Dalton
Transactions
PAPER
Cite this: Dalton Trans., 2015, 44,
8238
Received 17th February 2015,
Accepted 27th March 2015
DOI: 10.1039/c5dt00711a
www.rsc.org/dalton
A study on the thermal conversion of scheelite-
type ABO
4
into perovskite-type AB(O,N)
3
†
Wenjie Li,
a
Duan Li,
b
Xin Gao,
b
Aleksander Gurlo,
c
Stefan Zander,
d
Philip Jones,
e
Alexandra Navrotsky,
e
Zhijian Shen,
b
Ralf Riedel
a
and Emanuel Ionescu*
a
Phase-pure scheelite AMoO
4
and AWO
4
(A = Ba, Sr, Ca) were thermally treated under an ammonia atmos-
phere at 400 to 900 °C. SrMoO
4
and SrWO
4
were shown to convert into cubic perovskite SrMoO
2
N and
SrWO
1.5
N
1.5
, at 700 °C and 900 °C respectively, and to form metastable intermediate phases (scheelite
SrMoO
4−x
N
x
and SrWO
4−x
N
x
), as revealed by X-ray diffraction (XRD), elemental analysis and FTIR spectro-
scopy. High-temperature oxide melt solution calorimetry reveals that the enthalpy of formation for
SrM(O,N)
3
(M = Mo, W) perovskites is less negative than that of the corresponding scheelite oxides,
though the conversion of the scheelite oxides into perovskite oxynitrides is thermodynamically favorable
at moderate temperatures. The reaction of BaMO
4
with ammonia leads to the formation of rhombohedral
Ba
3
M
2
(O,N)
8
and the corresponding binary metal nitrides Mo
3
N
2
and W
4.6
N
4
; similar behavior was
observed for CaMO
4
, which converted upon ammonolysis into individual oxides and nitrides. Thus,
BaMO
4
and CaMO
4
were shown to not provide access to perovskite oxynitrides. The influence of the
starting scheelite oxide precursor, the structure distortion and the degree of covalency of the B-site-N
bond are discussed within the context of the formability of perovskite oxynitrides.
1 Introduction
Perovskite oxynitrides AB(O,N)
3
are typically synthesized via
ammonolysis of oxide precursors; thus they can be formally
represented as nitrogen-substituted perovskite-type oxides,
1,2
which are an emerging class of materials suitable for novel
applications in the fields of energy conversion, storage, non-
toxic pigments, dielectrics, etc.
3
Most perovskite-type oxynitrides are synthesized via conver-
sion of scheelite-type ABO
4
and pyrochlore-type A
2
B
2
O
7
upon
thermal treatment under an ammonia atmosphere. However,
not all scheelite- and pyrochlore-type oxides are able to afford
perovskite oxynitrides. For example, pyrochlore-type La
2
Zr
2
O
7
as
well as scheelite-type EuMO
4
(M = Nb and Ta) and SrMoO
4
provide perovskite-type LaZrO
2
N
4
as well as EuMO
2
N
5
and
SrMoO
2
N,
6
respectively, whereas other precursor oxides such as
scheelite-type ATaO
4
(A = Nd, Sm, Gd, Dy) and A
2
W
2
O
9
(A = Pr,
Nd, Sm, Gd, Dy) convert upon ammonolysis into pyrochlore-
type A
2
Ta
2
O
5
N
2
,
7
and scheelite-type AWO
3
N,
8
respectively.
According to our previous work, only a limited number of
perovskite-type oxynitrides are formable.
9
For instance,
SrMoO
2
N, SrWO
2
N, CaMoO
2
N and CaWO
2
N appear to be feas-
ible, while BaMoO
2
N and BaWO
2
N are not stable in the perovs-
kite-type structure. Although perovskite-type SrMo(O,N)
3
, SrW-
(O,N)
3
and CaMo(O,N)
36,10–14
have been reported in the litera-
ture (consistent with our prediction based on tolerance and
octahedral factors),
9
details of the structure evolution of the
oxides into perovskite-type oxynitrides are scarce. Furthermore,
the existence of perovskite-type BaMo(O,N)
3
is question-
able;
11,12
whereas perovskite-type BaW(O,N)
3
and CaW(O,N)
3
have not yet been synthesized.
2 Experimental methods
2.1 Synthesis
Scheelite-type oxide precursors (i.e., SrMoO
4
, SrWO
4
, BaMoO
4
,
BaWO
4
, CaMoO
4
and CaWO
4
) were synthesized via solvo-
thermal methods. Thus, Sr(NO
3
)
2
(Sigma-Aldrich, >99.0%),
†Electronic supplementary information (ESI) available: XRD, FTIR, lattice para-
meters and the phase composition obtained by Rietveld refinement, elemental
analysis and the enthalpies of formation results as well as Gibbs free energy cal-
culation. See DOI: 10.1039/c5dt00711a
a
Fachbereich Material- und Geowissenschaften Technische Universität Darmstadt,
Fax: +49 (0)6151 16 6346; Tel: +49 (0)6151 16 6342
b
Department of Materials and Environmental Chemistry, Arrhenius Laboratory,
Stockholm University, S-106 91 Stockholm, Sweden
c
Fachgebiet Keramische Werkstoffe, Institut für Werkstoffwissenschaften und –
technologien Fakultät III Prozesswissenschaften, Technische Universität Berlin,
Hardenbergstraße 40, 10623 Berlin, Germany
d
Helmholtz-Zentrum Berlin für Materialien und Energie, Department of
Crystallography, Hahn-Meitner-Platz 1, 14109 Berlin, Germany
e
Peter A. Rock Thermochemistry Laboratory and NEATORU, University of California
Davis, Davis, CA 95616-8779, USA
8238 |Dalton Trans.,2015,44, 8238–8246 This journal is © The Royal Society of Chemistry 2015
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
View Journal
| View Issue
Ba(NO
3
)
2
(Sigma-Aldrich, >99.0%) or Ca(NO
3
)
2
·4H
2
O (Sigma-
Aldrich, >99.0%) was mixed in an equimolar ratio with
Na
2
MoO
4
·4H
2
O (Sigma-Aldrich, >99.5%) or Na
2
WO
4
·4H
2
O
(Sigma-Aldrich, >99.5%) in ethylenediamine (FLUKA, >99.5%)
under vigorous stirring. Subsequently, the reaction mixture
was transferred into an autoclave with Teflon lining and
heated at 200 °C for 24 h. The resulting mixture was rinsed
5 times with deionized water and ethanol alternately. Centrifu-
gation and drying at 60 °C overnight led to the powdered
scheelite-type oxides.
