Synthesis, structure and magnetic properties of the pillared perovskites La5Re3MO16(M=Mg, Fe, Co, Ni)

Article abstract of DOI:10.1016/S0022-4596(02)00058-0

Four new compounds La₅Re₃MgO₁₆ La₅Re₃FeO₁₆ La₅Re₃CoO₁₆ La₅Re₃NiO₁₆ have been prepared by solid-state reaction and characterized by X-ray and neutron powder diffraction and SQUID magnetometry. Rietveld refinement revealed that the four compounds are isostructural with La₅Re₃MnO₁₆ and crystallize in space group C1 with cell parameters a=7.9370(3), 7.9553(5), 7.9694(7), and 7.9383(4) A; b=7.9998(3), 7.9960(6), 8.0071(8), and 7.9983(5) A; c=10.1729(4), 10.1895(7), 10.182(1), and 10.1732(6) A; α=90.190(3)°, 90.270(3)°, 90.248(4)°, 90.287(3)°; β=94.886(2)°, 95.082(3)°, 94.980(4)°, 94.864(3)°; γ=89.971(4)°, 90.001(5)°, 89.983(6)°, 89.968(4)° for Mg, Fe, Co, and Ni, respectively. The structures are related to a layered perovskite. The layers of corner-sharing octahedra Re⁵⁺M²⁺O₆ (M²⁺=Mg, Fe, Co, Ni) are pillared by diamagnetic edge-sharing octahedra dimers, Re₂O₁₀, involving a Re=Re double bond. Three crystallographically independent lanthanum atoms occupy the three-dimensional interstices. All compounds obey the Curie-Weiss law at sufficiently high temperatures with Curie constants or effective magnetic moments near the expected values for the combination of Re⁵⁺ (S=1) and M²⁺ (S=0, 2, 3/2, 1 for Mg, Fe, Co, and Ni, respectively). Weiss constants, θ_C, are negative (-575, -84, -71, and -217 K for Mg, Fe, Co, and Ni, respectively) indicating the predominance of antiferromagnetic exchange coupling. The phases for M=Fe, Co and Ni show long-range order at 155, 33, 36 and 14 K, respectively. Neutron diffraction discloses a magnetic structure for the Fe series member consisting of ferrimagnetic perovskite layers coupled antiparallel along the stacking c-axis, direction which is consistent with the magnetic structure found recently for La₅Re₃MnO₁₆.

Full text of DOI:10.1016/S0022-4596(02)00058-0

Journal of Solid State Chemistry 170 (2003) 165–175
Synthesis, structure and magnetic properties of the pillared
perovskites La5Re3MO16(M=Mg, Fe, Co, Ni)
Lisheng Chi,a A.E.C. Green,a R. Hammond,b C.R. Wiebe,c and J.E. Greedana,*
a Department of Chemistry, Brockhouse Institute for Materials Research, McMaster University, Hamilton, Ontario L8S 4M1, Canada
b Neutron Program for Materials Research, NRC, McMaster University, Canada
c Department of Physics and Brockhouse, Institute for Materials Research, McMaster University, Hamilton, Ontario L8S 4M1, Canada
Received 14 May 2002; received in revised form 9 September 2002; accepted 25 September 2002
Abstract
Four new compounds La5Re3MgO16 La5Re3FeO16 La5Re3CoO16 La5Re3NiO16 have been prepared by solid-state reaction and
characterized by X-ray and neutron powder diffraction and SQUID magnetometry. Rietveld refinement revealed that the four
%
compounds are isostructural with La5Re3MnO16 and crystallize in space group C1 with cell parameters a ¼ 7:9370ð3Þ; 7.9553(5),
(
(
7.9694(7), and 7.9383(4) A; b ¼ 7:9998ð3Þ; 7.9960(6), 8.0071(8), and 7.9983(5) A; c=10.1729(4), 10.1895(7), 10.182(1), and
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
(
10.1732(6) A; a ¼ 90:190ð3Þ ; 90.270(3) , 90.248(4) , 90.287(3) ; b ¼ 94:886ð2Þ ; 95.082(3) , 94.980(4) , 94.864(3) ; g ¼ 89:971ð4Þ ;
90.001(5)ꢀ, 89.983(6)ꢀ, 89.968(4)ꢀ for Mg, Fe, Co, and Ni, respectively. The structures are related to a layered perovskite. The layers
of corner-sharing octahedra Re5+ 2+O6 (M2+=Mg, Fe, Co, Ni) are pillared by diamagnetic edge-sharing octahedra dimers,
M
Re2O10, involving a Re=Re double bond. Three crystallographically independent lanthanum atoms occupy the three-dimensional
interstices. All compounds obey the Curie–Weiss law at sufficiently high temperatures with Curie constants or effective magnetic
moments near the expected values for the combination of Re5+(S ¼ 1) and M2+(S=0, 2, 3/2, 1 for Mg, Fe, Co, and Ni,
respectively). Weiss constants, yC; are negative (À575, À84, À71, and À217 K for Mg, Fe, Co, and Ni, respectively) indicating the
predominance of antiferromagnetic exchange coupling. The phases for M=Fe, Co and Ni show long-range order at 155, 33, 36 and
14 K, respectively. Neutron diffraction discloses a magnetic structure for the Fe series member consisting of ferrimagnetic perovskite
layers coupled antiparallel along the stacking c-axis, direction which is consistent with the magnetic structure found recently for
La5Re3MnO16.
r 2002 Elsevier Science (USA). All rights reserved.
