Conversion Reactions of Solids: From a Surprising Three-Step Mechanism towards Directed Product Formation

Article abstract of DOI:10.1002/chem.201505209

Directed conversion reactions from binary to multinary compounds are discovered from the reaction of Bi₂S₃ and Bi₂Se₃ with NiCl₂·6 H₂O in polyol media under basic conditions. Control of the synthesis conditions allows the preparation of NiBiSe and superconducting Ni₃Bi₂S₂ and Ni₃Bi₂Se₂. The formation of Ni₃Bi₂S₂ from Bi₂S₃ is found from an unexpected three-step reaction path with Bi and NiBi as intermediates. In the more complex Ni/Bi/Se system, the mechanism found can be used to selectively direct the reaction between the competing ternaries and to suppress side-product formation. Contrary to solid-state reactions (500-900 °C) control of product formation is reached at reaction temperatures and times between 166-300 °C and 0.5-10 h, respectively. The formation of different phases is discussed from results of DFT calculations.

Full text of DOI:10.1002/chem.201505209

DOI: 10.1002/chem.201505209
Full Paper
&
Nanotechnology
Conversion Reactions of Solids: From a Surprising Three-Step
Mechanism towards Directed Product Formation
[b, c]
[c]
[c]
[a, c]
Stefan Michael Rommel,
Alexander Krach, Philipp Peter, and Richard Weihrich*
Dedicated to Professor Henri Brunner on the occasion of his 80th birthday
Abstract: Directed conversion reactions from binary to
multinary compounds are discovered from the reaction of
Bi S and Bi Se with NiCl ·6H O in polyol media under basic
mechanism found can be used to selectively direct the reac-
tion between the competing ternaries and to suppress side-
product formation. Contrary to solid-state reactions (500–
9008C) control of product formation is reached at reaction
temperatures and times between 166–3008C and 0.5–10 h,
respectively. The formation of different phases is discussed
from results of DFT calculations.
2
3
2
3
2
2
conditions. Control of the synthesis conditions allows the
preparation of NiBiSe and superconducting Ni Bi S and
3
2 2
Ni Bi Se . The formation of Ni Bi S from Bi S is found from
3
2
2
3
2
2
2 3
an unexpected three-step reaction path with Bi and NiBi as
intermediates. In the more complex Ni/Bi/Se system, the
Introduction
organic solids can be attained at mild conditions and low tem-
[
10,19]
peratures.
They were developed as effective methods to
Inorganic chemistry of Bi compounds not only attracts atten-
produce phase-pure micro- or nanoscale inorganic com-
pounds. To study and direct reaction paths therein, conversion
[1]
[2]
tion because of a fascinating variety of clusters, nano, and
[3,4]
[9,20–22]
intermetallic structures.
but it also serves as a key element
and pseudomorph reactions turn out to be useful.
in innovative thermoelectric materials and topological insula-
Thereby, precursor particles are transformed into desired com-
pounds and properties can be changed or added. Examples
are the conversion of Co nanocrystals into Co S and Co S
9 8
[5–7]
tors, such as Bi Se , BaBiO , and Bi Rh I .
The latter com-
2
3
3
14
3 9
bines several structural features; to identify these and other Bi
compounds, computational chemistry is increasingly ap-
3
4
nanocrystals in the presence of sulfur in solution by a simple
[
5–9]
[22]
plied.
synthesis have been developed,
the study of metastable Bi-containing phases, as exemplified
To obtain them experimentally, new methods of
diffusion-based mechanism. By multistep paths, also multi-
[7,10]
[23]
which have even allowed
nary intermetallic compounds are obtained.
As part of our ongoing investigations into functional conver-
sions, SnS particles were recently shown to react with
[
11]
for NiBix. Related Bi compounds seem thus well suited to
further explore and elucidate directed reaction paths and
[
24]
NiCl ·6H O in alkaline ethylene glycol at 1978C. By adjusting
2
2
[
12]
[25]
mechanisms. Knowledge therein is not only desired in case
of competing product formation but also with respect to
reductive conditions Ni Sn S
2 2
and core-shell Ni@SnS particles
3
are obtained. Their formation and binary side products depend
on kinetics and equilibriums of dissolution and redox reactions.
A diffusion-driven transformation was concluded similar to the
[13–14]
environmental demands.
For a long time, mechanisms of formation of inorganic
solids remained a “black box”. A new decade therein started
with upcoming solution-based methods. By the polyol
[
26]
conversion of PbS to Ni Pb S . In both cases, binary chalco-
3
2 2
genide precursors AX are converted to perovskite-related
[
15–16]
[17–18]
[9,24–27]
route
or the synthesis in ionic liquids
a number of in-
M A X =M AX compounds
by adding the transition-
3
2
2
1.5
[
28]
metal Ni. As from solid-state reactions pyrite-type PtSnS is
formed from SnS+Pt, hypothetical “NiAX” was computed from
total-energy DFT calculations. For A=Sn, Pb and X=S, Se the
[
a] Prof. Dr. R. Weihrich
Departement of Chemistry, TU Munich
Lichtenbergstr. 4, 85474 Garching (Germany)
E-mail: richard.weihrich@chemie.uni-regensburg.de
NiAX phases are predicted to be less stable than the Ni AX
1.5
[
29]
compounds.
