Synthesis and redetermination of the crystal structure of NbF₅

Single crystals of NbF₅, niobium(V) fluoride, have been obtained by the reaction of niobium metal in a stream of dilute elemental fluorine at 473 K and subsequent sublimation. The as-obtained bulk phase compound was shown to be pure by powder X-ray diffraction at 293 K and by IR and Raman spectroscopy. A single-crystal X-ray analysis was conducted at 100 K. In comparison to the previously reported structure model [Edwards (1964). J. Chem. Soc. pp. 3714–3718], the lattice parameters and fractional atom coordinates were determined to much higher precision and individual, anisotropic displacement parameters were refined for all atoms.

1. Chemical context

NbF₅ was first synthesized by Ruff and Schiller (Ruff, 1909; Ruff & Schiller, 1911) from the reaction of Nb metal with elemental fluorine or from the reaction of NbCl₅ with anhydrous HF. By now, several alternative ways for its synthesis have also been described in the literature (Schäfer et al., 1965; O'Donnell & Peel, 1976). Niobium penta­fluoride is a colorless, hygroscopic solid that melts at 352.1 K and has a boiling point of 506.5 K (Junkins et al., 1952). The vapor pressure (Junkins et al., 1952; Fairbrother & Frith, 1951), the enthalpy of fusion (Junkins et al., 1952), and the electrical conductivity (Fairbrother et al., 1954) of liquid NbF₅ have also been determined. Infrared and Raman spectra of the solid were measured (Preiss & Reich, 1968; Beattie et al., 1969) and the structure of NbF₅ in the (supercooled) liquid, glassy state and the gas phase have been investigated by Raman spectroscopy (Boghosian et al., 2005; Papatheodorou et al., 2008). In a search for a suitable laboratory synthesis of NbF₅, we investigated several methods for its preparation. During our efforts, single crystals of several millimeters in size were obtained when hot NbF₅ re-sublimed at colder parts of our reaction setup (see Synthesis and crystallization). The former crystal structure published by Edwards (1964) is of lower precision compared to structure determinations possible nowadays and displacement parameters had not been refined anisotropically.

2. Structural commentary

The lattice parameters obtained from powder X-ray diffraction at 293 K [a = 9.62749 (19), b = 14.4564 (3) c = 5.12831 (10) Å, β = 95.8243 (4)°] agree with those determined by Edwards (1964) [a = 9.62 (1), b = 14.43 (2), c = 5.12 (1) Å, β = 96.1 (3)°]. Although the temperature was not explicitly stated in Edwards' work, it can be assumed that the structure was determined at room temperature. The powder X-ray diffraction pattern is shown in Fig. 1; crystallographic details of the Rietveld refinement are given in Table 1 and the supporting information.

Table 1
Selected crystallographic details for NbF₅ determined from single-crystal X-ray diffraction (SCXRD, middle column) and powder X-ray diffraction (PXRD, Rietveld refinement, right column)

