Ternary arsenides based on platinum-indium and palladium-indium fragments of the Cu3Au-type: Crystal structures and chemical bonding

Article abstract of DOI:10.1016/j.jallcom.2014.09.180

Three metal-rich palladium-indium and platinum-indium arsenides, Pd₅InAs, Pt₅InAs, and Pt₈In₂As, were synthesized using a high-temperature ampoule technique. Their crystal structures were determined from Rietveld analysis of powder diffraction data. All the compounds crystallize in tetragonal system with P4/mmm space group (Pd₅InAs: a = 3.9874(1) ?, c = 6.9848(2) ?, Z = 1, R_p = 0.053; R_b = 0.013; Pt₅InAs: a = 3.9981(2) ?, c = 7.0597(4) ?, Z = 1, R_p = 0.058, R_b = 0.016; Pt₈In₂As: a = 3.9872(3) ?, c = 11.1129(7) ?, Z = 1, R_p = 0.047; R_b = 0.014). The first two compounds belong to the Pd₅TlAs structure type, while the third one is isotypic with the recently discovered Pd₈In₂Se. Main structural units in all arsenides are indium-centered [TM₁₂In] cuboctahedra (TM = Pd, Pt) of the Cu₃Au type, single- and double-stacked along the c axis in TM₅InAs and Pt₈In₂As, respectively, alternating with [TM₈As] rectangular prisms. DFT electronic structure calculations predict all three compounds to be 3D metallic conductors and Pauli-like paramagnets. According to the bonding analysis based on the electron localization function and electron localizability indicator topologies, all compounds feature multi-centered interactions between transition metal and indium in their heterometallic fragments. Additionally, pairwise interactions between platinum atoms are also observed, indicating a somewhat more localized bonding in the case of platinum-based compounds.

Full text of DOI:10.1016/j.jallcom.2014.09.180

Accepted Manuscript
Ternary arsenides based on platinum-indium and palladium-indium fragments
of the Cu Au-type: crystal structures and chemical bonding
3
Elena Yu. Zakharova, Natalia A. Andreeva, Sergey M. Kazakov, Alexey N.
Kuznetsov
PII:
S0925-8388(14)02353-6
DOI:
http://dx.doi.org/10.1016/j.jallcom.2014.09.180
JALCOM 32297
Reference:
To appear in:
Journal of Alloys and Compounds
Received Date:
Revised Date:
Accepted Date:
13 August 2014
22 September 2014
23 September 2014
Please cite this article as: E.Y. Zakharova, N.A. Andreeva, S.M. Kazakov, A.N. Kuznetsov, Ternary arsenides based
on platinum-indium and palladium-indium fragments of the Cu Au-type: crystal structures and chemical bonding,
3
Journal of Alloys and Compounds (2014), doi: http://dx.doi.org/10.1016/j.jallcom.2014.09.180
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Ternary arsenides based on platinum-indium and palladium-indium fragments of the
Cu Au-type: crystal structures and chemical bonding
3
[
Elena Yu. Zakharova, Natalia A. Andreeva, Sergey M. Kazakov, and Alexey N.
a]
[a]
[a]
[a,b]*
Kuznetsov
[
a]
Department of Chemistry, Lomonosov Moscow State University, Leninskie Gory 1-3, GSP-1, 119991 Moscow, Russian
Federation
[
b]
N.S. Kurnakov Institute of General and Inorganic Chemistry of Russian Academy of Sciences, Leninskii pr. 31, 119991
Moscow, Russian Federation
*
Corresponding author: Dr Alexey N. Kuznetsov
Department of Chemistry, Lomonosov Moscow State University, Leninskie Gory 1-3, GSP-1, 119991 Moscow, Russian
Federation
Fax: +7(495)9390998
E-mail: alexei@inorg.chem.msu.ru
Keywords: intermetallics; transition metal alloys and compounds; crystal structure; electronic band
structure; X-ray diffraction; chemical bonding
Abstract
Three metal-rich palladium-indium and platinum-indium arsenides, Pd
Pt In As, were synthesized using a high-temperature ampoule technique. Their crystal structures
were determined from Rietveld analysis of powder diffraction data. All the compounds
crystallize in tetragonal system with P4/mmm space group (Pd InAs: a = 3.9874(1) Å, c =
.9848(2) Å, Z = 1, R =0.053; R =0.013; Pt InAs: a = 3.9981(2) Å, c = 7.0597(4) Å, Z = 1,
=0.058, R =0.016; Pt In As: a = 3.9872(3) Å, c = 11.1129(7) Å, Z = 1, R =0.047; R =0.014).
The first two compounds belong to the Pd TlAs structure type, while the third one is isotypic
with the recently discovered Pd In Se. Main structural units in all arsenides are indium-centered
TM12In] cuboctahedra (TM=Pd, Pt) of the Cu Au type, single- and double-stacked along the c
axis in TM InAs and Pt In As, respectively, alternating with [TM As] rectangular prisms. DFT
5 5
InAs, Pt InAs, and
8
2
5
6
p
b
5
R
p
b
8
2
p
b
5
8
2
[
3
5
8
2
8
electronic structure calculations predict all three compounds to be 3D metallic conductors and
Pauli-like paramagnets. According to the bonding analysis based on the electron localization
function and electron localizability indicator topologies, all compounds feature multi-centered
interactions between transition metal and indium in their heterometallic fragments. Additionally,
pairwise interactions between platinum atoms are also observed, indicating a somewhat more
localized bonding in the case of platinum-based compounds.
1

