Hydrogen impurity effects. A5Tt3 intermetallic compounds between A = Ca, Sr, Ba, Eu and Tt = Si, Ge, Sn with Cr5B3-like structures that are stable both as binary and as ternary hydride and fluoride phases

Article abstract of DOI:10.1006/jssc.2001.9144

All of the binary systems Ca, Sr, Ba, or Eu (A) with Tt (tetrel) = Si or Ge as well as Sr-Sn form both binary Cr₅B₃-type A₅Tt₃ phases and the corresponding ternary hydrides with stuffed Cr₅B₃- (Ca₅Sn₃F-) type structures. All of those tested, Ca-Si, Ba-Si, Ca-Ge, also yield the isotypic A₅Tt₃F_x phases. The tetragonal structures of Ca₅Si₃, Ca₅Si₃F_0.42, Sr₅Si₃, Eu₅Si₃H_x, Ca₅Ge₃, Ca₅Ge₃H_x, Ca₅Ge₃F_0.66(I4/mcm, No. 140) and of Ba₅Si₃F_0.16 (P4/ncc, Ba₅Si₃-type) were refined from single-crystal X-ray diffraction data. The interstitial H, F atoms are bound in a constricted tetrahedral (A²⁺)₄ cavity in the Cr₅B₃-type heavy atom structure, which can be described ideally as (A²⁺)₅(Tt₂)^6-(Tt)^4-. Many of 14 previous reports of the phases reported here were apparently hydrides according to lattice constant differences or, for Sr₅Si₃, the fractional coordinates of Sr2 about the tetrahedral site. An articulated model is developed that allows description of the relationship between the dimensions of the tetrahedral interstitial site and the cation cavity about Tt₂ and for some matrix effects in this structure type. The model suggests limitations on the stability of these binary A₅Tt₃ compounds for the heavier tetrels, as observed. The resistivities of Ca₅Ge₃ and Ca₅Ge₃H_x are both characteristic of poor metals, and Pauli-like magnetic susceptibilities are exhibited by Ca₅Ge₃, Ca₅Ge₃H_x, Ca₅Ge₃F_0.66, Sr₅Ge₃, and Sr₅Sn₃. The characteristic ideal Tt₂^6- dimers are evidently not realistic descriptions for these phases; rather at least some of the η*⁴ electrons in the dimers are delocalized in a conduction band. This effect appears to be greater in two europium salts. Bond lengths of dimers in the Ca-Si and Ca-Ge families appear to shorten slightly in three instances of their oxidation to form the hydride or the fluoride, as might be expected.

Full text of DOI:10.1006/jssc.2001.9144

Journal of Solid State Chemistry 159, 149}162 (2001)
doi:10.1006/jssc.2001.9144, available online at http://www.idealibrary.com on
Hydrogen Impurity Effects. A5Tt3 Intermetallic Compounds
between A ؍ Ca, Sr, Ba, Eu and Tt ؍ Si, Ge, Sn with Cr5B3-like
Structures that are Stable Both as Binary and as Ternary Hydride
and Fluoride Phases
E. Alejandro Leon-Escamilla and John D. Corbettꢀ
Department of Chemistry and Ames Laboratory~DOE,ꢁ Iowa State University, Ames, Iowa 50011
Received November 16, 2000; in revised form March 7, 2001; accepted March 15, 2001; published online May 11, 2001
Key Words: hydrogen impurities; Cr5B3-type structures;
alkaline-earth metal triels A5Tr3; structures of Cr5B3-type
phases; synthesis of Cr5B3-type compounds; interstitial binding
of H؊, F؊.