	NbF5 (SCXRD)	NbF5 (PXRD)
Empirical formula	NbF5	NbF5
Empirical formula moiety	Nb4F20	Nb4F20
Color and appearance	colorless block	colorless powder
Size (mm^3); capillary diameter (mm)	0.180 × 0.050 × 0.050	0.3
Mol­ecular mass (g mol^−1)	187.91	187.91
Crystal system	monoclinic	monoclinic
Space group (No.)	C2/m (12)	C2/m (12)
Pearson symbol	mC48	mC48
a (Å)	9.4863 (12)	9.62749 (19)
b (Å)	14.2969 (12)	14.4564 (3)
c (Å)	4.9892 (6)	5.12831 (10)
β (°)	97.292 (10)	95.8243 (4)
V (Å^3)	671.19 (13)	710.07 (3)
Z	8	8
Z′	2	2
ρcalc (g cm^−3)	3.719	3.515
λ (Å)	0.71073 (Mo Kα)	1.540596 (Cu Kα1)
T (K)	100	293
μ (mm^−1)	3.561	27.9495
2θ range measured (min, max, increment)	5.182, 60.76, –	6.885, 80.340, 0.015
2θ range refined (min, max)	–	10.005, 80.340
hklmax	−13 ≤ h ≤ 13	0 ≤ h ≤ 8
	−18 ≤ k ≤ 18	0 ≤ k ≤ 12
	−7≤ l ≤7	−4≤ l ≤4
Absorption correction	numerical	cylindrical
Tmax, Tmin	0.7778, 0.7760	–
Rint, Rσ	0.0318, 0.0172	–
Completeness	0.994	–
No. of unique reflections	1048	240
No. of parameters	60	74
No. of restraints	0	0
No. of constraints	0	0
Background parameters	–	20
Profile parameters	–	12^a
Rp, Rwp	–	0.0308, 0.0425
Rp^b, Rwp^b	–	0.0889, 0.0904
RBragg	–	0.0132
S (all data)	1.024	1.77
R(F) [I ≥ 2σ(I), all data]	0.0143, 0.0198	–
wR(F^2) [I ≥ 2σ(I), all data]	0.0315, 0.0323	–
Δρmax, Δρmin (e Å^−3)	0.544, −0.521	–
Notes: (a) refined profile parameters include spherical harmonics of order 4; (b) background-corrected R-factors.

Figure 1 Powder X-ray diffraction pattern and Rietveld refinement of NbF5: measured data points (black dots), calculated diffraction pattern (red line), background (green line) and difference curve (gray). The calculated reflection positions are indicated by the vertical bars at the bottom. Rp = 3.08, Rwp = 4.25%, RBragg = 1.32%, S = 1.77.

The single-crystal structure determination was performed at 100 K and thus resulted in smaller lattice parameters by about 1–3% compared to those determined at room temperature (see Table 1). Otherwise, there are no significant structural differences compared to the RT structure. The slight contraction of the lattice parameters is mainly due to the shortening of the distances between the Nb₄F₂₀ mol­ecules, while the intra­molecular F—Nb distances determined at 100 K differ only insignificantly from those determined at room temperature.

NbF₅ crystallizes in the space group C2/m (No. 12, Pearson code mC48, Wyckoff sequence j⁴i³h) with the lattice parameters a = 9.4863 (12), b = 14.2969 (12), c = 4.9892 (6) Å, β = 97.292 (10)°, Z = 8 at 100 K. NbF₅ crystallizes in the MoF₅ structure type (Edwards et al., 1962; Stene et al., 2018). The structure consists of NbF₅ units forming tetra­meric mol­ecules that can be described by the Niggli (Niggli, 1945) formula ⁰_∞{[NbF_2/2F_4/1]₄}. The structure of the Nb₄F₂₀ mol­ecule in the solid and the crystal structure of the compound are shown in Figs. 2 and 3. Two symmetry-independent niobium atoms reside on Wyckoff positions 4h (site symmetry 2, Nb1) and 4i (site symmetry m, Nb2) and are surrounded octa­hedron-like by six fluorine atoms. By edge-linking via two cis-positioned fluorine atoms, the NbF₆ units form square-like mol­ecules. The atomic distance between the Nb1 atom and the μ-bridg­ing fluorine atoms F4 is 2.0669 (9) Å, while the Nb2—μ-F4 distance is 2.0685 (10). Thus, both Nb—μ-F4 bond lengths are identical within their tripled standard uncertainty. The Nb1—μ-F4—Nb2 bridge is slightly bent by 172.94 (5)°, with the bridging fluorine atoms pointing towards the ring center (Wyckoff position 2c, site symmetry 2/m) of the planar Nb₄F₂₀ rings. The distances between the Nb and the F_trans atoms, Nb1—F6 and Nb2—F3, which are opposite to the μ-bridging F atoms, are 1.8157 (11) and 1.8121 (10) Å; also overlapping within the 3σ criterion. The μ-F—Nb—F_trans angles measure 172.83 (4) and 171.95 (4)°. The terminally bound fluorine ligands in axial positions (F1, F2 and F5) show slightly longer Nb—F bonds of 1.8577 (14), 1.8378 (14), and 1.8468 (10) Å compared to those oriented equatorially (F3 and F6), showing Nb—F distances of 1.8121 (10) and 1.8157 (11) Å. This phenomenon was observed to a similar extent for the structure of MoF₅ (Stene et al., 2018) and can be attributed to the structural trans effect (Coe & Glenwright, 2000; Shustorovich et al., 1975). The Nb atoms in a mol­ecule lie in a flat, nearly square plane and the crystallographic point group of the Nb₄F₂₀ mol­ecule is 2/m (C_2h). The intra­molecular Nb1⋯Nb2 distance is 4.1275 (4) Å while the Nb1⋯Nb2⋯Nb1 angle measures 89.62 (1)°. The distances between diagonally opposite Nb atoms in the ring are 5.8565 (8), and 5.8179 (6) Å. Thus, the four Nb atoms of the Nb₄F₂₀ mol­ecule do not form an ideal square. It is distorted in a diamond shape, which corresponds to a compression along the twofold axis of rotation. An overview of inter­atomic distances and angles in the structure of NbF₅ is given in Tables 2 and 3. The global crystal structure can be approximately described by a cubic close-packing of the fluorine atoms, in which 1/5th of the octa­hedral voids are occupied by Nb atoms in such a way that the Nb₄F₂₀ mol­ecules are obtained (Edwards, 1964; Müller, 2009).

