[Cu(HF₂)(pyz)₂]BF₄ (pyz = pyrazine): long-range magnetic ordering in a pseudo-cubic coordination polymer comprised of bridging HF₂⁻ and pyrazine ligands

Abstract

[Cu(HF₂)(pyz)₂]BF₄ consists of rare μ_1,3 bridging HF₂⁻ anions and μ-pyrazine ligands leading to a 3D pseudo-cubic framework that antiferromagnetically orders below 1.54(1) K.

---

The bifluoride ion, i.e. HF₂⁻, contains a two-coordinate H-atom exhibiting the strongest known hydrogen bond.¹ Molecular orbital theory² and various structural investigations³ suggest that this molecule can be best described as a hydrogen cation that is bridging two fluoride anions as in F⁻⋯H⁺⋯F⁻. Resonance modes consisting of coordinate covalent interactions such as F–H⋯F⁻ and F⁻⋯H–F may also be important.² In the KHF₂ and NH₄HF₂ salts, the H⋯F bond lengths in the HF₂⁻ anion are equivalent at 1.14 Å, which is substantially shorter than the H⋯F hydrogen bond distance of 1.57 Å found in the solid phase of HF.⁴

Pyrazine (pyz) has been extensively utilized as a neutral bridging ligand in the self-assembly of polymeric networks comprised of transition metal cations.^5,6 Furthermore, pyrazine is well known to facilitate long-range antiferromagnetic ordering with a range of exchange coupling constants. Two important examples, Cu(NO₃)₂(pyz)^5a and Cu(ClO₄)₂(pyz)₂,^5b which form one-dimensional (1D) chains and two-dimensional (2D) square lattices, respectively, have been identified as model low-dimensional S = ½ antiferromagnets. In a more recent example, bulk magnetic ordering has been reported in 2D layered Fe(NCS)₂(pyz)₂ which exhibits Ising-like magnetic behavior.⁶

Much of our recent work has focused on the use of pyrazine ligands and a variety of potentially coordinating anions such as dicyanamide (dca).⁷ We have begun to employ the HF₂⁻ anion in a similar role so as to study its coordinative properties and its ability to mediate spin exchange interactions between open-shell metal cations. Our design strategy is purposeful in that we are directly utilizing HF₂⁻ as a building block under mild conditions while previously reported compounds fortuitously generate this anion in situ when aqueous HF is used as a solvent.^3a,8 While pipzH₂[MnF₄(HF₂)] (pipz = piperazine) contains a bridging bifluoride anion, no magnetic interaction was observed.^3a Under appropriate conditions, we believe that HF₂⁻ may in fact be a suitable superexchange ligand and warrants further study. To this end, we have synthesized the novel coordination polymer, [Cu(HF₂)(pyz)₂]BF₄, by aqueous reaction of stoichiometric amounts of copper(II) tetrafluoroborate hydrate, ammonium bifluoride, and pyrazine.‡^,9

The crystal structure of [Cu(HF₂)(pyz)₂]BF₄ has been determined using conventional X-ray diffraction methods.§ Each Cu²⁺ center occupies a 422 symmetry site and lies at the center of a slightly elongated octahedron with four Cu–N and two Cu–F(2) bond distances of 2.040(4) and 2.207(4) Å, respectively. Owing to the high symmetry of the Cu site, all F(2)–Cu–N, N–Cu–N, and F(2)–Cu–F(2′) bond angles are regular, i.e. 90 and 180°. The HF₂⁻ anion features a symmetrically placed hydrogen atom located at a 422 site with H–F and F⋯F bond distances of 1.102(4) and 2.205(4) Å, respectively. Additionally, the H(2)–F(2)–Cu bond angle is 180° while C–N–Cu is 121.3(2)°. All intramolecular bond distances and angles associated with the pyrazine ligand are in agreement with reported values.^5,6

The extended structure of [Cu(HF₂)(pyz)₂]BF₄ is polymeric and consists of a three-dimensional (3D) framework shown in Fig. 1. Infinite 2D [Cu(pyz)₂]²⁺ sheets are formed in the ab-plane that are connected along the c-axis by linearly bridging μ_1,3-HF₂⁻ anions. By comparison, Cu(ClO₄)₂(pyz)₂ has a similar 2D network, however, the ClO₄⁻ anions do not bridge between the 2D [Cu(pyz)₂]²⁺ sheets but rather are terminal groups.^5b The ClO₄⁻ anions pack in between the layers in such a way as to cause the sheets to be mutually staggered. From Fig. 1, it can be seen that the pyrazine ligands in [Cu(HF₂)(pyz)₂]BF₄ are all tilted out of the ab-plane by 59.4(2)° in a pattern consistent with the 4-fold rotation symmetries of the copper sites. A similar rotation of pyz rings along M–pyz–M bridges has been reported in quasi-2D [Cu(NO₃)(pyz)₂]PF₆.¹⁰ Because the Cu⋯Cu separations along the Cu–pyz–Cu and Cu–FHF–Cu linkages are very similar [6.8519(3) and 6.6193(4) Å, respectively], the structure may be described as pseudo-cubic. The BF₄⁻ anions occupy the body-centered positions within each “cubic” pore, with the B-atom resting on a 4̄ 2m site.

