Single-crystal structures, optical absorptions, and electronic distributions of thorium oxychalcogenides ThOQ (Q = S, Se, Te)

Article abstract of DOI:10.1021/ic300510x

The compounds ThOS, ThOSe, and ThOTe have been synthesized, and their structures have been determined by means of single-crystal X-ray diffraction methods. All three compounds adopt the PbFCl structure type in the tetragonal space group D_4h

Full text of DOI:10.1021/ic300510x

Article
pubs.acs.org/IC
Single-Crystal Structures, Optical Absorptions, and Electronic
Distributions of Thorium Oxychalcogenides ThOQ (Q = S, Se, Te)
Lukasz A. Koscielski,† Emilie Ringe,† Richard P. Van Duyne,† Donald E. Ellis,†,‡ and James A. Ibers*,†
‡
†Department of Chemistry and Department of Physics and Astronomy, Northwestern University, 2145 Sheridan Road, Evanston,
Illinois 60208, United States
S
* Supporting Information
ABSTRACT: The compounds ThOS, ThOSe, and ThOTe have been synthesized, and their
structures have been determined by means of single-crystal X-ray diffraction methods. All three
compounds adopt the PbFCl structure type in the tetragonal space group D4h7 − P4/nmm. More
precise crystallographic data have been obtained for ThOS and ThOSe, which had previously only
been known from X-ray powder diffraction data. ThOS, ThOSe, and ThOTe are yellow-, orange-,
and black-colored, respectively. From single-crystal optical absorption measurements the band
gaps are 2.22, 1.65, and 1.45 eV, respectively. Optical band gaps, ionic charges, and densities of
states were calculated for the three compounds with the use of Density Functional methods.
chunks (Albrecht-Schmitt, Notre Dame) powdered according to a
INTRODUCTION
■
literature procedure.16 Both sources yielded ThOQ crystals.
The ThOQ (Q = S, Se, Te) compounds are isostructural to
The ThOQ compounds were synthesized as side products in
multiple reactions. Although other reactants were present, the
other AnOQ (An = U, Np, Pu) compounds,1−4 and crystallize
7
in the PbFCl structure type in space group D4h − P4/nmm of
the tetragonal system. The structure differs from those of the
parent ThQ2 compounds. ThS2 and ThSe2 crystallize in the
orthorhombic space group Pmna, and ThTe2, from X-ray
powder diffraction data, appears to crystallize in the hexagonal
system with cell constants a = 8.49 Å and c = 9.01 Å.5−7 The
ThOQ compounds were first identified and characterized
structurally from X-ray powder diffraction data.8,9 In addition,
the structure of ThOTe has been characterized by single-crystal
X-ray diffraction methods.10 Since then, there have been
numerous studies of these compounds, including their
syntheses, structures (including those at high pressures),
doping properties, and thermal properties.9,11−15 However, no
optical measurements have been performed on ThOQ
compounds despite their bright colors, nor have theoretical
studies been performed.
elemental compositions of the crystals were determined to be only
Th/O/Q on an EDX-equipped Hitachi S-3400 SEM. The crystals
were generally plate-like and were yellow, orange, or black for Q = S,
Se, or Te, respectively.
Listed below is one reaction for each chalcogen. Reactants were
loaded into carbon-coated fused-silica tubes in an Ar-filled glovebox.
Each tube was evacuated to near 10−4 Torr, flame-sealed, placed in a
computer-controlled furnace, and then heated according to the
specified temperature profile.
ThOS. A tube was loaded with Th (30.0 mg, 0.129 mmol), Ge (9.4
mg, 0.129 mmol), S (4.2 mg, 0.131 mmol), and NaBr (53.2 mg, 0.517
mmol). It was heated to 1273 K in 24 h, kept at 1273 K for 99 h,
cooled to 673 in 198 h, and then cooled to 298 K in 4 h. The tube
contained a yellow-colored amorphous solid and many well-formed
yellow plates of ThOS.
ThOSe. A tube was loaded with Th (30 mg, 0.129 mmol), Fe (7.2
mg, 0.129 mmol), Se (30.6 mg, 0.388 mmol), and CsCl (100 mg,
0.594 mmol). It was heated to 1173 K in 30 h, kept at 1173 K for 96 h,
cooled to 773 K in 144 h, kept at 773 K for 48 h, and then cooled to
298 K in 6 h. The tube contained orange crystals of ThOSe and
metallic pieces, which by EDX examination contained Cs, Fe, and Se.
