Pyrite formation via kinetic intermediates through low-temperature solid-state metathesis

Article abstract of DOI:10.1021/ja5081647

The preparation of materials with limited phase stabilities yet high kinetic activation barriers is challenging. Knowledge of their possible formation pathways aids in addressing these challenges. Metathesis reactions present an approach to circumvent these barriers; however, solid-state metathesis reactions are often too rapid from extensive self-heating to understand the reaction. The stoichiometric reaction of MCl₂ salts (M = Mn, Fe, Co, Ni, Cu, Zn) with Na₂S₂ enables the formation of pyrite (FeS₂), CoS₂, and NiS₂ at low temperatures (250-350 °C). Na₂S₂ has the same polyanionic dimer as found in the pyrite structure, which would suggest the possibility of a facile ion-exchange reaction. However, from high-resolution synchrotron X-ray diffraction and differential scanning calorimetry, the energetic driving force does not appear to result solely from NaCl formation but also from formation of intermediate and pyrite phases. It is apparent that the reaction proceeds through polyanionic disproportionation and formation of a low-density alkali-rich intermediate, followed by anionic comproportionation and atomic rearrangement into the pyrite phase. These results have profound implications for the use of low-temperature metathesis in achieving materials by design.

Full text of DOI:10.1021/ja5081647

Article
pubs.acs.org/JACS
Pyrite Formation via Kinetic Intermediates through Low-
Temperature Solid-State Metathesis
Andrew J. Martinolich and James R. Neilson*
Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523-1872, United States
ABSTRACT: The preparation of materials with limited phase stabilities yet
high kinetic activation barriers is challenging. Knowledge of their possible
formation pathways aids in addressing these challenges. Metathesis reactions
present an approach to circumvent these barriers; however, solid-state
metathesis reactions are often too rapid from extensive self-heating to
understand the reaction. The stoichiometric reaction of MCl salts (M = Mn,
2
Fe, Co, Ni, Cu, Zn) with Na S enables the formation of pyrite (FeS ), CoS ,
2
2
2
2
and NiS at low temperatures (250−350 °C). Na S has the same polyanionic
2
2 2
dimer as found in the pyrite structure, which would suggest the possibility of a
facile ion-exchange reaction. However, from high-resolution synchrotron X-ray
diffraction and differential scanning calorimetry, the energetic driving force
does not appear to result solely from NaCl formation but also from formation
of intermediate and pyrite phases. It is apparent that the reaction proceeds
through polyanionic disproportionation and formation of a low-density alkali-rich intermediate, followed by anionic
comproportionation and atomic rearrangement into the pyrite phase. These results have profound implications for the use of
low-temperature metathesis in achieving materials by design.
INTRODUCTION
metric completion due to the formation of many other sulfur
10,11
■
deficient compounds;
excess sulfur is often used to generate
The formation of solids illustrates the competition of
thermodynamics and kinetics: while there is often a
thermodynamic driving force to form a compound, the kinetic
barriers from solid-state diffusion make design of materials
the high fugacity needed to drive stoichiometric completion of
12
the reaction. Many transition metal disulfides form pyrite
polymorphs, ranging across the period from MnS to ZnS ;
2
2
1
however, MnS , CuS , and ZnS are all high-pressure
2 2 2
1
extremely challenging. Therefore, the solid-state chemist often
3−15
phases.
The properties of pyrite change dramatically
relies on high-temperature reactions in order to produce a
desired phase, assuming that the desired phase is thermody-
namically stable. Alternatively, self-heating solid-state meta-
thesis reactions have been used extensively in the formation of
refractory and superhard compounds, such as sulfides, nitrides,
borides, carbides, and silicides.
solid-state metathesis reaction does not proceed immediately
with rapid self-heating and instead only proceeds with the
application of heat over a period of time. Along these lines,
metathesis has been integrated into aerosol-assisted formation
pathways of nanomaterials. These formation pathways have
also been used to form transition metal chalcogenide materials,
and further integration of metathesis may allow control over
the size and shape of resulting nanoparticles, which may in turn
with metal substitution, ranging from antiferromagnetic
1
4,16−19
insulators to superconductors.
With the appropriate
synthetic tools in hand, it may become possible to prepare
novel related phases that would be otherwise unattainable
through metallurgical preparatory routes.
