Synthesis and Physical Properties of the Oxofluoride Cu2(SeO3)F2

Article pubs.acs.org/IC

Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

Synthesis and Physical Properties of the Oxofluoride Cu2(SeO3)F2 Eleni Mitoudi-Vagourdi,† Wassilios Papawassiliou,† Silvia Müllner,‡ Aleksander Jaworski,† Andrew J. Pell,† Peter Lemmens,‡ Reinhard K. Kremer,§ and Mats Johnsson*,† †

Department of Materials and Environmental Chemistry, Stockholm University, SE-106 91 Stockholm, Sweden Institute for Physics of Condensed Matter, TU Braunschweig, D-38106 Braunschweig, Germany § Max Planck Institute for Solid State Research, Heisenbergstrasse 1, D-70569 Stuttgart, Germany ‡

S Supporting Information *

ABSTRACT: Single crystals of the new compound Cu2(SeO3)F2 were successfully synthesized via a hydrothermal method, and the crystal structure was determined from single-crystal X-ray diffraction data. The compound crystallizes in the orthorhombic space group Pnma with the unit cell parameters a = 7.066(4) Å, b = 9.590(4) Å, and c = 5.563(3) Å. Cu2(SeO3)F2 is isostructural with the previously described compounds Co2TeO3F2 and CoSeO3F2. The crystal structure comprises a framework of corner- and edge-sharing distorted [CuO3F3] octahedra, within which [SeO3] trigonal pyramids are present in voids and are connected to [CuO3F3] octahedra by corner sharing. The presence of a single local environment in both the 19F and 77Se solid-state MAS NMR spectra supports the hypothesis that O and F do not mix at the same crystallographic positions. Also the specific phonon modes observed with Raman scattering support the coordination around the cations. At high temperatures the magnetic susceptibility follows the Curie−Weiss law with Curie temperature of Θ = −173(2) K and an effective magnetic moment of μeff ∼ 2.2 μB. Antiferromagnetic ordering below ∼44 K is indicated by a peak in the magnetic susceptibility. A second though smaller peak at ∼16 K is tentatively ascribed to a magnetic reorientation transition. Both transitions are also confirmed by heat capacity measurements. Raman scattering experiments propose a structural phase instability in the temperature range 6−50 K based on phonon anomalies. Further changes in the Raman shift of modes at ∼46 K and ∼16 K arise from transitions of the magnetic lattice in accordance with the susceptibility and heat capacity measurements.

■

The aim of the present work was to explore the Cu2+−Se4+− O−F system to search for new compounds. The work resulted in the successful preparation of Cu2(SeO3)F2 as the first compound synthesized in this system. In the related chlorooxohalide system there are several compounds described previously: Cu5Se2O8Cl2,7 Cu9Se4O14Cl6,8 triclinic and monoclinic forms of Cu3Se2O6Cl2,9,10 and Cu5Se4O12Cl2.11 In addition to the single crystal X-ray analysis of Cu2(SeO3)F2, 19F and 77Se solid-state MAS NMR was employed to investigate whether the O and F cosubstitute or if they exclusively occupy specific positions. Raman scattering spectra were collected in order to identify temperature-induced structural instabilities below room temperature. Magnetic susceptibility and heat capacity measurements were conducted to investigate the transitions of the magnetic lattice at low temperatures.

INTRODUCTION

Typically, the p-block elements with ns 2 p 0 electronic configuration (L), such as Sb3+, Se4+, and Te4+ have a onesided or distorted coordination of ligands due to the presence of lone-pair electrons. For example Se4+ exhibits a trigonal pyramidal [LO3] coordination, while Te4+ can exhibit [LO3], [LO3+1], or [LO3+2] coordinations. The stereochemically active lone-pair on L may impede the development of a covalent 3D network and instead lead to the formation of 2D layers and 1D chains with weak connections, or it can reside in a way such that the structure opens up to form channels;1 the diameter of such channels can be comparable to diameters in microporous compounds having catalytic properties.2,3 An additional factor that influences the dimensionality of the structure is that oxygen and halide ions have different bonding preferences; oxygen bonds to both p-block lone-pair elements and transition metals while the halide ions commonly bond only to transition metals if the stoichiometry allows.4,5 Accordingly, exotic properties can be expected for this kind of compounds due to the unusual coordination, and it has been reported that inorganic metal oxofluorides can be utilized in many technological applications such as energy storage, catalysis, and microelectronics.6 © XXXX American Chemical Society

■

EXPERIMENTAL SECTION

Single crystals of the new compound Cu2(SeO3)F2 were synthesized by a hydrothermal method. A mixture of 0.202 g (2 mmol) CuF2 (Aldrich, 99.9%) and 0.110 g (1 mmol) SeO2 (Alfa Aesar, 99.4%) were Received: February 9, 2018

A

DOI: 10.1021/acs.inorgchem.8b00372 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Magnetic susceptibility measurements were conducted with a SQUID magnetometer (MPMS, Quantum Design) at different magnetic fields on a polycrystalline sample (m ≈ 13.4 mg between 2 and 320 K by running field-cooled (fc) and zero-field cooled (zfc) cycles). Heat capacities were collected using the relaxation method (PPMS, Quantum Design) on a 4.9 mg polycrystalline sample intimately immersed in Apiezon N vacuum grease to ensure good thermal coupling. The heat capacities of the sample platform and the vacuum grease were measured in a separate run and subtracted from the total heat capacities.