Fig. 1 Crystal structure of [Cu(HF₂)(pyz)₂]BF₄ at 295 K. Pyrazine H-atoms and BF₄⁻ anions have been omitted for clarity. Thermal ellipsoids are drawn at the 35% probability level.

The magnetic properties of [Cu(HF₂)(pyz)₂]BF₄ were probed using a Quantum Design MPMS-7 SQUID magnetometer and data collected upon warming between 2 and 300 K. The main plot of Fig. 2 shows χvs.T while χTvs.T is plotted in the inset. At room temperature, the value of χT is 0.367 emuK/mol, which is slightly smaller than the expected value of 0.375 emuK/mol for isolated spin-½ Cu²⁺ cations. Upon cooling, χT decreases continuously until ∼50 K, below which it falls off much more rapidly owing to increasing antiferromagnetic correlations between Cu²⁺ sites.

Fig. 2 Plots of χ (main) and χT (inset) for [Cu(HF₂)(pyz)₂]BF₄.

χ reaches a broad maximum at 5.5 K, which is likely due to short-range magnetic interactions. Below χ_max, the data decrease continuously to the lowest measured temperature of 2 K, reaching a value of 0.017 emu/mol. The observed broad maximum, albeit at a lower temperature, is very similar to that found for 2D Cu(ClO₄)₂(pyz)₂.^5b A Curie–Weiss fit of χ between 6 and 300 K yielded g = 2.048(1) and θ = −8.1(1) K which indicates antiferromagnetic coupling between the spin-½ Cu²⁺ ions in the title compound. Within the limited temperature range of our instrumentation, there was no evidence for long-range magnetic ordering (LRO) above 2 K.

The magnetic Cu 3d_x²−y² orbital of each CuN₄F₂ octahedron lies in the CuN₄ plane so that the spin exchange interactions between neighboring Cu²⁺ ions occur through the σ-bonded pyz ligands. Therefore to a first approximation, the magnetic structure of [Cu(HF₂)(pyz)₂]BF₄ can be described by a 2D square lattice. A least-squares fit of χ to a spin-½ Heisenberg square lattice antiferromagnet¹¹ (H = −JΣS_i·S_j) yields a good agreement for g = 2.078(1) and J/k_B = −2.85(1) K between 6 and 300 K (solid line in Fig. 2). The J-value obtained from the fit is consistent with the mean-field result, |k_BT(χ_max)/JS(S + 1)| = 2.53, which gave |J/k_B| = 2.90.¹² Inclusion of a mean-field correction term (i.e., zJ′) makes a slight improvement to the quality of the fit, however its value is highly sensitive to the temperature range used.

This suggests the interactions between adjacent 2D square nets, which occur via the HF₂⁻ bridges, to be very weak, because each HF₂⁻ bridge that connects two Cu²⁺ ions lies on the 4-fold rotational axis of the Cu 3d_x²−y² magnetic orbital so that the F 2p_z orbitals cannot overlap with this metal orbital. Nevertheless, spin polarization arguments hint that the Cu–F–H–F–Cu pathway may lead to a weak ferromagnetic interaction, i.e. Cu(↑)–F(↓)–H(↑)–F(↓)–Cu(↑), although lower temperature magnetic susceptibility data are needed to confirm this. It should be noted that axial–axial Cu²⁺ connections could in turn promote a weak antiferromagnetic interaction as found in Cu(dca)₂(pym)·CH₃CN for example.¹³ While the sign of the exchange interaction along the HF₂⁻ bridge is presently unclear, LRO is expected to occur in [Cu(HF₂)(pyz)₂]BF₄, albeit likely at low temperature.

In order to search for LRO in [Cu(HF₂)(pyz)₂]BF₄, we have performed zero-field specific heat and muon-spin relaxation (ZF-μ⁺SR) studies. The C_p(T) measurements were conducted on a Quantum Design PPMS equipped with a 14 T magnet and low temperature insert while a sorption cryostat was utilized on the MuSR spectrometer at the ISIS Pulsed Muon Facility (Didcot, UK).

