Oligomeric Rare-Earth-Metal Halide Clusters
Inorganic Chemistry, Vol. 35, No. 24, 1996 7057
elements employed were powdered Ni, Ru, Rh, Pd, Os, Ir (Johnson-
Matthey, g99.9%), Re (Aldrich, 99.99%), Cr (AESAR, 99.5%), Fe
and Cu (reagent grade), chunks of Co (AESAR, 99.5%) and Mn
observed after quenching a reaction run at 1100 °C for 14 d,
demonstrating its relatively high stability. Observed product combina-
tions were sometimes in violation of the phase rule, indicating that
equilibrium had not been reached. Identification of all reaction products
was complicated by the number of phases found.
(
Johnson-Matthey, 99.99%), and Pt strips (government issue, reagent
grade).
The rare-earth-metal tribromides were prepared from the metals by
The analogous Sc16Br20
investigations of the corresponding Sc20-xBr28
Both occurred in more reduced (Sc-richer) systems, e.g., 20% Sc16
Br20Fe from compositions around Sc Br Fe, along with Sc-Fe binary
phases. The composition NaSc Br Fe at 840 °C for 24 d gave 60%
along with Sc and unreduced Na-Sc-Br compound(s) that
evidently served as a flux, but further heating of the product for 20 d
at 950 °C led to decomposition of the cluster phase. Likewise, a Sc
Z (Z ) Fe, Os) were encountered during
4
1
0
15
the ammonium bromide route, which proceeds via (NH
intermediates that are subsequently decomposed to form the RBr
which are then vacuum sublimed. Yttrium metal strips or dendrites
AESAR) were heated with a large excess of NH Br under Ar at 310
C until the evolution of H and NH YBr
ceased (∼3 d). The (NH
decomposed at 400 °C in vacuo to yield YBr (s), which was sublimed
was prepared in
4 3
) RBr
6
4
Z phases, as follows.
phases
-
3
4
3
3
(
°
4
4
4
)
4 3
6
Sc16Br20Fe
4
2
3
3
in a tube under dynamic vacuum at ∼800 °C. ScBr
3
3
-
1
3
a similar manner, with the corresponding formation, decomposition,
Br
amounts of an unknown, while a similar reaction at 950 °C for 2 w, or
of a NaSc Br Os composition at 840 °C or 950 °C, led to two different
unknowns and no Sc16Br20Os . The lattice parameters for Z ) Fe and
2 2 4 4
Os composition yielded ca. 50% Sc16Br20Os , Sc11Os , and small
and sublimation temperatures of 320, 550, and ∼700 °C, respectively.
Syntheses. Exploratory and directed synthetic reactions were carried
out on a scale of 150-200 mg within welded Nb tubes that were in
4
4
4
7
-9
Os determined from 22 and 34 indexed lines in the Guinier patterns
are a ) 10.996(1), 11.116(1) Å and c ) 16.021(4), 16.124(2) Å,
respectively.
turn jacketed by fused silica, as before.
gained if the interior of the loaded silica jacket was also cleaned with
an HF-H SO -HNO metal-cleaning solution and then rinsed just
before it was evacuated and sealed off. All transfers took place inside
a glovebox with e0.01 (vol) ppm of H O. Particular care was taken
to avoid the introduction of light impurities. Oxygen, usually from
O, was a persistent problem that inevitably manifested itself in the
Higher-purity products were
2
4
3
4 6
Y16Br24Ir . This was first observed after the composition CsY -
Br11Ir, had been heated to 850 °C, held there 4 d, and ramped (3 deg/
h) to 950 °C, where it was held for 2 w before cooling slowly. The
2
H
2
reaction product consisted of ∼50% Y16Br24Ir
black crystals grown in and around a mixture of YBr
(∼15%), and Cs Br (∼25%). A similar composition heated even
more slowly (15 d) to 950 °C followed by a 2 w reaction gave the
phases mentioned above plus Y20Br36Ir
4
as agglomerates of brittle
formation of the pervasive ROBr and upset the loaded stoichiometries.
This problem can be reduced by strong flaming of the walls of the
3
(∼10%), YOBr
Y
3 2
9
SiO
2 2
jacket under high vacuum just before sealing. Even so, H O
1
1
continues to evolve from the bulk silica and even to diffuse slowly
through the walls. Contamination from both sources increases with
time and temperature.
4
(below) in ∼20% yield. The
1
2
products formed appeared to depend on not only temperature but also
reaction time. The new phase was also synthesized in good yield in
analogous reactions with RbI or KI at 900 °C for 5 w. These products
Products were judged both visually and with the aid of Guinier X-ray
powder patterns, the yields of known phases being estimated (in terms
of equivalent scattering power) on the basis of calculated patterns and
their intensities. An internal standard of Si powder (NIST) added to
the ground products enabled lattice constants of known phases to be
refined from measured and indexed lines by least-squares means to a
9
also contained lines for Y
the presence of Rb Br
6
Br10Ir and Y
5
Ir
2
and weak lines evidencing
3
Y
2
9
or K
3
YBr
6
. (The formation of the ternary
alkali-metal yttrium(III) halide phases was generally favored over cluster
formation.) This oligomeric phase has been observed only in reactions
that contain alkali metals and therefore A
preferably with more YBr to compensate for this side product. There
is no evidence that any alkali metal is incorporated in the structure.
. Although the powder pattern of this compound was first
3 2 9 3 6
Y Br or A YBr as a flux,
4
3
few parts in ∼10 .
Y
16Br20Ru
4
and Analogues. The compound was first obtained from
, Y foil, and Ru powder loaded on stoichiometry
and heated at 980 °C for 24 d. This gave the new compound in 40%
yield as chunks and aggregates of black intergrown crystals, plus YBr
Y20Br36Ir
4
a reaction between YBr
3
observed several years ago, its quantification proved to be very
challenging owing to the difficulty of growing adequate single crystals.
3
,
1
3
These reaction mixtures ranged from Y
which gave either ∼20% each of nicely shaped black needles of Y
Br10Ir and aggregates of very small black Y20Br36Ir crystals or, with
larger Br:Y ratios, ∼40% of this phase with the balance as chunks of
YIr, YIr , Y Ir YOBr, and even small amounts of YBr . Reaction
mixtures in the range Y Br Ir to Y in
Br10Ir plus ∼15 mg of AlBr
each were heated very slowly and held at 750 °C, then 850 °C, and
3 2 2 2
Br Ir and Y Br Ir compositions,
YOBr (orange transparent plates), Y
another unidentified phase. Higher yields (60-85%) and better crystals
were obtained from reactions loaded as Y Br Ru, Y Br Ru, and Y
Br , which possibly
5
Ru
2
(irregular gray chunks), and
6
-
4
5
7
4
5
3
-
3
Ru, to all of which had been added ∼15 mg of AlBr
3
13
2
5
3
,
3
aided the formation of a gaseous reduced halide that transported metal
values. These systems were heated at 750 °C for 4 d, slowly ramped
to 850 °C and maintained there for 4 d, ramped again (3 deg/h) to 950
3
3
6
3
then 900-950 °C over a total of ∼24-39 d. The one loaded as Y
6
-
°C, and held there for 2 w, followed by a slow cooling to 700 °C before
Br10Ir gave nearly quantitative yield of that phase9 while those with
,16
turning off the furnace. (The slow heating was designed to lessen
nucleation rates, but there is no ensurance that this profile is essential.)
The products of particularly the bromine-richer runs included a small
amount of a compound that exhibited a powder pattern nearly identical
smaller Br:Y ratios gave moderate to high (g80%) yields of Y20Br36
Ir as well as small amounts (e25%) of an unknown phase (or phases).
Crystals of the target phase were still very small or poorly shaped.
Further heating of the products of a Y Br Ir reaction at 975 °C for 36
-
4
5
8
to that of Y20Br36Ir
theoretical yield from a reaction loaded as CsY
deficient) and heated under similar conditions. Other reaction products
included small amounts of YBr , YOBr, and, evidently, Cs Br as
the powder pattern contained the stronger lines calculated according
4
(below).
Y
16Br20Ru
4
was obtained in nearly
d followed by quenching gave instead ∼40% of an unidentified phase.
Temperature gradients or alkali-metal halide fluxes did not aid crystal
formation, the latter producing instead ternary alkali-metal yttrium-
6
Br10Ru (slightly Ru
3
3
Y
2
9
(
III) bromides and Y16Br24Ir
Ir composition heated at 975 °C for 36 d which gave ca. 50% of the
desired phase.
Structural Studies. Y16Br20Ru
4
. The best crystals came from a Y10Br13-
1
4
3
to the structure of Cs
Y
3 2
I
9
.
3 2 9
Cs Y Br (a ) 7.87 Å, c ) 19.68 Å)
was formed quantitatively in all reactions in which CsBr was present
and appeared to function as a useful flux. The oligomer was also
4
. Initial screening of crystals was
done with the aid of oscillation and Laue photographs. A blocklike
black chunk from a stoichiometric reaction was used for data collection
at room temperature on a Rigaku AFC6R diffractometer equipped with
Mo KR radiation. Twenty reflections from a random search indexed
to a primitive tetragonal cell, and Laue class 4/mmm was suggested by
the diffractometer software and confirmed by axial photos. A
hemisphere of data (h,(k,(l) was collected for 1° < 2θ < 50°, followed
by measurement of three ψ scans which indicated a relatively large
(
(
(
10) Meyer, G.; D o¨ tsch, S.; Staffel, T. J. Less-Common Met. 1987, 127,
155.
11) Holland, L. The Properties of Glass Surfaces; Chapman and Hall:
London, 1966; Chapter 4.
12) Schaeffer, H. A. In Concise Encyclopedia of Semiconducting Materials
and Related Technologies; Mahajan, S., Kimerling, L. C., Eds.;
Pergamon Press: Tarrytown, NY., 1992; p 457.
13) Villars, P.; Calvert. L. D. Pearson’s Handbook of Crystallographic
Data for Intermetallic Phases, 2nd ed.; American Society for Metals
International: Metals Park, OH, 1991; Vols. 3 and 4.
(
(15) Steinwand, S. J.; Corbett, J. D. Unpublished research.
(14) Guthrie, D. H.; Meyer, G.; Corbett, J. D. Inorg. Chem. 1981, 20, 1192.
(16) Payne, M. W.; Corbett, J. D. Inorg.Chem. 1990, 29, 2246.