New ternary and quaternary barium nitride halides; Synthesis and crystal chemistry

Article abstract of DOI:10.1021/ic201264u

New ternary and quaternary nitride halides, Ba₂N(X,X′) (X = F, Cl, Br; X′ = Br, I), have been synthesized from the high temperature reactions of barium subnitride with the respective barium halides under an inert atmosphere. The former include the first fully characterized barium nitride halides for X other than F, and the latter are the first examples of barium nitride mixed halides. The variation in structure with composition has been investigated by powder X-ray and powder neutron diffraction techniques. The heavier ternary and quaternary nitride halides (X, X′ = Cl, Br, I) crystallize in the hexagonal space group R3m, with the anti-α-NaFeO₂ structure. Ba₂NF forms with both an anti-α-NaFeO₂ structure, in which N^3- and F ^- are ordered and an anion-disordered simple rock salt structure. The hexagonal polymorph of Ba₂NF is the only example to date of a nitride fluoride adopting this layered structure. Both the ternary and the quaternary compounds display very weak, temperature independent paramagnetism.

Full text of DOI:10.1021/ic201264u

ARTICLE
pubs.acs.org/IC
New Ternary and Quaternary Barium Nitride Halides; Synthesis and
Crystal Chemistry
Andrew S. Bailey,†,^ Robert W. Hughes,‡ Peter Hubberstey,† Clemens Ritter,§ Ronald I. Smith,|| and
Duncan H. Gregory*,‡
†School of Chemistry, University of Nottingham, Nottingham NG7 2RD, United Kingdom
‡WestCHEM, School of Chemistry, University of Glasgow, Glasgow G12 8QQ, United Kingdom
§Institut Laue-Langevin, 6, rue Jules Horowitz, BP 156ꢀ38042, Grenoble, Cedex 9, France
ISIS Pulsed Neutron and Muon Source, Science and Technology Facilities Council, Rutherford Appleton Laboratory, Didcot,
OX11 0QX, United Kingdom
S Supporting Information
b
ABSTRACT: New ternary and quaternary nitride halides, Ba2N(X,X0)
(X = F, Cl, Br; X0 = Br, I), have been synthesized from the high
temperature reactions of barium subnitride with the respective barium
halides under an inert atmosphere. The former include the first fully
characterized barium nitride halides for X other than F, and the latter are
the first examples of barium nitride mixed halides. The variation in
structure with composition has been investigated by powder X-ray and
powder neutron diffraction techniques. The heavier ternary and qua-
ternary nitride halides (X, X0 = Cl, Br, I) crystallize in the hexagonal
space group R3m, with the anti-R-NaFeO2 structure. Ba2NF forms with
both an anti-R-NaFeO2 structure, in which N3ꢀ and Fꢀ are ordered and
an anion-disordered simple rock salt structure. The hexagonal poly-
morph of Ba2NF is the only example to date of a nitride fluoride
adopting this layered structure. Both the ternary and the quaternary compounds display very weak, temperature independent
paramagnetism.
’ INTRODUCTION
in some systems, different structural polymorphs have been
demonstrated to exist, for example, Ca2NF has been reported
variously to form with a simple cubic rock salt,17 a doubled cubic
rock salt,23 or an ordered rock salt-derived tetragonal cell.20
Moreover, variation in structure, stoichiometry, and anion dis-
tribution is likely to have profound effects on electronic proper-
ties. For example, in the MgꢀNꢀF system, one moves from
insulating MgF2 (calculated direct band gap, Eg, of 6.8 eV)
through Mg3NF3 (Eg = 3.6 eV) and LꢀMg2NF (Eg = 2.1 eV)
to semiconducting Mg3N2 with a calculated direct band gap of
1.6 eV (2.8 eV experimentally).24
The field of inorganic nitride chemistry has shown consider-
able growth over the past two decades.1ꢀ3 It is perhaps un-
expected that some of the more unusual and interesting nitrides
are formed by the alkali and alkaline earth metals.4,5 The group 2
metals (A) form nitrides with AꢀN bonding and structures that
change dramatically as one descends the group; the lighter metals
(Be to Ca) form ionic, insulating or semiconducting, salt-like
compounds6ꢀ8 whereas the heavier members of the group (Ca to Ba)
form subnitrides with low-dimensional structures and metallic
properties.9ꢀ11
In comparison to the other group 2 metals, nitride halides of
barium are rather less explored. Ehrlich et al. suggested the
existence of the compositions Ba2NCl, Ba2NBr, and Ba6NI9 from
the reaction of “Ba3N2” with BaX2.13 PXD produced patterns
that could not be matched to the starting materials, but no
structures were published. A simple rock salt structure for Ba2NF
had been described by Ehrlich et al.17 which has recently been
supported by a subambient temperature single crystal structure
determination by Seibel and Wagner.22
One of the simplest families of compounds produced experi-
mentally from the nitrides are the nitride halides A2NX (X =
FꢀI), and indeed some members have been known for
decades.12ꢀ17 More recently the structures of some of these
nitride halide materials have been examined in detail through
single-crystal diffraction techniques.18ꢀ23 Recent work has sug-
gested that the ternary nitride halide phase systems are more
complex than originally envisaged and that the structures and
anion distributions are sensitive to both the halide (X) and the
conditions of synthesis.20,21 The types of structure seen in the
nitride halides vary from simple rock salt analogues (for example,
Sr2NF)17 to layered materials (for example, Ca2NCl).18 Indeed
Received: June 12, 2011
Published: August 26, 2011
r
2011 American Chemical Society
9545
dx.doi.org/10.1021/ic201264u Inorg. Chem. 2011, 50, 9545–9553
|

Inorganic Chemistry
ARTICLE
Table 1. Ba2N(XX0) Samples Synthesized
before being slow cooled at 20 Khꢀ1. The crucibles were then opened in
a nitrogen filled glovebox with pipe cutters.
sample no.
reactants; ratio
nominal stoichiometry
Synthesis of Ba2NX1ꢀyX0y (X,X0 = Cl, Br, I; 0 < y < 1). All
manipulations were carried out in an argon-filled glovebox unless stated
otherwise. Quaternary barium nitride mixed halides were prepared
according to the following scheme:
1
2
3
4
5
6
7
8
Ba2N: BaF2; 2:1
Ba2NF
Ba2N: BaCl2; 2:1
Ba2NCl
Ba2N: BaCl2: BaBr2; 10:4:1
Ba2N: BaCl2: BaBr2; 10:3:2
Ba2N: BaCl2: BaBr2; 10:2:3
Ba2N: BaCl2: BaBr2; 10:1:4
Ba2N: BaBr2; 2:1
Ba2NCl0.8Br0.2
Ba2NCl0.6Br0.4
Ba2NCl0.4Br0.6
Ba2NCl0.2Br0.8
Ba2NBr
2Ba2N þ ð1 ꢀ yÞBaX2 þ yBaX02 f 2Ba2NX1ꢀyX0y þ Ba
ð2Þ
Samples were synthesized following a similar procedure to that
employed for the ternary compounds above with pelletized powders
sealed in Mo foil lined steel crucibles. All compounds were fired for 5
days at 998 K and slow cooled at 20 Khꢀ1. All crucibles were opened in a
nitrogen filled glovebox. The reagents, starting ratios and target products
(1ꢀ8) are summarized in Table 1.
Ba2N: BaBr2: BaI2; 10:4:1
Ba2NBr0.8I0.2
In our previous studies we examined the synthesis of ternary,
quaternary, and quintinary alkaline earth nitride halides of
calcium and strontium, and the effect that varying the halide
has on the structure.25ꢀ28 In this work we have probed the
existence, structure, and stoichiometry of compounds in the
bariumꢀnitrogenꢀhalogen systems and have successfully
synthesized a tranche of new compounds containing halides
varying from Fꢀ through Clꢀ and Brꢀ to Iꢀ. Powder X-ray and
neutron diffraction measurements have enabled us to determine
definitive structural models for a new hexagonal polymorph of
Ba2NF, the ternary compounds Ba2NCl and Ba2NBr, and a series
of new quaternary mixed halides, Ba2NX(1ꢀy)X0y (X = Cl, Br, I).
Powder X-ray Diffraction. The barium subnitride starting ma-
terial, ternary and quaternary products were initially characterized by
powder X-ray diffraction (PXD). Data were collected using a Philips
X-pert diffractometer operating with copper KR radiation. Scans were
carried out in BraggꢀBrentano geometry from 5ꢀ80ꢀ in 2θ with a step
size of 0.02ꢀ over 50 min. Additionally for 7 and 8, further scans were
carried out from 5ꢀ120ꢀ in 2θ with a step size of 0.02ꢀ over 14 h to
provide data suitable for Rietveld refinement. Because of the air-sensitive
nature of the products, a dedicated airtight aluminum sample holder with
Mylar windows was employed.29 Data were analyzed by Philips Auto-
mated Powder Diffraction software. Phase identification was carried out
using PhilipsPC-IDENTIFY to accessthe ICDD PDF database. Unit cells
were indexed and lattice parameters refined using DICVOL91.30
Powder Neutron Diffraction. Constant wavelength (CW) pow-
der neutron diffraction (PND) data for Ba2NF (1) were collected on the
D1A instrument at the Institut Laue-Langevin, Grenoble, France. A
sample of approximately 2 g was sealed inside a 10 mm diameter
vanadium sample can using a gold seal. A germanium monochromator
was used to select an incident wavelength of 1.909 Å.
’ EXPERIMENTAL SECTION
Caution! Ba2N and all derived nitride halide materials are extremely
air and moisture sensitive. Handling must be carried out in an argon
atmosphere, which presents difficulties in characterization. Even with
the use of a dedicated airtight powder X-ray diffraction (PXD) sample
holder,29 some sample hydrolysis was observed over longer scans.
Synthesis of Ba2N. Ba2N was prepared by reaction of barium with
nitrogen using liquid sodium as a solvent medium. All manipulations
were carried out in an inert atmosphere. In an argon filled glovebox a
piece of barium (approximately 15 g, Alfa, 99+ %) was cut from a larger
ingot, and the covering oxide layer was removed with a file. The clean
metal was submerged in molten sodium (Riedel-de Ha€en, >99%)
contained within a stainless steel crucible. The crucible was then sealed
inside a stainless steel reaction vessel fitted with a coldfinger and a gas-
line connection and removed from the glovebox. The vessel was
evacuated and filled with nitrogen gas before being heated for 48 h at
973 K. Upon cooling the vessel was placed under a vacuum of 10ꢀ4 Torr
and heated for 24 h at 723 K to remove the sodium. Once cooled to
room temperature the vessel was opened in an argon filled glovebox. The
product was a black crystalline solid deposited within and around the
stainless steel crucible. The product was ground to give a fine dark brown
powder, and PXD was used to confirm phase purity by comparison to
ICDD card 79-4939.
Time of flight (ToF) PND data for Ba2NCl (2) and Ba2NCl1ꢀyBry
(y = 0.2, 0.4, 0.6, 0.8) (3ꢀ6) were collected using the medium resolution,
high intensity POLARIS diffractometer at the ISIS facility, Rutherford
Appleton Laboratory, U.K. For each sample, 1ꢀ3 g of material was
sealed in an electron beam welded vanadium can, made airtight using an
indium wire gasket. All diffraction experiments were performed at 298 K
with collection times of 1ꢀ2 h per sample. Diffraction data were
3
collected using the He tube low angle and backscattering detector
banks at Æ2θæ = 35ꢀ and Æ2θæ = 145ꢀ and the ZnS scintillator detector
bank at Æ2θæ = 90ꢀ.
Rietveld Refinement. Rietveld refinement against all types of
diffraction data was performed using the General Structure Analysis
System (GSAS) via the windows-based EXPGUI interface.31,32 The
major phase in each sample (1ꢀ8) was fitted using the anti-R-NaFeO2
structure as an initial starting model. Lattice parameters obtained from
indexing PXD data were used for the initial values in the refinement.
Refinements followed a similar strategy for 1ꢀ8, although the back-
ground and peak shape functions selected were instrument-dependent.
For 1 collected on D1A, the first parameters refined were the back-
ground coefficients (GSAS Function 1, a Chebyschev polynomial) and
scale factor; this was followed by refinement of the unit cell parameters.
The atomic parameters, peak profile parameters, and isotropic tempera-
ture factors were varied subsequently. Modeling of the peak shapes was
carried out using CW function 2, the ThompsonꢀCoxꢀHastings
pseudo-Voigt function. Additional phases were added once refinement
of the main phases was almost complete, and refined using the same
approach. The anisotropic temperature factors of the main phase were
one of the last variables to be refined, while in the final refinement cycles
all parameters were simultaneously varied.
