furnace under an Ar atmosphere at 600 °C for 6 h in an alumina crucible to get the final divalent iron product (white powder). Along with mechanical ball milling, NaFe(PO3)3 was also prepared by hand mixing the precursors for ∼1 h. Then, a mixed metal Na(Fe1−xMnx)(PO3)3 (x = 0−1) metaphosphate series were prepared by solution combustion synthesis. To synthesize these metaphosphate solid solutions, stoichiometric amounts of NaH2PO4·H2O (Merck, ≥98%), Fe(NO3)3·9H2O (Merck, ≥98%), Mn(CH3COO)2·4H2O (Aldrich, ≥99%), and NH4H2PO4 (SDFine Chemicals, ≥99%) were used. A low-cost Fe3+ precursor (i.e., Fe(NO3)3·9H2O) was dissolved in distilled water and was converted into the Fe2+ state in solution by addition of a pinch of ascorbic acid (C6H8O6).9 This change in oxidation state can be inferred from the instantaneous change in color from brown (Fe3+) to colorless (Fe2+). Then, the remaining precursors were added into the same solution. Finally 1 mol of urea (fuel) was added to the Na:Fe:Mn:P mixture to propel the combustion reaction.16 This reaction mixture was heated at 120 °C on a hot plate with continuous stirring. Once the excess water evaporated, the temperature was increased to 400 °C to initiate combustion, which involves an exothermic chemical reaction involving gaseous evolution, leaving behind a dry black powder, also called combustion ash. This amorphous intermediate residue was later annealed at 600 °C for 6 h in Ar flow. It led to the one-step formation of nanostructured carboncoated Na(Fe 1−x Mn x )(PO 3 ) 3 (x = 0, 0.25, 0.50, 0.75, 1) metaphosphate end-members. The occurrence of carbon coating stems from the decomposition of carbonaceous precursors (such as ascorbic acid and urea) in the combustion synthesis. The final combustion reaction can be written as 5919
DOI: 10.1021/acs.inorgchem.7b00561 Inorg. Chem. 2017, 56, 5918−5929
Article
Inorganic Chemistry
Figure 1. NaFe(PO3)3 structure projected onto the b−c plane, showing interconnected FeO6 octahedra (green) and PO4 tetrahedra (blue). Crystallographically independent sites of (left) Fe (Fe1 to Fe4) and (right) P (P1 to P12) are illustrated.
Figure 2. (a) Representative metaphosphate framework with orthorhombic structure (space group: P212121) built from FeO6 octahedra (green), PO4 tetrahedra (blue), and Na toms (purple). High-resolution XRD profiles and pattern indexing of (b) NaFe(PO3)3 (Rp = 2.06, Rwp = 2.74, χ2 = 2.47), (c) NaFe0.5Mn0.5(PO3)3 (Rp = 2.17, Rwp = 2.86, χ2 = 2.40), and (d) NaMn(PO3)3 (Rp = 2.44, Rwp = 3.48, χ2 = 4.00). In each case, the experimental data points (red), simulated pattern (black), their difference (blue), and Bragg diffraction peaks (green) are presented. with some previous reports.25 In the initialization of the calculations, magnetic moments for Fe/Mn were set to high spin (MAGMOM = 7). The plane wave cutoff was set to 700 eV. In the current work, 2 × 2 × 2 k-points for relaxation of structure and 4 × 4 × 4 k-points for density of states (DOS) calculation were employed. The open-circuit voltage (OCV) was computed using the expression V = |E(NaM(PO3)3) − E(Na) − E(M(PO3)3)| (where M = Fe or Mn). The activation barrier for Na diffusion was calculated with the climbing nudged elastic band method with the VASP- transition state tool compilation. Sodium Intercalation Analysis. The electrochemical sodium (de)intercalation properties of NaFe(PO3)3 were examined with the
coin-type (CR2032) half-cell. In order to make the working electrode (cathode), the as-synthesized active material, acetylene black, and polyvinylidene fluoride were taken in a weight ratio of 80:15:5. They were ground to fine particles and were blended with a milling machine. To make a slurry, a few drops of N-methyl-2-pyrrolidone (NMP) solvent was added to this mixture before milling. This mixture was cast on an Al foil and dried at 120 °C overnight to remove excess NMP. Circular disks (diameter = 10 mm) were punched out inside an MBraun Labstar glovebox maintaining argon ambience (H2O/O2 level
