The O-O Bonding and Hydrogen Storage in the Pyrite-type PtO2

Article abstract of DOI:10.1021/acs.inorgchem.9b00046

We have synthesized pyrite-type PtO₂ (py-PtO₂) at 50-60 GPa and successfully recovered it at 1 bar. The observed O-O stretching vibration in Raman spectra provides direct evidence for inter-oxygen bonding in the structure. We also identified the O-H vibrations in py-PtO₂ synthesized from the lowerature areas, indicating hydrogenation, py-PtO₂H_x (x ≤ 1). Diffraction patterns are consistent with a range of degrees of hydrogenation controlled by temperature. We found that py-PtO₂ has a high bulk modulus, 314 ± 4 GPa. The chemical behaviors found in py-PtO₂ have implications for the hydrogen storage in materials with anion-anion bonding, and the geochemistry of oxygen, hydrogen, and transition metals in the deep planetary interiors.

Full text of DOI:10.1021/acs.inorgchem.9b00046

Article
pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
The O−O Bonding and Hydrogen Storage in the Pyrite-type PtO2
Huawei Chen,† Shuxiang Zhou,‡ Dane Morgan,‡ Vitali Prakapenka,§ Eran Greenberg,§
Kurt Leinenweber,∥ and Sang-Heon Shim*,†
†School of Earth and Space Exploration, Arizona State University, Tempe, Arizona 85287-6004, United States
‡Department of Materials Science and Engineering, University of Wisconsin−Madison, Madison, Wisconsin 53706, United States
§GeoSoilEnviroCars, University of Chicago, Chicago, Illinois 60439, United States
∥Eyring Materials Center, Arizona State University, Tempe, Arizona 85287-1604, United States
S
* Supporting Information
ABSTRACT: We have synthesized pyrite-type PtO2 (py-PtO2) at
50−60 GPa and successfully recovered it at 1 bar. The observed O−O
stretching vibration in Raman spectra provides direct evidence for
inter-oxygen bonding in the structure. We also identified the O−H
vibrations in py-PtO2 synthesized from the low-temperature areas,
indicating hydrogenation, py-PtO2Hx (x ≤ 1). Diffraction patterns are
consistent with a range of degrees of hydrogenation controlled by
temperature. We found that py-PtO2 has a high bulk modulus, 314
4 GPa. The chemical behaviors found in py-PtO2 have implications
for the hydrogen storage in materials with anion−anion bonding, and
the geochemistry of oxygen, hydrogen, and transition metals in the
deep planetary interiors.
moduli have been documented in OsO2 and RuO2 in similar
crystal structures (PdF2 or pyrite type).10−12 However, it still
remains controversial whether the anion-to-anion interaction is
responsible for the observations as well as the cation−anion
bonding.5 Strong anion−anion interaction has been also found
in the platinum group dichalcogenides (MQ2, where M = Pt or
Pd and Q = S or Se) with pyrite or distorted pyrite structures.13
Therefore, it is important to investigate a range of compounds
with pyrite-type crystal structures.
Recently, pyrite-type FeO2 has been synthesized from α-
FeOOH at 92 GPa and 2050 K.14 Subsequently, it was proposed
that py-FeO2 can be partially hydrogenated (py-FeO2Hx, 0 ≤ x
≤ 1).15 However, some later studies16,17 have questioned the
stabilities of py-FeO2 and its hydrogenated form and instead
argued for the existence of a fully hydrated form in a pyrite-type
structure, py-FeOOH. These two views lead to dramatically
different consequences for the deep planetary storage of
volatiles. The stability of py-FeO2Hx with the peroxide
bonding14,15 will result in a release of hydrogen at high
temperature, i.e., dehydrogenation. Consequently, the deep
interior could store a large amount of oxygen in a form of a
peroxide unit. In contrast, if fully hydrated without a peroxide
unit, py-FeOOH will be stable and therefore its thermal
decomposition will lead to a supply of H2O instead, i.e.,
dehydration, in the deep planetary interiors.16
INTRODUCTION
■
Several Pt oxides, such as PtO, β-PtO2, and Pt3O4, have been
synthesized through decomposition of chloroplatinic acid or
compression to high pressures.1−4 A multi-anvil study reported
that PtO2 synthesized at 17 GPa and 1273 K has a Pa3
structure.5 While the structure is very similar to pyrite, FeS2, it
was argued that Pa3-type PtO2 does not have sufficient anion-to-
̅
-type
̅
anion interaction expected in the pyrite type (Figure 1a). It was
instead assigned to a PdF2-type structure found at high pressure
(hp-PdF2 type), which does not include any significant anion-to-
anion interaction.
̅
The argument against the anion-to-anion interaction in Pa3-
type PtO2 is based on the bond distances constrained through
Rietveld refinements of the powder X-ray diffraction patterns
measured for the recovered sample at 1 bar and 300 K.5
Although higher-quality measurements are possible at 1 bar and
the symmetry of the phase is higher (and therefore a smaller
number of parameters to fit), powder diffraction can still suffer
from a range of problems for structure determination, including
large degree of preferred orientation which could be severe in
the materials synthesized at high pressures.
