Nb 2 O 2 F 3 : A Reduced Niobium (III/IV) Oxyfluoride. with a Complex Structural, Magnetic and Electronic. Phase Transition

Transcription

1 Supporting Information for: Nb 2 O 2 F 3 : A Reduced Niobium (III/IV) Oxyfluoride with a Complex Structural, Magnetic and Electronic Phase Transition T. Thao Tran,, Melissa Gooch,, Bernd Lorenz,, Alexander P. Litvinchuk,, Maurice G. Sorolla II,, Jakoah Brgoch,, Paul C. W. Chu,,, and Arnold M. Guloy,, * Department of Chemistry, Department of Physics and Texas Center for Superconductivity, University of Houston, Houston, Texas 77204, USA Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, California 94720, USA CORRESPONDING AUTHOR ADDRESS: aguloy@uh.edu S1

2 Table of Contents Experimental Section page S3-S7 Figure S1. Experimental and calculated powder X-ray diffraction patterns of monoclinic (M)-Nb 2 O 2 F 3 Figure S2. Le Bail refinement of the X-ray powder diffraction of monoclinic (M)-Nb 2 O 2 F 3 Figure S3. Results of EDS analysis for Nb 2 O 2 F 3 page S8 page S9 page S10 Figure S4. Magnetic susceptibilities, χ(t), of Nb 2 O 2 F 3 within the vicinity of the phase transition at 90K, under different magnetic fields (H = 1-5 Tesla) Figure S5. Temperature dependence (2 K to 150 K) of heat capacity (Cp) of Nb 2 O 2 F 3 in the vicinity of the phase transitions Figure S6. Raman scattering spectra of Nb 2 O 2 F 3 at 295, 70, and 8K Table S1. Crystallographic data and refinement results for monoclinic (M)-Nb 2 O 2 F 3 and triclinic (T)-Nb 2 O 2 F 3 Table S2. Atomic coordinates, equivalent isotropic and anisotropic displacement parameters (Å 2 ) for monoclinic (M)-Nb 2 O 2 F 3 Table S3. Atomic coordinates, equivalent isotropic and anisotropic displacement parameters (Å 2 ) for triclinic (T)-Nb 2 O 2 F 3 Table S4. Selected bond distances (Å) and angles ( ) for monoclinic (M)-Nb 2 O 2 F 3 and triclinic (T)-Nb 2 O 2 F 3 Table S5. Bond valence sum analysis for monoclinic (M)-Nb 2 O 2 F 3 Table S6. Bond valence analysis for triclinic (T)-Nb 2 O 2 F 3 page S11 page S12 page S13 page S14 page S15 page S16 page S17 page S18 page S19 S2

3 Experimental Section Reagents Nb (99.8%, Sigma-Alrich), SnF 2 (99.9%, Sigma-Alrich), SnO (99.9%, Sigma-Alrich) and Sn (99.9%, Sigma-Alrich) were used as starting materials. Synthesis The title compound was synthesized through the reaction of Nb, SnF 2 and SnO in Sn flux (4:4:3:8 molar ratio) in welded Nb containers within evacuated quartz jacket. The reaction was performed at 600 C for 7 days, followed by slowly cooling (3 C/hour) to 400 C. The reaction charge was then centrifuged while at 400 C to separate the crystals from the Sn flux, and then air-quenched to room temperature. Black and shiny crystals, subsequently determined to be Nb 2 O 2 F 3, were obtained. The title compound is relatively stable to air and moisture. No traces or any changes in chemical and physical properties were observed if the Nb 2 O 2 F 3 crystals were exposed to the air for a few days. All experimental manipulations and handling were performed under inert conditions: within a purified Ar-atmosphere glovebox with total H 2 O and O 2 levels of < 0.1 ppm. The products were stored in a glovebox, leaving the fresh and pure sample for further characterization. All polycrystalline samples (ground crystals) used in physical and chemical property measurements were determined to be phase pure by X-ray powder diffraction (Figure S1 and S2). Structure determination: Single-crystal X-ray diffraction and Bond valence calculations A black plate-shaped crystal ( mm) was selected for single-crystal diffraction analysis. Data were collected on a Bruker DUO platform diffractometer equipped with a 4K CCD APEX II detector using graphite-monochromated Mo-Kα radiation at room temperature and 83 K. A hemisphere of data (3061 frames at 6 cm detector distance) was S3