Here, the salt-exchange metathesis reactions of metal
dihalides with disodium disulfide
2
−6
With some precursors, the
7
Na S + MCl → MS + 2NaCl
(
1)
2
2
2
2
2+
2+
2+
8
where M is a divalent 3d transition metal (Mn , Fe , Co ,
2+
2+
2+
Ni , Cu , and Zn ) are investigated. The formal reaction, as
written, would correspond to a simple salt exchange, where the
2
−
+
dimeric persulfide [S ] polyanion exchanges two Na cations
2
2+
9
for M , as the pyrite lattice is comprised of the octahedral
20
alter their properties as the crystallite size decreases. The use
coordination of M2+ by [S2] dimers (Figure 1). This
metathetical reaction at low temperature (Trxn = 350 °C) yields
formation of high-purity compounds that are thermodynami-
cally stable at ambient pressures (FeS , CoS , and NiS ). While
2−
of solid-state metathesis at low temperatures, under conditions
without extensive self-heating, presents many opportunities for
the solid-state chemist to maintain increased control over
reactions to master the formation of designer materials.
A compound of great interest for energy applications and
2
2
2
this metathesis reaction reaches stoichiometric completion in
the end, high-resolution synchrotron powder X-ray diffraction
geosciences is iron pyrite, FeS . As a compound that melts
2
(
HRSXRD) data suggest a far more complex mechanism than
incongruently (Tdec = 743 °C) and is comprised of high-
melting iron and volatile sulfur, its preparation from the
elements at ambient pressure can require long reaction times at
high temperature and is not guaranteed to achieve stoichio-
Received: August 14, 2014
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/ja5081647 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Schematic description of the metathesis reaction. Upon heating, iron chloride and disodium disulfide react to form sodium chloride and
iron disulfide in the solid state. The reaction proceeds through metathesis, forming sodium iron sulfide and sulfur as reaction intermediates.
crystal chemistry would suggest, perhaps through reductive
recombination to the elements or ionic exchange as previously
hypothesized.
formation is kinetically slow below 100 °C, due to the requisite
diffusion through the solid matrix. Furthermore, a pressed
pellet of FeCl and Na S slowly darkens from pale yellow to
2
,7,21,22
2
2 2
black over the course of ∼2 months in either an inert or dry O
2
RESULTS
environment at room temperature, indicative that a reaction is
taking place very slowly.
■
From differential scanning calorimetry (DSC), significant
exothermic events occur on heating the reaction, even after
the formation of NaCl (Figure 2a). Both thermodynamic and
When an exotherm is triggered by heating to 175 °C, the
reaction is far from complete, even after 24 h of equilibration, as
indicated by the small mole fraction of FeS visible in the
2
product (Figure 2b,c and Table 1). At temperatures greater
than 175 °C, there is sufficient thermal energy (for diffusion
and other activated processes) as to allow for reaction
completion.
HRSXRD of the reaction of FeCl and Na S at 350 °C
2
2 2
reveals quantitative reaction completion and high-purity FeS2.
Rietveld analysis of the HRSXD data reveals 68.4(2) mol %
NaCl and 30.4(2) mol % FeS (Figure 3a). The remaining
2
23
1
.3(2) mol % is monoclinic NaFeS2. The NaFeS and NaCl
2
were washed away with anhydrous methanol, which resulted in
9.63(1)% phase-pure pyrite FeS , along with a 0.37(7) mol %
9
2
of Fe S impurity (Figure 2b); this small phase fraction was
7
8
further confirmed by determination of the ferromagnetic
response of the sample, which also confirms less than 1
mol % of a Fe S impurity.
7
8
The Fe S phase was not observed in the diffraction data
7
8
before washing and may have formed through decomposition
of the NaFeS or minor oxidation of the pyrite FeS upon
2
2
washing. Electronic transport measurements were performed
on cold-pressed pellets and show an increase in resistivity upon
cooling, confirming the material’s bulk semiconducting nature.