The significance of resolving this issue is that the anion−anion
bonding in a crystal structure could dramatically increase the
bulk modulus of materials (also possibly strength) and that
pyrite-type PtN2 has already been shown to contain a pernitride
unit, [N−N]4−, and consequently possess a very high bulk
modulus (347 GPa).6−9 Can Pt dioxide with the same type of
crystal structure possess an O−O bonding (similar to the N−N
bonding in py-PtN2) and a high bulk modulus? Very large bulk
Received: January 8, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.9b00046
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Crystal structure models of py-PtO2 and its hydrogenated forms. (a) py-PtO2, (b) py-PtO2Hx (0 < x < 1), and (c) py-PtO2H. The O−O
bondings are highlighted as green lines in (a). The hydrogen atom (blue spheres) may be inserted between the oxygen atoms in the hydrogenated form,
i.e., (b) and (c).
Figure 2. In situ X-ray diffraction patterns measured during an experiment on a Pt + H2O starting mixture at 60 GPa. (a) Before, (b) during, and (c)
after laser heating to 1250−1800 K. (d) We also measured diffraction patterns after decompression at 1 bar and 300 K. The wavelength of the X-ray
beam is 0.3344 Å. The insets show the 021 peak of py-PtO2.
experiments are listed in Table S1). Diamond anvils with a culet size of
200 μm were used to compress the samples to high pressure. A small
chip of pure elemental gold was loaded near the PtO2 sample foil for in
situ pressure measurements. For the Pt + H2O samples (runs 101 and
203), we used the equation of state of Pt to calculate pressure from its
EXPERIMENTAL METHODS
■
We prepared pure Pt or α-PtO2 (Alfa-Aesar) in a prepressed foil form.
The foil was then loaded in a 90 μm diameter hole in a preindented
rhenium gasket. We injected H2O as a medium to the sample chamber
and compressed in a symmetric-type diamond-anvil cell (DAC) (all the
B
DOI: 10.1021/acs.inorgchem.9b00046
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 3. X-ray diffraction patterns of py-PtO2 and py-PtO2Hx. (a) Diffraction measurements at 1 bar and 300 K. The sample was synthesized at 50
GPa and 1400 K. The bottom most diffraction pattern is for the hot spot, and the rest of the patterns are from areas outside of the hot spot as
schematically shown in an inset in (b). Diffraction peaks from α-PtO2 and the Re gasket are also labeled. (b) The 111 peaks also measured at different
distances from the laser-heated spot.
measured unit-cell volume.18 We placed several spacers less than 10 μm
(either pure Pt or α-PtO2) below and above the sample foil to form a
layer of H2O between the sample and diamond anvils for thermal
insulation during laser heating.
We performed density functional theory (DFT) calculations for the
crystal structure and bulk modulus of py-PtO2. All DFT calculations
were conducted with periodic boundary conditions with a plane-wave
basis set using the Vienna Ab initio Simulation Package (VASP).22,23
Initial atomic coordinates for PtO2 were obtained from the
We measured in situ X-ray diffraction (XRD) patterns in the laser-
heated DACs at beam-line 13ID-D in the GeoSoilEnviroCARS
(GSECARS) sector19 at the Advanced Photon Source (Figures 2 and
3). The monochromatic X-ray beam was focused to a beam size of 3 × 4
μm2 at the sample in the DAC. Two near-infrared laser beams (a
wavelength of ∼1 μm) were focused on the sample through two
opposite sides of the DAC with a hot spot size of 20−25 μm diameter.
Temperatures for laser heating were calculated by fitting the measured
thermal radiation spectra to the Planck equation from both sides of the
sample after subtracting backgrounds. We maintained the same
temperature for both sides of the sample by adjusting the laser
power. The laser beams were aligned coaxially with the X-ray beam, in
order to measure diffraction patterns at the center of the heating spot.
We have collected diffraction data with an X-ray wavelength of 0.3344
or 0.4133 Å. The diffraction setup, including X-ray energy and detector
tilts, was calibrated by measuring the diffraction patterns of a LaB6
standard.
We used a Pilatus detector for 2D diffraction image measurements.
In a typical run, we compressed the sample to a target pressure and then
heated for 10 min. After laser heating, we measured high-quality
diffraction patterns after removing laser mirrors. X-ray exposure time
varies between 1 and 5 s depending on the sample loading. We
integrated the 2D images into 1D diffraction patterns in the Dioptas
package.20 We performed phase identification and peak fitting to a
pseudo-Voigt profile shape function in the PeakPo package.21
We have measured Raman spectra for the recovered samples from
runs 101 and 417 at ASU. The Raman scattering was excited by a 532
nm wavelength beam from a frequency doubled Nd:YAG laser.
Considering the reflectivity and absorption from the optical
components in our Raman system and diamond anvils, the beam
intensity should be less than 30 mW at the surface of the sample. We
conducted measurements at spectral ranges of 400−1200 cm−1 for the
lattice vibrations and 2900−3600 cm−1 for the OH modes, using a 1800
grooves/mm grating. The spectrometer was calibrated using the neon
emission spectra. To remove the pixel-to-pixel sensitivity differences in
the CCD detector, we measured a spectrum of a glass with known
fluorescence intensities at different wavenumbers. Spectral data were
collected for 200 s.