added together with 2.5 mL deionized water in a 23 mL Teflon lined autoclave. An addition of 20 μL (2−3 drops) of HF acid (40%) in the mixture was found to aid the crystallization. Subsequently, the mixture was heated to 200 °C for 4 days. The synthesis product contained light-green plate-like crystals that were washed thoroughly first with deionized water and afterward with ethanol. The products were initially characterized in a scanning electron microscope (SEM, HITACHI-TM3000) with an energy dispersive spectrometer (EDS, Bruker Quantax 70). A small amount of a blue amorphous impurity phase was found among the single crystals of the new Cu2(SeO3)F2 oxo-fluoride. EDS measurements indicate that this blue impurity phase belongs to the Cu−O−F system. Single-crystal X-ray diffraction data were collected by using a Bruker D8 venture diffractometer mounted with a Mo Kα source with λ = 0.71073 Å. Software provided by the diffractometer manufacturer such as SAINT12 and SADABS13 was used for integration and multiscan absorption correction, respectively. The crystal structure was solved using the software Superflip14 and further refined by applying Jana2006.15 The final refinements of the crystal structure were done by full matrix on F2 and by using the software SHELXL version 2014/ 7.13 All atoms were refined with anisotropic displacement parameters. The structural figures were prepared with the program DIAMOND.16 Powder X-ray diffraction measurements were performed with a Panalytical X’Pert PRO diffractometer, and the collected powder pattern was compared with a simulated pattern of the crystal structure from the single crystal X-ray structure solution in order to confirm the phase purity of the product. A scanning mode of 2θ between 10° and 70° with a step size of 0.0131 was selected for the data collection. The 19F magic-angle-spinning (MAS) nuclear magnetic resonance NMR spectra were collected on a Bruker 600 Avance-III spectrometer operating at a Larmor frequency of 564.68 MHz using a 1.3 mm HX probe at a MAS frequency of 60 kHz. The two one-dimensional spectra were acquired by using the double-adiabatic spin−echo sequence using a pair of tanh/tan short high power adiabatic pulses (SHAPs)17−19 each sweeping through 5 MHz in 50 μs with a radiofrequency (RF) field amplitude of 200 kHz. For the first spectrum, 14336 scans were acquired with a recycle delay of 12 s in order to show all the components of the sample. The second spectrum was acquired with 1572864 scans using a recycle delay of 27 ms in order to emphasize the paramagnetic species of interest. Separation of the isotropic shifts and shift anisotropies was achieved with a twodimensional 19F adiabatic magic-angle turning (aMAT) experiment according to Clément et al.,20 which employed the same SHAPs as in the one-dimensional spectrum. The aMAT recoupling time was 4 rotor periods (66.67 μs), excluding the length of the SHAPs. 32 increments in the indirect dimension were acquired with a spectral width of 600 kHz and 131072 scans per increment. The chemical shifts are referenced to LiF at −204 ppm. The 77Se MAS NMR spectra were collected on a Bruker 400 Avance-III spectrometer at a Larmor frequency of 76.31 MHz with a 2.5 mm HX probe, at a MAS frequency of 30 kHz. As for the 19F spectra the double adiabatic spin−echo and aMAT pulse sequences were employed. The SHAPs swept through 5 MHz in 20 μs with an RF field amplitude of 76 kHz. 491520 scans were acquired using a recycle delay of 10 ms. For the aMAT experiment, the recoupling time was 7 rotor periods (116.67 μs), excluding the lengths of the SHAPs. 32 increments were acquired in the indirect dimension with a spectral width of 480 kHz and 32768 scans per increment. The chemical shifts are referenced to SeO2 at 1357.4 ppm.21 Raman scattering experiments were performed in quasi-backscattering geometry with a solid-state laser at an excitation wavelength of λexc = 532.1 nm and a laser power of Pexc = 300 μW. Selection rules are described using XX and YX, respectively, which indicate parallel and crossed polarization between the incident and the backscattered light within the (110) plane. The crystals were cooled on a coldfinger of a closed-cycle cryostat (Cryomech PT-405) in vacuum. The Raman spectra were collected in the temperature range 6−200 K with a triple spectrometer (Dilor-XY-500) at an acquisition time of 600 s accumulated for 10 times. The spectral resolution is 3−4 cm−1 with an entrance slit width of 100 μm.

■

RESULTS AND DISCUSSION Crystal Structure. The new framework compound, Cu2(SeO3)F2, crystallizes in the orthorhombic space group Pnma with unit cell parameters, a = 7.066(4) Å, b = 9.590(4) Å, and c = 5.563(3) Å. Experimental parameters are presented in Table 1. There is only one crystallographically independent Cu Table 1. Crystal Data and Refinement Parameters for Cu2(SeO3)F2 empirical formula formula weight (g/mol) temperature (K) wavelength (Å) crystal system space group α (Å) b (Å) c (Å) volume (Å3) Z density calc. (g cm−3) F(000) crystal color crystal habit crystal size (mm) θ range for data collection (deg) index ranges

reflections collected independent reflections data/restraints/parameters refinement method goodness-of-fit on F2 final R indices [I > 2σ(I)]a R indices (all data) a