1
. Introduction
When a metallic system is diluted with a typical non-metal, various combinations of
metallic and non-metallic fragments can be obtained as a result, including those featuring
anisotropic and potentially interesting properties. Applying this approach to the palladium-
indium system and introducing selenium as a typical non-metal into it, we have recently
discovered three new mixed palladium-indium selenides, Pd
2], all of which feature cuboctahedral [Pd12In] fragments that occur in binary intermetallics of
the Cu Au structure type. The former compound belongs to the Pd TlAs structure type [3],
which may be considered one of the less-studied ones and until recently included only a few
structurally characterized palldium- and platinum-based compounds, e.g. Pt MgP [3] and
Pd HgSe [4]. Several other palladium and platinum ternary pnictides, tetrelides and selenides,
belonging to the same type, were suggested in the same paper that established the Pd TlAs,
however, no structural data was provided and the compositions were not established properly.
This includes Pd-In-As and Pt-In-As systems, where the Pd TlAs-type phases were reported in
3] based solely on Guinier powder pattern indexing of non-phase-pure samples. Later the
5
InSe, Pd
8
In
2
Se [1], and Pd17In
4
Se
4
[
3
5
5
5
5
5
[
o
5
existence of Pt InAs was confirmed in the paper [5] reporting phase equilibria at 600 C in the Pt-
In-As systems, as well as the suggestion was made that the other compound with the proposed
composition of Pt In As exists in the same system. The latter was found to crystallize in the
tetragonal system, with a=3.987 Å, c=11.079 Å, and described as a superlattice variant of Pt In
based on three Pt In cells stacked one on top of the other [5], or Pd InAs with a Pt In cell
8
2
3
3
5
3
inserted in its center [5]. However, these suggestions were not supported by any structural
investigations. Here we report the results of our purposeful synthesis of the compounds of the
Pd
5
TlAs- and Pd
8
2
In Se-type in Pd-In-As and Pt-In-As systems, as well as their crystal end
electronic structure studies.
2. Experimental
2
.1. Synthetic and analytical procedures
Platinum (powder, 99.8%) and palladium (foil, 99.98%), indium (shot, 99.999%), and
elemental arsenic (shot, 99.99%) were used for sample preparation. Stoichiometric mixtures of
the elements (ca. 0.4–0.8 g in total per sample) were put into silica ampoules, sealed under
o
vacuum (ca. 20 mTorr), and annealed at 750°C for 7 to 10 days, and then at 600 C for another 30
2

to 40 days in case of palladium-containing samples, and for another 90-120 days for platinum-
containing ones, and then cooled down to room temperature. For all samples, a thorough
grinding in an agate mortar was used after each 10 to 14 days of annealing. The composition of
the products was studied by powder X-ray diffraction (XRD) using Stoe STADI-P automated
diffractometer (CuKα1). Energy-dispersive spectroscopic (EDS) analysis on separate
microcrystals from bulk samples was carried out with a Jeol JSM-6480 LV scanning microscope
equipped with an Oxford Instruments Inca+ Energy analytical module. The data were averaged
from 3–4 spots on the surface of each crystallite.
Chemical transport (CTR) technique was also employed in an attempt to grow single
crystals of platinum-containing phases. The annealed samples of both Pt
5
InAs and Pt
8
In
2
As
compositions were put into silica tubes (250 mm long, 15 mm in diameter) with a small amount
o
o
of iodine, and placed into a furnace with 650 C/700 C temperature gradient between the ends.
After 4 days of annealing, the only products of the CTR were the crystals of PtI
2
formed in the
cold end of each tube.
2
.2. Crystal structure determination
X-ray powder diffraction patterns were recorded using a Bruker D8 Advance
diffractometer (CuKα1-radiation, Ge(111) monochromator, reflection geometry, LynxEye strip
detector). Crystal structures of all three compounds were refined using the Rietveld method as
implemented in the TOPAS package [6].Crystal structures of Pt
from two poly-phase samples, with respective stoichiometries of Pt
structure of Pd InAs was refined from a phase-pure sample. Rietveld refinements were
performed using the fundamental parameter approach for the peak shape description. The
starting structural model for Pd InAs and Pt InAs was derived from the published data for the
Pd InSe [1], and the cell and coordinates for Pt In As were taken from the Pd In Se structure [1].
A small amount of impurities was found in the XRD data for the platinum-containing
compounds (<4% Pt, <3% PtAs in the Pt InAs sample; ~10% Pt In, <3% Pt in the Pt In As
5
InAs and Pt
8
In
2
As were refined
5
InAs and Pt
8
2
In As, while the
5
5
5
5
8
2
8
2
2
5
3
8
2
sample), and these phases were included into the refinement. The preferred orientation was
corrected using a spherical harmonics approach implemented in TOPAS.
3. Computational Details
Band structure calculations were performed on a density-functional theory (DFT) level
utilizing the all-electron full-potential linearized augmented plane wave method (FP-LAPW) as
3

