Crystal Structure of LaSr3Fe3O8 (OH) 2· x H2O

Jul 18, 2016 - La(NO3)3·6H2O (Sigma-Aldrich, >99.9%), Sr(NO3)2 (Sigma-Aldrich, > ..... Battle, P. D.; Green, M. A.; Laskey, N. S.; Millburn, J. E.; R...
0 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/IC

Crystal Structure of LaSr3Fe3O8(OH)2·xH2O Vegar Øygarden,*,† Helmer Fjellvåg,† Magnus H. Sørby,‡ and Anja O. Sjåstad*,† †

Department of Chemistry, Centre for Materials Science and Nanotechnology, University of Oslo, P.O. Box 1033, N-0315 Oslo, Norway ‡ Department of Physics, Institute for Energy Technology, N-2007 Kjeller, Norway S Supporting Information *

ABSTRACT: The crystal structure of the hydrated Ruddlesden−Popper (RP) phase LaSr3Fe3O8(OH)2·xH2O has been investigated with focus on the orientation of the hydroxide groups and intercalated water. Combined powder synchrotron X-ray and neutron diffraction techniques were used. On the basis of Rietveld refinements and Fourier maps, intercalated water was found to form a network within the rock-salt-type layers of the RP phase with a likely dynamic interchange between different orientations. The water content was determined at different temperatures using thermogravimetric analysis, with findings showing that the water occupation follows a linear temperature dependence. The magnetic properties of LaSr3Fe3O8(OH)2·xH2O are significantly influenced by hydration, but no long-range order was observed. The relationship between the physical properties and crystal structure is discussed in detail.

■

upon hydration and those forming body-centered structures.11 The properties of the large A cation were proposed to be the main driving force, with the size and charge as key parameters. The amount of intercalated water differs considerably between materials exhibiting primitive or body-centered symmetry. The insulating material LaSr3Fe3O8(OH)2·xH2O has been reported to exhibit proton conductivity comparable to that of Nafion, the state-of-the-art membrane in proton-exchangemembrane fuel cells.14,15 Understanding the crystal structure of LaSr3Fe3O8(OH)2·xH2O is imperative in order to understand its potential functional properties. The crystal structure of NdSr3Fe3O7.5+δ(OH)2·xH2O has been described based on X-ray diffraction (XRD) data;10 however, any detailed description of the hydration layer is still lacking. In general, there is scarce information at hand on reports describing the orientation of water molecules within the rock-salt-type layer in hydrated RP phases. Liu et al. recently published a detailed description of K2.5Bi2.5Ti4O13·H2O including a tentative description of the intercalated water.16 This structure differs from LaSr3Fe3O8(OH)2· xH2O by exhibiting A-site ordering and, furthermore, has a primitive symmetry, which leads to a different local environment for the intercalated water. Here, we provide a detailed description of the crystal structure of LaSr3Fe3O8(OH)2·xH2O using a combination of powder synchrotron X-ray and neutron diffraction techniques. We show that both LaSr3Fe3O8(OH)2·xH2O and the starting material LaSr3Fe3O9 are isostructural to their neodymium analogues. The atomic positions and orientations of the hydroxide groups, as well as the intercalated water molecules, are described for the

INTRODUCTION Ruddlesden−Popper (RP)-type oxides, An+1BnO3n+1, possess a wide range of technologically important properties; e.g., thermoelectricity, colossal magnetoresistance, mixed conductivity, multiferroics, high-temperature superconductivity, and photocatalytic properties.1−6 Certain RP materials display intercalation reactions where small molecules or ions such as water, carbonate, hydroxide, or oxonium are incorporated into layers of the oxide through a topotactic reaction.7−11 Such intercalations strongly affect the functional properties of the host material. The intercalated species locate at the rock-salt-type layers of the RP phase; however, the process and accompanying structural and physical changes are poorly understood. Nishi et al.12 found that the RP3 material (n = 3) LaSr3Fe3O10, after reduction to LaSr3Fe3O9, becomes prone to intercalation in the presence of water vapor. The related material NdSr3Fe3O9 is described as an interlayered structure of an RP1 (A2BO4) and a brownmillerite type (A2B2O5) structure, with Fe3+ in tetrahedral and octahedral coordination.13 Hydration of NdSr3Fe3O8.5+δ results in two different materials depending on the temperature; for lower temperatures, NdSr3Fe3O8(OH)2·xH2O is formed, while hydration at higher temperatures gives NdSr3Fe3O8(OH)2.10 Interestingly, the formation of NdSr3Fe3O8(OH)2 involves a transformation from body-centered to primitive symmetry, whereas the fully hydrated NdSr3Fe3O8(OH)2·xH2O keeps body-centered symmetry. In the primitive structure, octahedra of the perovskite blocks are eclipsed along the stacking direction, as opposed to the staggered configuration in the body-centered structure, where octahedra are shifted half a unit cell along [110] on either side of the interlayer. Lehtimäki et al. have categorized hydrating RP materials into those forming primitive structures © XXXX American Chemical Society