While the reaction achieves stoichiometric completion by
Trxn = 350 °C (i.e., with seemingly little activation barrier), it
does not do so through topochemical transformations or
diffusive salt exchange. There is extensive redox chemistry that
takes place, as indicated by the formation of several crystalline
intermediates. HRSXRD data of the products from incomplete
reactions (Trxn = 50 and 175 °C) yield insight into the reaction
pathway (Figure 2c). The diffraction patterns have appreciable,
diffuse, noncrystalline backgrounds, as well as broad humps that
sharpen at Trxn = 350 °C and index to the expected reflections
of monoclinic NaFeS . The observation of NaFeS is surprising,
Figure 2. (a) DSC of the reaction of FeCl with Na S shows multiple
2
2 2
exothermic processes as the temperature of the reaction increases. The
data are normalized per gram of total reaction mixture. (b) PXRD
patterns (black symbols) and corresponding fits determined through
the Rietveld method (orange lines) of the reaction at the temperatures
shown. NaCl formation is seen throughout, while FeS is not seen
2
until Trxn = 250 °C. (c) HRSXRD patterns of the same reaction
products show S formation at low reaction temperatures, as well as a
8
significant diffuse background component indicative of a disordered
phase, which crystallizes into NaFeS . [Data plotted as Q = 4π sin(θ)/
2
λ; λ = 1.5418 Å (b) and λ = 0.41367 Å (c).]
2
2
2
+
3+
as it would indicate the formal oxidation of Fe to Fe and
reduction of [S2]2 to 2 S ; there are no persulfide dimers in
−
2− 24
kinetic effects are required to reconcile the DSC and powder X-
ray diffraction (PXRD) data. From PXRD of a 24 h reaction,
NaCl forms by 50 °C (Figure 2b); however, the DSC trace with
the structure of NaFeS that would indicate a lesser negative
2
charge on the sulfur.
−1
a heating rate of 1 °C min shows that all exothermic
processes occur above ∼100 °C (Figure 2a). This suggests that
there is a low activation barrier for NaCl formation, but the
At both intermediate temperatures, there are also numerous,
sharp, low-Q reflections, all of which index to orthorhombic
25
2−
S , suggesting that there is also formal oxidation of the [S2]
8
B
dx.doi.org/10.1021/ja5081647 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Phase Fractions at Different Reaction Temperatures from Analysis of the Reaction Products of Eq 1 by High-
2
+
2+
Resolution Synchrotron PXRD (SXRD) and Laboratory PXRD (PXRD) for M = Fe and Co
2+
M = Fe ; mol % (PXRD/SXRD)
FeS2 S8
Trxn, °C
NaCl
NaFeS2
5
0
98(1)/86.1(1)
97(1)
0/0
0
1.4(2)/7.25(3)
0.8(2)/6.70(5)
0.45(4)
1
1
2
3
00
75
50
50
3.0(3)
88(2)/87.0(1)
60(3)
0/3.49(5)
39(2)
1.4(5)/3.97(3)
10.8(5)/5.59(4)
0
0
70(2)/68.3(1)
30(1)/30.37(6)
0/0
0/1.28(2)
2+
M = Co ; mol % (PXRD)
Trxn, °C
NaCl
CoS2
Na CoS2
2
CoS
100
175
250
350
97(1)
92(1)
85(2)
0
2.5(7)
8(1)
0
0
0
0.7(4)
a
14(3)
13(4)
0
b
73(1)
0
11.0(9)
a
b
Also includes 0.36(4) mol % Co S . Also includes 2.7(4) mol % Co S .
9
8
3 4
Figure 4. PXRD patterns of the products from the solid-state
metathesis reactions of CoS2 with Na2S2 (black circles) and
corresponding Rietveld analyses (orange lines) at various temper-
Figure 3. HRSXRD and Rietveld analysis of the products of reaction 1
a) before and (b) after washing with anhydrous methanol, showing
(
atures. At low temperatures, S8 and Na CoS2 are detected. At
2
the data (black circles), calculated fit (orange lines), and difference
between the two (black lines). Ticks below the data indicate expected
reflections. Asterisks in part a denote monoclinic NaFeS2 [1.3(2)
mol %]. Asterisks in part b denote Fe S [0.37(7) mol %]. Minority
increased temperatures, the formation of sulfur-deficient phases Co S ,
3
4
CoS, and Co S are detected, despite the stability of CoS up to
9
8
2
950 °C.
7
8
phases were incorporated into the refinements; their reflections were
omitted for clarity.
pathway may be altered by the increased mobility of the sulfur.
As the reaction temperature is increased, we infer that the
volatilization of the sulfur intermediate overcomes the
0
to 2S . However, crystalline sulfur is not present in the
HRSXRD data of the 350 °C reaction and did not condense in
the sealed ampule; this suggests that sulfur also participates as
an intermediate in the reaction mechanism. By HRSXRD, no
pyrite is observed at 50 °C, but crystallites are detected in the
Trxn = 175 °C products (Figure 2c, Table 1).
