Ln2Sr2PtO7+δ (Ln=La, Pr, and Nd): Three new Pt-containing [A2′ O1 + δ] [AnBn-1O3n]-type hexagonal perovskites

Article abstract of DOI:10.1016/j.jssc.2007.10.005

Polycrystalline samples of Ln₂Sr₂PtO_7+δ (Ln=La, Pr, Nd) were prepared by conventional solid state synthesis. The three compounds are new examples for n=2 members of the [A₂^′ O_{1 + δ}][A_nB_n-1O_3n] family of hexagonal perovskites containing platinum as the B-type cation. XRD Rietveld refinements show the platinates to crystallize in space group R over(3, -) and, in the case of Pr and Nd, revealed a complete ordering of Ln/Sr on the two distinct A-type positions, while for La a partial disorder was observed. By XANES investigations at the Pt-L_III threshold the oxidation state +4 for platinum was found. Thermogravimetry revealed a small oxygen excess for Ln=La and Pr (δ=0.13 and 0.07), pointing to the presence of peroxide ions as already observed for isostructural Ru- and Ir-based compounds. UV-Vis measurements were done for the yellow lanthanum and the green neodymium compound. They revealed two optical band gaps of 2.52 and 3.05 eV, respectively. Magnetic measurements showed La₂Sr₂PtO_7+δ to be diamagnetic as expected for Pt⁴⁺ with low-spin (t_{2 g}⁶) configuration. For Ln=Pr and Nd the observed strong paramagnetism can be explained solely by the magnetic moments of the rare earths.

Full text of DOI:10.1016/j.jssc.2007.10.005

ARTICLE IN PRESS
Journal of Solid State Chemistry 180 (2007) 3393–3400
www.elsevier.com/locate/jssc
Ln2Sr2PtO7+d (Ln ¼ La, Pr, and Nd): Three new Pt-containing
½A02O1þdꢀ [AnBnꢁ1O3n]-type hexagonal perovskites
Ã
Stefan G. Ebbinghausa, , Chasanoglou Erztoumenta, Ivan Marozaub
aLehrstuhl fu¨r Festko¨rperchemie, Institut fu¨r Physik, Universita¨t Augsburg, Universita¨tsstraXe 1, D-86159 Augsburg, Germany
bPaul Scherrer Institut, CH-5232 Villigen PSI, Switzerland
Received 20 July 2007; received in revised form 1 October 2007; accepted 1 October 2007
Available online 7 October 2007
Abstract
Polycrystalline samples of Ln2Sr2PtO7+d (Ln ¼ La, Pr, Nd) were prepared by conventional solid state synthesis. The three compounds
are new examples for n ¼ 2 members of the ½A02O1þdꢀ[AnBnꢁ1O3n] family of hexagonal perovskites containing platinum as the B-type
¯
cation. XRD Rietveld refinements show the platinates to crystallize in space group R 3 and, in the case of Pr and Nd, revealed a complete
ordering of Ln/Sr on the two distinct A-type positions, while for La a partial disorder was observed. By XANES investigations at the Pt-
LIII threshold the oxidation state +4 for platinum was found. Thermogravimetry revealed a small oxygen excess for Ln ¼ La and Pr
(d ¼ 0.13 and 0.07), pointing to the presence of peroxide ions as already observed for isostructural Ru- and Ir-based compounds. UV–Vis
measurements were done for the yellow lanthanum and the green neodymium compound. They revealed two optical band gaps of 2.52
and 3.05 eV, respectively. Magnetic measurements showed La2Sr2PtO7+d to be diamagnetic as expected for Pt4+ with low-spin
ðt62gÞ configuration. For Ln ¼ Pr and Nd the observed strong paramagnetism can be explained solely by the magnetic moments of
the rare earths.
r 2007 Elsevier Inc. All rights reserved.
Keywords: Hexagonal perovskites; Platinates; XRD Rietveld refinements; Thermogravimetry; Magnetic properties; Optical spectroscopy
1. Introduction
hexagonal cavities. These cavities can be occupied by oxide
ions (O2ꢁ), or peroxide ions (O22ꢁ). The few examples of
this family of hexagonal perovskites reported so far contain
the transition metals Mn, Nb, Ru, and Ir and belong to
n ¼ 6, 3, and 2, respectively. The only n ¼ 6 oxide known
to date is La4Ba2.6Ca1.4(Mn4Ca)O19 [1]. This oxide also
includes structural features typical for the family of cubic
perovskites, i.e. corner-sharing octahedra. Most com-
pounds described in literature belong to the n ¼ 3
structure: La2Ba0.8Sr0.6Ca1.6Mn2O10 [2], Ba5Nb2O9(O2)
[3], Ba5Ru2O10 [4], Ba5Ru2O9(O2) [5], and Ba3.44K1.56Ir2O10
[6]. The composition A4BO7+d (½A02O1þdꢀ[A2BO6], i.e.
