Extreme Sensitivity of a Topochemical Reaction to Cation Substitution: SrVO2H versus SrV1- xTixO1.5H1.5

Article abstract of DOI:10.1021/acs.inorgchem.8b00026

The anion-ordered oxide-hydride SrVO₂H is an antiferromagnetic insulator due to strong correlations between vanadium d electrons. In an attempt to hole-dope SrVO₂H into a metallic state, a strategy of first preparing SrV_1-xTi_xO₃ phases and then converting them to the corresponding SrV_1-xTi_xO₂H phases via reaction with CaH₂ was followed. This revealed that the solid solution between SrVO₃ and SrTiO₃ is only stable at high temperature. In addition, reactions between SrV_0.95Ti_0.05O₃ and CaH₂ were observed to yield SrV_0.95Ti_0.05O_1.5H_1.5 not SrV_0.95Ti_0.05O₂H. This dramatic change in reactivity for a very modest change in initial chemical composition is attributed to an electronic destabilization of SrVO₂H on titanium substitution. Density functional theory calculations indicate that the presence of an anion-ordered, tetragonal SrMO₂H phase is uniquely associated with a d² electron count and that titanium substitution leads to an electronic destabilization of SrV_1-xTi_xO₂H phases, which, ultimately, drives further reaction of SrV_1-xTi_xO₂H to SrV_1-xTi_xO_1.5H_1.5. The observed sensitivity of the reaction products to the chemical composition of initial phases highlights some of the difficulties associated with electronically doping metastable materials prepared by topochemical reactions.

Full text of DOI:10.1021/acs.inorgchem.8b00026

Article
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
pubs.acs.org/IC
Extreme Sensitivity of a Topochemical Reaction to Cation
Substitution: SrVO H versus SrV Ti O H
2
1−x x 1.5 1.5
†
†
‡
†
Midori Amano Patino, Dihao Zeng, Stephen J. Blundell, John E. McGrady,
,†
and Michael A. Hayward*
†Inorganic Chemistry Laboratory, Department of Chemistry, University of Oxford, South Parks Road, Oxford OX1 3QR, United
Kingdom
‡
Clarendon Laboratory, Department of Physics, University of Oxford, Parks Road, Oxford OX1 3PU, United Kingdom
*
S Supporting Information
ABSTRACT: The anion-ordered oxide−hydride SrVO H is an anti-
2
ferromagnetic insulator due to strong correlations between vanadium d
electrons. In an attempt to hole-dope SrVO H into a metallic state, a
2
strategy of first preparing SrV1−xTi O phases and then converting them to
x
3
the corresponding SrV1−xTi O H phases via reaction with CaH was
x
2
2
followed. This revealed that the solid solution between SrVO and SrTiO3
3
is only stable at high temperature. In addition, reactions between
SrV0.95Ti0.05O and CaH were observed to yield SrV0.95Ti0.05O H not
3
2
1.5 1.5
SrV0.95Ti0.05O H. This dramatic change in reactivity for a very modest
2
change in initial chemical composition is attributed to an electronic
destabilization of SrVO H on titanium substitution. Density functional
theory calculations indicate that the presence of an anion-ordered, tetragonal SrMO H phase is uniquely associated with a d
2
2
2
electron count and that titanium substitution leads to an electronic destabilization of SrV1−xTi O H phases, which, ultimately,
x
2
drives further reaction of SrV1−xTi O H to SrV Ti O H . The observed sensitivity of the reaction products to the chemical
x
2
1−x
x
1.5 1.5
composition of initial phases highlights some of the difficulties associated with electronically doping metastable materials
prepared by topochemical reactions.
INTRODUCTION
electron−electron repulsion terms (U) tend to localize the
unpaired metal d electrons on discrete metal cation sites.
Valence-electron localization of this type is frequently observed
in transition-metal oxide phaseseven those prepared via high-
temperature routesand is particularly common in systems
containing 3d transition-metal cations in integer oxidation
■
Low-temperature, structure-conserving reactions can be used to
manipulate the anion lattices of complex transition-metal
oxides, offering routes to highly metastable phases that cannot
be readily prepared via conventional high-temperature syn-
thesis. “Topochemical” reactions of this type have been used to
generate extended solids containing arrays of transition-metal
cations in highly unusual oxidation state/coordination geom-
13
states. Itinerant electrons can, however, be generated via
chemical “doping”, which drives the transition-metal ions into
2
+
1
−3
noninteger or mixed-valent states. A classic example is the Cu
phase La CuO , which is a simple antiferromagnetic insulator in
etry combinations.
For example, by utilizing binary metal
4
2
4
hydrides such as NaH, LiH, or CaH as reducing agents, it has
proved possible to deintercalate oxide ions from transition-
2
its pure form. However, doping the system via barium-for-
2
+/3+
lanthanum substitution leads to a mixed-valent Cu
state,
metal oxide perovskites to yield phases containing square-
2+
+
+
2+
5−8
which exhibits both metallic conductivity and superconductiv-
planar Fe , Co , Ni , or Ru cations.
In addition, hydride-
1
4
3+
4+
ity. Likewise, the Mn and Mn phases, LaMnO and
3
for-oxide anion-exchange reactions can also occur, leading to
15,16
CaMnO , are both antiferromagnetic insulators,
valent Mn
but mixed-
3
the formation of transition-metal oxide−hydride phases such as
3+/4+
II
members of the La1−xCa MnO solid solution
x
3
LaSrCoO H , Sr Co O
H
, A TiO H , and the
3
0.7
3
2
4.33 0.84
3−x
y
9
−12
exhibit metal−insulator transitions and large magnetoresistive
Srn+1V O2n+1H series,
all of which contain reduced
n
n
17
ratios.
transition-metal cations in oxide−hydride coordination spheres.