Similar trends are seen in the temperature-dependent
reaction of CoCl and Na S (Figure 4). At low temperatures,
Na CoS is observed, which has an analogous structure to
favorability of CoS formation, leading to these sulfur-deficient
2
phases.
In reactions with MCl , (M = Mn, Fe, Co, Ni, Cu, and Zn),
2
only FeS , CoS , and NiS pyrite are formed (Figure 5). Only
2
2
2
these pyrite-type compounds are found on the binary phase
1
0,26−30
diagrams determined for near ambient pressures.
While
2
2 2
pyrite MS phases have been reported at nonambient pressures
2
2
2
13−15
for manganese, copper, and zinc,
the thermodynamically
NaFeS ; the extra sodium compensates for the divalent charge
2
of Co2 with sulfide (S ) anions. While comparable to the
+
2−
favorable phases of MnS, CuS, and ZnS are observed, with the
remnant sulfur condensing on the side of the ampule as it
cooled according to the reaction
reaction with iron chloride, CoS formation is more sensitive to
temperature. At Trxn = 350 °C, CoS formation is accompanied
by the sulfur-deficient phases CoS and Co S . Despite CoS
2
2
3
4
2
26
Na S + MCl → MS + S(g) + 2NaCl
2
2
2
(2)
being stable up to 950 °C, progression through this reaction
C
dx.doi.org/10.1021/ja5081647 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
which is not likely slow enough to observe completion
formation of NaCl below 100 °C; therefore, the formation of
NaCl at lower temperatures is not likely. Furthermore, there is
a significant increase in the NaCl diffraction intensity (with a
decrease in peak width) when comparing the Trxn = 50 and 175
°
C diffraction data. The same trend is observed during the
formation of CoS (Figure 4). The comparison of the products
2
from these two reaction temperatures also shows a significant
increase in the amount of NaFeS (or Na CoS ) present in the
2
2
2
product (Table 1). Furthermore, there is not a substantial
increase in the intensity of the diffraction from NaCl between
the Trxn = 175 and 350 °C diffraction data in either the iron- or
cobalt-containing products; however, quantification of this
change is not straightforward. Together, these imply that the
initial exotherm is the initiation of the reaction through
formation of NaCl and NaFeS (or Na CoS ), with a low
2
2
2
activation barrier that is rapidly overcome by heating to 100 °C:
E ∼100°C
1
2
a
FeCl + Na S ⎯⎯⎯⎯⎯⎯⎯⎯⎯ ⎯→ NaFeS + NaCl + Cl (g)
2
2
2
2
2
(3)
The theoretical enthalpy of this reaction is ΔH = −2.8 kJ
Figure 5. PXRD patterns of the products from the solid-state
metathesis reactions of divalent transition metal chlorides with Na2S2
rxn
−1 33
−1
mol (since ΔH°(NaFeS ) = −331.3 kJ mol
). Such a
f
2
(
black circles) and corresponding Rietveld analyses (orange lines)
reaction is barely exothermic, which would explain the small
exotherm observed in DSC that is concomitant with the
display the formation of either MS or MS , as shown by the expected
2
reflections under each pattern. The formation of NaCl is conserved.
CuCl and wurtzite ZnS were also incorporated into the respective
Rietveld refinements; ticks were omitted for clarity.
observation of NaFeS and NaCl. The thermochemistry of the
2
reaction
1
2
NaFeS + Cl (g) → FeS + NaCl
2
2
2
(4)
The reaction with ZnCl produced a mixture of the high- and
2
low-temperature polymorphs of ZnS, with the wurtzite and zinc
blende structures, as previously shown.
is strongly exothermic relative to eq 3, as ΔH = −258.4 kJ
rxn
7
−1
mol . The presence (or absence of) Cl (g) [or S (s)] does not
2
8
influence the thermochemical calculation; therefore, it is clear
that the significant exotherm observed above T > 175 °C must
DISCUSSION
■
By correlating features in the DSC to changes in the observed
phases, we can elucidate the elusive mechanism for these low-
temperature solid-state metathesis reactions. Primarily, the
melting or decomposition of any reagents during the reaction
can be ruled out. Na S does not decompose until 470 °C, and
arise from the formation of FeS and NaCl.
2
From a control reaction, it was confirmed that Cl gas is not
2
evolved. The reaction was performed in the presence of a
sodium metal getter, with the reaction mixture in a crucible on
top of a crucible containing the sodium and both crucibles
2
2
3
1,32
FeCl does not melt or decompose until 677 °C.