n ¼ 2) represents the end-member of this family with
isolated BO6-units, while the 2H-structure can be consid-
ered the n ¼ N end-member. Examples for the n ¼ 2
structure are Ln2Ca2MnO7 (Ln ¼ La, Nd, Sm) [7,8] and the
corresponding peroxide La2Ca2MnO6(O2) [9]. Our group
reported on ruthenium containing samples, which were
found to be rather flexible with respect to their cationic
composition and can simultaneously contain oxide and
The family of hexagonal perovskites can be described by
hexagonal close packing of AO3-layers (stacking
a
sequence abababy) in which the B-cations occupy 1/4 of
the octahedral sites. This arrangement results in chains of
face-sharing BO6-octahedra. In the aristotype (the so-
called 2H structure) these chains are in principle infinite.
Many modified structures exist, in which the chains are
interrupted by, for example, trigonal prisms. These
structures can often be grouped in homologous series.
One specific family of hexagonal perovskites possesses the
general composition ½A02O1þdꢀ[AnBnꢁ1O3n], where nꢁ1
reflects the number of face-sharing BO6-octahedra. In
these compounds, the BO6-chains are interrupted by
½A02O1þdꢀ-layers in which six trigonal prisms form large
Ã
Corresponding author. Fax: +49 821 598 3002.
E-mail address: stefan.ebbinghaus@physik.uni-augsburg.de
(S.G. Ebbinghaus).
0022-4596/$ - see front matter r 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jssc.2007.10.005

ARTICLE IN PRESS
S.G. Ebbinghaus et al. / Journal of Solid State Chemistry 180 (2007) 3393–3400
3394
angular range 151p2yp1501. The step size was 0.0151 2y
with a counting time of 10 s per data point. For Rietveld
refinement calculations the program FullProf was used
[17]. The background was approximated by a linear
interpolation between approximately 30 data points in
regions with no Bragg reflections. The peak shape was
described by a pseudo-Voigt function.
Thermogravimetric measurements were performed on a
TA Q500 thermobalance. About 40 mg of the oxides were
heated in flowing forming gas (10% H2 in N2; 50 ml/min)
to 950 1C with a heating rate of 10 1C/min.
Optical properties of the polycrystalline powders were
investigated by diffuse reflectance spectroscopy. Reflectance
data were collected using a Varian Cary 500 Scan
UV–Vis–NIR scanning double-beam spectrometer equipped
with a 150 mm Labsphere DRA-CA-50 integrating sphere
over the spectral range of 250–2000nm (0.6–5 eV).
X-ray absorption spectroscopy at the Pt-LIII edge was
carried out at the beamline C1 (Cemo) of the Hamburger
Synchrotronstrahlungslabor (HASYLAB) at the Deutsches
Elektronensynchrotron (DESY). Approximately 35 mg of
each sample was mixed with polyethylene and pressed to a
thin pellet of 13 mm diameter. XANES measurements were
performed in transmission mode with a step width of 0.1 eV
and a counting time of 2 s per data point. The energy was
calibrated against the spectra of Pt metal, which was
measured simultaneously.
Fig. 1. Crystal structure of the [A02O1+d][A2BO6]-type hexagonal
perovskites. A0 cations in the trigonal prisms and A cations in the
A2BO6-layers are shown in light blue and dark blue, respectively. Oxygen
ions in the [A02O1+d]-layers are colored in magenta.
Magnetic measurements in the temperature range
1.8oTo300 K were done using a Quantum Design
MPMS7 SQUID magnetometer. Samples were field-cooled
(fc) in external fields of 0.1 T (Pr, Nd) and 1 T (La),
respectively. The obtained values were corrected for the
diamagnetic moment of the empty sample holder.
peroxide ions [10–12]. The composition of these samples
may be generalized as (Ln,Sr)4ꢁzRuO7ꢁd(O2)d. The only
n=2 compounds containing a 5d-transition metal are
La2.5K1.5IrO7 [13], La1.2Sr2.7IrO6.67(O2)0.33 [12] and the
very recently discovered platinates Ln3NaPtO7 (Ln ¼ La,
Nd) [14]. Some samples were found to exhibit interesting
physical properties like high dielectric constants [15] or
geometric magnetic frustration [16]. A representation of the
½A02O1þdꢀ[A2BO6] structure is shown in Fig. 1.
3. Results and discussion
In analogy to the previously described Ru and Ir
containing samples [11,12] we first tried to prepare
La1.2Sr2.7PtO7+d. The corresponding XRD pattern clearly
indicated the presence of the desired phase but also
revealed a considerable amount of Sr4PtO6. Further
increasing the reaction time and/or temperature did not
lead to significant changes in the relative peak intensities.
We therefore concluded that the desired phase possesses a
higher lanthanum content and in turn the oxidation state
of platinum must be lower than +5. Since in the impurity
phase platinum takes the oxidation state +4 it seemed
reasonable to assume that this is also the case in the desired
material, leading to the composition La2Sr2PtO7. The
attempt to prepare the corresponding compound resulted
in a phase pure sample of yellow color. Choosing the same
cationic compositions black and green samples were
obtained for Ln=Pr and Nd, respectively.