Received: May 3, 2016

A

DOI: 10.1021/acs.inorgchem.6b01085 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

were collected in the 2θ range of 4−130°. The deuterated LaSr3Fe3O8(OD)2·xD2O sample was packed and sealed in a 5 mm cylindrical vanadium sample holder in a helium atmosphere. Data were collected at room temperature and 8 K. Refinements according to the Rietveld method were carried out using the Topas v5.0 software.19 The powder X-ray and neutron data sets were simultaneously fitted to the same structural model. NdSr3Fe3O7.5+δ(OH)2·xH2O was used as the starting model.10 Thompson−Cox−Hastings pseudo-Voigt profile functions were used in combination with Chebychev background polynomials for both data sets. For the SR-XRD data set, a zero point and an absorption correction were employed. For the neutron data set, the correction parameters included a zero-point factor and a simple axial model factor with a known instrument value in order to correct for asymmetric peak shapes. Both data sets included a preferred orientation (PO) spherical harmonics correction with six coefficients to account for a systematic 2θ-dependent peak-shape broadening. The structure model included an occupational constraint to the water sites (O5, H2/D2, and H3/D3) to ensure that mass balance was upheld. The isotropic displacement parameters of O1a, O1b, O2, O3a, and O3b were constrained to be equal. Constraints were also included for the isotropic displacement parameters of La/Sr1, La/Sr2, Fe1, and Fe2. Fourier maps were calculated using Topas and plotted in VESTA.20 The magnetic measurements were carried out from 4 to 350 K using a Quantum Design Physical Property Measuring System with an ACMS sample insert. An applied field of 1−10 kG was used to measure the field-cooled (FC) and zero-field-cooled (ZFC) magnetic responses during heating. The magnetic hysteresis loop was measured at 2 K in a field of ±90 kG.

first time for a hydrated, body-centered RP structure. The amount of intercalated water has been determined by thermogravimetric analysis (TGA) as a function of the temperature, revealing how temperature correlates to the intercalated water content. The magnetic properties of this iron(III) compound are, furthermore, described and discussed in relation to the crystal structure and, in particular, to the orientation of water and hydroxide groups in the hydrated interlayers.

■

EXPERIMENTAL METHODS

LaSr3Fe3O8(OH)2·xH2O was synthesized stepwise using LaSr3Fe3O10 as the starting material. First, LaSr3Fe3O10 was synthesized using a wet-chemical approach. La(NO3)3·6H2O (Sigma-Aldrich, >99.9%), Sr(NO3)2 (Sigma-Aldrich, >99.9%), and Fe(NO3)·9H2O (Sigma-Aldrich, >99.9%) were dissolved in distilled water in 0.1 M concentrations. The accurate molarities were determined using a thermogravimetric standardization method in which known quantities of the solutions were heated in air to form well-defined oxides. Stoichiometric amounts of the solutions were added to a citric acid melt and evaporated until a viscous gel formed. The gel was dried in a heating cupboard at 180 °C overnight and then in a muffle furnace at 450 °C in air for 12 h. The resulting powder was ground in a mortar and further heat-treated at 900 °C in air for 12 h and finally at 1250 °C in air for 6 h with subsequent slow cooling at 2.5 °C/min to ensure a close-to-stoichiometric oxygen content. Cerimetric titration was used to determine the exact oxygen content of LaSr3Fe3O10.17 LaSr3Fe3O10 was reduced to LaSr3Fe3O9 using NbO (Sigma-Aldrich, >99.99%) as a reducing agent according to the following reduction reaction: 2 1 LaSr3Fe3O10 + NbO → LaSr3Fe3O9 + Nb2O5 3 3

