Gupta et al.
Tt ) Si, Ge.14 Ternary antimonide examples with rare-earth
elements that reinforced our interests in this area were the
three intergrowth structures reported by Ganguli, et al.15 in
the Mg-La-Sb system. These novel Zintl phases can be
described in terms of coherent intergrowths between either
Mg5Sb4 or Mg2Sb2 units derived from the Mg3Sb2 structure
and Lan+1Sbn slabs from LaSb (NaCl-type structure). The
longer and hence presumably weaker Mg(oct)-Sb bonds at
3.11 Å in comparison with stronger Mg(tetrah)-Sb bonds
that average 2.84 Å provide the required flexibility for the
rare-earth element substitution at the former site. However,
the observed (001)-(111) intergrowth was believed to be
additionally constrained to LnSb phases that have a/x2
values comparable to the a dimension of Mg3Sb2 (4.56 Å).
It was realized that further possibilities of deriving novel
ternary phases in this kind of system were not exhausted,
and this was the sole motivation behind the present work.
One possibility would be to substitute tin for antimony to
eliminate the cation defects, and another more obvious one
would be similar intergrowths with other rare-earth metal
antimonides. However, the latter proved unsuccessful for Ln
) Pr, Nd, Gd, and Tb, only Mg3Sb2 and LnSb forming as
major products. Lattice mismatch that induced strain was
thought to preclude intergrowths with these other trivalent
rare-earth element antimonides. On the other hand, attempted
substitution of divalent example Sm did afford the new
superstructure variation of Mg3Sb2 reported here.
600 °C yielded none of the new phase. Subsequently, a series of
three stoichiometries, that of the crystal composition and two that
were richer in Sm and poorer in Mg, were studied under three
different reaction conditions in order to obtain the new phase more
cleanly and in higher yields. Slow cooling from 850 °C yielded
moderate amounts of the product, some with noticeably large cells.
However, quenching of the samples from 950 °C followed by 10
day equilibrations at 700 °C gave the highest yields, in particular
85% 5-1-4 plus only SmSb from a Mg4Sm2Sb4 composition, the
latter suggesting that this loading was too rich in Sm. At the same
time, the cell volumes were observed to increase appreciably, from
275.3 Å3 for the structurally refined product through intermediate
values above 290 Å3 to the maximum seen for the last sample,
292.6 Å3 (a ) 4.686 Å, c ) 15.388 Å). The new phase is not very
sensitive to air at room temperature.
Substitution of Yb in the smaller parent structure was also
observed with different stoichiometries. However, Lu instead in
similar proportions and conditions did not lead to the superstructure
phase. Instead, only Mg3Sb2 and LuSb were formed, which suggests
that there are likely dipositive charge limitations to stabilizing
substitutional products in this structure.
2.2. Powder X-ray Diffraction. The powder diffraction patterns
of all products obtained with the aid of a Huber 670 camera and
Cu KR1 radiation were used for phase identification. Samples were
mounted inside a nitrogen-filled glovebox between two Mylar sheets
by means of a block-and-ring assembly and some vacuum grease
to hold and to seal the sample from the air. The presence of the
title and other phases were established by matching the observed
patterns with those of known phases generated from their lattice
parameters, space groups, and atomic positions, the last for the title
phase according to the single-crystal analysis.
2. Experimental Section
2.3. Single-Crystal Structure Determination. A few small black
irregular crystals of the title compound were isolated from the
2.1. Syntheses. Single crystals of a Mg5.231(8)Sm0.769(8)Sb4
(nominally 5-1-4) composition according to later single-crystal
analysis were obtained after a mixture of the high-purity elements
as Mg turnings (99.99%, Ames Laboratory), Sm powder (Research
Chemicals, 99.9%), and Sb chunks (99.99%, Johnson-Matthey)
was allowed to react at high temperature in Ta or Nb tubing (1/4
in. diam) welded under an Ar atmosphere. The tubes were further
jacketed in evacuated and sealed silica containers and heated in
resistance furnaces. The new phase was first recognized via high-
quality Guinier patterns obtained with the aid of a Huber 670
Guinier powder camera equipped with an area detector. The 5-1-4
phase was present in only ca. 30% yield following a reaction of
the nominal composition Mg2.5Sm0.5Sb2 in Ta at 850 °C for 7 days
followed by radiative cooling. However, four other phases were
also present, Mg3Sb2 (∼15%), Ta5Sb4 (∼10%), SmSb (∼20%), and
Sm3-xMgxSb2 (∼25%). The last, presumably with random substitu-
tion of Sm for Mg, was recognized (in several instances) from its
Mg3Sb2-like powder pattern, mainly through an increase in d values
of the reflections. This particular system was possibly not at
equilibrium. (The Ta5Sb4 side product from attack of the Ta
container was, as before in related systems,16 avoided thereafter
by the use of Nb.) An identical composition Mg2.5Sm0.5Sb2 yielded
∼80% of the new (5-1-4) phase along with ∼20% SmSb when
the elemental components were allowed to react at 700 °C for 10
days in a Nb container. On the other hand, the same reaction mixture
quenched from 850 or 1000 °C followed by annealing at 400 or
products from the first reaction (with a loaded composition Mg2.5
-
Sm0.5Sb2) and placed inside glass capillaries (0.3 mm) under a
nitrogen atmosphere. These were flame-sealed, and the crystals were
checked for singularity. An irregularly shaped one with approximate
dimensions of 0.08 mm × 0.07 mm × 0.07 mm was used for the
X-ray crystallographic analysis at 293 (2) K. Data were collected
on a Bruker SMART17 APEX CCD system equipped with an area
detector, a graphite monochromator, and a Mo KR fine-focus sealed
tube (λ ) 0.71073 Å). The detector was placed at a distance of
4.995 cm from the crystal. A total of 1800 frames were collected
for the indicated trigonal cell with ω - 2Θ scans, a scan width of
0.3°, and an exposure of 10 s/frame. The total data collection time
was approximately 8 h. The frames were integrated by means of
the Bruker SAINT software package17 and the narrow-frame
integration algorithm. The process yielded a total of 1865 reflections
to a maximum θ angle of 28.20° (0.75 Å resolution), of which 433
were independent and observed [I > 2σ(I)] (completeness ) 91.9%,
Rint ) 10.44%). Analysis of the data showed negligible decay
during the measurements. Data were corrected for absorption effects
with the aid of the multiscan technique SADABS.18 Some data
collection and refinement parameters are tabulated in Table 1.
The structure was solved by direct methods and refined by least-
2
squares methods on F0 with the Bruker SHELXTL (Version 6.1)
software package.19 Space group P3h (No. 147) was chosen on the
basis of the lack of any systematic absences and the E value
(14) Gladyshevskii, E. J.; Kripyakevich, P. I.; Bodak, O. I. Ukr. Phys. J.
1967, 12, 447.
(17) Bruker SMART and SAINT; Bruker AXS Inc.: Madison, WI, 1998.
(18) Sheldrick, G. M. SADABS user guide; University of Go¨ttingen:
Go¨ttingen, Germany, 1997.
(19) Sheldrick, G. M. SHELXTL, ver. 6.10; Bruker AXS, Inc.: Madison,
WI, 2001.
(15) Ganguli, A. K.; Kwon, Y, -U.; Corbett, J. D. Inorg. Chem. 1993, 32,
4354.
(16) Ganguli, A. K.; Gupta, S.; Corbett, J. D. Inorg. Chem., 2006, 45, 196.
8176 Inorganic Chemistry, Vol. 45, No. 20, 2006