In this paper, we describe the preparation and char-
acterization of new n ¼ 2 oxides containing Pt as B-type
cation. As will be shown, the three samples contain
diamagnetic Pt4+ ions and exhibit clear indications for a
¯
symmetry reduction from space group R 3 m to R 3.
¯
2. Experimental details
Polycrystalline samples of Ln2Sr2PtO7+d (Ln ¼ La, Pr,
Nd) were prepared by conventional solid state synthesis
from SrCO3, pre-dried Ln2O3 and Pt metal powder.
Appropriate amounts of these starting materials leading
to 3 g-batches were weighted with an accuracy better than
0.1 mg and thoroughly ground in an agate mortar under
isopropanol. After drying, the mixtures were placed in
alumina crucibles and heated for 36 h at 1100 1C in air in a
box furnace.
Rietveld refinements on the basis of X-ray diffraction
were carried out starting from the structural model for
¯
La1.2Sr2.7BO7+d (B=Ru, Ir) (space group R 3 m) described
in [12]. For Ln ¼ Nd a trace impurity of Sr4PtO6
X-ray diffraction data were collected on a Seifert XRD
3003-TT diffractometer using CuKa radiation in the

ARTICLE IN PRESS
S.G. Ebbinghaus et al. / Journal of Solid State Chemistry 180 (2007) 3393–3400
3395
(E1 wt%) was included in the calculations. The other two
samples were single phase.
found to be only slightly oblate and no significant
remaining scattering density near the c-axis was detected
in the difference Fourier map. It can therefore be
concluded that the Sr2+ ions only occupy the 6c site and
no disordered arrangement exists.
For all atoms except O2/O2a, refinement trials with
anisotropic displacement parameters were carried out. The
obtained ellipsoids for Ln (0, 0, 0.62) and Pt (0, 0, 0) were
found to be almost spherical for all three samples.
Therefore in the final runs these positions were refined
with isotropic displacement parameters. For O1 the
parameters b13 and b23 were fixed to 0 in order to reduce
the number of variables. This simplification was based on
the fact that neutron diffraction showed b13 and b23 to be
close to zero for the isostructural ruthenate and iridate.
For La1.2Sr2.7BO7.33 (B ¼ Ru, Ir) neutron diffraction
additionally revealed the oxygen atoms in the ½A02O1þdꢀ-
Ã
layer to occupy two positions (x, 1/2 x, 1/2) and (x, 0, 1/2),
with different occupation and distance to the c-axis. Due to
the small atomic form factor of oxygen, X-ray diffraction
cannot resolve such structural details. For the platinates
investigated here, we therefore assumed a toroidal electron
2
Ã
˚
Furthermore, b33 had to be fixed to 0.0012 A (the value
obtained from neutron diffraction data of the Ru and Ir
analogs) to avoid negative values in the case of Ln ¼ Nd.
density formed by equal amounts of oxygens on (x, 1/2 x,
Ã
1/2) and (Cos 30 x, 0, 1/2). For both sizes the Biso was fixed
2
to 1.0 A .
˚
¯
These initial calculations assuming space group R 3 m
Fig. 2 shows the Rietveld plot for La2Sr2PtO7+d with a
magnification of the angular range 901p2yp1501 as inset.
Results of the final structure refinement are listed in Tables
1 and 2. As an example Fig. 3 shows a part of the structure
of Pr2Sr2PtO7+d with displacement parameters given at the
80% confidence level.
resulted in extremely elongated displacement ellipsoids
for O1. For this reason, additional refinements in the less
¯
symmetric space group R 3 (No. 148) were performed. The
missing symmetry element (.m) allows to shift O1 from
(x, x/2, z) to (x, y, z) while keeping the cationic framework
¯
unchanged. Since in R 3 the bij values for O1 have no
The cell parameters of the three title compounds are
depicted in Fig. 4. As expected, both the a and c axes
become shorter with decreasing ionic radius of the
lanthanide ion. A comparison with the cell parameters of
restrictions, in principle six anisotropic displacement
parameters need to be refined, which seems hopeless on
the basis of powder X-ray diffraction. Therefore, the O1
atom was refined isotropically. In spite of the fewer free
˚
˚
La1.2Sr2.7RuO7.33 (a ¼ 5.753(1) A, c ¼ 18.351(3) A) and
˚
˚
¯
variables compared to the (anisotropic) model in R 3 m,
La1.2Sr2.7IrO7.33, (a ¼ 5.771(1) A, c ¼ 18.348(3) A) shows
that the a axes are of similar lengths while the c axes of the
platinates are significantly shorter.
identical or even slightly smaller residual parameters were
achieved. The estimated standard deviations (esds) for the
fractional coordinates of O1 were rather small, showing
that the atomic positions are well defined. In addition, the
¯
The lower symmetry of space group R 3 compared to
¯
R 3 m preserves the local 3 (D3d) point symmetry of the
¯
˚
obtained Biso values are in the order of 0.5–1 A, which is
reasonable for oxygen. For these reasons we conclude that
the structural model in R3 is to be preferred over the one in
platinum ions but strongly changes the coordination
¯
geometry of the lanthanide ions. In R 3 m the Ln cations
within one AO3 layer are surrounded by six equivalent O1
˚
ions at a distance of roughly 2.9 A. In R 3, on the other
¯
the higher symmetric space group R 3 m.