The resulting oxides were ground to fine powders (grain
size < 500 nm) and placed in a silica crucible. The thermal
treatments were carried out in flowing ammonia at tempera-
tures between 400 and 900 °C for 4–24 h. The Schlenk system
used for thermal ammonolysis is specifically limited to small
batch sizes (about 0.3–0.5 g) to maximize exposure to flowing
NH
3
and thus the product homogeneity.
2.2 Sample characterization
The crystalline phase composition of the as-synthesized
samples was analyzed by using powder X-ray diffraction (XRD,
STOE STADI P) with Mo Kαradiation (wavelength of 0.7093 Å).
The oxygen and nitrogen contents of the synthesized samples
were determined by hot gas extraction using a Leco TC436 ana-
lyzer. Fourier Transform Infrared (FT-IR) spectroscopy was per-
formed on a Varian 670-IR spectrometer. Thermogravimetric
analysis (TGA 92, SETARAM) under an ammonia atmosphere
was performed to obtain the weight change of samples. A high
resolution transmission electron microscope (HRTEM, JEOL
JEM-2100F) was used to assess the morphology and the local
crystallinity of the samples.
High temperature oxidative-solution calorimetry was used
to determine the enthalpies of formation of the prepared
oxynitride samples. This method is well developed
15–18
and
has been applied previously to study nitrides
19–21
and
oxynitrides.
22–24
Using this technique, ∼5 mg pellets, prepared
by pressing the powders into a die with a diameter of 1 mm,
were dropped from room temperature into molten sodium
molybdate (3Na
2
O·4MoO
3
) solvent at 701 °C using a custom
made Tian–Calvet twin microcalorimeter.
15,17
Neutron diffraction (ND) experiments were performed
using a high resolution powder diffractometer for thermal neu-
trons (HRPT)
25
located at the Swiss Spallation Neutron Source
(SINQ) of the Paul Scherrer Institute in Switzerland and the
Fine Resolution Powder Diffractometer (FIREPOD, E9)
26
at the
BERII of the Helmholtz-Zentrum Berlin (HZB), Germany. The
measurements were performed using a neutron wavelength of
λ= 1.494 Å at SINQ and λ= 1.308 Å at HZB. Crystallographic
parameters were confirmed by the individual Rietveld refine-
ments of the XRD and ND patterns. The peak shapes were
modeled with the pseudo-Voigt function for XRD and the
Thompson-Cox-Hastings pseudo-Voigt function
27
for ND pat-
terns. Isotropic thermal parameters of O/N were constrained to
the same value for the anions. All refinements were performed
with the Fullprof software.
28
3 Results and discussion
3.1 Ammonolysis of scheelite-type oxides
3.1.1 BaMoO
4
and BaWO
4
.The ammonolysis of the schee-
lite-type BaMoO
4
was performed at 600, 700 and 900 °C for
6 h. The sample treated at 600 °C already formed small
amounts of the Ba
3
Mo
2
(O,N)
8
oxynitride phase (structure iden-
tical to Ba
3
Mo
2
N
6
N
229,30
), as shown in Fig. S1.†At 700 °C,
Mo
3
N
231
and BaMoO
332
were observed. Up to 900 °C, only
small amounts of Mo
3
N
2
were detected besides the main phase
Ba
3
Mo
2
(O,N)
8
. The absence of the perovskite-type BaMo(O,N)
3
is consistent with the experimental work of Liu et al.
12
and our
previousprediction.
9
The crystallographic data and phase com-
positions of the samples obtained at 700 and 900 °C were ana-
lyzed by Rietveld refinement (Fig. S2a and b†). The refined
lattice parameter of BaMoO
3
was 4.0489 (6) Å, which is similar
to reported values.
11,33
No cubic perovskite BaMo(O,N)
3
was
formed. The lattice parameters of the rhombohedral
Ba
3
Mo
2
(O,N)
8
were 5.9670 (3) and 21.4812 (10) Å (Table S1†);
these values are smaller than those of Ba
3
Mo
2
O
6
N
2
(5.9706 (5)
and 21.5020 (6) Å)
34
probably because of the lower nitrogen
content in our as-synthesized oxynitrides (however, we balanced
eqn (1)–(3) based on Ba
3
Mo
2
O
6
N
2
and Ba
3
W
2
O
6
N
2
).
A noticeable reaction between BaWO
4
and NH
3
occurs at
700 °C (Fig. S3†). Compared to BaMoO
4
, BaWO
4
seems to be
rather more inert against ammonia, thus more than 50 wt% of
BaWO
4
still remained after ammonolysis at temperatures up to
850 °C (Table S2†). Hence, the ammonolysis of BaWO
4
at 700
and 850 °C leads to a mixture consisting of BaWO
4
,Ba
3
W
2
(O,
N)
8
and W
4.6
N
4
. The lattice parameters of Ba
3
W
2
(O,N)
8
and
W
4.6
N
4
assessed by Rietveld refinement of the XRD patterns
(Fig. S4†) are close to those reported in ref. 29 and 34 (see also
the ESI, Tables S2 and S4†). As we recently predicted,
9
the per-
ovskite-type BaWO
2
N cannot be formed.