Specific heat data (Fig. 3) show a broad peak in C/T near ∼2.5 K which is likely due to substantial short-range magnetic ordering. Noticeably absent is a λ-anomaly that would signify the onset of LRO. However, close inspection of the data reveals a subtle change in slope around 1.5 K and by plotting d(C/T)/dT as a function of temperature (Fig. 3, inset), a clear peak is resolved at 1.55 K. Theoretically, it has been found that the magnitude of the λ-anomaly depends on the strength of the 3D magnetic interaction. As the ratio J′/J approaches 0.01, the λ-anomaly vanishes.¹⁴ This places an upper limit of ∼0.03 K on the value of J′, however the susceptibility data suggest that J′ may be even smaller.

Fig. 3 Specific heat for [Cu(HF₂)(pyz)₂]BF₄ obtained at H = 0 plotted as C/Tvs.T (main) and d(C/T)/dTvs.T (inset).

Implanted muons act as microscopic magnetometers that can detect small internal magnetic fields in a material.¹⁵ In a magnetically ordered compound below T_N, the muon spins may precess coherently at a frequency, ν, that is proportional to the order parameter of the system. Following this behavior as a function of temperature allows us to probe the magnetic phase transition as shown in Fig. 4. For [Cu(HF₂)(pyz)₂]BF₄, we find two distinct precession frequencies owing to the presence of two muon stopping sites.^15,16 The lines in Fig. 4 are fits of the data to the power law, ν(T) = ν(0)[1 − (T/T_N)^α]^β, where ν₁(0) = 3.26(1) MHz, ν₂(0) = 0.94(1) MHz, T_N = 1.54(1) K, α = 2.2(1), and β = 0.29(2). The β-exponent obtained from the fit is similar to the value of 0.326 associated with 3D Ising behavior.¹⁷ Subsequently, the expression, ν_i = γ_μB_i/2π (γ_μ = 2π × 135.5 MHz/T), can be used to calculate the corresponding internal fields, B₁ and B₂, which are 24.1(1) and 7.0(1) mT, respectively. At present we are unable to unambiguously determine the type of magnetic ordering that occurs although we anticipate a collinear antiferromagnetic ordering due to the weak single-ion anisotropy of Cu²⁺ and the primary Cu–pyz–Cu exchange pathway that is antiferromagnetic.¹⁶

Fig. 4 Temperature-dependence of the two muon precession frequencies found in [Cu(HF₂)(pyz)₂]BF₄.

To determine whether the HF₂⁻ anion may contribute to the observed LRO, the ratio of T_N/J can be used as a rough guide. For [Cu(HF₂)(pyz)₂]BF₄, we obtain T_N/J = 0.53. This value compares well to those of 0.48, 0.56, and 0.58 calculated for quasi-2D Cu(ClO₄)₂(pyz)₂,^5c Cu(C₅H₅NO)₆(BF₄)₂,¹⁸ and (5-MAP)₂CuBr₄,¹⁹ respectively, and is slightly larger than that found for CuF₂·2H₂O (0.42).²⁰ As the T_N/J ratio increases, the more significant the 3D magnetic interactions become.²⁰ Hence, these comparisons seemingly suggest the presence of a non-zero (although weak) exchange interaction along the Cu–F–H–F–Cu pathway. We are currently studying the magnetochemical versatility of this system by substitution of BF₄⁻ for other counter-ions such as ClO₄⁻, PF₆⁻, and others. Additional synthetic and characterization studies are in progress.

This research was supported by an award from Research Corporation. Prof. J. L. Manson gratefully acknowledges helpful discussions with Prof. C. P. Landee and Prof. R. D. Willett. Work at ANL and at NCSU was supported by the BES, Division of Materials Sciences of the U.S. DOE under contract W-31-109-ENG-38 and grant DE-FG02-86ER45259, respectively. We are grateful to P. King (ISIS) for technical assistance. This work was also supported by the EPSRC. TL acknowledges support from the European Commission under the 6th Framework Program through the Key Action: Strengthening the European Research Area, Research Infrastructures, under contract RII3-CT-2003-505925.

