Cr1.45Tl1.87Mo15Se19, a monoclinic variant of the hexa-gonal In3Mo15Se 19 type

Article abstract of DOI:10.1107/S0108270109047301

The monoclinic compound Cr_1.45Tl_1.87Mo ₁₅Se₁₉ (chromium thallium penta-deca-molybdenum nona-deca-selenide) represents a variant of the hexa-gonal In₃Mo ₁₅Se₁₉ structure type. Its

Full text of DOI:10.1107/S0108270109047301

inorganic compounds
Acta Crystallographica Section C
Crystal Structure
Communications
ISSN 0108-2701
Cr1.45Tl1.87Mo15Se19, a monoclinic
variant of the hexagonal In3Mo15Se19
type
P. Gougeon,* D. Salloum and M. Potel
´
Laboratoire de Chimie du Solide et Inorganique Moleculaire, UMR CNRS No. 6226,
´
´ ´
Universite de Rennes I, Avenue du General Leclerc, 35042 Rennes Cedex, France
Correspondence e-mail: patrick.gougeon@univ-rennes1.fr
Received 4 September 2009
Accepted 9 November 2009
Online 25 November 2009
Figure 1
A view of Cr1.45(2)Tl1.870(4)Mo15Se19, along [101]. Displacement ellipsoids
are drawn at the 97% probability level.
The monoclinic compound Cr1.45Tl1.87Mo15Se19 (chromium
thallium pentadecamolybdenum nonadecaselenide) repre-
sents a variant of the hexagonal In3Mo15Se19 structure type.
Its crystal structure consists of an equal mixture of Mo6Se8Se6
and Mo9Se11Se6 cluster units. The Mo and Se atoms of the
median plane of the Mo9Se11Se6 unit, as well as three Cr ions,
lie on sites with m symmetry (Wyckoff site 2e). The fourth Cr
ion is in a 2b Wyckoff position with 1 site symmetry.
Comment
¨
Thirty years ago, Gruttner et al. (1979) reported the crystal
structures of the hexagonal compounds In2.9Mo15Se19 and
In3.3Mo15Se19, which were the first compounds containing a
transition metal cluster with a nuclearity higher than six,
namely the bioctahedral Mo9 cluster. This cluster, which
results from the face-sharing of two Mo6 octahedra, co-exists
with the octahedral Mo6 cluster in equal proportions. Both
clusters are surrounded by Se atoms to form Mo6Se8Se6 and
Mo9Se11Se6 units that share some of their Se atoms to create
the three-dimensional Mo–Se framework. The In atoms
occupy two crystallographically different positions, depending
on their formal oxidation state of +1 or +3. While the In+ site is
Figure 2
Plot showing the atom-numbering scheme and the inter-unit linkage of
the Mo9Se11Se6 and Mo6Se8Se6 cluster units. Displacement ellipsoids are
drawn at the 97% probability level.
fully occupied, the In3+ site presents a nonstoichiometry that
III
0.9
leads to compositions ranging from In InI2Mo15Se19 to
InI1I.I3InI2Mo15Se19. Interest in these Mo cluster compounds lies
not only in their structural aspects but also in their physical
properties, because they show superconductivity with high
critical magnetic fields at about 4 K. We present here the
synthesis and the crystal structure of Cr1.45Tl1.87Mo15Se19,
which constitutes a monoclinic variant of the In3Mo15Se19
structure type.
A view of the crystal structure of Cr1.45Tl1.87Mo15Se19 is
shown in Fig. 1. The Mo–Se framework is similar to that of the
In3Mo15Se19 compounds and consists of an equal mixture of
Mo6Se8Se6 and Mo9Se11Se6 cluster units interconnected
through Mo—Se bonds (Fig. 2). The first unit can be described
as an Mo6 octahedron surrounded by eight face-capping inner
Sei and six apical Sea ligands; for details of the i- and a-type
¨
ligand notation, see Schafer & Schnering (1964). The Mo9 core
of the second unit results from the one-dimensional trans face-
sharing of two octahedral Mo6 clusters. The Mo9 cluster is
surrounded by 11 Sei atoms capping the faces of the diocta-
hedron and six apical Sea ligands above the terminal Mo
atoms. The Mo6Se8Se6 and Mo9Se11Se6 units are centred at 2a
and 2e Wyckoff positions and thus have point group symme-
tries 1 and m, respectively, instead of 3.. and 6.. in the hexa-
gonal In3Mo15Se19 compounds. The Mo—Mo distances within
˚
the Mo6 clusters are between 2.6769 (8) and 2.7252 (7) A, and
Acta Cryst. (2009). C65, i87–i90
doi:10.1107/S0108270109047301
# 2009 International Union of Crystallography i87

inorganic compounds
Figure 3
(a) The distribution of Cr1, Cr2 and Cr3. (b) The environment of Tl1.
