Subscriber access provided by University of Sussex Library
Article 2-x
x
4
3
Modeling Ti/Ge distribution in LiTi Ge(PO) NASICON series by P MAS NMR and first principles DFT calculations 31
Virginia Diez-Gómez, Kamel Arbi, and Jesús Sanz J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b03583 • Publication Date (Web): 02 Jul 2016 Downloaded from http://pubs.acs.org on July 4, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 10
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of the American Chemical Society
Modeling Ti/Ge distribution in LiTi2-xGex(PO4)3 NASICON series by 31P MAS NMR and first principles DFT calculations. Virginia Diez-Gómez*, Kamel Arbi†, Jesús Sanz. Instituto de Ciencia de Materiales de Madrid, Consejo Superior de Investigaciones Científicas (ICMM-CSIC). Sor Juana Inés de la Cruz, 3. 28049 Madrid. Spain. KEYWORDS 31P NMR, DFT calculation, NASICON, disorder.
ABSTRACT: Ti/Ge distribution in rhombohedral LiTi2-xGex(PO4)3 NASICON series has been analyzed by 31P Magic-Angle Spinning Nuclear Magnetic Resonance (MAS-NMR) spectroscopy and first principles Density Functional Theory (DFT) calculations. Nuclear Magnetic Resonance results an excellent probe to follow Ti/Ge disorder, as it is sensitive to the atomic scale environment without long-range periodicity requirements. In samples considered here, PO4 units are surrounded by four Ti/Ge octahedra, then, five different components ascribed to P(OTi) 4, P(OTi)3(OGe), P(OTi)2(OGe)2, P(OTi)(OGe)3 and P(OGe)4 environments are expected in 31P MAS-NMR spectra of R͞3c NASICON samples. However, 31P MAS-NMR spectra of analyzed series display a higher number of signals, suggesting that although overall symmetry remains R͞3c, partial substitution causes a local decrement in symmetry. With the aid of first principles DFT calculations, ten detected 31P NMR signals have been assigned to different Ti4-nGen arrangements in the R3 subgroup symmetry. In this assignment, the influence of octahedra of the same or different R2(PO4)3 structural units has been considered. The influence of bond distances, angles and atom charges on 31P NMR chemical shieldings has been discussed. Simulation of the LiTi2-xGex (PO4)3 series suggests that detection of ten P environments is mainly due to the existence of two oxygen types, O1 and O2, whose charges are differently affected by Ge and Ti occupation of octahedra. From the quantitative analysis of detected components, a random Ti/Ge distribution has been deduced in next nearest neighbor (NNN) sites that surround tetrahedral PO4 units. This random distribution was supported by XRD data displaying Vegard’s law.
Introduction In the past decades, lithium-ion conducting solids have been investigated for potential applications as electrolytes in all-solid state lithium batteries1-4. For that, high lithium mobility and electrochemical stability are required. Most of lithium ion conductors, such as Li3N, Li- alumina, lithium lanthanum titanates (Li3xLa(2/3)−xTiO3), LISICON (Li2+2xZn1−xGeO4), LIPON (xLi2O–yP2O5–zPON), lithium sulphide glasses (Li2S–P2O5–LiI and Li2S–SiS2–Li3PO4) have either high ionic conductivity or high electrochemical stability, but rarely both4-10. In the case of LiR2(PO4)3, Li NASICON compounds have received much attention because their high ionic conductivity and stability against moisture.3,9,11-13 NASICON LiR2(PO4)3 framework is built up by R2(PO4)3 units, where two RO6 octahedra share oxygens with three PO4 tetrahedra. Tetrahedra and octahedra of contiguous R2(PO4)3 units are bounded to form the three-dimensional network of NASICON structure (Figure 1). In this network, each octahedron is connected to six tetrahedra and each tetrahedron to four octahedra. Symmetry of LiR2(PO4)3 compounds is usually rhombohedral R͞3c, although in some cases a triclinic distortion (space group C1̅) was detected14-18. In rhombohedral samples, Li+ ions occupy M1 sites surrounded by six oxygen atoms14,15; but in
triclinic phases, Li+ ions are at intermediate sites M12, coordinated to four oxygen atoms between M1 and M2 positions.16-18 In sodium NaR2(PO4)3 homologues, bigger Na cations only occupy M1 sites at ternary axes.19 In Li1+xRIV2-xSxIII(PO4)3 series, with R= Ti, Zr, Hf, Ge, Sn and S= Al, Ga, In, the increment detected in Li conductivity has been ascribed to the occupation of M3 (located in M2 cavities) at expenses of M1 sites.20-24
Figure 1. Detail of NASICON structure with R octahedra denoted in grey, P tetrahedra in orange, O atoms in red and Li atoms in purple
