Elsevier

Journal of Solid State Chemistry

Volume 207, November 2013, Pages 208-217
Journal of Solid State Chemistry

NbF5 and TaF5: Assignment of 19F NMR resonances and chemical bond analysis from GIPAW calculations

https://doi.org/10.1016/j.jssc.2013.09.001Get rights and content

Highlights

  • The 19F δiso values of NbF5 and TaF5 have been determined.

  • The 19F chemical shielding tensors have been calculated using the GIPAW method.

  • A confident assignment of the 19F NMR lines of NbF5 and TaF5 is obtained.

  • The relationships between the 19Fδiso values and the M–F bonds features are established.

Abstract

The 19F isotropic chemical shifts (δiso) of two isomorphic compounds, NbF5 and TaF5, which involve six nonequivalent fluorine sites, have been experimentally determined from the reconstruction of 1D 19F MAS NMR spectra. In parallel, the corresponding 19F chemical shielding tensors have been calculated using the GIPAW method for both experimental and DFT-optimized structures. Furthermore, the [M4F20] units of NbF5 and TaF5 being held together by van der Waals interactions, the relevance of Grimme corrections to the DFT optimization processes has been evaluated. However, the semi-empirical dispersion correction term introduced by such a method does not show any significant improvement. Nonetheless, a complete and convincing assignment of the 19F NMR lines of NbF5 and TaF5 is obtained, ensured by the linearity between experimental 19F δiso values and calculated 19F isotropic chemical shielding σiso values. The effects of the geometry optimizations have been carefully analyzed, confirming among other matters, the inaccuracy of the experimental structure of NbF5. The relationships between the fluorine chemical shifts, the nature of the fluorine atoms (bridging or terminal), the position of the terminal ones (opposite or perpendicular to the bridging ones), the fluorine charges, the ionicity and the length of the M–F bonds have been established. Additionally, for three of the 19F NMR lines of NbF5, distorted multiplets, arising from 1J-coupling and residual dipolar coupling between the 19F and 93Nb nuclei, were simulated yielding to values of 93Nb–19F 1J-coupling for the corresponding fluorine sites.

Graphical abstract

The complete assignment of the 19F NMR lines of NbF5 and TaF5 allow establishing relationships between the 19F δiso values, the nature of the fluorine atoms (bridging or terminal), the position of the terminal ones (opposite or perpendicular to the bridging ones), the fluorine charges, the ionicity and the length of the M–F bonds.

  1. Download : Download full-size image

Introduction

Unambiguous assignment of complex 19F solid-state NMR spectra of crystalline fluoride compounds having multiple crystallographic sites or fluoride glasses often remains a challenging task. In several studies [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], assignments of 19F NMR resonances to environments were based on similarities to crystalline model compounds (similar 19F isotropic chemical shifts (δiso) values intending to indicate similar structural environments). Nowadays, high resolution two-dimensional heteronuclear and homonuclear correlation experiments can be used for helping in the assignment of fluorine sites. The efficiency of these experiments to probe interatomic spatial proximities and through bond connectivities has been described in various studies [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28]. An alternative approach consists in the calculation of 19F isotropic chemical shieldings (σiso). These calculations have been first achieved in various crystalline fluoride materials [29], [30], [31], [32], [33], [34], [35] using a semi-empirical model, called superposition model [36]. Ab initio approaches devoted to molecular systems have been used for studying extended systems. In these cases, clusters of atoms were build such that the central atom experiences an environment similar to that in true extended solid, i.e., to mimic the crystalline periodic structures [14], [30], [37], [38], [39], [40], [41], [42], [43], [44]. Recent advances in the theoretical calculations of NMR parameters for extended solids lead to the development of the Gauge Including Projector Augmented Wave (GIPAW) method [45], [46] which integrates explicitly the periodic boundary conditions. This major breakthrough enables consequently the calculations of NMR tensors in solids [47], [48] and was applied on inorganic fluorides for the calculation of 19F σiso values [19], [20], [21], [22], [23], [27], [49], [50], [51], [52], [53].