1
[Symmetry codes are given in Table 1. Additionally: (viii) x; ꢂy þ ; z;
2
1
2
1
2
(ix) x þ 1; ꢂy þ ; z; (xi) x þ 1; ꢂy þ ; z þ 1.]
those within the Mo9 clusters are between 2.6278 (9) and
˚
2.7491 (7) A. These values are comparable with those
observed in the In3Mo15Se19 compounds.
The Se atoms bridge either one (Se1 to Se7) or two (Se8,
Se9 and Se10) Mo triangular faces of the clusters. Moreover,
atoms Se1, Se2, Se3, Se5, Se6 and Se7 are linked to an Mo
atom of a neighbouring cluster. The Mo—Se bond distances
˚
range from 2.5436 (9) to 2.6610 (6) A within the Mo6Se8Se6
˚
unit and from 2.5421 (7) to 2.7001 (8) Awithin the Mo9Se11Se6
unit. Each Mo9Se11Se6 unit is connected to six Mo6Se8Se6 units
(and vice versa) via Mo—Sea-i bonds to form the three-
dimensional Mo–Se framework, the connective formula of
Figure 4
Difference Fourier maps around the Tl1 mean position (large central
circles), showing (a) harmonic refinement (minimum and maximum
i-a
i-a
which is Mo9Se5iSe Sea-i , Mo6Se2iSe Sea-i . As a result of this
6/2 6/2
6/2 6/2
arrangement, the shortest intercluster Moꢀ ꢀ ꢀMo distance
ꢂ3
densities: ꢂ9.84 and 11.02 e A , respectively; step: 0.5 e Aꢂ3), and (b)
˚
˚
˚
between the Mo6 and Mo9 clusters is 3.4409 (8) A, indicating
only a weak metal–metal interaction. This distance is a little
anharmonic refinement (minimum and maximum densities: ꢂ2.08 and
ꢂ3
2.92 e A , respectively; step: 0.5 e Aꢂ3).
˚
˚
˚
shorter than the value of 3.512 (8) A observed for
In3.3Mo15Se19.
The main difference between Cr1.45Tl2Mo15Se19 and the
indium compounds lies in the distribution of the trivalent
cations. Indeed, in addition to the triangular group of distorted
octahedral sites located between two consecutive Mo6Se8Se6
units, the Cr atoms partially occupy a supplementary site
located at the mid-point of the a axis (2b position). In the
monoclinic compound Cr1.45Tl1.87Mo15Se19, atoms Cr1, Cr2
and Cr3 occupy the distorted octahedral sites located between
two consecutive Mo6Se8Se6 units that are occupied by tri-
valent indium in the In3Mo15Se19 compounds (Fig. 3a). The
loss of the threefold axis leads to three crystallographically
independent sites with site-occupancy factors of 0.619 (6),
0.313 (6) and 0.125 (6), compared with the trivalent indium
with site occupancies of 0.29 (5) for In2.9Mo15Se19 and 0.44 (2)
for In3.3Mo15Se19. The Cr—Se distances are in the ranges
2.515 (2)–2.850 (3), 2.524 (3)–2.798 (2) and 2.501 (8)–
˚
2.9635 (6) A for the Cr1, Cr2 and Cr3 sites, respectively.
Atoms Cr4 in the 2b Wyckoff position also occupy a distorted
octahedral site of Se atoms with Cr—Se distances ranging
˚
from 2.3096 (6) to 2.9635 (6) A.
The Tl+ cation of Cr1.45Tl1.87Mo15Se19 replaces the mono-
valent In+ in the large ten-coordinated sites of the Se atoms
(Fig. 3b). While the In+ site centred on the threefold axis has
3.. symmetry, the Tl+ site has no imposed symmetry. The Tl—
ꢁ
i88 Gougeon et al.