Notes and references

1. J. Clark, J. Emsley, D. Jones and R. Overill, J. Chem. Soc., Dalton Trans., 1981, 1219.
2. J. Emsley, Chem. Soc. Rev., 1980, 9, 91.
3. (a) U. Bentrup, K. Harms, W. Massa and J. Pebler, Solid State Sci., 2000, 2, 373; (b) T. R. R. McDonald, Acta Crystallogr., 1960, 13, 113; (c) R. K. Brown, P. L. Johnson, T. J. Lynch and J. M. Williams, Acta Crystallogr., Sect. B, 1978, 34, 1965.
4. (a) S. W. Peterson and H. A. Levy, J. Chem. Phys., 1952, 20, 704; (b) M. Atoji and W. N. Lipscomb, Acta Crystallogr., 1954, 7, 173; (c) S. P. Habuda and Y. P. Gagarinsky, Acta Crystallogr., Sect. B, 1971, 27, 1677.
5. (a) A. Santoro, A. D. Mighell and C. W. Reimann, Acta Crystallogr., Sect. B, 1970, 26, 979; (b) J. Darriet, M. S. Haddad, E. N. Duesler and D. N. Hendrickson, Inorg. Chem., 1979, 18, 2679; (c) T. Lancaster, S. J. Blundell, M. L. Brooks, P. J. Baker, F. L. Pratt, J. L. Manson, F. Xiao, C. P. Landee, A. Chaves, S. Soriano and M. A. Novak, manuscript in preparation.
6. H. N. Bordallo, L. C. Chapon, J. L. Manson, J. Hernandez-Velasco, D. Ravot, W. M. Reiff and D. N. Argyriou, Phys. Rev. B, 2004, 69, 224405.
7. J. L. Manson, Q.-z. Huang, J. W. Lynn, H.-J. Koo, M.-H. Whangbo, R. Bateman, T. Otsuka, N. Wada, D. N. Argyriou and J. S. Miller, J. Am. Chem. Soc., 2001, 123, 162.
8. e.g. J. Vicente, J. Gil-Rubio, D. Bautista, A. Sironi and N. Masciocchi, Inorg. Chem., 2004, 43, 5665.
9. Anal. Calcd (%) for C₈H₉N₄F₆BCu: C, 27.49; H, 2.59; N, 16.03. Found: C, 27.64; H, 2.60; N, 15.78.
10. M. M. Turnbull, A. S. Albrecht, G. B. Jameson and C. P. Landee, Mol. Cryst. Liq. Cryst., 1999, 335, 245.
11. (a) G. S. Rushbrooke and P. J. Wood, Mol. Phys., 1958, 1, 257; (b) C. P. Landee, S. A. Roberts and R. W. Willett, J. Chem. Phys., 1978, 68, 4574.
12. L. J. de Jongh and A. R. Miedema, Adv. Phys., 1974, 23, 1.
13. I. Riggio, G. A. van Albada, D. D. Ellis, A. Spek and J. Reedijk, Inorg. Chim. Acta, 2001, 313, 120 . pym = pyrimidine.
14. P. Sengupta, A. W. Sandvik and R. R. P. Singh, Phys. Rev. B, 2003, 68, 094423.
15. S. J. Blundell, Contemp. Phys., 1999, 40, 175.
16. Muons implant into highly electronegative regions of a material and/or those sites of high symmetry. Owing to the high electronegativity of fluorine, we may be able to use F–μ–F resonances to make dipole field calculations in order to elucidate the type of LRO. Further investigation is needed which is beyond the scope of the present contribution.
17. A. L. Talapov and H. W. J. Blöte, J. Phys. A, 1996, 29, 5727.
18. R. Navarro, H. A. Algra, L. J. de Jongh, R. L. Carlin and C. J. O'Connor, Physica B, 1977, 86, 693.
19. F. M. Woodward, A. S. Albrecht, C. W. Wynn, C. P. Landee and M. M. Turnbull, Phys. Rev. B, 2002, 65, 144412 . 5-MAP = 2-amino-5-methylpyridinium cation.
20. K. Koyama, H. Nobumasa and M. Matsura, J. Phys. Soc. Jpn., 1987, 56, 1553.

---

Footnotes

† Electronic supplementary information (ESI) available: IR, and TGA plots and CIF for [Cu(HF₂)(pyz)₂]BF₄. See DOI: 10.1039/b608791d

‡ NH₄HF₂ (0.2895 g, 5 mmol) and pyrazine (0.7870 g, 10 mmol) are dissolved together in 3 mL of water. To this solution is added, while stirring, a filtered 5-mL aqueous solution containing 1.1860 g (5 mmol) of Cu(BF₄)₂·yH₂O. A dark blue precipitate is obtained immediately. The solid is collected via vacuum filtration and dried in vacuo for 3 hours. Crystals suitable for structural studies are prepared by a similar method although employing 20-fold diluted solutions. Upon slow solvent evaporation, single crystals are obtained after 2–3 weeks. NOTE: ammonium bifluoride is highly corrosive and should be handled with care!

§ Crystal data: C₈H₉N₄F₆BCu, M = 349.54, tetragonal, a = b = 9.6901(4), c = 6.6193(4) Å, U = 621.54(5) ų, T = 298 K, space group P4/nbm (no. 125), Z = 4, μ(Mo-Kα) = 1.825 mm⁻¹, 5131 reflections measured, 423 unique (R_int = 0.0238) which were used in all calculations. The final wR(F²) was 0.1220. CCDC 611880. For crystallographic data in CIF or other electronic format see DOI: 10.1039/b608791d

---