When interested in the prediction of δiso values, the calculated 19F σiso values have to be converted into the isotropic chemical shift scale. Assuming that the 19F σref can be obtained, the calculated 19F σiso values can be converted into “calculated” 19F δiso values applying the relation δisoσrefσiso [14], [19], [42], [49], [54]. Calculated 19F σiso values can also be converted into “calculated” 19F δiso values [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [43] using an experimental absolute scale for fluorine [55], [56]. To avoid these referencing problems and possible errors coming from the calculation of the 19F σref value, “calculated” 19F δiso values can be deduced from the linear regression established between calculated 19F σiso values and experimental 19F δiso values for the compounds under study [20], [21], [23], [50], [51], [53], implying that an assignment was already done. This third approach seems to become the standard practice [47], [48]. Alternatively a linear regression previously established on various compounds [50] can be used [22], [27], [52]. For many nuclei, deviations from the theoretically expected slope of minus one have been reported for GIPAW calculations [48] using GGA (generalized gradient approximation) for the exchange and correlation potential. 19F is not an exception with a large dispersion of reported slopes: −0.68 [20], −0.80 [50], 0.86 [49] and 0.83 [21], the last two positive values being obtained when “calculated” and experimental δiso are compared. Pedone et al. [53] were able to reach a nearly ideal slope value of −0.963 but only on few compounds.

The two binary pentafluorides NbF5 and TaF5 are isomorphs and consist of tetrameric structural units [M4F20] (M=Nb,Ta), build up of MF6 octahedra linked to each other by corners in a cis-configuration. Both the compounds involve six fluorine sites, 2 sites with multiplicity 4 and 4 sites with multiplicity 8 [57], [58]. These two compounds, known to be very hygroscopic, have not been yet studied by solid state NMR. In the present work, we report for the first time the one dimensional (1D) 19F magic angle spinning (MAS) NMR spectra of NbF5 and TaF5, allowing us to determine their experimental 19F δiso values. Due to the hygroscopic character and reactivity (chemical attack of the caps of the rotors by NbF5) of NbF5 and TaF5, we choose to avoid 19F–19F correlation experiments since they require long acquisition time. Anyway, in these cases, only fragmentary information could be obtained since some of the fluorine atoms have similar environments and since at least two 19F NMR resonances overlap. Consequently, the only way to complete the initial partial assignment, based on NMR line relative intensities and 19F δiso and chemical shift anisotropy (δcsa) values, was to perform calculations of the 19F chemical shielding tensors. These DFT (density functional theory) computations have been performed using the GIPAW [45], [46] method implemented in the NMR-CASTEP code [59], [60]. Since the agreement between the calculated and experimental NMR parameters is generally significantly improved after a DFT structural optimization [16], [21], [23], [31], [47], [48], [49], [51], [52], [61], GIPAW calculations have been achieved on both experimental and optimized structures. In NbF5 and TaF5, however, [M4F20] units are held together by van der Waals (vdW) interactions and those dispersion forces are not described by most of the exchange-correlation DFT functional used for the GIPAW calculations. A recently proposed method to work around this problem consists in adding a semi-empirical dispersion energy term to the conventional Kohn–Sham DFT energy [62]. For this purpose, structural optimizations were also done with the semi-empirical DFT-D2 approach of Grimme as implemented in the VASP package [63], [64]. The semi-empirical parameters needed for describing the three elements under study (F, Nb and Ta) were taken from the work of Grimme [62].

In the present work, we show that GIPAW calculations enable a complete and convincing assignment of the lines of the 19F NMR spectra of NbF5 and TaF5. Moreover, the effects of the geometry optimisations are carefully examined and the structural features deeply discussed. Relations are also clearly established between the fluorine chemical shifts and the nature of the fluorine atoms (bridging or terminal), the position of the terminal ones (opposite or perpendicular to the bridging ones), the fluorine Mulliken charges or the ionicity and length of the M–F bonds. Additionally, for three of the 19F NMR lines of NbF5, distorted multiplets, arising from 1J-coupling and residual dipolar coupling between the 19F and 93Nb nuclei, were simulated yielding to values of 93Nb–19F 1J-coupling for the corresponding fluorine sites.

Section snippets

Materials and methods

The samples of TaF5 and NbF5 were purchased from Alfa Aesar (lot numbers D20L30 and l11T005, respectively) and were kept in a dry glove box under nitrogen atmosphere. As both compounds are very hygroscopic, the rotors were filled inside the glove box to avoid any hydration of the samples.