Cu2(SeO3)F2 292.05 295(2) 0.71073 orthorhombic Pnma 7.066(4) 9.590(4) 5.563(3) 377.0(5) 4 5.146 536 green plate 0.4 × 0.6 × 0.3 4.23 to 41.18 −12 ≤ h ≤ 10 −17 ≤ k ≤ 17 −10 ≤ l ≤ 9 9206 1286 [R(int) = 0.0763] 1286/0/40 full-matrix least-squares refinement on F2 1.106 R1 = 0.0403 wR2 = 0.0.0553 R1 = 0.0789 wR2 = 0.0604

R1 = ∑||Fo| − |Fc||/∑|Fo|; wR2 = {∑[w(F2o − F2c )2]/∑[w(F2o)2]}1/2.

and Se atom with the oxidation states +2 and +4, respectively; see Figure 1. These oxidation states both provide charge balance and are supported by bond valence sum (BVS) calculations. The BVS calculations are performed according to Brown and Alternatt, and the results are given in the ESI.22 The Se atom is connected to three O atoms to form a [SeO3] trigonal pyramid with Se − O bond length in the range 1.695(2)−1.753(3) Å where the stereochemically active lone pair completes a distorted tetrahedra. The [SeO3] groups do not commonly polymerize in selenite structures.23 The Cu2+ cation bonds to three O and three F atoms to form a distorted B


Article

Inorganic Chemistry

Figure 3. (a) Chain consisting of distorted [CuO3F3] octahedra connected to each other via edge sharing at two alternating F and two O2 atoms. (b) The chains are connected to each other via corner sharing F atoms to form layers parallel to (010). (c) The layers connect by corner sharing at O1 to build the framework structure of Cu2(SeO3)F2.

Figure 1. Asymmetric unit and equivalents of Cu2(SeO3)F2. (i) x, 1.5 − y, z; (ii) 0.5 + x, y, 0.5 − z; (iii) 0.5 − x, 1 − y, 0.5 + z; (iv) 1 − x, 1 − y, −z; (v) −0.5 + x, y, 0.5 − z.

[CuO3F3] octahedron. The Cu−O and Cu−F bond lengths are in the ranges 1.978(1)−2.260(2) Å and 1.933(2)−2.380(2) Å, respectively. An overview of the crystal structure is given in Figure 2.

signals at −147 and −83 ppm stem from impurities formed during the synthesis process, namely a blue amorphous phase. Under these experimental conditions these impurity phases dominate the NMR spectrum as the resonances are sharper and relax more slowly than the signals due to the paramagnetic material. To emphasize the latter, state-of-the-art methods in paramagnetic solid state NMR are employed, which are optimized for signals with large shifts and shift anisotropies and very short relaxation times.19 In Figure 4b, the 19F MAS spectrum recorded with a shorter recycle delay 27 ms is shown, in order to highlight the 19F nuclei directly bonded to Cu2+, which are generally very difficult to observe.24 Here, the sharper resonances due to the paramagnetic impurities are saturated and so appear with intensities that are considerably reduced. Nevertheless, the overlapping spinning sideband manifolds still hinder the identification of the isotropic shift of the broad resonance due to the paramagnetic phase, which has a large Fermi-contact shift contribution due to spin transfer of the unpaired 3d electron into the F s orbital through the Cu−F bond25,26 and a broad spinning-sideband manifold due to the longer-range through-space spin-dipolar interaction with these same unpaired electrons. This problem is resolved in the 19F aMAT spectrum presented in Figure 5.20 The indirect dimension reveals the isotropic spectrum, which is correlated with the conventional MAS spectrum. Here, a broad spinning sideband manifold with an isotropic peak at 36.9 ppm is observed in addition to the isotropic peaks at −193, −147.3, and −83 ppm which were observed previously in Figure 4a. The signal at 36.9 ppm is attributed to 19F with a direct bond to Cu2+ in the structure. To support this assignment, the spin− lattice relaxation times T1 of these 19F local environments were measured, with the resonance at 36.9 ppm having a comparatively short T1 = of 5.4 ms due to the paramagnetic relaxation enhancement (PRE) from the unpaired 3d electron of Cu2+. By contrast the other components have longer T1values of 2.5 s. The 77Se MAS NMR spectrum is presented in Figure 6. The isotropic shift of 3890 ppm is located outside the typical window between +2400 ppm and −900 ppm of shifts from diamagnetic systems.27 The deviation is again attributed to the Fermi-contact shift due to the unpaired electrons of nearby Cu2+ ions, this time via a two-bond spin transfer via Cu−O−Se.

Figure 2. The crystal structure of Cu2(SeO3)F2 is made up of edgeand corner-sharing [CuO3F3] octahedra, forming an open framework where the channels extending along [100] are occupied by Se4+ and its stereochemically active lone pair.

The distorted [CuO3F3] octahedra connect to each other via edge sharing alternating at O2 and F1 atoms to form [CuO2F2]∞ chains extending along [001], as shown in Figure 3a. The chains further connect to each other via corner sharing so that each octahedron connects to four other octahedra to form [CuO2F]∞ layers parallel to (010); see Figure 3b. The layers are then connected via corner sharing at the common O1 in order to form the 3D open framework structure as shown in Figure 3c. The presence of the stereochemically active lonepairs of Se4+ and the way they reside in the structure enhance the formation of channels along the [100] axis. The 19F solid-state MAS NMR spectrum presented in Figure 4a was acquired with a long recycle delay of 12 s, so that the integrals are quantitative. The spectrum exhibits three main resonances which appear at −193, −147.3, and at −83 ppm. These can be attributed to diamagnetic impurities in the sample. More specifically the signal at −193 ppm is attributed to hydrogen fluoride adsorbed on the surface,24 whereas the C


Article

Inorganic Chemistry

Figure 4. 19F MAS spectra of the Cu2(SeO3)F2 crystals with recycle delays of (a) 12 s and (b) 27 ms.