Synthesis of Ba2NX (X = F, Cl, Br). All manipulations were
carried out in an argon-filled glovebox unless stated otherwise. Ternary
barium nitride halides were prepared by the reaction of barium nitride,
Ba2N, with the relevant barium halide, BaX2 (X = F, Cl, Br):
2Ba2N þ BaX2 f 2Ba2NX þ Ba
ð1Þ
The barium halides (X = F (Aldrich, 99.9%), Cl (Aldrich, 99.9%), Br
(Strem, 99%), I (Alfa, 99%)) were dried prior to use. Each was heated
under a dynamic vacuum (10ꢀ4 atm) for 24 h at 423 K.
Powders of Ba2N and BaX2 were intimately mixed and pelletized. The
pellets were wrapped in a molybdenum foil liner before being welded
inside a stainless steel crucible. The crucible was then fired for 5 days at
998 K under an argon atmosphere (to prevent crucible oxidation),
The POLARIS ToF PND data taken from the three detector banks at
low angle Æ2θæ = 35ꢀ, Æ2θæ = 90ꢀ and backscattering angles at Æ2θæ = 145ꢀ
9546
dx.doi.org/10.1021/ic201264u |Inorg. Chem. 2011, 50, 9545–9553

Inorganic Chemistry
ARTICLE
Figure 1. Structural representation of the A2NX structure (anti-R-NaFeO2 type). The [A2N]+ layers are composed of edge sharing NA6 octahedra.
Halide ions, Xꢀ, reside between the positively charged layers.
were used in the refinements of 2ꢀ 6. The background was fitted using
the exponential expansion function (Function 4 in GSAS), and the peak
shape was modeled with ToF Function 3, a convolution of back-to-back
exponentials with a pseudo-Voigt function with Lorentzian broadening.
For 7 and 8 the refinements were carried out against PXD data. The
background was modeled with GSAS Function 1, a Chebyschev poly-
nomial, and the peakshape with the ThompsonꢀCoxꢀHastings pseu-
do-Voigt function (Function 2). A March Dollase function was used to
model (00l) preferred orientation arising from the flat plate experiments.
Magnetic Measurements. Variable temperature magnetic sus-
ceptibility measurements were performed for 1ꢀ8 (ca. 10ꢀ25 mg for
each sample) using a Quantum Design MPMS-XL 5T SQUID magnet-
ometer. All samples were loaded in a nitrogen-filled, recirculating
glovebox. Data were collected between 5ꢀ300 K under fields of 1000
Oe with points at 1 K intervals from 5ꢀ30 and 5 K intervals between
30ꢀ300 K. Data were corrected for core diamagnetism and the
diamagnetic contribution of the sample holders (gelatin capsules).
hexagonal cell (a = 4.00(5), c = 19.7(2) Å) in space group R3m,
which is consistent with the anti-R-NaFeO2 structure adopted by
other alkaline earth nitride halides A2NX (A= Ca, Sr. X = Cl,
Br).26,27 This structure is a filled anti-CdCl2-type formed from
[NBa2]+ slabs and halide ions, Fꢀ, which reside in the octahedral
voids between the positively charged layers. This thus creates
alternating edge sharing layers of NBa6 and FBa6 octahedra in a
cubic close packed arrangement. A representation of this struc-
ture can be seen in Figure 1, and the structure is discussed in
more detail elsewhere.26,27 This is the first incidence of a nitride
fluoride crystallizing with this structure. Previous studies of A2NF
(A = MgꢀBa) have revealed compounds which adopt cubic
(simple or ordered) rock salt structures or tetragonal (ordered,
rock salt-derived) structures only.16,17,20ꢀ23
Given the mixture of light and heavy atoms in the structure and
the similarity of form factor for isoelectronic N3ꢀ and Fꢀ,
obtaining a complete structure solution for 1 from PXD data
alone was not trivial. PND data were collected to address this
problem since there is sufficient contrast in the coherent neutron
scattering lengths to obtain a complete description of the
structure (Ba b = 5.07 fm; N b = 9.36 fm, F b = 5.65 fm).39
The refinement was performed employing the anti-R-NaFeO2
structure, as adopted by Ca(Sr)2NCl(Br) compounds, to model
the majority phase, utilizing the cell parameters from the indexed
PXD data. The refinement progressed smoothly and converged
to produce a good fit to the data. Attempts to refine occupancies
of the anion sites to investigate possible N/F disorder reduced
the quality of the fit, increased R-factors, and caused some
parameters to become unstable. The evidence suggests, there-
fore, that an ordered model is appropriate as has been seen for
other layered A2NX compounds.26,27
’ RESULTS AND DISCUSSION
Ba2N Starting Material. After distillation of the sodium from
the reaction, the Ba2N product was left as a brown/black solid
which became a dark brown fine powder upon grinding. PXD data
confirmed a pure phase product by comparison to the ICDD
database. (PXD data can be found in the Supporting Information,
Figure SI1). The cell parameters were refined to a = 4.1146(3) Å
and c = 22.603(2) Å. This is slightly larger in a than previously
reported (a = 3.99ꢀ4.05 Å) although the c parameter falls within
previously reported limits (c = 22.08ꢀ22.70 Å).33ꢀ38
Ba2NF. Reactions consistently led to powdered black/gray
products. During attempts to synthesize powder samples of
Ba2NF, PXD revealed that two phases were present in addition
to a BaF2 impurity phase. Some of the outstanding reflections
could be indexed to cubic Ba2NF (space group Fm3m),22 but
patterns also contained a significant number of additional higher
intensity peaks. It was possible to index these reflections to a
As alluded to above, the fit was improved by the addition of
two secondary phases; the first was a cubic Ba2NF phase
following the single crystal model of Seibel and Wagner,22 while
the second was BaF2. The quality of the diffraction data allowed
9547
dx.doi.org/10.1021/ic201264u |Inorg. Chem. 2011, 50, 9545–9553

Inorganic Chemistry
ARTICLE
Table 2. Refined Crystallographic Parameters for Ba2NF (1, 10) from CW PND data at 298 K
Ba2NF (1)
Ba2NF (10)
diffraction experiment, instrument
CW PND, D1A
hexagonal
R3m (No. 166)
4.0242 (2)
19.975 (1)
280.14 (2)
3
CW PND, D1A
cubic
crystal system
space group
Fm3m (No. 225)
5.7945 (6)
a/Å
c/Å
V/Å3
194.56 (7)
1
Z
M/g molꢀ1
307.659
307.659
5.252
calculated density, Fx/g cmꢀ3
z (Ba) (6c; 0,0,z)
Ue ꢁ 100/Å2a
5.471
0.2453(4)
Ba
N
F
6c (0,0,z)
1.78b
2.40b
4.43b
4a (0,0,0)
1.12 (3)
2.54 (2)
2.54 (2)
3a (0,0,1/2)
3b (0,0,0)
4b (1/2,1/2,1/2)c
4b (1/2,1/2,1/2)c
phase fraction
no. of data
no. of parameters
Rwp
76.7(1)%
16.4(1)%
1459
18
0.0460
0.0340
5.397
Rp
reduced χ2
a Ue, the equivalent thermal displacement factor, is defined as one-third of the sum of the diagonal elements of the orthogonalized Uij tensor Uij =
exp[ꢀ2π2(U11h2a*2 + U22k2b*2 + U33l2c*2 + 2U12hka*b* + 2U23klb*c* + 2U13hla*c*)]. b Displacement factors refined anisotropically. c Site occupied
statistically by 50% N and 50% F.
us to obtain an accurate structural model for the secondary
nitride fluoride phase (in addition to the Ba2NF major phase).
The anion distribution in the minority Ba2NF phase was con-
firmed as a disordered one with N and F statistically occupying
the 4b (1/2,1/2,1/2) site. The results of the refinement can be
found in Table 2. Anisotropic thermal displacement parameters
are detailed in the Supporting Information. The fit to the data is
shown in the profile plot in Figure 2.
The refined anti-R-NaFeO2 structure of the majority phase
Ba2NF has a c parameter that is notably shorter than related
compounds. The unfilled subnitride Ba2N33 has a c parameter of
22.425(4) Å, decreasing to 20.493(4) Å in the hydride Ba2NH40
and 19.975(1) Å in the fluoride reported here. This is due to the
repulsion between the [Ba2N]+ layers being reduced relative to
Ba2N by the inclusion of increasingly electronegative anions
(vs size effects). Our CW PND refinement demonstrated that our
sample contained approximately 16% by weight of cubic Ba2NF
(10). The refined cubic a parameter of the rock salt phase is larger
than that reported by Seibel and Wagner (5.6796(19) Å) but it
should be noted that our PND data were collected at 298 K, while
the previously reported single crystal X-ray data were collected at
100 K.22 Ehrlich et al. had earlier reported a cubic phase of Ba2NF
with a lattice parameter of 5.69 Å but there is no record of the
temperature at which their data were collected.17 Ba2NF might
be regarded loosely as a “pseudo-oxide” (given that the nitride
fluoride and oxide are isoelectronic) and the cubic phase is
isostructural, of course with BaO. In this respect, it is noteworthy
that the cubic lattice parameter in the nitride fluoride is larger
than that of barium oxide (5.517 Å)41 (and hence the Baꢀ(N,F)
bond longer than the equivalent BaꢀO distance).
Figure 2. Rietveld refinement profile for Ba2NF from CW PND data.
Crosses indicate the observed data, the upper continuous line shows the
calculated profile and the lower continuous line the difference profile.
Tick marks show reflection positions for hexagonal Ba2NF (lower), BaF2
(middle), and cubic Ba2NF (upper).
cubic structure reported here (4.097(1) Å) are greater than those in
the hexagonal phase (4.024(1) Å), by about 2%. The BaꢀBa
distance in the literature structure is marginally reduced because of
the lower data collection temperature.22 The cubic rock salt phase
has one anion site occupied statistically by N3ꢀ and Fꢀ and the
Baꢀ(N,F) distance (2.8973(4) Å) in the cubic phase is close to the
mean of the BaꢀFandBaꢀN values (2.859(1) Å) inthehexagonal
polytype. Interestingly and given also the unusually low coordina-
tion number of 6 for Ba in Ba2NX compounds, the BaꢀF distances
in the hexagonal Ba2NF polytype (1) are significantly longer than
the bond lengths in either ambient or high pressure forms of BaF2
(2.6831 (2) Å and 2.663 (1) Å (at 1.5 GPa) respectively).42,43
There is also an increase in the BaꢀN distances compared to those
Table 3 gives a comparison of important bond lengths between
the hexagonal and cubic polytypes of Ba2NF and the previously
reported data for the cubic variant. The BaꢀBa distances in the
9548
dx.doi.org/10.1021/ic201264u |Inorg. Chem. 2011, 50, 9545–9553

Inorganic Chemistry
Table 3. Selected Interatomic Distances and Angles in Hexagonal and Cubic Ba2NF (1, 10) from CW PND Data at 298 K
ARTICLE
Ba2NF (1)
Ba2NF (10)
Ba2NF22a
crystal system, space group
BaꢀBa/Å
hexagonal, R3m (No. 166)
3.9100(1) ꢁ 3
4.0242(2) ꢁ 6
2.8055(1) ꢁ 3
2.9128(1) ꢁ 3
91.649 (2)
cubic, Fm3m (No. 225)
cubic, Fm3m (No. 225)
BaꢀBa/Å
4.0974(4) ꢁ 12
2.8973(4) ꢁ 6
2.8973(4) ꢁ 6
90.0 ꢁ 6
4.016 (4) ꢁ 12
2.840 (2) ꢁ 6
2.840 (2) ꢁ 6
90.0 ꢁ 6
BaꢀN/Å
BaꢀX/Å
BaꢀNꢀBa/deg
BaꢀXꢀBa/deg
a Data collected at 100 K.
87.386 (3)
90.0 ꢁ 6
90.0 ꢁ 6
Table 4. Refined Crystallographic Parameters for Ba2NCl(1ꢀy)Bry (2ꢀ6) from ToF PND Data at 298 K
Ba2NCl (2)
Ba2NCl0.8Br0.2 (3)
Ba2NCl0.6Br0.4 (4)
Ba2NCl0.4Br0.6 (5)
Ba2NCl0.2Br0.8 (6)
diffraction experiment, instrument
TOF PND, POLARIS
hexagonal
R3m (No. 166)
4.1212 (1)
22.9580 (4)
337.681(7)
3
crystal system
space group
a/Å
4.1023 (1)
22.6679 (2)
329.487 (3)
3
4.1202 (1)
22.7619 (5)
334.644 (10)
3
4.1348 (1)
23.1668 (4)
343.017 (7)
3
4.1388 (1)
23.3045 (4)
345.717 (7)
3
c/Å
V/Å3
Z
M/g molꢀ1
calculated density, Fx/g cmꢀ3
324.113
4.900
333.004
4.956
341.894
350.784
5.104
359.674
5.197
5.074
z (Ba) (6c; 0,0,z)
Ue ꢁ 100/Å2a,b
0.7682 (1)
0.73 (1)
1.63 (2)
1.75 (2)
1:0
0.7686 (1)
1.13 (2)
2.23 (3)
1.67 (3)
0.801: 0.199 (7)
13035
0.7695 (1)
1.10 (3)
0.7701 (1)
0.95 (2)
1.95 (2)
1.68 (3)
0.386: 0.614 (5)
13481
0.7705 (1)
1.08 (2)
2.14 (2)
1.77 (3)
0.178: 0.822 (6)
13692
Ba (6c)
N (3a)
X (3b)
2.12 (3)
1.75 (4)
Cl: Br
0.554: 0.446 (8)
13481
no. of data
no. of parameters
Rwp
13696
25
30
27
30
30
0.0176
0.0392
1.543
0.0230
0.0273
0.0230
0.0295
Rp
0.0434
0.0574
0.0508
0.0485
reduced χ2
2.987
3.307
2.047
5.474
a Ue, the equivalent thermal displacement factor, is defined as one-third of the sum of the diagonal elements of the orthogonalized Uij tensor
Uij = exp[ꢀ2π2(U11h2a*2 + U22k2b*2 + U33l2c*2 + 2U12hka*b* + 2U23klb*c* + 2U13hla*c*)]. b Displacement factors refined anisotropically.
in the unfilled subnitride, Ba2N (in which Ba is three-coordinate to
nitrogen).33 It was our intention to probe the structure by CW PND
at variable temperature to explore the phase change behavior on
heating and cooling. Unfortunately the Ba2NF material is highly
reactive at elevated temperature and reacted vigorously with the
vanadium sample can during our variable temperature neutron
experiment, compromising the sample environment and preventing
elucidation of any phase change behavior at this time.
The range of A2NF materials in the literature form a variety of
structures, although mainly rock salt or rock salt-related in
nature. In simple rock salt structures, the N3ꢀ and Fꢀ anions
must be disordered to maintain the simple cubic cell. Ordering of
the anions leads to structures with larger cells, such as the
tetragonal cell formed by L-Mg2NF16 and Ca2NF,20 or the cubic
cell, doubled along all axes, exhibited by Sr2NF21 and Ca2NF.23
Compositional variation may also lead to ordered structures such
as the rock salt related cubic cell of Mg3NF3 where 1/4 of the
magnesium sites are empty in an ordered fashion.16 The synth-
esis regime almost certainly plays a role in determining the
structure formed. Heat and pressure (1100 ꢀC, 25 kbar) trans-
form the tetragonal, ordered LꢀMg2NF into the disordered
cubic HꢀMg2NF.16 By analogy in oxide rock salt materials,
cation disorder is seen to increase with heating. For example, in
oxide series such as Li2(Ti1ꢀxZrx)O3 it is possible to quench the
disordered structures from high temperature for compounds of
certain values of x.44 In nitride fluoride systems, ordered struc-
tures are typically seen in single crystal samples which have
been cooled slowly from the melt, favoring ordering of the
anions.20,21,23 By contrast, simple rock salt phases were originally
reported for powders prepared by solid state methods, although
magnesium ordered structures exist synthesized as powders and
single crystals of rock salt structured Ba2NF exhibit anion
disorder.22 The Ba2NF phase we have reported here (1; hex-
agonal R-NaFeO2 structure) is another example of an ordered
structure formed on slow cooling. Uniquely among alkaline earth
nitride fluorides to date, hexagonal Ba2NF is also a phase in
competition with a disordered cubic rock salt (10). Later below,
we consider the possible effects of cation and anion composition
on structural stability in group 2 metal nitride halides, but
systematic probing of the phase space at and away from thermal
equilibrium would clarify the relationships between the different
ordered and disordered phases observed for A2NF materials.
9549
dx.doi.org/10.1021/ic201264u |Inorg. Chem. 2011, 50, 9545–9553

Inorganic Chemistry
ARTICLE
Table 5. Selected Interatomic Distances and Angles for Ba2NCl(1ꢀy)Bry (2ꢀ6) from ToF PND Data at 298 K
2
3
4
5
6
crystal system space group
3 ꢁ BaꢀBa/Å
hexagonal, R3m (No. 166)
3.77639 (4)
3.77962 (2)
4.10225 (2)
2.78900 (1)
3.29808 (2)
94.688 (1)
3.79477 (4)
4.12023 (3)
2.80074 (2)
3.31958 (3)
94.709 (1)
3.77708 (3)
4.13484 (3)
2.80015 (2)
3.38344 (2)
95.178 (1)
3.78009 (4)
4.13881 (4)
2.80262 (2)
3.40062 (3)
95.187 (1)
6 ꢁ BaꢀBa/Å
4.12118 (4)
3 ꢁ BaꢀN/Å
2.79487 (2)
3 ꢁ BaꢀX/Å
3.35131 (3)
BaꢀNꢀBa/deg
95.011 (1)
Table 6. Refined Crystallographic Parameters for Ba2NBry
I(1ꢀy) from PXD Data at 298 K
Ba2NBr (7)
Ba2NBr0.8I0.2 (8)
diffraction experiment, instrument
CW XRD, Xpert CW XRD, Xpert
hexagonal hexagonal
R3m (No. 166) R3m (No. 166)
crystal system
space group
a/Å
4.1534 (1)
23.348 (1)
352.50 (12)
3
4.1578 (2)
23.538 (1)
357.74 (3)
3
c/Å
V/Å3
Z
M/g molꢀ1
calculated density, Fx/g cmꢀ3
368.565
5.291
377.965
5.256
z (6c; 0,0,z)
Uiso ꢁ 100/Å2
0.7698 (4)
3.09 (2)
5.42 (2)
3.81 (4)
1: 0
0.7710 (1)
3.63 (2)
8.27 (1)
3.35 (3)
0.812: 0.188 (9)
5750
Ba (6c)
N (3a)
X (3b)
Figure 3. Rietveld refinement profile for Ba2NCl from ToF PND data
(Æ2θæ = 145ꢀ detector bank). Crosses indicate the observed data, the
upper continuous line shows the calculated profile, and the lower
continuous line the difference profile. Tick marks show reflection
positions for Ba2NCl (lower) and BaO (upper).
Br: I
no. of data
no. of parameters
Rwp
5750
14
23
Ba2N(X,X0) (X, X0 = Cl, Br, I). The series Ba2NCl(1ꢀy)Bry
(2ꢀ6) produced powdered products that were varying shades
of purple reaching a brown coloration by y = 1, and all appeared
from initial inspection of the PXD data to be single phase anti-R-
NaFeO2-type structures. Indexing the data showed an increase in
both a and c parameters with increasing y commensurate with the
progressive introduction of the larger halide anion into the
system (ionic radii of 1.82 Å for Brꢀ vs 1.67 Å for Clꢀ).45 Initial
short scans of the ternary nitride bromide, Ba2NBr (7) and the
brown quaternary bromide iodide, Ba2NBr0.8I0.2 (8), showed
reflections corresponding to anti-R-NaFeO2 type structures to
be present, and data indicated that both samples were single
phase by PXD. It is notable, however, that when reactions with
higher ratios of BaI2/BaBr2 were attempted (i.e., y > 0.2 in eq 2),
the diffraction data became progressively poorer and additional
unidentified reflections grew in intensity with increasing y. The
origin of these reflections is under current investigation. By y = 1
there were no peaks that could be assigned to an anti-R-NaFeO2
type structure. No convincing match was made when comparing
reflections to a pattern of the Ba6NI9 phase proposed by Ehrlich
et al.13 Magnetometry measurements revealed that each of the
samples 1ꢀ8 exhibited very weak temperature independent para-
magnetic behavior (typically χM ∼ 5 ꢁ 10ꢀ6 emu molꢀ1) suggestive
of intrinsically diamagnetic materials with low levels of alkaline earth
metal impurities (as corroborated by Rietveld refinement data) and
consistent with other alkaline earth nitride halides.26ꢀ28
0.1497
0.1153
1.526
0.1347
Rp
0.1022
reduced χ2
1.667
The inclusion of minor impurity phases improved the fits significantly,
and many of these phases are not unexpected given that barium is
expected to be generated in the reactions from eq 3 . The impurity
phases were found to be barium metal (3ꢀ 6), barium oxide (2ꢀ 6),
and barium oxide chloride, Ba4Cl6O (3 and 6) respectively.
The final refined crystallographic parameters for 2ꢀ6 are pre-
sented in Table 4 and selected interatomic distances and angles in
Table 5. Refined anisotropic thermal displacement parameters
for 2ꢀ6 are listed in the Supporting Information. An exemplar
observed, calculated, and difference profile plot for Ba2NCl (2) is
shown in Figure 3 (profile plots for the other nitride halides, 3ꢀ6
are included in the Supporting Information).
In previous PND experiments on Ca(Sr)2N(X,X0) compounds
we found no ordering of halides, with the anions statistically
occupying the interlayer 3b (0, 0, 0) site.26ꢀ28 A similar situation is
seen in these ToF PND experiments for 2ꢀ6. There is sufficient
contrast in the neutron scattering lengths of chlorine and bromine
(9.58 and 6.80 fm respectively) that any ordering should be
apparent in the diffraction data, and we find no evidence to
suggest that a model in which Clꢀ and Brꢀ are completely
disordered across the 3b site is not the correct one. Likewise we
see no evidence for disorder across the 3b and 3a sites, and so we
can conclude that just as in calcium and strontium nitride chlorides
and bromides, N3ꢀ uniquely occupies the 3a site.
ToF PND data were collected for 2ꢀ6 and refinements per-
formed using anti-R-NaFeO type A2NX structures as starting
2
models. All refinements progressed smoothly to convergence.
9550
dx.doi.org/10.1021/ic201264u |Inorg. Chem. 2011, 50, 9545–9553

Inorganic Chemistry
ARTICLE
Table 7. Selected Interatomic Distances and Angles for
Ba2NBryI(1ꢀy) from PXD Data at 298 K
Table 8. Layer Thickness (t), Interlayer Distances (d), Mean
X Anionic Radius (Xav), and Mean X Pauling Electronegativity
(χP) for Ba2N(X,X0) Nitride Halides (1ꢀ8)
7
8
compound
c/a
t/Å
d/Å t/d Xav radius/Å χP (Xav)
crystal system space group
3 ꢁ BaꢀBa/Å
hexagonal, R3m (No. 166)
Ba2NF (1)
Ba2NCl (2)
4.961 3.142 3.518 0.89
5.526 2.951 4.604 0.64
1.19
1.67
1.70
1.73
1.76
1.79
1.82
1.87
3.98
3.16
3.12
3.08
3.04
3
3.8011 (3)
4.1489 (4)
2.8134 (2)
3.3990 (2)
95.011 (4)
3.8115 (1)
4.1790 (2)
2.8281 (1)
3.4506 (1)
95.267 (2)
6 ꢁ BaꢀBa/Å
Ba2NCl0.8Br0.2 (3) 5.524 2.945 4.641 0.63
Ba2NCl0.6Br0.4 (4) 5.571 2.932 4.722 0.62
Ba2NCl0.4Br0.6 (5) 5.603 2.931 4.793 0.61
Ba2NCl0.2Br0.8 (6) 5.631 2.929 4.840 0.61
3 ꢁ BaꢀN/Å
3 ꢁ BaꢀX/Å
BaꢀNꢀBa/deg
Ba2NBr (7)
Ba2NBr0.8 0.2
5.621 2.939 4.831 0.61
5.661 2.935 4.912 0.60
2.96
2.9
I
(8)
Figure 4. Rietveld refinement profile for Ba2NBr0.8 0.2
I
from PXD data.
Crosses indicate the observed data, the upper continuous line shows the
calculated profile, and the lower continuous line the difference profile.
Tick marks show reflection positions for Ba2NBr0.8 0.2
I
(lower) and the
aluminum sample holder (upper).
Rietveld refinements performed using PXD data collected
over longer periods for 7 and 8 converged smoothly to yield good
fits to the data. Peaks unaccounted for by the model could be
attributed to aluminum reflections from the flat plate sample
holder. The refinement results for 7 and 8 can be found in
Table 6 and selected interatomic distances and angles in Table 7.
The Rietveld refinement profile for 8 is shown in Figure 4. As was
observed from indexing the initial X-ray data, lattice parameters
increased universally with increasing bromide content as would
be expected as one replaces anions with those of a larger effective
ionic radius. The increase is more marked in the c parameter, with
an increase across the series 2ꢀ7 of 3%, compared to a 1.2%
increase in a. This is not unexpected, as the anions are inserted
between the [Ba2N]+ layers, and one would expect anisotropic
expansion. Indeed a similar pattern of increased relative expansion
of c compared with a is observed in the analogous Ca2NCl(1ꢀy)Bry
and Sr2NCl(1ꢀy)Bry (0 e y e 1) series.26,27 With the introduc-
tion of iodide, the increasing size of the halide anion again results
in larger lattice parameters, with a slightly more marked relative
increase in c than a. Between 7 and 8 there is a 0.45% increase in
c compared with a 0.1% increase in a.
Figure 5. Plot of average X site Pauling electronegativity (χp) against
interlayer separation (d). Squares represent the Ba2N(XX0) series
(1ꢀ8). Triangles represent a series of Sr2N(XX0) values obtained from
ref 27. Circles show a series of Ca2N(XX0) values from ref 26. Lines show
best linear fits to the data series.
Equally the changes in BaꢀN bond lengths are quite small across
the whole series, from 2.8790(1) in 2 through to 2.8281(1) Å in
8. The corresponding length in the unfilled nitride is 2.7675 Å.
More substantial changes occur in the BaꢀX distances as might
be expected on intercalating larger halide ions between the layers.
The BaꢀCl distance in 2 is 3.2981(1) Å. The binary chloride
BaCl2 has two chlorine sites and a total of nine BaꢀCl
distances.46 Seven of these are shorter (2.86(8)ꢀ3.25(8) Å),
with two longer out of plane distances (3.58(8) Å). Therefore,
the bond distance in our layered material lies between these
shorter and longer values from the binary chloride. BaBr2 and
BaI2 are isostructural withthe chloride, and a similar pattern is seen
with the BaꢀX distance in the layered nitride halide having a value
between the shorter and longer BaꢀX value in the corresponding
binary material.
Considering the important interatomic distances and angles
for 2ꢀ8 (Tables 5 and 7), the configuration of the BaꢀN layers
in the materials does not change greatly on going across the
series. The BaꢀBa distances in the ab plane increase from
4.1023(1) to 4.1790(1) Å from 2ꢀ8 (cf. 4.0290 Å in Ba2N).
In the c direction, within the layer, the BaꢀBa distance changes
from 3.7796(1) to 3.8115(1) Å from 2ꢀ8 (3.7952 Å in Ba2N).
An alternative perspective when assessing the evolution in
structure with X (and A) in these systems is to consider the
variations in layer thickness (t) and interlayer distance (d). These
are summarized in Table 8 for all the compounds, 1ꢀ8. In
general, increasing the size of the intercalated halide decreases t
9551
dx.doi.org/10.1021/ic201264u |Inorg. Chem. 2011, 50, 9545–9553

Inorganic Chemistry
ARTICLE
material Ba2NF, this represents the first example of an alkaline
earth metal nitride fluoride crystallizing with a hexagonal layered
structure. This phase appears to form in competition with the
previously reported cubic rock salt phase of Ba2NF (which is
present in this study as a minor component). Further under-
standing of the phase relationship in this and other A2NF materials
might be forthcoming using variable temperature diffraction.
Compounds containing the heavier halides form hexagonal R-
NaFeO2-type structures by analogy to the equivalent calcium and
strontium compounds. A complete series of quaternary nitride
halides is formed between the end members Ba2NCl and Ba2NBr
with the c parameter increasing as the Brꢀ anion replaces Clꢀ.
The solubility of Iꢀ in this structure type reaches a limit close
to the composition Ba2NBr0.8I0.2. Attempts to synthesize mate-
rials containing higher proportions of Iꢀ led to the growth of an
additional phase that we are currently investigating. It is possible
that the new phase is related to the Ba6NI9 composition
proposed by Ehrlich et al. from melting point analysis of mixtures
of “Ba3N2” and BaI2, although there is no direct match of the
reflections we observe to the earlier diffraction patterns. All of the
investigated nitride halides exhibit weak, temperature-independent
magnetism.
Figure 6. Schematic overview of ternary AꢀNꢀX (A = MgꢀBa;
X = FꢀI) compounds by composition and structure type.
’ ASSOCIATED CONTENT
while increasing d. Remarkable is the value of the t/d ratio for
Ba2NF compared with other nitride halides. In the former case, the
values of t and d are quite similar reflecting the near parity in
anionic radius between N3ꢀ and Fꢀ.44 At these values of t/d (c/a),
one would thus assume, it becomes favorable for anionic disorder
and transition from the R-NaFeO2 structure to a simple rock salt
structure. The observation that all the compounds containing
heavier halides, 2ꢀ8, crystallize in the R3m space group is perhaps
not unexpected since all have a t/d ratio less than the nitride
fluorides but above 0.5, which correlates to the minimum value for
the stability of the anti-R-NaFeO2 structure.26 For t/d < 0.5, the
layered materials adopt the anti-β-RbScO2 structure (e.g., Ca2NI),
with an HCP rather than CCP arrangement of the layers.
S
Supporting Information. PXD analysis of Ba2N starting
b
material, tables of anisotropic temperature factors from powder
neutron diffraction refinements for (1ꢀ6), and additional profile
plots from the refinements. This material is available free of
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: duncan.gregory@glasgow.ac.uk.
Present Addresses
^AWE, Aldermaston, Reading, Berkshire RG7 4PR, United Kingdom.
One can also consider the value of d with respect to the Pauling
electronegativity (χp) of the halide X (X0). The average electro-
negativity of the site can be calculated for Ba2N(X,X0) as
’ ACKNOWLEDGMENT
χpaverage ¼ ðχpðXÞð1 ꢀ yÞÞ þ ðχpðX0ÞyÞ
ð3Þ
We thank the ILL, France and ISIS, RAL, U.K., for the
provision of beamtime associated with this work. D.H.G. and
P.H. thank ENEA C.R., Brasimone, for funding a studentship for
A.S.B. D.H.G. thanks WestCHEM for funding R.W.H. as a
WestCHEM fellow.
The relationship of χp against d is represented for 1ꢀ8 in
Figure 5, together with related strontium and calcium materials.
There is a linear relationship between the electronegativity and
the layer separation with a series of almost parallel lines displaced
to lower d as one ascends group 2. It is possible through
extrapolation of the linear fits to estimate values for the interlayer
spacings for hexagonal Sr2NF (∼3.03 Å) and Ca2NF (∼2.88 Å).