The samples obtained upon ammonolysis of BaMoO
4
and
BaWO
4
were also investigated by FTIR spectroscopy; both show
an absorption band around 975 cm
−1
for oxynitride (Fig. S5†),
which was assigned to a stretching mode (ν(M–N)) in
(WO
3
N)
3−
/(MoO
3
N)
3−
, having W
6+
/Mo
6+
in tetrahedral coordi-
nation, as reported by Herle et al.
30
Thus ammonolysis of the scheelite-type oxides BaWO
4
and
BaMoO
4
leads to non-perovskite oxynitride products following
the paths proposed in eqn (1)–(3):
10BaMoO4þ38
3NH3)
700°C 3Ba3Mo2ðO;NÞ8þBaMoO3þMo3N2
þ19H2Oþ14
3N2
ð1Þ
9BaMoO4þ12NH3)
900°C 3Ba3Mo2ðO;NÞ8þMo3N2þ18H2Oþ2N2
ð2Þ
69BaWO4þ92NH3)
850 °C 23Ba3W2ðO;NÞ8þ5W4:6N4þ138H2O
þ13N2
ð3Þ
Dalton Transactions Paper
This journal is © The Royal Society of Chemistry 2015 Dalton Trans.,2015,44,8238–8246 | 8239
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
3.1.2 SrMoO
4
and SrWO
4
.A similar ammonolysis pro-
cedure was applied to the scheelite-type SrMoO
4
and SrWO
4
.
XRD measurements confirm that the ammonolysis of
SrMoO
4
at 700 °C for 4 h leads to the formation of SrMoO
2
N
(Fig. 1). The change of the O/N ratio with annealing time in
SrMoO
2
N was shown to decrease from 2.3 upon annealing
time of 4 h (the empirical chemical composition of the oxy-
nitride SrMoO
2.09(1)
N
0.91(1)
) to 1.89 after 12 h (SrMoO
1.96(1)
N
1.04(1)
)
and to 1.54 after 24 h of ammonolysis (SrMoO
1.82(1)
N
1.18(1)
).
However, the nitrogen incorporation seems to have limitations
under the conditions used, thus nitrogen-rich compositions (e.g.,
SrMoON
2
,withanO/Nratioof0.5)arenotaccessibleinthisway.
The ammonolysis of scheelite-type SrWO
4
at 900 °C leads to
the corresponding perovskite-type oxynitride as well (Fig. 2).
However, the temperature required to obtain phase-pure Sr,W-
based oxynitrides was higher than that used for SrMoO
4
. The
nitrogen content of the SrW(O,N)
3
increases slightly with the
increasing temperature and holding time. Moreover, the O/N
ratio in SrW(O,N)
3
seems to be more constant than that in
SrMo(O,N)
3
and appears to be independent of the annealing
time. Thus, the O/N ratio decreases only slightly as the anneal-
ing time was extended from 4 h (SrWO
1.50(6)
N
1.50(6)
; O/N ratio
1.08) to 12 h (SrWO
1.42(2)
N
1.58(2)
; O/N ratio 0.98) and to 24 h
(SrWO
1.39(2)
N
1.61(2)
; O/N ratio 0.86). Interestingly, the Sr,W-
based system can accommodate more nitrogen than its analo-
gous Sr,Mo-based system. Nevertheless the O/N ratio still
cannot be pushed down to 0.5.
3.1.3 CaMoO
4
and CaWO
4
.The ammonolysis of CaMoO
4
was found to proceed in a different way, leading to the for-
mation of CaO and various molybdenum nitrides (including
Mo
2
N, Mo
3
N
2
and MoN, depending on the temperature, time
and ammonia flow) (Fig. S6†) and consequently CaMoO
4
was
not considered further as a precursor for the corresponding
perovskite-type oxynitrides.
Ammonolysis of CaWO
4
at 900 °C for 6 h leads to complete
decomposition into Ca
3
WO
6
and W
4.6
N
4
and no oxynitride
phase was observed (Fig. S6†). For both CaMoO
4
and CaWO
4
,
the corresponding perovskite oxynitrides did not form and
thus their conversion into an oxide/nitride mixture is assumed
to occur as follows:
6CaMoO4þ12NH3!
700°C 6CaO þMo2NþMo3N2þMoN
þ18H2Oþ4N2ð4Þ
6:9CaWO4þ9:2NH3!
900°C 11:5Ca3WO6þW4:6N4þ13:8H2O
þ2:6N2
ð5Þ
3.2 Intermediate oxynitride phase during the conversion of
SrMoO
4
into perovskite-type SrMoO
2
N
An interesting phenomenon during the ammonolysis of
SrMoO
4
at 600 °C relates to the incorporation of 2.23 wt%
nitrogen without the formation of any new crystalline phase;
thus, the color of the sample changed from white to light-
grayish and the FTIR spectrum showed a new absorption band
at 978 cm
−1
related to (MoO
3
N)
3−
units in tetrahedral coordi-
nation (Fig. 3a),
30
as also observed in Ba
3
Mo
2
(O,N)
8
. Tetra-
coordinated Mo
6+
in scheelite-type SrMoO
4
can be identified
by FTIR spectroscopy via a very broad band around 822 cm
−1
representing the antisymmetric stretching vibrations of Mo–O
in (MoO
4
)
2−
tetrahedral units.