The resonance also exhibits a broad spinning-sideband pattern that is due to the anisotropic spin-dipolar interaction with the unpaired electrons. The aMAT spectrum shown in Figure S1 agrees with those results, exhibiting a single isotropic shift with a broad spinning sideband pattern. The presence of a single local environment in the 19F and 77Se NMR spectra of the paramagnetic species is in agreement with the proposed crystal structure from the single crystal X-ray diffraction experiments. We can compare the new material Cu2(SeO3)F2 with previously described analogues. For instance, the two previously described compounds Co2TeO3F2 and Co2SeO3F2 are isostructural with the new Cu2(SeO3)F2, with the former also being synthesized by hydrothermal methods.5 The Co−O and Co−F bond lengths in Co2SeO3F2 are in the range 2.073(5)−2.105(5) Å and 2.039(4)−2.130(4) Å, respectively. The Se−O bond lengths are within the ranges 1.692(5)− 1.747(7) Å. However, while the two compounds Co2TeO3Cl2 and Co 2 TeO 3 Br 2 have analogous formulas with the new Cu2(SeO3)F2 compound,28 their crystal structures are completely different and the Cl and Br atoms are in terminating positions, rather than bridging, and protruding out from layers that are only connected by weak van der Waal interactions. The differences in electronegativity between F and Cl/Br thus play an important role in this family of oxohalide compounds in determining if the compound is layered with weak connections in between the layers or a 3D framework structure. However, in the present compound F acts like Cl/Br and bonds to only Cu2+ and not to Se4+. Magnetic Susceptibility and Heat Capacity. Figures 7a−b show the temperature dependence of the magnetic susceptibility of Cu2(SeO3)F2 collected in a magnetic field of 1 T. The magnetic susceptibility is characterized by a broad shoulder centered at ∼125 K and a sharp spike at ∼44 K. The shape of this spike depends on the thermal history and exhibits a splitting of the field-cooled and zero-field cooled susceptibilities which is more pronounced for smaller field but disappears above 1 T (see Figure 7b). The spike is ascribed to the onset of long-range magnetic ordering possibly into a spin-canted state with a weak ferromagnetic component. An additional shallow peak appears at ∼16 K which may indicate some reorientation of the magnetic moments in the ordered

Figure 5. 19F aMAT spectrum of the Cu2(SeO3)F2 crystals. The isotropic shift from the paramagnetic phase is observed at 36.9 ppm. The signals observed at −193 ppm and at −147.3 ppm are due to impurities. The inset presents one-dimensional slices taken at the isotropic peak positions. The isotropic shift from the paramagnetic phase is observed at 36.9 ppm. The signals observed at −193 ppm and at −147.3 ppm are due to hydrogen fluoride adsorbed on the surface and a blue amorphous phase, respectively.

Figure 6. 77Se MAS NMR spectrum of the Cu2(SeO3)F2 crystals, showing a single resonance with an isotropic shift of 3890.0 ppm.

D


Article

Inorganic Chemistry

Figure 7. (a) Field-cooled (fc) and zero-field cooled (zfc) magnetic susceptibility of Cu2(SeO3)F2 per formula unit measured in a magnetic field of 1 T and (inset) inverse magnetic susceptibility. The splitting between zfc and fc susceptibility disappears for magnetic fields above 1 T. (b) The history and field dependence of the spike near 44 K.

electrons in closed electronic shells and from the van Vleck susceptibility of the Cu2+ cations. The latter typically amounts to contributions of the order of 100 × 10−6 cm3/mol per Cu2+ cation and often just compensates the diamagnetic contributions.29 Fitting the experimental data at high temperatures to eq 1 was only possible if the g-factor or χ0 were fixed. Assuming a g-factor g ∼ 2.25 consistent with g-factors typically found for Cu2+ cations,30 the best-fit value of the Curie−Weiss temperature, Θ, is −173(2) K, indicating predominantly antiferromagnetic spin exchange between the Cu2+ magnetic dipole moments. The temperature independent susceptibility, χ0, was determined to be 485(25) × 10−6 cm3/mol, which is somewhat too large to account for the van Vleck and the diamagnetic susceptibilities of the closed shell atoms. For the latter, using Pascal’s increments,31 Cu2+: −11 × 10−6 cm3/mol, Se4+: −8 × 10−6 cm3/mol, F−: −11 × 10−6 cm3/mol, and O2−: −12 × 10−6 cm3/mol, one obtains χdia = −88 × 10−6 cm3/mol, which leaves a positive susceptibility contribution from the van Vleck susceptibilities of the order of χ0 ≈ 100 × 10−6 cm3/mol. This is substantially smaller than the result found from the Curie−Weiss fits. The origin of this discrepancy is currently not clear. A closer inspection of the crystal structure gives an essential hint to the origin of the broad feature in the magnetic susceptibility temperature curve at 125 K which we earlier have tentatively ascribed to short-range magnetic ordering. Figure 9 displays the connectivity of the Cu atoms via F and O atoms to a corrugated isolated two-leg ladder array. The Cu−F−Cu bonding angle amounts to 130.60° and the Cu−O−Cu bonding angle to 123.03°, allowing antiferromagnetic spinexchange interaction.32 Accordingly, we have analyzed the magnetic susceptibility of Cu2(SeO3)F2 assuming a Heisenberg Hamiltonian for a spin S = 1/2 chain with nearest-neighbor antiferromagnetic spinexchange according to eq 3

state. Both of these anomalies (the latter shallow peak and the sharp spike) have counterparts in the heat capacity displayed in Figure 8 supporting magnetic ordering and/or magnetic reorientation at these temperatures.