All of the binary systems Ca, Sr, Ba, or Eu (A) with Tt
(tetrel)؍ Si or Ge as well as Sr+Sn form both binary Cr5B3-type
A5Tt3 phases and the corresponding ternary hydrides with stu4ed
Cr5B3- (Ca5Sn3F-) type structures. All of those tested, Ca+Si,
Ba+Si, Ca+Ge, also yield the isotypic A5Tt3Fx phases. The
tetragonal structures of Ca5Si3, Ca5Si3F0.42, Sr5Si3, Eu5Si3Hx,
Ca5Ge3, Ca5Ge3Hx, Ca5Ge3F0.66 (I4/mcm, No. 140) and of
Ba5Si3F0.16 (P4/ncc, Ba5Si3-type) were re5ned from single-crystal
X-ray di4raction data. The interstitial H, F atoms are bound in
a constricted tetrahedral (A2 ؉)4 cavity in the Cr5B3-type heavy
atom structure, which can be described ideally as
1. INTRODUCTION
Compounds formed between alkali, alkaline-earth, or
rare-earth metals and the main group p-elements have been
major features in the development of diverse structural and
(A2 ؉)5(Tt2)6 ؊(Tt)4 ؊. Many of 14 previous reports of the phases bonding aspects of solid state chemistry. The main-group
reported here were apparently hydrides according to lattice con-
components most often feature the tetrels (Si family, ¹t),
stant di4erences or, for Sr5Si3, the fractional coordinates of Sr2
about the tetrahedral site. An articulated model is developed that
allows description of the relationship between the dimensions of
the tetrahedral interstitial site and the cation cavity about Tt2 and
for some matrix e4ects in this structure type. The model suggests
limitations on the stability of these binary A5Tt3 compounds for
the heavier tetrels, as observed. The resistivities of Ca5Ge3 and
Ca5Ge3Hx are both characteristic of poor metals, and Pauli-like
magnetic susceptibilities are exhibited by Ca5Ge3,
Ca5Ge3Hx, Ca5Ge3F0.66, Sr5Ge3, and Sr5Sn3. The characteristic
pnictogens (P family principally), and chalcogens. One com-
mon feature of many of these products is the predominance
of valence or closed shell compounds, the so-called Zintl
phases (1), although many of these have been so classi"ed
solely on the basis of their structures (2). Contradictions or
inconsistencies between our results and those in literature,
or, equally often, an inability to obtain a reported com-
pound in high yield, have repeatedly forced us into a consid-
eration of phase stabilization via unrecognized impurities.
ideal Tt62 ؊ dimers are evidently not realistic descriptions for these Carbon, nitrogen, and oxygen are naturally common candi-
phases; rather at least some of the p 4 electrons in the dimers are
*
dates, but hydrogen has turned out to be a particularly
delocalized in a conduction band. This e4ect appears to be
greater in two europium salts. Bond lengths of dimers in the
Ca+Si and Ca+Ge families appear to shorten slightly in three
instances of their oxidation to form the hydride or the 6uoride, as
might be expected. ( 2001 Academic Press
pervasive problem, both because its presence is usually not
seriously considered and the fact that the atom is essentially
undetected in heavy atom compounds by common X-ray
di!raction methods for phase detection or analysis.
Our "rst systematic exploration stemmed from an inabil-
ity to sort out the formation of Mn Si - from b-Yb Sb -
ꢀTo whom correspondence should be addressed. E-mail: jcorbett@
iastate.edu.
ꢂ ꢃ
ꢂ
ꢃ
type pnictides (Pn) of the alkaline-earth and dipositive rare-
ꢁThis research was supported by the O$ce of the Basic Energy Sciences, earth metals. The problem was solved with the discovery
Materials Sciences Division, U. S. Department of Energy. The Ames
Laboratory is operated by DOE by Iowa State University under Contract
No. W-7405-Eng-82. Accordingly, the U.S. Government retains a nonex-
clusive, royalty-free license to publish or reproduce the published form
of the contribution, or to allow others to do so for U.S. Government
purposes.
that all nine examples assigned to the latter structure type
''
were in fact valence hydrides (A ) Pn H (3, 4) in what is
ꢂ
ꢃ
better termed the orthorhombic Ca Sb F structure type (5).