experimental results (space group: Pa3
̅
, No. 205).5 The projector
augmented wave (PAW) method24 as implemented by Kresse and
Joubert25 is used as the electron−ion interactions. The valence electron
configuration for the Pt pseudopotential was 5p65d96s1. The exchange
correlation functional parametrized in the generalized gradient
approximation (GGA)26 by Perdew, Burke, and Ernzerhof (PBE)27
was used, with and without the Heyd−Scuseria−Ernzerhof (HSE)
screened hybrid functional.28 The plane-wave cutoff energy was set to
450 eV, and the stopping criteria for self-consistent loops were 1 meV
per cell for electronic and ionic relaxation. The conventional cell for py-
PtO2 (4Pt and 8O atoms) is used in the structure relaxation and the
EOS calculations with a 15 × 15 × 1 Monkhorst−Pack29 k-point grid
for GGA and a 5 × 5 × 5 Monkhorst−Pack k-point grid for HSE. The
volume and energy convergence of HSE are checked with a 9 × 9 × 9
Monkhorst−Pack k-point grid. The volume difference of the relaxed
crystal between 5 × 5 × 5 and 9 × 9 × 9 k-point grids is 0.003 Å3/atom,
and the energy difference between them is <0.15 meV/atom.
Therefore, the calculation results under 5 × 5 × 5 k-point grids are
̅
converged to the given scales. The bulk modulus of Pa3-type RuO2 is
also obtained from the same calculation approaches and settings to
further compare the results from different potential functionals.
RESULTS AND DISCUSSION
■
In our experiments in a laser-heated diamond-anvil cell, several
new peaks appeared in diffraction patterns within 10 min of
heating for a Pt + H2O starting mixture at 1250−1900 K and 60
GPa (run 101 in Table S1). These sharp peaks can be well-
indexed with a cubic unit-cell. The existence of the 021
̅
diffraction line at ∼9.2° 2θ confirms a Pa3-type structure for the
new phase rather than a face-centered cubic structure (Figure 2).
During the laser heating, the peaks gained intensities but no
additional lines appeared in the patterns. After quenching to 1
bar and 300 K, the diffraction pattern remained essentially the
same except for the shifts in the peak positions to lower 2θ angles
due to decompression, demonstrating that the new structure is
C
DOI: 10.1021/acs.inorgchem.9b00046
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
quenchable to ambient conditions. For this pattern, we
conducted Rietveld refinements30 on the crystal structure at 1
bar as shown in Figure 4. We obtained a unit-cell volume of
118.45(1) Å3. The shortest oxygen-to-oxygen distance was 2.13
Å (run 101 in Table S1).
also successfully quenched to 1 bar and 300 K (Figures 3 and
S2). Within the laser-heated area, all the spots we examined
showed a zero pressure unit-cell volume of 118.136(3) Å3. This
̅
value is essentially the same as the value reported for Pa3-type
PtO2 synthesized at 17 GPa and 1273 K under anhydrous
conditions,5 118.1418(2) Å3. However, we found a systemati-
cally higher unit-cell volume (by 0.3%) for the same phase
synthesized from the Pt + H2O starting material, 118.45(1) Å3
̅
(Figure 4). This volume is similar to that reported for Pa3-type
PtO2 synthesized at lower pressure (118.1418(2) Å3),5 but
larger than our samples from run 417 (118.136(3) Å3)
considering the estimated uncertainty. The larger unit-cell
̅
volume of the Pa3-type PtO2 phase in run 101 likely results from
hydrogen atoms incorporated into the crystal structure, which
we will discuss later.
Possible O−O bonding has been inferred from the O−O
distances measured from X-ray diffraction.31 The shortest O−O
distance from our Rietveld refinements is 2.13 Å at 1 bar (Figure
4). Our O−O distance is slightly smaller than the value reported
̅
for the Pa3-type PtO2 phase synthesized at lower pressures
under anhydrous conditions.5 However, as we discussed earlier,
it is challenging to constrain the bond distance reliably and the
inference to the existence of bonding is indirect and uncertain.
Raman spectroscopy is sensitive to covalent bonds. In pyrite-
type PtN2, N−N bonding had been successfully identified with
the technique.8 The quenchability of Pa3
̅
-type PtO2 allows us to
measure the Raman spectra reliably at 1 bar. Raman spectra
show two intense vibrational modes (860 and 748 cm−1)
together with two much weaker modes (674 and 891 cm−1 in
Figure 5). An intense mode at 860 cm−1 has a frequency near the
O−O symmetric stretch mode (Ag) (750−850 cm−1) reported
for a number of solid peroxides.32 For hydrogen peroxide, the
mode exists at 880 cm−1, again supporting the assignment of the
Figure 4. Rietveld refinement of a diffraction pattern measured for py-
PtO2 after recovery to 1 bar and 300 K. The unit-cell parameter is a =
4.91103(1) Å. The shortest O−O bond distance is 2.13334(1) Å.
mode at 860 cm−1 in Pa3
(therefore, now we call pyrite-type PtO2 instead of Pa3
PtO2). The Eg mode at 748 cm−1 has also been documented in
̅
-type PtO2 to O−O bonding
In a separate experiment (run 417) with an oxidized starting
material instead, α-PtO2 + H2O, we also observed the formation
of the same phase at 1300−1500 K and 55 GPa. The sample was
̅
-type
Figure 5. Raman spectrum of the recovered PtO2 samples at 1 bar and 300 K. (a) The lattice mode region for py-PtO2 from run 101 with the smallest
V0, and (b) the lattice mode and (c) the OH mode.