However, the question appeared on directed
[
b] S. M. Rommel
Institute for Materials Resource Management, University of Augsburg
Universitätsstraße 1a, 86159 Augsburg (Germany)
syntheses of possible metastable or competing products on
the path AX+Ni.
[
c] S. M. Rommel, A. Krach, P. Peter, Prof. Dr. R. Weihrich
Institute of Inorganic Chemistry, University of Regensburg
Universitätsstraße 31, 93040 Regensburg (Germany)
This question is picked up subsequently for conversion
reactions of the binary semiconductors Bi S and Bi Se with
2
3
2
3
NiCl ·6H O in ethylene glycol (eg) or tetraethylene glycol (TEG).
2
2
Supporting information and the ORCID identification number(s) for the au-
thor(s) of this article can be found under http://dx.doi.org/10.1002/
chem.201505209.
[
30]
First, the reaction mechanism of the reported transformation
[
25,31,32]
of Bi S to superconducting Ni Bi S
2 2
is studied in ethyl-
2
3
3
Chem. Eur. J. 2016, 22, 6333 – 6339
6333
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ene glycol. A proposed diffusion mechanism of nickel into
Bi S seemed unlikely because of the change in the Bi/S ratio
2
3
and the unexplained formation of elementary Bi that is ob-
served in the diffraction pattern. Subsequently presented
polyol-mediated reactions show a surprising multistep mecha-
nism for the transformation of Bi S to Ni Bi S . Therein, the for-
2
3
3
2 2
[11,33]
mation of Bi and NiBi
plays an important role. In a next
step, knowledge of the mechanism is applied for directed con-
2
+
versions of Bi Se with Ni . The aim was to find separate reac-
2
3
tion pathways to the competing products Ni Bi Se and NiBiSe
3
2
2
and to suppress other side products. By a similar approach, the
synthesis of still unknown NiBiS was attempted. The
competing phase stability of the 3:2:2 and 1:1:1 phase was
analyzed by total-energy DFT calculations.
The reported results not only serve as an example for direct-
ed reactions to adjust the M content along the AX+M path-
way, they also show the successful use of binary intermetallic
Figure 2. Schematic representation of the found reaction pathways from
Bi X (X=S, Se) to Bi, NiBi, NiBiSe, and Ni Bi X .
2 3 3 2 2
“
templates” for the formation of ternary products, which is still
[
34]
a rare method. The evaluated multistep reaction mechanism
may help to study and direct further reactions of Bi-containing
materials in “soft” solid-state chemistry.
ite (Ni Bi Se ) and pyrite (NiBiSe, Figure 1, paths 2 and 3). Upon
3 2 2
a possible transformation of NiBiSe to Ni Bi Se (=Ni BiSe),
3
2
2
3/2
the Ni content is increased (Figure 1, path 4).
All compounds under consideration are summarized in
Figure 2 to point out the structural differences of educts and
products. Contrary to the two-dimensional BiÀX (X=S, Se)
structures in Bi S and Bi Se , Bi and Se form BiÀSe dumbbells
Results and Discussion
2
3
2
3
Studied reaction paths
[9,28]
in pyrite related NiSbS-type NiBiSe.
Ni atoms are found in
octahedrons NiBi Se . The Ni Bi X compounds exhibit CsCl-like
The studied compositional and structural transformations are
rationalized from the phase diagram (Figure 1). Along the line
AX+M, bulk powders of cubic PbS and orthorhombic SnS can
be transformed with Ni at elevated temperatures (T>4008C,
path 1) in pure solid-state reactions into trigonal-layered shan-
dite-type Ni Pb S and isotypic Ni Sn S . This path is also used
3
3
3
2 2
BiX substructures with Ni atoms in half of the available tetrag-
onal bipyramides X Bi . Because of the structural relations to
2
4
O-deficite perovskites, the monoclinic parkerites (C2/m) and
trigonal shandites (R 3¯ m) were classified as half anti-
[
27]
perovskites.
They differ in Ni-site occupation and contain
3
2
2
3
2 2
[
9c]
low-dimensional [Ni X ] substructures related to P and As
under mild conditions (T=166–1978C) from SnS and PbS
3 2
0
2
+ [24,26]
with the rare situation of Ni in linear XÀNiÀX coordination.
Conversion chemistry in the BiÀNiBiÀNi Bi S system
nanoparticles in a reductive solution of Ni
.
A diffusion-
driven mechanism was proven for the formation of the terna-
ries, whereas core-shell particles Ni@SnS are formed by a more
rapid reduction. A similar mechanism for the conversion of the
orthorhombic low-gap semiconductors Bi S and Bi Se to the
3
2 2
The unknown mechanism for path 2 with X=S is studied first
2
3
2
3
and compared to the results of Shao and co-workers who pro-
layered monoclinic superconductor Ni Bi X (X=S, Se), howev-
3
2 2
[
30]
posed a simple diffusion-driven reaction of nickel into Bi2S
3.
er, requires a change in the Bi:X ratio. For the reaction of Bi Se
2
3
2
+
However, the formation of Ni Bi S thereby seems strange as
with Ni , even two products compete with the known parker-
3
2 2
not only an element (here Ni) is added but also one must be
lost (expected: S). A different mechanism is indicated by the
occurrence of elementary Bi as a side product that is observed
in the diffraction pattern, but not commented. Based on this
ambiguity, our first aim was to clarify a more probable
mechanism of the reaction.