ThOTe. A tube was loaded with Th (30.0 mg, 0.129 mmol) and Te
(33.0 mg, 0.258 mmol). It was heated to 1273 K in 24 h, kept at 1273
K for 72 h, and then cooled to 298 K in 9 h. The tube contained black
plates of ThOTe in nearly 100 wt % yield and some black powder,
which was identified as ThOTe by an X-ray powder diffraction
measurement.
Here we report single-crystal X-ray diffraction structures of
all three compounds as well as their optical properties and
computational results on electronic distributions and densities
of states.
EXPERIMENTAL METHODS
■
Caution! 232Th is an α-emitting radioisotope and as such is considered a
health risk. Its use requires appropriate infrastructure and personnel
trained in the handling of radioactive materials.
Syntheses. Fe (Aesar, 99.5%), Ge (Aldrich, 99.99%), S
(Mallinckrodt, 99.6%), Se (Cerac, 99.999%), Te (Aldrich, 99.8%),
NaBr (Aldrich, 99.99+%), and CsCl (Aldrich, 99.9%) were used as
received. Two sources of Th were used: Th (MP Biomedicals) and Th
Received: March 7, 2012
Published: July 16, 2012
© 2012 American Chemical Society
8112
dx.doi.org/10.1021/ic300510x | Inorg. Chem. 2012, 51, 8112−8118

Inorganic Chemistry
Article
Crystal Structure Determinations. Single-crystal X-ray diffrac-
tion data for ThOQ (Q = S, Se, Te) were collected with the use of
graphite-monochromatized MoKα radiation (λ = 0.71073 Å) at 100 K
on a Bruker APEX2 diffractometer.17 For ThOS and ThOSe, data
were collected by a scan of 0.3° in ω in groups of 606 frames at φ
settings of 0°, 90°, 180°, and 270°. Crystal decay was monitored by
recollecting 50 initial frames at the end of the data collection. For
ThOTe the data collection strategy was obtained from an algorithm in
COSMO in the program APEX217 as a series of 0.3° scans in φ and ω.
For all data collections the crystal-to-detector distance was 6.0 cm and
the exposure time was 10 s/frame. The collection of intensity data as
well as cell refinement and data reduction were carried out with the
use of the program APEX2.17 Face-indexed absorption, incident beam,
and decay corrections were performed with the use of the program
SADABS.18 The structures were solved with the direct-methods
program SHELXS and refined with the least-squares program
SHELXL.19 The atomic positions were standardized with the program
STRUCTURE TIDY.20 Additional experimental details are given in
Table 1 and in the Supporting Information.
Periodic spin-polarized band structure calculations were performed
with the use of the first principles DFT program VASP (Vienna ab
initio simulation package); pseudopotentials were applied with a plane-
wave basis.22−25 The exchange correlation potential was chosen as the
generalized gradient approximation (GGA) in a projector augmented
wave (PAW) method, specifically GGA (PW91).26,27 Relativistic core
pseudopotentials were used. In selected cases a perturbative treatment
of spin−orbit coupling was made, after lattice relaxations had
converged. Automatically generated Monkhorst−Pack grids were
used to carry out Brillouin zone integrations.28 6 × 6 × 6 k-point
meshes were chosen for relaxations, total energy calculations,
establishing convergence, energy comparisons, and DOS analysis.
Ionic relaxation convergence was established when Hellmann−
Feynman forces on each ion relaxed below 0.02 eV/Å. Additionally
for ThOS, an onsite Coulomb correction to the Th 5f shell, that is the
Hubbard U term,29 was implemented in the rotationally invariant
approach where the onsite Coulomb term U and onsite exchange term
J were treated together as Ueff = U − J.30 Because the 5f shell is
nominally unoccupied, we expected that the effects of Ueff on optical
properties, such as the band gap, would be minimal. It was
nevertheless important to check this by direct calculation.
In the calculations, the electrons described as core in the PAW
potentials were those composed of [Xe]5d104f14 for Th, leaving 12
valence elections per atom as 6s2p6d27s2; [He] for O, leaving six
valence electrons as 2s2p4; [Ne] for S, leaving six valence electrons as
3s2p4; [Ar]3d10 for Se, leaving six valence electrons as 4s2p4; and
[Kr]4d10 for Te, leaving six valence electrons as 5s2p4. Calculations
were conducted on the 6-atom periodic crystallographic unit cell in the
tetragonal space group P4/nmm; atomic positions within the fixed 100
K unit cell were relaxed to their lowest energy positions.