implemented in the ELK code [7]. The Brillouin zone sampling was performed using 84 (Pt
3
In),
1
80 (Pd InAs, Pt InAs), and 168 (Pt In As) irreducible k-points. The PBESol exchange-
5
5
8
2
correlation functional [8] of the GGA-type was used in all the calculations. The muffin-tin sphere
radii for the respective atoms are (Bohr): 2.60 (In), 2.55 (Pt), 2.50 (Pd), 2.17 (As). The maximum
moduli for the reciprocal vectors kmax were chosen so that RMTkmax=10.0. The convergence of the
total energy with respect to the k-point sets was checked. Atomic charges were analyzed
according to Bader’s QTAIM approach [9]. The electron localization function (ELF) and electron
localizability indicator (ELI-D) were calculated according to [10] and [11], respectively, using
DGrid package [12]. The calculations were performed using the Intel Core-i7-based laboratory
cluster and the MSU Lomonosov supercomputer [13]. Visualization of the electron localization
indicators and their topological analysis was performed using ParaView package [14].
4. Results and Discussion
4
.1. Remarks on the synthesis of ternary compounds
In both palladium- and platinum-based systems, target compositions were chosen as
InAs and TM In As (TM=Pd, Pt), analogous to the previously obtained selenides [1], with
the synthetic conditions chosen accordingly. XRD data from the Pd-In-As system have shown,
however, that only the first homologue, Pd InAs, is obtained, and the samples are phase-pure,
while the ones with the Pd In As target composition contain the mixture of Pd InAs and Pd In.
Therefore we can conclude that “Pd In As” either cannot be produced or requires very different
TM
5
8
2
5
8
2
5
3
8
2
synthetic approach. In the Pt-In-As system, both target homologues were synthesized, according
to the XRD data. However, the situation with sample phase content mirrored closely what was
previously observed in the Pd-In-Se system, and also for the two Pd
namely, the joint presence of both homologues. Thus, in the samples with the Pt
composition, the dominant component was indeed found to be Pt InAs, but with an admixture of
the second homologue, Pt In As, just as in Pd-In-Se system. More metal-rich compound,
Pt In As, can be synthesized without its lower homologue present in the samples, however, in
3
In polytypes [15-17],
5
InAs
5
8
2
8
2
that case it will unavoidably coexist with its parent binary intermetallic. It has to be noted that
o
the synthetic conditions given in [3] (1 day of annealing at 600-700 C) are absolutely insufficient
to produce the samples with high content of target compounds. In order to minimize the amount
of impurities we had to use rather extreme annealing times, as long as four months of high-
temperature treatment with multiple intermediate homogenizations (see Section 2.1).
Considering that the structures of isotypic selenides were well-established from Rietveld data
4