Solid-state NMR experiments were performed on an Avance 300 Bruker spectrometer operating at 7.0 T (19F Larmor frequency of 282.2 MHz), using a 2.5 mm CPMAS probehead. 19F one dimensional (1D) MAS NMR spectra were

Calculation

The chemical shift tensor is described by three parameters, the isotropic chemical shift (δiso), the chemical shift anisotropy (δcsa) and the asymmetry parameter (ηcsa), determined experimentally, and defined as δiso(ppm)=(1/3)(δxx+δyy+δzz), δcsa(ppm)=δzzδiso, ηcsa=(δyyδxx)/δcsa, with the principal components defined in the sequence |δzzδiso||δxxδiso||δyyδiso|.

The chemical shielding tensor is also described by three parameters, the isotropic chemical shielding (σiso), the chemical

Results and discussion

TaF5 and NbF5 are isomorphs and crystallize in a monoclinic cell (space group C2/m, a=9.62 Å, b=14.43 Å, c=5.12 Å and β=96.1° for NbF5 [57] (ICSD [71] file no. 26647) and a=9.5462 Å, b=14.3678 Å, c=5.0174 Å and β=97.086° for TaF5 [58] (ICSD [71] file no. 171155)). They consist of tetrameric structural units, [M4F20], build up of MF6 octahedra linked to each other by corners in a cis-configuration (Fig. 1). The structures of both compounds contain six F crystallographic sites, two sites of

Conclusions

The 19F δiso values in NbF5 and TaF5 are determined from the reconstruction of 1D 19F MAS NMR spectra. An initial partial assignment is achieved, based on NMR line relative intensities and δiso values previously determined in several compounds containing also bridging and terminal fluorine atoms bonded to Nb and Ta atoms.

The atomic position optimizations and full geometry optimizations were performed using the VASP package [63] with and without vdW corrections (i.e. with and without using the

Acknowledgments

The authors thank the Région Pays de la Loire for the financial support of the RMN3MPL project, especially M. Biswal (doctoral grant) and A. Sadoc (post-doctoral fellowship). The computational presented in this work have been carried out at the Centre Régional de Calcul Intensif des Pays de la Loire (CCIPL), financed by the French Research Ministry, the Région Pays de la Loire, and Nantes University. We thank CCIPL for CASTEP licenses financial support.

References (96)

  • J.M. Miller

    Prog. Nucl. Magn. Reson. Spectrosc

    (1996)
  • L.-S. Du et al.

    J. Solid State Chem.

    (1998)
  • B. Bureau et al.

    J. Non-Cryst. Solids

    (1999)
  • J.F. Stebbins et al.

    J. Non-Cryst. Solids

    (2000)
  • J.C.C. Chan et al.

    J. Non-Cryst. Solids

    (2001)
  • T.J. Kiczenski et al.

    J. Non-Cryst. Solids

    (2002)
  • T.J. Kiczenski et al.

    J. Non-Cryst. Solids

    (2004)
  • R.E. Youngman et al.

    Solid State Nucl. Magn. Reson.

    (2005)
  • C. Martineau et al.

    J. Solid State Chem.

    (2013)
  • G. Silly et al.

    C.R. Chim.

    (2004)
  • M. Body et al.

    J. Solid State Chem.

    (2005)
  • B. Bureau et al.

    Chem. Phys.

    (1999)
  • S.-H. Cai et al.

    Chem. Phys. Lett.

    (1999)
  • S.-H. Cai et al.

    Chem. Phys. Lett.

    (2002)
  • T. Charpentier

    Solid State Nucl. Magn. Reson.

    (2011)
  • M. Body et al.

    Chem. Phys. Lett.

    (2006)
  • F. Marchetti et al.

    J. Fluorine Chem.

    (2010)
  • K. Radan et al.

    J. Fluorine Chem.

    (2011)
  • F. Marchetti et al.

    J. Organomet. Chem.

    (2011)
  • R.K. Harris et al.

    Prog. Nucl. Magn. Reson. Spectrosc

    (1992)
  • F. Marchetti et al.

    Inorg. Chim. Acta

    (2011)
  • D. Massiot et al.

    C.R. Chim.

    (2010)
  • B. Bureau et al.

    J. Phys. Condens. Matter

    (1997)
  • Q. Zeng et al.

    Am. Mineral.

    (2000)
  • R.E. Youngman et al.

    J. Am. Ceram. Soc.

    (2002)
  • S. Chaudhuri et al.

    J. Am. Chem. Soc.

    (2002)
  • C.D. Martin et al.

    Am. Mineral.

    (2005)
  • T.M. Alam et al.

    Chem. Mater.