35
Thus, the formation of
(MoO
3
N)
3−
is considered to be a result of the substitution of
one oxygen with nitrogen in (MoO
4
)
2−
tetrahedra. Therefore,
we assume that an intermediate scheelite-type oxynitride
phase SrMoO
4−x
N
x
(x= 0.39 in our experiment, as obtained
from elemental analysis and Rietveld refinement, Fig. 4) forms
at 600 °C, which subsequently rearranges into the perovskite
structure while taking up more nitrogen. The absence of the
absorption band of (MoO
3
N)
3−
in the samples obtained upon
Fig. 1 XRD patterns of SrMoO
4
after heating at 400, 600 and 700 °C
for different times under an ammonia flow in forming gas (a mixture of 5
vol% H
2
and 95 vol% N
2
). The arrow indicates the diffraction pattern of
the oxynitride obtained upon ammonolysis of SrMoO
3
which was syn-
thesized by reducing SrMoO
4
under an ammonia flow at 700 °C.
Fig. 2 XRD patterns of SrWO
4
after heating at 400, 600, 700 and
900 °C for different times under an ammonia flow.
Paper Dalton Transactions
8240 |Dalton Trans.,2015,44,8238–8246 This journal is © The Royal Society of Chemistry 2015
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
ammonolysis at temperatures above 700 °C might be related to
the strong absorption of the black sample.
Yang et al.
36
reported the formation of analogous scheelite-
type EuWO
4−y
N
y
oxynitride as the intermediate phase during
the nitridation from Eu
2
W
2
O
9
to EuWO
1+x
N
2−x
. However, in
their case, the nitrogen substitution is compensated by the
partial oxidation of Eu
2+
to Eu
3+
(y= 0.04 in EuWO
4−y
N
y
;i.e.,
Eu
2+1−y
Eu
3+y
WO
4−y
N
y36,37
). In our system, Sr
2+
is not able to be
oxidized to Sr
3+
,soadifferent mechanism must be responsible
for the formation of the nitrogen-containing scheelite-based
phase. A likely explanation is that the generation of oxygen
vacancies compensates the extra negative charge due to the
replacement of oxygen by nitrogen within the pre-formed crys-
tallites, which usually occurs for nitrogen-doped TiO
238,39
or
HfO
2
.
40
As shown in the HRTEM image within the FFT pattern
(Fig. 5), the crystalline phase in the sample obtained after
ammonolysis of SrMoO
4
at 600 °C for 4 h was indexed as tetra-
gonal (I41/a,i.e. the same as scheelite-type SrMoO
4
) and
exhibited the presence of pores. Some defect regions with
different fringe distances were observed as well probably due
to the distortion of the lattice. Interestingly, thermogravimetric
analysis of the SrMoO
4
in ammonia revealed a slight mass
increase of the sample at temperatures up to 600 °C (Fig. 6),
indicating that the oxygen, which is expected to be released
from SrMoO
4
upon ammonolysis, might be stored at inter-
mediate temperatures in the pores or interstitially in the struc-
ture as molecular oxygen
13
before being released (as shown by
the mass loss of SrMoO
4
at temperatures beyond 700 °C, see
Fig. 6). This was shown to be the reason for anomalous mag-
netic behavior at T=−219 °C (54 K) as reported by Logvinovich
et al. The sharp weight loss above 650 °C is attributed to the
complete conversion from scheelite to perovksite resulting in
1 mol oxygen release. Elemental analyses confirm the expected
oxygen loss for samples heated in NH
3
between 600 and
700 °C and are in agreement with the measured mass loss,
indicating that nitrogen is already incorporated into the
sample at 600 °C. (Tables 1 and S6†).
Based on all these observations, we conclude that the nitri-
dation of SrMoO
4
occurs prior to the reduction of W
6+
during
ammonolysis, thus scheelite-type SrMoO
4−x
N
x
forms as an
intermediate phase and decomposes fast according to the fol-
lowing paths (σstands for the amount of oxygen vacancies):
SrMoO4þxNH3!
600°C SrMoO4σxV
OσNxþσ
2
O2þxH2O
þx
2H2
ð6Þ
SrMoO4σxV
OσNxþNH3!
>600°C SrMoO3xyNxþyþH2OþH2
ð7Þ
3.3 Ammonolysis of SrMoO
4
vs. SrMoO
3
In order to investigate the influence of the oxide precursor on
the final oxynitride, we converted the scheelite-type SrMoO
4
Fig. 3 FTIR spectrum of the as-synthesized scheelite oxide: (a) SrMoO
4
,
(b) SrWO
4
and the resulting oxynitrides from ammonolysis at different
temperatures (400, 600, 700 and 900 °C) for 6 h.
Fig. 4 Rietveld patterns of the X-ray powder diffraction data of the
sample obtained upon ammonolysis of SrMoO
4
at 600 °C for 4 h. Blue
tick marks are Bragg peak positions of the related phase SrMoO
3.61(3)
-
N
0.39(3)
(the ratio of O/N was fixed based on the results of the elemental
analysis). The green line at the bottom denotes the difference in intensi-
ties between the observed and calculated profiles. Table S3†summar-
izes the results of the structure refinement.
Dalton Transactions Paper
This journal is © The Royal Society of Chemistry 2015 Dalton Trans.,2015,44, 8238–8246 | 8241
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
into SrMo(O,N)
3
via a two-step process as well. In the first step,
the scheelite-type oxide SrMoO
4
was easily reduced to the
perovskite-type SrMoO
3
(Fig. S7†) upon thermal annealing at
900 °C for 6 h under forming gas (a mixture of 5 vol% H
2
and
95 vol% N
2
). In the subsequent step, SrMoO
3
underwent
ammonolysis at 700 °C for 4 h (the same conditions as for
SrMoO
4
) to obtain SrMo(O,N)
3
. Interestingly, the nitrogen
content of the phase-pure perovskite-type oxynitride (empirical
formula SrMoO
2.77(3)
N
0.23(3)
, see the Rietveld refinement data
of the neutron diffraction pattern shown in Fig. 7a) obtained
from perovskite-type SrMoO
3
was significantly lower than that
of the oxynitride obtained under the same conditions from
SrMoO
4
(SrMoO
2.19(2)
O
0.81(2)
). This obviously relates to the oxi-
dation state of Mo in SrMoO
4
and SrMoO
3
and its evolution
under an ammonia atmosphere which will be discussed later.
Moreover, the attempt to synthesize perovskite-type SrWO
3
upon reducing SrWO
4
at high temperatures was unsuccessful.
3.4 Structural verification of perovskite oxynitrides
The neutron powder diffraction data measured at room temp-
erature for SrMo(O,N)
3
and SrW(O,N)
3
were refined by the Riet-
veld method on the basis of the cubic Pm3
ˉ
mperovskite-type
structure (Fig. 7 and Table 2). The refined O/N content of
SrMoO
2.19(2)
N
0.81(2)
(700 °C for 4 h) and SrWO
1.50(6)
N
1.50(6)
(900 °C for 4 h) is consistent with the results of elemental ana-
lysis (ESI, see Table S6†).
The enthalpies of dissolution (ΔH
ds
) and formation (ΔH
f
)
of scheelite-type SrMoO
4
and SrWO
4
and the corresponding
perovskite-type oxynitride samples measured by high tempera-
ture oxide melt solution calorimetry are listed in Table 3.
The enthalpies of formation of the oxides and oxynitrides
from the elements were calculated using the thermodynamic
cycles shown in Tables S7 and S8†and are given in Table 3.
The enthalpy of formation of SrMoO
4
(−260.2 ± 0.5 kJ per
g-atom) is ∼36 kJ per g-atom more exothermic than that of
SrMoO
1.96
N
1.04
(−223.8 ± 0.7 kJ per g-atom). Likewise, the
enthalpy of formation of SrWO
4
(−273.4 ± 0.5 kJ per g-atom) is
∼83 kJ per g-atom more exothermic than that of SrWO
1.5
N
1.5
Fig. 5 HRTEM micrographs of SrMoO
4
after heating at 600 °C for 4 h.
Table 1 Experimental and calculated mass loss of SrMoO
4
upon
ammonolysis at 600 and 700 °C. The calculated mass loss relies on
the evolution of the chemical composition of the sample upon
ammonolysis
Specimens Experiment Empirical formula Calculated
SrMoO
4
_NH600_4H 0.4218 g SrMoO
3.61
N
0.39
SrMoO
4
_NH700_12H 0.3925 g SrMoO
1.96
N
1.04
Mass loss (wt %) 6.95 6.97
Fig. 6 TG curve of SrMoO
4
under an ammonia atmosphere from room
temperature to 800 °C.
Paper Dalton Transactions
8242 |Dalton Trans.,2015,44, 8238–8246 This journal is © The Royal Society of Chemistry 2015
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
(−190.4 ± 0.7 kJ per g-atom). Thus, perovskite-type oxynitrides
show less favorable enthalpies of formation than their corres-
ponding scheelite-type oxides. Furthermore, the difference of
the enthalpy of formation for Sr–W is larger than that of Sr–
Mo. This suggests that the formation of Sr–W oxynitrides is
less favorable and requires higher temperatures (as observed),
probably for both thermodynamic and kinetic reasons.
In order to gain further insights into the energetics of the
conversion of SrMO
4
into SrM(O,N)
3
(M = Mo, W) under an
ammonia atmosphere, the Gibbs free energy (ΔG) of reactions
(8) and (9) was calculated (Tables S9 and S10†). Since the
entropies of SrMoO
2
N and SrWO
1.5
N
1.5
are not available, we
estimated them as 5/6 of the entropy of the corresponding
scheelite-type oxide. Eqn (9) and (11) describe the temperature
evolution of the Gibbs free energy of the reaction of SrMO
4
with NH
3
to give SrM(O,N)
3
, indicating that the reaction is
spontaneous at temperatures exceeding 992 K (i.e., 719. °C) for
SrWO
4
, whereas for SrMoO
4
the reaction seems to be thermo-
dynamically favorable at any of the temperatures used for its
ammonolysis (Fig. 8). It is worth pointing out that only a
thermodynamic consideration might not be enough to
describe the ammonolysis processes of the scheelite oxides.
The kinetics (e.g., activation energy) of the ammonolysis prob-
ably also play an important role and thus might explain why
the conversion of SrMoO
4
into the perovskite oxynitride needs
temperatures exceeding 600 °C and proceeds through an inter-
mediate phase.
SrMoO4þ2NH3¼SrMoO2Nþ2H2OþH2þ1
2N2ð8Þ
ΔGSr–Mo ðkJ mol1Þ¼50:404 0:197Tð9Þ
SrWO4þ2NH3¼SrWO1:5N1:5þ5
2H2Oþ1
2H2þ1
4N2ð10Þ
ΔGSr–WðkJ mol1Þ¼175:615 0:177Tð11Þ
The negative temperature dependence of the free energy
reflects positive entropy of the reaction because 1.5 moles of
gas are produced.
3.5 Factors affecting the formation of perovskite-type
oxynitrides
As addressed above, the experimental results related to the
conversion of BaMoO
4
, BaWO
4
, SrMoO
4
and SrWO
4
into per-
ovskite-type oxynitrides are consistent with our prediction.
9
However, CaMoO
4
and CaWO
4
do not appear to be converted
to oxynitrides.