supported by electron diffraction and transmission electron microscopy [1], and they were rather
straightforward and had no problematic features, for the structure refinement of platinum-based
compounds we used the poly-phase samples with minimal amounts of impurities.
4.2. Crystal structure description
Crystallographic data, atomic coordinates and selected bond distances for the three new
arsenides are given in Tables 1−3. Final Rietveld refinement plots are given in Figure 1. The
crystal structures of the new compounds are depicted in Figure 2. According to the Rietveld
refinement data, Pd
5 5 5
InAs and Pt InAs are fully isotypic with the Pd InSe compound [1] and
represent the Pd TlAs structure type [3]. The results of the EDS analysis of separate
5
microcrystals from both samples are in good agreement with the structural data: averaged
compositions of the respective compounds are Pd73(3)In14(2)As13(2) and Pt71(2)In15(1)As14(1), which
corresponds to the respective formulae Pd5.2(2)In1.0(1)As0.9(1) and Pt5.1(1)In1.07(8)As1.00(8). The
5
composition of the Pd InAs phase of ‘Pd72In16As12’, suggested in [3] for a reportedly phase-pure
sample according to the Guinier data, implies that there is a 25% shift towards indium from 1:1
ratio between indium and arsenic content. Our data, from both the structure refinement and EDS,
show no significant deviation from said 1:1 ratio between those elements. Comparing our unit
cell parameters for Pd
5
InAs (a = 3.9874(1) Å, c = 6.9848(2) Å) with the ones from [3]
(
a=3.96(6) Å, c=6.93(1) Å) we see that the c parameter for the latter is ca. 0.04-0.05 Å smaller
than in our case. However, the precision of the unit cell determination in [3], judging from large
margins of error, looks rather low for such small cell dimensions, which makes the quality of
indexing questionable. Therefore, we cannot unequivocally conclude whether the differences in
the composition discussed above arise from possible homogeneity range for Pd
reflect experimental errors in earlier studies.
5
InAs, or they just
As in the cases of other compounds of this type, the structures are built from indium-
centered [Pd12In] or [Pt12In] cuboctahedra, joined via common faces along the (001) plane to
form heterometallic layers, alternating along the c axis with the layers of rectangular prisms
[
Pd
and prisms are connected in the c-direction via common TM(2)-TM(2) edges. The Pt
compound is isotypic with Pd In Se [1] and shares the principal structural features with the
compounds of the Pd TlAs-type, except for the distinction of cuboctahedral fragments in its
structure being doubly stacked along the c axis, thus forming heterometallic layers of double
8
As] or [Pt
8
As] of the PtHg
2
structure type, that are joined via common edges. Cuboctahedra
8
2
In As
8
2
5
thickness with respect to Pt
5
InAs and others. The composition of the compound from the EDS
5

data (Pt73(2)In18(2)As9(1), or Pt8.1(2)In2.0(2)As1.0(2)) is in good agreement with structural data. We
have to note here that the initial guess of its structure made in [5] was both close to the reality
and wrong, since its structure is related to the Pt
hardly be viewed as a superlattice of tripled Pt
cuboctahedral fragments alternating with [Pt As] prisms, and the latter do not comply with the
Pt In structure.
5
InAs motif rather closely, however, it can
3
In, since it includes doubled [Pt12In]
8
3
To estimate the degree of distortion of the polyhedra, comprising the structures of ternary
arsenides, along the c axis, we use the height-to-width ratio for cuboctahedra and square prisms
that we have established for selenides in [1]. Evidently, for a cubic Pt In this parameter is
precisely 1, as for any undistorted cuboctahedron. For Pd In (TiAl -type), which can be regarded
as a parent intermetallic for our Pd InAs, this ratio is 0.91, showing that cuboctahedra are
3
3
3
5
significantly ‘squashed’ in a vertical direction. For all our ternary arsenides reported here, this
ratio is found to be equal to 1.01, which means that they all feature almost perfect cuboctahedra,
ever so slightly elongated along the c axis. Interestingly, this is also true for the original Pd
whose height-to-width ratio for cuboctahedra is 1.06, while both ternary palladium-indium
selenides have this parameter at 0.97. As far as [TM As] prisms are concerned, they are
elongated in both selenides and arsenides (their height-to-width ratios in Pd InAs, Pt InAs, and
Pt In As are 1.05, 1.07, and 1.08, respectively), the only exception being the Pd TlAs compound
itself, where [Pd As] polyhedra are almost perfectly cubic, with height-to-width ratio being
.995.
5
TlAs,
8
5
5
8
2
5
8
0
The distances between palladium or platinum atoms (see Table 3) in ternary arsenides
are slightly longer than in respective metals and are very close to those in their parent
intermetallics Pd In (2.77-2.89 Å [16]) and Pt In (2.82 Å [18]) , and also to those in Pd HgSe
2.84-2.85 Å [4]). Transition metal to indium distances in arsenides are also very similar to the
ones found in ternary selenides (2.80-2.83 Å [1]) and binary intermetallics Pd In and Pt In (2.76-
.87 Å and 2.82 Å, respectively). Palladium-arsenic contacts (2.48 Å) almost perfectly match
3
3
5
(
3
3
2
those in PdAs
2
(2.49 Å [19]), and platinum-arsenic distances in square prisms (2.50-2.51 Å) are
(2.49 Å [20]).
just slightly longer than in PtAs
2
4.3. Electronic structure and bonding
5 5 8 2 3
Electronic structures of Pd InAs, Pt InAs, and Pt In As, along with Pt In that can be
regarded a parent intermetallic for the two latter compounds, were evaluated based on the DFT
6