ꢂ
ꢃ
This was the "rst instance in our work in which #uoride,
which is easily located by modern X-ray di!raction means,
149
0022-4596/01 $35.00
Copyright ( 2001 by Academic Press
All rights of reproduction in any form reserved.

150
LEON-ESCAMILLA AND CORBETT
was utilized as a good stand-in for hydride, although this well as for the stu!ed version as the hydride, which is
relationship has been known for much longer (6). Evidently, another future topic (8).
the most prevalent sources of hydrogen impurities have
The presence of a signi"cant tetrahedral cavity in the
been commercial alkaline-earth metals, although most ac- Cr B structure type was "rst signaled by the serendipitous
ꢂ ꢃ
tive metals can become contaminated with careless hand- discovery of a stu!ed Cr B -type structure for La Pb O
ꢂ ꢃ
ꢂ
ꢃ
ling. (Hydrogen is commonly retained during their reactions (via dirty lead) (13). The lead atoms here are all monomeric
with adventitious water.) Examples in the same Mn Si - in the presence of these electron-richer cations, and the
ꢂ ꢃ
type systems in which the pnictide hydride forms "rst, or analogous nitride would appear to be electron-precise. (A
only, by utilizing a well-characterized interstitial hole (7) in similar disposition of monoanions is also known in the
that structure type will be described in a forthcoming manu- distinctive so-called In Bi subfamily (14).) We choose to
ꢂ
ꢃ
script (8). Other speci"c hydride errors have included the distinguish the structure of La Pb O and its absence of
ꢂ
ꢃ
Zintl phases Ba Ga H (9) and Ba M O H
(M"Si, a dimer from the situation in the new Ca Sn F structure
ꢂ
ꢄ ꢁ
ꢁꢀ ꢁ ꢂ ꢁꢁ>V
ꢂ ꢃ
Ge, Ga, etc., x"2,0) (10), the structures of which were type where the dimeric anion is retained, although both
originally reported without the hydrogen components.
Similar events have shadowed the chemistry of the related
types occur in the same space group and are isopointal.
The protocol developed earlier for studies of the
A ¹t phases of the tetrel (¹t) elements Si}Pb. These are Mn Si H- vs Ca Sb F-type phase problems (4) was again
ꢂ
ꢃ
ꢂ ꢃ
ꢂ
ꢃ
frequently encountered with tetragonal Cr B -type struc- applied to test for the formation of hydride derivatives as
ꢂ ꢃ
tures with the characteristic equal numbers of monomeric well as to ensure the absence of impurity hydrogen in checks
and dimeric anions, meaning that those with divalent ca- for the formation of true Cr B types and the ternary
ꢂ ꢃ
tions formally meet the structural criterion (2) for Zintl #uorides. Although it is possible to clean the alkaline-earth
>ꢁ
\ꢄ
\ꢅ
phases, i.e., (A ) (¹t )(¹t ) in terms of oxidation states. metals, etc., aforehand by careful dissociation of hydrogen
ꢂ
ꢁ
One might even imagine that these closed-shell con"gura- in vacuum, it is generally far easier to simply load the
tions should be immune to oxidation and to the inclusion of reactants in tantalum tubing, weld this shut, and heat the
hydride without a necessary change of structure type. In assembly in high vacuum at, say, 900}12003C. The metals
fact, the problems here are particularly insidious because and any hydride compounds evidently all have appreciable
not only is oxidation of the substrate by hydrogen generally H dissociation pressures under these conditions, and tanta-
ꢁ
possible, but the hydride formed is also readily incorporated lum is permeable to only hydrogen in our systems above
into a suitable cavity already present in the parent Cr B
structure. We have recently reported on seven instances gen-free system can then be made at suitable temperatures.
550}6003C. Subsequent equilibrations of a virtually hydro-
ꢂ ꢃ
with tin or lead for which only the stu!ed Cr B structure On the other hand, Ta sample containers loaded with hy-
ꢂ ꢃ
type is stable in the new Ca Sn F structure type (or the drogen sources, e.g., CaH , are equilibrated only after their
ꢂ
ꢃ
ꢁ
hydride equivalent), namely, for the Ca}Sn, Ba}Sn, Eu}Sn, further enclosure in sealed fused silica jackets in order to
Yb}Sn, Sr}Pb, Ba}Pb and Eu}Pb systems (11). All that retain the hydrogen. Additional smaller amounts of hydro-
were tested also form the corresponding isotypic A ¹t F. gen can still originate from water dissociation from the
ꢂ
ꢃ
Thus, major hydride contamination was evidently the fused silica over long periods or at higher temperatures.
source of ten reports of these particular binary combina- These two routes will hereafter be designated as the dy-
tions as simple tetragonal A (¹t) phases. In these systems, namic vacuum (dv) and sealed (SiO ) container (sc)
ꢂ
ꢃ
ꢁ
the truly binary A ¹t compositions occur in another struc- methods.
ꢂ
ꢃ
ture type or, in one instance, Ba}Sn, that composition exists
as a mixture of two other binary phases. In fact, such serious
hydride errors had evidently precluded the discovery of the
2. EXPERIMENTAL SECTION
oligomeric Sr Pb , which has a very similar composition,
Materials utilized were: Ca, Sr, Ba (Ae) from APL Engin-
eering, distilled and sealed in glass; Eu and Yb, Ames Labor-
ꢃꢀ ꢁꢆ
60.8 at.% Sr vs 62.5 at.% for Sr Pb , until 1993 (12).
ꢂ
ꢃ
The present article reports on nine more complicated atory 599.95% total; Si, electronic grade, Ge and Sn,
instances in which both the binary and the ternary (hydride) Aesar 5}9's or better. The AH phases were prepared by
ꢁ
phases occur with the same heavy atom Cr B structure direct reaction of H with the metals held in a Mo boat
ꢂ ꢃ
ꢁ
type. This situation naturally means that X-ray lattice di- within a closed vacuum system that was equipped with
mensions are probably the quickest and best way to tell a diaphragm manometer. The reactions were run to comple-
them apart and, in retrospect, usually the only means avail- tion at 420 to 7003C under 600}700 torr H , depending on
ꢁ
able. Of course, comparison of dimensional data from di!er- the dissociation characteristics of the particular hydrides
ent time periods and equipment naturally entails some (15). All were treated as stoichiometric AH phases, as
ꢁ
uncertainties. It should be noted that still six more members veri"ed by their cell dimensions (15, 16). The AeF reagents
ꢁ
of the A ¹t family, apparently only with A"Sm and Yb, were precipitated from AeCO slurries with excess HF(aq),
ꢂ
ꢃ
ꢃ
exhibit a Mn Si type structure for the binary (usually) as "ltered, washed, and vacuum-dried at 4003C.
ꢂ ꢃ

HYDROGEN IMPURITY EFFECTS
151
In particular cases, the as-received alkaline-earth metals given from 14 literature reports of these as binary phases.
were dehydrogenated after sealing them in Ta containers. All in this group have the particular problem that each
These were heated in silica tubing under high vacuum system contains both a binary A ¹t and a ternary
ꢂ
ꢃ
\ꢂ
(410 torr), namely, at 6503C for Ca and at 7103C for the A ¹t H with the same heavy atom structure, namely for,
ꢂ
ꢃ V
remainder for &12 h or until the vacuum stayed below that Ca, Sr, Ba, Eu with either Si or Ge, plus Sr}Sn. Hydride
producing a Tesla coil discharge after a small increase in formation can be best judged on a routine basis by the
temperature.