■

RESULTS AND DISCUSSION LaSr3Fe3O10 and LaSr3Fe3O9 were synthesized as single-phase materials with close-to-stoichiometric oxygen contents. The oxygen content of the synthesized LaSr3Fe3O10 was 9.971, while 9.013 was found for LaSr3Fe3O9, according to cerimetric titration. The precisions between the titration parallels for these samples are ±0.002 and ±0.003, respectively. When likely systematic errors were included, an absolute standard deviation of ±0.02 was estimated for both samples. Hydration of LaSr3Fe3O9 resulted in two different materials depending on the temperature:

(1)

This allowed for highly accurate reductions. Two crucibles containing stoichiometric amounts of LaSr3Fe3O10 and NbO were sealed in a quartz ampule under vacuum and heated to 900 °C. The ampule was thereafter opened in a glovebox to avoid reoxidation or uncontrolled intercalation reactions. The oxygen content of the LaSr3Fe3O9 product was determined by cerimetric titration. In order to rule out the possibility of any unwanted reaction with CO2 during the hydration reaction toward LaSr3Fe3O8(OH)2·xH2O, LaSr3Fe3O9 was loaded in an autoclave with degassed water in a separate compartment. The autoclave was sealed and heated to 50 °C and left overnight to form the final product of LaSr3Fe3O8(OH)2·xH2O. Because of the high incoherent scattering factor of hydrogen, a deuterated sample was prepared for collection of the neutron data. A route similar to that described above was followed using D2O instead of H2O, forming LaSr3Fe3O8(OD)2·xD2O. The hydration reaction of LaSr3Fe3O9 was studied as a function of the temperature using a Netzsch Jupiter STA 449 F1 thermogravimetric (TG/DSC) analyzer equipped with a steam generator. The hydration reactions were carried out at isothermal conditions by switching from dry N2 to wet N2 and then back to dry N2 while logging the change in mass. The selected time scale of the hydration ensured that the oxide was fully hydrated before the atmosphere was changed back to dry N2. A background correction was carried out at each temperature using the same hydration program but with an empty crucible. The partial pressure of water in wet N2 was set to 80% relative humidity at 45 °C, with a flow rate over the sample of 100 mL/min. Powder synchrotron X-ray diffraction (SR-XRD) data were collected at the BM01A beamline (Swiss−Norwegian Beamlines, SNBL) at the European Synchrotron Radiation Facility (ESRF) in Grenoble, France, using a 2D Pilatus detector and a wavelength of 0.69687 Å. The data were collected in the 2θ range of 1.8−48°. The 2D diffraction data were rebinned to 1D data sets by means of an in-house script using Fit2D as the converting software.18 Powder neutron diffraction (PND) data were collected with the PUS two-axis, high-resolution powder diffractometer at the JEEP II reactor at Kjeller, Norway, using a wavelength of 1.5538 Å. The data

LaSr3Fe3O9 + (x + 1)H 2O(g) T < ∼ 100 ° C

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯→ LaSr3Fe3O8(OH)2 ·x H 2O

(2)

T > ∼ 100 ° C

LaSr3Fe3O9 + H 2O(g) ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯→ LaSr3Fe3O8(OH)2

(3)

Figure 1 gives a visual representation and correlation between the precursor oxides LaSr3Fe3O10 and LaSr3Fe3O9 and the resulting hydrated structures of LaSr3Fe3O8(OH)2·xH2O and LaSr3Fe3O8(OH)2. The progress of hydration was monitored in situ by TGA (see Figure 2), and the water uptake was quantified. At 50 °C, there was approximately 2.7 mol of H2O per 1 mol of starting oxide, giving a stoichiometry of LaSr3Fe3O8(OH)2·1.7H2O according to eq 2. Hydrations carried out at different temperatures revealed that the amount of intercalated water decreased linearly with increasing temperature (see the inset in Figure 2). Extrapolating the linear trend suggests that the structure may achieve full occupancy of water (2.0 mol of H2O) at ∼33 °C. Because of slow kinetics, no direct proof is provided. At 100 °C, which is just below the dehydration temperature of LaSr3Fe3O8(OH)2·xH2O → LaSr3Fe3O8(OH)2, the water content was close to 1, equaling that of LaSr3Fe3O8(OH)2·H2O. Hydrations carried out at temperatures higher than 100 °C gave the oxide B

DOI: 10.1021/acs.inorgchem.6b01085 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. Schematic visualization of LaSr3Fe3O10, LaSr3Fe3O9, LaSr3Fe3O8(OH)2·xH2O, and LaSr3Fe3O8(OH)2. LaSr3Fe3O9 is shown in a pseudotetragonal setting for comparison. Note that the Fe−O apical bonds toward the rock-salt layers are strongly stretched for all structures except LaSr3Fe3O10.