Cr1.453Tl1.87Mo15Se19
Acta Cryst. (2009). C65, i87–i90

inorganic compounds
Crystal data
3
˚
V = 1608.22 (4) A
Z = 2
Cr1.453Tl1.87Mo15Se19
Mr = 3396.7
Monoclinic, P21=m
Mo Kꢁ radiation
ꢂ = 36.91 mmꢂ1
T = 293 K
0.10 ꢄ 0.05 ꢄ 0.04 mm
˚
a = 9.7347 (1) A
˚
b = 19.2639 (3) A
˚
c = 9.7987 (1) A
ꢀ = 118.9305 (8)ꢃ
Data collection
Nonius KappaCCD area-detector
diffractometer
Absorption correction: analytical
(de Meulenaer & Tompa, 1965)
Tmin = 0.097, Tmax = 0.331
36020 measured reflections
10161 independent reflections
6942 reflections with I > 2ꢃ(I)
Rint = 0.087
Refinement
R[F2 > 2ꢃ(F2)] = 0.049
wR(F2) = 0.087
S = 1.49
231 parameters
Áꢄmax = 4.01 e A
Áꢄmin = ꢂ3.97 e A
ꢂ3
˚
ꢂ3
˚
10161 reflections
Figure 5
Anharmonic probability density function map of Tl1 in the bc plane
Table 1
Selected bond lengths (A).
˚
(minimum and maximum densities: ꢂ0.086 and 8.7905 Aꢂ3, respectively;
˚
step: 0.5 Aꢂ3).
˚
Tl1—Se1i
Tl1—Se2ii
Tl1—Se3
3.6290 (11)
3.4600 (11)
3.7414 (8)
3.2114 (11)
3.1803 (7)
3.1152 (11)
4.1426 (13)
4.1250 (12)
3.2515 (11)
3.8353 (6)
2.6769 (8)
2.7252 (7)
2.6992 (6)
2.7098 (8)
3.4409 (8)
2.5771 (8)
2.6214 (10)
2.5704 (10)
2.5575 (7)
2.6610 (6)
3.8047 (9)
2.6967 (9)
2.6906 (8)
3.4797 (8)
2.5574 (7)
2.5654 (9)
2.6055 (6)
2.5760 (9)
2.6456 (9)
3.8142 (8)
3.5231 (6)
2.6149 (8)
2.5821 (8)
2.5555 (8)
2.5436 (9)
2.6346 (9)
2.6441 (6)
2.6278 (9)
2.7347 (5)
2.6931 (7)
2.6523 (9)
2.6306 (8)
2.6382 (8)
2.6873 (7)
2.5507 (8)
2.6431 (8)
2.7095 (6)
3.8244 (6)
Mo5—Mo9
Mo5—Se2
Mo5—Se5
Mo5—Se6
Mo5—Se10
Mo5—Se11
Mo6—Mo7
Mo6—Mo8
Mo6—Mo9
Mo6—Se3vii
Mo6—Se6
Mo6—Se7
Mo6—Se9
Mo6—Se11
Mo7—Mo8
Mo7—Mo9
Mo7—Se5
Mo7—Se5viii
Mo7—Se8
Mo7—Se10
Mo8—Mo9
Mo8—Se7
Mo8—Se7viii
Mo8—Se8
Mo8—Se9
Mo9—Se6
Mo9—Se6viii
Mo9—Se9
Mo9—Se10
Cr1—Se4i
Cr1—Se6ii
Cr1—Se8
2.7300 (7)
2.6890 (6)
2.6337 (10)
2.6285 (11)
2.6808 (5)
2.5421 (7)
3.8222 (9)
2.7491 (7)
2.7032 (5)
2.6742 (9)
2.6536 (7)
2.5952 (6)
2.7001 (8)
2.5511 (8)
2.7127 (10)
2.6851 (13)
2.5885 (8)
2.5885 (8)
2.6093 (9)
2.5939 (10)
2.6863 (9)
2.5594 (6)
2.5594 (6)
2.5981 (15)
2.6152 (15)
2.5775 (8)
2.5775 (8)
2.6073 (11)
2.6096 (11)
2.515 (2)
2.750 (2)
2.850 (3)
2.821 (3)
2.524 (3)
2.798 (2)
2.788 (6)
2.738 (6)
2.501 (8)
2.782 (7)
2.859 (8)
2.886 (8)
2.3096 (6)
2.9635 (6)
2.7815 (8)
3.231 (5)
Tl1—Se5
˚
Se distances range from 3.1152 (11) to 4.2214 (9) A. The large
Tl1—Se6ii
Tl1—Se7iii
Tl1—Se8
Uiso value for Tl, which is about five times higher than those
for the Mo atoms and three to four times higher than those for
the Se and Cr atoms, is due to the large voids that it occupies
and does not result from an artefact due to the correlations
(0.506 in our case) between the atomic displacement para-
meters (ADPs) and the site-occupation factor of the Tl atom.
This is corroborated by the fact that similar ratios are
observed in other reduced molybdenum chalcogenides such as
Tl2Mo6Se6 (Potel et al., 1980a) or Tl2Mo9S11 (Potel et al.,
1980b), in which the Tl sites are fully occupied. While in the
In3Mo15Se19 structure type, the monovalent In+ site is fully
occupied, in the title compound the Tl+ site is only 0.935 (2)
occupied. This probably results from the higher temperature
used during the crystal growth process, which leads to a loss of
TlSe because of its high volatility at 1773 K. It is interesting to
note that a similar site-occupation factor of 0.94 (1) is found
when refining using SHELXL (Sheldrick, 2008) a model with
two Tl positions and harmonic ADPs and with only one
twinning matrix.
Tl1—Se10ii
Tl1—Se11iv
Mo1—Mo1v
Mo1—Mo2
Mo1—Mo2v
Mo1—Mo3
Mo1—Mo3v
Mo1—Mo5
Mo1—Se1v
Mo1—Se2
Mo1—Se3
Mo1—Se4
Mo1—Se5
Mo2—Mo2v
Mo2—Mo3
Mo2—Mo3v
Mo2—Mo6iii
Mo2—Se1
Mo2—Se2v
Mo2—Se3
Mo2—Se4
Mo2—Se6iii
Mo3—Mo3v
Mo3—Mo4vi
Mo3—Se1
Mo3—Se2
Mo3—Se3v
Mo3—Se4
Mo3—Se7vi
Mo4—Mo5
Mo4—Mo6
Mo4—Mo7
Mo4—Mo8
Mo4—Se1i
Mo4—Se5
Mo4—Se7
Mo4—Se8
Mo4—Se11
Mo5—Mo6
Mo5—Mo7
Mo5—Mo8
Cr1—Se9ii
Cr2—Se4
Experimental
Cr2—Se5
Cr2—Se9iii
Cr2—Se10
Cr3—Se4
Single crystals of Cr1.45Tl1.87Mo15Se19 were prepared from a mixture
of Cr2Se3, MoSe2, TlSe and Mo with the nominal composition
Cr1.5Tl2Mo15Se19. Before use, Mo powder was reduced under flowing
H2 gas at 1273 K for 10 h in order to eliminate any trace of oxygen.
The binaries Cr2Se3, MoSe2 and TlSe were obtained by heating
stoichiometric mixtures of the elements in sealed evacuated silica
tubes for about 2 d. All handling of materials was carried out in an
argon-filled glove-box. The initial mixture (ca 5 g) was cold pressed
and loaded into a molybdenum crucible, which was then sealed under
a low argon pressure using an arc-welding system. The charge was
heated at a rate of 300 K hꢂ1 to 1773 K, held at that temperature for
48 h, cooled at 100 K hꢂ1 to 1373 K and finally furnace cooled.
Cr3—Se7vi
Cr3—Se8vi
Cr3—Se10
Cr4—Se1
Cr4—Se5vi
Cr4—Se11vi
Cr1—Cr2i
Cr1—Cr3i
Cr2—Cr3
3.216 (12)
3.295 (10)
Symmetry codes: (i) x ꢂ 1; y; z ꢂ 1; (ii) x; y; z ꢂ 1; (iii) x þ 1; y; z; (iv) ꢂx ꢂ 1; ꢂy; ꢂz;
1
2
(v) ꢂx; ꢂy; ꢂz þ 1; (vi) x þ 1; y; z þ 1; (vii) x ꢂ 1; y; z; (viii) x; ꢂy þ ; z.
ꢁ
Acta Cryst. (2009). C65, i87–i90
Gougeon et al. Cr1.453Tl1.87Mo15Se19 i89

inorganic compounds
0.947 (2), 0.008 (1), 0.011 (1), 0.0262 (4), 0 and 0.0079 (4) for matrices
T1–T6, respectively. Refinement of the occupancy factors of Tl and
Cr led to the final stoichiometry Cr1.45(2)Tl1.870(4)Mo15Se19.
Numerous crystals were tested on a Nonius KappaCCD diffrac-
tometer. Most of them presented a substantial mosaic spread, leading
to unreliable lattice parameters having values close to those of a
hexagonal system. For these crystals, the mosaicity as calculated by
DENZO (Otwinowski & Minor, 1997) was of about 2ꢃ, indicating that
they are strongly twinned. For the studied crystal, the mosaicity was
only 0.496 (1)ꢃ, as usually observed for good single crystals. On the
other hand, all the Bragg spots found using the ’/ꢅ procedure
(Duisenberg et al., 2000) could be indexed, indicating that if the
crystal were twinned the twinning was very minor. Consequently, the
integration of the Bragg spots from the collected frames was made
with the orientation matrix deduced from the ’/ꢅ procedure and, in
order to take into account the possible existence of non-overlapping
spots at high diffraction angles, large integration boxes were used.