While
sealed into an evacuated SiO ampule. The expectation would
2
2
3
1
Na S undergoes a structural phase transition at 170 °C,
be that if chlorine gas were being produced, NaCl would form
on the surface of the liquid sodium metal. After reaction at 350
2
2
crystalline Na S is not observed in any of the X-ray diffraction
2
2
2−
3+
patterns. This precludes formation of either S or Fe ions
through decomposition of the reactants and must result from
their reaction.
°C for 24 h, no NaCl was formed, but instead Na S was found
2
in the getter crucible. Accordingly, the reaction mixture of
FeCl and Na S formed FeS and NaCl. Since Cl was not
2
2
2
2
Sodium chloride provides a driving force to initiate the
reaction; however, it does not provide a force to drive the
reaction to completion. NaCl formation is observed by XRD at
lower temperatures (Trxn = 50 °C) than the exothermic release
just above T = 100 °C observed from DSC (Figure 2).
Furthermore, the exothermic processes shown here contrast the
formation of pyrite NiS2 from the previously reported
metathesis reaction of K NiF and 2Na S ·0.06H O. There, a
detected, it provides evidence that the chlorine must be
integrated into the low-density NaFeS intermediate.
2
After heating the mixture of Na S and FeCl at 175 °C for
2
2
2
24 h, one observes NaFeS , S , and NaCl. From the analysis of
2
8
density functional theory calculations, NaFeS is reported to be
2
unstable with respect to decomposition into FeS , FeS, and
2
34
Na S. This is further confirmed by our attempts to prepare
2
NaFeS , which were unsuccessful. Preparation of related
2
6
2
5
2
2
vigorous exothermic reaction occurs around T ∼ 65 °C, with
compounds is best achieved by reaction of the alkali with the
transitional metal chalcogenide as a prereaction, followed by
rxn
−1
21
a peak heat flow of ∼1700 mW g . Here, the maximum heat
−1
35−39
flow is ∼200 mW g , reflecting a weaker overall driving force
and the ability to follow the reaction intermediates. Taken
together, these results indicate that a small amount of thermal
energy is required for diffusion, which is sufficient to initiate the
formation of NaCl, but not the pyrite phase.
homogenization and reaction at elevated temperature.
Reaction of FeS with sodium metal yielded only FeS , Na S,
2
2
2
and a sulfur-deficient FeS(2−x) phase. Instead, preparation of
NaFeS2 requires the leaching of aqueous suspensions or
2
3,24,33
electrochemical means.
These findings further implicate
The exothermic release just above T = 100 °C corresponds
to the formation of NaCl and of a poorly ordered NaFeS2
phase. In the DSC experiment, the heating rate was 1 °C/min,
NaFeS as an intermediate phase in the reaction.
Collectively, these data provide insight into the complete
mechanism of pyrite formation via metathesis. The metathesis
2
D
dx.doi.org/10.1021/ja5081647 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
reaction does not appear to be limited by solid-state diffusion,
even at room temperature, and the reaction does not appear to
be susceptible to thermal runaway. Rather than the simple salt
CONCLUSION
■
Solid-state metathesis can be used to prepare pyrite metal
disulfides at low temperature and without extensive self-heating
or thermal runaway. While robust for a number of other
transition metal disulfides, it appears that a major driving force
for product formation is not primarily the formation of the
metathesis byproduct, NaCl. Instead, from a combination of
HRSXRD and DSC, formation of the thermodynamically stable
pyrite phases drives the overall reaction, while a low activation
barrier for NaCl formation and favorable intermediate
formation enables initiation of the reaction, without extensive
self-heating. Through careful analysis of the reaction
intermediates, the metathesis involves significant atomic
rearrangements in the formation of alkali-rich metal sulfides
at low temperatures. This approach provides wide implications
for the preparation of novel compounds or low decomposition
temperature materials that have substantial activation barriers
that prevent their facile formation.
2
+
2−
exchange between Fe and [S ] , there appears to be a more
2
complicated reaction mechanism predicated on the formation
of low-density crystalline intermediates. From XRD, it is
apparent that NaCl forms first, while the Fe and S combine to
form the intermediates NaFeS and S: FeCl + Na S → xNaCl
2
2
2 2
+
yNaFeS + zS → FeS + 2NaCl. However, this does not
2 2
stoichiometrically balance with chlorine, and chlorine gas was
not detected.