calculations. Total and projected densities of states (DOS) near the Fermi level for all
compounds are shown in Fig.3. As can be seen from the DOS plots, the principal features for all
compounds, including binary intermetallics, are rather similar. Main contributions near the Fermi
level arise from the transition metal d-states, which are almost completely filled and close to the
10
d -configuration. The contributions from arsenic 4p-states and indium 5s-states reside in a
significantly lower energy range, below -0.2 Hartree (ca. -5.5 eV). Indium 5p-states also show a
contribution to a region above the Fermi level. This picture is very typical for Cu
compounds of palladium and platinum and has been observed previously for a number of binary
and ternary compounds (Pd7-δSnTe [21], Pd17In Se [2], Pd In, Pd InSe, and Pd In Se [1]).
Relatively low yet non-zero DOS at the Fermi level indicates 3D metallic conductivity and a
likely temperature-independent paramagnetism, as was observed for other Cu Au-based
3
Au-based
2
4
4
3
5
8
2
3
chalcogenides. The fact that the DOS for ternary compounds are so similar to those of their
parent intermetallics indicates that the heterometallic fragments are largely responsible for the
main features and peculiarities of electronic structures of ternary compounds.
5 8 2
Band structures of Pt InAs and Pt In As (see Fig.4) are also consistent with 3D metallic
behavior. A slight spatial anisotropy in the band dispersion is observed, manifesting in an
unequal number of bands crossing the Fermi level along different directions in the k-space,
particularly for Pt
8
2
In As, which might be taken as an indication of a minor anisotropy of electric
conductivity, but not to the extent that it would significantly affect an overall metallic properties.
Charge density for all compounds was calculated and analyzed using R. Bader’s QTAIM
approach [9]. Derived atomic charges are given in Table 4. As seen from the data, there is a
noticeable charge transfer from indium atoms centering the cuboctahedra towards transition
metal atoms, particularly pronounced in the case of platinum, pointing out to a polar transition
metal-indium interactions. Arsenic atoms also carry a partial positive charge, also indicating
possible degree of polarity of the Pd-As or Pt-As interactions, especially in the case of the latter.
It has to be noted that in the isotypic selenides no significant charge transfer was observed
between selenium and palladium atoms [1,2], which might be taken as an indication of higher
polarity of platinum and palladium bonds with arsenic as compared to selenium. However,
charge density partitioning according to Bader’s scheme can only provide rather arbitrary
information on the bonding patterns, particularly in solid state, since charge density has only
atomic attractors and thus not very well-suited for the description of covalent or metallic
interactions.
7

In order to gain more insight into the chemical bonding in the compounds in question,
both binary intermetallics and ternary arsenides, we have performed topological analysis of the
electron localization function (ELF) [10] and electron localizability indicator (ELI-D) [11], both
of which have a benefit of featuring bonds and lone pairs as attractors in addition to the atomic
ones. It must be noted that the topologies of both functions turned out to be identical and only
differed in the absolute values (ELF is normalized to cover a range from 0 to 1, while ELI-D
does not have such limits). Both functions are analyzed in the same way, by going down from
higher values of localization parameters (η or Υ, respectively) and observing arising features, so
we will be mainly concentrating on discussing ELI-D topologies. ELI-D can also be partitioned,
in the spirit of the QTAIM, into space-filling basins of the respective attractors (atoms, lone
pairs, or bonds), the integration of charge density over which can provide more quantitative
bonding description.
As is typical for compounds based on d-metals, at high and medium values of Υ (or η) we
only observe atomic shells. Other features start to appear below Υ~0.94 (η~0.3) in the
intermetallic part of the structures. Fig. 5 displays ELI-D iso-surfaces for Pt
arsenides. As evident from the figure, there is a distinct difference in bonding in the palladium-
and platinum-based compounds. ELI-D iso-surface for Pd InAs (Fig.5D) shows 8 localization
3
In and three ternary
5
domains (we will label this type of domains Ω1) placed symmetrically around each indium atom
in the centers of octants and corresponding to the four-centered 3Pd+In interactions, as it has
been already established for binary palladium-indium intermetallics as well as ternary selenides
[
1,2]. The integration of charge density over ELI-D basins corresponding to Ω1 gives us ca. 0.9
electrons (e) per each tetrasynaptic basin. The polarity of the bonding can also be estimated
according to [22]. For this purpose, an ELI-D bonding basin intersection by charge density
basins is calculated, and then the charge density is integrated over each intersection part. Thus,
the Ω1 ELI-D basin is intersected by 3 Pd and one In charge density basins with the ratio of
75.2%:24.8% by volume, thus confirming the four-centered nature of the bonding. However, the
integration of the charge density over each intersection shows that the charge density is
distributed between Pd and In atoms as 79.6%:20.4%, showing that the charge density is not
distributed uniformly across the basins, so that the basin volume alone cannot be taken as a
measure of bond strength or polarity, and a charge density integration must always be performed,
and also confirming the polar metallic interactions between palladium and indium atoms.
Moving towards the platinum-based compounds, we also observe these tetrasynaptic
basins corresponding to multi-centered metal-metal bonding, however, they are not the first
8