Figure 8. Molar heat capacity divided by temperature per formula unit of Cu2(SeO3)F2. The vertical red arrows highlight the anomalies which correspond to anomalies in the magnetic susceptibilities.

At sufficiently high temperatures (T > 200 K) the magnetic susceptibility of Cu2(SeO3)F2 follows a Curie−Weiss law according to χmol (T ) =

C + χ0 T−Θ

(1)

where C is the Curie constant given by C=

NAg 2μB2 S(S + 1) 3kB

(2)

and NA is Avogadro’s constant, g is the electronic g-factor, and μB is the Bohr magneton. S is the spin of a Cu2+ cation associated with the electronic configuration (3d9), i.e. S = 1/2, kB is the Boltzmann constant, and Θ is the Curie−Weiss temperature. The second term in eq 1 arises from the temperature independent contributions to the magnetic susceptibility due to the diamagnetic susceptibilities of the

/ chain = Jchain ∑ Sk⃗ Sk⃗ + 1 k

(3)

as well as a Hamiltonian for a two-leg ladder with nearestneighbor antiferromagnetic Heisenberg spin-exchange according to E


Article

Inorganic Chemistry

Figure 9. Section of the crystal structure of Cu2(SeO3)F2 highlighting the connectivity of the magnetic Cu2+ cations via the F− and O2− anions to chains (Jrung = 0) or to a two-leg ladder with the spin-exchange along the ladders denoted as Jleg and the exchange connecting the legs as Jrung. Se atoms are omitted.

/ ladder = Jleg

Table 2. Spin-Exchange Parameters, Jchain and Jleg, along the Chains and Legs, respectively, and along the Rungs, Jrung, and Temperature Independent Susceptibilities, χ0, as Obtained from Fits of Eqs 3 and 4 to the Experimental Data above 100 Ka

∑ (S1⃗ kS1⃗ k+ 1 + S2⃗ kS2⃗ k+ 1) + Jrung ∑ S1⃗ kS2⃗ k k

k

(4)

For both Hamiltonians very precise results for the temperature dependence of the magnetic susceptibility have been obtained by Quantum Monte Carlo simulations and transfermatrix density-matrix renormalization group calculations. These results have been encoded into Padé approximants by Johnston et al. the coefficients of which are listed in their articles.33,34 The predictions for the chain and the two-leg ladder were fitted to the experimental data between 100 and 320 K. For consistency with the Curie−Weiss fit but also to avoid numerical instabilities especially in the case of the two-leg ladder fit, we fixed the g-factor in both cases to 2.25. Figure 10 shows the comparison of the experimental data with the theoretical results. The fitted J-values are summarized in Table 2.

model

Jleg/chain (K)

chain ladder

177(3) 165(10)

Jrung (K)

g-factor

χ0 (10−6 cm3/mol)

ΘMF (K)

71(40)

2.25 2.25

243(5) 299(20)

89 100

a

The rightmost column lists the Curie Weiss temperature expected from molecular field theory according to eqs 5 and 6a,6b.

As can be seen in Figure 10, both models predict the same temperature dependence for the magnetic susceptibility above ∼70 K and match the experimental data above 100 K. The fits indicate antiferromagnetic spin-exchange parameters along the chains or legs of ∼170 K. According to the two-leg ladder fit, exchange coupling between the chains/legs is also antiferromagnetic but by a factor of 3 smaller than the interchain exchange. The temperature independent susceptibilities, χ0, are smaller by a factor of 1.5 to 2 than the values obtained from the Curie−Weiss fits and more consistent with the expected value (see the discussion above). Also given in Table 2 are the Curie−Weiss temperatures, ΘMF, expected from molecular field theory. They have been calculated according to 1 ΘMF = − S(S + 1) ∑ zmJm 3 (5) m where zm is the number of mth nearest neighbors of a given atom and Jm is the spin exchange interaction between mth neighbors.35 With S = 1/2 eq 5 gives for a chain with two nearest neighbors:

Figure 10. (○) Magnetic susceptibility of Cu2(SeO3)F2 together with fits assuming a spin S = 1/2 chain with uniform nearest-neighbor Heisenberg spin exchange (solid blue line) and a spin S = 1/2 two-leg ladder with Heisenberg spin exchange (solid red line). The inset highlights the comparison in an enlarged scale.