Oxidation states as ionic charges were determined by volume
integration of electron density with the use of both atomic sphere
integrations with radii, RWS, and Bader’s topological atom
method.31−33 Rather than dividing space into hard spheres (RWS),
the approach of Bader is to divide space into atomic regions
determined by zero-flux charge-density surfaces.31 Using these two
methods, we define the oxidation state as the difference between the
number of valence electrons contained within a volume and the
number assigned to the neutral atom. The values of RWS were initially
set to the standard radii34 and then increased to fill the unit-cell
volume. Final values of RWS were chosen as 1.5, 1.6, 1.8, 1.8, and 1.8 Å
for Th, O, S, Se, and Te, respectively. The radii of S, Se, and Te were
set to the same value to enable a consistent comparison among the
chalcogenides.
a
Table 1. Crystal Data and Structure Refinements for ThOQ
compound
color
ThOS
ThOSe
orange
ThOTe
black
yellow
Fw
280.10
327.00
375.64
a (Å)
3.9580(2)
6.7468(3)
105.694(9)
8.801
4.0176(4)
7.0198(6)
113.31(2)
9.584
4.1173(3)
7.5289(6)
127.63(2)
9.775
c (Å)
V (Å3)
ρc (g cm−3
)
μ (mm−1
)
71.08
81.45
69.25
b
R(F)
0.0134
0.0177
0.0130
c
Rw(F2)
0.0298
0.0455
0.0345
a
For all structures Z = 2, space group = P4/nmm, λ = 0.71073 Å, T =
b
c
2
2
100(2) K. R(F) = ∑||Fo| − |Fc||/∑|Fo| for Fo > 2σ(Fo ). Rw(F2) =
{∑[w(Fo 2 − Fc2)2]/∑wFo }1/2 for all data. w−1 = σ2(Fo ) + (qFo )2 for
4
2
2
2
2
2
Fo ≥ 0; w−1 = σ2(Fo ) for Fo < 0. q = 0.0127 for ThOS, 0.0249 for
ThOSe, and 0.0135 for ThOTe.
Optical Measurements. Single-crystal optical absorption meas-
urements were performed at visible frequencies from 3.2 eV (387 nm)
to 1.5 eV (827 nm) and at IR frequencies from 1.8 eV (668 nm) to
1.08 eV (1148 nm) at 298 K. A single crystal of ThOQ (Q = S, Se, Te)
mounted on a goniometer head was inserted on a custom-made holder
fitted to a Nikon Eclipse Ti2000−U inverted microscope. The crystal
was positioned at the focal plane above the 20× objective of the
microscope and illuminated with a tungsten-halogen lamp. The
transmitted light was spatially filtered with a 200 μm aperture. For
visible measurements, light was dispersed by a 150 groove/mm grating
in an Acton SP2300i imaging spectrometer, and collected on a back-
illuminated, liquid nitrogen-cooled CCD (Spec10:400BR, Princeton
Instruments). For IR measurements, light was dispersed by a 150
groove/mm grating in an Acton SP2300i imaging spectrometer
configured for IR wavelengths, and collected on a liquid nitrogen-
cooled InGaAs Array detector (OMA V 1024−1.7, Princeton
Instruments). Spectra of ThOS and ThOSe at visible wavelengths
and of ThOTe at IR wavelengths can be found in Figure 1. Spectra of
ThOTe at visible wavelengths and of ThOS at IR wavelengths were
acquired as controls and are available in the Supporting Information.
Theoretical Calculations. Bond valences were calculated from
standard parameters21 as a simple empirical measure of oxidation state.
The bond valence of an atom, V, is defined as the sum of the individual
bond valences, vi, surrounding the atom: V = ∑vi . Individual bond
valences were calculated from experimental bond lengths, Ri, and
empirical parameters, R0, which are unique to each atom pair: vi =
exp[(R0 − Ri)/0.37]. The parameter R0 for the atom pairs Th−O,
Th−S, Th−Se, and Th−Te was 2.167, 2.64, 2,76, and 2.94 Å,
respectively.21
RESULTS AND DISCUSSION
■
Syntheses. Because of the oxophylicity of Th and the
stability of the ThOQ compounds, these are often the major
products in reactions involving Th and a chalcogen with or
without a variety of other elements when such reactions are
carried out in fused-silica tubes, even if such tubes are carbon-
coated. This is apparent in the present syntheses.