Table 2
Selected inter­atomic distances (Å) for the crystal structure of NbF₅

Nb1—F4	2.0669 (9)	Nb2—F3	1.8121 (10)
Nb1—F5	1.8468 (10)	Nb2—F4	2.0685 (10)
Nb1—F6	1.8157 (11)	Nb1—Nb2	4.1275 (4)
Nb2—F1	1.8577 (14)	Nb1—Nb1^iii	5.8179 (6)
Nb2—F2	1.8378 (14)	Nb2—Nb2^ii	5.8565 (8)
Symmetry codes: (i) x, −y, z; (ii) −x, y, 1 − z, (iii) −x, −y, 1 − z.

Table 3
Selected inter­atomic angles (°) for the crystal structure of NbF₅

F6—Nb1—F6^ii	97.52 (7)	F3—Nb2—F3^i	98.46 (7)
F6—Nb1—F5	95.18 (5)	F3—Nb2—F2	96.78 (5)
F6^i—Nb1—F5	95.61 (5)	F3^i—Nb2—F2	96.78 (5)
F6—Nb1—F5^ii	95.61 (5)	F3—Nb2—F1	94.53 (5)
F6^ii—Nb1—F5^ii	95.18 (5)	F3^i—Nb2—F1	94.53 (5)
F5—Nb1—F5^ii	163.61 (7)	F2—Nb2—F1	162.63 (6)
F6—Nb1—F4^ii	172.83 (4)	F3—Nb2—F4	89.47 (5)
F6^ii—Nb1—F4^ii	89.59 (4)	F3^i—Nb2—F4	171.95 (4)
F5—Nb1—F4^ii	83.14 (5)	F2—Nb2—F4	83.54 (5)
F5^ii—Nb1—F4^ii	84.63 (4)	F1—Nb2—F4	83.43 (4)
F6—Nb1—F4	89.59 (4)	F3—Nb2—F4^i	171.95 (4)
F6^ii—Nb1—F4	172.83 (4)	F3^i—Nb2—F4^i	89.47 (5)
F5—Nb1—F4	84.62 (4)	F2—Nb2—F4^i	83.54 (5)
F5^ii—Nb1—F4	83.14 (5)	F1—Nb2—F4^i	83.43 (4)
F4^ii—Nb1—F4	83.32 (5)	F4—Nb2—F4^i	82.57 (5)
Nb1—Nb2—Nb1^iii	89.62 (1)	Nb1—F4—Nb2	172.94 (5)
Nb2—Nb1—Nb2^ii	90.38 (1)
Symmetry codes: (i) x, −y, z; (ii) −x, y, 1 − z, (iii) −x, −y, 1 − z.

