78
A.K. Azad et al. / Journal of Alloys and Compounds 364 (2004) 77–82
terial of the type BaMn0.5Mo0.5O3 showing a hexagonal
structure with a = 11.238 and c = 15.758 Å, and a diffuse
ferroelectric behavior [13]. To our knowledge no struc-
tural characterization has previously been done for ordered
Ba2MnMoO6.
In the present study we have described the synthesis of
these materials, prepared by solid state reaction procedures,
and the results of high-resolution neutron powder diffrac-
tion (NPD) studies on well-crystallized samples. We have
reported complete structural data for these materials and
described the magnetic properties from the magnetization
measurements.
2.3. Magnetisation measurement
Magnetization measurements were carried out using a
Quantum Design Superconducting Quantum Interference
Device (SQUID) magnetometer. Magnetization versus tem-
perature curves were measured between 0 to 300 K in
field-cooled (FC) and zero-field-cooled (ZFC) modes with
applied fields (0H) of 20 Oe.
3. Results and discussions
3.1. Crystal structure
2. Experimental details
The crystal structure determination was performed on
the neutron powder diffraction data collected at room
temperature with a wavelength λ = 1.47 Å. In case of
Ba2MnMoO6, the structure was cubic with the unit cell
parameter a = 8.1680(1) Å at 295 K. The space group
was Fm3m and the Rietveld refinement resulted in good
R-factors (Rp = 3.31%, Rwp = 4.24%, RBragg = 2.71% and
χ2=1.85). The x coordinate of oxygen was varied during
the refinement and determined to be 0.2643(1). The possible
presence of oxygen non-stoichiometry was not confirmed
during the final Rietveld refinements. The cation occupan-
cies were also found to be very close to their nominal values.
From the Rietveld analysis we obtain an elemental ratio
Ba:Mn:Mo:O = 0.332(3):0.167(1):0.165(1):1.000, which is
in excellent agreement with the nominal composition of the
sample, Ba2MnMoO6. The Mn and Mo ions were ordered
at the B-sites and form together with oxygen a NaCl-type
lattice where both the cations show perfect octahedral an-
ion coordination. The MnO6 octahedra are larger than the
MoO6 octahedra, an observation in accordance with the
larger ionic size of Mn2+ (rMn2+ = 0.83 Å) compared to
Mo6+ (rMo6+ = 0.59 Å). The volumes of the octahedra
are calculated to be 13.41(1) Å3 for MnO6 and 9.51(1) Å3
for MoO6. Each barium ion is coordinated to 12 oxygen
ions, being part of the ccp layers. The average Ba–O bond
lengths at 295 K compare well with the expected values cal-
culated as the sum of the ionic radii. The average observed
Mn–O and Mo–O bond distances are 2.158(1) and 1.926(1)
Å, respectively. Bond valence sum (BVS) [17] calculation
from the average observed bond lengths (d ≤ 3.5 Å) be-
tween Mn or Mo and O shows that the charges of Mn and
Mo cations are +2.214 and +5.706, respectively. The tol-
erance factor was calculated to be 1.05 using the Shannon
ionic radii [18]. If we calculate the parameters considering
completely ordered and pure oxidation state for Mn2+ and
Mo6+ using SpuDS software [19], we get the lattice param-
eter a = 8.207 Å, bond lengths between Mn–O = 2.1965
Å and Mo–O = 1.907 Å, and a tolerance factor, t = 1.016,
which are in good agreement with the experimental values.
BVS calculation using Mn3+ and Mo5+ gives the value
2.046 and 5.718 Å, respectively, in which very different
from the expected value. The small disagreements may be
2.1. Sample preparation
The Ba2MnMoO6 and Sr2MnMoO6 compounds were pre-
pared as polycrystalline powders by the solid state reac-
tion technique. Stoichiometric amounts of analytical grade
BaCO3 or SrCO3, MnO and MoO3 were mixed together by
ethanol. The finely mixed powders were pre-sintered in a
furnace for 15 h at 950 ◦C. The powders were mixed in an
agate mortar again and pressed into small discs and heated
up to 1200 ◦C for 48 h. The sample was again reground and
heated at 1350 ◦C for 48 h and finally at 1400 ◦C for 48 h. All
heat treatments were conducted under nitrogen environment.
2.2. X-ray diffraction, neutron diffraction and
Rietveld analysis
X-ray powder diffraction measurements of the samples
were carried out to check the quality and phase distribu-
tion of the samples using a Guinier Camera (Cu K␣1 radi-
ation) at the KEMLAB of Studsvik Neutron Research Lab-
oratory, Sweden. Silicon (NBS 640b) was used as an inter-
nal standard and a computerized line scanner for evaluation
of the film. Indexing and refinement of the lattice param-
eters were made using TREOR90 [14] and Checkcell [15]
software, respectively. Neutron diffraction data were col-
lected at the NPD instrument at the R2 reactor, Sweden. The
double monochromator system consists of two copper crys-
tals aligned in (220) mode giving a wavelength of 1.470 Å.
The neutron flux at the sample position was ∼106 neutron
cm−2 −1. The step scan covered the 2θ range 4◦–139.92 ◦
s
with a step size 0.08◦. NPD data sets were refined by the
Rietveld method using FullProf software [16] including the
coherent scattering length supplied by the software. Diffrac-
tion peak shapes were quantified by a pseudo-Voigt func-
tion, with a peak asymmetry correction applied at angles
below 45◦ in 2θ. Background intensities were described by
a Chebyshev polynomial with six coefficients. Each struc-
tural model was refined to convergence, with the best result
selected on the basis of agreement factors, chemical sense
and stability of the refinement.