4 collected using a narrow-frame algorithm with scan widths of 0.30 in omega and an exposure time of 20 seconds per frame. Nb 2 O 2 F 3 crystal was found to be non-merohedrally twinned, with two major (approximately 90%) and several minor domains. The data reduction was processed as two-domain twin with the ratio of 4:1 by volume. The data were integrated using the SAINT- V7.23A program, 1 with the intensities corrected for Lorentz factor, polarization, air absorption, and absorption attributable to the variation in the path length through the detector faceplate. TWINABS and SADABS absorption corrections were applied based on the entire data sets collected at room temperature and low temperature 83 K, respectively. 1 Redundant reflections were averaged. Final cell constants for monoclinic (M)-Nb 2 O 2 F 3 and triclinic (T)-Nb 2 O 2 F 3 were refined using 4218 and 2527 reflections, respectively, having I > 10σ(I). It should be noted that the two data sets obtained at room temperature and 83 K were collected on the same single crystal. The positions of the niobium atoms were determined by Direct methods using SHELXS-97 2 and the remaining atoms were located by difference Fourier maps and least-squares cycles utilizing SHELXL All calculations were performed using SHELXL-97 crystallographic software package. 3-4 Relevant crystallographic data for Nb 2 O 2 F 3 are given in Tables S1. Since O and F atoms cannot be well differentiated by X-ray diffraction, refinements with varying mixed O/F occupancies during the least-squares refinements were attempted. However, the refinement of the two data sets (room temperature and low temperature) did not proceed very nicely and the thermal displacement parameters of some O/F atoms were not well-behaved (non-positive definite). The structural modeling was then improved and completed by refining O and F separately at the different sites. For the monoclinic structure: F was satisfactorily refined at the S4

5 X1 and X2 sites, and O was refined at the X3 site. For the low temperature triclinic structure: F was refined at the X1, X2 and X3 sites, and O was refined at the X4 and X5 sites. Bond valence sum (BVS) calculations, based on the empirical expression S i = exp[(r 0 -R i )/B] where S i =valence of bond i and B=0.37, 5 were performed for both crystal structures. For monoclinic (M)-Nb 2 O 2 F 3, BVS of Nb 3.5+, O 2- (X3) and F - (X1,X2) resulted in values of 3.39, 1.64, and , respectively (Table S5). For triclinic (T)-Nb 2 O 2 F 3, BVS of Nb 3+, Nb 4+, O 2- (X4, X5) and F - (X1, X2, X3) resulted in values of 3.10, 3.66, , and , respectively (Table S6). Nevertheless, we still cannot completely discount disorder between O and F atoms in the ligand sites (X). Thus, we assign X as O/F disorder in the structural description. Atomic coordinates, equivalent isotropic and anisotropic displacement parameters, selected bond distances and angles for Nb 2 O 2 F 3 are presented in Tables S2-S4. Powder X-ray diffraction Powder X-ray diffraction (PXRD) experiments on Nb 2 O 2 F 3 were performed using a PANalytical X Pert PRO diffractometer equipped with Cu Kα radiation. Data were collected in the 2θ range of 5-90 with a step size of and a scan time of 0.3s. Preferred orientations in the PXRD of Nb 2 O 2 F 3 arise from the plate-shaped crystal. Precise unit cell parameters were obtained from Le Bail refinements using JANA2006 software. 6 No impurities were observed and the experimental and calculated PXRD are in excellent agreement (Figure S1-S2). Energy-dispersive X-ray spectroscopy (EDS) Analysis A JEOL JSM 6330F field emission scanning electron microscope equipped with an electron dispersive spectrometer was used to determine the presence of Nb, O and F and confirm the absence of Sn above 0.03 wt.% level. The data crystal of Nb 2 O 2 F 3 was mounted on one flat face. Intensity data were processed by EDAX TEAM software. Analyses on this sample were S5

6 obtained with a focused beam of 15 kev accelerating voltage and 12 µa emission current (Figure S3). Raman spectroscopy Raman scattering spectra of Nb 2 O 2 F 3 were measured using the triple spectrometer Jobin-Yvon T64000, equipped with an optical microscope and charge-coupled-device detector. A 100x objective was used to both focus the incident laser beam into a spot of about 2 micrometer in diameter and collect the scattered light in the backward scattering geometry. The nm Ar+ laser line was used for the excitation. Laser power did not exceed 0.5 mw in order to avoid heating of the sample. The spectral resolution did not exceed 1.5 cm -1. Sample was mounted on a cold finger of a helium flow cryostat (Microstat, Oxford) and temperature was control to within 0.1K. Magnetic and transport measurements The magnetic properties were studied in a Magnetic Property Measurement System (MPMS, Quantum Design) in magnetic fields up to 5 Tesla and temperatures between 2 K and room temperature (~300 K). Several single crystals with random orientations were used for magnetization measurements. The resistivity (transport) measurements were performed on a compressed polycrystalline pellet using the standard four-probe method, employing a Keithley 220 dc current source and a Keithley 182 nano voltmeter. Thin platinum wires were attached to the sample using silver paint. Heat capacity measurements The heat capacity was measured using a relaxation method implemented as an option into the Physical Property Measurement System (PPMS, Quantum Design). Several crystals of Nb 2 O 2 F 3 with a total mass of a few milligrams were mounted on the sample platform for measurement. S6