Figure 2 Structure of the Nb4F20 mol­ecule as it appears in the crystal structure of NbF5. Atom labeling in accordance with Edwards et al. (1962). Displacement ellipsoids are shown at the 70% probability level at 100 K. [Symmetry codes: (i) x, −y, z; (ii) −x, y, 1 − z; (iii) −x, −y, 1 − z.]

Figure 3 Crystal structure of NbF5 viewed along the c axis. Displacement ellipsoids are shown at 70% probability level at 100 K.

In addition to X-ray powder diffraction, the bulk phase was also investigated by IR and Raman spectroscopy. The obtained spectra, which are given in the supporting information, agree with those reported in the literature (Preiss & Reich, 1968; Beattie et al., 1969; Papatheodorou et al., 2008), and indicate a phase pure sample.

3. Conclusion

NbF₅ was synthesized from F₂ and Nb metal and obtained as a colorless, phase-pure solid and by sublimation as single crystals. The previous structure model was significantly improved with much more precise atomic coordinates and all atoms refined anisotropically, giving much better bond lengths and angles for the Nb₄F₂₀ mol­ecules.

4. Synthesis and crystallization

Niobium penta­fluoride was synthesized from the elements directly using the apparatus sketched in Fig. 4. Therein, niobium metal sheets (17.28g, 185.9mmol, TANIOBIS GmbH) were loaded in a corundum boat, which was placed inside a tube furnace. One side of the inner corundum tube of the furnace was connected to a metal Schlenk line via a PTFE sealed copper fitting, allowing control of the fluorine supply, as well as evacuating and purging the system with argon. The other side was connected to a U-shaped, 3/4-inch PFA tube via a copper pipe, followed by a PFA gas wash bottle filled with perfluoro polyether (Hostinert 216) and an absorber column filled with soda lime (Carl Roth). The copper pipe, all fittings and valves were surrounded by heating sleeves or wires and heated to 473 K to prevent resublimation of solid NbF₅ inside. Before use, the apparatus was thoroughly baked out and passivated using diluted fluorine (F₂/Ar, 20:80 v/v, Solvay). For the reaction a stream of diluted fluorine (F₂/Ar, 20:80 v/v, approx. 36 mL min⁻¹) was applied and the furnace temperature was set to 473 K. The first single crystals of resublimed NbF₅ were obtained within several minutes in the U-shaped PFA tube. After 16 h the reaction was complete, giving 34.2 g (182.0 mmol, 98%) NbF₅ as a colorless, crystalline solid (see Fig. 5).

Figure 4 Scheme of the apparatus used for the synthesis of NbF5. (a) Connection to a metal Schlenk line for evacuation, purging with inert gas, and fluorine supply, (b) tube furnace, (c) copper pipe surrounded by a heating sleeve, (d) PFA U-trap for product collection equipped with Monel connectors and diaphragm valves (Hoke), (e) PFA gas wash bottle with steel fitting filled with perfluoro polyether, (f) outlet connected to the absorber.

Figure 5 Photo of colorless crystalline NbF5 accumulated in the U-shaped PFA tube during the reaction (left, photo was taken inside a glove box) and corundum boat containing niobium metal: before (top right) and during the reaction (bottom right).