The reaction that yielded ThOS contained, in addition to Th
and S, Ge and NaBr. The Ge was present because the reaction
was intended as an exploration of the Th/Ge/S system with
NaBr as flux. The reaction that yielded ThOSe contained, in
addition to Th and Se, Fe and CsCl. The Fe was present
because the reaction was intended to synthesize ThFeSe3. The
reaction that yielded ThOTe contained only Th and Te and
was intended to synthesize ThTe2. The most likely oxygen
source in the reactions was the fused-silica tubes, because the
reaction tubes were etched despite being carbon-coated.
Experimental Structures. The ThOQ compounds crys-
tallize with two formula units in the tetragonal space group P4/
nmm. They adopt the PbFCl structure with Th, O, and Q in
place of Pb, F, and Cl, respectively. The site symmetries of Th,
O, and Q are 4mm, 4mm, and 4m2, respectively. The unit-cell
̅
8113
dx.doi.org/10.1021/ic300510x | Inorg. Chem. 2012, 51, 8112−8118

Inorganic Chemistry
Article
Figure 1. (A) Absorbance vs energy (eV), (B) the spectra calculated for a direct band gap, and (C) spectra calculated for an indirect direct band gap
for ThOS, ThOSe, and ThOTe. The spectra calculated for an indirect band gap more closely resemble the observed absorbance spectra.
dimensions are listed in Table 1. These are plotted along with
those for AnOQ (An = U, Np; Q = S, Se)1 in Figure 2. The
Figure 2. Comparison of axial lengths for AnOQ (An = Th, U, Np; Q
= S, Se).
Figure 3. Structure of ThOQ viewed down the b-axis. The unit cell is
outlined.
contraction of the actinide radius with Z is evident from these
lattice constants. As seen in Figure 3, each Th atom is
connected to four O and five Q atoms in a distorted
monocapped square antiprism. Adjacent ThO4Q5 units share
one O and two Q atoms with each other in the ab plane to
form a layer. Each capping Q atom also forms one of the
corners of the base of an ThO4Q4 antiprism in the layer above
to create a double layer of ThO4Q5 units that share two Q
atoms between the layers.
Interatomic distances for ThOQ are listed in Table 2. Each
Th atom is equidistant from four Q atoms with a fifth Q atom
as a cap. In ThOSe and ThOTe the capping distance is longer
and in ThOS it is shorter than the distance to the other four Q
atoms.
The Th−O and Th−Q distances are similar to those in
ThO2 and the ThxQy binaries. The Th−O distance in ThOQ
varies between 2.3981(2) and 2.4287(3) Å whereas in ThO2 it
8114
dx.doi.org/10.1021/ic300510x | Inorg. Chem. 2012, 51, 8112−8118

Inorganic Chemistry
Article
The average bond valences for the atoms Th, O, and Q, are
+4.01, −2.06, and −1.95, respectively, which support the
formulation of ThOQ as Th4+O2−Q2−.
Table 2. Interatomic Distances in ThOQ (Q = S, Se, Te)
distance (Å)
ThOS
ThOSe
ThOTe
Th − O × 4
Th − Q × 4
Th − Q (cap)
2.3981(2)
3.002(1)
2.952(3)
2.4095(4)
3.0916(8)
3.139(2)
2.4287(3)
3.2593(5)
3.486(1)
Table 4. Oxidation State Calculations in ThOQ: Bader,
Wigner-Seitz (W−S), and Bond Valence
is 2.4237(1) Å.35 The Th−Q (noncapping Q) distances in
ThOS, ThOSe, and ThOTe are 3.002(1), 3.0916(8), and
3.2593(5) Å, respectively. In ThS2, ThSe2, and Th7Te12 the
Th−Q distance varies from 2.796(1) to 3.2353(9), 2.8598(1)
to 3.2882(1), and 3.187(2) to 3.459(2) Å, respectively.5,6,36
The present metrical data for ThOQ (Q = S, Se, Te) agree with
data previously obtained for the compounds, mainly from X-ray
powder diffraction studies. Only for ThOTe was there an
earlier single-crystal X-ray diffraction study, and the comparison
to this work is made in Table 3.
