Investigations of the New Solids In5S5Cl, In5Se5Cl, In5S5Br and In5Se5Br
orientation matrices for the twin components. Reciprocal planes
calculated from imaging plate data (SPACE [14]) show, that the
twinning can be described as a twinning by reticular pseudomero-
hedry, reflections h0l with h ϭ 5n coincide exactly (see above).
Due to the existence of partly overlapping reflections beside com-
pletely overlapping and separated ones, a refinement using the
SHELX HKLF 5 Format [17] is unfavorable. The Program TWIN
[14] allows a simultaneous integration of reflections from all do-
mains by rejecting all completely and partly overlapping ones.
Using this strategy, the results of the refinement of both structures
(In5Ch5Cl (Ch ϭ S, Se)) are satisfactory although the datasets are
incomplete.
The observed systematic extinctions are consistent with the space
groups P21 and P21/m. The latter which is centrosymmetric was
chosen. This choice was supported by means of the distribution of
the normalized structure factors (E-statistic, N(Z)-test). The large
maximum residual electron density and the relatively high R-values
can be ascribed to the incomplete data sets and to a variety of
crystal defects described above (HRTEM investigations).
In case of the bromine compounds (In5Ch5Br (Ch ϭ S, Se)) a close
examination of reciprocal space shows no anomalies in contrast to
the chlorine compounds. Again the programs INDEX [14] and
CELL [14] were used to index the lattices and to obtain the orienta-
tion matrix for In5S5Br and In5Se5Br, respectively. The following
integration had to be applied by the program INTEGRATE [14].
The observed systematic extinctions are consistent with the space
groups Pmmn and Pmn21. Whereas the structure solution and the
refinement converged well in the latter acentric space group. At-
tempts to solve the structure in the centrosymmetric space group
Pmmn did not lead to satisfactory result. The choice of the non-
centrosymmetric space group was also confirmed by means of the
distribution of the normalized structure factors (E-statistic, N(Z)-
test). Both structures of the bromine crystals were refined as an
inversion twin with a fractional contribution of 0.38(2) (In5S5Br)
and 0.49(2) (In5Se5Br), respectively.
Experimental Part
Synthesis: Single crystals of In5Ch5X were prepared by heating stoi-
chiometric mixtures of the elements In, S or Se together with the
required amount of the binary compound InX3 (X ϭ Cl, Br) in
evacuated dry quartz glass ampoules (e.g. In ϩ Se ϩ InCl3 in the
ratio 14:15:1). The ampoule content was heated up to 773 K and
annealed at this temperature for one week. The resulting fine, nee-
dle-shaped, not markedly air-sensitive crystals were reddish brown
in the case of In5S5X and dark brown to black for In5Se5X. In all
cases the products are inhomogeneous. The by-products are black
needles of In6Ch7 (Ch ϭ S, Se) and crystals of In2S3 for In5S5X
samples. Using a special temperature program (heating with 2 K/h
up to 773 K, annealing for two weeks, cooling with 5 K/h) bigger
crystals were obtained, nevertheless the samples were inhomogene-
ous. Almost homogeneous products were achieved by annealing
pellets of binary compounds InCh, In2Ch3 and InX in the ratio
2:1:1. The composition of all compounds was verified by quantita-
tive Energy Dispersive X-ray (EDX) measurements. For In5S5Br
and In5Se5Cl the results agree with the refined compositions of the
X-ray experiments. For In5S5Cl and In5Se5Br, where the anions are
indistinguishable in the single crystal X-ray diffraction experiment,
the expected composition was confirmed too.
Electron microscopy: Microcrystalline samples with nominal com-
positions In5S5Cl and In5S5Br were crushed under n-butanol. Cop-
per grids were covered with the suspension, leaving the crystallites
in random orientations after drying. These sample carriers were
fixed in a side-entry, double-tilt holder with a maximum tilt of
25° in two directions. HRTEM and SAED were performed in a
Philips CM30ST (300 kV, LaB6 cathode). The SAED patterns were
obtained using a diaphragm which limited the diffraction to a cir-
cular area of approximately 105 pm in diameter. The EMS program
package, see [13], served for the simulation of HRTEM micro-
graphs (spread of defocus: 7000 pm, illumination semiangle: 1.2
mrad) and SAED patterns (kinematical approximation). All images
were recorded with a Multiscan CCD Camera (Gatan, Software:
Digital Micrograph 3.6.1, Gatan). A small amount of noise resul-
ting from an amorphous surface layer on the crystals can be redu-
ced by filtering Fourier transforms with a suitable band-pass mask.
It was checked for all cases whether image processing leads to a
loss of essential real structure information. EDX was performed in
the scanning- and nanoprobe mode of Philips CM30ST with a Si/
Li-EDX detector (Noran, Vantage System).
The trial structures for In5Ch5X (Ch ϭ S, Se; X ϭ Cl, Br) were
obtained by direct methods (SHELXS97 [16]) and refined using
SHELXL97 [17]. A summary of crystal and structure refinement
data for In5Ch5X (Ch ϭ S, Se; X ϭ Cl, Br) is reported in Tables
1Ϫ3.
Further details of the crystal structure investigation are available
from the Fachinformationszentrum Karlsruhe, D-76344 Eggenstein-
Leopoldshafen (Germany), on quoting the depository number CSD-
414218 (In5Se5Br), -414219 (In5S5Br), -414220 (In5Se5Cl), -414221
(In5S5Cl), the name of the author(s), and citation of the paper.
Crystallography: X-ray powder diagrams show good agreements to
calculated ones assuming the structure models of the single crystal
investigations. Especially the patterns of In5S5Cl and In5Se5Cl give
no evidence for additional orthorhombic phases.
Conclusion
Crystals of suitable size for X-ray structure determination were
separated from the bulk product under a polarizing microscope.
The crystals were fixed at the top of glass capillaries and mounted
on a STOE IPDS I diffractometer using graphite-monochromated
Mo-Kα radiation. Experimental details are summarized in Table 1.
The STOE IPDS Program package [14] was used to analyze the
measured data. A close examination of the reciprocal space using
the program RECIPE [14] shows that all examined chlorine com-
pounds (In5Ch5Cl (Ch ϭ S, Se)) have overlapping and interpenetra-
ting lattices. In the case of In5S5Cl the above mentioned results
were obtained from a systematically twinned crystal with the ap-
proximate volume ratio of 1:1. The data for In5Se5Cl are based on
a crystal which consist of two systematically twinned domains with
an approximate volume ratio of 3:1. The programs INDEX [14]
and CELL [14] were used to index the lattices and to obtain the
The initial point of our investigation was the occurrence of
similar (cis and trans edge-sharing (InCh6)-octahedra con-
nected by ethane analogues In2Ch6 unites) and different
two dimensional building units in the mixed valence com-
pounds MIn5Ch6, MIn5Ch7 and MIn7Ch9.
We expected new structure types based on different com-
binations of these building units in the sense of a construc-
tion kit. In fact, the presented compounds fulfill this con-
cept in several ways. On the one hand both new structure-
types can be classified in the framework of this concept. On
the other hand, nanoscale defects of the real structure of
In5Ch5Cl can easily be explained by combinations of ele-
ments of the construction kit. Our HRTEM investigations
Z. Anorg. Allg. Chem. 2004, 630, 2319Ϫ2328
zaac.wiley-vch.de
2004 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
2327
Deiseroth, Hans-Joerg