5. Structure determination

5.1 Single crystal structure determination: A crystal of NbF₅ was selected under pre-dried perfluorinated oil (Fomblin YR 1800) and mounted using a MiTeGen loop. Intensity data of a suitable crystal were recorded with an IPDS 2 diffractometer (Stoe & Cie). The diffractometer was operated with Mo Kα radiation (0.71073 Å, graphite monochromator) and equipped with an image plate detector. Evaluation, integration and reduction of the diffraction data was carried out using the X-AREA software suite (X-AREA V1.90; Stoe & Cie, 2020). A numerical absorption correction was applied with the modules X-SHAPE and X-RED32 of the X-AREA software suite. The structures were solved with dual-space methods (SHELXT; Sheldrick, 2015a), and refined against F² (SHELXL) within the ShelXle GUI (Sheldrick, 2015b; Hübschle et al., 2011). All atoms were refined with anisotropic displacement parameters. The highest residual electron density after the final refinement was 0.80 Å distant from atom F6. Representations of the crystal structures were created with the DIAMOND software (Brandenburg & Putz, 2022).

5.2 Powder X-ray diffraction: For powder X-ray diffraction, the sample was ground using a glassy carbon mortar and filled into a quartz capillary with a diameter of 0.3 mm. The powder X-ray pattern was recorded with a StadiMP diffractometer (Stoe & Cie) in Debye-Scherrer geometry. The diffractometer was operated with Cu Kα₁ radiation (1.5406 Å, germanium monochromator) and equipped with a MYTHEN 1K detector.

Rietveld refinements (Rietveld, 1969) were performed using the TOPAS-Academic software (version 7; Coelho, 2018). The structural model derived from single-crystal X-ray diffraction was used as the starting point for the refinement. A shifted Chebyshev polynomial was used to describe the background of the powder pattern, the peak profiles were fitted with a modified Thompson–Cox–Hastings pseudo-Voigt (`TCHZ') function as implemented in TOPAS, and the zero offset was refined. To account for absorption, an intensity correction for cylindrical samples was applied as implemented in TOPAS. A weak preferential orientation of the crystallites was taken into account by means of a fourth-order spherical-harmonics function. The final refinement cycles converged with free refinement of all background, profile, and lattice parameters, including the coordinates of all atoms, the isotropic displacement parameters of the F atoms and anisotropic displacement parameters of the Nb atoms. Further details concerning the Rietveld refinement are given in Table 1 and in the supporting information. Crystal data, data collection and structure refinement details are summarized in Table 4.

Table 4
Experimental details

Crystal data
Chemical formula	Nb4F20
Mr	187.91
Crystal system, space group	Monoclinic, C2/m
Temperature (K)	100
a, b, c (Å)	9.4863 (12), 14.2969 (12), 4.9892 (6)
β (°)	97.292 (10)
V (Å^3)	671.19 (13)
Z	8
Radiation type	Mo Kα
μ (mm^−1)	3.56
Crystal size (mm)	0.18 × 0.05 × 0.05
Data collection
Diffractometer	Stoe IPDSII
Absorption correction	Numerical (X-RED32 and X-SHAPE; Stoe & Cie, Stoe & Cie, 2020)
Tmin, Tmax	0.776, 0.778
No. of measured, independent and observed [I > 2σ(I)] reflections	5819, 1048, 903
Rint	0.032
(sin θ/λ)max (Å^−1)	0.712
Refinement
R[F^2 > 2σ(F^2)], wR(F^2), S	0.014, 0.032, 1.02
No. of reflections	1048
No. of parameters	60
Δρmax, Δρmin (e Å^−3)	0.54, −0.52
Computer programs: X-AREA and X-RED32 (Stoe & Cie, 2020), SHELXT2014/5 (Sheldrick, 2015a), SHELXL2018/3 (Sheldrick, 2015b) and DIAMOND (Brandenburg & Putz, 2020).