3
features to appear in the ELI-D after the atomic shells. For Pt In (see Fig.5A), below Υ=0.937 we
first observe the appearance of disynaptic domains (Ω2) corresponding to the pairwise Pt-Pt
interactions, and only at lower localization parameter values (Υ=0.928) we see the domains of
the Ω1 type appearing, which might indicate that the pairwise Pt-Pt interactions are better
localized than the four-centered 3Pt+In ones. This picture holds true for both ternary arsenides
with one extra feature: in addition to the Ω2-type domains corresponding to the bonds between
platinum atoms in the same crystallographic position (bases of the cuboctahedra, Pt(2)-Pt(2)) we
also get the domains (Ω2’) corresponding to the bonds between platinum atoms in different
crystallographic positions (diagonals, Pt(1)-Pt(2) and Pt(2)-Pt(3), see fig.5B, 5C). The
integration of the charge density over the bonding basins for platinum-based compounds gives us
3 5
the following results: Pt In – 1.6 e (Ω1), 0.4 e (Ω2); Pt InAs - 1.6 e (Ω1), 0.6 e (Ω2), 0.4 e
(
Ω2’); Pt In As – 1.4 e (Ω1), 0.6 e (Ω2), 0.3 e (Ω2’). Basin intersection calculations similar to
8
2
those described above provide us with the following information. All basins of Ω2 and Ω2’ type
feature almost equal charge distribution (~51%:49%, with Pt(2) atoms featuring slightly less
electrons involved in the pairvise interactions), while basins of the Ω1-type show the bias
towards more polar interactions than that in the case of palladium: 83%:17% charge distribution
3
for the four-centered 3Pt+In bonds in Pt In, and ~85%:15% charge distribution in ternary
platinum-indium arsenides. As we can see from these data, despite pairwise interactions
appearing to be more localized than the four-centered ones, the number of electrons involved in
the latter is still greater, so the multi-centered metallic bonding is still the dominant effect in the
heterometallic fragments here, as it was found to be in the case of palladium-based compounds.
Pairwise Pt-Pt interactions, even though featuring smaller number of electrons involved, trhough
the additional strengthening of the bonding, might provide extra stability for the structure. For
instance, this effect might help to account for the lack of the Pd
8 2
In As compound isotypic to the
discovered Pt InAs. The polarity of the heterometallic bonding in platinum-based compounds
5
also exceeds that of the platinum-based ones, indicating slightly larger portion of ionic
interactions in the case of the former compounds.
5. Conclusions
Using the high-temperature ampoule technique, we have synthesized three block-type
metal-rich arsenides, Pd InAs, Pt InAs, and Pt In As, and determined their crystal structures
from powder XRD data. All compounds are built from cuboctahedral [TM12In] fragments
5
5
8
2
(
TM=Pd, Pt) of tetragonally distorted Cu
3
Au-type structure, single-stacked in Pd
9
5
InAs and

Pt
prisms with PtHg
discovered palladium-indium metal-rich selenides, Pd
5
InAs, and double-stacked in Pt
structure. The compounds were found to be fully isotypic with recently
InSe and Pd In Se [1]. DFT calculations
8 2 8
In As, alternating along the c axis with [TM As] rectangular
2
5
8
2
predict all compounds to be 3D metallic conductors and Pauli-like paramagnets. According to
the bonding analysis based on the ELF and ELI-D topologies and Bader’s QTAIM approach, all
compounds feature multi-centered transition metal-indium interactions in their heterometallic
fragments and significant charge transfer from indium to transition metal atoms, indicating the
polarity of bonding, particularly in the case of platinum-based compounds. The distinguishing
feature of the latter compounds is the additional pairwise Pt-Pt bonding within heterometallic
fragments, that was not observed in the cases of palladium-based compounds and might account
for the differences in the stability between the two groups.
Acknowledgments
This work was partially supported by Russian Foundation for Basic Research (grants 14-
03-31802 and 12-03-00833a) and the Presidential Programme of Russian Academy of Sciences
(
grant 8P23). The use of the resources of MSU Supercomputer Center is kindly acknowledged.
References
[
1] E.Yu. Zakharova , S.M. Kazakov , A.A. Isaeva , A.M. Abakumov , G. Van Tendeloo, A.N.
Kuznetsov, J. Alloys Compd. 589 (2014) 48–55.
2] E.Yu. Zakharova, A.V. Churakov, Th. Doert, A.N. Kuznetsov, Eur. J. Inorg. Chem. (2013)
164–6169.
[
6
[
[
3] M. El-Boragy, K. Schubert, Z. Metallkunde, 61 (1970) 579-584.
4] F. Laufek, A. Vymazalová, M. Drábek, J. Navrátil, T. Plecháček, J. Drahokoupil, Solid State
Sci. 14 (2012) 1476-1479.
[
[
5] D. Swenson, Y.A. Chang, Mater. Sci. Eng. 22 (1994) 267-273.
6] TOPAS V3, General Profile and Structure Analysis Software for Powder Diffraction Data,
User’s Manual, Bruker AXS, Karlsruhe, Germany, 2003.
7] ELK, an all-electron full-potential linearised augmented-plane wave (FP-LAPW) code, ver.
.2.10. http://elk.sourceforge.net.
8] J.P. Perdew, A. Ruzsinszky, G.I. Csonka, O.A. Vydrov, G.E. Scuseria, L.A. Constantin, X.
Zhou, K. Burke, Phys. Rev. Lett. 100 (2008) 136406.
9] R. F. W. Bader, Atoms in Molecules: A Quantum Theory, Oxford University Press, Oxford,
[
2
[
[
10