D
DOI: 10.1021/acs.inorgchem.9b00046
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
pyrite-type phases.8 We also measured Raman spectra for α-
PtO2, which was one of our starting materials. α-PtO2 has two
modes at 505 and 544 cm−1. Thus, none of the existing Raman
modes for py-PtO2 are related to starting α-PtO2. It is notable
that the Raman spectrum of py-PtO2 is very similar to that of py-
PtN2.8 It is likely because the [N−N]4− bonding9 is isoelectronic
to [O−O]2−. Therefore, our Raman data further support the
pyrite type for the high-pressure polymorph of PtO2. A former
higher at 1 bar. The lattice modes from the lower-temperature
areas also show systematic shifts to lower frequencies (Figure
5c). Such decreases can be interpreted as a consequence of
hydrogen incorporation, because hydrogen will likely be located
between the oxygen atoms in the structure, weakening the O−O
bonding. An additional mode was found in this sample at 690
cm−1. Pyrite type has five Raman active modes37 and this may be
the fifth mode, which was not observed in our much less
hydrogenated py-PtO2 sample (Figure 5a). It is feasible that
hydrogen incorporation and consequent distortion in the crystal
structure may make the mode more visible in more hydro-
genated py-PtO2. Therefore, our Raman results support our
interpretation for the observed gradual unit-cell volume increase
in XRD: an increase in hydrogenation of py-PtO2 at lower
temperatures (py-PtO2Hx).
From the observations of the unit-cell volume lying between
those expected for py-FeO2 and py-FeOOH, incorporation of
varying amounts of hydrogen has been inferred for py-FeO2.15
Such a comparison is only possible at pressures above 80 GPa for
py-FeO2, because of its limited stability at lower pressures. A
later study did not find such volume behaviors16 and argued that
other factors, such as deviatoric stresses and pressure gradient, in
high-pressure experiments can contribute to the observations. In
the case of PtO2, owing to the quenchability of the pyrite-type
phase to 1 bar, we were able to characterize such volume
behaviors more reliably in the absence of stress. More
importantly, the existence of hydrogen or hydroxyl is probed
directly through Raman spectroscopy for py-PtO2, while such
measurements remain to be performed for py-FeO2.
Because the unit-cell volume is unknown for fully hydro-
genated py-PtO2H, it is difficult to estimate the degree of
hydrogenation of our py-PtO2 samples through the measured
unit-cell volumes. However, owing to the quenchability, more
direct quantification of hydrogen can be performed in future
experiments for py-PtO2H through a series of hydrothermal
syntheses at different temperatures. The peak splittings found in
some of the diffraction patterns in Figure 3a are likely due to the
existence of py-PtO2 phases with different amounts of hydrogen.
It remains an interesting question whether py-PtO2 can be
hydrogenated gradually or stepwise (existence of a few
energetically favored amounts for hydrogen). However, the
behavior of the most intense 111 peak shown in the figure seems
to indicate a more gradual change in the amount of hydrogen in
py-PtO2. The most robust observation for PtO2 is the existence
of both pyrite-type structured peroxide and its hydrogenated
version.
We measured the unit-cell volume of py-PtO2 synthesized in
runs 101 and 203 during decompression from 65 GPa to 1 bar.
The data were fit to the Birch−Murnaghan equation38 with the
measured volume at 1 bar (V0) and a fixed pressure derivative of
bulk modulus of 4. We found indeed a high bulk modulus of
314(4) GPa (Table 1 and Figure 6). Because these samples are
slightly hydrogenated (inferred from their larger unit-cell
volume at 1 bar) and hydrogenation may decrease the bulk
modulus,39 we believe the bulk modulus for the dehydrogenated
form, py-PtO2, can be even greater. The value we found for py-
PtO2Hx is similar to the high bulk modulus of pyrite structured
materials, in particular PtN2 (347 GPa).8
̅
study assigned hp-PdF2 as a possible structure for Pa3-type
PtO2.5 However, the study did not measure Raman spectra for
the sample and the synthesis conditions are different from ours:
much higher pressure and hydrous conditions in our study.
In the laser-heated diamond-anvil cell, the size of the heated
area is limited, in our case, 20−25 μm in diameter. Outside of the
heated area, the temperature decreases with distance and
reaches ultimately 300 K at a sufficient distance (inset in Figure
3). We measured diffraction patterns as we moved away from the
heating area for the sample recovered from run 417. In Figure 3,
the bottommost diffraction pattern is for the hot spot and the
rest of the patterns are from areas outside of the hot spot as
schematically shown in an inset in (b). With an increase in
distance, the temperature should decrease roughly by 50−100
K/μm. The diffraction lines of the py-PtO2 shift to lower 2θ
angles (or higher d-spacings) with an increase in distance from
the heated area, suggesting an increase in unit-cell volume at
lower-temperature spots (Figure 3). For example, at a spot 9 μm
away from the heated area (estimated temperature of 500−1000
K during heating), we found the largest unit-cell volume (1.6%
higher than that at the heated area) for the py-PtO2 at 1 bar and
300 K.19 While we move out of the heated spot, we found 2−3
additional diffraction lines. They can be indexed with the peaks
expected for unreacted α-PtO2 and Re (gasket). We note that
those peaks did not exist within the heated area at high pressure
(Figure S2).
At pressures above ∼80 GPa, FeO2 and FeOOH are shown to
have the same crystal structure, pyrite type. However, py-
FeOOH has an ∼1.6 Å3/f.u. higher volume than py-FeO2,
suggesting a volume expansion by hydrogen incorporation in the
crystal structure.14,16 Therefore, we can attribute the volume
difference found between the cold and hot areas in run 417 to
the existence of hydrogenated py-PtO2 at low-temperature areas
(py-PtO2Hx).