ACS Paragon Plus Environment
Journal of the American Chemical Society
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
In order to understand physical and chemical properties of NASICON compounds, a detailed knowledge of the local structure and cation order/disorder is required. Solid-State Nuclear Magnetic Resonance (SS-NMR) can be used to study local atomic arrangements without longrange order. Subtle differences in local environment produce changes on chemical shielding, achieving, in some cases, enough resolution to differentiate MAS-NMR components. In the case of NASICON LiR2(PO4)3 materials, each P tetrahedra is surrounded by four octahedra, which permits to follow the main features of octahedral cation distribution around P atoms. Density Functional Theory (DFT) calculations can be used to analyze the influence of next nearest neighbor (NNN) sites on the chemical shift of atoms.25 The comparison between experimental NMR and DFT-calculated parameters has previously allowed the assessment of atomic arrangements, where diffraction data were not conclusive.26,27 Optimization of structural parameters makes possible a better analysis of electronic arrangement in chemical bonds.28,29 In particular, the influence of bond angles and lengths on 31P NMR chemical shielding has been previously tested with DFT techniques.30-32 The aim of the present work is the analysis of Ti/Ge distribution in the LiTi2-xGex(PO4)3 series using 31P MASNMR spectroscopy. For a better understanding of deduced parameters, a parallel NMR and DFT study of this series has been performed and the assignment of experimental signals to specific environments done. With the aid of DFT, the sensitivity of NMR chemical shift to atoms charges, bond distances and bond angles has been investigated. The quantitative analysis of 31P NMR spectra has been used to deduce main characteristics of the Ti/Ge distribution. Experimental Samples LiTi2-xGex(PO4)3 samples, with 0 ≤ x ≤ 2, were prepared by the ceramic route. Stoichiometric amounts of Li2CO3, TiO2/GeO2 and (NH4)2HPO4 were first dried at 393 K for 10h and then mixed in a platinum crucible and heated at increasing temperatures as described elsewhere21,33. After each treatment, the mixture was ground in an agate mortar and analyzed by X-ray diffraction to assess the formation of single phases of prepared compounds. Samples were stored for later use when characteristic peaks of NASICON phases were detected and those of reagents or intermediate pyrophosphates eliminated from XRD patterns. Techniques XRD powder patterns were recorded with the Cu-K radiation (=1.540598 Å) in the 10-70º 2Ө range with a step size of 0.02º and a counting time of 0.1s·step-1 in a PW 1050/25 Phillips diffractometer. Unit cell parameters of formed phases were deduced by using the Fullprof program34 (pattern matching technique). 31P NMR spectra were recorded at room temperature in an AVANCE Bruker spectrometer (9.4 Tesla). The fre-
Page 2 of 10
quency used was 161.97 MHz. NMR signals were obtained after /2 pulse irradiation (3.5 s) with a recycling time of 600 s (single pulse experiments). The number of scans was 16. To improve the experimental resolution, samples were spun at 10 kHz around an axis inclined 54º44’ with respect to the external magnetic field (MAS technique) during spectra recording. 31P chemical shifts were referred to an 85% H3PO4 aqueous solutions. The fitting of NMR spectra was performed with the DMFIT software package.35 DFT simulations Calculations were carried out using the CASTEP 7.0 code36. In this work, Density Functional Theory (DFT), which employs the gauge including projector augmented wave (GIPAW)37 algorithm, enabled the reconstruction of all-electron wave functions in presence of a magnetic field. The Generalized Gradient Approximation (GGA) PBE38 functional was used and the core-valence interactions were described with ultrasoft reported pseudopotentials.39 Structural parameters (unit cell and atomic positions) reported in ICSD 9597940 (for Ti) and 6976314 (for Ge) were used in DFT calculations. Crystal structures were reproduced by using periodic boundary conditions. For geometrical optimization and NMR parameter calculation, numerical integrals were performed over the Brillouin zone, using a Monkhorst– Pack grid with a k-point spacing of 0.04 Å-1. Wavefunctions were expanded in plane waves with kinetic energy smaller than the cut-off energy, 1200 eV. With the chosen kinetic energy cutoff and k-point grid, the total energy converged to