Therefore, we speculate that the reaction may proceed
through the formation of NaCl and attack of the FeCl lattice
2
by sulfur through dispropotionation of the persulfide anion,
2−
2−
0
[
S ] → S + S :
2
E ∼100°C
a
FeCl + Na S ⎯⎯⎯⎯⎯⎯⎯⎯⎯ ⎯→ (2 − x)NaCl
2
2 2
+
Na1
FeS
Cl + xS
−0.5x
(2−x)
x
(5)
(6)
EXPERIMENTAL SECTION
■
∼
175°C
All reagents were purchased from Sigma-Aldrich or NOAH
Technologies at a purity of 98% or higher. Sulfur was additionally
purified through vapor transport in an evacuated sealed tube with a
temperature gradient from 550 to 400 °C. The impurities were left in
the hot end of the tube while sulfur condensed in the cooler end of the
tube as the vessel cooled.
⎯
⎯⎯⎯⎯⎯ ⎯→ FeS + 2NaCl
2
The first step, eq 5, has a low activation barrier; it can take place
even at room temperature over the course of several months.
Evidence in favor of such a mechanism with the incorporation
Na S was prepared from the elements in a manner with which to
of Cl into NaFeS is provided by the poor crystallinity of the
2 2
2
avoid direct contact of the elements, which could produce an
exothermic thermite reaction. In an argon-filled glovebox, sodium and
sulfur were placed in separate alumina crucibles and sealed in the same
silica tube, with the sulfur on top. The tube was then evacuated to <10
mTorr and sealed without exposure to the atmosphere. The tube was
then heated to 200 °C for 24 h to allow sulfur vapor to corrode the
sodium. The resulting heterogeneous solid was homogenized and
pelleted in the glovebox, resealed, and heated for an additional 12 h at
300 °C.
observed NaFeS phase, in addition to the distinct, larger lattice
2
parameters of the monoclinic NaFeS phase relative to the
2
−
previously reported compound, as to accommodate the Cl
ions, which provide less electrostatic screening between metal
4
0,41
cations.
Such a mechanism, derived by considering only the
observed crystalline products, requires just partial oxidation of
2
+
3+
Fe to Fe during the early stages of FeCl attack, as the iron
2
valence is given by Fe(
3−x/2)+
, where x is given in eq 5.
For the solid-state metatheses of MS , anhydrous MCl and Na S
2
2
2 2
However, our observations and analysis do not preclude the
formation of additional amorphous intermediates.
Furthermore, these redox activities can be rationalized as a
reaction between the two intermediates, elemental sulfur and
were ground to a homogeneous powder and pelletized in an argon-
filled glovebox. The pellet was then placed in an alumina crucible and
sealed in a quartz tube, evacuated to <10 mTorr without exposure to
the atmosphere. The reaction vessel was then heated at 1 °C min− in
a box furnace to the specified temperature for 24 h and the furnace
cooled. The product was then recovered in an argon glovebox, washed
five times with anhydrous methanol under an inert atmosphere using
Schlenk techniques, dried under vacuum at room temperature, and
then recollected in the glovebox.
1
NaFeS . Comproportionation of the sulfur can occur when the
2
2
−
3+
0
S
anion reduces both the Fe and S to form pyrite:
S2− + 2Fe3+ + S → 2FeS2
0
3
(7)
Purity of the Na S was confirmed by X-ray diffraction using a
2
2
From Rietveld refinements of the HRSPXRD data, the nominal
Bruker APEX II single crystal diffractometer (Mo Kα radiation, λ =
0.7107 Å) with powder in a sealed quartz capillary; the two-
dimensional data were radially integrated for analysis. Powder X-ray
diffraction data (PXRD) were collected on a Scintag X-2
diffractometer with Cu Kα radiation. High-resolution synchrotron
powder X-ray diffraction data were collected from powders sealed in
quartz capillaries loaded into Kapton capillaries on the beamline 11-
3+
2−
0
ratios of Fe :S :S are 1:2:1 at 50 °C and 1.375:2.75:1 at 175
C. Therefore, discrepancies between the stoichiometry of the
°
balanced chemical reaction and the observed values may be
attributed to the poor crystallinity or sulfur deficiency of the
NaFeS intermediate.
2
The removal of NaFeS is predicated on the formation of
42
2
BM-B (λ = 0.413 673 Å). Rietveld analyses were performed using
NaCl, which leads to the formation of FeS . The reactions with
43,44
2
EXPGUI/GSAS.