Θchain MF = −

Jchain

(6a) 2 and for the ladders with two nearest neighbors along the legs and one along the rungs: F


Article

Inorganic Chemistry 1 Θ ladder = − (2Jleg + Jrung ) MF 4

of modes into one at lower temperature gives evidence for a structural phase transition. The heat capacity measurements confirm that apart from the phase transition from the paramagnetic into the antiferromagnetic phase transition at 42 K, a second magnetic phase transition exists at 16 K. This additional phase transition at 16 K is of first order, which implies a structural phase transition. There exists a broad background in the frequency range between 100 and 1100 cm−1 with a maximum at about Emax ≈ 600 cm−1 and only a weak dependence on light polarization. This continuum of scattering is a candidate for magnetic Raman scattering. The integrated intensity of the continuum is determined by fitting Lorentzian lines to the phonon excitations and subtracting them from the experimental data, as shown in Figure 12. The continuum increases in intensity

(6b)

Here, we have considered exclusively spin exchange interactions within the chains and along the legs and the rungs of the ladders disregarding any interchain or interladder spin exchange. With the spin exchange values obtained from the fits of the chain or the ladder model (cf. Table 2), the molecular field values for the Curie−Weiss temperature are by about 50% too low compared to the result of the Curie−Weiss fit. The difference provides a first estimate of the interchain or interladder exchange which finally leads to long-range order below ∼44 K. Raman Scattering. The irreducible representation of Raman-active vibrational modes is given by Γ = 13Ag + 11B1g + 13B2g + 11B3g, out of which the Ag and B1g modes are visible with light polarizations within the (110) plane (point group D2h). The B2g and B3g modes need light polarization with a component parallel to the crystallographic c-axis. The calculated 13Ag + 11B1g modes coincide nicely with the 24 observed modes in the ab-plane. An overview of the phonons, which are distributed within a frequency range of 80−900 cm−1, is seen in Figure 11.

Figure 12. Two-magnon Raman scattering process. The intensity of this scattering process increases with decreasing temperatures, and its maximum energy position is at approximately 600 cm−1. The inset shows the integrated intensity of the continuum as a function of temperature.

with decreasing temperatures and is in accordance with the behavior of a two-magnon scattering process. Furthermore, several phonons close to Emax show a deviation from a Lorentzian line width. Such an effect is denoted as a Fano line shape and often observed in low-dimensional spin systems with spin phonon coupling.38 Emax and the line width give information about the dynamics of the scattering process. Within a mean-field approximation, the typical energy of the spin-flip processes can be estimated as follows: Emax = 3Jtwo‑magnon yielding an exchange coupling energy Jtwo‑magnon = 290 K. This value is larger than but of the same order of magnitude as the estimate from magnetic susceptibility (J = 170 K). We attribute this to a larger weight than spectroscopic probes give compared to thermodynamic probes to larger couplings in a distribution of exchange paths.

Figure 11. Overview of the phonon spectra from a Cu2(SeO3)F2 crystal at 6 and 200 K in XX polarization. The inset shows the temperature dependence of the weak anharmonicity of the phonon at 160 cm−1 which combines with the neighboring phonons at T = 6 K.

The observed phonons distribute into four frequency ranges belonging to the building units. These specific phonon modes originate from the [SeO3] and the [CuO3F3] groups and are distributed as follows: 80−170 cm−1 [CuO3F3], 170−290 cm−1 [CuO3F3], 300−570 cm−1 [CuO3F3], and 600−900 cm−1 [SeO3], respectively. One B1g and two Ag modes are predicted for the Se atom at the Wyckoff position 4c. Due to the presence of the lone-pair electrons in Se 4+ , a high scattering intensity of the corresponding phonon is expected. [SeO3F]−[K]+ is known to have an A1 mode at 899 cm−1.36 It is therefore probable that the three phonons with the highest Raman shift at 835.3, 826.0, and 690.1 cm−1 correspond to the [SeO3] group as they also have a high scattering intensity. [CuF4] and [CuO4] groups have modes at the low energy Raman shift approximately at 300 to 500 cm−1.37 This supports that the low energy phonons originate from the [CuO3F3] groups. The figure inset shows how three modes at approximately 155 cm−1 combine into a single mode at 6 K. The combination

■

CONCLUSIONS The new oxo-fluoride compound Cu2(SeO3)F2 has been synthesized via a hydrothermal method using CuF2 and SeO2 as starting materials. The crystal structure was determined from single crystal X-ray diffraction data. The compound crystallizes in the orthorhombic system, space group Pnma. The Cu2+ cations adopt a distorted [CuO3F3] octahedral coordination and Se4+ a [SeO3] trigonal pyramid forming a distorted [SeO3E] tetrahedron when including the stereochemically G


Article

Inorganic Chemistry

RTG 1952/1, the NTHSchool for Contacts in Nanosystems, and Quanomet. PL and SM thank A. Möller for discussions. We thank E. Brücher and G. Siegle for expert experimental assistance.

active lone-pair (E). The [CuO3F3] octahedra are connected to each other via edge and corner sharing to build a framework structure with channels along [100] where the Se4+ cations reside. The proposed atom coordination’s around the cations found from the X-ray study are supported by the 19F and 77Se solid-state MAS NMR investigations. The different bonding preferences of O and F have resulted in unusual building blocks of the structure and have led to an open framework with a channel diameter of 5.2 Å. This kind of material may have catalytic properties. Cu2(SeO3)F2 exhibits long-range antiferromagnetic order below ∼44 K, clearly evidenced by magnetic susceptibility and heat capacity measurements. An additional anomaly in the heat capacity and the magnetic susceptibility at 16 K is ascribed to a reorientation of the magnetic structure. The high-temperature susceptibility data evidence antiferromagnetic short-range order which has been analyzed in terms of a spin chain with uniform nearest neighbor Heisenberg exchange or a two-leg ladder model. Our analyses strongly support 1D magnetism with antiferromagnetic spin exchange of the order of ∼170 K. However, interchain or interladder spin exchange is sizable and leads to long-range antiferromagnetic order below ∼44 K. The Raman scattering spectra also indicate a structural instability most probably related to the magnetic lattice. Combination of modes into one is observed at 42 and 16 K. Both modes correspond to low energy phonons originating from the [CuO3F3] groups.