1990.
[
10] a) A. D. Becke, K. E. Edgecombe, J. Chem. Phys. 92 (1990) 5397-5403; b) M. Kohout, A.
Savin, Int. J. Quantum Chem. 60 (1996) 875–882; c) A. Savin, O. Jepsen, J. Flad, O.K.
Andersen, H. Preuss, H.G. von Schnering, Angew. Chem. Int. Ed. Engl. 31 (1992) 187-188.
[
11] a) M. Kohout, Int. J. Quantum Chem. 97 (2004) 651-658; b) M. Kohout, Faraday Discuss.
35 (2007) 43–54; c) M. Kohout, K. Pernal, F.R. Wagner, Yu. Grin, Theor. Chem. Acc. 112
1
(
[
[
2004) 453–459.
12] M. Kohout, DGrid, ver. 4.6, Radebeul, 2011.
13] V. Sadovnichy, A. Tikhonravov, Vl. Voevodin, and V. Opanasenko in Contemporary High
Performance Computing: From Petascale toward Exascale (Chapman & Hall/CRC
Computational Science), CRC Press, Boca Raton, USA, 2013, pp.283-307.
[
[
[
[
[
[
[
[
14] ParaView 4.1.0. Open Source Scientific Visualization. www.paraview.org.
15] S. Bhan, K. Schubert, J. Less-Common Met. 17 (1969) 73–90.
16] H. Kohlmann, C. Ritter, Z. Anorg. Allg. Chem. 635 (2009) 1573–1579.
17] H. Kohlmann, C. Ritter, Z. Naturforsch. 62b (2007) 929–934.
18] K. Schubert, S. Bhan, T.K. Biswas, K. Frank, P.K. Panday. Naturwiss. 55 (1968) 542-543.
19] N.E. Brese, H.G. von Schnering, Z. Anorg. Allg. Chem. 620 (1994) 393-404.
20] J.T. Szymanski, Can. Mineral. 17 (1979) 117-123.
21] S.V. Savilov, A.N. Kuznetsov, B.A. Popovkin, V.N. Khrustalev, P. Simon, J. Getzschmann,
Th. Doert, M. Ruck, Z. Anorg. Allg. Chem. 631 (2005) 293-301.
22] S. Raub, G. Jansen, Theor. Chem. Acc. 106 (2001) 223–232.
[
11

Figure captions
Figure 1. Observed, calculated and difference Rietveld profiles for the Pd
5
InAs (top), Pt
5
InAs
(
8 2
middle) and Pt In As (bottom) samples.
Figure 2. Polyhedral representation of the crystal structures of Pd
5
InAs, Pt InAs (left) and
5
Pt In As (right). Palladium/platinum atoms are shown as cyan, indium as blue, and arsenic as
8
2
light brown spheres.
3
Figure 3. Total (TDOS) and projected (PDOS) densities of states near the Fermi level for Pt In
(
5 8 2 5
top left) and, Pt InAs (top right), Pt In As (bottom left), and Pd InAs (bottom right).
Figure 4. Band structures near the Fermi level of Pt
5
InAs (top) and Pt
8
2
In As (bottom) from DFT
calculations.
Figure 5. ELI-D isosurfaces for: A – Pt
3
In (Υ=0.927), B – Pt
8
In
2
As (Υ=0.905), C – Pt InAs
5
(
Υ=0.932), D – Pd InAs (Υ=0.932). For the explanation of Ω1, Ω2, and Ω2’ see the text.
5
12

Table 1. Data collection and Rietveld analysis parameters, and final residuals.
Refinement 1
Pt In As
18.4%
Refinement 2
Pt In As
Refinement 3
Pd InAs
Compound
Pt
5
InAs
8
2
8
2
5
Mass fraction
Data collection
75.6%
87.2%
Bruker D8 Advance
100%
Radiation type, source
X-ray, CuKα1
Data collection
temperature, K
295
Range in 2θ, step size (°)
7-92, 0.02
InAs, Z=1 Pt In
7.5-112, 0.02
P4/mmm (№123)
Z=1
8-97.5, 0.02
Space group
Z
Pt
5
8
2
As, Z=1
Z=1
Formula weight (g / mol)
Unit cell parameters
a (Å)
1165.130
1865.183
1865.183
721.840
3.9981(2)
7.0597(4)
112.85(1)
4.0005(9)
10.974(4)
175.6(1)
3.9872(3)
11.1129(7)
176.67(3)
4.71, 6.93
1.41
3.9874(1)
6.9848(2)
111.055(8)
5.31, 8.32
1.61
c (Å)
3
V (Å )
R
p
, Rwp (%)
5.77, 8.16
RBragg (%)
1.34
2.54
GoF
1.78
1.80
2.12
13

*
Тable 2. Refined coordinates and isotropic displacement parameters for Pt
Pd InAs.
5
InAs, Pt
8
In As, and
2
5
5
Pt InAs
2
Atom
Pt(1)
Pt(2)
In
Site
1a
4i
1c
1b
x/a
0
0
1/2
0
y/b
0
1/2
1/2
0
z/c
0
0.2861(2)
0
1/2
Biso, Å
0.6(1)
1.1(1)
1.3(2)
0.8(2)
As
8 2
Pt In As
2
Atom
Pt(1)
Pt(2)
Pt(3)
In
Site
2f
4i
2g
2h
1b
x/a
0
0
0
1/2
0
y/b
1/2
1/2
0
1/2
0
z/c
Biso, Å
0
0.99(6)
0.67(6)
0.67(8)
0.3(1)
0.3624(1)
0.1817(4)
0.1774(1)
1/2
As
1.0(2)
5
Pd InAs
2
Atom
Pd(1)
Pd(2)
In
Site
1a
4i
1c
1b
x/a
0
0
1/2
0
y/b
0
1/2
1/2
0
z/c
Biso, Å
0
0.9(1)
0.7(1)
0.5(1)
0.7(1)
0.2887(1)
0
1/2
As
*
-
Refinement 2, respectively.
5 8 2
coordinates and thermal displacement parameters for Pt InAs and Pt In As are taken from Refinement 1 and
14

Table 3. Selected interatomic distances for the Pt
5
InAs, Pt
8
In
2
As, and Pd InAs structures.
5
Atoms
Distance,
Å
Atoms
Distance,
Å
Atoms
Distance,
Å
5
Pt InAs
Pt(1)-Pt(2) 2.842(1)
Pt(2)-Pt(2) 2.827(1)
Pt(1)-In
Pt(2)-In
2.827(1)
2.842(1)
8 2
In As
Pt(2)-As
Pt(2)-As
2.505(1)
2.513(1)
Pt
2
2
2
2
.819(1)
.838(1)
.819(1)
.829(1)
2.804(1)
Pt(1)-Pt(1)
Pt(1)-Pt(3)
Pt(2)-Pt(2)
Pt(2)-Pt(3)
Pt(1)-In
2.864(1)
2.820(1)
Pt(2)-In
Pt(3)-In
Pd
5
InAs
2.819(1)
2.835(1)
Pd(1)-Pd(2) 2.835(1)
Pd(2)-Pd(2) 2.819(1)
Pd(1)-In
Pd(2)-In
Pd(2)-As
2.481(1)
15

Table 4. Calculated QTAIM atomic charges in Pt
3
In, Pd
5 5 8 2
InAs, Pt InAs, and Pt In As.
Pt In Pd InAs Pt
3
5
5
InAs
Pt
8
2
In As
Atom
Charge
Atom
Pd(1)
Pd(2)
-
Charge
-0.27
-0.18
-
Atom
Pt(1)
Pt(2)
-
Charge Atom
Charge
-0.37
Pt
-
-0.34
-0.29
-0.28
-
Pt(1)
Pt(2)
Pt(3)
In
-
-0.27
-
-
-0.28
In
-
+1.02
-
In
+0.80
+0.18
In
+1.03
+0.47
+1.05
+0.46
As
As
As
16

•
•
•
•
Three metal-rich platinum- indium and palladium-indium arsenides were synthesized.
Their crystal structures were determined from powder XRD.
Electronic structures and bonding were studied using DFT/FP-LAPW calculations.
Multi-centered platinum-indium or palladium-indium bonding was revealed using ELF
and ELI-D analysis.
•
The key difference in bonding between Pd- and Pt-based compounds are extra pairwise
Pt-Pt interactions, which are not observed for Pd.
17