The reported (one-pot) low-temperature self-template route
to Ni Bi S starts with Bi S /NiCl ·6H O (molar ratio 1:3) in an
3
2
2
2
3
2
2
ethylene glycol (eg)/ethylenediamine (en) mixture and addi-
tional NaOH to increase the reduction potential of the solu-
[
30]
tion.
By a reproduction of the one-pot approach under
slightly modified conditions, the starting material Bi S is al-
2
3
ready entirely converted to Ni Bi S and Bi as a side phase after
3
2 2
3
0 min at 1668C (Figure S1a and S1b, Supporting Information).
This shows that the reaction already starts at even lower tem-
peratures than thought before. After extending the reaction
Figure 1. Investigated reaction paths in the MÀAÀX system.
Chem. Eur. J. 2016, 22, 6333 – 6339
www.chemeurj.org
6334
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
time to 10.5 h at 1978C, Bi cannot be detected anymore in the
XRD. It seems that the excess Bi was dissolved and washed
out. In principle one could conclude on a self-template mecha-
nism as proposed by Shao and co-workers. The ribbon-shaped
structure of the reactant Bi2S3 and the obtained product
Ni Bi S can be seen in Figure S2, Supporting Information.
3
2 2
However, after sintering the product in an evacuated ampoule
Scheme 1. Multistep mechanisms to different ternaries in the evaluated
NiÀBiÀS/Se system.
at 300–5008C, Bi S and Bi are still found as side phases, which
2
3
were not crystalline enough to be detected before. One must
conclude that the reaction is not just nickel diffusion into Bi S ,
2
3
[33]
but proceeds over multiple, different steps. Therefore Bi2S3
might not act as the template during the reaction.
nism of the one-pot synthesis most probably takes the same
route over NiBi as the intermediate and reactive template, as
the occurring Bi side phase can be explained hereby. This is
To elucidate the questioned mechanism of formation, the re-
evaluated one-pot synthesis was split into three parts
underlined as NiBi and NiBi compounds also occur in the one-
3
(
Figure 3). The exact amount of solvents and reagents that de-
pot synthesis under large Ni excess (SR4). The morphologies of
viate from the standard one-pot synthesis, described in the ex-
perimental part, are summarized in the Supporting Information
the BiÀNiBiÀNi Bi S particles are shown in Figure S3, Support-
3
2 2
ing Information. No rodlike structure, as observed during the
one-pot synthesis can be seen. The particles seem to be more
roundish and agglomerated. This can most likely be attributed
to the fact that the elemental Bi had to be ground after the
first reaction step as it clumped together to one big glob.
During the one-pot reaction, the transformations seem to be
fast enough, so that no pronounced agglomeration occurs and
the rodlike morphology is not destroyed. However, there is
considerable morphology retention between NiBi and Ni Bi S .
3
2 2
This proves that intermetallic NiBi can possibly not only act as
a reactive but also as a morphological template in the
production of ternary chalcogenides. This will be the work of
future studies.
From the obtained results, one must conclude the reactions
as summarized in Equations (1–4). Bi S decomposes in the
2
3
3
+
2À
eg/en/NaOH mixture, Bi is reduced by the polyol and S is
complexed by en, leading to formal [S(en) ]. In a second step
Figure 3. XRD patterns of the decoupled one-pot reaction a) SR1: Collected
Bi after reduction of Bi
2
+
S
3
in an eg/en/NaOH mixture b) SR2: NiBi after
reaction of Bi with Ni in an eg/en/NaOH mixture c) SR3: Ni Bi +Bi after
S at 1978C in an eg/en/NaOH mixture (molar
x
2
2+
[36,37]
3
2
S
2
Ni is reduced to Ni
to form NiBi with the previously
reaction of NiBi with S or Na
ratio 1:1 or 3:2).
2
emerged Bi. Finally, NiBi reacts with [S(en) ] to Ni Bi S and Bi is
x
3
2 2
evolved. Excess [S(en) ] can just be washed out. Elemental
x
sulfur can also be applied, as the reductive power of the
polyol solution is reported to be strong enough to reduce it to
(
SR1–SR8; they are referenced with SRn with n as the number
2
À [38]
of the reaction in the Supporting Information). In a first step,
S .
Consequently, the observed reaction from Bi2S3 to
Bi S was reacted at 1978C without the addition of NiCl ·6H O,
Ni Bi S is also related to a multistep process involving the
2
3
2
2
3
2 2
which led to elementary Bi as the only product after a short
time (SR1, Supporting Information and Figure 3a). This is due
to the temperature-dependent reduction potential of basic
ethylene glycol. It is known that Bi S can be reduced in ethyl-
formation of Bi and NiBi.
Bi S þ 3 x en ! 2 Bi þ 3 ½SðenÞ Š
ð1Þ
2
3
x
2
3
2À
ene glycol at temperatures above 1508C and S ions are
HOCH CH OH ! CH CHO þ H O
ð2aÞ
ð2bÞ
ð3Þ
2
2
3
2
[
10,35]
evolved.