The very unusual electronic structures of these metastable
solid-state compounds mean that they are of intrinsic chemical
interest, but their electronic and magnetic properties, which
arise from the cooperative interactions between the para-
magnetic centers, also offer the potential for novel materials
properties. Unfortunately, the majority of the reported anion-
vacancy-ordered reduced oxides and oxides−hydrides prove to
be antiferromagnetic insulators, simply because large on-site
The remarkable differences in physical properties of the
doped and undoped oxides provide strong motivation to
explore the effects of doping on topochemically prepared
phases. The oxide−hydride SrVO H is an ideal candidate in
2
12
this respect. As shown in Figure 1, SrVO H adopts an anion-
2
Received: January 4, 2018
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
stoichiometric ratios of SrO (prepared by heating SrCO3 under
vacuum at 1100 °C), V O (99.7%), and TiO were thoroughly mixed
2
3
2
inside an argon-filled glovebox, pressed into pellets, and transferred to
the arc-melt furnace without being exposed to air. The furnace was
purged with argon, and the pellets were melted a total of four times in
an electrical discharge arc, on a water-cooled copper hearth, with each
pellet being turned between heatings. Samples were allowed to cool for
1
5 min inside the furnace under flowing argon before immediately
being transferred to an argon-filled glovebox.
Reduction of SrV1−xTi O . Samples of SrV Ti O were reacted
x
3
1−x
x
3
with CaH , as described previously for the synthesis of SrVO H from
2
2
1
2
SrVO3. Powder samples of SrV1−xTi O were ground together with 2
x
3
mol equiv of CaH in an argon-filled glovebox and then sealed under
2
vacuum in amorphous-silica ampules. Samples were then heated at 610
°
C for multiple 2-day periods (as described in the text below) with
regrinding between heating periods. Reaction products were washed,
under an inert atmosphere, with 4 × 100 mL of a 0.1 M solution of
NH Cl in methanol, followed by 4 × 100 mL of clean methanol in
4
order to remove unreacted CaH and reaction byproducts (CaO).
2
Figure 1. Crystal structure of SrVO H. Green, gray, red, and yellow
spheres represent Sr, V, O, and H atoms, respectively.
2
Characterization. Powder X-ray diffraction data were collected
from samples sealed under argon in gastight sample holders using a
PANalytical X’Pert diffractometer incorporating an X’celerator
position-sensitive detector (monochromatic Cu Kα1 radiation).
High-resolution synchrotron powder X-ray diffraction data were
collected using instrument I11 at Diamond Light Source Ltd.
Diffraction patterns were collected using silicon-calibrated X-rays
with an approximate wavelength of 0.825 Å, from samples sealed in
0.5-mm-diameter borosilicate glass capillaries. All samples were diluted
with an equal volume of amorphous boron to limit absorption.
Rietveld profile refinements were performed using the GSAS suite of
ordered structure in which the V3+ cations reside within trans-
VO H local coordination sites, which then share corners to
4
2
form a 3D network. The ordered arrangement of oxide and
hydride anions lifts the degeneracy of the π-symmetry
vanadium d orbitals (d , d , and d ), giving a local
xz
yz
xy
2
0
0
0
2
2
2
(
dxz/yz) (d ) (d ) (d
)
configuration. Substantial on-site
band, leading to a Mott−
xy
z
x −y
repulsions split the half-filled d
xz/yz
22
1
2,18,19
programs. Magnetization data were collected using a Quantum
Hubbard insulating state.
In very recent work, we have
+
Design MPMS SQUID magnetometer. The μ SR experiments were
demonstrated that SrVO H undergoes a pressure-induced
2
carried out using the GPS instrument at the Swiss Muon Source, Paul
20
insulator-to-metal transition at ∼50 GPa. The resulting
metallic state exhibits a significant two-dimensional character
because the hydride ions block dispersion along the z axis of the
π-symmetry bands that span the Fermi level. A detailed study of
this pressure-induced metallic state is, unfortunately, exper-
imentally very challenging, and so we hope to gain an
alternative perspective on this material via hole doping with
Ti, which should partially depopulate the dxz/yz band and hence
generate a metallic state at ambient pressure. Here we describe
+
Scherrer Institut, Villigen, Switzerland. In a μ SR experiment, spin-
polarized muons were implanted in the bulk of the material, and the
time dependence of their polarization was monitored by recording the
angular distribution of the subsequent positron decay. Thermogravi-
metric measurements were performed by heating powder samples at a
−1
rate of 5 °C·min under an oxygen atmosphere, using a Mettler-
Toledo MX1 thermogravimetric microbalance. Manganometric
titrations were performed by dissolving samples in 5 M H SO and
2
4
then titrating the resulting solution against KMnO4.
Density Functional Theory (DFT) Calculations. All DFT
our attempts to hole-dope SrVO H, by first preparing samples
2
calculations were performed using the VASP software package
2
3
of titanium-substituted SrV1 Ti O and then reducing them to
−x
x
3
(VASP 5.3), with the Perdew−Burke−Ernzerhof density func-
2
4
form oxide−hydride phases. In the course of these studies, we
observed that the low-temperature anion-exchange reaction
that forms the oxide−hydride phases is extremely sensitive to
even very small changes in the composition of the “parent”
tional, and an effective Hubbard U (U ) of 2.0 eV on Ti and V
eff
centers was used to describe the effect of strong correlations. A plane-
wave cutoff of 600 eV was imposed in all cases, and the Brillouin zone
was sampled on an 8 × 8 × 7 Γ-centered grid. In order to investigate
different models of anion disorder in SrTiO H, we consider a
2
oxide. The titanium-doped analogue of SrVO H,
2
2
×
2 × 2 expansion of the primitive unit cell that contains four
SrV1−xTi O H, is only transiently stable, reacting further to
x
2
transition-metal centers (Figure 6). All the configurations were
modeled as G-type antiferromagnets. Geometry optimizations were
performed with the unit cell volume constraint released and all ions
allowed to relax.
form SrV1−xTi O H , the doped analogue of the inaccessible
x
1.5 1.5
SrVO H phase. This extreme sensitivity to chemical
1
.5 1.5
substitution has important implications for doping strategies
of other topochemically prepared materials.
RESULTS
EXPERIMENTAL SECTION
■
■
Synthesis of SrV1−xTi O . The synthesis of SrV Ti O samples
SrVO
3
−SrTiO Solubility. Powder X-ray diffraction data
3
x
3
1−x
x
3
was attempted in two different ways, a “ceramic” synthesis route and
an “arc-melt” route. The ceramic synthesis of SrV1−xTi O utilized a
collected from SrV1 Ti O3 (x = 0.05, 0.4, 0.5) samples
−x
x
x
3
prepared by the ceramic route (Figures 2 and S1) could be
indexed on the basis of two primitive cubic phases, with lattice
parameters in close agreement with the reported parameters of
modified version of the procedure previously described to prepare
21
SrVO3. Suitable ratios of SrCO (99.99%), V O (99.99%), and TiO
3
2
5
2
(
99.995%, previously dried at 900 °C) were ground together and then
21
25
SrVO (3.840 Å) and SrTiO (3.905 Å). Thus, two-phase
3
3
heated in air at 900 °C for 12 h to decompose the carbonate and
models based on the structures of SrVO and SrTiO were
3
3
prepare Sr2V2−2xTi O7−δ, as confirmed by powder X-ray diffraction.
2x
refined against the data to yield the lattice parameters and
phase fractions listed in Table 1 (top). The very close
correspondence between the lattice parameters of the two
phases and those of SrVO and SrTiO , combined with the
Samples were then heated at 1075 °C, under flowing H (100%), for
multiple periods of 12 h with intermediate grinding, until no further
changes were observed in powder X-ray diffraction patterns.
2
SrV1−xTi O3 samples were also prepared via arc-melting. Suitable
x
3
3
B
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Lattice Parameters and Phase Fractions from the
Structural Refinement of SrV1−xTi O Samples Prepared by
x
3
Ceramic and Arc-Melt Routes
Ceramic Samples
phase 1
phase 2
Ti content (%)
5
a (Å)
fraction (%)
a (Å)
fraction (%)
3.8471(1)
3.8439(2)
3.8463(1)
95.5(3)
57.2(5)
48.3(4)
3.9063(1)
3.9041(2)
3.9035(1)
4.5(3)
42.8(5)
51.7(4)
4
5
0
0
Arc-Melt Samples
phase 1
phase 2
fraction
(%)
fraction
(%)
Ti content (%)
a (Å)
a (Å)
5
0
0
3.8513(1)
3.8610(1)
3.8742(1)
3.8510(1)
3.8432(1)
97.3(1)
38.1(3)
37.9(4)
97.2(1)
94.7(2)
3.9214(1)
3.8832(1)
3.8900(1)
3.9210(1)
3.9020(1)
2.7(1)
61.9(3)
62.1(4)
2.8(1)
4
5
5
(annealed 600 °C)
(annealed 1100 °C)
5
5.3(2)
route, there is some solubility between SrVO and SrTiO
3
3
under the arc-melt conditions and that the sample of nominal
composition SrV0.95Ti0.05O (i.e., that formed from 5% Ti
3
content in the starting material) does contain some titanium
dissolved in the majority phase. For the sake of simplicity, this
majority phase will henceforth be referred to as
“
SrV0.95Ti0.05O ”, although the Ti content is lower than this
3
formula indicates. To investigate the stability of these
SrV1−xTi O solid solutions, two aliquots of the arc-melt-
x
3
prepared SrV0.95Ti0.05O3 sample were sealed in evacuated
amorphous-silica ampules, heated, one at 600 °C and the
other at 1100 °C, for 7 days, and then cooled at 2 °C·min− to
ambient temperature. Refinement of the powder X-ray
diffraction data from the annealed samples shows that the
solid solution is stable at 600 °C but not at 1100 °C, where it
reverts to a mixture of SrVO and SrTiO (as indicated by the
1
3
3
parameters in Table 1). Manganometric titrations of the
ceramic and arc-melted samples of SrV0.95Ti0.05O indicated
3
that both samples had average vanadium oxidation states of
3
.99(2)+ and 4.00(2)+, respectively, and were thus oxygen-
Figure 2. (top) Powder X-ray diffraction data collected from samples
stoichiometric, as described in detail in the Supporting
Information.
of nominal composition SrV Ti O prepared by ceramic and arc-
0.6 0.4
3
melt routes. (bottom) Observed, calculated, and difference plots from
the two-phase refinements against synchrotron powder X-ray
diffraction data of samples of nominal composition SrV0.95Ti0.05O3
prepared by ceramic and arc-melt routes.
Topochemical Reduction of “SrV0.95Ti0.05O ”. Samples
3
of “SrV0.95Ti0.05O ” prepared via both ceramic and arc-melt
routes were reacted with CaH as described above. Figure 3
3
2
(
top) shows a series of powder X-ray diffraction patterns
correspondence between the fraction of phase 2 (SrTiO ) and
collected from the products of reaction between the ceramic
3
the Ti content of the original mixture, suggests that there is
almost no solubility between SrVO and SrTiO under these
sample and CaH as a function of reaction time. These data are
2
consistent with the ceramic sample of SrV0.95Ti0.05O being a
3
3
3
“ceramic” synthesis conditions.
simple mixture of SrVO and SrTiO , which is transformed by
3
3
Powder X-ray diffraction data collected from SrV1−xTi O3
the action of CaH into a mixture of SrVO H (marked in red)
x
2
2
samples prepared by the arc-melt route are shown in Figures 2
and S1. Two primitive cubic phases are again required to model
the data, but the refined lattice parameters (listed in Table 1)
do not correspond to either SrVO or SrTiO . Moreover, unlike
and a small amount of SrTiO3−xH (black), precisely as would
x
be expected from previous studies of these perovskite phases in
1
2,26
isolation.