Figure 2. Mass change measured with TGA for the hydration reaction of LaSr3Fe3O9 at 50 °C. The inset shows the amount of intercalated water in LaSr3Fe3O8(OH)2·xH2O as a function of the temperature.

Figure 3. Sections of the powder SR-XRD patterns of LaSr3Fe3O10, LaSr3Fe3O9, and LaSr3Fe3O8(OH)2·1.7H2O at 298 K. λ = 0.69687 Å. Very weak additional low-angle peaks are due to λ/3 contributions from the monochromator.

hydroxide LaSr3Fe3O8(OH)2 with no intercalated water and primitive crystal symmetry. Titrations were also carried out for LaSr3Fe3O8(OH)2·1.7H2O, which confirmed that no oxidation or reduction of Fe3+ occurred during hydration, meaning that the intercalated water is neutral H2O rather than H3O+ or OH−, according to eq 2. Synchrotron X-ray diffractograms of as-synthesized LaSr3Fe3O10, reduced LaSr3Fe3O9, and hydrated LaSr3Fe3O8(OH)2· 1.7H2O are presented in Figure 3. The crystal structures of LaSr3Fe3O9 and LaSr3Fe3O8(OH)2·1.7H2O appear isostructural to the neodymium analogues previously reported. The powder SR-XRD data for LaSr3Fe3O10 and LaSr3Fe3O9 were indexed in the tetragonal space group I4/mmm (a = 3.8659 Å and c = 28.0421 Å) and orthorhombic space group Bbmm (a = 5.5479 Å, b = 5.4846 Å, and c = 28.8846 Å), respectively.13,21 The nonstandard Bbmm setting of Cmcm was chosen to ease comparison with literature data for the neodymium analogue. For the crystal structure of LaSr3Fe3O8(OH)2·1.7H2O, supplementary PND data were collected for the deuterium

analogue LaSr3Fe3O8(OD)2·1.7D2O to accurately determine the atomic positions of the intercalated water and hydroxide groups. The PND and SR-XRD data were fitted simultaneously to the same structural model during Rietveld refinements, using space group I2/m as reported by Pelloquin et al.10 Initial refinements were conducted with a uniform isotropic displacement parameter, Beq, for all atomic sites. However, the refinement was significantly improved when Beq was only constrained to be equal for La/Sr1, La/Sr2, Fe1, and Fe2, the common O sites (O1a, O1b, O2, O3a, and O3b), the hydroxide group (O4 and H1/D1), and the water sites (O5, H2/D2, and H3/D3). Refinements with free Beq for all sites were also attempted, but this did not lead to any significant improvement of the refinement. The refinement converged to give a global Rwp = 3.39, Rp = 2.36, and GOF = 1.70. A significant improvement of the refinement was obtained when a PO spherical harmonics correction with six coefficients was included in the refinement in order to C

DOI: 10.1021/acs.inorgchem.6b01085 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