Surprisingly, intermetallic NiBi is obtained when
2þ
þ
reacting the obtained Bi particles with NiCl ·6H O at 1978C in
2 Ni þ 2 CH CHO ! 2 Ni
þ CH COCOCH þ 2 H
2
2
3
in situ
3
3
an eg/en/NaOH mixture for more than 10 h (SR2, Supporting
Information, Figure 3b). This step is very sensitive to the syn-
Niin situ þ Bi ! NiBi
thesis conditions, as NiBi is often contaminated with NiBi , pos-
3
3
NiBi þ 3 ½SðenÞ Š ! Ni Bi S þ Bi þ ½SðenÞ Š
ð4Þ
x
3
2
2
x
sibly due to agglomerated Bi particles derived from the first re-
ductive step. Reaction of the collected NiBi with Na S or S
2
Conversion chemistry in the BiÀNiBiÀNiBiSeÀNi Bi Se
3
2
2
(
molar ratio 1:1 or 3:2) in an eg/en/NaOH solution leads to
system
One-pot attempts to synthesize the related Ni Bi Se com-
Ni Bi S and the obligatory Bi (SR3; Figure 3c and Scheme 1a).
3
2 2
During the whole process no other possible binary or ternary
compounds like Ni S or NiBiS occurred. Trials to convert NiBi
3
2
2
pound starting from Bi Se and NiCl ·6H O in an eg/en/NaOH
3
2
3
2
3
2
2
into ternaries were not successful. We suggest that the mecha-
mixture resulted in a blend of NiSe, BiSe, NiBiSe, and Ni Bi Se
3 2 2
Chem. Eur. J. 2016, 22, 6333 – 6339
www.chemeurj.org
6335
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(
SR5). This joint appearance speaks for a close stability of binar-
to Ni Bi Se under the mentioned harsh conditions which sup-
3
2
2
ies and ternaries with 1:1, 1:1:1, and 3:2:2 stoichiometries. The
binary phases cannot be eliminated by a change in tempera-
ture and reaction time, which indicates an equilibrium
situation that is not easily shifted completely towards one of
the ternary products under these conditions.
press the 1:1:1 phase (SR9). We suppose that there is a narrow
stability range for NiBiSe.
Again one must conclude on a series of reactions [Eq. (5–
3
+
10)]. During the one-pot reaction, Bi Se decomposes, Bi is
2
3
2À
reduced, and Se is complexed to [Se(en) ]. As a consequence,
x
After a successful application and identification of NiBi as
a reactive template for the synthesis of ternary Ni Bi S , it was
the binary NiSe and BiSe side phases are formed due to
2À
a strong affinity of Se to the metal ions. One part of the
3
2 2
3
+
2+
now used for directed synthesis of its selenium homologues.
Starting the reaction from NiBi, the controlled synthesis of ter-
nary compounds is possible and no binary nickel or bismuth
selenides occur in the end products. Interestingly, the 1:1:1
and 3:2:2 phases still occur as competing phases. At tempera-
tures around 1978C and moderate alkalinity both phases are
detected (SR6). Our attempts to obtain phase pure NiBiSe re-
mained difficult, as for all evaluated conditions (no matter if
available Bi is reduced to Bi , which reacts further to BiSe;
the other part is reduced to the elemental state, which is able
to form NiBi in a second step. NiBi can afterwards either form
NiBiSe or Ni Bi Se , depending on the harshness of the reaction
3
2
2
conditions. The key to the reaction is found in intact NiBi. By
2
+
3+
these means, no Ni and Bi are available in the reaction
medium and binaries can be avoided. The exact mechanism of
the activation of the elemental selenium during the reaction
could not be determined in this work. It is known that elemen-
tal selenium in ethylenediamine is able to react with elemental
or ionic metals at room temperature or at least under solvo-
thermal conditions, but no intermediate reaction products are
3
:2 or 1:1 stoichiometry) small amounts of Ni Bi Se and Bi
3 2 2
were present. Nevertheless, by working without NaOH and at
lower temperatures of 1908C, NiBiSe is nearly the sole product
(
SR7, Figure 4a, Scheme 1c). Increasing the pH, temperature,
and reaction time (SR8, Figure 4b, Scheme 1b) to 3008C for
4.5 h completely suppressed the formation of the 1:1:1 phase.
[
38,39]
known.
Xie and co-workers propose a nucleophilic reac-
1
tion between en and selenium and show that selenium is dis-
[
40]
In general, shorter reaction times, lower temperatures, and/or
less NaOH led to an increase of the NiBiSe phase. Additionally,
the once-formed 1:1:1 phase can be completely transformed
solved in ethylenediamine under solvothermal conditions.
This brown solution is shown to be a suitable selenium feed
stock. Furthermore, en probably also activates the metal sur-
[
39]
face of Bi and NiBi. In our case, the proposed activation of
selenium with en is additionally supported by the reductive
power of the polyol solution [Eq. (2a–b)].