¯
Attempts to refine a mixed occupation of Ln/Sr on the
two A cation sites led to occupation factors close to 1 and
0, respectively, for Ln ¼ Pr and Nd, indicating a complete
cationic ordering. In the case of La, on the other hand, we
found a significant disorder with 8.5% of the Ln3+ position
occupied by Sr2+ and vice versa. The reason for this
structural difference between the three compounds may
result from the different sizes of the rare-earth cations.
Assuming a nine-fold oxygen coordination, the ionic radii
hand, the six oxygens split into two sets of three ions each
˚
with bond lengths of E2.5 and E3.3 A, respectively. For
Ln ¼ La, Pr, and Nd the shorter Ln–O1 distance decreases
6000
Observed
Calculated
75
Difference
Bragg
4000
positions
3+
according to Shannon [18] are La3+: 1.216 A, Pr
:
the radius is
˚
3+
1.179 A, and Nd : 1.163 A, while for Sr
2+
0
˚
˚
3+
˚
1.31 A. It therefore seems likely that only La
is large
2000
enough to allow a (partial) mixing with the strontium ions,
while the other two rare earths are too small. In this
context it is worth mentioning that also in La1.2Sr2.7BO7.33
(B ¼ Ru, Ir) a mixed La/Sr occupation was observed for
the La1 site.
-75
100
120
140
0
Earlier Rietveld refinements of La1.2Sr2.7BO7.33 (B ¼ Ru,
Ir) gave hints for a split model with parts of the Sr2 ions
located on (0, 0, z) and parts on a three-fold position close
to the c-axis. For the platinates studied in this work, on the
other hand, the displacements ellipsoids of the Sr2 site were
20
40
60
80
[°]
100
120
140
2
θ
Fig. 2. XRD Rietveld refinement plot for La2Sr2PtO7+d. The inset shows
a zoom of the angular range 901p2yp1501.

ARTICLE IN PRESS
S.G. Ebbinghaus et al. / Journal of Solid State Chemistry 180 (2007) 3393–3400
3396
Table 1
Cell parameters, atomic coordinates and (equivalent) isotropic displacement parameters for Ln2Sr2PtO7+d
w2 1.51
Biso/eq (A )
˚
a (A) 5.7996(1)
˚
c (A) 18.0750(4)
La
Atom
Rp 8.68
z
Rwp 11.2
SOF
2
˚
Site
x
y
Pt
O1
3a
0
0.3145(15)
0
0.2169(18)
0
1
0.74(3)
0.56(24)
0.99(4)
1.38(6)a
1.0
18f
6c
0.0621(4)
0.6233(1)
0.1731(1)
0.5
1
La1/Sr1
Sr2/La2a
O2
0
0
0.915/0.085(6)
0.915/0.085(6)
0.082(3)
6c
0
0
18f
18f
0.1258(52)
0.1452(62)
0
0.0726(31)
O2a
0.5
0.082(3)
1.0
w2 1.34
Biso/eq (A )
˚
a (A) 5.7523(1)
˚
c (A) 17.8919(3)
Pr
Atom
Rp 9.26
z
Rwp 11.8
SOF
2
˚
Site
x
y
Pt
O1
3a
0
0.3215(19)
0
0.2324(19)
0
1
0.52(3)
0.78(28)
1.20(4)
1.61(7)a
1.0
18f
6c
0.0657(4)
0.6230(1)
0.1730(1)
0.5
1
Pr1
Sr2a
O2
0
0
1
6c
0
0
1
18f
18f
0.1015(55)
0.1173(65)
0
0.0587(32)
0.090(3)
0.090(3)
O2a
0.5
1.0
w2 1.43
Biso/eq (A )
˚
a (A) 5.7427(1)
˚
c (A) 17.8770(4)
Nd
Atom
Rp 9.06
z
Rwp 11.6
SOF
2
˚
Site
x
y
Pt
O1
3a
0
0.3237(17)
0
0.2370(9)
0
1
0.42(3)
0.95(28)
0.98(4)
1.44(6)a
1.0
18f
6c
0.0641(4)
0.6234(1)
0.1729(1)
0.5
1
Nd1
Sr2a
O2
0
0
1
6c
0
0
1
18f
18f
0.1218(55)
0.1407(65)
0
0.0703(33)
0.083(3)
0.083(3)
O2a
0.5
1.0
¯
Space group R 3(No. 148). Parameters given without standard deviations were fixed during the refinements.
aRefined with the anisotropic displacement parameters listed in Table 2.