The py-PtO2 sample was examined further with Raman
spectroscopy (Figure 5c). At the lower-temperature areas, we
detected two peaks at 3255 and 3550 cm−1 in the O−H
stretching vibration region at 1 bar. The wavenumber range is
well-known for the O−H vibration in a range of compounds
with a hydroxyl group. The existence of those two modes is
direct evidence for the structurally incorporated hydrogen
bonded with oxygen atoms in py-PtO2. Hydrogen may attack the
O−O bonds in the pyrite-type structure and be incorporated
there at the center of the O−O bond (i.e., a bond
symmetrization), as suggested in computations.16,33,34 Hydro-
gen bond symmetrization has also been observed in high-
pressure H2O ice35 where hydrogen is bonded with two oxygen
atoms in contrast to a single oxygen atom at 1 bar in water. The
O−H stretching vibration frequency for such a configuration is
expected to be much lower.36 However, we note that our Raman
measurements were conducted for the samples metastably
decompressed to 1 bar. Therefore, even if symmetrized O−H−
O existed at high pressure, the hydrogen atom can move away
from the center position during decompression and con-
sequently the O−H vibrational frequency can be significantly
We calculated the bulk modulus of py-PtO2 using density
functional (DFT) theory (both GGA and HSE; Table 1). We
also calculated for RuO2 in the Pa3
̅
structure, which is also
-type
known to have a high bulk modulus.31 We note that the Pa3
̅
RuO2 has the same crystal structure as py-PtO2, but the
E
DOI: 10.1021/acs.inorgchem.9b00046
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
type RuO2.31 It was also found that the bulk modulus of py-PtO2
from HSE agrees better with our experimental value and
therefore supports a high bulk modulus of py-PtO2.
Table 1. Equation of State and Crystal Structure of py-PtO2
Constrained from Our High-Pressure Experiments (Expt)
and Density Functional Theory (DFT) Calculations (Calc)
a
a0 (Å) V0 (Å3) O(8c) B0 (GPa)
B′0
CONCLUSION AND IMPLICATIONS
■
pyrite Pa3
̅
-type PtO2
Through Raman spectroscopy and X-ray diffraction, we have
shown that py-PtO2 has a strong O−O interaction. In addition,
we showed that py-PtO2 can store hydrogen. Together with the
similar case suggested for py-FeO2HX,14,15 it can be hypothe-
sized that the energy difference may be sufficiently small
between the peroxide configuration, [O−O]2−, and the
symmetrized hydrogen configuration, [O−H−O]3−, in the
transition metal dioxides (Figure 1c). In addition, the flexible
oxidation states (+2, +3, and +4, respectively) of transition
metals should also contribute to such behaviors. Considering the
Expt5
Expt, this work
Calc (GGA)40
Calc (GGA), this work
Calc (HSE), this work
4.91
4.91
5
4.97
4.89
118.1
118.3
125
122.7
0.34
0.375
0.341
0.345
0.343
314
4.0*
237
312
5.9
5.3
116.1
̅
Pa3-type RuO2
Expt5
Expt10,31
Calc (GGA)40
Calc (GGA)41
Calc (GGA), this work
Calc (HSE), this work
4.85
4.83
4.9
4.88
4.88
4.81
114.2
112.8
117.5
116.5
116
0.35
0.347
0.351
0.344
0.35
399
3.5
299
279
326
3.9
4.8
4.8
̅
fact that the Pa3-type phase has already been found in some
transition metal dioxides,5 it is of particular interest if this
behavior can be common among the transition metal dioxides.
It is of interest if the hydrogen atom in the pyrite-type
transitional dioxides can be symmetrized at high pressure. The
oxygen−oxygen bonding distance at 1 bar is about 2.13(1) Å in
py-PtO2, and the distance at 80 GPa is 1.94 Å in py-FeO2.14
These distances are larger than the O−O dimer distance (1.49
Å) at 1 bar,32 which might be caused or favored by having a
hydrogen atom between the oxygen atoms in the crystal
structure. The [O−H−O]3− could be symmetrized at high
pressure as the symmetric hydrogen bonding will have an O−O
distance of 2.42 Å for the case of H2O ice.35 However, with
decompression, it is more likely to have asymmetric regular
hydrogen bonding as indicated in the case of high-pressure H2O
ice35 and our Raman observation for py-PtO2Hx.
111.3
0.351
a
̅
We also performed similar calculations for Pa3-type RuO2 for
comparison. We present other works for comparisons.31,40,41 a0: unit-
cell parameter; V0: unit-cell volume; O(8c): the fractional coordinate
for O in 8c, (x, x, x); B0: bulk modulus; and B′0: pressure derivative of
bulk modulus at 1 bar (* fixed during fitting).
̅
PtO2 with a Pa3 structure has been synthesized at pressures as
low as 17 GPa.5 Given the fact that the stability of the lower-
pressure polymorphs had been so far demonstrated only up to
6−7 GPa,4 py-PtO2 could be made at even lower pressures than
17 GPa, opening up possibilities for synthesizing larger amounts
of the sample. The most intriguing property we found is the
partial hydrogenation of py-PtO2 together with the incompres-
sibility. The synthesis of the partially hydrogenated version of
py-PtO2 was demonstrated to be straightforward: hydrothermal
synthesis with temperature control as shown in our experiments.