Bi Se þ 3 x en ! 2 Bi þ 3 ½SeðenÞ Š
ð5Þ
2
3
x
Ni þ x en Ð ½NiðenÞxŠ2þ; Ni þ ½SeðenÞ Š ! NiSe þ x en ð6Þ
2
þ
2þ
x
2
þ
Bi þ ½SeðenÞ Š ! BiSe þ x en
ð7Þ
ð8Þ
x
Niin situ þ Bi ! NiBi
3
NiBi þ 3 ½SeðenÞ Š ! NiBiSe þ x en
ð9Þ
x
NiBiSe þ x en ! Ni Bi Se þ Bi þ ½SeðenÞ Š
ð10Þ
3
2
2
x
Quantum chemical calculations
For an interpretation of the experimental results, the conduct-
ed reactions were simulated in terms of energetic classifica-
tions of the educts and the products by quantum chemical cal-
culations. As only elemental and metallic reactants have been
used with NiBi as the starting material, the idea of our recently
developed scheme to systematically evaluate phase stabilities
[
9,29,9c]
of competing systems with different compositions
is ap-
plied. Therein, the total electronic energy for educts and possi-
ble products is calculated under stoichiometric conditions, that
is, the number of atoms is maintained according to paths (1)
and (2). The present calculations are, however, restricted to the
formation of the 1:1:1 and 3:2:2 phases from NiBi. A full ap-
proach that will take into account all possible side products
will be published elsewhere. It is noteworthy, that although
our approach does not require any experimental preinforma-
tion, the calculations reflect the experimental results, which
Figure 4. a) XRD results of SR7; synthesis of NiBiSe, starting from NiBi at
1
908C without NaOH (small side phase of Ni
marked with #). b) XRD results of SR8; synthesis of Ni
from NiBi at 3008C in tetraethylene glycol (TEG), (Bi is marked with #).
3
Bi
2
Se
2
marked with + nd Bi
3
Bi Se and Bi, starting
2
2
Chem. Eur. J. 2016, 22, 6333 – 6339
www.chemeurj.org
6336
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the case of smaller differences between different binaries and
ternaries, control of conversion to different stoichiometry is
only possible by applying NiBi as an intermetallic precursor, as
2
+
seen for the selenium system. Intact NiBi avoids reactive Ni
3
+
and Bi ions that otherwise are able to form different binaries.
In the selenium system, two different ternaries, Ni Bi Se and
3
2
2
NiBiSe appear in the phase diagram. Primarily, NiBiSe is formed
when no NaOH is given to the reaction solution and lower
temperatures of around 1908C are applied (Scheme 1c). Under
more harsh conditions of high pH and high temperatures
Ni Bi Se can be obtained (Scheme 1b). Furthermore, the
3
2
2
present results also show that the conversion of one ternary
compound into the other is possible along the M+AX line in
the ternary phase diagram for NiBiSe to Ni Bi Se .
Figure 5. Energetic classification of educts and products for path (1):
NiBi+2X!Ni Bi +Bi and path (2): NiBi+X!NiBiX.
3
3
2 2
X
3
2
2
Elucidation of these reaction pathways provides a view into
the steps involved in the formation of solid-state compounds.
The present conversion reaction provides methods for the di-
rected synthesis of various intermetallics and novel functional
multinary compounds.
will be shown hereinafter. The zero value in Figure 5 represents
the energy of educts, that is, a combination of the values for
NiBi and the respective chalcogenide X that equals the utilized
stoichiometry. The gain in energy with respect to the products
is depicted for X=S, Se by bars according to the following
scheme:
Experimental Section
3
NiBi þ 2 X ! Ni Bi X þ Bi
pathð1Þ
pathð2Þ
3
2
2
General
NiBi þ X ! NiBiX
The following chemicals have been used without further purifica-
tion: Bismuth(III)nitrate (Merck), thiourea (Aldrich), ethylene glycol
(Fluka, >99.5%), tetraethylene glycol (Aldrich, 99%), ethylenedia-
mine (Aldrich, >99%), nickel(II)chloride hexahydrate (Alfa Aesar,
98%), selenium black (Merck, 99.5%), sulfur powder (Aldrich,
99.998%), Sodium hydroxide pellets (Merck, 99%), bismuth pieces
In the case of the X=S reaction, path (1) to Ni Bi S is pre-
3
2 2
ferred to path (2) to a still unknown NiBiS by more than
À1
1
20 kJmol . For X=Se, the formation of Ni Bi Se is also fa-
3
2
2
vored over NiBiSe. However, the difference in energy for the
À1
(
ChemPur, 99.999%).
3
:2:2 phase is less than 70 kJmol . From a thermodynamic
point of view the calculations give a hint that the difference in
energy is related to the observation that both phases can be
obtained experimentally for X=Se, but not for X=S and that
NiBiSe can be converted to Ni Bi Se under more harsh condi-
Synthesis of Bi S nanorods:
2
3
The Bi S nanorods were synthesized according to ref. [41].
2
3
3
2
2
tions. In case of X=S, only the more stable product is directly
obtained (Ni Bi S ) by the present approach. However, a final
Synthesis of Bi Se
2 3
3
2
2
answer on the question if NiBiS can be obtained or not will be
given from subsequent detailed calculations on the entire
Elemental Bismuth powder (0.522 g) and Se black powder (0.296 g)
were stirred in ethylenediamine (6 mL) overnight. The resulting
precipitate was washed with water and ethanol and dried at 808C
in a drying cabinet. To increase crystallinity, the obtained powder
was annealed at 2508C for 3 h under flowing argon according to
[8,9]
energy landscape with all possible side products
novel experiments.
and/or
[39]
Li.