Scheelite-type ABO
4
oxides are rather common precursors
for the synthesis of perovskite oxynitrides, e.g. Nd
3+
V
4+
O
2
N,
41
Eu
2+
Nb
5+
O
2
N
5
,La
3+
Nb
4+
O
2
N,
42
Ca
2+x
Sr
2+1−x
W
5+
O
2
N
43
and so
on. The formation of hydrogen due to the dissociation of
ammonia at high temperatures is beneficial for the reduction
of the B-site cation in scheelite-type oxides (e.g. from A
2+
B
6+
O
4
to A
2+
B
5+
O
2
NorfromA
3+
B
5+
O
4
to A
3+
B
4+
O
2
N). In the case of
using perovskite oxides as precursors for perovskite-type oxy-
nitrides, the B-site cation has to be oxidized in order to com-
pensate for the increase of the negative charge resulting from
nitrogen incorporation (e.g. from Sr
2+
Mo
4+
O
3
to Sr
2+
Mo
5+
O
2
N).
Thus, it seems that scheelite-type oxide precursors are more
favorable for the synthesis of perovskite-type oxynitrides.
Fig. 7 Rietveld patterns of the neutron powder diffraction data of the
sample obtained upon ammonolysis of (a) SrMoO
3
at 700 °C for 4 h
(FIREPOD, E9); (b) SrMoO
4
at 700 °C for 4 h (HRPT, SINQ) and (c) SrWO
4
at 900 °C for 4 h (FIREPOD, E9). Blue tick marks are Bragg peak posi-
tions of related phases (a) SrMoO
2.77(3)
N
0.23(3)
; (b) SrMoO
2.19(2)
N
0.81(2)
and
(c) SrWO
1.50(6)
N
1.50(6)
. The green line at the bottom denotes the differ-
ence in intensities between the observed and calculated profiles.
Dalton Transactions Paper
This journal is © The Royal Society of Chemistry 2015 Dalton Trans.,2015,44, 8238–8246 | 8243
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
Moreover, parameters such as the tolerance factor (describ-
ing the distortion of the cubic perovskite structure) were
shown to be crucial for the formability of perovskite-type oxy-
nitrides.
9
As defined by Goldschmidt,
44
the tolerance factor (t)
in ABX
3
is expressed as:
tO¼ðrArXÞ
ffiffiffi
2
pðrBrXÞð12Þ
r
A
,r
B
and r
X
being the ionic radii of A, B and X atoms,
respectively.
In our previous work,
9
the formability of perovskite-type
oxynitrides was also rationalized upon assessing the values of
the tolerance factor, defined as in eqn (13) (see Table 4, as for
the O/N ratio 2, i.e. ABO
2
N):
toxy ¼½ðrAþrOÞ8ðrAþrNÞ41=12
ffiffiffi
2
p½ðrBþrOÞ4ðrBþrNÞ21=6ð13Þ
For a general consideration of the formability of perovskite-
type oxynitrides, we compared their tolerance factors with
those of the corresponding perovskite-type oxides. The values
of the tolerance factors t
o
and t
oxy
calculated from the ionic
radii
45
are shown in Table 4 and indicate that the formal
substitution of O
2−
with N
3−
in SrMoO
3
, SrWO
3
, CaMoO
3
and
CaWO
3
reduces the structural distortion (i.e., the tolerance
factor becomes closer to unity), which suggests that the for-
mation of the corresponding perovskite-type oxynitrides is
favorable. This is in agreement with the experiment for Sr–Mo/
Sr–W compounds and does not fit the experimental obser-
vations for Ca–Mo/Ca–W compositions. Large basic cations
like Ca typically stabilize higher oxidation states of the tran-
sition metals (Mo, W as in our case),
46,47
thus this may explain
why the Ca scheelite-type oxides cannot be converted into
oxynitrides.
In contrast, incorporation of nitrogen into BaMoO
3
and
BaWO
3
increases the structural distortion; thus, the formation
of BaMoO
2
N and BaWO
2
N would be less favorable. This is in
agreement with our synthetic observation.
Moreover, the higher covalent character of the B-site-N
bond than that of the B-site-O bond might also induce struc-
tural distortion into the perovskite structure of oxynitrides as
Table 3 Thermochemical data obtained by drop-solution calorimetry of scheelite-type oxides and their corresponding perovskite-type oxynitrides
Composition Crystal structure ΔH
ds
(kJ mol
−1
)ΔH
f
(kJ mol
−1
)ΔH
f
(kJ per g-atom)
SrMoO
4
Tetragonal/scheelite 161.8 ± 1.5 −1561.3 ± 3.1 −260.2 ± 0.5
SrMoO
1.96
N
1.04
Cubic/perovskite −291.9 ± 2.3 −1119.1 ± 3.6 −223.8 ± 0.7
SrWO
4
Tetragonal/scheelite −162.8 ± 1.5 −1641.2 ± 3.1 −273.4 ± 0.5
SrWO
1.5
N
1.5
Cubic/perovskite −537.2 ± 1.9 −952.9 ± 3.6 −190.4 ± 0.7
Fig. 8 Gibbs free energy (ΔG) for the ammonolysis of SrMoO
4
and
SrWO
4
(eqn (8) and (10), respectively) as a function of temperature.
Table 2 Crystal structure data of AB(O,N)
3
perovskite oxynitrides
Specimens and
parameters SrMoO
2.77(3)
N
0.23(3)
SrMoO
2.19(2)
N
0.81(2)
SrWO
1.50(6)
N
1.50(6)
S.G. Pm3
ˉ
m, Nr. 221 Pm3
ˉ
m, Nr. 221 Pm3
ˉ
m, Nr. 221
Z111
a,b,c(Å) 3.9744(3) 3.9756(1) 3.9856(2)
Sr x,y,z0.5, 0.5, 0.5 0.5, 0.5, 0.5 0.5,0.5, 0.5
B
iso
(Å
2
) 0.666(25) 0.879(21) 0.738(56)
Occ. 1 1 1
Mo/W x,y,z0.0, 0.0, 0.0 0.0, 0.0, 0.0 0.0, 0.0, 0.0
B
iso
(Å
2
) 0.298(23) 0.693(18) 0.880(57)
Occ. 1 1 1
O/N x,y,z0.5, 0.0, 0.0 0.5, 0.0, 0.0 0.5, 0.0, 0.0
B
iso
(Å
2
) 0.748(18) 0.799(12) 0.798(32)
Occ. 2.77(3)/0.23(3) 2.19(2)/0.81(2) 1.50(6)/1.50(6)
Paper Dalton Transactions
8244 |Dalton Trans.,2015,44,8238–8246 This journal is © The Royal Society of Chemistry 2015
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
compared to their analogous perovskite oxides, i.e. B(O,N)
6
octahedra are expected to be more distorted than their analo-
gous BO
6
octahedra. This structural distortion might be quite
pronounced, as for compounds which exhibit so-called
second-order Jahn–Teller distortion
48
(i.e. d
0
B-site octahedra
such as in LaZrO
2
N, NdTiO
2
N or LaTiO
2
N).