Synchrotron powder X-ray diffraction data
collected from the washed reaction products after 10 days of
heating confirm that these are the reaction products (Figure S2
and Table S3). Thermogravimetric reoxidation of this sample
under flowing oxygen resulted in a 14.5% gain in mass (Figure
S3), consistent with the composition detailed in Table S3 (full
details of the experiment are described in the Supporting
Information).
3
3
the phases generated by ceramic synthesis, the refined lattice
parameters are dependent on sample composition. Finally, the
fractions of the two phases do not correlate with the Ti content
of the samples as closely as those from the ceramic route. This
accumulated evidence indicates that while perfect, single-
phased solid solutions were not prepared via the arc-melt
C
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 3 also shows diffraction data collected from the
products of reaction between arc-melt-prepared SrV0.95Ti0.05O3
and CaH . The diffraction pattern taken after 4 days of heating
2
looks very similar to the analogous data from the ceramic
SrV0.95Ti0.05O sample, containing a mixture of CaO, CaH , and
3
2
unreacted cubic SrV0.95Ti0.05O along with a small amount of a
3
new tetragonal phase that we formulate as “SrV0.95Ti0.05O H”
2
(
red). The a lattice parameter of this new phase is very similar
to that of undoped SrVO H [3.929(1) vs 3.92331(3) Å], but a
2
distinct elongation along the c axis, V−H−V, is apparent
[3.738(4) vs 3.6671(3) Å]. After 8 days of heating, the peaks
due to the tetragonal phase reduce in intensity and are replaced
by a new cubic phase [a = 3.888(1) Å; green] that eventually
becomes dominant after 10 days. Synchrotron powder X-ray
diffraction data collected from a sample after 10 days of heating,
washed to remove CaH and CaO (Figure 4), can be fitted by a
2
three-phase model consisting of the dominant cubic phase
[
[
SrV0.95Ti0.05O H , 84.3(2)%], a tetragonal phase
x y
SrV0.95Ti0.05O H, 9.7(4)%], and a minor vanadium metal
2
impurity [5.6(4)%] (Table 2). Thermogravimetric reoxidation
of this sample under flowing oxygen resulted in a 17.2% gain in
mass (Figure S4), consistent with a composition of
SrV0.95Ti0.05O H for the majority cubic phase, as described
1.5
y
in detail in the Supporting Information. Manganometric
titrations indicate an average vanadium/titanium oxidation
state of 2.12(3)+ for the sample (Table S4). When the
presence of SrV0.95Ti0.05O H and vanadium metal is taken into
2
account, this indicates that the average vanadium/titanium
oxidation state in SrV0.95Ti0.05O H is 2.47(4)+, consistent with
x
y
a composition for the majority cubic phase of
SrV0.95Ti0.05O H and indicating reduction substantially
1.5
1.5
3+
beyond the V level.
Physical Characterization of SrV0.95Ti0.05O H . Zero-
1.5
1.5
Figure 3. Powder X-ray diffraction data collected during the reaction
field-cooled and field-cooled magnetization data collected from
SrV0.95Ti0.05O H in an applied field of 100 Oe (Figure 5)
between CaH and SrV0.95Ti0.05O prepared by ceramic (top) and arc-
2
3
1.5
1.5
melt (bottom) routes. Red symbols indicate positions and Miller
can be fitted to the Curie−Weiss law [χ = C/(T − θ) + K] to
indices of tetragonal “Sr(V/Ti)O H” phases, and green symbols
2
−
2
3
−1
yield values of C = 1.99(3) × 10 cm ·K·mol , θ = 2.27(7) K,
indicate positions and Miller indices of cubic Sr(V/Ti)O H . Data
1.5 1.5
and K = 8.5(1) × 10− cm ·K·mol . The parameters indicate
4
3
−1
sets marked with “(w)” have been washed to remove CaH and CaO.
2
Figure 4. Observed, calculated, and difference plots from the refinement of SrV0.95Ti0.05O H against synchrotron powder X-ray diffraction data.
1.5
1.5
Lower tick marks indicate peak positions of the majority phase, middle tick marks SrV0.95Ti0.05O H, upper tick marks elemental vanadium. The
2
excluded region removes a contribution from the “amorphous” boron used to dilute the sample.
D
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
12
Table 2. Structural Details from the Three-Phase Refinement
as expected from previous studies of this material). This
of SrV0.95Ti0.05O H against Synchrotron Powder X-ray
Diffraction Data
combination of physical data indicates that while no local
paramagnetic moment can be observed for the V centers, there
is no indication of long-range magnetic order in
SrV0.95Ti0.05O H . Transport measurements performed on
1.5
1.5
a
SrV0.95Ti0.05O1.5H1.5
1.5
1.5
atom
x
y
z
fraction
Uiso (Å2)
cold-pressed powder samples of SrV0.95Ti0.05O H indicate
1.5
1.5
Sr
1/2
1/2
1/2
1
0.0038(2)
0.0024(1)
0.0017(2)
that it is insulating.
V/Ti
O/H
0
0
0
0.95/0.05
0.5/0.5
1
/2
0
0
DISCUSSION
■
̅
SrV0.95Ti0.05O1.5H : space group Pm3m (No. 221)
1.5
SrVO −SrTiO Solubility. The lack of solubility between
3
3
formula weight: 163.90 g·mol− , Z = 1
mass fraction: 84.3(2)%, a = 3.888(1) Å
SrV0.95Ti O H
1
SrVO and SrTiO under “ceramic” conditions is surprising in
3
3
light of previous reports in the literature describing the
complete SrV1−xTi O solid solution, studied both as powder
0.05
2
x
3
Uiso (Å2)
27,28
atom
x
y
z
fraction
and single-crystal samples.
In the former study, powder
Sr
1/2
1/2
1/2
1
0.0042(3)
0.0021(2)
0.0019(3)
samples were prepared at higher temperatures than those
employed under our “ceramic” conditions (1500 °C, 10% H in
V/Ti
O
0
0
0
0.95/0.05
2
1
argon). However, when we attempted to replicate these high-
temperature conditions, the samples prepared contained large
/2
0
0
1
1
1/
a
H
0
0
0.002
2
quantities of SrV O , presumably due to strontium volatility,
and so were unsuitable for topochemical reduction studies. In
the arc-melting route used here, high temperatures (T > 2000
SrV0.95Ti0.05O H: space group P4/mmm (No. 123)
10 15
2
formula weight: 170.40 g·mol−1, Z = 1
mass fraction: 9.7(4)%, a = 3.929(1) Å, c = 3.738(4) Å
°
C) are applied only for a brief time, and as a result, the reagent
Vanadium
volatility is less of a problem. The increased SrVO −SrTiO
3
3
̅
vanadium: space group Im3m (No. 229)
formula weight: 50.94 g·mol 1, Z = 2
−
solubility observed in samples prepared by arc-melting, in
combination with the previous high-temperature synthesis
reports, suggests that the SrV1−xTi O solid solution is only
mass fraction: 5.6(4)%, a = 3.032(3) Å
radiation source: rynchrotron X-ray, λ = 0.82626(1)
temperature: 298 K
x
3
stable at high temperature but can be partially quenched as a
metastable phase by rapid cooling. This low-temperature
metastability is confirmed by the observation that when
annealed at 1100 °C, “quenched” SrV1−xTi O samples
2
χ = 9.625; wR = 3.43%; R = 2.31%
p
p
a
Thermal parameter of hydride is set at 0.002 Å2.
x
3
completely separate into simple mixtures of SrVO3 and
SrTiO . As noted above, the arc-melted SrV Ti O solid
3
1−x
x
3
solutions still exhibit a degree of phase separation. We attribute
the multiphase nature of the arc-melted samples to the limited
cooling rate that can be achieved in the arc furnace, which
allows for a degree of phase separation upon cooling. Despite
this partial phase separation, it is clear that, in the arc-melted
SrV0.95Ti0.05O3 sample, the majority cubic “SrVO ” phase
3
contains some dissolved titanium. Furthermore, annealing
studies show that the SrV0.95Ti0.05O solid solution is kinetically
3
stable at 600 °C, so the topochemical reduction/anion
exchange of SrV1−xTi O phases can be studied.
x
3
Influence of Doping on Reaction Products. The change
in reactivity between SrVO and CaH upon titanium doping is
3
2
striking. Unsubstituted SrVO reacts readily with CaH to yield
3
2
SrVO H, but attempts to advance this reaction further by
2
raising the reaction temperature lead to nontopochemical
reactions and, ultimately, to the formation of SrO and metallic
vanadium. In contrast, the reaction between CaH and the arc-
2
melted sample of nominal composition SrV0.95Ti0.05O3
proceeds all the way to cubic SrV0.95Ti0.05O H , with the
Figure 5. Zero-field-cooled and field-cooled magnetization data
collected from SrV0.95Ti0.05O H in an applied field of 100 Oe.
The inset shows an approximate fit to the Curie−Weiss law after a
large temperature-independent term has been removed.
1.5
1.5
1.5
1.5
direct analogue of SrVO H, tetragonal “SrV0.95Ti0.05O H”,
2
2
appearing only transiently as an intermediate phase. These
very different reaction outcomes suggest that the presence of
even small amounts of titanium dopant destabilizes
that the majority of the observed magnetization is temperature-
independent, with the value of the Curie constant correspond-
ing to 0.2% of that expected for S = 1, V centers. μ SR data
SrV0.95Ti0.05O H in some way, allowing it to react further.
While the different reactivity patterns have frustrated our initial
2
3
+
+
goal of forming stable hole-doped analogues of SrVO H, they
2
collected from SrV0.95Ti0.05O H (Figure S5) show only a
raise some important questions about the origin of stability in
these mixed anion lattices. The lattice parameters of the
1.5
1.5
relaxation of the polarization and no precession signal that
would be associated with long-range magnetic order (data
collected from the reduced ceramic sample do show oscillations
“SrV0.95Ti0.05O H” intermediate phase offer some clues to the
2
origin of its relative instability. We noted above that while the
that are due to the presence of magnetically ordered SrVO H,
lattice parameter a is marginally shorter than that in
2
E
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 6. (top) Trans and cis unit cells used in the computational study. Sr, V, O, and H atoms are in dictated by green, gray, red, and yellow
spheres, respectively. (bottom) Calculated density of states for trans-SrVO H and cis-SrTiO H.
2
2
unsubstituted SrVO H [3.929(1) vs 3.9331(4) Å], c is
trans along the c direction. In the cis isomer, the hydrides are
moved into the ab plane, and in the arrangement shown in
Figure 6, they are eclipsed in successive ab planes. The
alternative where the hydrides are staggered in successive
planes is computed to be higher in energy in all cases and is not
considered further here. We acknowledge that the cis
arrangement represents only one possible arrangement
among the many that could contribute to the anion-disordered
cubic phase, but nevertheless a comparison of the two unit cells
allows us to assess the energetic cost of moving the hydrides
2
1
2
significantly elongated [3.738(4) vs 3.6671(3) Å]. These
changes are consistent with a degree of anion disorder in the
titanium-containing phase, with small concentrations of oxide
ions along c and hydride ions in the ab plane causing elongation
and contraction, respectively. We note in this context that
SrCrO H adopts a cubic, anion-disordered structure quite
2
29
different from that of SrVO H, suggesting that the anion-
ordered tetragonal phase is uniquely stable for a d electronic
2
2
configuration: deviations to either lower (SrV0.95Ti0.05O H,
2
1
.95
3
d
) or higher (SrCrO H, d ) electron counts appear to drive
around the lattice. In the case of SrVO H, the trans
2
2
the system into alternative anion-disordered phases. The partial
disordering of the anion lattice adds a large degree of strain to
the framework because V−O−V connections are inserted into a
lattice optimized for V−H−V links and vice versa, which
significantly destabilizes SrV Ti O H with respect to anion-
arrangement of hydride ligands proves to be substantially
more stable than the cis alternative (ΔE = +0.26 eV per formula
unit), consistent with its established preference for the
tetragonal phase. In SrTiO H, in contrast, the order is reversed,
2
0
.95 0.05
2
with a strong preference for the cis arrangement over trans (ΔE
ordered SrVO H.