Physical properties measurements were per-
M = Co2 or Ni also have suitable intermediates with the
+
2+
formed using a Quantum Design, Inc. Dynacool PPMS, using the
electronic transport option and vibrating sample magnetometer.
Differential scanning calorimetry was performed on samples hermeti-
cally sealed in the glovebox using a TA Instruments 2920 modulated
alkali-rich phases related to the structures of Na CoS or
2
2
K Ni S , the former of which was detected in the low-
2
3 4
temperature reactions of CoCl and Na S . The resulting metal
2
2 2
−1
DSC with a 1 °C min ramp rate in order to replicate the reaction
sulfides are all thermodynamically stable phases at ambient
pressure, but not necessarily at ambient temperature, as shown
by the formation of both zinc blende and wurtzite ZnS, along
with NaCl and sulfur. This reveals the possibility of a
synthetically controlled route toward formation of nonambient
and ambient phases through metathesis.
conditions.
AUTHOR INFORMATION
■
Corresponding Author
james.neilson@colostate.edu
E
dx.doi.org/10.1021/ja5081647 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Notes
(29) Lee, B.-J.; Sundman, B.; Kim, S. I.; Chin, K.-G. ISIJ Int. 2007,
7, 163−171.
4
(
The authors declare no competing financial interest.
30) Sharma, K.; Chang, Y. J. Phase Equilib. 1996, 17, 261−266.
(
31) Sangster, J.; Pelton, A. J. Phase Equilib. 1997, 18, 89−96.
ACKNOWLEDGMENTS
■
(32) Patnaik, P. Handbook of Inorganic Chemicals; McGraw-Hill: New
The authors thank the excellent staff at the 11-BM-B beamline
for their rapid and efficient collection of the synchrotron X-ray
diffraction data. The research reported in this publication was
supported by the Energy Institute at Colorado State University.
Use of the Advanced Photon Source at Argonne National
Laboratory was supported by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences, under
Contract No. DE-AC02-06CH11357.
York, 2003.
(
33) Lassin, A.; Piantone, P.; Crouzet, C.; Boden
Appl. Geochem. 2014, 45, 14−24.
34) Jain, A.; Ong, S. P.; Hautier, G.; Chen, W.; Richards, W. D.;
́
an, F.; Blanc, P.
(
Dacek, S.; Cholia, S.; Gunter, D.; Skinner, D.; Ceder, G.; Persson, K.
A. APL Mater. 2013, 1, 011002.
(
35) Neilson, J. R.; McQueen, T. M. J. Am. Chem. Soc. 2012, 134,
750−7757.
36) Caron, J. M.; Neilson, J. R.; Miller, D. C.; Llobet, A.; McQueen,
T. M. Phys. Rev. B 2011, 84, 180409(R).
37) Caron, J. M.; Neilson, J. R.; Miller, D. C.; Arpino, K. E.; Llobet,
A.; McQueen, T. M. Phys. Rev. B 2012, 85, 180405(R).
38) Neilson, J. R.; Llobet, A.; Stier, A. V.; Wu, L.; Wen, J.; Tao, J.;
7
(
(
REFERENCES
■
(
1) Committee for an Assessment of and Outlook for New Materials
(
Synthesis and Crystal Growth, National Research Council (US).
Frontiers in Crystalline Matter: From Discovery to Technology; The
National Academies Press: Washington, DC, 2009.
Zhu, Y.; Tesanovic, Z. B.; Armitage, N. P.; McQueen, T. M. Phys. Rev.
B 2012, 86, 054512.
(39) Neilson, J. R.; McQueen, T. M.; Llobet, A.; Wen, J.; Suchomel,
(
2) Bonneau, P. R.; Jarvis, R. F.; Kaner, R. B. Nature 1991, 349, 510−
M. R. Phys. Rev. B 2013, 87, 045124.
5
12.
(
(
40) Boller, H.; Blaha, H. Monatsh. Chem. 1983, 114, 145−154.
41) Neilson, J.; Kurzman, J.; Seshadri, R.; Morse, D. Chem.Eur. J.
(
(
3) Gillan, E. G.; Kaner, R. B. Chem. Mater. 1996, 8, 333−343.
4) Chung, H.-Y.; Weinberger, M. B.; Levine, J. B.; Kavner, A.; Yang,
2
010, 16, 9998−10006.
J.-M.; Tolbert, S. H.; Kaner, R. B. Science 2007, 316, 436−439.