compound
ThOS
element
Bader
W−S
bond valence
Th
O
+2.49
−1.30
−1.18
+2.39
−1.33
−1.06
+2.25
−1.33
−0.93
+3.56
−2.24
−0.94
+3.56
−2.18
−0.50
+3.56
−2.08
+0.30
+4.07
−2.14
−1.93
+4.07
−2.08
−1.99
+3.89
−1.97
−1.92
S
ThOSe
ThOTe
Th
O
Se
Th
O
Te
Table 3. Comparison of Two Single-Crystal X-ray Structure
Determinations of ThOTe
A more quantitative analysis follows from integration of the
self-consistent electronic charge density. The calculated ionic
charge of each atom from both RWS volume integration and the
Bader topological method may be found in Table 4. The RWS
integration summed to 23.62, 23.12, and 22.22 electrons out of
a total of 24 valence electrons/cell for the compounds ThOS,
ThOSe, and ThOTe, respectively. The “missing” 0.38 to 1.78
electrons are to be found in the interstitial region. The increase
in the calculated interstitial charge S < Se < Te results from the
selection of equal chalcogen radii. This choice of RWS allows
constant-volume comparisons, which verify the sequential
expansion of electron density about the chalcogen site. This
is evident in Table 4 as the RWS charge on the chalcogen
becomes more positive on going from S to Te, whereas the
charge on Th is unchanged and that on O barely changes (0.16
e−). At the same time, the Th−O distance stays nearly constant
in the three structures, whereas the Th−Q distance increases
on going from S to Se and even more from Se to Te. This
demonstrates that the Th−Q bond becomes more covalent, S <
Se < Te, with increased interstitial and shared electron density
between Th and Q.
reference
a (Å)
this work
ref. [10]
4.1173(3)
7.5289(6)
2.4287(3)
3.2593(5)
3.486(1)
100(2)
4.128(2)
7.558(7)
2.435(1)
3.268(1)
3.507(3)
293 ?
c (Å)
Th−O (Å)
Th−Te (Å)
Th−Te (cap) (Å)
T (K)
R(F)
0.0130
0.042
Optical Measurements. ThOS, ThOSe, and ThOTe are
yellow-, orange-, and black-colored, respectively. Representative
photographs of crystals are shown in Figure 4.
Figure 4. From left to right: Crystals of ThOS, ThOSe, and ThOTe.
The Bader method divides space into atomic regions
determined by zero-flux charge-density surfaces.31 The volume
of integration is not constant for the same atom in all three
compounds as it was in the RWS method. However, the entire
cell volume is taken into account, and all electrons are
accounted for. Calculation of formal oxidation states using the
Bader charges gives unreasonable (i.e., nontraditional)
oxidation states: averaging the values for Th, O, and Q, gives
Th2.4+O1.3−Q1.1−. Despite the problematic charges, conclusions
can be drawn about the electron distribution. As seen from the
data in Table 4, as the size of Q increases, S < Se < Te, Th gains
electrons, O does not change, and Q loses electrons. For O to
retain the same charge, either the electron volume must change
(with a corresponding change in electron density) or the
volume (and electron density) appropriated to O must stay the
same. The latter is more likely as the Th−O distance and the
environment about O are constant in each structure. Because
the Th−Q bond distance increases S < Se < Te, the volume
appropriated to both Th and Q increases. As O and Q are only
bonded to Th and not to each other (the shortest Q−Q
distance being 3.3639(1) Å) it is evident that electron density
moves from Q to Th. This means that more electrons are
localized on the Theither closer to Th than Q in the Th−Q
A fundamental absorption edge is clearly visible for
transmission through the (001) crystal face of single crystals
of ThOS, ThOSe, and ThOTe (Figure 1A). Extrapolation to
the absorption edge shows that ThOQ (Q = S, Se, Te) are
semiconductors. Comparison of plots of absorbance vs energy
to plots of (αhν)1/2 and (αhν)2 vs energy (hν) provides insight
on the nature of the optical transition. Because of the superior
resemblance of the original absorbance data to an indirect
transition plot ((αhν)1/2 vs hν), indirect optical band gaps of
2.22, 1.65, and 1.46 eV are assigned for ThOS, ThOSe, and
ThOTe, respectively. Fits to the direct transition data yield
slightly different values for ThOS and ThOSe of 2.25 and 1.89
eV, respectively. The bandgaps shown in Figures 1B and 1C
were averaged over three measurements for a single crystal of
ThOS and over two measurements for single crystals of ThOSe
and ThOTe (see Supporting Information). No polarization
dependence was observed when the polarization of visible light
was scanned from 0 to 180°, and the experimental values are
consistent with the colors of the compounds.
Theoretical Calculations. Calculated bond valence sums
for the atoms in ThOQ (Q = S, Se, Te) are shown in Table 4.
8115
dx.doi.org/10.1021/ic300510x | Inorg. Chem. 2012, 51, 8112−8118

Inorganic Chemistry
Article
bond or in nonbonding orbitalsand the Th−Q bond
becomes more covalent, S < Se < Te.
Densities of States. Plots of the partial densities of states
correlated with Th−Q covalency. The upper valence band
(UVB, mostly p-character) for Q varies somewhat in shape but
not in overall width. The lower valence bands (LVB, mostly s-
character) are rather narrow owing to lack of overlap/
interaction with Th or O states.
(PDOS) for each atom type in ThOQ are found in Figures 5, 6,
The UVB states from −6 to 0 eV are almost exclusively due
to p-orbitals of O and Q; these heavily overlap, suggesting
strong hybridization between these orbitals. Starting at
approximately −11 eV one finds the rather narrow LVB with
a large Q s-orbital contribution and nearly zero overlap from
Th and O. The ∼2 eV LVB bandwidth is attributed mostly to
Q−Q interactions. The deep-lying states from −21 to −14 eV
are almost exclusively from the p-orbitals mixing with s-, p-, and
(a small amount of) d-orbitals of O. The p- and d-orbitals of O
overlap with the p-orbital of Th but not with their own s-
orbital; that is, there seems to be little O s-p hybridization. The
O 2s-orbital dominates the contribution of overlap with the Th
p-orbital.
The unoccupied states from 0 to 4 eV above EF are
dominated by empty Th f-orbitals with a small but non-
negligible contribution from empty d-orbitals of Th, O, and Q.
Theoretical Band Gaps. In all three compounds the
smallest band gap is direct and occurs at the k-space zone
center Γ with values of 1.09, 0.75, and 0.17 eV for ThOS,
ThOSe, and ThOTe, respectively. The three next highest band
gaps for all three compounds are two indirect transitions from
Γ to (0 0 1/6) = ζ and a ζ − ζ direct gap. The values for these
band gaps are 1.18, 1.19, and 1.28 eV for ThOS; 0.83, 0.85, and
0.93 eV for ThOSe; and 0.23, 0.27, and 0.33 eV for ThOTe.
The low-lying excitations are essentially from chalcogen Q p-
states to Th f-states, in accord with simple ionic models with
the Th4+6p66d07s05f0 configuration. In agreement with experi-
ment, one sees a trend to lower energy down the chalcogen
series. The calculated values are systematically smaller than
experiment, as frequently observed in DFT modeling. The
origins of the band gap errors in DFT are well-known, and
related to the difference in effective potentials seen by occupied
and vacant states. The so-called DFT+U methodology
introduces an explicit correlation term, generally parametrized,
to expand the gap.
For ThOS, electronic states were recalculated after the
introduction of the (somewhat large) Hubbard U parameter of
value 4 eV for Th f-states. In actinide compounds, the Hubbard
U parameter is typically used to introduce a gap between
occupied and vacant f-states. As expected for Th, introducing
the U term for Th f-states had essentially no effect on the
spectra as described by the PDOS (Figures 8A and 8B), as Th
has virtually no occupied f orbitals. The difference in the
calculated band gap turned out to be negligible; the value for
the lowest band gap was 1.13 eV with the Hubbard U term and
1.09 eV without. The same lack of a major effect is expected for
the introduction of the U term in ThOSe and ThOTe. This is
clearly a case where a single-site correlation correction for the
excited state energies is inadequate; multiple-site corrections
are beyond the scope of this work.
Figure 5. Density of States for Th in ThOQ.
Figure 6. Density of States for O in ThOQ.
and 7. Each figure compares the PDOS of a different element
for all three compounds. The PDOS are further separated into
contributions for orbitals of s, p, d, and f character. The PDOS
for both Th and O vary little among the three compounds, with
the shapes and intensities of the bands being the same. They do
vary in energy relative to the Fermi energy, EF, (set to 0 eV) in
the sequence Te < Se < S; a net shift of ∼2 eV clearly is
As described above, low-lying transitions have been
calculated at approximately 1.1 eV for ThOS. However,
examination of the PDOS shows that the spectral weight of
these transitions is very small, and that the main absorption
onset occurs from a Q p-band to a Th f-band at approximately
2 eV, close to values extrapolated from experimental data.
Similarly, the main absorption as evident from the PDOS of
ThOSe and ThOTe occurs at approximately 1.5 and 0.75 eV,
Figure 7. Density of States for Q in ThOQ.