Table 2
Anisotropic
displacement
parameters
(exp [ꢁ(b11h2+b22k2+b33l2
+2b12hk+2b13hl+2b23kl)]) for the atom Sr2
Ln
b11
b22
b33
b12
b13
b23
La
Pr
0.0164(11)
0.0215(12)
0.0201(11)
0.0164(11)
0.0215(12)
0.0201(11)
0.00065(9)
0.00045(10)
0.00027(9)
0.0082(5)
0.0107(6)
0.0100(6)
0
0
0
0
0
0
Nd
˚
(2.578, 2.497, and 2.469 A) while simultaneously the longer
˚
distance is elongated (3.265, 3.312, and 3.324 A). In turn
the coordination geometry becomes successively more
irregular.
For the La and Nd containing samples the O2/O2a
coordinates lead to oxygen–oxygen distances which are in
reasonable agreement with the value expected for peroxide
ions, while for the Pr compound the interatomic distance
is significantly smaller. The site occupation factors for
O2/O2a lead to total oxygen contents of 6.98(4), 7.08(4),
and 7.00(4) for Ln ¼ La, Pr, Nd. Within a 2s range the
Rietveld refinements therefore provide no clear proof for
the presence of peroxide ions.
As a second, independent method for determining the
oxygen stoichiometry thermogravimetry (TG) was used.
During heating in reducing atmosphere (e.g. mixtures of H2
with N2 or Ar) most perovskites are reduced to the
Fig. 3. Section of the crystal structure of Pr2Sr2PtO7+d as derived from
XRD Rietveld refinement. The displacement parameters are shown with
80% probability.
corresponding metals and/or simple binary oxides, which
can be identified by X-ray diffraction. From the observed
weight loss and the known reaction products the oxygen
content can easily be calculated. Fig. 5 gives the TG results
for the three title compounds. As reduction products,

ARTICLE IN PRESS
S.G. Ebbinghaus et al. / Journal of Solid State Chemistry 180 (2007) 3393–3400
3397
18.10
18.05
18.00
17.95
17.90
100
99
98
97
96
95
La
Pr
Nd
17.85
5.80
La
5.78
5.76
5.74
5.72
100 200
300 400
500 600
700
800
900
Temperature [°C]
Fig. 5. Thermogravimetric investigation of the reduction of
Ln2Sr2PtO7+d. The horizontal line corresponds to the expected weight
loss for Ln ¼ Pr assuming d ¼ 0.
Pr
Nd
1.22
1.20
1.18
1.16
former method yielded an oxygen contents of 7.13, the
latter resulted in a d value close to zero. It should not be
forgotten though that due to the small X-ray scattering
power of oxygen, the site occupation factors for O2/O2a
show rather high esd values. Furthermore, the O2/O2a
occupation factor strongly correlates with the displacement
Ln radius [Å]
Fig. 4. Evolution of the cell parameters with changing ionic radius of the
lanthanide ions. Estimated standard deviations are smaller than the size of
the symbols.
2
˚
Ln2O3 and Pt metal were identified directly after the
reduction, while SrO appeared as SrCO3 after keeping the
reaction products in air for some days. The straight line in
the figure corresponds to the expected weight loss for d ¼ 0
in the case of Ln ¼ Pr. For the other two rare earths the
calculated values are very similar and therefore not shown
in the figure for clarity. From Fig. 5 it is obvious that for
Ln ¼ La and Pr the weight loss is larger than expected. The
observed Dm/m values of 4.46%, 4.33%, and 4.19% for La,
Pr and Nd, lead to the following compositions: La
d ¼ 0.13, Pr d ¼ 0.07, and Nd d ¼ 0.02. The uncertainties
of these values are estimated to be 0.03. Following the
discussion in [11] an oxygen excess (d40) can be combined
with the observed oxidation state of +4 of platinum (see
below) by a mixed oxygen/peroxide stoichiometry accord-
ing to Ln2Sr2PtO7ꢁd(O2)d. This means that the large
hexagonal cavities in the ½A02O1þdꢀ-layers are occupied by
both oxide (O2ꢁ) and peroxide (O22ꢁ) ions. It is noteworthy
that the d values ( ¼ peroxide contents) of the platinates
studied in this work are significantly lower than those for
La1.2Sr2.7BO7+d (B ¼ Ru, Ir). For these two compounds a
d-value of 0.33 was found. In addition, in the present
samples d decreases from 0.13 to 0 going from La to Nd.
The reason for this may be the shrinking of the a parameter
(see Table 1), which reduces the available size for the O22ꢁ
ions in the ½A02O1þdꢀ-layers.
parameter, which was somewhat arbitrarily fixed to 1 A .
Because of these reasons, the oxygen contents determined
by thermogravimetry seem more reliable. Neutron diffrac-
tion experiments are planned for the future to verify the
presence of peroxide ions in the samples.