The high incompressibility of py-PtO2 makes such hydrogen
storage capability even more intriguing. Furthermore, the partial
hydrogenation may open up the possibility to control the
properties of py-PtO2, such as strength, through hydrogenation
(i.e., through weakening of the O−O bonding). It is important
to further investigate if hydrogen storage can be found in other
transition metal dioxides or even dinitrides, particularly with
lighter transition metals, with anion−anion bonds in future
studies. For designing such materials, it will be key to understand
the electronic configurations of the cations and the anion-anion
dimers in pyrite-type structures. While such efforts have been
made recently for py-FeO2,42,43 because of its limited stability
only at high pressures, detailed investigation is difficult. The
quenchability and the accessible synthesis pressure for a large
amount of sample in py-PtO2 will provide unique opportunities
to understand the properties we discussed above in much more
detail at room conditions.
Figure 6. Compressibility of the pyrite-type PtO2. We plotted the
normalized unit-cell volume with respect to the volume at 1 bar, V/V0,
as a function of pressure for pyrite-type PtO2. The red circles are
experimentally measured volumes at high pressure and 300 K, and the
red curve is the fitting result with the Birch−Murnaghan equation. The
yellow and blue curves are the compression of py-PtO2 from our GGA
and HSE calculations, respectively.
existence of strong O−O interaction is still under debate
̅
(therefore, Pa3-type RuO2). Our calculation results of the
relaxed crystal structures show good agreements with both
experiments and other calculations.5,31,40,41 However, the bulk
moduli calculated by HSE are 50−80 GPa larger than those by
GGA for both Pa3
̅
-type RuO2 and py-PtO2. This discrepancy is
If such anion-to-anion covalency enhances with compression,
it would also have implications for Earth and the large Earth-like
exoplanets, super-earths. Together with the implications for the
oxygen and hydrogen geochemistry discussed in a previous
study,14 if such behavior is common in transition metal dioxides,
because the calculated energies by GGA and HSE have a large
discrepancy from the use of a hybrid functional in HSE.
Although still lower, HSE provides a bulk modulus in better
̅
agreement with the experimentally measured value for the Pa3-
F
DOI: 10.1021/acs.inorgchem.9b00046
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(3) Moore, W. J.; Pauling, L. The Crystal Structures of the Tetragonal
Monoxides of Lead, Tin, Palladium, and Platinum. J. Am. Chem. Soc.
1941, 63 (5), 1392−1394.
the increased covalent anion-to-anion bonding at high pressure
will impact the geochemical behaviors, such as partitioning of
transition metals (Ir, Ru, Rh, and Os) which are often used as
tracers to understand the deep processes. Oxygen is the main
anion for the phases stable in such planets. Therefore, the
possibility of increased covalency would affect the equations of
state of the phases (such as compressibility) and consequently
alter the mass-radius relations, which are fundamental tools in
inferring the composition of the exoplanets in astrophysical
measurements. It can also imply that the geochemical cycle of
the super-earths could be fundamentally different from that of
smaller versions (e.g., Earth) if such pressure driven processes
(dimerization of anions) can alter the chemical behaviors of
materials at high pressures.
(4) Range, K.-J.; Rau, F.; Klement, U.; Heyns, A. M. β-PtO2: High
Pressure Synthesis of Single Crystals and Structure Refinement. Mater.
Res. Bull. 1987, 22 (11), 1541−1547.
(5) Shirako, Y.; Wang, X.; Tsujimoto, Y.; Tanaka, K.; Guo, Y.;
Matsushita, Y.; Nemoto, Y.; Katsuya, Y.; Shi, Y.; Mori, D.; et al.
Synthesis, Crystal Structure, and Electronic Properties of High-
Pressure PdF2-Type Oxides MO2 (M = Ru, Rh, Os, Ir, Pt). Inorg.
Chem. 2014, 53 (21), 11616−11625.
(6) Zhao, Z.; Xu, B.; Tian, Y. Recent Advances in Superhard Materials.
Annu. Rev. Mater. Res. 2016, 46 (1), 383−406.
(7) Salamat, A.; Hector, A. L.; Kroll, P.; McMillan, P. F. Nitrogen-Rich
Transition Metal Nitrides. Coord. Chem. Rev. 2013, 257 (13), 2063−
2072.
(8) Crowhurst, J. C.; Goncharov, A. F.; Sadigh, B.; Evans, C. L.;
Morrall, P. G.; Ferreira, J. L.; Nelson, A. J. Synthesis and Character-
ization of the Nitrides of Platinum and Iridium. Science 2006, 311
(5765), 1275−1278.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.9b00046.
■
S
(9) Wessel, M.; Dronskowski, R. Nature of N-N Bonding within High-
Pressure Noble-Metal Pernitrides and the Prediction of Lanthanum
Pernitride. J. Am. Chem. Soc. 2010, 132 (7), 2421−2429.
(10) Haines, J.; Leger, J. M.; Schmidt, M. W.; Petitet, J. P.; Pereira, A.
S.; Da Jornada, J. A. H.; Hull, S. Structural Characterisation of the Pa3-
Type, High Pressure Phase of Ruthenium Dioxide. J. Phys. Chem. Solids
1998, 59 (2), 239−243.
X-ray diffraction pattern from Run 417. Raman spectra of
α-PtO2 starting material. Table summary for experimental
runs. Method for Rietveld refinements (PDF)
́
(11) Leger, J. M.; Haines, J.; Blanzat, B. Materials Potentially Harder
AUTHOR INFORMATION
Corresponding Author
*E-mail: SHDShim@asu.edu.