Conclusions
Reproduction of the one-pot synthesis of Ni Bi S
2 2
In this paper, we identified and tested a plausible multistep re-
action pathway that results in the formation of ternary mixed
metal chalcogenides in the Ni/Bi/X (X=S, Se) system. Our ex-
periments identified three different pathways that can take
place: 1) AX!M AX; 2) A X !M AX; and (3) A X !MAX.
3
The one-pot synthesis of the ternary Ni Bi S was performed ac-
3
2 2
[30]
cording to a slightly modified report of Shao. A mixture of Bi2S3
(0.129 g) and NiCl ·6H O (0.178 g, molar ratio 1:3) in ethylene
2
2
glycol (70 mL) was added to a 100 mL round-bottomed flask, mag-
netically stirred, and treated with ultrasound. Then NaOH (0.545 g)
and ethylenediamine (5.05 mL) were added. After the NaOH had
completely dissolved the flask was heated and the mixture was re-
fluxed at 1978C for 0.5–10.5 h in air. After cooling to room temper-
ature, the products were centrifuged and washed several times
with water and ethanol and dried at 608C.
3
/2
2
3
3/2
2 3
According to the results found for A=Bi, Bi is formed first
when the reaction is started from Bi S or Bi Se . Bi is then con-
2
3
2
3
verted into intermetallic NiBi that, in turn, can react to ternary
compounds Ni BiX with different z(Ni).
z
According to calculated high differences in the energy of
formation, the more stable of the competing products is di-
rectly obtained (Ni Bi S , Scheme 1a). No binaries are formed,
All other modifications of the synthesis conditions can be found in
the Supporting Information. They are referenced with SRn (n
corresponds to the number of the respective reaction.
3
2 2
independent from which reactant is applied (Bi S or NiBi). In
2
3
Chem. Eur. J. 2016, 22, 6333 – 6339
www.chemeurj.org
6337
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Analysis
[7] a) H. Zhang, C.-X. Liu, X.-L. Qi, X. Dai, Z. Fang, S.-C. Zhang, Nat. Phys.
2009, 5, 438–442; b) B. Yan, M. Jansen, C. Felser, Nat. Phys. 2013, 9,
X-ray powder diffraction: XRD was performed with a Huber G670
diffractometer equipped with an imaging plate with monochro-
matic CuKa1 radiation (l=1.54060 Š, Ge-monochromator, increment
Dq=0.0058, Bragg angular range 48ꢀ1008). Morphologies of the
as-prepared samples were investigated with SEM using a Zeiss
Digital Scanning Microscope DSM 950.
709–711.
[8] a) M. Jansen, I. V. Pentin, J. C. Schçn, Angew. Chem. Int. Ed. 2012, 51,
132–135; Angew. Chem. 2012, 124, 136–139; b) V. L. Deringer, R. Dron-
skowski, Angew. Chem. Int. Ed. 2015, 54, 15334–15340; Angew. Chem.
2015, 127, 15550–15557; c) G. Kuhn, S. Mankovsky, H. Ebert, M. Regus,
W. Bensch, Phys. Rev. B 2013, 87, 085113.
9] a) F. Bachhuber, J. Rothballer, T. Sçhnel, R. Weihrich, J. Chem. Phys. 2013,
[
Quantum chemical calculations: The first-principles calculations
were carried out within the framework of DFT with exchange-cor-
relation functionals in the generalized gradient approximation
1
39, 214705; b) F. Bachhuber, J. Rothballer, T. Sçhnel, R. Weihrich,
Comput. Mater. Sci.. 2014, 89, 114–121; c) F. Bachhuber, J. von Appen,
R. Dronskowski, P. Schmidt, T. Nilges, A. Pfitzner, R. Weihrich, Angew.
Chem. Int. Ed. 2014, 53, 11629–11633; Angew. Chem. 2014, 126, 11813–
11817.
[42]
(
GGA) according to Perdew–Burke–Ernzerhof (PBE). Full geome-
try optimizations were executed with the Vienna Ab initio Simula-
[43]
tion Package (VASP), atomic site parameters and cell constants
were, therefore, allowed to fully relax with the conjugant gradient
algorithm. The interactions between the ions and the electrons are
described by the projector-augmented-wave (PAW) method with
scalar-relativistic potentials and a cutoff energy of 500 eV. All struc-
ture optimizations were performed in three subsequent steps with
an initial k-grid mesh of 4”4”4 rising to 8”8”8 and 12”12”12
to reach sufficient accuracy. Particularly for large unit cells, this
offers an additional verification of the performed calculations. A
structure optimization was considered to be converged with a dif-
[10] a) M. Kaiser, B. Rasche, M. Ruck, Angew. Chem. Int. Ed. 2014, 53, 3254–
258; Angew. Chem. 2014, 126, 3319–3323; b) M. F. Groh, M. Heise, M.