4
However, we con-
sider in our compound SrM(O,N)
3
the first-order Jahn–Teller
effect is relevant and thus the contribution of the B-site-N
covalency to the distortion might not be significant.
4 Conclusions
In the present study, preparative possibilities to access perov-
skite-type oxynitrides AM(O,N)
3
(A = Ba, Sr, Ca; B = Mo, W)
phases upon thermal ammonolysis of scheelite-type AMO
4
oxide precursors were investigated. The as-synthesized results
of perovskite-oxynitrides are consistent with our previous pre-
diction in general.
The experimental data reveal that both scheelite-type
SrMoO
4
and SrWO
4
transform into a scheelite-type oxynitride
intermediate phase, SrMO
4−x
N
x
(M = Mo, W), which sub-
sequently converts fast into perovskite-type SrM(O,N)
3
at temp-
eratures above 600 °C and are in agreement with the high
temperature oxide melt solution calorimetry experiments
which indicate that the conversion of scheelite SrMO
4
into
perovskite SrM(O,N)
3
is thermodynamically favorable at the
ammonolysis temperatures used.
Furthermore, the formability of the perovskite-type oxy-
nitrides depends on the structure of the oxide precursor used
(scheelite seems to be favorable, except for large basic A
cations) and on the structural distortion described by the toler-
ance factor.
Acknowledgements
The authors acknowledge Dr Samuel Bernard (IEM, University
Montpellier 2) for the TGA of SrMoO
4
under an ammonia
atmosphere. This work was supported by Dr Denis Sheptyakov
and based on experiments performed at the Swiss Spallation
Neutron Source (SINQ), Paul Scherrer Institute, Villigen, Swit-
zerland. This research was funded by the European Union
Seventh Framework Programme (FP7/2007–2013) under the
grant agreement FUNEA –Functional Nitrides for Energy
Applications. The calorimetry at UC Davis was supported by
the U.S. Dept. of Energy, Office of Basic Energy Sciences, grant
DE-FG02-03ER46053.
Notes and references
1 A. Fuertes, J. Mater. Chem., 2012, 22, 3293–3299.
2 Y. I. Kim and P. M. Woodward, J. Solid State Chem., 2007,
180, 3224–3233.
3 S. G. Ebbinghaus, H. P. Abicht, R. Dronskowski, T. Müller,
A. Reller and A. Weidenkaff,Prog. Solid State Chem., 2009,
37, 173–205.
4 S. J. Clarke, B. P. Guinot, C. W. Michie, M. J. C. Calmont
and M. J. Rosseinsky, Chem. Mater., 2002, 14, 288–294.
5 A. B. Jorge, J. Oró-Solé, A. M. Bea, N. Mufti, T. T. M. Palstra,
J. A. Rodgers, J. P. Attfield and A. Fuertes, J. Am. Chem. Soc.,
2008, 130, 12572–12573.
6 D. Logvinovich, R. Aguiar, R. Robert, M. Trottmann,
S. G. Ebbinghaus, A. Reller and A. Weidenkaff,J. Solid State
Chem., 2007, 180, 2649–2654.
7 P. Maillard, F. Tessier, E. Orhan, F. Chevire and
R. Marchand, Chem. Mater., 2005, 17, 152–156.
8 F. Chevire, F. Tessier and R. Marchand, Mater. Res. Bull.,
2004, 39, 1091–1101.
9 W. J. Li, E. Ionescu, R. Riedel and A. Gurlo, J. Mater. Chem.
A, 2013, 1, 12239–12245.
10 Y. I. Kim, P. M. Woodward, K. Z. Baba-Kishi and C. W. Tai,
Chem. Mater., 2004, 16, 1267–1276.
11 D. Logvinovich, M. H. Aguirre, J. Hejtmanek, R. Aguiar,
S. G. Ebbinghaus, A. Reller and A. Weidenkaff,J. Solid State
Chem., 2008, 181, 2243–2249.
12 G. Liu, X. H. Zhao and H. A. Eick, J. Alloys Compd., 1992,
187, 145–156.
13 D. Logvinovich, J. Hejtmanek, K. Knizek, M. Marysko,
N. Homazava, P. Tomes, R. Aguiar, S. G. Ebbinghaus,
A. Reller and A. Weidenkaff,J. Appl. Phys.,2009,105, 023522.
14 R. M. Po Antoine, Y. Lament, C. Michel and B. Raveau,
Mater. Res. Bull., 1988, 23, 953–957.
15 A. Navrotsky, Phys. Chem. Miner., 1977, 2,89–104.
16 J. M. McHale, G. R. Kowach, A. Navrotsky and F. J. DiSalvo,
Chem. –Eur. J., 1996, 2, 1514–1517.
Table 4 The tolerance factors for ABO
3
and ABO
2
N calculated with eqn (12) and (13), respectively
a
Oxide BaMoO
3
BaWO
3
SrMoO
3
SrWO
3
CaMoO
3
CaWO
3
t
O
1.03 1.027 0.98 0.975 0.945 0.941
t
oxy
1.053 1.048 0.995 0.989 0.959 0.955
Oxynitride BaMoO
2
N BaWO
2
N SrMoO
2
N SrWO
2
N CaMoO
2
N CaWO
2
N
Predicted
b,9
NNPPP P
Experiment
c
NNPPP
11
N
a
P: perovskite; N: non-perovskite.
b
The predicted formability of oxynitrides.
c
The formability of oxynitrides for this work.
Dalton Transactions Paper
This journal is © The Royal Society of Chemistry 2015 Dalton Trans.,2015,44, 8238–8246 | 8245
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
17 A. Navrotsky, Phys. Chem. Miner., 1997, 24, 222–241.
18 A. Navrotsky, J. Alloys Compd., 2001, 321, 300–306.
19 S. H. Elder, F. J. Disalvo, L. Topor and A. Navrotsky, Chem.
Mater., 1993, 5, 1545–1553.
20 M. R. Ranade, F. Tessier, A. Navrotsky, V. J. Leppert,
S. H. Risbud, F. J. DiSalvo and C. M. Balkas, J. Phys. Chem.
B, 2000, 104, 4060–4063.
21 M. R. Ranade, F. Tessier, A. Navrotsky and R. Marchand,
J. Mater. Res., 2001, 16, 2824–2831.
22 J. J. Liang, A. Navrotsky, V. J. Leppert, M. J. Paskowitz,
S. H. Risbud, T. Ludwig, H. J. Seifert, F. Aldinger and
M. Mitomo, J. Mater. Res., 1999, 14, 4630–4636.
23 F. Tessier and A. Navrotsky, Chem. Mater., 2000, 12,
148–154.
24 I. Molodetsky, A. Navrotsky, F. DiSalvo and M. Lerch,
J. Mater. Res., 2000, 15, 2558–2570.
25 P. Fischer, G. Frey, M. Koch, M. Konnecke, V. Pomjakushin,
J. Schefer, R. Thut, N. Schlumpf, R. Burge, U. Greuter,
S. Bondt and E. Berruyer, Physica B, 2000, 276, 146–147.
26 D. M. Tobbens, N. Stusser, K. Knorr, H. M. Mayer and
G. Lampert, Epdic 7: European Powder Diffraction, Pts 1 and
2, 2001, 378–3, 288–293.
27 P. Thompson, D. E. Cox and J. B. Hastings, J. Appl. Crystal-
logr., 1987, 20,79–83.
28 J. Rodriguezcarvajal, Physica B, 1993, 192,55–69.
29 Z. G. Pinsker, Acta Crystallogr., 1957, 10, 775–775.
30 P. S. Herle, M. S. Hegde and G. N. Subbanna, J. Mater.
Chem., 1997, 7, 2121–2125.
31 V. V. Klechkovskaya, N. V. Troitskaya and Z. G. Pinsker, Sov.
Phys. Crystallogr. (Engl. Transl.), 1965, 10,28–35.
32 R. Scholder and L. Brixner, Z. Naturforsch., B: Chem. Sci.,
1955, 10, 178–179.
33 L. H. Brixner, J. Inorg. Nucl. Chem., 1960, 14, 225–230.
34 M. T. Weller and S. J. Skinner, Int. J. Inorg. Mater., 2000, 2,
463–467.
35 A. P. A. Marques, M. T. S. Tanaka, E. Longo, E. R. Leite and
I. L. V. Rosa, J. Fluoresc., 2011, 21, 893–899.
36 M. Yang, J. Oro-Sole, A. Kusmartseva, A. Fuertes and
J. P. Attfield, J. Am. Chem. Soc., 2010, 132, 4822–4829.
37 A. Kusmartseva, M. Yang, J. Oró-Solé, A. M. Bea, A. Fuertes
and J. P. Attfield, Appl. Phys. Lett., 2009, 95, 02210.
38 J. B. Varley, A. Janotti and C. G. Van de Walle, Adv. Mater.,
2011, 23, 2343–2347.
39 A. K. Rumaiz, J. C. Woicik, E. Cockayne, H. Y. Lin,
G. H. Jaffari and S. I. Shah, Appl. Phys. Lett., 2009, 95.
40 M. Yang, J. H. Bae, C. W. Yang, A. Benayad and H. Baik,
J. Anal. Atom. Spectrom., 2013, 28, 482–487.
41 J. Oró-Solé, L. Clark, W. Bonin, J. P. Attfield and A. Fuertes,
Chem. Commun., 2013, 49, 2430–2432.
42 D. Logvinovich, S. C. Ebbinghaus, A. Reller, I. Marozau,
D. Ferri and A. Weidenkaff,Z. Anorg. Allg. Chem., 2010, 636,
905–912.
43 Y. Masatomo, F. Uhi, N. Hiromi, O. Kazuki and
R. H. James, J. Phys. Chem. C, 2013, 117, 18529–18539.
44 V. M. Goldschmidt, Skrifer Norske Videnskaps-Akad,I.Mat.
Naturv., Oslo, 1926.
45 R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect. B:
Struct. Crystallogr. Cryst. Chem., 1969, 25, 925–946.
46 K. Kamata, T. Nakamura and T. Sata, Chem. Lett., 1975,
81–86.
47 K. Kamata, T. Nakamura and T. Sata, Mater. Res. Bull.,
1975, 10, 373–378.
48 R. G. Pearson, Proc. Natl. Acad. Sci. U. S. A., 1975, 72,
2104–2106.
Paper Dalton Transactions
8246 |Dalton Trans.,2015,44, 8238–8246 This journal is © The Royal Society of Chemistry 2015
Published on 30 March 2015. Downloaded by TU Berlin - Universitaetsbibl on 12/06/2017 13:58:45.
View Article Online