■

■

(1) Hu, S.; Mace, A.; Johnsson, M.; Law, J. M.; Gnezdilov, V.; Lemmens, P.; Tap, J.; Möller, A. Crystal structure and magnetic properties of the S = 1/2 quantum spin system Cu7(TeO3)6F2 with Mixed Dimensionality. Inorg. Chem. 2014, 53, 7661−7667. (2) Lok, B. M.; Messina, C. A.; Patton, R. L.; Gajek, R. T.; Cannan, T. R.; Flaningen, E. M. Silicoaluminophosphate Molecular Sieves: Another New Class of Microporous Crystalline Inorganic Solids. J. Am. Chem. Soc. 1984, 106, 6092−6093. (3) Davis, M. E. Ordered porous materials for emerging applications. Nature 2002, 417, 813−821. (4) Becker, R.; Prester, M.; Berger, H.; Lin, P. H.; Johnsson, M.; Drobac, D.; Zivkovic, I. Crystal structure and magnetic properties of two new cobalt selenite halides: Co5(SeO3)4X2 (X= Cl, Br). J. Solid State Chem. 2007, 180, 1051−1059. (5) Hu, S.; Johnsson, M. Synthesis and crystal structure of two synthetic oxofluoride framework compounds − Co2TeO3F2 and Co2SeO3F2. Dalton Trans. 2012, 41, 12786−12789. (6) Leblanc, M.; Maisonneuve, V.; Tressaud, A. Crystal Chemistry and Selected Physical Properties of Inorganic Fluorides and OxideFluorides. Chem. Rev. 2015, 115, 1191−1254. (7) Galy, J.; Bonnet, J. J.; Andersson, S. The Crystal Structure of a New Oxide chloride of copper(II) and selenium(IV): Cu5Se2O8Cl2. Acta Chem. Scand. 1979, A33, 383−389. (8) Krivovichev, S. V.; Filatov, S. K.; Semenova, T. F.; Rozhdestvenskaya, I. V. Crystal chemistry of inorganic compounds based on chains of oxocentered tetrahedra I. Crystal structure of chloromenite, Cu9O2(Se03)4Cl6. Z. Kristallogr. 1998, 213, 645−649. (9) Millet, P.; Bastide, B.; Johnsson, M. Cu3(SeO3)2Cl2: a new oxochloride of copper(II) and selenium(IV). Solid State Commun. 2000, 113, 719−723. (10) Becker, R.; Berger, H.; Johnsson, M. Monoclinic Cu3(SeO3)2Cl2: an oxohalide with an unusual CuO4Cl trigonal− bipyramidal coordination. Acta Crystallogr. 2007, C63, i4−i6. (11) Zhang, D.; Berger, H.; Kremer, R.; Wulferding, D.; Lemmens, P.; Johnsson, M. Synthesis, Crystal Structure, and Magnetic Properties of the Copper Selenite Chloride Cu5(SeO3)4Cl2. Inorg. Chem. 2010, 49, 9683−9688. (12) APEX3, Version 2016.1-0; Bruker AXS Inc: Madison, WI, 2012. (13) Sheldrick, G. M. Crystal structure refinement with SHELXL. Acta Crystallogr., Sect. C: Struct. Chem. 2015, C71, 3−8. (14) Palatinus, J. L.; Chapuis, G. SUPERFLIP− a computer program for the solution of the crystal structures by charge flipping in arbitrary dimensions. J. Appl. Crystallogr. 2007, 40, 786−790. (15) Petricek, V.; Dusek, M.; Palatinus, L. Crystallographic Computing System Jana2006: General Features. Z. Kristallogr. Cryst. Mater. 2014, 229, 345−352. (16) Bergerhoff, G.; Berndt, M.; Brandenburg, K. Evaluation of Crystallographic Data with the Program DIAMOND. J. Res. Natl. Inst. Stand. Technol. 1996, 101, 221−225. (17) Kervern, G.; Pintacuda, G.; Emsley, L. Fast adiabatic pulses for solid-state NMR of paramagnetic systems. Chem. Phys. Lett. 2007, 435, 157−162. (18) Hwang, T. L.; van Zijl, P. C.; Garwood, M. Fast Broadband Inversion by Adiabatic Pulses. J. Magn. Reson. 1998, 133, 200−203. (19) Pell, A. J.; Pintacuda, G. Broadband solid-state MAS NMR of paramagnetic systems. Prog. Nucl. Magn. Reson. Spectrosc. 2015, 84−85, 33−72. (20) Clement, R. J.; Pell, A. J.; Middlemiss, D. S.; Strobridge, F. C.; Miller, J. K.; Whittingham, M. S.; Emsley, L.; Grey, C. P.; Pintacuda, G. Spin-transfer pathways in paramagnetic lithium transition-metal phosphates from combined broadband isotropic solid-state MAS

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.8b00372. Tables of atomic coordinates, bond lengths, bond valence calculations, and 77Se aMAT spectrum (PDF) Accession Codes

CCDC 1823718 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing data_ [email protected], or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Reinhard K. Kremer: 0000-0001-9062-2361 Mats Johnsson: 0000-0003-4319-1540 Notes

The authors declare no competing financial interest. Further details on the crystal structural investigations can be obtained from the Fachinformationszentrum Karlsruhe, Abt. PROKA, 76344 Eggenstein-Leopoldshafen, Germany (fax +497247-808-666; E-mail: crysdata@fizkarlsruhe.de) on quoting the depository number: CSD-434121.

