FIVE TERNARY ZINTL PHASES IN THE SYSTEMS A}In}Bi
437
Five ofthese compounds are the subjects ofthe present
products were always multiphase mixtures. Perhaps, some
paper. The corresponding systems ofalkaline-earth }In}Bi ofthe indium is lost as solid solutions with the niobium
and alkali-metal}Ga}Bi were also explored. Until now, all container (no Nb}In compounds are known).
tested reactions proved unsuccessful, yielding compounds of
known structure types (below). However, the A}In}Bi
phases (A"Na, K, Rb), surprisingly, dissolve in
ethylenediamine, and, in the presence of2,2,2-crypt, com-
pounds with exactly the proposed above deltahedral clus-
ters of[In Bi ]ꢂ\ can be crystallized from the solution (17).
PHASE AND ELEMENTAL ANALYSIS
Phase analysis ofthe products was carried out by powder
X-ray di!raction on an Enraf-Nonius Guinier camera (un-
der vacuum) with CuKꢀ radiation. For that matter a repre-
ꢀ
ꢁ
ꢆ
sentative portion ofthe "nely ground sample is isolated
ofa compound in the Cs }In}Bi system, Cs In Bi (18). Here
between two pieces ofcellophane tape together with NBS
NIST) silicon as an internal standard. The tape protects the
specimen from contact with oxygen and moisture during
transfer from the glove box to the camera.
ꢃ
ꢀ ꢄ
ꢂ
we report the synthesis and structures ofNa In Bi (I),
(
ꢅ ꢂ
Na InBi (II), K In Bi (III), K In Bi (IV), and
ꢂ
ꢅ
ꢀ ꢄ
ꢃ
ꢀ
ꢄ
ꢆꢆ
ꢄ ꢇ
Rb In Bi (V).
ꢃ
Qualitative microprobe analysis was performed on single
crystals from all samples and con"rmed the presence of
alkali metal, In, and Bi only.
SYNTHESIS
All manipulations were performed in a glove box with
puri"ed N (oxygen and moisture levels below 1 ppm).
The elements (all from Alfa-Aesar, 99.9#% pure) were
STRUCTURE DETERMINATION
ꢅ
Single-Crystal X-Ray Diwraction
used as received. The surfaces of Na and K were cleaned
with a scalpel immediately before use. Mixtures of the
elements scaled to about 500 mg per reaction were loaded
in niobium containers which were then closed by arc-
welding under argon, and were sealed in fused-silica am-
poules under high vacuum. These assemblies were heated at
Several crystals from each sample were picked in the
glove box and sealed in thin-walled glass capillaries. Over-
all, all phases were very poorly di!racting, and many crys-
tals had to be checked in order to select relatively good ones
for data collection. Data were collected for I through IV on
an Enraf-Nonius CAD4 single-crystal di!ractometer with
monochromated Mo Kꢀ radiation at room temperature
8
503C for 3 days and were then slowly cooled to room
temperature at a rate of5 3C/h. Mixtures ofelements with
compositions of A In Bi (A"Na, K, Rb) were reacted
(
ꢁ}2ꢂ scans, 2ꢂꢀ "503). Compound V is isostructural
ꢁ
ꢁ ꢀ
ꢁꢂ
initially in an attempt to synthesize heteroatomic clusters of
with III and only the lattice parameters were determined.
The data sets were corrected for Lorentz and polarization
e!ects, and for absorption with the aid of 6ꢃ-scans for I, and
[
In Bi ]ꢁ\ isoelectronic with the known homoatomic
ꢁ
ꢀ
nido-¹tꢀ\ (¹t"Si, Ge, Pb) (9}13). Instead, the sodium
ꢇ
reaction yielded two coexisting phases identi"ed later as
3
ꢃ-scans for II, III, and IV. Subsequently, the data were
Na In Bi (I) and Na InBi (II). The potassium analog
further treated for absorption using the Xabs (DIFABS)
software but some problems with absorption persisted,
and this is re#ected in some ofthe R-factors (19). The
unit cell parameters for all compounds were con"rmed by
axial photographs. The structures were solved by direct
methods and were re"ned on Fꢅ with the aid ofthe
collection and structure re"nements are given in Table 1,
while positional along with equivalent isotropic displace-
respectively.
ꢂ
ꢅ
ꢂ
ꢂ
ꢅ
a!orded the simultaneous formation of K In Bi (III) and
ꢃ
ꢀ ꢄ
K In Bi (IV), while the rubidium reaction yielded only
ꢆꢆ
Rb In Bi (V) and traces ofan amorphous phase. All ofthe
ꢄ
ꢇ
ꢃ
ꢀ ꢄ
products also contained traces ofNaBi, KBi , or RbBi and,
ꢅ
ꢅ
often, unreacted In and/or Bi.
After the structures and compositions of the I, II, III, and
IV were determined from single-crystal X-ray di!raction,
many reactions were loaded with the exact compositions in
an attempt to produce each compound as a single phase and
with better crystallinity. Also, a variety oftemperature pro-
"
les were probed including rapid heating to 9503C followed
by a rapid quenching to !1963C (breaking the hot quartz
ampoule under liquid N ), slow heating to moderate tem-
Na In Bi . A full sphere of data was collected on
ꢅ
ꢂ ꢅ ꢂ
peratures ofca. 550 3C, followed by annealing for prolonged a dark-to-black, irregularly shaped single crystal of
periods ofca. 3 }4 weeks and then slowly cooling to room I (0.25;0.15;0.10 mm). Axial photographs along the or-
temperature. Attempts were also made to synthesize each thorhombic axes con"rmed mmm Laue class which, together
compound starting from binary precursors such as KBi
with the observed systematic absences, indicated two pos-
ꢅ
and KIn , for example. In addition, an excess of one of the sible space groups, Pnma and Pna2 . The intensity statistics
ꢀ
ꢆ
components was used to presumably serve as possible suggested the centrosymmetric space group, and the struc-
&#ux.'' Nevertheless, despite all these e!orts, the resulting ture was successfully solved and re"ned in Pnma (R "4.14,
&
ꢆ
Bobev, Svilen