ORCID
Huawei Chen: 0000-0001-8463-6855
Dane Morgan: 0000-0002-4911-0046
Notes
■
than Diamond: Quenchable High-Pressure Phases of Transition Metal
Dioxides. J. Mater. Sci. Lett. 1994, 13 (23), 1688−1690.
̈
(12) Lundin, U.; Fast, L.; Nordstrom, L.; Johansson, B.; Wills, J. M.;
Eriksson, O. Transition-Metal Dioxides with a Bulk Modulus
Comparable to Diamond. Phys. Rev. B: Condens. Matter Mater. Phys.
1998, 57 (9), 4979−4982.
́
(13) Soulard, C.; Rocquefelte, X.; Petit, P.-E.; Evain, M.; Jobic, S.; Itie,
J.-P.; Munsch, P.; Koo, H.-J.; Whangbo, M.-H. Experimental and
Theoretical Investigation on the Relative Stability of the PdS - and
2
The authors declare no competing financial interest.
Pyrite-Type Structures of PdSe . Inorg. Chem. 2004, 43 (6), 1943−
2
1949.
ACKNOWLEDGMENTS
(14) Hu, Q.; Kim, D. Y.; Yang, W.; Yang, L.; Meng, Y.; Zhang, L.; Mao,
H.-K. FeO2 and FeOOH under Deep Lower-Mantle Conditions and
Earth’s Oxygen-Hydrogen Cycles. Nature 2016, 534 (7606), 241−244.
(15) Hu, Q.; Kim, D. Y.; Liu, J.; Meng, Y.; Yang, L.; Zhang, D.; Mao,
W. L.; Mao, H. Dehydrogenation of Goethite in Earth’s Deep Lower
Mantle. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (7), 1498−1501.
(16) Nishi, M.; Kuwayama, Y.; Tsuchiya, J.; Tsuchiya, T. The Pyrite-
Type High-Pressure Form of FeOOH. Nature 2017, 547 (7662), 205−
208.
(17) Yuan, L.; Ohtani, E.; Ikuta, D.; Kamada, S.; Tsuchiya, J.; Naohisa,
H.; Ohishi, Y.; Suzuki, A. Chemical Reactions Between Fe and H 2 O up
to Megabar Pressures and Implications for Water Storage in the Earth’s
Mantle and Core. Geophys. Res. Lett. 2018, 45 (3), 1330−1338.
(18) Ye, Y.; Prakapenka, V.; Meng, Y.; Shim, S.-H. Intercomparison of
the Gold, Platinum, and MgO Pressure Scales up to 140 GPa and 2500
K. J. Geophys. Res. Solid Earth 2017, 122, 3450−3464.
(19) Prakapenka, V. B.; Kubo, A.; Kuznetsov, A.; Laskin, A.;
Shkurikhin, O.; Dera, P.; Rivers, M. L.; Sutton, S. R. Advanced Flat
Top Laser Heating System for High Pressure Research at GSECARS:
Application to the Melting Behavior of Germanium. High Pressure Res.
2008, 28 (3), 225−235.
(20) Prescher, C.; Prakapenka, V. B. DIOPTAS: A Program for
Reduction of Two-Dimensional X-Ray Diffraction Data and Data
Exploration. High Pressure Res. 2015, 35 (3), 223−230.
(21) Shim, S.-H. PeakPo - A python software for X-ray diffraction
analysis at high pressure and high temperature; 2017. https://zenodo.
org/record/810200.
■
We thank three anonymous reviewers for their helpful
comments. The work has been supported by the Keck
Foundation (PI: P. Buseck), NASA (80NSSC18K0353), and
NSF (EAR1338810). The results reported herein benefit from
collaborations and/or information exchange within NASA’s
Nexus for Exoplanet System Science (NExSS) research
coordination network sponsored by NASA’s Science Mission
Directorate. The synchrotron experiments were conducted at
GSECARS, Advanced Photon Source (APS). GSECARS is
supported by NSF-Earth Science (EAR-1128799) and DOE-
GeoScience (DE-FG02-94ER14466). APS is supported by
DOE-BES, under contract DE-AC02-06CH11357. Computa-
tions in this work benefit from the use of the Extreme Science
and Engineering Discovery Environment (XSEDE), which is
supported by National Science Foundation − United States,
grant no. OCI-1053575. The experimental data for this paper are
available by contacting hchen156@asu.edu.
REFERENCES
■
(1) Adams, R.; Shriner, R. L. Platinum Oxide as a Catalyst in the
Reduction of Organic Compounds. III. Preparation and Properties of
the Oxide of Platinum Obtained by the Fusion of Chloroplatinic Acid
with Sodium Nitrate. J. Am. Chem. Soc. 1923, 45 (9), 2171−2179.
(2) Muller, O.; Roy, R. Formation and Stability of the Platinum and
Rhodium Oxides at High Oxygen Pressures and the Structures of
Pt3O4, β-PtO2 and RhO2. J. Less-Common Met. 1968, 16 (2), 129−
146.
(22) Kresse, G.; Furthmuller, J. Efficient Iterative Schemes for Ab
̈
Initio Total-Energy Calculations Using a Plane-Wave Basis Set. Phys.
Rev. B: Condens. Matter Mater. Phys. 1996, 54 (16), 11169−11186.