customary decrease in cell dimensions and volumes (0.3 to
All reactions between the elements with or without #uor- 2%) following the reaction of a A ¹t H composition in
ꢂ
ꢃ ꢁꢇꢆ
ide or hydride were carried out in welded Ta containers that a sealed container relative to those for Ae ¹t binary phases
ꢂ
ꢃ
were heated either under vacuum or within evacuated, well- synthesized under dynamic vacuum conditions to ensure the
#amed, and sealed silica jackets, depending on the target. absence of signi"cant hydrogen. Furthermore, all of those
All reagents and products were handled in gloveboxes with tested, Ca Si , Ba Si , and Ca Ge , also form the character-
ꢂ ꢃ
ꢂ ꢃ
ꢂ
ꢃ
0.2}0.4 ppm (vol) H O levels. These techniques as well as istic #uoride A ¹t F (dv), and their #uoride contents
ꢁ
ꢂ
ꢃ V
those for Guinier powder pattern measurements (with NIST have all been quanti"ed by single crystal X-ray di!raction
Si as an internal standard) and lattice constant least-squares studies.
re"nement methods have been described in more detail
This condition is not true for the hydrides for which their
before (4, 17, 18). A CuKa wavelength of 1.54056 A was hydrogen contents were not quanti"ed. The magnitudes of
ꢀ
used for this purpose. The phase distributions in mixed the hydrogen dissociation pressures are unavailable in all
products were estimated in terms of equivalent X-ray scat- cases, either at 6503C, the "nal temperature of the equilibra-
tering powers in their powder patterns with the aid of tions, or at some higher temperature should the uptake rate
calculated patterns for known phases. Although the accu- become insigni"cant relative to a cooling rate of 103C/h.
racies of these estimates vary perhaps by $2}5%, they are (Substantially full hydrogen occupancy in Ca Sn D was
ꢂ
ꢃ
internally more precise then that. All phases seen are re- established earlier from a neutron di!raction study (11).) Of
ported. Single crystals of eight of the compounds reported course, a parallel uncertainly also exists with regard to the
here were selected under low magni"cation and mounted in rate of change of cell dimensions with variation of the
thin-walled glass capillaries in a glovebox for X-ray di!rac- hydrogen content. Assessments as to whether, by compari-
tion. These were "rst checked for quality with Laue photo- son, earlier reports of the supposedly binary A ¹t phases
ꢂ
ꢃ
graphs. The data collection and re"nement parameters are were troubled by unrecognized hydrogen impurities are also
detailed in Table 1. Absorption was corrected for with the naturally dependent on the accuracies of the earlier data,
aid of three psi-scans collected at s5803 and, after iso- whatever the di!raction technique. But earlier studies of
tropic re"nement, with DIFABS except for Ca Ge . Struc- both the Cr B -type A ¹t (11) and &&b-Yb Sb -type''
ꢂ
ꢃ
ꢂ ꢃ
ꢂ
ꢃ
ꢂ
ꢃ
tural re"nements were made with the TEXSAN (19) A Pn (4) systems in which only the ternary hydrides are
ꢂ
ꢃ
package. Lattice dimensions secured from calibrated stable suggest that contamination has been pervasive. Of
Guinier powder pattern data were used in all distance course, the occurrence of a negative result in a really hydro-
calculations from single-crystal results. The F /F listings gen-free system of this latter type is virtually impossible to
ꢀ
#
are available from J.D.C.
assess unless a phase diagram is reported or a virtually
Magnetic susceptibility data were secured for seven quantitative yield of another product is obtained. We also
single-phase weighed samples, each held between two fused assume that a continuous variation of dimensions occurs in
silica rods that "t snugly inside a 5-mm o.d. tube in a con- the present single phase A ¹t }A ¹t H systems because
ꢂ
ꢃ
ꢂ
ꢃ V
tainer sealed under He. The measurements were made with we have never seen otherwise.
the aid of a Quantum Design MPMS SQUID magneto-
meter at 3 T and between 6 and 300 K. The data were
corrected for the susceptibilities of both the apparatus and
3.2. Results
the atomic cores in the sample. Resistivities of two phases
were measured by the Q-method at &35 MHz on an im- Eu all form A Si H hydrides from loaded compositions
proved apparatus (20).
3.2.1. Silicides. The A Si phases for A"Ca, Sr, Ba,
ꢂ ꢃ
ꢂ ꢃ V
A Si H when "rst reacted (sc) at 11003C and then slowly
ꢂ ꢃ ꢁꢇꢆ
cooled to 6503C (conditions in Table 2). We assume x41.0
3. RESULTS AND DISCUSSION
because of structural limitations (below). In order to get
good yields for binary Ca}Si and Sr}Si and ternary
Ca}Si}F compositions, we found it necessary to heat these
3.1. Syntheses
Table 2 lists conditions of 22 representative reactions run "rst to 13003C, near the melting points, before slow cooling
for these systems, the products identi"ed from by powder to 6003C. This follows our earlier discovery that reactions to
patterns, and the Guinier lattice constants of the tetragonal form Ca Sb F and Ca Bi F from the elements and CaF
ꢂ
ꢃ
ꢂ
ꢃ
ꢁ
Cr B -type phases obtained (34). Dimensional data are also were very incomplete when carried out at 9003C, even
ꢂ ꢃ

152
LEON-ESCAMILLA AND CORBETT
TABLE 1
Data Collection and Re5nement Parameters
Ca Si
Ca Si F
Sr Si
Ba Si F
ꢂ
ꢃ
ꢂ
ꢃ ꢆꢇꢅꢁ
ꢂ
ꢃ
ꢂ
ꢃ ꢆꢇꢀꢄ
Space group
I4/mcm (No. 140)
2.178
0.12;0.20;0.32
Enraf}Nonius
CAD4
I4/mcm
2.333
0.16;0.30;0.36
Enraf}Nonius
CAD4
I4/mcm
3.368
0.15;0.18;0.26
Enraf}Nonius
CAD4
P4/ncc (No. 130)
4.445
0.12;0.25;0.30
Enraf}Nonius
CAD4
ꢃ ?
Density calculations (g/cm )
Crystal dimensions (mm)
Di!ractometer
Collected octants
Scan type
2h limit (deg.)
h,!k,$l
u!2h
54
!h,!k,$l
u!2h
54
!h,!k,$l
u!2h
54
h,!k,$l
u!2h
54
\ꢀ
Absorption coe$cient (cm
Rel. transm. coe!. range
Number of re#ections
)
33.3
0.833}1.00
33.7
0.797}1.00
252.2
0.697}1.00
167.0
0.752}1.00
@
Meas. (obs. )
2182 (837)
700 (275)
5.55
2173 (812)
807 (299)
7.79
2571 (744)
937 (287)
2898 (1893)
1115 (673)
9.75
@
Indep. (obs. )
A
R
(%)@
15.8 (7.85 )
!ꢁ%
No. of re"ned variables
16
21
16
23
R/R (%)B
1.5/1.9
1.120
2.5/2.6
0.946
3.5/3.7
1.166
3.2/3.6
1.927
ꢂ
Goodness of "t
\ꢈ
Sec. extinc. coe!. (10
Max./min. in *F map (e
)
11.6(9)
0.25/!0.26
14(2)
0.43/!0.42
4.5(7)
1.35/!1.21
6.7(4)
0.240/!1.03
\
\ꢃ
A
)
Eu Si H
Ca Ge
Ca Ge H
Ca Ge F
ꢂ
ꢃ
V
ꢂ
ꢃ
ꢂ
ꢃ
V
ꢂ
ꢃ ꢆꢇꢄꢄ
Space group
I4/mcm
I4/mcm
I4/mcm
I4/mcm
ꢃ ?
Density calculations (g/cm )
Crystal dimensions (mm)
Di!ractometer
5.887
3.134
3.185
3.303
0.18;0.24;0.32
Enraf}Nonius
CAD4
0.11;0.16;0.22
Enraf}Nonius
CAD4
0.16;0.22;0.25
Enraf}Nonius
CAD4
0.17;0.20;0.28
Enraf}Nonius
CAD4
Collected octants
Scan type
2h limit (deg.)