The obtained fits from the combined Rietveld refinements of the powder SR-XRD and PND data for LaSr3Fe3O8(OH)2· 1.7H2O and LaSr3Fe3O8(OD)2·1.7D2O are presented in Figures 4 and 5, respectively. The weak additional peaks at low angles, shown by arrows in Figure 4, are most likely not of magnetic origin because there is no sign for any ordering in the magnetization data (see below). The unit cell parameters and atomic positions are summarized in Table 1, and the crystal structure is visualized in Figure 6. While the interlayer distance (see Figure 1) increased marginally from 2.71 to 2.78 Å upon reduction of LaSr3Fe3O10 to LaSr3Fe3O9, the hydration of LaSr3Fe3O9 to LaSr3Fe3O8(OH)2·1.7H2O caused a significant expansion of the unit cell along the stacking direction [001], thereby increasing the interlayer distance from 2.78 to 6.26 Å. There were no major changes in the a and b dimensions. There were very minor differences in the c axes (35.6541 vs 35.6327) and β angles (88.788 vs 88.001) for the hydrated and deuterated samples synthesized for the SR-XRD and PND data, respectively. The a and b axes were the same. This difference may result from slightly different levels of intercalated water in the samples. For this reason, the unit cell parameters were not constrained for the two data sets. The crystal structure of LaSr3Fe3O8(OH)2·1.7H2O contains two Fe sites within the perovskite blocks (see Figure 6). Fe1 is located in the central layer of the perovskite block within a deformed FeO6 octahedron with four shorter bonds in the equatorial plane: Fe1−O1a (1.941 Å), Fe1−O1b (1.942 Å), and two longer Fe1−O2 bonds (2.034 Å) in the stacking direction [001]. The Fe2 site bonds very weakly to the hydroxyl group, which is reflected in the elongated bond distance (Fe2−O4 = 2.417 Å). The bond lengths in the equatorial plane are Fe2−O3a:0 = 1.920 Å, Fe2−O3a:1 = 1.973 Å, and 2× Fe2− O3b = 1.978 Å, while a shorter bond was found in the stacking direction with Fe2−O2 = 1.834 Å. The valence-bond method was used to evaluate the bond valence of the cationic sites.22 The results are given in Table 2. The calculation gave a slight overestimation for the average charge of the Fe sites (3.21+ vs expected value 3+) and a small underestimation of the La/Sr sites (2.15+ vs expected value 2.25+). The overall average charge of 2.55+ was in good agreement with the expected value of 2.57+.

correct for 2θ-dependent peak broadening in the SR-XRD and PND data sets. The cause of the broadening is discussed below.

Figure 4. Observed, calculated, and difference intensity profiles based on the Rietveld refinement of PND data for LaSr3Fe3O8(OD)2· 1.7D2O measured at 298 K. The inset shows the 2θ range of 13−21°. λ = 1.5538 Å. The arrows point to additional Bragg reflections, arguably not caused by magnetic order (see the text).

Figure 5. Observed, calculated, and difference intensity profiles based on the Rietveld refinement of SR-XRD data for LaSr3Fe3O8(OH)2· 1.7H2O measured at 298 K. λ = 0.69687 Å. The inset shows the 2θ range of 24−48° in more detail.

Table 1. Crystallographic Data for LaSr3Fe3O8(OH)2·1.7H2O and LaSr3Fe3O8(OD)2·1.7D2O from Combined Rietveld Refinements of Powder SR-XRD and PND Dataa atom

Wyckoff

x

La/Sr1 La/Sr2 Fe1 Fe2 O1a O1b O2 O3a O3b O4 H1/D1 O5 H2/D2 H3/D3

4i 4i 2a 4i 2c 2b 4i 4i 4i 4i 4i 4i 4e 4e

0.489(1) 0.473(1) 0 0.984(2) 1 /2 0 0.020(5) 0.495(6) 0.975(5) 0.982(4) 0.977(4) 0.002(7) 1 /4 3 /4

y

z

occupancy

Beq

/2 1 /2 0 0 0 1 /2 0 0 1 /2 0 0 0 1 /4 1 /4

0.0576(1) 0.1615(1) 0 0.1085(2) 0 0 0.0547(4) 0.1119(5) 0.1174(5) 0.1749(5) 0.2044(4) 0.2677(7) 1 /4 1 /4

0.25/0.75 0.25/0.75 1 1 1 1 1 1 1 1 1 1 0.85(3) 0.85(3)

0.82(6) 0.82(6) 0.46(8) 0.46(8) 2.9(1) 2.9(1) 2.9(1) 2.9(1) 2.9(1) 2.1(3) 2.1(3) 5.8(4) 5.8(4) 5.8(4)

1

a

Calculated standard deviations are in parentheses. Space group: I2/m (No. 12). Unit cell dimensions: a = 3.8819(2) Å, b = 3.8843(3) Å, c = 35.655(2) Å, and β = 88.740(8)°. D

