SUPPLEMENTARY INFORMATION

Transcription

1 In the format provided by the authors and unedited. DOI: /NMAT4976 Reversible magnesium and aluminium ions insertion in cation-deficient anatase TiO 2 Toshinari Koketsu 1, Jiwei Ma 2,3, *, Benjamin J. Morgan 4, Monique Body 5, Christophe Legein 5, Walid Dachraoui 3,6, Mattia Giannini 6,7,8, Arnaud Demortière 3,6,7, Mathieu Salanne 2,3, François Dardoize 2, Henri Groult 2, Olaf J. Borkiewicz 9, Karena W. Chapman 9, Peter Strasser 1,*, and Damien Dambournet 2,3,* 1 The Electrochemical Energy, Catalysis, and Materials Science Laboratory, Department of Chemistry, Chemical Engineering Division, Technical University Berlin, Berlin, Germany. 2 Sorbonne Universités, UPMC Univ Paris 06, CNRS UMR 8234, Laboratoire PHENIX, 4 place Jussieu, F Paris, France. 3 Réseau sur le Stockage Electrochimique de l'energie (RS2E), FR CNRS 3459, France. 4 Department of Chemistry, University of Bath, BA2 7AY Bath, United Kingdom. 5 Universite Bretagne Loire, Universite du Maine, UMR CNRS 6283, Institut des Mole cules et des Mate riaux du Mans (IMMM), Avenue Olivier Messiaen, Le Mans Cedex 9, France. 6 Laboratoire de Réactivité et Chimie des Solides, CNRS UMR 7314, Université de Picardie Jules Verne, 33 rue Saint Leu, Amiens Cedex, France. 7 ALISTORE-European Research Institute, FR CNRS 3104, Amiens, France. 8 Thermo Fisher Scientific, Materials and Structural Analysis, Achtseweg Noord 5, Eindhoven 5651 GG, the Netherlands. 9 X-ray Science Division, Advanced Photon Source, Argonne National Laboratory, 9700 South Cass Avenue, Argonne, Illinois 60439, United States. These authors contributed equally to this work. * To whom correspondence should be addressed. address: jiwei.ma@upmc.fr (J.M.), pstrasser@tu-berlin.de (P.S.), damien.dambournet@upmc.fr (D.D.) NATURE MATERIALS 1

2 Figure S1. a) XRD patterns and b) PDF refinements for anatase TiO 2 and hydroxyfluorinated anatase Ti O 1.12 F 0.40 (OH) 0.48 obtained by solvothermal synthesis. Table S1. Structural parameters extracted from PDF refinements for TiO 2 anatase and hydroxyfluorinated anatase Ti O 1.12 F 0.40 (OH) TiO 2 Ti O 1.12 F 0.40 (OH) 0.48 a (Å) c (Å) (7) 9.494(3) 3.802(1) 9.522(6) V (Å 3 ) (3) (6) Coherence length (nm) 6.04(16) 4.98(15) Ti [4a] occupancy 1.00(1) 0.74(4) Figure S2. Cyclic voltammetry of THF solution containing 0.2 mol L -1 2PhMgCl-AlCl 3 (designated APC) (scan rate: 50 mv s -1, working electrode: Pt plate). NATURE MATERIALS 2

3 0.2 mol L -1 2PhMgCl-AlCl 3 in THF electrolyte solution 1 was prepared by dissolving the predetermined amount of 2 mol L -1 PhMgCl/THF (Sigma-Aldrich) and 0.5 mol L -1 AlCl 3 /THF (Sigma-Aldrich) in THF under stirring for 12 hours. The preparation was conducted in an argon-filled glove box (Mbraun, Germany) with H 2 O and O 2 below 1 ppm. Figure S2 shows the electrochemical behaviors of the solution determined by cyclic voltammetry, exhibiting two characteristics. One is the electrochemical reversible Mg deposition/dissolution, and other one is the electrochemical stability window of the electrolyte solution of more than 2.7 V vs. Mg 2+ /Mg. Figure S3. Cyclic voltammetry of AlCl 3 /[EMIm]Cl ionic liquid electrolyte with the molar ratio of EMImCl:AlCl 3 =1:1.1 (scan rate: 50 mv s -1, working electrode: Pt plate). A chloroaluminate-based room-temperature ionic liquid ([EMIm]Cl-AlCl 3 ) electrolyte was prepared by slowly mixing 1-ethyl-3-methylimidazolium chloride ([EMIm]Cl, 98%, Aldrich) and anhydrous AlCl 3 (99.99%, Ultra-dry, Alfa Aesar) under stirring with a mole ratio of 1:1.1. [EMIm]Cl was dried at 120 C under vacuum (<1 Pa) for 12 h prior to use. The preparation was conducted in an argon-filled glove box (Mbraun, Germany) with H 2 O and O 2 below 1 ppm. Figure S3 shows the electrochemical behaviors of the AlCl 3 /[EMIm]Cl ionic liquid electrolyte determined by cyclic voltammetry, exhibiting two characteristics. One is the reversible electrochemical Al deposition/dissolution, and other one is the electrochemical stability window of the electrolyte of more than 2.0 V vs. Al 3+ /Al. NATURE MATERIALS 3