7 Lattice dynamics calculations The first principles density functional calculations of the electronic ground state of both structure modifications of Nb 2 O 2 F 3 were performed within the local density approximation with the Perdew-Wang local functional, 7 using the Dmol3 code. 8,9 The integration over the Brillouin zone was performed over the 5x5x5 (5x5x3 for the triclinic phase) Monkhorst-Pack grid in the reciprocal space. Prior to calculations the internal atomic coordinates were relaxed so that the forces on atoms in equilibrium position did not exceed 1 mev/a. The finite displacement method was further used to access the lattice dynamics properties. S7

8 Figure S1. Experimental and calculated powder X-ray diffraction patterns of monoclinic (M)- Nb 2 O 2 F 3. Preferred orientations in the PXRD of Nb 2 O 2 F 3 arise from the plate-shaped morphology of the crystal. S8

9 Figure S2. X-ray powder diffraction Le Bail refinement of monoclinic (M)-Nb 2 O 2 F 3 Unit cell parameters Le Bail refinements Single-crystal refinements a(å) (7) (1) b(å) (8) (1) c(å) β( ) (5) (5) (2) (1) S9

10 Figure S3. Results of EDS analysis for Nb 2 O 2 F 3 Element Atomic % Experimental Atomic % Theoretical Nb 27.94(3) O 30.52(9) F 41.54(7) S10

11 Figure S4. Magnetic susceptibilities, χ(t), of Nb 2 O 2 F 3 within the vicinity of the phase transition at 90K, under different magnetic fields (H = 1-5 Tesla) S11

12 Figure S5. Temperature dependence (2 K to 150 K) of heat capacity (Cp) of Nb 2 O 2 F 3 in the vicinity of the phase transitions S12

13 Figure S6. Raman scattering spectra of Nb 2 O 2 F 3 at 295, 70, and 8K S13

14 Table S1. Crystallographic data and refinement results for monoclinic (M)-Nb 2 O 2 F 3 and triclinic (T)-Nb 2 O 2 F 3 (M)-Nb 2 O 2 F 3 (T)-Nb 2 O 2 F 3 M/gmol T/K 296(2) 296(2) 83(2) λ/ Å Crystal system Monoclinic Monoclinic Triclinic Space group [setting] I2/a (No. 15) [preferred] C2/c (No. 15) [standard] P1 (No. 2) a/ Å (1) (2) (5) b/ Å (1) (1) (6) c/ Å (2) (1) (7) α/ deg (3) β/ deg (1) (1) (2) γ/ deg (3) V/ Å (11) (11) (3) Z D c /gcm µ/mm θ max / Number of reflections Number of parameters R int GOF R(F) a R w (F 2 o ) b a R(F) = Σ F o - F c /Σ F o. b R w (F o 2 ) = [Σw(F o 2 - F c 2 ) 2 /Σw(F o 2 ) 2 ] 1/2 S14

15 Table S2. Atomic coordinates, equivalent isotropic and anisotropic displacement parameters (Å 2 ) for monoclinic (M)-Nb 2 O 2 F 3 I2/a (in preferred setting) Atom x y Z a U eq Nb (2) (2) (2) (5) X (2) (2) (6) (2) X (2) (2) X (2) (2) (6) (2) a U eq is defined as one-third of the trace of the orthogonalized U ij tensor Wyckoff 8f 8f 4e 8f Atom U 11 U 22 U 33 U 23 U 13 U 12 Nb (6) (6) (6) (2) (3) (2) X (3) (3) (3) (2) (2) (2) X (4) (4) (3) (3) X (3) (3) (3) (2) (2) (2) Atomic coordinates, equivalent isotropic and anisotropic displacement parameters (Å 2 ) for monoclinic (M)-Nb 2 O 2 F 3 C2/c (in standard setting) Atom x y z a U eq Nb (7) (2) (2) (5) X (6) (2) (2) (2) X (2) (2) X (6) (2) (2) (2) a U eq is defined as one-third of the trace of the orthogonalized U ij tensor Wyckoff 8f 8f 4e 8f Atom U 11 U 22 U 33 U 23 U 13 U 12 Nb (6) (6) (6) (2) (4) (2) X (3) (3) (2) (2) (2) (2) X (3) (4) (4) (3) X (3) (3) (3) (2) (2) (2) S15