DOI: 10.1021/acs.inorgchem.6b01085 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

suggests dynamic movements of the water molecules, which in terms of the D atoms could be rationalized in an average site between two O5 sites. The dynamic movement is reflected in large isotropic displacement parameters for O5, H2/D2, and H3/D3 (see Table 1). For this reason, two 4e H/D sites at (1/4, 1/4, 1/4) and (3/4, 1/4, 1/4) were included in the combined SR-XRD and PND structural model. The occupancy of the intercalated water found from the Rietveld refinement converged to a value of 0.85(3), corresponding to the composition LaSr3Fe3O8(OD)2·1.7D2O, in excellent agreement with the TGA data (see Figure 2). Figure 8a shows the orientation of the hydroxide groups and the crystal water for the average structure of the rock-salt layer in an ac projection, while Figure 8b shows their orientations in the ab plane. The intercalated water is located with the O atoms in a jagged configuration in the ab plane, with the opening of the D−O−D angle facing away from the La/Sr2 site. The hydroxide groups are pointing toward the O5 sites with a O5−D1 separation distance of 2.26(3) Å, suggesting the presence of hydrogen bonds. The average positions of the D2 and D3 sites allow for several different orientations of the water molecules. Parts c−j of Figure 8 illustrate the four possible configurations based on the average D sites in Figure 8b. The average electron density given in the Fourier difference map (Figure 7) suggests that the water molecules do not favor one single configuration but are instead constantly switching between different configurations. The neutron data collected for LaSr3Fe3O8(OD)2·1.7D2O at 8 K were refined to evaluate whether any of the configurations in Figure 8 were preferred. However, no conclusions could be drawn. Hence, there is an apparent lack of long-range ordering in the hydration layer, which indicates dynamical changes between the different configurations. A broadening of the diffraction peaks was observed for both the powder SR-XRD and PND data, with increased broadening at higher 2θ angles. Previous studies of similar hydrated materials suggest the broadening to be due to lowered crystallinity.10 However, the clear 2θ dependence might suggest that strain due to variable water content is the cause of the observed peak broadening. This highly anisotropic expansion during hydration might indicate another cause for the 2θ-dependent broadening. The intercalation of one monolayer of water into the rock-salt layer implies a maximum of two water molecules due to space restrictions. TGA shows that the water content (corresponding to the site occupancy for water molecules in the structure model) depends on the conditions, with time and temperature for hydration the most important parameters. An occupancy of less than 1 could imply that certain rock-salt interlayers are not hydrated at all and that potentially staging may also occur. This will add to peak broadening and diffuse scattering. Moreover, the water molecules could be found in several possible configurations, as shown in Figure 8. The atomic positions of the sites surrounding the water layer will be influenced by the orientation of the water molecules. For each of the possible configurations, a slight variation of the atomic coordinates of the O4, D1, and

Figure 6. Visual representation of the LaSr3Fe3O8(OD)2·1.7D2O crystal structure. Green spheres are La/Sr, red O, and white D, and brown polyhedra represent coordination at Fe sites. The labeling of atoms is in accordance with Table 1.

During the hydration reaction, the O4 position, originally shared between the rock-salt interlayer and perovskite octahedra surrounding the Fe2 site in the LaSr3Fe3O9 structure, moves toward the hydration layer to form hydroxide groups. The O−H bond distance in the hydroxide group was found to be 1.05(2) Å. The O5 position, located between the O4 and La/Sr2 sites in the stacking direction, represents the O atom of the intercalated water molecule. From the PND data for LaSr3Fe3O8(OD)2·1.7D2O, the electron density connected with the deuterium positions for the intercalated water appeared to be distributed over more sites. Fourier difference maps revealed the electron density between the O5 sites (see Figure 7). This Table 2. Calculated Bond Valencesa

site charge a b

La/Sr1

La/Sr2

1.96

2.15

Fe1

Fe2

3.38

3.12

∑La/Sr,avg

∑Fe,avg

∑tot,avg

∑exp,avgb

2.05

3.21

2.55

2.57

The bond valence parameters 1.759 for Fe −O and 2.129 for 0.25La /0.75Sr −O ∑exp,avg = (2La/Sr1 + 2La/Sr2 + Fe1 + 2Fe2)/∑(A and B sites). 3+

2−

3+

E

2+

2−

were collected from ref 22 and used in the calculations. DOI: 10.1021/acs.inorgchem.6b01085 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 7. (a) Difference Fourier map for LaSr3Fe3O8(OD)2·1.7D2O based on the PND data. The yellow volumes represent the electron density associated with the D atoms. (b) 2D slice through the O5 sites of the hydration layer showing additional electron density.