(
42) Wang, J.; Toby, B. H.; Lee, P. L.; Ribaud, L.; Antao, S. M.;
(
5) Nartowski, A. M.; Parkin, I. P.; Craven, A. J.; MacKenzie, M. Adv.
Kurtz, C.; Ramanathan, M.; Von Dreele, R. B.; Beno, M. A. Rev. Sci.
Instrum. 2008, 79, 085105.
Mater. 1998, 10, 805−808.
(
(
6) Parkin, I. P. Chem. Soc. Rev. 1996, 25, 199−207.
(
(
43) Toby, B. H. J. Appl. Crystallogr. 2001, 34, 210−213.
7) Shaw, G. A.; Morrison, D. E.; Parkin, I. P. J. Chem. Soc., Dalton
44) Larson, A.; Von Dreele, R. General Structure Analysis System
Trans. 2001, 1872−1875.
8) Mann, A. K. P.; Skrabalak, S. E. Chem. Mater. 2011, 23, 1017−
022.
9) Motl, N. E.; Mann, A. K. P.; Skrabalak, S. E. J. Mater. Chem. A
(
GSAS); Los Alamos National Lab: Los Alamos, NM, 2004.
(
1
(
2
(
(
013, 1, 5193−5202.
10) Waldner, P.; Pelton, A. J. Phase Equilib. Diffus. 2005, 26, 23−38.
11) Yu, L.; Lany, S.; Kykyneshi, R.; Jieratum, V.; Ravichandran, R.;
Pelatt, B.; Altschul, E.; Platt, H. A. S.; Wager, J. F.; Keszler, D. A.;
Zunger, A. Adv. Energy Mater. 2011, 1, 748−753.
(
12) Seefeld, S.; Limpinsel, M.; Liu, Y.; Farhi, N.; Weber, A.; Zhang,
Y.; Berry, N.; Kwon, Y. J.; Perkins, C. L.; Hemminger, J. C.; Wu, R.;
Law, M. J. Am. Chem. Soc. 2013, 135, 4412−4424.
(
13) Bither, T.; Prewitt, C.; Gillson, J.; Bierstedt, P.; Flippen, R.;
Young, H. Solid State Commun. 1966, 4, 533−535.
14) Bither, T. A.; Bouchard, R. J.; Cloud, W. H.; Donohue, P. C.;
Siemons, W. J. Inorg. Chem. 1968, 7, 2208−2220.
15) Bither, T.; Donohue, P.; Cloud, W.; Bierstedt, P.; Young, H. J.
Solid State Chem. 1970, 1, 526−533.
16) Corliss, L. M.; Elliott, N.; Hastings, J. M. J. Appl. Phys. 1958, 29,
(
(
(
3
(
(
91−392.
17) Ennaoui, A.; Tributsch, H. Sol. Cells 1984, 13, 197−200.
18) Cheng, S. F.; Woods, G. T.; Bussmann, K.; Mazin, I.; Soulen, R.
J., Jr.; Carpenter, E.; Das, B.; Lubitz, P. J. Appl. Phys. 2003, 93, 6847−
849.
19) Gautier, F.; Krill, G.; Lapierre, M.; Panissod, P.; Robert, C.;
6
(
Czjzek, G.; Fink, J.; Schmidt, H. Phys. Lett. A 1975, 53, 31−33.
(20) Goodenough, J. B. J. Solid State Chem. 1971, 3, 26−38.
(21) Bonneau, P. R.; Shibao, R. K.; Kaner, R. B. Inorg. Chem. 1990,
2
(
2
(
9, 2511−2514.
22) Bonneau, P. R.; Jarvis, R. F.; Kaner, R. B. Inorg. Chem. 1992, 31,
127−2132.
23) Taylor, P.; Shoesmith, D. W. Can. J. Chem. 1978, 56, 2797−
802.
2
(
(
24) Taft, C. A. J. Phys. (Paris) 1977, 38, 1161−1162.
25) Abrahams, S. C. Acta Crystallogr. 1955, 8, 661−671.
(
26) Kuznetsov, V.; Sokolova, M.; Palkina, K.; Popova, Z. Inorg.
Mater. 1965, 1, 617−632.
27) Franzen, H. In Binary Alloy Phase Diagrams, 2nd ed.; Massalski,
T., Ed.; 1990; Vol. 3, pp 2593−2597.
28) Waldner, P.; Pelton, A. D. Z. Metallkd. 2004, 95, 672−681.
(
(
F
dx.doi.org/10.1021/ja5081647 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX