ARTICLE IN PRESS
A.M. Abakumov et al. / Journal of Solid State Chemistry 177 (2004) 2855–2861
2856
diffractometer (CuKa-radiation, reflection geometry,
proportional counter). X-ray powder diffraction data
for crystal structure determination were collected on a
STADI-Pdiffractometer (Cu Ka1-radiation, curved Ge
monochromator, transmission mode, linear PSD). CSD
program package was used for ab initio structure
determination and the Rietveld refinement from powder
X-ray diffraction data [3]. The profile reliability factors
were calculated after background intensity subtraction.
Samples for electron microscopy were made by
grinding the powder sample in ethanol and depositing
it on a holey carbon grid. Electron diffraction (ED)
patterns and EDX spectra were obtained using a Philips
CM20 electron microscope. EDX analysis was per-
formed with Pb(Ma), Sb(La) and Mn(Ka) lines, using
Pb2MnO4 and Sb2MnO4 as standards. High-resolution
electron microscopy (HREM) observations were made
on a JEOL 4000EX instrument, having a point-to-point
resolution of 1.7 A. Image simulations were carried out
with the MacTempas software.
of Pb5Sb2MnO11 was indexed on an orthorhombic
lattice with cell parameters
a ¼ 9:0660ð8ÞA,
b=11.489(1)A, c ¼ 10:9426ð9ÞA. The observed h k l :
h þ k ¼ 2n and h 0 l : h; l ¼ 2n reflection conditions
suggest Cmcm as the highest space group for this
compound. Syntheses starting from combinations
of other oxides of constituent metals (e.g.,
PbO2þ3Pb3O4þ2MnO þ 2Sb2O3 or 3PbO2+7PbO+
Mn2O3+2Sb2O3) that provide the same total stoi-
chiometry also allow to prepare samples with
Pb5Sb2MnO11 as a main constituent, but with a larger
amount of admixture phases (up to 7–9% of the
pyrochlore, PbO and Mn3O4). Deviation from the
stoichiometric composition by decreasing the Pb content
by 10 at% or increasing the oxygen content up to
Pb5Sb2MnO11.2 result in a drastic increase of the
amount of impurity phases and yield the pyrochlore
phase as the main phase in the samples. Increasing the
Mn content (‘‘Pb4.5Sb2Mn2O12’’ sample), however,
allows to suppress the formation of the pyrochlore
admixture, but results in the appearance of 3–4% of
Mn3O4 and an unknown phase. The unit cell parameters
of Pb5Sb2MnO11 remain the same within the range of
two standard deviations for all samples, which indicates
that the compound composition does not depend on the
choice of the initial reagents and synthesis conditions.
The cation composition of the Pb5Sb2MnO11 phase was
confirmed by EDX analysis in the transmission
electron microscope performed at 12 points on four
different crystallites. The cation ratio Pb:Sb:Mn=
5.03(7):1.98(7):0.99(3) was found to be very close to
the nominal one.
Magnetic susceptibility measurements were per-
formed with a Quantum Design SQUID magnetometer
(MPMS-XL) in a 1000 Oe applied field in the tempera-
ture range of 5–385 K.
Electronic structure calculation and electron localiza-
tion function (ELF) analysis were carried out with the
TB-LMTO-ASA program [4]. The Barth–Hedin LDA
functional [5] with additional Langreth–Mehl–Hu non-
local part was used for exchange-correlation treating in
the SCF calculation. The k-space integration was
performed by the standard tetrahedron method using
14 k-points within the irreducible Brillouin zone. The
basis set consisted of 6s; 6p-orbitals for Pb, 5s; 5p-
orbitals for Sb, 4s; 4p; 3d-orbitals for Mn and 2p-
orbitals for O. 6d; 5f -orbitals for Pb, 5d; 4f -orbitals for
Sb, 3s; 3d-orbitals for O were treated using a down-
folding technique. The spin-polarized case was consid-
ered; all Mn sites were treated as equivalent. The ELF
distribution was calculated using the intrinsic procedure
of the TB-LMTO-ASA program complex.
3.2. Crystal structure and ELF distribution analysis
The crystal structure of Pb5Sb2MnO11 was solved ab
initio from X-ray powder diffraction data. The integral
intensities of the reflections over the angular range 10–
80ꢀ 2y were determined by profile fitting and then used
for the direct structure solution. The Cmcm space group
was assumed for the initial structure model. The analysis
of the resulting electron density map revealed the
positions of the Pb atoms. The atomic coordinates of
the Sb, Mn and oxygen atoms were found by a sequence
of difference Fourier syntheses. Prior to the Rietveld
refinement of the obtained structure model the impacts
of six impurity peaks of Mn3O4 and the unknown phase
were subtracted from the raw data using their fitted
profiles since they did not overlap with the peaks from
the main phase. The refinement was carried out in an
isotropic approximation for the atomic displacement
parameters. Preferred orientation along the [110] direc-
tion was taken into account using the March–Dollase
function. After sequential iterations, a good agree-
ment between the experimental and calculated profiles
was achieved: RI ¼ 0:045; RP ¼ 0:059; RwP ¼ 0:052: An
3. Results and discussion
3.1. Sample preparation
The target compound was prepared starting from the
4PbO2+PbO+MnO2+Sb2O3 mixture. After annealing
at 650ꢀC for 50 h, the Pb5Sb2MnO11 phase with a large
amount of the Pb2Sb2O7 pyrochlore, PbO and Mn3O4
was obtained. The subsequent annealing for 50 h after
regrinding the sample significantly improved the sample
quality leaving only a small amount of the pyrochlore
admixture (the intensity of the strongest peak is
Irel ¼ 4:0%) and traces of metallic Pb. Pb5Sb2MnO11
has a yellow-greenish color. The X-ray powder pattern