1. Introduction
Compared to 3d transition metal oxides, their 4d and
5d counterparts have been much less explored. The 4d
and 5d transition metal oxides exhibit different crystal
structures and crystal chemistry from 3d oxides owing to
the more extended nature of the 4d and 5d orbitals, the
predominance of higher oxidation states and a tendency
to form metal–metal bonds, among others. Recently, a
new perovskite-related structure type has been reported
independently by two groups. [3,4] The composition is
Ln5M3M0O16 with Ln=a lanthanide and M and
M0=Mo or M=Re and M0=Mn. The structure
features perovskite-like corner-sharing octahedral
layers, Mo2O6 or MnReO6, which are pillared by
Mo2O10 or Re2O10 dimeric units involving metal–metal
double bonds. Very recent results for the Mo-based
series indicate that divalent transition metal ions can
partially occupy one of the intralayer Mo sites [5]. The
Ln3+ ions occupy interstitial sites. The interlayer
In recent decades, layered perovskite-type transition
metal oxides have attracted great attention in the fields
of chemistry, physics and materials. Obvious examples
are the cuprate superconductors which are structurally
related to ABO3 perovskites and the alkali earth-doped
rare-earth manganate perovskites La1ÀXAXMnO3
(X=Sr, Ca, Ba) [1] which exhibit fascinating physical
phenomena such as colossal magnetoresistance, metal–
insulator transitions, charge ordering and orbital order-
ing. Another perovskite-type compound Sr2FeMoO6 [2]
is a so-called half-metallic ferromagnet with a transition
temperature far above room temperature.
*Corresponding author.
E-mail address: greedan@mcmail.cis.mcmaster.ca (J.E. Greedan).
0022-4596/03/$ - see front matter r 2002 Elsevier Science (USA). All rights reserved.
PII: S 0 0 2 2 - 4 5 9 6 ( 0 2 ) 0 0 0 5 8 - 0

166
L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
(
separation is B10 A but, surprisingly, both series of
compounds show evidence for long-range magnetic
order at rather high temperatures, 160–180 K [4,6,7]. In
this work, the crystal chemistry and structure/property
relationships of related phases in which various divalent
metal ions M2+=Mg, Fe, Co, Ni are substituted for
Mn2+ in La5Re3MnO16 are explored. The details of
their synthesis, structure, magnetism and in the case of
the Fe material, the magnetic structure are reported.
2.3. Magnetic properties
Magnetic susceptibility data were collected using a
Quantum Design MPMS SQUID magnetometer. Zero-
field cooled and field cooled (ZFC/FC ) and isothermal
magnetization measurements were performed on sam-
ples in gelatin capsule containers. An oven insert was
used to collect high-temperature(300–600 K) suscept-
ibility data.
3. Results and discussion
2. Experimental
3.1. Structural characterization
2.1. Synthesis
As no single crystals were available, structure analysis
was carried out with powder diffraction data, both X ray
and neutron. The Guinier powder patterns for M=Mg,
Fe, Co and Ni were observed to be similar to that for
La5Re3MnO16 and all reflections could be indexed on a
C-centered triclinic cell. Consequently, the refinements of
the room temperature powder diffraction data were
carried out with the La5Re3MnO16 model in space group
The starting reagents La2O3(99.9%, Aldrich),
CoO(99.5%, CERAC), NiO(99.995%, CERAC),
MgO(99.95%, Alfa Aesar), ReO3 (rhenium alloys) and
Re (rhenium alloys) were used to synthesize polycrystal-
line title compounds La5Re3MO16 (M=Mg, Fe, Co, Ni)
according to the reaction
5La2O3 þ 5ReO3 þ Re þ MO
-2La5Re3MO16ðM ¼ Mg; Fe; Co; NiÞ:
%
C1 [4]. The FULLPROF program [9] in the WIN-
PLOTR format was used with both X-ray and neutron
data. Not surprisingly, analysis of the neutron data
yielded superior results with respect to the metal–oxygen
distances and geometry and these are presented here. The
unit-cell constants and agreement indices are shown in
Table 1 with the final atomic coordinates displayed in
Tables 2–5. Selected bond lengths and angles are listed
in Table 6. The observed, calculated and difference
diffraction profiles are plotted in Figs. 1a–d.
As can be seen from examination of Tables 2–5, it was
not always possible, even with the neutron data to
refine, individually, the displacement parameters for all
of the atoms to positive, definite values. In fact this was
accomplished only for the M=Fe case. In the case of
La5Re3CoO16 the displacement factors were fixed to the
values of La5Re3MnO16.
La2O3 was dried at 800ꢀC for overnight before use. FeO
was prepared by firing Fe2O3 (Fisher) and Fe at 900ꢀC
for 12 h. Because of the well-known nonstoichiometry
for FeO [8], it was necessary to add excess Fe to obtain a
stoichiometric product for La5Re3FeO16 reaction.
Stoichiometric amounts La2O3, ReO3, Re and MO
(M=Mg, Fe, Co, Ni) were accurately weighed, ground
together, loaded in a platinum crucible, sealed into a
quartz tube under high vaccum (410À5 mmHg), and
reacted at 1050ꢀC for 48 h.
2.2. X-ray and neutron diffraction
X-ray powder diffraction data were obtained using a
Guinier-Hagg
camera
with
CuKa1
radiation
(
(l ¼ 1:54056 A) and high-purity silicon powder as an
internal standard. As well, diffraction data were
collected on a Bruker D8 diffractometer using CuKa
radiation in the 2y range from 10ꢀ to 80ꢀ. A 2y stepsize
of 0.025ꢀ and a steptime of 12 s were used. Neutron
diffraction data collections were performed on the C2
diffractometer operated by the Neutron Program for
Materials Research at the AECL Chalk River Labora-
tories. The angular ranges 10ꢀp2yp116.9ꢀ at room
temperature (298 K) and 3ꢀp2yp81.9ꢀ at low tempera-
ture (below 160 K) were scanned with the same step
interval of 0.1ꢀ. Two different wavelengths, 1.328 and
From these results it is clear that all four compounds
La5Re3MO16 (M=Mg, Fe, Co, Ni) are isostructural
with La5Re3MnO16, which has a pillared perovskite
structure as illustrated for La5Re3FeO16 in Fig. 2. The
upper part shows the perovskite plane consisting of iron
and rhenium ions which are octahedrally coordinated.
Intralayer cation order is evident. The iron and
rhenium(2) octahedra are interlinked through O7 along
the b direction and O8 along the a direction. As
determined for the M=Mn material, the formal
oxidation states of Re and M are +5 and +2,
respectively[4]. The unit-cell volume scales roughly with
the M2+ radius as seen in Fig. 3.
(
2.372 A, were used, respectively, to collect room
temperature and low temperature data for crystal
structure refinement (short wavelength) and magnetic
structure determination (long wavelength).
The perovskite layers are stacked along c direction
and separated or pillared by dimeric Re(1)2O10 units
formed by Re(1)O6 octahedra sharing a common edge

L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
167
Table 1
The final refinement results for La5Re3MO16 (M=Mg, Fe, Co, Ni)
Compound
La5Re3MgO16
La5Re3FeO16
Neutron
La5Re3CoO16
Neutron
La5Re3NiO16
Neutron
Data source
Space group
Cell parameters
Neutron
%
C1
%
C1
%
C1
%
C1
(
(
(
(
a ¼ 7:9359ð5Þ A
a ¼ 7:9553ð5Þ A
a ¼ 7:9694ð7Þ A
a ¼ 7:9383ð4Þ A
(
(
(
(
b ¼ 7:9963ð5Þ A
b¼ 7:9960ð6ÞA
b ¼ 8:0071ð8Þ A
b ¼ 7:9983ð5Þ A
(
(
(
(
c ¼ 10:1683ð7Þ A
a ¼ 90:196ð4Þꢀ
b ¼ 94:878ð3Þꢀ
g ¼ 89:966ð6Þꢀ
c ¼ 10:1895ð7Þ A
a ¼ 90:270ð3Þꢀ
b ¼ 95:082ð3Þꢀ
g ¼ 90:001ð5Þꢀ
c ¼ 10:182ð1Þ A
c ¼ 10:1732ð6Þ A
a ¼ 90:287ð3Þꢀ
b ¼ 94:864ð3Þꢀ
g ¼ 89:968ð4Þꢀ
a ¼ 90:248ð4Þꢀ
b ¼ 94:980ð4Þꢀ
g ¼ 89:983ð6Þꢀ
3
3
3
3
(
(
(
(
V ¼ 642:92ð7Þ A
V ¼ 645:61ð8Þ A
V ¼ 647:3ð1Þ A
V ¼ 643:60ð6Þ A
Agreement factors
Rp ¼ 2:92
Rwp ¼ 3:93
w2 ¼ 3:20
RB ¼ 3:59
Rp ¼ 3:77
Rwp ¼ 4:92
w2 ¼ 1:24
RB ¼ 5:76
Rp ¼ 2:72
Rwp ¼ 3:55
w2 ¼ 1:21
RB ¼ 5:71
Rp ¼ 3:22
Rwp ¼ 4:25
w2 ¼ 2:95
RB ¼ 4:37
Table 2
Atomic coordinates and displacement factors, B (A ), for La5Re3
MgO16
Table 4
Atomic coordinates for La5Re3CoO16
2
(
Name
x
y
z
2
B (A )
(
Name
x
y
z
La1
La2
La3
Re1
Re2
Co
0.227(1)
0.233(1)
0.50000
0.746(1)
0.264(1)
0.50000
0.499(1)
0.00000
0.50000
0.508(2)
0.499(2)
0.502(2)
0.010(2)
0.735(2)
0.266(2)
0.243(1)
0.048(1)
0.8024(9)
0.800(1)
0.50000
La1
La2
La3
Re1
Re2
Mg
O1
0.228(1)
0.230(1)
0.50000
0.746(1)
0.263(1)
0.50000
0.499(1)
0.00000
0.50000
0.506(2)
0.499(2)
0.502(2)
0.006(2)
0.734(1)
0.264(1)
0.247(1)
0.0404(9)
0.8027(7)
0.798(8)
0.50000
0.1(1)
0.5(1)
0.5(1)
0.1
0.0565(8)
0.00000
0.00000
0.3956(6)
0.00000
0.00000
0.0566(6)
0.00000
0.00000
0.3936(4)
0.00000
0.00000
0.1
0.4(2)
0.4(1)
0.5(1)
0.2(1)
0.4(1)
0.2(2)
0.7(2)
0.9(2)
0.5(2)
O1
O2
0.187(1)
0.280(1)
À0.040(1)
0.072(1)
0.036(2)
0.041(2)
À0.055(1)
0.236(1)
0.5662(9)
0.323(1)
0.199(1)
0.184(1)
0.367(1)
0.367(1)
À0.003(1)
À0.034(1)
0.1925(9)
0.2793(9)
À0.0393(8)
0.0720(9)
0.037(2)
0.039(1)
À0.048(1)
0.237(1)
0.5653(7)
0.3230(7)
0.1977(7)
0.1845(8)
0.365(1)
O2
O3
O3
O4
O4
O5
O5
O6
O6
O7
0.369(1)
À0.001(1)
À0.0346(8)
O7
O8
O8
Table 3
Atomic coordinates and displacement factors, B (A ), for La5Re3
FeO16
Table 5
Atomic coordinates and displacement factors, B (A ), for La5Re3
NiO16
2
2
(
(
2
B (A )
2
B (A )
(
(
Name
x
y
z
Name
x
y
z
La1
La2
La3
Re1
Re2
Fe
0.232(1)
0.231(1)
0.50000
0.745(1)
0.264(1)
0.50000
0.499(1)
0.00000
0.50000
0.504(2)
0.498(2)
0.502(2)
À0.001(2)
0.738(1)
0.268(2)
0.243(2)
0.043(1)
0.8007(9)
0.798(1)
0.50000
0.1(1)
0.7(2)
0.4(2)
0.04(8)
0.04(8)
0.7(2)
0.6(2)
0.5(2)
0.4(2)
0.9(2)
0.3(2)
0.8(2)
1.0(2)
0.7(2)
La1
La2
La3
Re1
Re2
Ni
0.230(1)
0.229(1)
0.50000
0.7457(9)
0.2634(8)
0.50000
0.4985(9)
0.00000
0.50000
0.507(1)
0.499(1)
0.501(1)
0.005(1)
0.736(1)
0.265(1)
0.246(1)
0.0446(8)
0.8027(7)
0.7999(7)
0.50000
0.3(1)
0.3(1)
0.6(1)
0.2
0.0567(8)
0.00000
0.00000
0.393(5)
0.00000
0.00000
0.0572(5)
0.00000
0.00000
0.3936(4)
0.00000
0.00000
0.2
0.4(1)
0.42(4)
0.42(4)
0.42(4)
0.42(4)
0.42(4)
0.42(4)
0.42(4)
0.42(4)
O1
O2
0.190(1)
0.277(1)
À0.041(1)
0.071(1)
0.036(2)
0.037(2)
À0.049(1)
0.238(1)
0.5679(9)
0.3239(9)
0.200(1)
0.182(1)
0.365(1)
0.368(1)
0.001(1)
À0.034(1)
O1
O2
0.1906(9)
0.2805(8)
À0.0423(8)
0.0706(8)
0.036(1)
0.5664(7)
0.3248(7)
0.1972(7)
0.1849(7)
0.3632(9)
0.3685(9)
À0.0004(8)
À0.0349(7)
O3
O4
O3
O4
O5
O6
O5
O6
0.040(1)
À0.0505(9)
0.2366(9)
O7
O8
O7
O8

168
L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
Table 6
Selected bond distances and angles for La5Re3MO16 (M=Mg, Fe, Co, Ni)
La5Re3MgO16
La5Re3FeO16
La5Re3CoO16
La5Re3NiO16
La(1)–O(2)
La(1)–O(3)
La(1)–O(4)
La(1)–O(5)
La(1)–O(6)
La(1)–O(7)
La(1)–O(7)
La(1)–O(8)
La(2)–O(2)
La(2)–O(3)
La(2)–O(4)
La(2)–O(5)
La(2)–O(6)
La(2)–O(8)
La(2)–O(8)
La(3)–O(1)
La(3)–O(2)
La(3)–O(5)
La(3)–O(6)
Re(1)–Re(1)
Re(1)–O(1)
Re(1)–O(1)
Re(1)–O(2)
Re(1)–O(3)
Re(1)–O(5)
Re(1)–O(6)
Re(2)–O(4)
Re(2)–O(7)
Re(2)–O(8)
O(4)–Re(2)–O(7)
O(4)–Re(2)–O(7)
O(4)–Re(2)–O(8)
O(4)–Re(2)–O(8)
O(7)–Re(2)–O(8)
O(7)–Re(2)–O(8)
M–O(3)
2.41(2)
2.48(2)
2.56(1)
2.42(2)
2.50(2)
2.50(2)
2.41(2)
2.49(2)
2.60(2)
2.42(1)
2.47(1)
2.55(1)
2.64(1)
2.63(1)
2.61(2)
2.63(2)
2.67(2)
2.63(2)
2.62(1)
2.64(1)
2.57(1)
2.55(1)
2.59(2)
2.57(2)
2.55(2)
2.52(2)
2.56(1)
2.54(1)
2.87(1)
2.42(2)
2.91(1)
2.43(2)
2.93(1)
2.44(2)
2.90(1)
2.45(1)
2.42(2)
2.42(1)
2.41(2)
2.46(2)
2.42(2)
2.38(2)
2.40(1)
2.44(1)
2.58(1)
2.61(1)
2.57(2)
2.62(2)
2.62(2)
2.59(2)
2.57(1)
2.62(1)
2.46(1)
2.86(1)
2.47(1)
2.85(1)
2.42(1)
2.80(1)
2.43(1)
2.83(1)
2.583(8) ꢁ 2
2.402(7) ꢁ 2
2.55(1) ꢁ 2
2.53(1) ꢁ 2
2.412(6)
1.973(8)
2.056(9)
1.963(8)
2.072(9)
1.91(1)
2.617(9) ꢁ 2
2.407(9) ꢁ 2
2.53(1) ꢁ 2
2.57(1) ꢁ 2
2.429(8)
1.99(1)
2.64(1) ꢁ 2
2.40(1) ꢁ 2
2.54(1) ꢁ 2
2.57(1) ꢁ 2
2.380(8)
1.95(1)
2.603(7) ꢁ 2
2.385(7) ꢁ 2
2.553(9) ꢁ 2
2.549(9) ꢁ 2
2.419(6)
1.975(8)
2.043(8)
1.96(1)
2.04(1)
1.95(1)
2.02(1)
1.99(1)
2.05(1)
1.94(1)
1.87(2)
2.08(1)
1.93(2)
1.89(2)
2.085(8)
1.93(1)
1.88(1)
1.89(1)
1.915(8) ꢁ 2
2.01(1) ꢁ 2
1.969(8) ꢁ 2
91.4(7)
1.89(1) ꢁ 2
1.98(1) ꢁ 2
1.98(1) ꢁ 2
92.4(9)
1.91(1) ꢁ 2
2.00(1) ꢁ 2
1.98(1) ꢁ 2
91.4(9)
1.917(7) ꢁ 2
2.009(8) ꢁ 2
1.974(8) ꢁ 2
91.7(6)
88.6(7)
87.8(6)
87.6(9)
88.1(8)
88.6(9)
87.3(8)
88.3(6)
88.3(5)
92.2(6)
91.2(6)
88.8(6)
91.9(8)
91.4(8)
88.6(8)
92.7(8)
91.1(8)
88.9(8)
91.7(5)
90.8(5)
89.2(5)
2.060(7) ꢁ 2
2.06(1) ꢁ 2
2.111(8) ꢁ 2
88.1(7)
2.09(1) ꢁ 2
2.09(1) ꢁ 2
2.11(1) ꢁ 2
87.2(8)
2.08(1) ꢁ 2
2.10(1) ꢁ 2
2.14(1) ꢁ 2
88.1(8)
2.062(7) ꢁ 2
2.070(8) ꢁ 2
2.121(8) ꢁ 2
87.3(6)
M–O(7)
M–O(8)
O(3)–M–O(7)
O(3)–M–O(7)
O(3)–M–O(8)
O(3)–M–O(8)
O(7)–M–O(8)
O(7)–M–O(8)
Re(2)–O(7)–M
Re(2)–O(8)–M
92.0(7)
86.1(5)
92.8(8)
85.6(7)
91.9(8)
85.7(7)
92.7(6)
85.6(5)
93.9(5)
88.2(6)
94.4(7)
88.7(8)
94.3(7)
88.4(7)
94.4(5)
88.7(5)
91.8(6)
158.4(4)
152.9(3)
91.3(8)
157.7(5)
152.6(4)
91.6(7)
155.2(5)
150.8(4)
91.3(5)
157.3(3)
151.5(3)
O1–O1 as seen in the lower part of the figure. Within the
Re(1)2O10 unit, the Re(1)–Re(1) distances are 2.380(8),
dimer. A final important feature of the structure is the
connectivity between the perovskite layers. As seen from
Fig. 2, the dimers connect directly only the M2+ sites
and not the Re5+ sites in the layers.
(
2.397(5), 2.429(8), and 2.419(6) A for Co, Mg, Fe, and
Ni, respectively. Again, Re is formally Re+5, 5d2. A
similar, short Re–Re distance has been found in
(
Nd4Re2O11(Re–Re:2.421(1) A) [10] with the same for-
3.2. Magnetic measurements
mal Re oxidation state. It is reasonable to associate such
a Re–Re distance with a Re=Re double bond. In the
present compounds La5Re3MO16 (M=Mg, Fe, Co, Ni),
the four electrons in the Re2 pair can be used to form a
sigma and a pi bond. The existence of a Re=Re double
bond quenches the Re magnetic moment within the
3.2.1. La5Re3MgO16
Susceptibility data covering the range 2–600 K are
shown in Fig 4. Turning first to the high temperature
results seen in the inset, adherence to the Curie–Weiss
law, w ¼ C=ðT À yÞ; is found only above B500 K with

L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
169
12000
10000
8000
6000
4000
2000
0
2800
2400
2000
1600
1200
800
400
0
-400
-2000
0
20
40
60
2θ(o)
80
100
120
0
20
40
60
80
100
120
2θ(o)
(a)
(b)
10000
3500
3000
2500
2000
1500
1000
500
8000
6000
4000
2000
0
0
-2000
-500
0
20
40
60
80
100
120
0
20
40
60
80
100
120
2θ(o)
2θ(o)
(c)
(d)
Fig. 1. (a–d) Rietveld refinements of neutron diffraction data for La5Re3MO16. (a) M=Mg, (b) M=Fe, (c) M=Co, and (d) M=Ni.
fitting
parameters
C ¼ 1:00ð1Þ emu K/mol
and
In fact, a more careful analysis suggests that frustration
may play a dominant role in the intraplanar magnetic
interactions. As is seen in Fig. 6 below, a Re5+ spin will
interact with both four nearest neighbor (nn) at 5.63A
(
and four next-nearest neighbor (nnn) at 7.93A. Both
y ¼ À575ð12Þ K where C is Curie constant and y is the
Weiss constant. These values are consistent with a spin-
only moment for Re5+, 5d2, and surprisingly large, net,
antiferromagnetic spin–spin correlations, given that the
Re–Re interactions are mediated by O–Mg–O linkages.
The regime from 2 to 350 K is dominated by two
features, a strong paramagnetic response at the lowest
temperatures and a very broad feature beginning below
B300K. It is reasonable to interpret the broad feature in
terms of short-range magnetic correlations. This is
consistent with the layered nature of the crystal
structure and the very weak interlayer connectivity
between Re5+ ions as discussed earlier. A first effort to
characterize the short-range order involved an attempt
to fit to the S ¼ 1; two-dimensional model of lines which
is appropriate for a quadratic lattice. [11]. This is shown
in Fig. 5, where the low-temperature Curie–Weiss tail
has been subtracted. The fit is only fair, the g-factor is
unreasonably small at 0.77 and it was necessary to
include a large TIP term of B8 ꢁ 10À4 emu/mol. It is
clear that this approach is not satisfactory.
(
interactions involve similar superexchange pathways
involving Re–O–Mg–O–Re linkages and it is not
unreasonable to assume, given the very large, negative
Curie–Weiss temperature, that both the nn and nnn
interactions are antiferromagnetic and comparable in
magnitude. In such a circumstance frustration arises,
Fig. 6, as the Re spins are subject to competing exchange
constraints. Such a spin system might be expected to seek
an unconventional ground state such as a spin liquid or
spin glass. Note from Fig. 5 that no evidence for a ZFC/
FC divergence, a common signature for the spin glass or
other types of statically ordered state, is seen even down
to 2 K. As well, the isothermal magnetization data at 2 K
(not shown) show a nearly linear dependence with no
divergences out to 5.5 T. This suggests an unusual ground
state and studies are currently underway to characterize
this material in more detail.

170
L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
1200
1100
1000
900
3.0
2.5
2.0
1.5
1.0
800
300
400
500
600
Temperature(K)
0
100
200
300
400
Temperature(K)
Fig. 4. Molar susceptibility data for La5Re3MgO16 as a function of
temperature from 2 to 600 K. The applied field was 1.0 T. The inset
shows the inverse susceptibility up to 600 K and the fit to the Curie–
Weiss law above 450 K. Only the ZFC curve is shown, there is no
difference between the FC and ZFC curves down to 2 K.
0.0014
0.0012
0.0010
00080.
Fig. 2. Crystal structure of La5Re3FeO16. TopA perovskite-like layer
of corner-sharing FeO6 and ReO6 octahedra showing the chemical
ordering of the Fe2+ and Re5+ ions. Bottom: A view showing the
Re2O10 dimers which connect the perovskite-like layers. The solid
spheres are the La atoms. Note that the connection pathway involves
directly only the Fe2+ ions.
La5Re3MgO16
(Curie Tail Subtracted)
0.0006
0.0004
0.0002
0.0000
-0.0002
Fit to Lines' 2D S = 1 Model
656
Mn
654
652
650
0
50
100
150
200
250
300
350
400
T(K)
Fig. 5. An attempt to fit the susceptibility data for La5Re3MgO16
(after subtraction of a low temperature Curie tail) to the Lines’
function for the S=1 square lattice.
648
646
644
642
Co
Mg
Fe
Ni
200 K and reach a maximum at 155 K. When the applied
field is decreased to 100 Oe, the FC–ZFC divergence is
more significant than that with 1000 Oe and there is a
broad shoulder between 65 and 125 K. The Curie–Weiss
law fit in the temperature range 520–600 K (Fig. 8)
yields y ¼ À84 K, C ¼ 3:60ð2Þ(emu/mol K), comparable
to the sum of Fe2+(d6, C ¼ 3:00(emu/mol K)) and
Re5+(d2, C ¼ 1:00 (emu/mol K)). Fig. 8 shows the iso-
thermal magnetic field dependence of the magnetization
at 5, 60, 140 and 250 K. At 5 K, the magnetization
increases only gradually to H=4 T followed by a weak
upturn to 5.5 T. At 60 and 140 K, both magnetization
curves have sharpincreases at 2000 and 5000 Oe,
respectively, and reach saturation magnetic moments,
0.68 0.70 0.72 0.74 0.76 0.78 0.80 0.82 0.84
Radius(Å)
Fig. 3. Variation of unit cell volume with the M2+ radius [16].
3.3. La5Re3FeO16
The ZFC–FC magnetic susceptibilities of La5Re3
FeO16 in applied fields of 100 and 1000 Oe are plotted
as a function of temperature in Fig. 7. For both applied
fields, the ZFC and FC data begin to diverge below

L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
171
1.6
5K
60K
140K
250K
1.2
0.8
0.4
0.0
0
20000
40000
60000
Applied field(Oe)
Fig. 8. Isothermal magnetization versus applied field behavior for
La5Re3FeO16. Note the appearance of sharp jumps at low applied
fields for 140 and 60 K.
Fig. 6. Re–O–Mg–O–Re linkages between nearest neighbors and next-
nearest neighbors in the perovskite layers of La5Re3MgO16. The
arrows represent the spin on Re5+ ions. Note that if JnnBJnnn and
both are antiferromagnetic, the spins will be frustrated.
structure at least at zero applied field. For example,
the M vs H curves show sharptransitions at moderate
fields, features indicative of metamagnetic behavior, and
the best saturation magnetization values are more
consistent with AF Fe2+–Re5+ coupling. (msat:-2 mB)
rather than F coupling(msat:-6 mB).
40
200
Neutron diffraction is of course the most powerful
technique for the characterization of magnetic order and
experiments were carried out to determine the magnetic
structure of La5Re3FeO16. Fig. 9 shows neutron diffrac-
tion patterns in the low-angle region at selected
temperatures, emphasizing the development of the
major Bragg reflections. In the pattern at 13 K, the
most intense of the low-angle magnetic Bragg peaks
occur at 2y=24.60ꢀ, 25.44ꢀ, 30.50ꢀ and 32.53ꢀ and can
160
30
120
80
20
300
400
500
600
Temperature(K)
ZFC (1000Oe)
FC (1000Oe)
ZFC (100Oe)
10
0
be indexed on a magnetic unit cell with amag ¼ a; bmag
¼
FC (100Oe)
b; cmag ¼ 2c; where a; b; c are the axes of the chemical
unit cell, i.e., with ordering wave vector k ¼ ð0; 0; 1=2Þ:
This is the same result as for La5Re3MnO16 which
exhibited an A-type magnetic structure consisting of
ferrimagnetically coupled perovskite layers coupled
antiparallel along the c-axis. Using the same model,
the refinement of the magnetic structure of La5Re3
FeO16 resulted in the agreement factors with Rp ¼ 6:22;
Rwp ¼ 8:45; Rmag ¼ 33:7 Rexp ¼ 3:33 and the structure is
shown in Fig. 10 where the Fe2+ moments of 3.0(2) mB
are tilted from the c-axis by an angle of 18(1)o, while
Re5+ moments of 1.5(1) mB are nearly parallel to the
c-axis, indicating that the Re5+ and Fe2+ moments
are ferrimagnetically coupled within the layer and
antiferromagnetically coupled between the layers, i.e.,
an A-type ordering. As shown in Fig. 11, the tempera-
ture dependence of the intensities of the magnetic
reflections indicates Tc ¼ 155 K which is consistent with
the bulk susceptibility data of Fig. 7.
0
100
200
300
400
Temperature(K)
Fig. 7. ZFC and FC susceptibilities for La5Re3FeO16 for applied fields
of 100 and 1000 Oe. The inset shows the inverse susceptibility up to
600 K and the fitted line indicates a Curie–Weiss law above B 500 K.
1.4 and 0.9 mB ,respectively at 4 T. At 250 K, the curve is
linear which is a signature of paramagnetic behavior.
Note that yC; while still negative, is considerably smaller
than that for La5Re3MgO16. This suggests that the
Fe2+–O2À–Re5+ superexchange interaction is ferro-
magnetic. This is in qualitative agreement with the
Goodenough–Kanamori rules for (t42g eg2)–(t22g) super-
exchange [12,13]. Nonetheless, the bulk magnetic data
are not consistent with an overall ferromagnetic

172
L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
300
(-1,±1,1/2)
(1,±1,1/2)
(-1,±1,3/2)
(1,±1,3/2)
8000
4000
160K
200
100
0
150K
140K
130K
90K
50K
13K
0
0
20
40
2θ(°)
0
20
40
60
80
100
120
140
160
180
(
Fig. 9. Neutron diffraction patterns for La5Re3FeO16 (l=2.372 A) at
selected temperatures showing the development of magnetic Bragg
peaks (arrows).
Temperature(K)
Fig. 11. The temperature dependence of the major magnetic Bragg
peaks for La5Re3FeO16 indicating Tc=155 K.
50
ZFC
FC
40
200
30
160
20
120
10
0
80
300
400
500
600
Temperature(K)
0
100
200
300
400
Temperature(K)
Fig. 12. ZFC-FC susceptibilities for La5Re3CoO16 at an applied field
of 1000 Oe. Note the sharpmaximum at 33 K. The inset shows inverse
susceptibilities up to 600 K and the fitted line shows Curie–Weiss
behavior above 300 K.
600 K and fitting yields yC ¼ À71:2ð5Þ K, and
C ¼ 3:578ð3Þ emu/mol K, which is slightly larger than
that expected for Co2+(d7), C ¼ 1:87 emu/mol K and
Re5+(d2), C ¼ 1 emu/mol K. This might be due to a
partial orbital contribution to the magnetic moment of
Co2+. Such an orbital contribution has been used to
explain the unusual magnetic moment value in CoWO4
[14]. The magnetization is plotted as a function of
applied field at 5, 23, 33, 60 and 100 K in Fig. 13. At 5
and 23 K, both magnetization curves have a sharp
increase at 5000 Oe and approach a saturation magnetic
moment of about 1.4 mB by 4 T. At 33 K, the magnetiza-
tion curve shows an approach to saturation behavior. At
60 and 100 K, the magnetization curves indicate para-
magnetic behavior.
Fig. 10. The magnetic structure of La5Re3FeO16 refined from neutron
diffraction data at 13 K.
3.4. La5Re3CoO16
Fig. 12 presents the ZFC–FC magnetic susceptibilities
of La5Re3CoO16 as a function of temperature in an
applied field of 1000 Oe. There is a sharp cusp at 33 K
which can be taken as the ordering temperature, Tc:
Curie–Weiss-like behavior sets in between 300 and

L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
173
3.5. La5Re3NiO16
C ¼ 2:29ð1Þ emu/mol K, comparable to the sum of
Ni2+(d8) Re5+(d2)
and
C ¼ 1:00 emu/mol K
The ZFC–FC magnetic susceptibilities are plotted as
a function of temperature in an applied field of 1000 Oe
in Fig. 14. The ZFC curve shows two maxima at 36 and
14 K, respectively, while for the FC curve only the sharp
cusp at 14 K is apparent but a ZFC–FC divergence
occurs at 36 K. The Curie–Weiss law (inset) obtains
between 450 and 600 K with yC ¼ À217ð3Þ K and
C ¼ 1:00 emu/mol K. Fig. 15 exhibits the magnetization
as a function of applied field at 5, 12, 34, and 70 K. At
5 K, the magnetization does not approach saturation
even at the highest magnetic field of 5.5 T. At 12 K, the
magnetization deviates only slightly from linearity in the
measured range. At 34 and 70 K, the linear behavior of
the magnetization indicates paramagnetic behavior.
2.0
1.2
5K
23K
33K
5K
12K
1.0
1.5
60K
100K
34K
70K
0.8
1.0
0.5
0.0
0.6
0.4
0.2
0.0
0
10000
20000
30000
40000
50000
60000
0
10000
20000
30000
40000
50000
Applied field(Oe)
Applied field(Oe)
Fig. 15. Isothermal magnetization versus applied field at selected
temperatures for La5Re3NiO16. Note the appearance of relatively
sharp jumps at applied fields 420 000 Oe as the temperature is
decreased.
Fig. 13. Isothermal magnetization versus applied field for La5Re3
CoO16 at several temperatures. Note the appearance of sharp jumps in
the magnetization for low applied fields as the temperature decreases.
12
ZFC
FC
360
8
320
280
240
200
4
300
400
500
600
Temperature(K)
0
0
100
200
Temperature(K)
300
400
Fig. 14. ZFC-FC susceptibilities for La5Re3NiO16 at an applied field of 1000 Oe. Note the sharp maxima in the ZFC data at 36 and 14 K. The inset
shows inverse susceptibilities up to 600 K. The fitted line indicates Curie–Weiss behavior above 450 K.

174
L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
Table 7
Collected magnetic data for La5Re3MO16
.
a
CSO (emu K/mol)
M
yC (K)
Tc (K)
Tc/S2 (K)
Cobs (emu K/mol)
Mn
Fe
À48(5)
161
155
33
26
39
15
36
—
4.43(4)
3.60(2)
3.578(3)
2.29(1)
1.00(1)
5.38
4.00
2.87
2.00
1.00
À84(4)
Co
Ni
À71.2(5)
À217(3)
À575(12)
36,14
—
Mg
a The Curie constant based on the spin-only model.
4. Summary and conclusions
considerably less negative. This suggests that the
Re5+–M2+ magnetic interaction is overall ferromag-
netic. This is in fact consistent with the Goodenough–
Kanamori rules. Re5+ has a t22g configuration. All of the
other ions have 1/2 filled or more than 1/2 filled t2g levels
and partially filled eg levels. The eng-t22geg0 transfers from
M-Re would be ferromagnetic and the overall effect
should be at least weakly ferromagnetic. Nonetheless,
the intralayer Re–M coupling is unambiguously anti-
parallel (ferrimagnetic). Ultimately, perhaps, the most
remarkable material of all is the M=Mg2+ member.
In spite of very large antiferromagnetic exchange
already noted, this material does not appear to order
down to 2 K, while deviations from the Curie–Weiss
law set in near 300 K. As mentioned, the intraplanar
spin–spin correlations are subject to frustration and
the interplanar Re–Re superexchange pathway is
extremely convoluted. This compound is under further
investigation.
Finally, it is worth making some comparisons with
the newly reported phases, La5Mo3+xMxO16, wherein
the M site is partially substituted by a divalent ion. [5] In
this system substitution levels of xB0:75 have been
achieved using both molten salt electrolysis and solid
state reaction methods. For some cases, Tc can be
estimated from DC and AC susceptibility data. The
observed dependence of Tc on M is less dramatic
with Tc remaining near 100 K for x ¼ 0:75 and M=Mn,
Fe and Co. These values are reduced significantly
from the parent M=Mo compound which shows
Tc ¼ 200 K.
The isostructural series, La5Re3MO16 shows proper-
ties consistent with local moment magnetism at the
Re(2) site and the M site in the perovskite-like layers.
Curie constants or effective moments obtained from
Curie–Weiss fits at high temperatures are consistent with
nearly spin-only values for Re5+ and M2+. All except
M=Mg2+ show evidence for long-range magnetic order
at the temperatures given in Table 7. These temperatures
are surprisingly high given that the interlayer separation
(
is B10 A for all cases. There is also a strong dependence
of Tc on the identity of M2+ with high values for Mn
and Fe and quite low values for Co and Ni. For layered
magnets a simple mean field argument suggests that
TcBx2intraJinterS2=k [15], where xintra is the intraplanar
correlation length, Jinter is the interplanar exchange
coupling and S is the spin. The S2 dependence can be
removed by dividing into Tc and the result, Table 7
,indicates no additional systematics with respect to M:
One notes that Tc=S2 for Co is anomalously small.
The magnetic structures for M=Mn2+ and Fe2+
have been solved and both show essentially the same
result with ferrimagnetically coupled (but slightly
canted) Re and M moments within the perovskite layers
which are then coupled antiferromagnetically along the
layer stacking direction. Although this result is for zero
field only, it provides a means to understand the
metamagnetic-like transitions observed for all of the
phases in the isothermal M vs H data. That is, the
relatively weak interplanar exchange coupling is over-
come by the application of modest fields of the order of
0.5 T in most cases, to give parallel interlayer coupling.
The intralayer configuration remains ferrimagnetic and
this is consistent with the values of the ‘‘saturation’’
moments for all materials which fall far below the limits
for ferromagnetic coupling which are 7, 6, 5, and 4 mB
for M=Mn, Fe, Co, and Ni, respectively.
Acknowledgment
We thank the Natural Science and Engineering
Research Council of Canada for a Research Grant to
J.E.G.
There does exist a minor paradox with respect to the
observation of ferrimagnetic intraplanar coupling and
the observed yC values. It is clear that the Re5+–Re5+
intraplanar exchange is very large and antiferromagnetic
from the observed yC ¼ À575 K for M=Mg2+, while
corresponding temperatures for the other M are
References
[1] M. Imada, A. Fujimori, Y. Tokura, Rev. Mod. Phys. 70 (1998)
1039.

L. Chi et al. / Journal of Solid State Chemistry 170 (2003) 165–175
175
[2] K.-I. Kobayashi, T. Kimura, H. Sawada, K. Terakura, Y.
Tokura, Nature (London) 395 (1998) 677.
[9] T. Roisnel, J. Rodriguez-Carvajal, Mater. Sci. Forum 378–381
(2001) 118.
[3] M. Ledesert, Ph. Labbe, W.H. McCarroll, H. Leligny, B. Raveau,
J. Solid State Chem. 105 (1993) 143.
[10] K.A. Wilhelmi, E. Lagervall, O. Muller, Acta Chem. Scand. 24
(1970) 3406.
[4] C.R. Wiebe, A. Gourrier, T. Langet, J.F. Britten, J.E. Greedan,
J. Solid State Chem. 151 (2000) 31.
[11] M.E. Lines, J. Phys. Chem. Solids 1 (1970) 101.
[12] J.B. Goodenough, Magnetism and the Chemical Bond, Wiley,
New York, 1963.
[5] K.V. Ramanujchary, S.E. Lofland, W.H. McCarroll,
T.J. Emge, M. Greenblatt, M. Croft, J. Solid State Chem. 164
(2002) 60.
[13] J. Kanamori, Phys. Chem. Solids 10 (1959) 87.
[14] J.B. Forsyth, C. Wilkinson, J. Phys.: Condens. Matter 6 (1994)
3073.
[6] K.V. Ramanujchary, M. Greenblatt, W.H. McCarroll, J.B.
Goodenough, Mater. Res. Bull. 28 (1993) 1256.
[7] A.E.C. Green, C.R. Wiebe, J.E. Greedan, Solid State Sci. 4
(2002) 305.
[15] L.J. de Jongh, in: L.J. deJongh (Ed.), Magnetic Properties of
Layered Transition Metal Compounds, Kluwer Academic Pub-
lishers, Dordrecht, 1990, p. 19.
[8] E.R. Jette, F. Foote, J. Chem. Phys. 1 (1933) 29.
[16] R.D. Shannon, Acta Crystallogr. A 32 (1976) 751.