3
Kaiser, M. Ruck, Nachr. Chem. 2013, 61, 26–29; c) M. F. Groh, M. Heise,
M. Ruck, GIT Labor-Fachz. 2014, 6, 48–50.
[
[
11] a) M. Kaiser, A. Isaeva, M. Ruck, Angew. Chem. Int. Ed. 2011, 50, 6178–
6
180; Angew. Chem. 2011, 123, 6302–6304; b) M. Ruck, T. Sçhnel, Z. Na-
turforsch. B 2006, 61, 785–791; c) M. Ruck, Z. Anorg. Allg. Chem. 1999,
25, 2050–2054.
6
12] a) N. Pienack, W. Bensch, Angew. Chem. Int. Ed. 2011, 50, 2014–2034;
Angew. Chem. 2011, 123, 2062–2083; b) C. Feldmann, Angew. Chem. Int.
Ed. 2013, 52, 7610–7611; Angew. Chem. 2013, 125, 7762–7763; c) A.
Schiener, A. Magerl, A. Krach, S. Seifert, H.-G. Steinrück, J. Zagorac, D.
Zahn, R. Weihrich, Nanoscale 2015, 7, 11328–11333.
13] a) S. Krohns, P. Lunkenheimer, S. Meissner, A. Reller, B. Gleich, A. Rath-
geber, T. Gaugler, H. U. Buhl, D. C. Sinclair, A. Loidl, Nat. Mater. 2011, 10,
899–901; b) A. Reller, Z. Anorg. Allg. Chem. 2012, 638, 1552–1552.
14] H. Dong, Y.-C. Chen, C. Feldmann, Green Chem. 2015, 17, 4107–4132.
15] C. Feldmann, Adv. Funct. Mater. 2003, 13, 101–107.
À6
ference in total energy of less than 1”10 eV and a maximum
À4
À1
Hellmann-Feynmann force of 1”10 eVŠ . The final values of the
total energies of the investigated systems were obtained with
energy differences (between last and second to last step) of less
[
À3
than 1”10 eV per formula unit.
[
[
[
16] F. Fievet, J. Lagier, B. Blin, B. Beaudoin, M. Figlarz, Solid State Ionics
1
989, 32, 198–205.
17] D. Freudenmann, S. Wolf, M. Wolff, C. Feldmann, Angew. Chem. Int. Ed.
011, 50, 11050–11060; Angew. Chem. 2011, 123, 11244–11255.
[18] E. Ahmed, J. Breternitz, M. F. Groh, M. Ruck, CrystEngComm 2012, 14,
874–4885.
Acknowledgements
[
2
The work was supported by the Deutsche Forschungsgemein-
schaft (DFG) (WE 4284/1-2 and WE 4284/3-2). Prof. Dr. A.
Pfitzner is acknowledged for the use of technical equipment
and premises. We thank the computer centre of the University
of Regensburg for technical support.
4
[19] a) Y. Lin, S. Dehnen, Inorg. Chem. 2011, 50, 7913–7915; b) Y. Lin, W.
Massa, S. Dehnen, Chem. Eur. J. 2012, 18, 13427–13434.
[20] a) R. E. Cable, R. E. Schaak, J. Am. Chem. Soc. 2006, 128, 9588–9589;
b) Y. Vasquez, A. E. Henkes, C. J. Bauer, R. E. Schaak, J. Solid State Chem.
2008, 181, 1509–1523.
[
21] C. Schçttle, P. Bockstaller, R. Popescu, D. Gerthsen, C. Feldmann, Angew.
Chem. Int. Ed. 2015, 54, 9866–9870; Angew. Chem. 2015, 127, 10004–
10008.
Keywords: formation mechanism · nanoparticles · polyol
process · synthetic methods · ternary chalcogenide
[
[
22] Y. Yin, C. K. Erdonmez, A. Cabot, S. Hughes, A. P. Alivisatos, Adv. Funct.
Mater. 2006, 16, 1389–1399.
23] a) N. H. Chou, R. E. Schaak, J. Am. Chem. Soc. 2007, 129, 7339–7345;
b) M. Leonard, R. E. Schaak, J. Am. Chem. Soc. 2006, 128, 11475–11482.
24] S. M. Rommel, R. Weihrich, Chem. Eur. J. 2015, 21, 9863–9867.
25] P. Gütlich, K. J. Range, C. Felser, C. Schultz-Münzenberg, W. Tremel, D.
Walcher, M. Waldeck, Angew. Chem. Int. Ed. 1999, 38, 2381–2384;
Angew. Chem. 1999, 111, 2524–2527.
[
1] a) S. Scharfe, F. Kraus, S. Stegmaier, A. Schier, Th. F. Fässler, Angew.
Chem. Int. Ed. 2011, 50, 3630–3670; Angew. Chem. 2011, 123, 3712–
[
[
3
754; b) M. Ruck in Reference Module in Chemistry, Molecular Sciences
and Chemical Engineering, Elsevier, 2015; c) R. Ababei, J. Heine, M. Hoły-
n´ ska, G. Thiele, B. Weinert, X. Xie, F. Weigend, S. Dehnen, Chem.
Commun. 2012, 48, 11295–11297; d) J. M. Goicoechea, S. C. Sevov,
Angew. Chem. Int. Ed. 2006, 45, 5147–5150; Angew. Chem. 2006, 118,
[
[
26] H. Wang, M. Shao, S. Wang, H. Gao, G. Shao, J. Cryst. Growth 2008, 310,
3508–3511.