La/Sr2 (see Figure 6) sites is to be expected. This would lead to slight variations in the interplanar spacings (d values), which could add to the observed 2θ-dependent peak broadening. The PND data show no evident peaks that could be ascribed to long-range magnetic ordering. The intensities of the weak peaks indicated in the inset in Figure 4 for the 2θ range of 17−22° did not change significantly between 8 and 295 K and are, hence, unlikely to stem from magnetic ordering. The magnetic susceptibilities of LaSr3Fe3O9 and LaSr3Fe3O8(OH)2·1.7H2O were measured as a function of the temperature (see Figure 9). Insignificant differences were observed between

Figure 8. Sketch of the hydration layer of LaSr3Fe3O8(OD)2·1.7D2O. Parts a and b show the projected average structure, while parts c−j show four different orientations of the water molecules. Dotted lines indicate possible hydrogen bonds.

Figure 9. Magnetic (left) and inverse magnetic (right) susceptibility for the ZFC sample of (a) LaSr3Fe3O9 between 4 and 350 K in a measuring field of 10 kG and (b) LaSr3Fe3O8(OH)2·1.7H2O between 4 and 300 K in a measuring field of 1 kG. F

DOI: 10.1021/acs.inorgchem.6b01085 Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

■

ACKNOWLEDGMENTS This work was supported by the Research Council of Norway [Grant 221905 (FRIPRO) and Project NOFCO]. The authors acknowledge the assistance of the research team at the SNBL, ESRF, and thank Dr. Susmit Kumar for his skilled assistance with the collection of the magnetic data.

the FC (not shown) and ZFC samples for both materials. The measured susceptibility of LaSr3Fe3O9 was low with an upturn from about 250 K, indicating antiferromagnetic (AFM) ordering with TN higher than the measured temperature interval. Related iron(III) compounds like LaFeO3 and Sr2Fe2O5 both have AFM ordering with high TN.23,24 The measured susceptibility of LaSr3Fe3O8(OH)2·1.7H2O differs significantly from that of LaSr3Fe3O9, indicating paramagnetic behavior; however, no temperature range is consistent with paramagnetic Curie−Weiss behavior. The inverse susceptibility exhibits a downturn over a fairly large temperature interval approaching 4 K. This behavior may suggest magnetic contributions from short-range ordering effects. The different magnetic behavior between LaSr3Fe3O9 and LaSr3Fe3O8(OH)2·1.7H2O demonstrates how anisotropic expansion along the c axis upon hydration appears to decouple any long-range magnetic ordering between the FeIII sites found in the perovskite blocks.

■

CONCLUSIONS LaSr3Fe3O8(OH)2·xH2O was prepared in a stepwise synthesis by hydration of LaSr3Fe3O9, which was obtained by the reduction of LaSr3Fe3O10. The intercalated water content varies linearly as a function of the temperature. At 50 °C, the water content was 1.7H2O, decreasing to 1.0H2O at 100 °C. Heating beyond this temperature led to phase transformation into LaSr3Fe3O8(OH)2. Extrapolation of the linear dependence indicates that the material would become fully hydrated (2.0 mol of H2O) at ∼33 °C; however, the kinetics is slow. The combined refinement of the SR-XRD and PND patterns by the Rietveld method provides a complete description of the crystal structure of LaSr3Fe3O8(OH)2·xH2O including the positions of the hydroxide groups and intercalated water. The water molecules form a network within the rock-salt layer. The O sites of the water molecules displayed long-range ordering, while the H sites were found to be elongated between the O sites. This suggested a dynamic movement and rotation of the water molecules, resulting in four possible orientations. Rietveld refinement of room temperature and 8 K PND data was attempted with the goal of determining whether any particular configuration was favored; however, no definite conclusions could be drawn. The magnetic susceptibility indicated significant decoupling due to separation of the perovskite layers by the hydrated rocksalt layers in the [001] stacking direction. Although the magnetic data appeared purely paramagnetic, no temperature range could be fitted to Curie−Weiss behavior. ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.6b01085. Neutron diffraction data collected at 8 K, selected bond lengths, and magnetic hysteresis curve (PDF) Structural data for LaSr3Fe3O8(OH)2·xH2O in CIF format (CIF)