These distances are significantly smaller than those in any of the
hexagonal R-NaFeO2 structured nitride halides (and those in the
A2N parent binary subnitrides) and this (i.e., minimization of
[A2N]+ layer repulsion) is probably one of the driving forces
behind the favorability of the three-dimensional, simple, and
superstructured rock salt phases adopted by strontium and
calcium nitride fluorides versus the observation of both layered
and simple rock salt structures for Ba2NF (Figure 6).
’ REFERENCES
(1) DiSalvo, F. J. Science 1990, 247, 649.
(2) DiSalvo, F. J.; Clarke, S. J. Curr. Opin. Solid State Mater. Sci. 1996,1, 241.
(3) Gregory, D. H. J. Chem. Soc., Dalton Trans. 1999, 259.
(4) Simon, A. Coord. Chem. Rev. 1997, 163, 253.
(5) Gregory, D. H. Coord. Chem. Rev. 2001, 215, 301.
(6) von Stackelberg, M.; Paulus, R. Z. Phys. Chem. 1933, B22, 305.
(7) Partin, D. E.; Williams, D. J.; O’Keefe, M. J. Solid State Chem.
1997, 132, 56.
(8) Gregory, D. H.; Barker, M. G.; Edwards, P. P.; Siddons, D. J.
Inorg. Chem. 1995, 34, 5195.
(9) Gregory, D. H.; Bowman, A.; Baker, C. F.; Weston, D. P. J. Mater.
Chem. 2000, 10, 1635.
’ CONCLUSIONS
(10) Brese, N. E.; O’Keeffe, M. J. Solid State Chem. 1990, 87, 134.
(11) Reckeweg, O.; DiSalvo, F. J. Solid State Sci. 2002, 4, 575.
(12) Ehrlich, P.; Deissmann, W. Angew. Chem. 1958, 70, 656.
A series of new barium nitride halide layered materials have
been synthesized and characterized. In the case of the fluoride
9552
dx.doi.org/10.1021/ic201264u |Inorg. Chem. 2011, 50, 9545–9553

Inorganic Chemistry
ARTICLE
(13) Ehrlich, P.; Deissmann, W.; Koch, E.; Ullrich, V. Z. Anorg. Allg.
Chem. 1964, 328, 243.
(14) Emons, H.-H.; Anders, D.; Roewer, R.; Vogt, F. Z. Anorg. Allg.
Chem. 1964, 333, 99.
(15) Emons, H.-H.; Grothe, W.; Seyfarth, H.-H. Z. Anorg. Allg.
Chem. 1968, 363, 191.
(16) Andersson, S. J. Solid State Chem. 1970, 1, 306.
(17) Ehrlich, P.; Linz, W.; Seifert, H. J. Naturwissenschaften 1971, 4,
219.
(18) Hadenfeldt, C.; Herdej€urgen, H. Z. Anorg, Allg. Chem. 1987,
545, 177.
(19) Hadenfeldt, C.; Herdej€urgen, H. Z. Anorg, Allg. Chem. 1988,
558, 35.
(20) Nicklow, R. A.; Wagner, T. R.; Raymond, C. C. J. Solid State
Chem. 2001, 160, 134.
(21) Wagner, T. R. J. Solid State Chem. 2002, 169, 13.
(22) Seibel, H.; Wagner, T. R. J. Solid State Chem. 2004, 177, 2772.
(23) Jack, D. R.; Zeller, M.; Wagner, T. R. Acta Crystallogr. 2005,
C61, i6.
(24) Fang, C. M.; Ramanujachary, K. V.; Hintzen, H. T.; de With, G.
J. Alloys Compds. 2003, 351, 72.
(25) Bowman, A.; Mason, P. V.; Gregory, D. H. Chem. Commun.
2001, 351, 1650.
(26) Bowman, A.; Smith, R. I.; Gregory, D. H. J. Solid State Chem.
2005, 178, 1807.
(27) Bowman, A.; Smith, R. I.; Gregory, D. H. J. Solid State Chem.
2006, 179, 130.
(28) Brogan, M. A.; Hughes, R. W.; Smith, R. I.; Gregory, D. H.
Dalton Trans. 2010, 39, 7153.
(29) Barker, M. G.; Begley, M. J.; Edwards, P. P.; Gregory, D. H.;
Smith, S. E. J. Chem. Soc., Dalton Trans. 1996, 1.
(30) Boultif, A.; Louer, D. J. Appl. Crystallogr. 1991, 24, 987.
(31) Larson, A. C.; von Dreele, R. B. The General Structure Analysis
System; Los Alamos National Laboratories: Los Alamos, NM, 2000.
(32) Toby, B. H. J. Appl. Crystallogr. 2001, 34, 210.
(33) Reckeweg, O.; DiSalvo, F. J. Z. Kristallogr. NCS 2005, 220, 519.
(34) K€unzel, H.-T. Doctoral Thesis, Universit€at Stuttgart, Stuttgart,
Germany, 1980.
(35) Seeger, O. Doctoral Thesis, Universit€at Stuttgart, Stuttgart,
Germany, 1994.
(36) Steinbrenner, U. Doctoral Thesis, Universit€at Stuttgart, Stuttgart,
Germany 1997.
(37) Schultz-Coulon, V. Doctoral Thesis, Universit€at Bayreuth,
Bayreuth, Germany 1998.
(38) Vajenine, G. V.; Grzechnik, A.; Syassen, K.; Loa, I.; Hanfland,
M.; Simon, A. C. R. Chim. 2005, 8, 1897.
(39) Sears, V. F. Neutron News 1992, 3 (3), 26.
(40) Wegner, B.; Essman, R.; Bock, J.; Jacobs, H. Eur. J. Solid State
Inorg. Chem. 1992, 29, 1217.
(41) Gerlach, W. Z. Phys. 1922, 9, 184–192.
(42) Radtke, A.; Brown, G. Am. Mineral. 1974, 59, 885–888.
(43) Leger, J.; Haines, J.; Atouf, A.; Schulte, O.; Hull, S. Phys. Rev. B
1995, 52, 13247.
(44) Campos, A.; Quintana, P.; West, A. J. Solid State Chem. 1990,
86, 129–130.
(45) Shannon, R. D. Acta Crystallogr. 1976, A32, 751.
(46) Brackett, E.; Brackett, T.; Sass, R. J. Phys. Chem. 1963, 67,
2132–2135.
9553
dx.doi.org/10.1021/ic201264u |Inorg. Chem. 2011, 50, 9545–9553