4 Figure S4. (a) Rate capability, (b) and (c) long-term cycling performance at high current densities of 150 ma g -1 and 300 ma g -1 (obtained after activation of the cell at 20 ma g -1 for first 3 cycles) in the potential range of V vs. Mg 2+ /Mg for Ti O 1.12 F 0.40 (OH) 0.48 electrode. NATURE MATERIALS 4

5 Ex-situ high-energy synchrotron diffraction on magnesiated/de-magnesiated electrodes Figure S5. Ex situ high energy XRD patterns at different discharged and charged states from points i viii. NATURE MATERIALS 5

6 Figure S6. Refinements of the PDFs at different discharged and charged states using anatase crystal structure for Ti O 1.12 F 0.40 (OH) 0.48 electrodes in Mg-ion batteries. Table S2. The extracted lattice constant and Mg occupancy in Ti [4a] from PDF fits for initial electrode and various discharged and charged electrodes. Initial Discharge Charge a (Å) c (Å) V (Å 3 ) Mg Occ. in Ti [4a] i ii iii iv v vi vii viii (0.001) (0.001) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.0057) (0.0056) (0.0063) (0.0062) (0.0076) (0.0054) (0.0049) (0.0053) N.A (0.03) (0.04) (0.04) (0.04) (0.04) (0.03) NATURE MATERIALS 6

7 Quantification of the valence state of Ti in the reduced sample The Ti L 2,3 EEL spectrum was evaluated in terms of Ti valence state according to a calibration method available in literature. 2 The spectrum was first subtracted of an exponential background and of a double-arctangent function, which removes the contributions from transitions to unoccupied continuum states (Figure S7a). Subsequently, the integrated intensity ratio between two 1 ev windows cantered on the white lines of Ti (Figure S7b) was measured and used to calculate the valence state of Ti. 2 Figure S7. Ti L 2,3 EEL spectrum of Mg0.13 before (a) and after (b) double-arctangent background subtraction. Green line: subtracted background; dashed vertical lines: integration windows. NATURE MATERIALS 7

8 Structural analysis of the chemically magnesiated sample Figure S8. PDF refinement of the chemically magnesiated Ti O 1.12 F 0.40 (OH) 0.48 sample. Table S3. Structural parameters extracted from the PDF refinement of chemically magnesiated Ti O 1.12 F 0.40 (OH) 0.48 sample. Magnesiated sample (0.13 eq. Mg) a (Å) 3.819(4) c (Å) 9.53(3) V (Å 3 ) (6) Mg 2+ in [4a] site (Ti vacancies) 0.12(6) Mg 2+ in [4b] site (Interstitial) 0.03(3) NATURE MATERIALS 8

9 Solid-state 19 F MAS NMR For comparison purpose, we acquired the 19 F NMR spectrum of TiF 3 (Figure S9). The effect of dipolar coupling interactions between the unpaired electrons of Ti and the 19 F nuclei accounts for the large number of intense spinning sidebands of the 19 F MAS NMR spectra of TiF 3 (this effect being not explicitly considered in the fits of the spectra, they are fitted with very large "chemical shift anisotropies") and for the large width of the 19 F NMR resonance. Assuming that, at this spinning frequency (34 khz), homonuclear 19 F- 19 F dipolar couplings are significantly reduced, this broadening arises mainly from this effect. The 19 F NMR shift value of TiF 3 is δ iso = ppm which is in the range of the iso values of diamagnetic MF n fluorides. 3,4 The NMR shift of TiF 3 increases when the temperature increases (~17 ppm between 57 C (30 khz) and 70 C (34 khz)). Figure S9. Experimental (in blue) and fitted (in red) 19 F MAS (34 khz) NMR spectra of TiF 3. The parameters used for this fit are the following: δ iso = ppm, Line Width = 24.3 ppm, δ csa = 493 ppm, csa = NATURE MATERIALS 9