8116
dx.doi.org/10.1021/ic300510x | Inorg. Chem. 2012, 51, 8112−8118

Inorganic Chemistry
Article
their colors. Extrapolation to the absorption edge shows that
each compound is a semiconductor. Calculated bond valence
sums for the atoms in ThOQ (Q = S, Se, Te) lead to average
bond valences for the atoms Th, O, and Q of +4.01, −2.06, and
−1.95, respectively, which support the formulation of ThOQ as
Th4+O2−Q2−. Periodic spin-polarized band structure calcula-
tions were performed with the use of the first principles DFT
program VASP. The calculated ionic charge of each atom from
both RWS volume integration and the Bader topological method
demonstrates that the Th−Q bond becomes more covalent, S <
Se < Te, with increased interstitial and shared electron density
between Th and Q. The calculated band gaps are systematically
smaller than experiment, as is frequently observed in DFT
modeling.
ASSOCIATED CONTENT
■
S
* Supporting Information
Details on the bandgaps shown in Figures 1B and 1C. Spectra
of ThOTe at visible wavelengths and of ThOS at IR
wavelengths. The crystallographic files in CIF format for
ThOS, ThOSe, and ThOTe. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: ibers@chem.northwestern.edu.
Notes
The authors declare no competing financial interest.
Figure 8. Comparison of Density of States for ThOQ (A) without and
(B) with the Hubbard U parameter of 4 eV.
ACKNOWLEDGMENTS
■
The research was supported at Northwestern University by the
U.S. Department of Energy, Basic Energy Sciences, Chemical
Sciences, Biosciences, and Geosciences Division and Division of
Materials Sciences and Engineering Grant ER-15522. L.A.K.
was also supported by the Nuclear Energy University Programs
of the DOE Office of Nuclear Energy. Funding was also
provided by the National Science Foundation Grant CHE-
0911145 and the NSF MRSEC (DMR-0520513) at the
Materials Research Center of Northwestern University. We
thank Prof. Thomas E. Albrecht-Schmitt at the University of
Notre Dame for his donation of Th metal.
respectively. A summary of calculated and experimental band
gaps is found in Table 5.
Table 5. Summary of Theoretical and Experimental Band
Gaps (eV) Found in ThOQ (Q = S, Se, Te)
type
ThOS
ThOSe
ThOTe
a
calculated (four lowest)
direct
1.09
1.18
1.19
1.28
0.75
0.83
0.85
0.93
∼1.5
1.65
1.89
0.17
0.23
indirect
indirect
direct
0.27
0.33
strong onset from PDOS
∼2
2.22
∼0.75
1.46
REFERENCES
b
■
from optical
indirect
direct
(1) Jin, G. B.; Raw, A. D.; Skanthakumar, S.; Haire, R. G.; Soderholm,
L.; Ibers, J. A. J. Solid State Chem. 2010, 183, 547−550.
(2) Trzebiatowski, W.; Niemiec, J.; Sepichowska, A. Bull. Acad. Pol.
Sci., Ser. Sci. Chim. 1961, 9, 373−377.
(3) Marcon, J.-P. C. R. Seances Acad. Sci., Ser. C 1967, 265, 235−237.
(4) Gorum, A. E. Acta Crystallogr. 1957, 10, 144.
2.25
1.46
a
Lowest band gap from the calculation containing the Hubbard U
b
term was 1.13 eV. The indirect transition fits absorbance data better.
(5) Amoretti, G.; Calestani, G.; Giori, D. C. Z. Naturforsch. A: Phys.
Phys. Chem. Kosmophys. 1984, 39, 778−782.
CONCLUSIONS
■
The compounds ThOS, ThOSe, and ThOTe were synthesized
as side products in multiple reactions. Their structures have
been determined by means of single-crystal X-ray diffraction
methods. All three compounds adopt the PbFCl structure type
(6) D’Eye, R. W. M. J. Chem. Soc. 1953, 75, 1670−1672.
(7) Graham, J.; McTaggart, F. K. Aust. J. Chem. 1960, 13, 67−73.
(8) Zachariasen, W. H. Acta Crystallogr. 1949, 2, 291−296.
(9) D’Eye, R. W. M.; Sellman, P. G.; Murray, J. R. J. Chem. Soc. 1952,
74, 2555−2562.