Supporting information

Computing details

Niobium(V) fluoride

Crystal data

Nb4F20	F(000) = 688
Mr = 187.91	Dx = 3.719 Mg m^−3
Monoclinic, C2/m	Mo Kα radiation, λ = 0.71073 Å
a = 9.4863 (12) Å	Cell parameters from 8094 reflections
b = 14.2969 (12) Å	θ = 2.6–30.9°
c = 4.9892 (6) Å	µ = 3.56 mm^−1
β = 97.292 (10)°	T = 100 K
V = 671.19 (13) Å^3	Block, colorless
Z = 8	0.18 × 0.05 × 0.05 mm

Data collection

Stoe IPDSII diffractometer	1048 independent reflections
Radiation source: sealed X-ray tube, 12 x 0.4 mm long-fine focus	903 reflections with I > 2σ(I)
Detector resolution: 6.67 pixels mm^-1	Rint = 0.032
rotation method, ω scans	θmax = 30.4°, θmin = 2.6°
Absorption correction: numerical (X-RED32 and X-SHAPE; Stoe & Cie, Stoe & Cie, 2020)	h = −13→13
Tmin = 0.776, Tmax = 0.778	k = −18→20
5819 measured reflections	l = −7→7

Refinement

Refinement on F^2	0 restraints
Least-squares matrix: full	Primary atom site location: dual
R[F^2 > 2σ(F^2)] = 0.014	Secondary atom site location: difference Fourier map
wR(F^2) = 0.032	w = 1/[σ^2(Fo^2) + (0.0196P)^2] where P = (Fo^2 + 2Fc^2)/3
S = 1.02	(Δ/σ)max = 0.001
1048 reflections	Δρmax = 0.54 e Å^−3
60 parameters	Δρmin = −0.52 e Å^−3

Special details

Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
Nb1	0.000000	0.20347 (2)	0.500000	0.01085 (6)
Nb2	0.26047 (2)	0.000000	0.24168 (4)	0.01094 (6)
F1	0.34193 (16)	0.000000	0.6005 (3)	0.0169 (3)
F2	0.13043 (15)	0.000000	−0.0646 (3)	0.0148 (3)
F3	0.36790 (12)	0.09599 (7)	0.1443 (2)	0.0179 (2)
F4	0.12213 (11)	0.09546 (6)	0.3786 (2)	0.01482 (19)
F5	−0.10961 (12)	0.18505 (7)	0.17107 (19)	0.0163 (2)
F6	0.11939 (13)	0.28718 (7)	0.3733 (2)	0.0171 (2)

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
Nb1	0.01143 (11)	0.00957 (9)	0.01173 (9)	0.000	0.00212 (7)	0.000
Nb2	0.01106 (10)	0.01044 (9)	0.01157 (10)	0.000	0.00242 (7)	0.000
F1	0.0168 (7)	0.0195 (6)	0.0141 (6)	0.000	0.0011 (5)	0.000
F2	0.0145 (6)	0.0152 (6)	0.0145 (6)	0.000	0.0013 (5)	0.000
F3	0.0177 (5)	0.0169 (5)	0.0196 (5)	−0.0046 (4)	0.0043 (4)	0.0012 (4)
F4	0.0150 (4)	0.0130 (4)	0.0170 (4)	0.0021 (4)	0.0039 (3)	−0.0013 (4)
F5	0.0176 (5)	0.0163 (4)	0.0146 (4)	−0.0003 (4)	0.0008 (4)	−0.0004 (4)
F6	0.0175 (5)	0.0152 (4)	0.0189 (5)	−0.0026 (4)	0.0036 (4)	0.0018 (4)

Geometric parameters (Å, º)