16 Table S3. Atomic coordinates, equivalent isotropic and anisotropic displacement parameters (Å 2 ) for triclinic (T)-Nb 2 O 2 F 3 Atom x y z a U eq Wyckoff Nb (9) (8) (7) (3) 2i Nb (9) (8) (7) (3) 2i X (6) (5) (5) (6) 2i X (6) (5) (5) (6) 2i X (6) (5) (4) (6) 2i X (7) (6) (5) (7) 2i X (5) (6) (6) (7) 2i a U eq is defined as one-third of the trace of the orthogonalized U ij tensor Atom U 11 U 22 U 33 U 23 U 13 U 12 Nb (4) (4) (4) (2) (2) (2) Nb (4) (4) (4) (2) (2) (2) X (13) (13) (13) (10) (11) (10) X (13) (13) (13) (11) (11) (10) X (14) (13) (13) (11) (11) (11) X (16) (16) (16) (13) (13) (12) X (16) (15) (15) (12) (13) (12) S16

17 Table S4. Selected bond distances (Å) and angles ( ) for monoclinic (M)-Nb 2 O 2 F 3 and triclinic (T)-Nb 2 O 2 F 3 (M)-Nb 2 O 2 F 3 (T)-Nb 2 O 2 F 3 Nb Nb (1) Nb1 Nb (9) Nb X1 Nb X1 Nb X2 Nb X3 Nb X3 Nb X (7) (7) (4) (8) (8) (8) Nb1 X1 Nb1 X2 Nb1 X3 Nb1 X4 Nb1 X4 Nb1 X (3) 2.137(3) 2.087(3) 2.067(3) 2.077(3) 2.038(3) X1 Nb X2 X1 Nb X3 X1 Nb X3 X2 Nb X3 X2 Nb X3 X3 Nb X3 X3 Nb X3 Nb X1 Nb Nb X2 Nb Nb X3 Nb 90.32(2) 90.78(3) 88.20(3) 94.59(3) 85.12(3) 92.21(1) 94.85(2) (4) (5) 77.78(3) Nb2 Nb (9) Nb2 X (3) Nb2 X (3) Nb2 X (3) Nb2 X (4) Nb2 X (3) Nb2 X (3) X1 Nb1 X2 X1 Nb1 X4 X3 Nb1 X4 X3 Nb1 X4 X4 Nb1 X5 X4 Nb1 X5 Nb1 X2 Nb2 Nb1 X3 Nb2 Nb1 X4 Nb (11) 81.40(12) 91.97(12) 90.77(12) 90.76(13) 95.20(14) (15) (14) 74.20(12) X1 Nb2 X3 X1 Nb2 X4 X2 Nb2 X5 X3 Nb2 X4 X3 Nb2 X5 X4 Nb2 X5 X4 Nb2 X5 Nb2 X5 Nb (11) 98.63(13) 88.84(13) 90.26(13) 83.89(12) 93.63(13) 95.00(14) 81.70(12) S17

18 Table S5. Bond valence analysis for monoclinic (M)-Nb 2 O 2 F 3 a Atom O 2 F1 2 F2 Σcation Nb Σanions a Bond valence sums calculated with the formula: S i = exp[(r 0 -R i )/B], where S i =valence of bond i and B=0.37. S18

19 Table S6. Bond valence analysis for triclinic (T)-Nb 2 O 2 F 3 a Atom O1 O2 F1 F2 F3 Σcation Nb Nb Σanions a Bond valence sums calculated with the formula: S i = exp[(r 0 -R i )/B], where S i =valence of bond i and B=0.37. S19

20 REFERENCES (1) SAINT; 7.23A ed.; Bruker AXS Inst. Inc.: Madison, WI: (2) Sheldrick, G. M. In Program for Solution of Crystal Structures; University of Gottingen, Germany: (3) Sheldrick, G. M. In Program for Refinement of Crystal Structures; University of Gottingen, Germany: (4) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2008, 64, 112. (5) Brown, I. D. Chemical Reviews 2009, 109, (6) Petricek, V.; Dusek, M.; Palatinus, L. Z. Kristallogr. - Cryst. Mater. 2014, 229, 345. (7) Perdew, J. P.; Wang, Y. Phys. Rev. B 1992, 45, (8) Delley, B. J. J. Phys. Chem. 1996, 100, (9) Delley, B. J. J. Chem. Phys. 2000, 113, S20