    (2008)
  • C. Martineau et al.

    Chem. Commun.

    (2007)
  • C. Martineau et al.

    Dalton Trans.

    (2008)
  • C. Martineau et al.

    Inorg. Chem.

    (2008)
  • C. Martineau et al.

    Phys. Chem. Chem. Phys.

    (2009)
  • Q. Wang et al.

    Phys. Chem. Chem. Phys.

    (2009)
  • J.M. Griffin et al.

    J. Am. Chem. Soc.

    (2010)
  • C. Martineau et al.

    Inorg. Chem.

    (2011)
  • A.-L. Rollet et al.

    Inorg. Chem.

    (2012)
  • A. Cadiau et al.

    J. Mater. Chem.

    (2011)
  • C. Martineau et al.

    Phys. Chem. Chem. Phys.

    (2011)
  • C. Martineau et al.

    Chem. Mater.

    (2011)
  • C. Martineau et al.

    J. Phys. Chem. C

    (2012)
  • C. Martineau et al.

    Dalton Trans.

    (2013)
  • M. Body et al.

    Inorg. Chem.

    (2004)
  • C. Martineau et al.

    Inorg. Chem.

    (2006)
  • C. Legein et al.

    Inorg. Chem.

    (2006)
  • F. Le Berre et al.

    Dalton Trans.

    (2007)
  • L. Sronek et al.

    J. Phys. Chem. C

    (2008)
  • M. Mortimer et al.

    J. Chem. Soc., Faraday Trans.

    (1996)
  • S.-H. Cai et al.

    J. Phys. Chem. A

    (2002)
  • Cited by (24)

    • Advances in the computation of nmr parameters for inorganic nuclides

      2023, Comprehensive Inorganic Chemistry III, Third Edition
    • Progress in Our Understanding of <sup>19</sup>F Chemical Shifts

      2018, Annual Reports on NMR Spectroscopy
      Citation Excerpt :

      Exact assignment of complex 19F solid-state NMR spectra of either fluoride compounds with multiple crystallographic sites or fluoride glasses is a difficult task. Recently, researchers have begun using high-resolution 2-D heteronuclear or homonuclear NMR correlation experiments for this task [174]. In the past studies, it has been reported that double-quantum single-quantum (DQ–SQ) MAS NMR correlation experiments [175], with homonuclear dipolar couplings [176], were able to determine the fluorine–fluorine spatial proximities in other fluoride materials [177–180].

    • Identify OH groups in TiOF<inf>2</inf> and their impact on the lithium intercalation properties

      2017, Journal of Solid State Chemistry
      Citation Excerpt :

      The lines located at higher chemical shift values, i.e. ~ 146 and 182 ppm (relative intensities equal to 18%), are assigned to fluorine atoms close to titanium vacancies Ti-F-□. These assignments are in agreement with the increase of the chemical shift of the anion when the number of surrounding cations decreases, as observed for 19F NMR in NbF5 and TaF5 [24] and in the anatase Ti0.78□0.22O1.12(OH)0.48F0.40 [7] and for 17O NMR in TiO2(B) [25]. The 19F δiso values of the fluorine atoms close to titanium vacancies Ti-F-□ are significantly larger than those measured for F-Ti□2 in the anatase Ti0.78□0.22O1.12(OH)0.48F0.40 (98 ppm [7]) and for terminal F in K2TiF6 (71.4 ppm [26]) and in hybrid hydroxyfluorotitanates (IV) (76–82 ppm [27]).

    • NMR parameters in column 13 metal fluoride compounds (AlF<inf>3</inf>, GaF<inf>3</inf>, InF<inf>3</inf> and TlF) from first principle calculations

      2014, Solid State Nuclear Magnetic Resonance
      Citation Excerpt :

      These findings therefore clearly confirm that the previously proposed relationship [9] could allow predicting 19F NMR spectra for a broad range of crystalline metal fluoride compounds with a relatively good accuracy. For completeness, it should be noted that Pedone et al. [51] obtain a slope of −0.963 on five compounds among which three contain Ca2+, by applying a larger shift of the local potential of the USPP of Ca2+ and that a slope close to −1 has been obtained for NbF5 [50]. As mentioned above, a nice agreement between experimental quadrupolar parameters and PAW [19,20] calculated values for the IS structures of α-AlF3 and InF3 is observed (Table 2), showing the accuracy of these crystalline structures.

    View all citing articles on Scopus
    View full text