= −0.21 eV). The mixed-metal system SrV Ti O H is
2
0.75 0.25
2
These observations are supported by a series of DFT
intermediate between these two limits, with the trans structure
being favored, but to a much lesser degree than in the all-
vanadium analogue (ΔE = +0.09 eV). The calculations
therefore indicate that the trans arrangement is optimal for
calculations performed on the 2 × 2 × 2 expanded unit
cells shown in Figure 6, each of which contains four transition-
metal centers. We have studied the two limiting stoichiome-
tries, SrVO H and SrTiO H, as well as SrV0.75Ti0.25O H, the
2
3+
the d configuration of V and is progressively destabilized
2
2
2
3
+
closest model to the SrV0.95Ti0.05O H phase that is accessible
relative to its cis analogue as the concentration of Ti is
increased. We can analyze these differences in terms of two
competing factors, the first of which is the strong trans
influence of the hydride ligand, which will in general disfavor
2
within a unit cell of these dimensions. In the trans unit cell, the
anions are arranged in the same ordered array as that found in
native SrVO H, i.e., with the two hydrides aligned mutually
2
F
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
mutually trans arrangements of hydride ligands. This effect is
mediated by the σ framework and stems from the fact that only
a single metal d orbital can participate in both metal hydride σ
bonds when the ligands are trans to each other. Superimposed
upon this are striking differences in the π character of the two
ligands: a hydride is a pure σ donor, while an oxide is also a
strong π donor, and there is a strong driving force to
accumulate oxide ligands in the same plane as vacant metal-
SrV1−xTi O H. This, in turn, raises the activation energy for the
x
2
topochemical reaction that converts SrVO H into SrVO H ,
2
1.5 1.5
allowing the former to be trapped as a stable species while
SrV1−xTi O H is only a transiently stable intermediate. Indeed,
x
2
the absence of detectable amounts of SrVO H indicates that
1.5 1.5
the activation energy for the formation of this compound from
SrVO H is greater than that for its subsequent decomposition
2
to SrO and metallic vanadium. Even small amounts of titanium
substitution significantly degrade the special stability of
based orbitals with dπ character. In the case of SrVO H, with a
2
2
d configuration at the metal, these two competing factors are
SrVO H, making the system more reactive and leading to the
2
diametrically opposed: the presence of a single vacant dπ orbital
observed changes in reaction outcome.
(
d ) means that the O−V π bonding is optimal in the
Attempts to dope other phases prepared via topochemical
reactions have revealed similar but less dramatic responses to
cation substitution. For example, attempts to electron-dope the
xy
tetragonal phase, where all four oxides are aligned in a single
plane, but this necessarily forces the two hydrides into a
mutually trans arrangement. The observed preference for the
tetragonal phase indicates that in this case the enhanced π1
donation from the oxides is the dominant factor. For a d
configuration, in contrast, the presence of two vacant dπ
orbitals allows for effective O−Ti π bonding in either cis or
trans arrangements, and the σ effects (the trans influence) then
dominate, imposing a preference for the cis arrangement of
2+
Fe phase SrFeO
reveal that RE Sr1−xFeO3−δ is reduced to RE
maintaining an oxidation state of Fe in the reduced phase.
However, it should also be noted that the sensitivity to cation
substitution observed for SrVO H and SrFeO is not universal
in topochemical reduction reactions. For example, the
conversion of La1−x MnO (A = Ba, Sr, Ca) perovskites to
MnO2.5 brownmillerite phases containing mixtures of
via rare-earth substitution for strontium
2
Sr1−xFeO2+x/2
,
x
x
2+
30
2
2
A
x
3
hydrides. In the SrV0.75Ti0.25O H system, the arrangement of
La1−xA
x
2
Mn2
+/3+
is tolerant of a range of A-site compositions and can be
anions necessarily represents a compromise because it is not
possible to simultaneously satisfy the geometric preferences of
vanadium (trans) and titanium (cis) in a single lattice.
readily achieved for 0.15 < x < 0.5 for A = Sr and 0.22 < x < 0.5
31
for A = Ca.
The tendency of increasing quantities of titanium in the
lattice to reduce the stability of the tetragonal phase suggests
that the presence of a stable, anion-ordered tetragonal phase is
CONCLUSIONS
Topochemical reduction of SrVO yields the anion-ordered,
tetragonal phase SrVO H. However, the analogous reaction
conditions applied to hole-doped SrV0.95Ti0.05O yields cubic
■
3
2
uniquely associated with a d configuration. Moreover, the
2
properties of the SrV0.95Ti0.05O H phase indicate that even a
2
3
relatively small depletion of the electron density in the dπ band
is sufficient to tip the balance in favor of anion disorder. This
provides a framework for understanding the different paths
SrV0.95Ti0.05O H not tetragonal SrV0.95Ti0.05O H. This
1.5
1.5
2
product, SrV0.95Ti0.05O H , is an insulating mixed-valent
1.5
1.5
2
+/3+
2−
−
V
phase, suggesting that the O /H disorder acts to
+
followed by the topochemical reactions of CaH with SrVO3
localize the vanadium d electrons. Magnetization and μ SR data
show respectively that the V centers exhibit no local magnetic
moment and that there is no indication of (antiferro)magnetic
order. This unusual combination of transport and magnetic
properties leads us to suggest that the system may adopt a
network of V−V bonds, although further characterization will
be required to rigorously demonstrate this.