5
271–5274.
2] R. Boldt, M. Kaiser, D. Kçhler, F. Krumeich, M. Ruck, Nano Lett. 2010, 10,
08–210.
3] U. Schwarz, S. TencØ, O. Janson, C. Koz, C. Krellner, U. Burkhardt, H.
Rosner, F. Steglich, Y. Grin, Angew. Chem. Int. Ed. 2013, 52, 9853–9857;
Angew. Chem. 2013, 125, 10038–10042.
27] a) R. Weihrich, S. F. Matar, V. Eyert, F. Rau, M. Zabel, M. Andratschke, I.
Anusca, Th. Bernert, Prog. Solid State Chem. 2007, 35, 309–320; b) R.
Weihrich, W. Yan, J. Rothballer, P. Peter, S. M. Rommel, S. Haumann, F.
Winter, C. Schwickert, R. Pçttgen, Dalton Trans. 2015, 44, 15855–15864;
c) R. Weihrich, S. F. Matar, V. Eyert, I. Anusca, F. Bachhuber, S. Seidlmayer,
J. Solid State Chem. 2011, 184, 797–804.
[
[
2
[
[
4] A. Isaeva, M. Ruck, K. Schäfer, U. Ch. Rodewald, R. Pçttgen, Inorg. Chem.
2
015, 54, 885–889.
5] a) B. Chen, C. Uher, L. Iordanidis, M. G. Kanatzidis, Chem. Mater. 1997, 9,
655–1658; b) J. Koenig, M. Winkler, T. Dankwort, A. Hansen, H. Pernau,
V. Duppel, M. Jaegle, K. BartholomØ, L. Kienle, W. Bensch, Dalton Trans.
015, 44, 2835–2843.
[28] R. Weihrich, D. Kurowski, A. C. Stückl, S. F. Matar, F. Rau, T. Bernert, J.
Solid State Chem. 2004, 177, 2591–2599.
[29] F. Bachhuber, A. Krach, A. Furtner, T. Sçhnel, J. Rothballer, R. Weihrich, J.
Solid State Chem. 2015, 226, 29–35.
[30] G. Qian, M. Shao, Y. Tong, Y. Ni, J. Cryst. Growth 2005, 284, 412–416.
[31] X. Lin, S. L. Budko, P. C. Canfield, Philos. Mag. 2012, 92, 2436–2447.
[32] T. Sakamoto, M. Wakeshima, Y. Hinatsu, J. Phys. Condens. Matter 2006,
18, 4417–4426.
1
2
[
6] B. Rasche, A. Isaeva, M. Ruck, S. Borisenko, V. Zabolotnyy, B. Büchner, K.
Koepernik, C. Ortix, M. Richter, J. van den Brink, Nat. Mater. 2013, 12,
4
22–425.
Chem. Eur. J. 2016, 22, 6333 – 6339
www.chemeurj.org
6338
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
33] F. Dawood, Template directed synthesis and characterization of nanocrys-
talline metal oxides and chalcogenides, Dissertation 2010, The Pennsyl-
vania State University.
34] M. P. Brady, S. K. Wrobel, T. A. Lograsso, E. A. Payzant, D. T. Hoelzer, J. A.
Horton, L. R. Walker, Chem. Mater. 2004, 16, 1984–1990.
[41] Z. Liu, S. Peng, Q. Xie, Z. Hu, Y. Yang, S. Zhang, Y. Qian, Adv. Mater.
2003, 15, 936–940.
[42] a) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865–
3868; b) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1997, 78,
1396–1396.
[43] a) G. Kresse, J. Hafner, Phys. Rev. B 1994, 49, 14251–14269; b) G. Kresse,
J. Furthmüller, Phys. Rev. B 1996, 54, 11169–11186; c) G. Kresse, J. Non-
Cryst. Solids 1995, 192–193, 222–229; d) G. Kresse, J. Furthmüller,
Comput. Mater. Sci. 1996, 6, 15–50.
[
[
35] G. Valiulien e˙ , A. Zieliene, B. Simk u¯ naite, L. Naruskevicius, L. T. Tamasi u¯ -
ˇ
˙
ˇ
˙
ˇ
ˇ
ˇ
nait e˙ , V. Pak sˇ tas, A. Selskis, J. Solid State Electrochem. 2010, 14, 203–
2
12.
[
[
[
36] A. Kumar, P. V. Ramamurthy, Int. J. Chem. Kinet. 2000, 32, 286–293.
37] Y. Gao, H. Niu, C. Zeng, Q. Chen, Chem. Phys. Lett. 2003, 367, 141–144.
38] Z. Quan, C. Li, X. Zhang, J. Yang, P. Yang, C. Zhang, J. Lin, Cryst. Growth
Des. 2008, 8, 2384–2392.
[
[
39] Y. Li, Z. Wang, Y. Ding, Inorg. Chem. 1999, 38, 4737–4740.
40] J. Lu, Y. Xie, F. Xu, L. Zhu, J. Mater. Chem. 2002, 12, 2755–2761.
Received: December 15, 2016
Published online on March 21, 2016
Chem. Eur. J. 2016, 22, 6333 – 6339
www.chemeurj.org
6339
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim