Suppression of magnetic order in Sr3Fe2O 4Cl2 by Fe-site substitution by cobalt

Article abstract of DOI:10.1021/ic200832n

The low-temperature topotactic reduction of Sr₃Fe _2-xCo_xO₅Cl₂ oxychloride phases with LiH allows the preparation of phases of composition Sr₃Fe _2-xCo_xO₄Cl₂ (0 ≥ x ≥ 1). The reduced phases adopt body-centered tetragonal structures which are isostructural with Sr₃Fe₂O₄Cl₂ and contain square-planar (Fe/Co)O₄ centers connected into apex-linked sheets, analogous to the CuO₂ sheets present in superconducting cuprate phases. As the cobalt concentration in Sr₃Fe_2-xCo _xO₄Cl₂ is increased the antiferromagnetic order of the Sr₃Fe₂O₄Cl₂ host phase is suppressed, ultimately leading to spin-glass behavior, at low temperature, in Sr₃Fe_2-xCo_xO₄Cl₂ phases with x ≥ 0.8. The limited influence of cobalt substitution on the reactions which form the Sr₃Fe_2-xCo_xO₄Cl ₂ phases is discussed and contrasted to that of the related SrFeO_3-δ-SrFeO₂ system.

Full text of DOI:10.1021/ic200832n

ARTICLE
pubs.acs.org/IC
Suppression of Magnetic Order in Sr Fe O Cl by Fe-Site Substitution
3
2 4 2
by Cobalt
Edward Dixon and Michael A. Hayward*
Inorganic Chemistry Laboratory, Department of Chemistry, University of Oxford, South Parks Road, Oxford OX1 3QR,
United Kingdom
S Supporting Information
b
ABSTRACT: The low-temperature topotactic reduction of
Sr Fe Co O Cl oxychloride phases with LiH allows the
3
2ꢀx
x
5
2
preparation of phases of composition Sr Fe Co O Cl
0 e x e 1). The reduced phases adopt body-centered
tetragonal structures which are isostructural with Sr Fe O Cl
2
3
2ꢀx
x
4
2
(
3
2
4
and contain square-planar (Fe/Co)O centers connected into
4
apex-linked sheets, analogous to the CuO sheets present in
2
superconducting cuprate phases. As the cobalt concentration in
Sr Fe Co O Cl is increased the antiferromagnetic order of
3
2ꢀx
x
4
2
the Sr Fe O Cl host phase is suppressed, ultimately leading to
3
2
4
2
spin-glass behavior, at low temperature, in Sr Fe Co O Cl
phases with x g 0.8. The limited influence of cobalt substitution on the reactions which form the Sr Fe Co O Cl phases is
3
2ꢀx
x
4
2
3
2ꢀx
x
4
2
discussed and contrasted to that of the related SrFeO3 ꢀSrFeO system.
ꢀδ
2
’
INTRODUCTION
Since the discovery of high-temperature superconductivity in
coordination sites, is attributable to the favorable crystal-field
stabilization energy associated with the d electron count of these
9
ions in such a coordination environment. It is however possible
to prepare phases with extended sheets of corner-linked MO4
centers in the absence of such favorable interactions, as demon-
strated by the reduction of SrFeO3 and Sr Fe O to the
complex copper oxides, there has been great interest in preparing
materials which contain square-planar MO transition-metal
centers, particularly when these centers are arranged into cor-
ner-linked sheets, analogous to the CuO sheets present in the
superconducting cuprate phases. However despite extensive
investigations, very few thermodynamically stable phases have
been discovered which exhibit this structural motif in the absence
of copper.
Given the apparent low thermodynamic stability of extended
oxide phases containing square-planar transition metal centers, a
logical synthetic approach is to exploit kinetically controlled
chemical processes to prepare suitable metastable phases. To-
potactic reduction reactions, which exploit the large difference
between cation and anion mobilities in complex oxides to
deintercalate oxide ions from extended solids, facilitate the
preparation of kinetically stabilized phases which contain ex-
tended arrays of transition-metal centers in unusual coordination
environments and oxidation states. This includes phases which
contain apex-linked sheets of planar MO centers. Thus, for
example, the reduction of the Ni(III) oxides LaNiO and
LaSrNiO at low-temperature yields the Ni(I) phases LaNiO
and LaSrNiO , respectively, which contain extended arrays of
4
3
2 6
2
5,6
respective Fe(II) phases SrFeO and Sr Fe O , both of which
2
3
2 5
II
contain arrays of apex-linked Fe O planar centers as shown in
4
Figure 1.
In this instance, it is not clear what directs SrFeO and
2
Sr Fe O to adopt anion deficient structures which contain
3
2 5
Fe(II) centers in these highly unusual square-planar coordination
6
sites. The high-spin d electronic configuration adopted by the
Fe(II) centers in these phases is offered no obvious stabilization
by square-planar coordination geometries. Likewise the large
scale rearrangement of the anion lattice during the course of the
reduction reactions that lead to the formation of both SrFeO2
and Sr Fe O , make it hard to rationalize the structures adopted
3
2 5
by these phases on the basis of the coordination preferences of
2+
the Sr A-cations, in contrast to a number of related anion
7,8
deintercalation reactions.
4
To shed some light on the structural selectivity exhibited
3
during the formation of SrFeO , a recent study by Kageyama
2
4
2
1,2
et al. investigated the effect of replacing some of the iron centers
3
9
I
with cobalt or manganese. They observed that at low levels of
corner-linked Ni O square-planes. Specifically LaNiO forms
the so-called infinite-layer structure and is isostructural with
Ca0.84Sr0.16CuO2 while LaSrNiO adopts a one-dimensional
structure isostructural with Sr CuO3 as shown in Figure 1.
4
2
substitution the reduction of SrFe1 (Mn/Co) O substrate
ꢀx
x
3ꢀδ
3
phases led as expected to the formation of the corresponding
3
4
2
The formation of anion deficient phases such as LaNiO2
Received: April 21, 2011
Published: June 29, 2011
and LaSrNiO , which contain Ni(I) centers in square-planar
3
r 2011 American Chemical Society
7250
dx.doi.org/10.1021/ic200832n
|Inorg. Chem. 2011, 50, 7250–7256

Inorganic Chemistry
ARTICLE
(
0 e x e 1) to complement and contrast those of the analogous
SrFe1 Co O (0 e x e 0.3) phases.
ꢀx
x 2
’
EXPERIMENTAL SECTION
Preparation of Sr Fe Co O Cl Phases. Samples of Sr Fe2ꢀx-
3
3
2ꢀx
x
5
2
Co
x 5 2
O Cl (x = 0.2, 0.6, 0.8, 1) were prepared by the direct-combination
route previously described by Weller et al. for the synthesis of Sr Fe-
3
1
1
CoO
SrCO
Fe O (99.99%), and Sr Co O (prepared using the method described by
5
Cl
2
.
Suitable quantities of SrO (prepared by the decomposition of
3
at 1100 °C under vacuum), SrCl
2
(dried at 180 °C under vacuum),
2
3
2
2 5
12
Grenier et al. ) were thoroughly mixed using an agate pestle and mortar in
an argon filled glovebox (O and H O levels <1 ppm). The resulting
mixtures were then sealed under vacuum in silica ampules and heated at
50 °C for 2 periods of 24 h with one intermediate grinding. X-ray powder
diffraction data sets collected from the Sr Fe2ꢀxCo Cl samples were
consistent with single phase materials with lattice parameters which span the
range between the values reported for Sr Fe Cl and Sr FeCoO Cl as
2
2
8
3
x
O
5
2
3
2
O
5
2
3
5
2
11
detailed in the Supporting Information.
Reduction of Sr Fe Co O Cl Phases. The reduction of the
Sr Fe Co O Cl (x = 0.2, 0.6, 0.8, 1) phases was performed using LiH
3
2ꢀx
x
5
2
3
2ꢀx
x
5
2
1
3
as a solid-state reducing agent in a similar manner to that reported
1
0
for the topotactic reduction reaction of Sr
3
Fe
2
O
5
Cl
2
.
Small samples
(
∼ 300 mg) of the required Sr Fe Co O Cl phase were ground
3
2ꢀx
x
5
2
together in a 1:4 molar ratio with LiH in an argon filled glovebox. These
mixtures were then sealed under vacuum in Pyrex ampules and heated at
temperatures between 210 and 400 °C to monitor the temperature
dependence of the reduction reactions.
Because of the hazards associated with the production of hydrogen
1
3
gas when using LiH as a reducing agent, large-scale samples suitable for
characterization by neutron powder diffraction were synthesized in a
sealed Parr Series 4740 pressure vessel which is capable of withstanding
the hydrogen gas pressures generated by these reactions. Approximately
4
g of the required Sr
3
x 5 2
Fe2ꢀxCo O Cl phase was mixed, in an argon
filled glovebox, with 4 mol equivalents of LiH and loaded into an alumina
finger which was then sealed inside the pressure vessel. The pressure
vessel was then evacuated, sealed, and heated so that the sample was held
at a temperature of 300 °C for 4 periods of 24 h with intermediate
grinding. After reaction, samples were washed with 4 ꢁ 100 mL aliquots
of methanol, under nitrogen, to remove any lithium containing phases
2
Figure 1. (a) Infinite layer structure adopted by LaNiO , Ca0.84Sr0.16-
CuO , and SrFeO . (b) The structure of LaSrNiO . (c) The structure of
2
2
3
Sr
3
Fe
2
O
5
. (d) The structural transformation which occurs during
Fe Cl to Sr Fe Cl
topotactic reduction of Sr
3
2
O
5
2
3
2
O
4
2
.
(
LiH and Li
2
O) before being dried under vacuum.
infinite layer SrFe1 (Mn/Co) O phases. However, at substitu-
tion levels of greater than 30%, reduction reactions led to
alternative products which are yet to be identified. The sensitivity
ꢀx
x 2
Characterization. X-ray powder diffraction data were collected
from samples contained within homemade gastight sample holders using
a PANalytical X’Pert diffractometer incorporating an X’celerator posi-
tion sensitive detector (monochromatic Cu KR1 radiation). Neutron
powder diffraction data were collected from samples contained within
vanadium cans which had been sealed under argon with indium washers.
Table 1 details the instruments, temperature ranges, and incident
wavelengths at which data were collected. Rietveld profile refinement
of the reduction reactions of SrFe1 (Mn/Co) O phases to
ꢀx
x
3ꢀδ
6
cation substitution does suggest that the d electronic configura-
tion of the Fe(II) centers plays some role in directing the
II
formation of anion deficient structures with square-planar Fe O4
coordination sites. However, a more immediate consequence of
the limited range of substituted SrFe1 (Mn/Co) O phases
14
ꢀx
x 2
against these data was performed using the GSAS suite of programs.
which can be prepared is the frustration of the second goal of the
Kageyama investigation: to probe the electronic behavior of
Magnetization measurements were collected using a Quantum Design
MPMS SQUID magnetometer. Magnetization data indicated that all the
reduced samples contained small quantities of ferromagnetic elemental
iron and cobalt, which masked the magnetic behavior of the bulk phases.
The paramagnetic susceptibility of samples was therefore measured
substituted SrFeO phases as a function of a wide range of
2
9
d-electron count.
Recently we reported the preparation of the Fe(II) phase
15
Sr Fe O Cl via the topotactic reduction of Sr Fe O Cl with
using a “ferromagnetic subtraction” technique described previously
and described fully in the Supporting Information.
3
2
4
2
3
2
5
2
1
0
LiH. In this instance the formation of the anion deficient
structure adopted by Sr Fe O Cl , which also contains sheets of
3
2
4
2
II
apex-linked planar Fe O centers (Figure 1), can be seen to be
directed by the coordination preferences of the Sr A-cations. As
a result the reactions which form Sr Fe O Cl are much more
4
’
RESULTS
2+
Reactivity of Sr Fe O Cl . X-ray powder diffraction data
collected from the products of reactions between Sr Fe Cox-
3 2ꢀx
O Cl (x = 0.2, 0.6, 0.8, 1) phases and LiH reveal that at tem-
5 2
3
2
4
2
3
2
4
2
tolerant to cation substitution than those which form SrFeO2.
Here we report the effects of cobalt substitution on the crystal
structure and magnetic behavior of Sr Fe Co O Cl phases
peratures below 300 °C no reactions occur. Reactions performed
3
2ꢀx
x
4
2
7
251
dx.doi.org/10.1021/ic200832n |Inorg. Chem. 2011, 50, 7250–7256

Inorganic Chemistry
ARTICLE
Table 1. Details of the Neutron Powder Diffraction Experi-
Structural Characterization. Neutron powder diffraction
ments Performed on the Sr Fe Co O Cl Samples
data collected from the reduced Sr Fe Co O Cl (x = 0.2,
3
2ꢀx
x
4
2
3 2ꢀx x 4 2
0
.6, 0.8, 1) samples could be readily indexed on the basis of body-
instrument in temperature range
centered tetragonal cells, with lattice parameters similar to those
of Sr Fe O Cl . Structural models based on the refined structure
3
2
4
2
5
< T/K < 300
298 K
D1a, ILL
300 < T/K < 323
of Sr Fe O Cl were therefore constructed for the Sr Fe Cox-
3
2
4
2
3
2ꢀx
Sr
3
Sr
3
Sr
3
Sr
3
Fe1.8Co0.2
Fe1.4Co0.6
Fe1.2Co0.8
O
O
O
4
Cl
4
Cl
4
Cl
2
2
2
D1b, ILL
λ = 2.52 Å
D1b, ILL
λ = 2.52 Å
D2b, ILL
λ = 1.59 Å
D2b, ILL
O Cl phases and refined against these neutron powder diffrac-
4 2
λ = 1.91 Å
D1a, ILL
λ = 1.59 Å
tion data sets. As noted above the samples contained small
quantities of Sr Fe Co O Cl impurities; therefore, structural
3
2ꢀx
x
5
2
models based on Sr Fe Co O Cl phases were added to the
λ = 1.91 Å
D2b, ILL
λ = 1.59 Å
3
2ꢀx
x
5
2
structural refinements to account for these diffraction features.
Further additional diffraction features were observed in the
room-temperature data collected from the x = 0.2 sample, which
were consistent with the presence of long-range magnetic order,
4
FeCoO Cl
2
POLARIS, ISIS POLARIS, ISIS
time of flight time of flight
10
as observed previously for Sr Fe O Cl . A magnetically or-
3
2
4
2
dered model was therefore added to the refinement performed
against this data set, as described more fully below.
The structural refinements against all 4 room-temperature
data sets converged readily to give good statistical fits to the data.
To confirm the Sr Fe Co O Cl composition of the bulk phases,
3
2ꢀx
x
4
2
the fractional occupancies of the oxygen and chlorine sites were
refined and were observed to remain at fully occupancy, within
1
1
error. In addition an extra oxide ion site was added at ( / , / , 0) to
2
2
model partial oxidation of the bulk phases. This site refined to zero
occupancy for all samples, demonstrating that the mild oxidation
observed during the methanol wash resulted in a mixture of
Sr Fe Co O Cl and Sr Fe Co O Cl not some partially oxi-
3
2ꢀx
x
4
2
3
2ꢀx
x
5
2
dized Sr Fe Co O Cl phase. Full details of the structures
refined for the Sr Fe Co O Cl phases are given in Table 2 with
selected bond lengths in Table 3. Observed, calculated and difference
3
2ꢀx
x
5ꢀy
2ꢀx
2
3
x
4
2
plots from the refinement of Sr FeCoO Cl against neutron powder
3
4
2
diffraction data are shown in Figure 3, with analogous plots from
the room-temperature structural refinements of Sr Fe Co O Cl
2
3
2ꢀx
x
4
(
x = 0.2, 0.6, 0.8) given in the Supporting Information.
Magnetic Characterization. Variable temperature neutron
powder diffraction data were collected from the Sr Fe Cox-
Figure 2. X-ray powder diffraction data collected from Sr
Cl (top), Sr Cl prior to methanol wash
middle), and Sr Fe1.8Co0.2O Cl after the methanol washing proce-
3
Fe1.8-
3
2ꢀx
Co0.2
(
O
5
2
3
Fe1.8Co0.2
O
4
2
O Cl (x = 0.2, 0.6, 0.8, 1) samples, as described in Table 1, to
4
2
3
4
2
characterize the long-range magnetic order in these phases. As
dure (bottom). The arrow marks additional scattering due to the
presence of Sr Cl
reported previously, Sr Fe O Cl adopts an antiferromagneti-
3
2
4
2
3
Fe1.8Co0.2
O
5
2
.
10
cally ordered state below TN ∼ 378 K. Examination of the
variable temperature neutron powder diffraction data collected
from Sr Fe Co O Cl and Sr Fe Co O Cl revealed dif-
at temperatures above 350 °C result in the decomposition of the
Sr Fe Co O Cl (0.2 e x e 1) substrate phases to form a
3
1.8
0.2
4
2
3
1.4
0.6
4
2
√
√
0
0
3
2ꢀx
x
5
2
fraction features consistent with the a = 2 ꢁ a, b = 2 ꢁ b,
0
complex mixture of binary metal oxides, chlorides, and elemental
iron and cobalt. In the temperature range 300 < T/°C < 350
reactions resulted in the formation of body-centered tetragonal
c = c magnetic cell refined for Sr Fe O Cl ; therefore, magnetic
3
2
4
2
models analogous to that refined for Sr Fe O Cl were refined
3
2
4
2
against these data sets. Complete descriptions of these refine-
ments are given in the Supporting Information.
10
phases with lattice parameters similar to that of Sr Fe O Cl ,
3
2
4
2
indicating that topotactic reduction reactions of the Sr Fe
Figure 4 shows a plot of the ordered magnetic moment refined
3
2ꢀx-
Co O Cl (0.2 e x e 1) substrate phases had occurred.
for the Sr Fe Co O Cl (x = 0, 0.2, and 0.6) phases as a
x
5
2
3 2ꢀx x 4 2
Examination of X-ray powder diffraction data collected from
samples before and after the methanol washing procedure
function of temperature. Fits of these data to an I = A(1 ꢀ
β
(T/T )) power law yielded values of β = 0.494, 0.495, and
N
(Figure 2) revealed a number of additional weak diffraction
0.394; T = 378 K, 340 K, and 300 K for the x = 0, 0.2, and 0.6
N
reflections in the latter data sets. These additional features are
consistent with the presence of small amounts of oxidized
samples, respectively. There was no indication of magnetic order
in either the x = 0.8 or x = 1 sample down to the lowest
temperature measured (5 K).
Sr Fe Co O Cl phases formed by mild oxidation during
3
2ꢀx
x
5
2
the washing procedure. The quantity of these oxidized phases
could be minimized by the use of dry solvents and rigorously
anaerobic conditions; however, with the exception of the x = 1
sample, their presence could not be completely eliminated.
Magnetic susceptibility data collected from Sr FeCoO Cl
3
4
2
using the “ferromagnetic subtraction” method described above,
are shown in Figure 5. These data can be fitted to the Cur-
ieꢀWeiss law (χ = C/(T ꢀ θ)) in the temperature range 35 e
3
ꢀ1
This high level of air sensitivity exhibited by Sr Fe Cox-
T/K e 300, to yield values of C = 0.778(4) cm K mol and
θ = ꢀ9.7(3) K. It should be noted that this value of the Curie
constant is significantly lower than would be expected for a simple
3
2ꢀx
O Cl phases is consistent with the pyrophoric behavior of
4
2
1
0
Sr Fe O Cl .
3
2
4
2
7
252
dx.doi.org/10.1021/ic200832n |Inorg. Chem. 2011, 50, 7250–7256

Inorganic Chemistry
ARTICLE
Table 2. Structural Parameters Refined for Sr Fe Co O Cl Phases against Room-Temperature Neutron Powder
3
2ꢀx
x
4
2
a
Diffraction Data
Sr
3
x 4 2
Fe2ꢀxCo O Cl
x = 0
x = 0.2
x = 0.6
x = 0.8
x = 1
a (Å)
c(Å)
4.0095(1)
22.6365(5)
0.0054(2)
0.1524(1)
0.0087(1)
0.4263(1)
0.0048(1)
0.4214(1)
0.0128(4)
0.2950(1)
0.0135(3)
0
4.0160(1)
22.6275(9)
0.0072(18)
0.1526(1)
0.0093(13)
0.4250(1)
0.0061(7)
0.4215(1)
0.0167(8)
0.2948(1)
0.0121(7)
15%
4.0178(1)
22.552(1)
0.0136(21)
0.1526(2)
0.0082(15)
0.4258(2)
0.0032(8)
0.4218(2)
0.0141(8)
0.2954(1)
0.0122(8)
17%
4.0162(1)
22.595(1)
0.0104(15)
0.1528(2)
0.0141(12)
0.4263(2)
0.0023(2)
0.4218(1)
0.0135(6)
0.2956(1)
0.0134(6)
13%
4.0145(1)
22.4795(7)
0.0038(4)
0.15309(7)
0.0066(3)
0.42676(9)
0.0021(2)
0.42152(6)
0.0110(2)
0.29492(6)
0.0129(2)
0
2
Sr(1) Uiso (Å )
Sr(2) z
2
Sr(2) Uiso (Å )
Fe/Co z
2
Fe/Co Uiso (Å )
O z
2
O Uiso (Å )
Cl z
2
Cl Uiso (Å )
weight percent Sr
3 x 5 2
Fe2ꢀxCo O Cl
2
χ
4.6
1.88
1.529
2.318
1.284
diffractometer
D2B
D1A
1
D1A
D2B
POLARIS
a
Sr(1) is located at (0,0,0); Sr(2) at (0,0,z); Fe/Co at (0,0,z); O at ( /
2
,0,z); Cl at (0,0,z). Data for the x = 0 sample are taken from reference 10 for
comparison.
a
Table 3. Selected Bond Lengths Extracted from the Refinements of Sr Fe Co O Cl Phases
3
2ꢀx
x
4
2
Sr Fe2ꢀxCo
3
x
O
4
Cl
2
FeꢀO (Å)
FeꢀCl (Å)
Sr(1)ꢀO (Å)
Sr(2)ꢀO (Å)
Sr(2)ꢀCl (Å)
Sr(2)ꢀCl (Å)
x = 0
2.008(1)
2.010(1)
2.011(1)
2.011(1)
2.011(1)
2.972(3)
2.946(3)
2.941(5)
2.940(5)
2.964(2)
2.680(2)
2.681(1)
2.673(3)
2.670(1)
2.672(1)
2.610(2)
2.616(2)
2.617(4)
2.617(3)
2.616(1)
3.075(1)
3.079(1)
3.074(2)
3.068(2)
3.070(1)
3.228(3)
3.218(3)
3.220(5)
3.212(5)
3.188(2)
x = 0.2
x = 0.6
x = 0.8
x = 1
a
Data for the x = 0 sample are taken from reference 10 for comparison.
3
ꢀ1
paramagnet (C
=4.87cm Kmol assuming high-spin, S=2
oxide ions from these materials to yield products of composition
Sr Fe Co O Cl which are isostructural to the previously
expected
2+
3
2+
Fe and high-spin, S = / Co ). Below this temperature range
2
3
2ꢀx
x
4
2
10
there is a maximum in the magnetic susceptibility at T ∼ 25 K,
indicative of a magnetic transition. This feature is associated with
a sharp increase in the saturated ferromagnetic moment of the
sample, from an essentially temperature independent value of
1000 emu mol in the range 35 < T/K < 300 (attributable
to the presence of elemental iron and cobalt), to 2500 emu mol at
K. Magnetization-field isotherms collected from Sr FeCoO Cl
Figure 6) reflect this change in saturated ferromagnetic moment.
Furthermore close inspection of the field-cooled isotherm collected at
K (inset to Figure 6) reveal the magnetization-fieldloopisdisplaced
reported phase Sr Fe O Cl . The observation that the partial
3 2 4 2
substitution of iron with cobalt does not change the structural
selectivity of the reaction between Sr Fe O Cl and LiH is in
3
2
5
2
agreement with the strong lattice energy driven preference for
ꢀ
1
∼
the removal of the bridging oxide ion from Sr Fe O Cl , des-
3
2
5
2
ꢀ1
10
cribed previously. Furthermore the formation of square planar
Co(II) centers in the resulting Sr Fe Co O Cl phases ap-
5
(
3
4
2
3
2ꢀx
x
4
2
pears to present no energetic barrier to the reaction. This is
consistent with the relative prevalence of square planar Co O
II
4
5
centers in thermodynamically stable extended solids (in Sr Co-
2
16,17
to higher magnetization, consistent with spin-glass type behavior.
O Cl and Ba CoO Cu S for example)
compared to the
2
2
2
2
2 2
II
This suggests that the anomalies observed in the magnetic data at T∼
paucity of the analogous Fe O centers. The limited influence of
4
2
5 K are associated with a transition to a spin-glass like behavior.
cobalt on the formation of Sr Fe Co O Cl phases is in strong
3 2ꢀx x 4 2
Magnetic susceptibility data collected from the remaining
Sr Fe Co O Cl (x = 0.2, 0.6, 0.8) samples exhibit the same
general features as the data collected from Sr FeCoO Cl
2
contrast to that of the SrFeO3 ꢀSrFeO system, in which
ꢀδ
2
replacing more than 30% of the iron centers with cobalt or
3
2ꢀx
x
4
2
9
manganese changes the structure of the reduced product. This
3
4
(
Supporting Information). Table 4 lists the Curie constants and
difference strongly suggest that different factors are directing the
Weiss temperatures extracted from these data in the temperature
range 35 e T/K e 300.
anion vacancy order in the Sr Fe O Cl ꢀSr Fe O Cl and
3
2
5
2
3
2
4
2
SrFeO3 ꢀSrFeO systems, the A-cation lattice in the former
ꢀδ
2
case and some electronic factor in the latter.
The influence of cobalt substitution on the structural para-
’
DISCUSSION
meters of Sr Fe O Cl is minimal. The a-lattice parameter and
3
2
4
2
The reaction of phases in the composition range Sr Fe
in-plane CoꢀO bonds appear unchanged within error across the
3
2ꢀx-
Co O Cl (0 e x e 1) with LiH results in the deintercalation of
Sr Fe Co O Cl (0 e x e 1) series (Table 3). The c-lattice
x
5
2
3
2ꢀx
x
4
2
7
253
dx.doi.org/10.1021/ic200832n |Inorg. Chem. 2011, 50, 7250–7256

Inorganic Chemistry
ARTICLE
Figure 4. Plot of the refined ordered magnetic moment as a function of
temperature for Sr Fe Co O Cl (x = 0, 0.2, 0.6) phases. Inset shows
3
2ꢀx
x
4
2
the antiferromagnetically ordered structure of the Sr
3
Fe2ꢀxCo
x 4 2
O Cl
(
x = 0, 0.2, 0.6) phases. Data for the x = 0 sample are taken from ref 10.
data collected from Sr Fe Co O Cl samples are significantly
3
2ꢀx
x
4
2
smaller than would be expected for simple paramagnetic beha-
vior, representing between 6 and 25% of the value predicted for
spin-only behavior. These low values are entirely expected for
Sr Fe Co O Cl and Sr Fe Co O Cl as these phases
3
1.8
0.2
4
2
3
1.4
0.6
4
2
exhibit long-range antiferromagnetic order at temperatures be-
low 340 K and 300 K, respectively (Figure 4). In this instance the
observed paramagnetic susceptibility is attributable to uncoupled
spins at grain boundaries and other defect or surface sites. The
weak paramagnetic responses of the remaining Sr Fe Cox-
3
2ꢀx
O Cl (x g 0.8) phases are more surprising, as they imply that
4
2
even in the absence of long-range magnetic order there are still
strong antiferromagnetic correlations between spins. When com-
bined with the displaced magnetization-field loops at low tempera-
ture (Figure 6), this suggests that the substitution of iron with cobalt
Figure 3. Observed, calculated, and difference plots from the structural
refinement of Sr FeCoO Cl against room-temperature neutron pow-
3 4 2
der diffraction data collected using the POLARIS diffractometer. The
adds additional magnetic interactions into the Sr Fe Co O Cl
3
2ꢀx
x
4
2
lower set of tick marks indicate peak positions for Sr FeCoO Cl , upper
3
4
2
system which compete with and frustrate long-range antiferro-
magnetic order, ultimately leading to glassy magnetic behavior.
As noted previously the local electronic configuration of
tick marks indicate peak positions for the vanadium sample holder.
2
2
1
1
parameter contracts by ∼0.7% across the compositional range
associated with a general contraction of the structure in this
direction. The slight scatter in the data in Tables 2 and 3 as a
function of cobalt concentration is attributable to the systematic
errors introduced by collecting the neutron powder diffraction
data utilized in the structural refinements, using three different
diffractometers (Table 1).
Fe(II) centers in Sr Fe O Cl is (d ) (d , d ) (d ) (d
z2
2
2)
3
2
4
2
xz yz
xy
x ꢀy
1
0,18
as shown in Figure 7a.
The GoodenoughꢀKanamouri
19
rules predict strong antiferromagnetic superexchange within
the FeO planes of Sr Fe O Cl through 180°, σ-type Fe
2
3
2
4
2
1
1
(d
2
2
) ꢀO (2p)ꢀFe (d
2
2
) coupling (Figure 7c). In addi-
x ꢀy
x ꢀy
tion we would expect antiferromagnetic direct σ-exchange
2
coupling between the two full Fe (dz ) orbitals, expectations
2
The influence of cobalt substitution on the magnetic behavior
which are consistent with the high-temperature antiferromag-
netic order observed in the all-iron phase. The electronic con-
of Sr Fe O Cl is more striking. As shown in Figure 4, substitu-
3
2
4
2
II
tion of iron with cobalt leads to as steady depression of the
antiferromagnetic ordering temperature of Sr Fe Co O Cl
figuration of square-planar Co O centers is predicted to be
4
2
3
1
1
2
(dz
2
) (d , d ) (d ) (d
2
) (Figure 7b) by analogy to other
structurally related phases containing square-planar Co(II)
centers. Thus the partial substitution of Fe(II) with Co(II) in
Sr Fe Co O Cl phases retains local electronic configurations
on all transition metal centers suitable for strong antiferromag-
netic σ-exchange coupling (i.e., a full (dz2) orbital and a half-full
) orbital). However, the presence of an additional elec-
tron in the degenerate d dyz pair of orbitals of Co(II) centers
xz,
3
2ꢀx
x
4
2
xz yz
xy
x ꢀy
phases followed by a complete suppression of long-range mag-
netic order for samples with x g 0.8. The mechanism by which
cobalt substitution suppresses long-range magnetic order can be
elucidated by examining the paramagnetic susceptibility data
17
3
2ꢀx
x
4
2
2
collected from Sr Fe Co O Cl samples and considering the
3
2ꢀx
x
4
2
1
magnetic coupling interactions present in these phases. As shown
(d
2
2
x ꢀy
in Table 4, the Curie constants extracted from the susceptibility
7
254
dx.doi.org/10.1021/ic200832n |Inorg. Chem. 2011, 50, 7250–7256

Inorganic Chemistry
ARTICLE
Figure 6. Magnetization-field isotherms collected at 300 K (red) and
5
3 4 2
K (black) from Sr FeCoO Cl . Inset shows an expanded region of the 5 K
isotherm emphasizing that the data are displaced to higher magnetizations.
Table 4. Parameters Extracted from Fits to the Magnetic
Susceptibility Data Collected from Sr Fe Co O Cl
Samples in the Temperature Range 35 e T/K e 300
3
2ꢀx
x
4
2
a
3
ꢀ1
3
ꢀ1
Sr
3
x
Fe2ꢀxCo O
4
Cl
2
C
obs (cm K mol
)
C
exp (cm K mol
)
θ (K)
x = 0.2
x = 0.6
x = 0.8
x = 1
0.357(3)
1.213(5)
0.732(9)
0.788(4)
5.775
5.325
5.10
ꢀ2.4(6)
ꢀ2.3(3)
ꢀ5.6(4)
ꢀ9.7(3)
4.875
a
Expected Curie constants are calculated using the spin-only moments
of high-spin Fe (S = 2) and high-spin Co (S = / ).
Figure 5. Magnetization data collected from Sr FeCoO Cl using the
3
4
2
2+
2+
3
2
“
ferromagnetic subtraction” method. Upper panel shows the paramag-
netic susceptibility as a function of temperature. Inset shows a fit to the
CurieꢀWeiss law in the temperature range 35 e T/K e 300. Lower
panel shows the saturated ferromagnetic moment as a function of
temperature.
of cobalt concentration. Room-temperature M €o ssbauer spectra
collected from these phases allow the authors to estimate that a
3
0% cobalt substitution level reduces the antiferromagnetic
9
ordering temperature by “no more than 50 K”. While it is hard
to accurately quantify the influence of the cobalt concentration
2
introduces the possibility of π-type Co (d /d ) ꢀO (2p )ꢀFe
xz yz
z
1
(
d /d ) (Figure 7d) superexchange coupling which would be
on the magnetic behavior of SrFe Co O from the data
xz yz
1ꢀx x 2
expected to be ferromagnetic. This ferromagnetic interaction will
presented, it does appear that the substitution of cobalt for iron
compete with the antiferromagnetically coupled σ-exchange
in SrFeO has a weaker disruptive influence on the antiferro-
2
1
network and other π-type couplings such as Co(d /d ) ꢀ
xz yz
magnetic order exhibited by the undoped phase, than the analo-
gous substitution in the Sr Fe Co O Cl system (at 30% sub-
1
O(2p )ꢀFe(d /d ) (Figure 7c) which are also antiferromag-
z
xz yz
3
2ꢀx
x
4
2
netic. We propose that it is this competition between the ferro-
magnetic π-superexchange coupling (attributable to the extra
valence electron associated with Co(II) compared with Fe(II))
and the underlying network of antiferromagnetic interactions
which leads to the suppression of long-range magnetic order on
stitution ΔT = 80 K for Sr Fe Co O Cl ; ΔT < 50 K for
N 3 2ꢀx x 4 2 N
SrFe1 Co O ). The greater robustness to disruption by sub-
ꢀx
x 2
stitution of the underlying antiferromagnetically ordered net-
work of SrFeO is attributable to the more three-dimensional
2
nature of the SrFeO structure compared to the more two-
2
cobalt substitution, and ultimately glassy behavior in Sr Fe
-
dimensional Sr Fe O Cl This is also consistent with the observed
3
2ꢀx
3 2 4 2.
Co O Cl (x g 0.8) phases.
antiferromagnetic ordering temperatures of 473 and 378 K for
SrFeO and Sr Fe O Cl , respectively. We would therefore
x
4
2
Recently Kageyama et al. have reported the successful pre-
2
3
2
4
2
9
paration of cobalt substituted SrFeO , the three-dimensional
predict that cation substitutions at the iron coordination sites
2
6
structural analogue of Sr Fe O Cl . In this study it is observed
in the one-dimensional phase Sr Fe O would have an even
3
2
4
2
3
2 5
that in the limited substitution range that can be prepared
0 e x e 0.3) cobalt doping has only a minor effect on the
magnetic behavior of SrFe1 Co O phases. All compositions
exhibit long-range antiferromagnetic order at room-temperature
with observed ordered moments which are almost independent
greater effect on the magnetic behavior of this material than those
(
observed for either SrFeO or Sr Fe O Cl .
2
3
2
4
2
In conclusion, we have prepared phases of composition
Sr Fe Co O Cl (0 e x e 1) which contain sheets of apex-
ꢀx
x 2
3
2ꢀx
x
4
2
linked square-planar (Fe/Co)O centers. Addition of cobalt
4
7
255
dx.doi.org/10.1021/ic200832n |Inorg. Chem. 2011, 50, 7250–7256

Inorganic Chemistry
ARTICLE
(2) Crespin, M.; Landron, C.; Odier, P.; Bassat, J. M.; Mouron, P.;
Choisnet, J. J. Solid State Chem. 1992, 100, 281.
3) Siegrist, T.; Zahurak, S. M.; Murphy, D. W.; Roth, R. S. Nature
988, 334, 231.
(
1
(4) Weller, M. T.; Lines, D. R. J. Solid State Chem. 1089, 82, 21.
(5) Tsujimoto, Y.; Tassel, C.; Hayashi, N.; Watanabe, T.; Kageyama,
H.; Yoshimura, K.; Takano, M.; Ceretti, M.; Ritter, C.; Paulus, W. Nature
007, 450, 1062.
6) Hiroshi, K.; Takashi, W.; Yoshihiro, T.; Atsushi, K.; Yuji, S.;
2
(
Kazuyoshi, K.; Kazuyoshi, Y.; Naoaki, H.; Shigetoshi, M.; Mikio, T.;
Monica, C.; Werner, P.; Clemens, R.; Gilles, A. Angew. Chem. 2008,
4
2
1
7, 5740.
(7) O’Malley, M.; Lockett, M. A.; Hayward, M. A. J. Solid State Chem.
007, 180, 2851.
(
8) Seddon, J.; Suard, E.; Hayward, M. A. J. Am. Chem. Soc. 2010,
32, 2802.
9) Seinberg, L.; Yamamoto, T.; Tassel, C.; Kobayashi, Y.; Hayashi,
(
N.; Kitada, A.; Sumida, Y.; Watanabe, T.; Nishi, M.; Ohoyama, K.;
Yoshimura, K.; Takano, M.; Paulus, W.; Kageyama, H. Inorg. Chem.
2
011, 50 (9), 3988–3995.
(
(
10) Dixon, E.; Hayward, M. A. Inorg. Chem. 2010, 49, 9649.
11) Knee, C. S.; Field, M. A. L.; Weller, M. T. Solid State Sci. 2004,
2
+
Figure 7. (a) Local electronic state of square-planar Fe centers.
(
6, 443.
(12) Grenier, J. C.; Ghodbane, S.; Demazeau, G.; Pouchard, M.;
Hagenmuller, P. Mater. Res. Bull. 1979, 14, 831.
(13) Adkin, J. J.; Hayward, M. A. Inorg. Chem. 2008, 47, 10959.
(14) Larson, A. C.; Von Dreele, R. B. General Structure Analysis
System (GSAS), Los Alamos National Laboratory Report LAUR 86-
2+
b) The local electronic state of square-planar Co centers. (c) A repre-
sentation of the orbitals involved in antiferromagnetic σ-type Fe/Co
) superexchange. (d) A representa-
tion of the orbitals involved in ferromagnetic π-type Co (d /d ) ꢀO
xz yz
1
1
(d
2
2
) ꢀO (2p)ꢀFe/Co (d
2
2
x ꢀy
x ꢀy
2
1
(2p
z
)ꢀFe (dxz/dyz) superexchange.
7
48; Los Alamos National Laboratory: Los Alamos, NM, 2000.
(
15) Hayward, M. A.; Rosseinsky, M. J. Chem. Mater. 2000, 12, 2182.
suppresses the antiferromagnetic order of the Sr Fe O Cl host
(16) Knee, C. S.; Weller, M. T. Phys. Rev. B 2004, 70, 144406.
(17) Smura, C. F.; Parker, D. R.; Zbiri, M.; Johnson, M. R.; Gal, Z. A.;
3
2
4
2
phase leading to spin-glass behavior in Sr Fe Co O Cl
phases with x g 0.8.
3
2ꢀx
x
4
2
Clarke, S. J. J. Am. Chem. Soc. 2011, 133, 2691.
18) Xiang, H. J.; Wei, S.-H.; Whangbo, M. H. Phys. Rev. Lett. 2008,
00, 167207.
19) Goodenough, J. B. Magnetism and the Chemical Bond; Wiley:
New York, 1963.
(
1
’
ASSOCIATED CONTENT
Supporting Information. Description of the magnetic
(
S
b
measurement procedure for samples containing elemental cobalt
impurities. The refined lattice parameters of Sr Fe Co O Cl
3
2ꢀx
x
5
2
(0 e x e 1) phases. Observed, calculated, and difference plots
from the room-temperature structural refinements of Sr Fe
-
2ꢀx
3
Co O Cl (x = 0.2, 0.6, 0.8) and the structural and magnetic
x
4
2
refinements of Sr Fe Co O Cl (x = 0.2, 0.6) against data
3
2ꢀx
x
4
2
collected at 5 K. Magnetization data collected from Sr Fe
-
3
2ꢀx
Co O Cl (x = 0.2, 0.6, 0.8). This material is available free of
x
4
2
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
’
*
Phone: +44 1865 272623. Fax: +44 1865 272690. E-mail:
michael.hayward@chem.ox.ac.uk.
’
ACKNOWLEDGMENT
We thank R. Smith and E. Suard for assistance in collecting the
neutron diffraction data. Experiments at the ISIS pulsed neutron
facility were supported by a beam time allocation from the
Science and Technology Facilities Council. E.D. thanks the
EPSRC for a studentship.
’
REFERENCES
(1) Hayward, M. A.; Green, M. A.; Rosseinsky, M. J.; Sloan, J. J. Am.
Chem. Soc. 1999, 121, 8843.
7
256
dx.doi.org/10.1021/ic200832n |Inorg. Chem. 2011, 50, 7250–7256