The structure was solved using SIR97 (Altomare et al., 1999) and
subsequent difference Fourier syntheses, and was refined in an
anisotropic approximation down to R = 0.077. Because the lattice
parameters are close to those of a hexagonal system, the existence of
twin domains was envisaged in the studied crystal. As the space group
P21/m is a subgroup of index 3 of the space group P63/m, six twin
domains can exist, according to the procedure implemented in
Data collection: COLLECT (Nonius, 1998); cell refinement:
COLLECT; data reduction: EVALCCD (Duisenberg, 1998);
program(s) used to solve structure: SIR97 (Altomare et al., 1999);
ˇ´ˇ
program(s) used to refine structure: JANA2006 (Petrıcek et al., 2006);
molecular graphics: DIAMOND (Brandenburg, 2001); software used
to prepare material for publication: JANA2006.
Supplementary data for this paper are available from the IUCr electronic
archives (Reference: LG3024). Services for accessing these data are
described at the back of the journal.
References
Altomare, A., Burla, M. C., Camalli, M., Cascarano, G. L., Giacovazzo, C.,
Guagliardi, A., Moliterni, A. G. G., Polidori, G. & Spagna, R. (1999). J.
Appl. Cryst. 32, 115–119.
Bachmann, R. & Schulz, H. (1984). Acta Cryst. A40, 668–675.
Brandenburg, K. (2001). DIAMOND. Version 2.1e. Crystal Impact GbR,
Bonn, Germany.
Duisenberg, A. J. M. (1998). PhD thesis, University of Utrecht, The
Netherlands.
Duisenberg, A. J. M., Hooft, R. W. W., Schreurs, A. M. M. & Kroon, J. (2000).
J. Appl. Cryst. 33, 893–898.
Gru¨ttner, A., Yvon, K., Chevrel, R., Potel, M., Sergent, M. & Seeber, B. (1979).
Acta Cryst. B35, 285–292.
Johnson, C. K. & Levy, H. A. (1974). International Tables for X-ray
Crystallography, edited by J. A. Ibers & W. C. Hamilton, Vol. IV, pp. 311–
336. Birmingham: Kynoch Press.
Meulenaer, J. de & Tompa, H. (1965). Acta Cryst. 19, 1014–1018.
Nonius (1998). COLLECT. Nonius BV, Delft, The Netherlands.
Otwinowski, Z. & Minor, W. (1997). Methods in Enzymology, Vol. 276,
Macromolecular Crystallography, Part A, edited by C. W. Carter Jr & R. M.
Sweet, pp. 307–326. New York: Academic Press.
ˇ´ˇ
JANA2006 (Petrıcek et al., 2006), which decomposes a point group
into co-sets with respect to a subgroup. Consequently, six twin
matrices corresponding to the identity (T1) and the clockwise (T2)
and anticlockwise (T3) threefold rotations about b, and the three
inverse matrices of these (T4, T5, and T6), were introduced into the
refinement. This lowered the R factor to 0.055. At this stage, analysis
of the difference Fourier map revealed positive and negative residual
peaks of 11.02 and ꢂ9.84 e Aꢂ3, respectively, around the Tl1 atom
˚
(Fig. 4a). Fourth-order tensors in the Gram–Charlier expansion
(Johnson & Levy, 1974) of the thallium displacement factor were used
to better describe the electronic density around the cationic site. The
residual R value dropped to 0.0489 and the residual difference
electron-dꢂen3sity extrema in the vicinity of Tl1 dropped to 2.92 and
ˇ´ˇ
ˇ
Petrıcek, V., Dusek, M. & Palatinus, L. (2006). JANA2006. Institute of Physics,
Czech Academy of Sciences, Prague, Czech Republic.
Potel, M., Chevrel, R. & Sergent, M. (1980a). Acta Cryst. B36, 1545–1548.
Potel, M., Chevrel, R. & Sergent, M. (1980b). Acta Cryst. B36, 1319–
1322.
˚
ꢂ2.08 e A (Fig. 4b). Fig. 5 shows the anharmonic probability
density function map for Tl1 in the bc plane. The absence of a
significant negative region indicates that the refined model can be
considered valid (Bachmann & Schulz, 1984). At the end of the
refinement, the fractional volumes of the twin domains were
¨
Schafer, H. & Schnering, H. G. (1964). Angew. Chem. 76, 833–849.
Sheldrick, G. M. (2008). Acta Cryst. A64, 112–122.
ꢁ
i90 Gougeon et al.
Cr1.453Tl1.87Mo15Se19
Acta Cryst. (2009). C65, i87–i90