The oxidation state of platinum was determined by
X-ray absorption spectroscopy, making use of the so-called
‘valence shift’. This shifting of the absorption edge to
higher energies with increasing oxidation number results
from a reduced shielding of the positive charge of the
nucleus by the smaller number of valence electrons. In turn,
the inner electrons are more tightly bound and higher
energies are required for ionization. The usefulness
of XANES investigations for the determination of oxida-
tion states has been described in a series of publications
for other platinum group metals like Ru and Ir
[10,11,16,19–23].
Fig. 6 shows the normalized absorption spectrum of
Pr2Sr2PtO7+d in comparison to PtO2, which served as
Pt4+-reference. As can be seen, the energetic positions of
the so-called ‘white lines’ of these two compounds are very
similar, already pointing to an oxidation state of +4 for
platinum. White lines result from dipole-allowed transi-
tions to empty orbitals with predominantly d character.
Due to the crystal field splitting in an octahedral
coordination, the white lines split into two sub-peaks
corresponding to the t2g and eg orbitals, respectively, if the
t2g orbitals are only partly filled with electrons.
For Ln ¼ Pr and Nd the d values obtained by thermo-
gravimetry and XRD Rietveld refinements agree very well.
In the case of Ln ¼ La, on the other hand, there is a
discrepancy between the TG and XRD results. While the
For a quantitative analysis, the XANES spectra were
fitted according to the procedure described in [22]. For

ARTICLE IN PRESS
S.G. Ebbinghaus et al. / Journal of Solid State Chemistry 180 (2007) 3393–3400
3398
Pt4+ with its ðt2gÞ configuration only one peak at the
Pt-LIII threshold is expected, corresponding to the transi-
tion of the 2p electron to the empty eg orbitals. As can be
seen in Fig. 6 it was in fact possible to fit the white lines of
our samples (and also of the PtO2 reference) with only one
pseudo-Voigt function for this transition and an arctangent
function describing the absorption edge itself. For the
energy positions of the pseudo-Voigt function the follow-
ing values were obtained: La2Sr2PtO7+d 11.56794 keV,
Pr2Sr2PtO7+d 11.56798 keV, and Nd2Sr2PtO7+d 11.56795 keV.
These values agree well with the one found for PtO2
(11.56774 keV). The deviation of roughly 0.2 eV lies well
within the estimated uncertainty of the method at these
energies [12]. It can therefore be concluded that platinum
possesses the oxidation state +4 in Ln2Sr2PtO7+d, as
expected.
Fig. 7, these two compounds exhibit almost identical
strong paramagnetic signals. This similarity can be
explained using the free ion approximation for the
lanthanide ions. For Pr3+ with its 4f 2 electronic config-
uration the ground term is 3H4. Calculating the corre-
6
sponding Lande
´
factor yields g ¼ 0.8, and the resulting
pffiffiffiffiffi
magnetic moment becomes m=mB ¼ 0:8 20 ¼ 3:578. The
same considerations for Nd3+ (4f 3 configuration) results
4
in a I9/2 ground term with g ¼ 0.7272 and m=mB ¼
pffiffiffiffiffiffiffiffiffiffi
0:7272 99=4 ¼ 3:618. The magnetic moments of these
two ions are therefore expected to differ by only 1%
despite of their completely different electronic configura-
tion and ground terms.
The inset in Fig. 7 shows the experimental 1/w vs. T
values for Ln ¼ Pr and Nd. The data were fitted in the
temperature range 50–300 K by a modified Curie–Weiss
law w ¼ ðN2Lm0m2=3RðT ꢁ T0ÞÞ þ w0 yielding the values
listed in Table 3. The obtained magnetic moments per
lanthanide ion are slightly smaller than the expected ones,
which is not surprising taking into account that the free ion
approximation neglects both crystal field effects and the
contribution of higher exited states. Besides these slight
discrepancies our results clearly indicate that the magnetic
behavior of the two samples can be explained only by the
contributions of the f-elements involved, thus confirming
the presence of diamagnetic Pt4+ ions in Ln2Sr2PtO7+d, as
already deduced from the XANES results.
For Ln ¼ La and Nd, the samples had a yellow and
green color while Pr2Sr2PtO7+d was black. The optical
properties of La2Sr2PtO7+d and Nd2Sr2PtO7+d were
investigated by measurements of their diffuse reflectance
in the wavelength range 250–2000 nm. The corresponding
spectra are shown in Fig. 8. While La2Sr2PtO7+d shows
basically no absorption above approximately 500 nm, the
spectrum of Nd2Sr2PtO7+d contains various absorption
bands due to the intra-atomic f–f transitions of the Nd3+
Further evidence for the oxidation state stems from
magnetic measurements. For Pt+4 a (t2g)6 low-spin con-
figuration is expected. As a result La2Sr2PtO7+d should be
diamagnetic. The SQUID measurements shown in Fig. 7 in
fact reveal a small, nearly temperature-independent nega-
tive susceptibility (right scale of Fig. 7) due to the core
diamagnetism of the corresponding ions. The slight upturn
below 25 K may either result from the sample holder,
which exhibited an increasing paramagnetic signal at these
low temperatures or from traces of paramagnetic centers.
For technical reasons La2Sr2PtO7+d was measured in an
external magnetic field of 1 T, while for the other two
samples a field of 0.1 T was applied. As can be seen from
Pr2Sr2PtO7
1.2
Pseudo Voigt
Arctan
PtO2
1.0
0.8
0.6
0.4
0.2
0.0
6
5
0.0024
0.0020
0.0016
0.0012
0.0008
0.0004
0.0000
-0.0004
12
10
8
Pr
Nd
Pr
Nd
La
4
6
3
4
2
2
0
1
0
50 100 150 200 250 300
0
-1
11.55
11.56
Energy [keV]
11.57
11.58
0
50
100 150 200 250 300
T [K]
Fig. 6. Normalized XANES spectrum of Pr2Sr2PtO7+d in comparison to
PtO2. The dotted lines represent the pseudo-Voigt and arctangent function
used to approximate the white line and absorption edge, respectively.
Fig. 7. Magnetic susceptibility of Ln2Sr2PtO7. The very small diamagnetic
susceptibility for Ln ¼ La is shown on the right scale. The inset shows 1/w
for Ln ¼ Pr and Nd together with the fit by a Curie–Weiss behavior.

ARTICLE IN PRESS
S.G. Ebbinghaus et al. / Journal of Solid State Chemistry 180 (2007) 3393–3400
3399
Table 3
Magnetic results obtained by Curie–Weiss fits for 50 KoTo300 K
The platinates reported in this work are in principle
isostructural with earlier described ruthenates and iridates.
Nevertheless, some remarkable structural deviations were
observed. For example, we found evidences for a symmetry
m/mB
T0 (K)
w0 (cm3 molꢁ1
)
¯
¯
Pr2Sr2PtO7
Nd2Sr2PtO7
3.040(13)
3.281(9)
ꢁ7.1(7)
ꢁ0.0045(7)
reduction from space group R 3 m to R 3: In the higher
symmetric space group the oxygen ions forming the PtO6-
octahedra (O1, positioned at roughly (0.32, 0.16, 0.06))
exhibited extremely elongated displacement ellipsoids while
ꢁ16.3(4)
0.0175(5)
¯
in R 3 the refinements indicated a shift of these ions to
Energy [eV]
approximately (0.32, 0.22, 0.06). In addition, a complete
ordering of Ln3+ and Sr2+ on the two distinct crystal-
lographic A-sites was found for Pr and Nd, while for La a
slightly disordered arrangement was observed. Further-
more, no indications for vacancies on the Sr2 sites could be
found and the Sr2+ ions occupy only the 6c site (0, 0, z), in
contrast to La1.2Sr2.7BO7+d (B ¼ Ru, Ir) where a fraction
of Sr2+ ions takes a (2x, x, z) position. Finally, the d values
are significantly lower than in the corresponding Ru- and
Ir-based compounds, indicating a much smaller content of
peroxide ions, which decreases from 0.13 to zero for
Ln ¼ La and Nd.
4
3
2
1
100
80
60
40
20
80
60
40
20
La
Nd
Magnetic measurements proved La2Sr2PtO7+d to be
diamagnetic in accordance with the expected low-spin (t2g)6
electronic configuration of Pt4+. For Ln ¼ Pr and Nd the
observed paramagnetism can well be explained by the
magnetic moment of the respective rare-earth ions.
The oxidation state of platinum was additionally
investigated using X-ray absorption spectroscopy. By
comparing the energy position of the white line at the
Pt-LIII absorption edge with the PtO2 reference, platinum
was found to be tetravalent in all three samples, in
accordance with the magnetic measurements. UV–Vis
diffuse reflectance data were recorded for the yellow La
and green Nd samples. Both compounds yielded optical
band gaps of roughly 3.05 eV with an additional transition
at 2.5 eV.
400 600 800 1000
1400 1600 1800 2000
1200
Wavelength [nm]
Fig. 8. UV–Vis spectra of La2Sr2PtO7+d and Nd2Sr2PtO7+d
.
ions. The diffuse reflectance data were converted according
2
to the Kubelka–Munk function FðRÞ ¼ ðK=SÞ ¼ ð1 ꢁ RÞ =
2R. Band gap energies were estimated by extrapolating the
onset of absorption to the wavelength axis. For both
compounds the reflectance spectra show a pronounced
shoulder at approximately 415 nm. Therefore, two band
gap energies can be calculated. The obtained values are
La2Sr2PtO7+d: 2.54 eV/3.07 eV; Nd2Sr2PtO7+d: 2.50 eV/
3.03 eV. Within an estimated uncertainty of 0.05 eV the
results are identical for both compounds.
As described above, a good structural model in space
¯
group R 3 was obtained. Nevertheless, X-ray diffraction
The color of the samples in combination with the general
catalytic activity of platinum makes these samples poten-
tially interesting candidates for photocatalytic applications.
Investigations of the photocatalytic gas phase decomposi-
tion of simple volatile organic molecules are currently
carried out in our laboratory.
does not allow an unambiguous determination of the
correct space group. Also the site occupation of the
O2/O2a position and in turn the total oxygen content
cannot be determined with the desired accuracy. These
questions will be addressed in the near future using neutron
diffraction.
Acknowledgments
4. Conclusions
Thanks are due to HASYLAB for allocating synchro-
tron beamtime. This work was supported by the Deutsche
Forschungsgemeinschaft (DFG) through SFB 484.
By conventional solid state synthesis three new com-
pounds of the composition Ln2Sr2PtO7+d with Ln ¼ La,
Pr, and Nd were prepared. Besides the recently discovered
Ln3NaPtO7 (Ln ¼ La, Nd) these samples are the first
examples for n ¼ 2 members of the ½A02O1þdꢀ[AnBnꢁ1O3n]
series of hexagonal perovskites containing platinum as B-
type cation. The short list of transition metals known to
form this special structural family (Mn, Nb, Ru, and Ir)
was thus extended by one additional element.
References
[1] L.J. Bie, Y.X. Wang, J.H. Lin, C.K. Loong, J.W. Richardson,
L.P. You, C. Dong, Chem. Mater. 15 (2003) 516.
[2] L. Bie, J. Lin, Y. Wang, C.-K. Loong, Inorg. Chem. Commun. 5
(2002) 966.

ARTICLE IN PRESS
S.G. Ebbinghaus et al. / Journal of Solid State Chemistry 180 (2007) 3393–3400
3400
[3] F. Grasset, M. Zakhour, J. Darriet, J. Alloys Compd. 287 (1999) 25.
[4] C. Dussarrat, J. Fompeyrine, J. Darriet, Eur. J. Solid State Inorg.
Chem. 31 (1994) 289.
[14] T.J. Hansen, R.B. Macquart, M.D. Smith, H.-C. zur Loye, Solid
State Sci. 9 (2007) 785.
[15] P. Lunkenheimer, T. Gotzfried, R. Fichtl, S. Weber, T. Rudolf,
¨
[5] F. Grasset, C. Dussarrat, J. Darriet, J. Mater. Chem. 7 (1997) 1911.
[6] S.J. Kim, M.D. Smith, J. Darriet, H.-C. zur Loye, J. Solid State
Chem. 177 (2004) 1493.
A. Loidl, A. Reller, S.G. Ebbinghaus, J. Solid State Chem. 179 (2006)
3965.
[16] S.G. Ebbinghaus, E.-W. Scheidt, T. Go
144414.
[17] J. Rodrı
¨
tzfried, Phys. Rev. B 75 (2007)
[7] Y. Wang, J. Lin, Y. Du, R. Qin, B. Han, C.K. Loong, Angew. Chem.
Int. Ed. 39 (2000) 2730.
´
guez-Carvajal, Abstracts of the Satellite Meeting on Powder
[8] Y. Wang, L. Bie, Y. Du, J. Lin, C.-K. Loong, J. Richardson, L. You,
J. Solid State Chem. 177 (2004) 65.
Diffraction of the XV Congress of the IUCr, 1990, p. 127.
[18] R.D. Shannon, Acta Crystallogr. A 32 (1976) 751.
[19] J.-H. Choy, J.-Y. Kim, S.-H. Hwang, S.-J. Kim, G. Demazeau, Int. J.
Inorg. Mater. 2 (2000) 61.
[9] E. Gaudin, G. Goglio, A. Besnard, J. Darriet, J. Solid State Chem.
175 (2003) 124.
[10] S.G. Ebbinghaus, J. Solid State Chem. 177 (2004) 817.
[20] J.-H. Choy, D.-K. Kim, G. Demazeau, D.-Y. Jung, J. Phys. Chem. 98
(1994) 6258.
[11] T. Gotzfried, A. Reller, S.G. Ebbinghaus, Solid State Sci. 6 (2004)
¨
1205.
[21] J.-H. Choy, D.-K. Kim, S.-H. Hwang, G. Demazeau, D.-Y. Jung,
J. Am. Chem. Soc. 117 (1995) 8557.
[12] T. Gotzfried, A. Reller, S.G. Ebbinghaus, Inorg. Chem. 44 (2005)
¨
6550.
[22] S. Ebbinghaus, Z. Hu, A. Reller, J. Solid State Chem. 156 (2001) 194.
[23] Z. Hu, H. von Lips, M.S. Golden, J. Fink, G. Kaindl, F.M.F. de
Groot, S. Ebbinghaus, A. Reller, Phys. Rev. B 61 (2000) 5262.
[13] S.J. Mugavero III, M.D. Smith, H.-C. zur Loye, J. Solid State Chem.
178 (2005) 3176.