■

REFERENCES

(1) Battle, P. D.; Green, M. A.; Laskey, N. S.; Millburn, J. E.; Radaelli, P. G.; Rosseinsky, M. J.; Sullivan, S. P.; Vente, J. F. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 15967−15977. (2) Doig, K. I.; Peters, J. J. P.; Nawaz, S.; Walker, D.; Walker, M.; Lees, M. R.; Beanland, R.; Sanchez, A. M.; McConville, C. F.; Palkar, V. R.; Lloyd-Hughes, J. Sci. Rep. 2015, 5. (3) Jorgensen, J. D.; Dabrowski, B.; Pei, S.; Richards, D. R.; Hinks, D. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1989, 40, 2187−2199. (4) Shimizu, K.; Itoh, S.; Hatamachi, T.; Kodama, T.; Sato, M.; Toda, K. Chem. Mater. 2005, 17, 5161−5166. (5) Moon, S. J.; Jin, H.; Kim, K. W.; Choi, W. S.; Lee, Y. S.; Yu, J.; Cao, G.; Sumi, A.; Funakubo, H.; Bernhard, C.; Noh, T. W. Phys. Rev. Lett. 2008, 101, 226402. (6) Lee, K. H.; Kim, S. W.; Ohta, H.; Koumoto, K. J. Appl. Phys. 2006, 100, 063717. (7) Grimaud, A.; Mauvy, F.; Bassat, J. M.; Fourcade, S.; Marrony, M.; Grenier, J. C. J. Mater. Chem. 2012, 22, 16017−16025. (8) Matvejeff, M.; Lehtimaki, M.; Hirasa, A.; Huang, Y. H.; Yamauchi, H.; Karppinen, M. Chem. Mater. 2005, 17, 2775−2779. (9) Pelloquin, D.; Barrier, N.; Flahaut, D.; Caignaert, V.; Maignan, A. Chem. Mater. 2005, 17, 773−780. (10) Pelloquin, D.; Hadermann, J.; Giot, M.; Caignaert, V.; Michel, C.; Hervieu, M.; Raveau, B. Chem. Mater. 2004, 16, 1715−1724. (11) Lehtimaki, M.; Yamauchi, H.; Karppinen, M. J. Solid State Chem. 2013, 204, 95−101. (12) Nishi, T.; Toda, K.; Kanamaru, F.; Sakai, T. Key Eng. Mater. 1999, 169−170, 235−238. (13) Barrier, N.; Pelloquin, D.; Nguyen, N.; Giot, M.; Bouree, F.; Raveau, B. Chem. Mater. 2005, 17, 6619−6623. (14) Takeguchi, T.; Yamanaka, T.; Takahashi, H.; Watanabe, H.; Kuroki, T.; Nakanishi, H.; Orikasa, Y.; Uchimoto, Y.; Takano, H.; Ohguri, N.; Matsuda, M.; Murota, T.; Uosaki, K.; Ueda, W. J. Am. Chem. Soc. 2013, 135, 11125−11130. (15) Zawodzinski, T. A.; Derouin, C.; Radzinski, S.; Sherman, R. J.; Smith, V. T.; Springer, T. E.; Gottesfeld, S. J. Electrochem. Soc. 1993, 140, 1041−1047. (16) Liu, S.; Avdeev, M.; Liu, Y.; Johnson, M. R.; Ling, C. D. Inorg. Chem. 2016, 55, 1403−1411. (17) Jantsky, L.; Okamoto, H.; Demont, A.; Fjellvag, H. Inorg. Chem. 2012, 51, 9181−9191. (18) Hammersley, A. P.; Svensson, S. O.; Hanfland, M.; Fitch, A. N.; Hausermann, D. High Pressure Res. 1996, 14, 235−248. (19) Coelho, A. A. Topas V5.0; Bruker AXS: Karlsruhe, Germany, 2014. (20) Momma, K.; Izumi, F. J. Appl. Crystallogr. 2011, 44, 1272−1276. (21) Lee, J. Y.; Swinnea, J. S.; Steinfink, H.; Reiff, W. M.; Pei, S.; Jorgensen, J. D. J. Solid State Chem. 1993, 103, 1−15. (22) Brese, N. E.; O'Keeffe, M. Acta Crystallogr., Sect. B: Struct. Sci. 1991, 47, 192−197. (23) Peterlinneumaier, T.; Steichele, E. J. Magn. Magn. Mater. 1986, 59, 351−356. (24) Schmidt, M.; Campbell, S. J. J. Solid State Chem. 2001, 156, 292−304.

■

■

Article

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest. G

DOI: 10.1021/acs.inorgchem.6b01085 Inorg. Chem. XXXX, XXX, XXX−XXX