■

ACKNOWLEDGMENTS This work has been carried out with support from the Swedish Research Council (VR) grant 2016-03441 and 2014-3931, Deutsche Forschungsgemeinschaft DFG LE967/16-1, DFGH


Article

Inorganic Chemistry NMR spectroscopy and DFT calculations. J. Am. Chem. Soc. 2012, 134, 17178−17185. (21) Griffin, J. M.; Knight, F. R.; Hua, G.; Ferrara, J. S.; Hogan, S. W. L.; Woollins, J. D.; Ashbrook, S. E. 77Se Solid-State NMR of Inorganic and Organoselenium Systems: A Combined Experimental and Computational Study. J. Phys. Chem. C 2011, 115, 10859−10872. (22) Brown, I. D.; Alternatt, D. Bond-Valence Parameters Obtained from a Systematic Analysis of the Inorganic Crystal Structure Database. Acta Crystallogr., Sect. B: Struct. Sci. 1985, B41, 244−247. (23) Mao, J.-G.; Jiang, H.-L.; Kong, F. Structures and Properties of Functional Metal Selenites and Tellurites. Inorg. Chem. 2008, 47, 8498−8510. (24) Hua, X.; Robert, R.; Du, L. S.; Wiaderek, K. M.; Leskes, M.; Chapman, K. W.; Chupas, P. J.; Grey, C. P. Comprehensive Study of the CuF2 Conversion Reaction Mechanism in a Lithium Ion Battery. J. Phys. Chem. C 2014, 118, 15169−15184. (25) McConnell, H. M.; Robertson, R. E. Isotropic Nuclear Resonance Shifts. J. Chem. Phys. 1958, 29, 1361−1365. (26) Kim, J.; Middlemiss, D. S.; Chernova, N. A.; Zhu, B. Y. X.; Masquelier, C.; Grey, C. P. Linking Local Environments and Hyperfine Shifts: A Combined Experimental and Theoretical 31P and 7Li Solid-State NMR Study of Paramagnetic Fe(III) Phosphates. J. Am. Chem. Soc. 2010, 132, 16825−16840. (27) Duddeck, H. Selenium-77 nuclear magnetic resonance spectroscopy. Prog. Nucl. Magn. Reson. Spectrosc. 1995, 27, 1−323. (28) Becker, R.; Berger, H.; Johnsson, M.; Prester, M.; Mahronic, Z.; Miljiak, M.; Herak, M. Crystal structure and magnetic properties of Co2TeO3Cl2 and Co2TeO3Br2. J. Solid State Chem. 2006, 179, 836− 842. (29) Banks, M. G.; Kremer, R. K.; Hoch, C.; Simon, A.; Ouladdiaf, B.; Broto, J. M.; Rakoto, H.; Lee, C.; Whangbo, M.-H. Magnetic ordering in the frustrated Heisenberg chain system cupric chloride CuCl2. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 80, 024404-1− 24404-15. (30) Carlin, R. L. Magnetochemistry; Springer − Verlag: Berlin, Heidelberg, New York, Tokyo, 1986. (31) Selwood, P. W. Magnetochemistry, 2nd ed.; Interscience Publishers, Inc.: New York, 1956. (32) Hay, P. J.; Thibeault, J. C.; Hoffmann, R. J. Orbital Interactions in Metal Dimer Complexes. J. Am. Chem. Soc. 1975, 97, 4884−4899. (33) Johnston, D. C.; Kremer, R. K.; Troyer, M.; Wang, X.; Klümper, A.; Bud’ko, L.; Panchula, A. F.; Canfield, P. C. Thermodynamics of spin S = 1/2 antiferromagnetic uniform and alternating-exchange Heisenberg chains. Phys. Rev. B: Condens. Matter Mater. Phys. 2000, 61, 9558−9606. (34) Johnston, D. C.; Troyer, M.; Miyahara, S.; Lidsky, D.; Ueda, K.; Azuma, M.; Hiroi, Z.; Takano, M.; Isobe, M.; Ueda, Y.; Korotin, M. A.; Anisimov, V. I.; Mahajan, A. V.; Miller, L. L. Magnetic Susceptibilities of Spin-1/2 Antiferromagnetic Heisenberg Ladders and Applications to Ladder Oxide Compounds. 2000, arXiv:cond-mat/0001147v1. (35) Smart, J. S. Effective Feld Theories of Magnetism; W. B. Saunders Company: Philadelphia & London, 1966. (36) Weidlein, J.; Müller, U.; Dehnicke, K. Schwingungsfrequenzen I. Hauptgruppen-elemente G.; Thieme Verlag: Stuttgart, 1981. (37) Weidlein, J.; Müller, U.; Dehnicke, K. Schwingungsfrequenzen II. Nebengruppen-elemente. G.; Thieme Verlag: Stuttgart, 1986. (38) Lemmens, P.; Güntherodt, G.; Cros, C. Magnetic Light Scattering in low-dimensional Quantum Spin Systems. Phys. Rep. 2003, 375, 1−103.

I


Synthesis and Physical Properties of the Oxofluoride Cu2(SeO3)F2

Recommend Documents