7
in the tetragonal space group D4h − P4/nmm. More precise
(10) Beck, H. P.; Dausch, W. Z. Anorg. Allg. Chem. 1989, 571, 162−
164.
crystallographic data have been obtained for ThOS and ThOSe,
which had previously only been known from X-ray powder
diffraction data. ThOS, ThOSe, and ThOTe are yellow-,
orange-, and black-colored, respectively. From single-crystal
optical absorption measurements the indirect band gaps are
2.22, 1.65, and 1.45 eV, respectively; these are consistent with
(11) Eastman, E. D.; Brewer, L.; Bromley, L. A.; Gilles, P. W.;
Lofgren, N. L. J. Am. Chem. Soc. 1951, 73, 3896−3898.
(12) Boelsterli, H. U.; Hulliger, F. J. Mater. Sci. 1968, 3, 664−665.
(13) Gensini, M.; Gering, E.; Benedict, U.; Gerward, L.; Staun, O. J.;
Hulliger, F. J. Less-Common Met. 1991, 171, L9−L12.
8117
dx.doi.org/10.1021/ic300510x | Inorg. Chem. 2012, 51, 8112−8118

Inorganic Chemistry
Article
(14) Amoretti, G.; Giori, D. C.; Varacca, V.; Spirlet, J. C.; Rebizant, J.
Phys. Rev. B 1979, 20, 3573−3578.
(15) Amoretti, G.; Blaise, A.; Collard, J. M.; Hall, R. O. A.; Mortimer,
M. J.; Troc, R. J. Magn. Magn. Mater. 1984, 46, 57−67.
(16) Witt, R. H.; Nylin, J.; McCullough, H. M. A Study of the Hydride
Process for Producing Thorium Powder; United States Atomic Energy
Commission, Atomic Energy Division, Sylvania Electric Products, Inc.:
Bayside, NY, 1956.
(17) Bruker APEX2 Version 2009.5-1 and SAINT version 7.34a Data
Collection and Processing Software; Bruker Analytical X-Ray Instru-
ments, Inc.: Madison, WI, 2009.
(18) Bruker SMART Version 5.054 Data Collection and SAINT-Plus
Version 6.45a Data Processing Software for the SMART System; Bruker
Analytical X-Ray Instruments, Inc.: Madison, WI, 2003.
(19) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr.
2008, 64, 112−122.
́
(20) Gelato, L. M.; Parthe, E. J. Appl. Crystallogr. 1987, 20, 139−143.
(21) Brese, N. E.; O’Keeffe, M. Acta Crystallogr., Sect. B: Struct. Sci.
1991, 47, 192−197.
(22) Kresse, G.; Hafner, J. Phys. Rev. B 1993, 47, 558−561.
(23) Kresse, G.; Hafner, J. Phys. Rev. B 1994, 49, 14251−14271.
(24) Kresse, G.; Forthmuller, J. Comput. Mater. Sci. 1996, 6, 15−50.
̈
(25) Kresse, G.; Furthmuller, J. Phys. Rev. B 1996, 54, 11169−11186.
̈
(26) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758−1775.
(27) Perdew, J. P. In Electronic Structure of Solids; Ziesche, P.; Eschrig,
H., Eds.; Akademie Verlag: Berlin, Germany, 1991; pp 11−20.
(28) Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188−5192.
(29) Hubbard, J. Proc. R. Soc. London, Ser. A 1963, 276, 238−257.
(30) Dudarev, S. L.; Botton, G. A.; Savrasov, S. Y.; Humphreys, C. J.;
Sutton, A. P. Phys. Rev. B 1998, 57, 1505−1509.
(31) Henkelman, G.; Arnaldsson, A.; Jon
2006, 36, 354−360.
́
sson, H. Comput. Mater. Sci.
(32) Sanville, E.; Kenny, S. D.; Smith, R.; Henkelman, G. J. Comput.
Chem. 2007, 28, 899−908.
(33) Bader, R. F. W. Atoms in molecules: a quantum theory; Oxford
University Press, Inc.: New York, 1990.
(34) Shannon, R. D. Acta Crystallogr., Sect. A: Cryst. Phys. Diffr. Theor.
Gen. Crystallogr. 1976, 32, 751−767.
(35) Taylor, D. Br. Ceram. Trans. J. 1984, 83, 32−37.
(36) Tougait, O.; Potel, M.; Noel, H. Inorg. Chem. 1998, 37, 5088−
̈
5091.
8118
dx.doi.org/10.1021/ic300510x | Inorg. Chem. 2012, 51, 8112−8118