2
and SrV1−xTi O . Figure 7 shows schematic reaction-energy
x
3
profiles for the two cases, where the key difference is that the
tetragonal SrVO H phase lies in a deeper well on the potential
2
energy surface than the corresponding tetragonal phase for
While the hole doping of SrVO did not, ultimately, achieve
3
the initially intended outcome of producing hole-doped
analogues of SrVO H, the dramatic differences in reactivity in
2
response to seemingly minor changes in composition have
encouraged us to consider the factors that determine the
outcome of reactions in the solid state. Under typical high-
temperature synthetic conditions, the products of a reaction are
determined by thermodynamic stability, and small composi-
tional changes tend not to change the thermodynamic
parameters to the extent that some other competing phase
becomes more stable. As a result, it is generally possible to
make small compositional changes to solid phases to tune their
physical and chemical behavior; that is to say, these phases are
generally “dopable”. In contrast, the products of low-temper-
ature topochemical reactions are selected under kinetic control;
the product phase is the one that forms fastest and is not
necessarily the most thermodynamically stable. As a result, the
products of such reactions depend sensitively on the micro-
scopic mechanism, followed by a reaction. The change in
reaction product from SrVO H to SrV Ti O H upon
2
1−x
x
1.5 1.5
substituting ∼5% of the vandium in SrVO for titanium shows
3
Figure 7. Schematic diagram showing the difference in activation
that even small changes in the composition can dramatically
alter the relative rates of the competing pathways that exist
energies for the reduction of SrVO and SrV Ti O .
3
1−x
x
3
G
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
under low-temperature conditions. Given that this observed
sensitivity of the reaction mechansim to chemical composition
is likely to be a general feature of topochemical reactions, we
must conclude that a large number of the metastable phases
prepared via topochemical reactions will, ultimately, prove not
to be dopable.
(9) Hayward, M. A.; Cussen, E. J.; Claridge, J. B.; Bieringer, M.;
Rosseinsky, M. J.; Kiely, C. J.; Blundell, S. J.; Marshall, I. M.; Pratt, F.
L. The hydride anion in an extended transition metal oxide array -
LaSrCoO H. Science 2002, 295, 1882.
3
(
10) Helps, R.; Rees, N.; Hayward, M. Sr Co O4.33H0.84: An extended
3 2
transition metal oxide-hydride. Inorg. Chem. 2010, 49, 11062−11068.
(11) Kobayashi, Y.; Hernandez, O. J.; Sakaguchi, T.; Yajima, T.;
Roisnel, T.; Tsujimoto, Y.; Morita, M.; Noda, Y.; Mogami, Y.; Kitada,
A.; Ohkura, M.; Hosokawa, S.; Li, Z. F.; Hayashi, K.; Kusano, Y.; Kim,
J. E.; Tsuji, N.; Fujiwara, A.; Matsushita, Y.; Yoshimura, K.; Takegoshi,
ASSOCIATED CONTENT
■
*
S
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.8b00026.
K.; Inoue, M.; Takano, M.; Kageyama, H. An oxyhydride of BaTiO
3
exhibiting hydride exchange and electronic conductivity. Nat. Mater.
2012, 11, 507−511.
(12) Denis Romero, F.; Leach, A.; Moller, J. S.; Foronda, F.; Blundell,
Powder X-ray diffraction data from SrV Ti O
3
samples, structural, compositional, and μ SR character-
ization of “reduced” ceramic SrV0.95Ti0.05O3 and
0
.5 0.5
S.; Hayward, M. A. Strontium vanadium oxide-hydrides: ‘square-
planar’ two-electron phases. Angew. Chem., Int. Ed. 2014, 53, 7556−
7559.
+
SrV0.95Ti0.05O H , and details of DFT calculations
1.5
1.5
(13) Goodenough, J. B. Localized to itinerant electronic transition in
(
PDF)
perovskite oxides; Springer-Verlag: Berlin, 2001.
(
14) Bednorz, J. G.; Muller, K. A. Possible high-Tc superconductivity
in the Ba-La-Cu-O system. Z. Phys. B: Condens. Matter 1986, 64, 189−
93.
15) MacChesney, J. B.; Williams, H. J.; Potter, J. F.; Sherwood, R. C.
AUTHOR INFORMATION
■
1
(
Corresponding Author
E-mail: michael.hayward@chem.ox.ac.uk.
*
Magnetic study of manganate phases - CaMnO3 Ca4Mn O
3
10
Ca Mn O Ca MnO . Phys. Rev. 1967, 164, 779.
ORCID
3
2
7
2
4
(
16) Matsumoto, G. Study of (La Ca )MnO magnetic strucutre of
1‑x x 3.1
Michael A. Hayward: 0000-0002-6248-2063
Author Contributions
LaMnO . J. Phys. Soc. Jpn. 1970, 29, 606.
3
(17) Schiffer, P.; Ramirez, A. P.; Bao, W.; Cheong, S. W. Low-
The manuscript was written through contributions of all
authors.
temperature magnetoresistance and the magnetic phase-diagram of
La Ca MnO . Phys. Rev. Lett. 1995, 75, 3336−3339.
1
‑x
x
3
(18) Liu, K.; Hou, Y. S.; Gong, X. G.; Xiang, H. J. Orbital
Notes
delocalization and enhancement of magnetic interactions in perovskite
oxyhydrides. Sci. Rep. 2016, 6, XX.
The authors declare no competing financial interest.
(
19) Wei, Y. F.; Gui, H.; Li, X.; Zhao, Z. J.; Zhao, Y. H.; Xie, W. H.
ACKNOWLEDGMENTS
■
The effect of hydrogen ordering on the electronic and magnetic
properties of the strontium vanadium oxyhydride. J. Phys.: Condens.
Matter 2015, 27, 206001.
M.A.P. thanks CONACYT and the Balliol College, Oxford, for
a scholarship. Experiments at the Diamond Light Source were
performed as part of the Block Allocation Group award
(
20) Yamamoto, T.; Zeng, D.; Kawakami, T.; Arcisauskaite, V.; Yata,
K.; Patino, M. A.; Izumo, N.; McGrady, J. E.; Kageyama, H.; Hayward,
M. A. The role of π-blocking hydride ligands in a pressure-induced
insulator-to-metal phase transition in SrVO H. Nat. Commun. 2017, 8,
“
Oxford Solid State Chemistry BAG to probe composition−
structure−property relationships in solids” (EE13284). Part of
this work was carried out at the Swiss Muon Source, Paul
Scherrer Institut, Villigen, Switzerland.
2
1
(
217.
21) Rey, M. J.; Dehaudt, P.; Joubert, J. C.; Lambertandron, B.;
Cyrot, M.; Cyrotlackmann, F. Preparation and structure of the
REFERENCES
■
compounds SrVO and Sr VO . J. Solid State Chem. 1990, 86, 101−
3
2
4
(
1) Hayward, M. A. Soft chemistry synthesis of oxides. In
Comprehensive Inorganic Chemistry II; Reedijk, J., Poeppelmeier, K.
R., Eds.; Elsevier: Oxford, U.K., 2013; Vol. 2, pp 417−453.
2) Schaak, R. E.; Mallouk, T. E. Perovskites by design: A toolbox of
1
08.
(
22) Larson, A. C.; Von Dreele, R. B. General Structure Analysis
System. Los Alamos National Laboratory Report LAUR 86-748, 2000.
23) Kresse, G.; Furthmuller, J. Efficient iterative schemes for ab
(
(
solid-state reactions. Chem. Mater. 2002, 14, 1455−1471.
3) Gopalakrishnan, J. Chimie douce approaches to the synthesis of
metastable oxide materials. Chem. Mater. 1995, 7, 1265−1275.
4) Yamamoto, T.; Kageyama, H. Hydride reduction of transition
metal oxides. Chem. Lett. 2013, 42, 946.
5) Tsujimoto, Y.; Tassel, C.; Hayashi, N.; Watanabe, T.; Kageyama,
initio total-energy calculations using a plane-wave basis set. Phys. Rev.
B: Condens. Matter Mater. Phys. 1996, 54, 11169−11186.
(
(
24) Perdew, J. P.; Burke, K.; Wang, Y. Generalized gradient
(
approximation for the exchange-correlation hole of a many-electron
system. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 16533−
(
1
6539.
H.; Yoshimura, K.; Takano, M.; Ceretti, M.; Ritter, C.; Paulus, W.
Infinite-layer iron oxide with a square-plannar coordination. Nature
(25) Schmidbauer, M.; Kwasniewski, A.; Schwarzkopf, J. High-
precision absolute lattice parameter determination of SrTiO , DyScO
3
3
2
(
007, 450, 1062−1065.
6) Seddon, J.; Suard, E.; Hayward, M. Topotactic reduction of
YBaCo O and LaBaCo O : square-planar Co(I) in an extended oxide.
and NdGaO single crystals. Acta Crystallogr., Sect. B: Struct. Sci. 2012,
3
6
(
8, 8−14.
2
5
2
5
26) Sakaguchi, T.; Kobayashi, Y.; Yajima, T.; Ohkura, M.; Tassel, C.;
J. Am. Chem. Soc. 2010, 132, 2802−2810.
Takeiri, F.; Mitsuoka, S.; Ohkubo, H.; Yamamoto, T.; Kim, J. E.; Tsuji,
N.; Fujihara, A.; Matsushita, Y.; Hester, J.; Avdeev, M.; Ohoyama, K.;
Kageyama, H. Oxyhydrides of (Ca,Sr,Ba)TiO3 perovskite solid
solutions. Inorg. Chem. 2012, 51, 11371−11376.
(
7) Hayward, M. A.; Green, M. A.; Rosseinsky, M. J.; Sloan, J.
Sodium hydride as a powerful reducing agent for topotactic oxide
deintercalation: synthesis and characterisation of the nickel (I) oxide
LaNiO . J. Am. Chem. Soc. 1999, 121, 8843−8854.
2
(
8) Denis Romero, F.; Burr, S. J.; McGrady, J. E.; Gianolio, D.; Cibin,
(27) Hong, K. Y.; Kim, S. H.; Heo, Y. J.; Kwon, Y. U. Metal-insulator
2+
G.; Hayward, M. A. SrFe Ru O : sqaure-planar Ru in an extended
transitions of SrTi1‑x
V
O3 solid solution system. Solid State Commun.
x
0.5
0.5
2
oxide. J. Am. Chem. Soc. 2013, 135, 1838−1844.
2002, 123, 305−310.
H
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
28) Tsuiki, H.; Kitazawa, K.; Fueki, K. The donor level of V4+ and
the metal-nonmetal transtion in SrTi V O . Jpn. J. Appl. Phys. Part 1
1
‑x
x
3
1
(
983, 22, 590−596.
29) Tassel, C.; Goto, Y.; Kuno, Y.; Hester, J.; Green, M.; Kobayashi,
Y.; Kageyama, H. Direct synthesis of chromium perovskite oxyhydride
with a high magnetic-transition temperature. Angew. Chem., Int. Ed.
2
(
014, 53, 10377−10380.
30) Yamamoto, T.; Ohkubo, H.; Tassel, C.; Hayashi, N.; Kawasaki,
S.; Okada, T.; Yagi, T.; Hester, J.; Avdeev, M.; Kobayashi, Y.;
Kageyama, H. Impact of lanthanoid substitution on the structural and
physical properties of an infinite-layer iron oxide. Inorg. Chem. 2016,
5
(
5, 12093−12099.
31) Parsons, T. G.; D’Hondt, H.; Hadermann, J.; Hayward, M. A.
The synthesis and structural characterisation of of La A MnO (A =
1
‑x
x
2.5
Ba, Sr, Ca) phases − mapping the variants of the brownmillerite
structure. Chem. Mater. 2009, 21, 5527.
I
DOI: 10.1021/acs.inorgchem.8b00026
Inorg. Chem. XXXX, XXX, XXX−XXX