Nb1—F6	1.8157 (11)	Nb2—F3	1.8121 (10)
Nb1—F6^i	1.8158 (10)	Nb2—F3^ii	1.8122 (10)
Nb1—F5	1.8468 (10)	Nb2—F2	1.8378 (14)
Nb1—F5^i	1.8468 (10)	Nb2—F1	1.8577 (14)
Nb1—F4^i	2.0669 (9)	Nb2—F4	2.0685 (10)
Nb1—F4	2.0669 (9)	Nb2—F4^ii	2.0685 (10)
F6—Nb1—F6^i	97.52 (7)	F3—Nb2—F2	96.78 (5)
F6—Nb1—F5	95.18 (5)	F3^ii—Nb2—F2	96.78 (5)
F6^i—Nb1—F5	95.61 (5)	F3—Nb2—F1	94.53 (5)
F6—Nb1—F5^i	95.61 (5)	F3^ii—Nb2—F1	94.53 (5)
F6^i—Nb1—F5^i	95.18 (5)	F2—Nb2—F1	162.63 (6)
F5—Nb1—F5^i	163.61 (7)	F3—Nb2—F4	89.47 (5)
F6—Nb1—F4^i	172.83 (4)	F3^ii—Nb2—F4	171.95 (4)
F6^i—Nb1—F4^i	89.59 (4)	F2—Nb2—F4	83.54 (5)
F5—Nb1—F4^i	83.14 (5)	F1—Nb2—F4	83.43 (4)
F5^i—Nb1—F4^i	84.63 (4)	F3—Nb2—F4^ii	171.95 (4)
F6—Nb1—F4	89.59 (4)	F3^ii—Nb2—F4^ii	89.47 (5)
F6^i—Nb1—F4	172.83 (4)	F2—Nb2—F4^ii	83.54 (5)
F5—Nb1—F4	84.62 (4)	F1—Nb2—F4^ii	83.43 (4)
F5^i—Nb1—F4	83.14 (5)	F4—Nb2—F4^ii	82.57 (5)
F4^i—Nb1—F4	83.32 (5)	Nb1—F4—Nb2	172.94 (5)
F3—Nb2—F3^ii	98.46 (7)

Symmetry codes: (i) −x, y, −z+1; (ii) x, −y, z.

Selected interatomic angles (°) for the crystal structure of NbF₅

F6^i—Nb1—F5	95.61 (5)	F3^i—Nb2—F2	96.78 (5)
F6—Nb1—F6^ii	97.52 (7)	F3—Nb2—F3^i	98.46 (7)
F6—Nb1—F5	95.18 (5)	F3—Nb2—F2	96.78 (5)
F6^i—Nb1—F5	95.61 (5)	F3^i—Nb2—F2	96.78 (5)
F6—Nb1—F5^ii	95.61 (5)	F3—Nb2—F1	94.53 (5)
F6^ii—Nb1—F5^ii	95.18 (5)	F3^i—Nb2—F1	94.53 (5)
F5—Nb1—F5^ii	163.61 (7)	F2—Nb2—F1	162.63 (6)
F6—Nb1—F4^ii	172.83 (4)	F3—Nb2—F4	89.47 (5)
F6^ii—Nb1—F4^ii	89.59 (4)	F3^i—Nb2—F4	171.95 (4)
F5—Nb1—F4^ii	83.14 (5)	F2—Nb2—F4	83.54 (5)
F5^ii—Nb1—F4^ii	84.63 (4)	F1—Nb2—F4	83.43 (4)
F6—Nb1—F4	89.59 (4)	F3—Nb2—F4^i	171.95 (4)
F6^ii—Nb1—F4	172.83 (4)	F3^i—Nb2—F4^i	89.47 (5)
F5—Nb1—F4	84.62 (4)	F2—Nb2—F4^i	83.54 (5)
F5^ii—Nb1—F4	83.14 (5)	F1—Nb2—F4^i	83.43 (4)
F4^ii—Nb1—F4	83.32 (5)	F4—Nb2—F4^i	82.57 (5)
Nb1—Nb2—Nb1^iii	89.62 (1)	Nb1—F4—Nb2	172.94 (5)
Nb2—Nb1—Nb2^ii	90.38 (1)

Symmetry codes: (i) x, -y, z; (ii) -x, y, 1 - z, (iii) -x, -y, 1 - z.