10 The intercalation of one Mg 2+ ion being simultaneous with the reduction of two Ti 4+ ions into Ti 3+ ions, the number of fluorine environments dramatically increases from the pristine compound to the chemically magnesiated samples; Ti IV 2 -F can be converted into Ti III 2 -F, Ti IV Mg -F, Ti IV Mg 2 -F, Ti III Mg -F and Ti III Mg 2 -F, Ti IV 2 -F can be converted into Ti IV Ti III -F, Ti III 2 -F, Ti IV 2Mg-F, Ti IV Ti III Mg-F and Ti III 2Mg-F and Ti IV 3-F can be converted into Ti IV 2Ti III -F, Ti IV Ti III 2-F and Ti III 3-F. So, based on the 19 F NMR chemical shift values of the environments Ti IV 2 -F, Ti IV 2 -F and Ti IV 3-F in Ti O 1.12 F 0.40 (OH) 0.48, Mg 3 -F in MgF 2 and Ti III 2-F in TiF 3 and on the decrease of the chemical shift value of fluoride anions when the number of surrounding cations increases, the intercalation of Mg 2+ ion in a titanium vacancy and the reduction of a Ti 4+ ion into a Ti 3+ ion are both expected to induce a decrease of the iso value of the neighboring fluorine. A decrease (an increase) of the relative intensities of the NMR resonances of higher (smaller) chemical shift, the vanishing of the NMR line assigned to Ti IV 2 -F and the occurrence of a broad NMR line at lower chemical shift are indeed observed when Mg 2+ ions are inserted. Furthermore, due to dipolar coupling interactions between the unpaired electrons of Ti 3+ and the 19 F nuclei, the NMR lines broaden, in agreement with the presence of Ti 3+ ions. The expected evolution of the 19 F δ iso values of the species which may exist in the chemically magnesiated sample (Table S4) compared to the 19 F δ iso values of the species existing in Ti O 1.12 F 0.40 (OH) 0.48 (Figure S10, Table S5), led us to tentatively assign the 19 F NMR resonances to the various species in the chemically magnesiated sample (Figure S11, Table S6). The chemical shift value of the Ti III 3 F species is assumed to be the lowest one and we assume that this species does not occur in the chemically magnesiated sample. Table S4. Species occurring in Ti O 1.12 F 0.40 (OH) 0.48, 19 F δ iso values (ppm) and relative intensities (I, %) of the associated NMR lines of these species, conversion of these species in chemically magnesiated samples, expected evolution of their 19 F δ iso values and potential occurrence from the 19 F NMR spectra of the chemically magnesiated samples. Species in Ti O 1.12 F 0.40 (OH) 0.48 Ti IV 3-F Ti IV 2 -F Ti IV 2 -F F δ iso I Conversion Expected evolution of 19 F δ iso Potential occurrence Ti IV 3-F = x (low) Ti IV 2Ti III -F x (low) Ti IV Ti III 2-F x (low) Ti III 3-F Ti IV 2 -F = x Ti IV Ti III -F x Ti III 2 -F x Ti IV 2Mg-F x Ti IV Ti III Mg-F x Ti III 2Mg-F x Ti IV 2 -F = Ti III 2 -F x Ti IV Mg -F x Ti IV Mg 2 -F x Ti III Mg -F x Ti III Mg 2 -F x NATURE MATERIALS 10

11 Figure S10. Experimental (in blue) and fitted (in red) 19 F MAS (60 khz) NMR spectra of Ti O 1.12 F 0.40 (OH) The individual resonances used for the fit are shown below (see Table S5). Table S5. Isotropic chemical shifts δ iso (ppm), line widths LW (ppm), and relative intensities I (%) of the NMR lines obtained from the reconstruction of the 19 F solid state MAS (60 khz) NMR spectrum of Ti O 1.12 F 0.4 (OH) 0.48 and assignment of these NMR lines. δ iso (±0.5) LW (±0.5) I (±0.5) Assignment Ti 3 -F Ti 2 -F Ti 2 -F NATURE MATERIALS 11

12 Figure S11. Experimental (in blue) and fitted (in red) 19 F MAS (34 khz) NMR spectra of chemically magnesiated sample (0.13 eq. Mg). The individual resonances used for the fit are shown below (see Table S6). Table S6. Isotropic chemical shifts δ iso (ppm), line widths LW (ppm), and relative intensities I (%) of the NMR lines obtained from the reconstruction of the 19 F solid state MAS (34 khz) NMR spectrum of chemically magnesiated sample (0.13 eq. Mg) and tentative assignment of these NMR lines. δ iso (±0.5) LW (±0.5) I (±0.5) Assignment Ti IV Ti III 2-F, Ti III 2Mg-F, Ti III Mg 2 -F Ti IV 2Ti III -F, Ti III 2 -F, Ti IV Ti III Mg-F, Ti IV Mg 2 -F Adsorbed F Ti IV 3-F, Ti IV Ti III -F, Ti IV 2Mg-F, Ti III Mg -F Ti IV 2 -F, Ti IV Mg -F, Ti III 2 -F NATURE MATERIALS 12

13 References: 1. Aurbach, D., Suresh, G. S., Levi, E., Mitelman, A., Mizrahi, O., Chusid, O. & Brunelli, M. Progress in Rechargeable Magnesium Battery Technology. Adv. Mater. 19, (2007). 2. Stoyanov, E., Langenhorst, F. & Steinle-Neumann, G. The effect of valence state and site geometry on Ti L 3,2 and O K electron energy-loss spectra of Ti x O y phases. Am. Mineral. 92, (2007). 3. Sadoc, A., Body, M., Legein, C., Biswal, M., Fayon, F., Rocquefelte, X. & Boucher, F. NMR parameters in alkali, alkaline earth and rare earth fluorides from first principle calculations. Phys. Chem. Chem. Phys. 13, (2011). 4. Sadoc, A., Biswal, M., Body, M., Legein, C., Boucher, F., Massiot, D. & Fayon, F. NMR parameters in column 13 metal fluoride compounds (AlF 3, GaF 3, InF 3 and TlF) from first principle calculations. Solid State Nucl. Magn. Reson , 1-7 (2014). NATURE MATERIALS 13