G
DOI: 10.1021/acs.inorgchem.9b00046
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(23) Kresse, G.; Hafner, J. Ab Initio Molecular Dynamics for Liquid
Metals. Phys. Rev. B: Condens. Matter Mater. Phys. 1993, 47 (1), 558−
561.
(24) Blochl, P. E. Projector augmented-wave method. Phys. Rev. B:
Condens. Matter Mater. Phys. 1994, 50 (24), 17953.
(25) Kresse, G.; Joubert, D. From Ultrasoft Pseudopotentials to the
Projector Augmented-Wave Method. Phys. Rev. B: Condens. Matter
Mater. Phys. 1999, 59 (3), 1758−1775.
(26) Langreth, D. C.; Perdew, J. P. Theory of Nonuniform Electronic
Systems. I. Analysis of the Gradient Approximation and a General-
ization That Works. Phys. Rev. B: Condens. Matter Mater. Phys. 1980, 21
(12), 5469−5493.
(27) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient
Approximation Made Simple. Phys. Rev. Lett. 1996, 77 (18), 3865−
3868.
(28) Krukau, A. V.; Vydrov, O. A.; Izmaylov, A. F.; Scuseria, G. E.
Influence of the Exchange Screening Parameter on the Performance of
Screened Hybrid Functionals. J. Chem. Phys. 2006, 125 (22), 224106.
(29) Monkhorst, H. J.; Pack, J. D. Special Points for Brillouin-Zone
Integrations. Phys. Rev. B 1976, 13 (12), 5188−5192.
(30) Toby, B. H.; Von Dreele, R. B. GSAS-II: The Genesis of a
Modern Open-Source All Purpose Crystallography Software Package. J.
Appl. Crystallogr. 2013, 46 (2), 544−549.
(31) Haines, J.; Leger, J. M.; Schulte, O. Pa3 Modified Fluorite-Type
Structures in Metal Dioxides at High Pressure. Science 1996, 271
(5249), 629−631.
(32) Eysel, H. H.; Thym, S. RAMAN Spectra of Peroxides. Z. Anorg.
Allg. Chem. 1975, 411 (2), 97−102.
(33) Zhu, S.; Hu, Q.; Mao, W. L.; Mao, H.; Sheng, H. Hydrogen-Bond
Symmetrization Breakdown and Dehydrogenation Mechanism of
FeO2H at High Pressure. J. Am. Chem. Soc. 2017, 139 (35), 12129−
12132.
(34) Tsuchiya, J.; Tsuchiya, T. First-Principles Prediction of a High-
Pressure Hydrous Phase of AlOOH. Phys. Rev. B: Condens. Matter
Mater. Phys. 2011, 83 (5), 054115.
(35) Goncharov, A. F.; Struzhkin, V. V.; Somayazulu, M. S.; Hemley,
R. J.; Mao, H. K. Compression of Ice to 210 Gigapascals: Infrared
Evidence for a Symmetric Hydrogen-Bonded Phase. Science 1996, 273
(5272), 218−220.
(36) Tsuchiya, J.; Tsuchiya, T.; Wentzcovitch, R. M. Vibrational
Properties of -AlOOH under Pressure. Am. Mineral. 2008, 93 (2−3),
477−482.
(37) Vogt, H.; Chattopadhyay, T.; Stolz, H. J. Complete First-Order
Raman Spectra of the Pyrite Structure Compounds FeS2, MnS2 AND
SiP2. J. Phys. Chem. Solids 1983, 44 (9), 869−873.
(38) Birch, F. Finite Elastic Strain of Cubic Crystals. Phys. Rev. 1947,
71 (11), 809−824.
(39) Nisr, C.; Leinenweber, K.; Prakapenka, V.; Prescher, C.;
Tkachev, S.; Shim, S.-H. D. Phase Transition and Equation of State
of Dense Hydrous Silica up to 63 GPa. J. Geophys. Res. Solid Earth 2017,
122, 6972−6983.
́
(40) Tse, J. S.; Klug, D. D.; Uehara, K.; Li, Z. Q.; Haines, J.; Leger, J.
M. Elastic Properties of Potential Superhard Phases of RuO2. Phys. Rev.
B: Condens. Matter Mater. Phys. 2000, 61 (15), 10029−10034.
(41) Jain, A.; Ong, S. P.; Hautier, G.; Chen, W.; Richards, W. D.;
Dacek, S.; Cholia, S.; Gunter, D.; Skinner, D.; Ceder, G. Commentary:
The Materials Project: A Materials Genome Approach to Accelerating
Materials Innovation. APL Mater. 2013, 1 (1), 011002.
(42) Liu, J.; Hu, Q.; Bi, W.; Yang, L.; Xiao, Y.; Chow, P.; Meng, Y.;
Prakapenka, V. B.; Mao, H.-K.; Mao, W. L. Altered Chemistry of
Oxygen and Iron under Deep Earth Conditions. Nat. Commun. 2019,
10 (1), 153.
(43) Boulard, E.; Harmand, M.; Guyot, F.; Lelong, G.; Morard, G.;
Cabaret, D.; Boccato, S.; Rosa, A. D.; Briggs, R.; Pascarelli, S. Ferrous
Iron Under Oxygen-Rich Conditions in the Deep Mantle. Geophys. Res.
Lett. 2019, 46 (3), 1348−1356.
H
DOI: 10.1021/acs.inorgchem.9b00046
Inorg. Chem. XXXX, XXX, XXX−XXX