!h,!k,$l
u!2h
54
h, k,$l
h, k,$l
h,$k$l
u!2h
56
u!2h
u!h
56
56
\ꢀ
Absorption coe$cient (cm
Rel. transm. coe!. range
Number of re#ections
)
327.1
0.673}1.00
127.2
0.546}1.00
129.0
0.661}1.00
128.5
0.761}1.00
@
Meas. (obs. )
2164 (828)
821 (301)
4.30
16
2.1/2.4
1.926
2443 (817)
824 (281)
6.70
16
1.9/2.2
1.171
2397 (849)
1184 (430)
7.23
16
2.9/3.0
1.527
3104 (1176)
832 (321)
10.9
21
2.1/2.4
0.981
@
Indep. (obs. )
R
(%)@
!ꢁ%
No. of re"ned variables
R/R (%)B
ꢂ
Goodness of "t
\ꢈ
Sec. extinc. coe!. (10
)
7.4(2)
4.0(5)
1.3(6)
9(1)
\
\ꢃ
A
Max./min. in *F map (e
)
1.10/!1.08
0.67/!0.59
0.94/!0.75
0.62/!0.88
?Lattice parameters obtained from calibrated Guinier powder pattern data are in Table 2.
@All data I'0.
AObserved re#ections; I53p .
'
BR"+""F "!"F ""/+"F "; R "[+w("F "!"F ")ꢁ/+w(F )ꢁ]ꢀꢉꢁ; w"p\ꢁ.
ꢀ
#
ꢀ
ꢂ
ꢀ
#
ꢀ
$
though a liquid phase is in equilibrium with Ca Sb (s) ours. This probably happened because their thermal analy-
ꢂ
ꢃ
(incongr. m.p. &9963C) for the former (35), whereas a few sis samples were sealed in Ta and "rst melted ('12403C)
hours at 12503C was much more e!ective (5). Lattice con- through induction heating, which should have removed any
stants given in the 1974 report of Ca Si (21) (Table 2), if hydrogen. The single-crystal X-ray di!raction results for
ꢂ ꢃ
accurate, indicate a substantial hydride impurity was pres- Ca Si and Ca Si F will be described later.
ꢂ ꢃ
ꢂ ꢃ ꢆꢇꢅ
ent, whereas the very recent report on the Ca}Si phase
Analogous results pertain to Sr Si and Sr Si H com-
ꢂ ꢃ
ꢂ ꢃ ꢁꢇꢆ
diagram by Manfrinetti and coworkers (22) gave Guinier- positions treated under dv and sc conditions, respectively,
based lattice constants that are in good agreement with Table 2. The cell volume from the 1967 report of the binary

HYDROGEN IMPURITY EFFECTS
153
TABLE 2
Reaction and Lattice Data for A5Tt3 Phases with Cr5B3-Type Structures that Are Stable as Binary as well as
Ternary Hydrides and Fluorides
@
ꢃ
Approximate yields, %
Lattice parameters of the Cr B -type phase (A,A )
ꢂ ꢃ
?
Loaded
Ca Si
Condition
C
Other
a
c
<
a/c
dv
sc
1
2
1
90
10 CS
7.6495(4)
7.6386(3)
7.6354(4)
7.64
14.835(2)
14.747(1)
14.829(2)
14.62
868.1(1)
860.5(1)
864.5(1)
853
0.516
0.518
0.515
ꢂ
ꢃ
Ca Si H
'95
ꢂ
ꢃ ꢁꢇꢆ
ꢃ
Ca Si F
dv
75
10CS, 15CaF
ꢂ
ꢁ
Ref. (21)
Ref. (22)
7.641(1)
14.876(2)
868.5
0.514
Sr Si
dv
sc
1
2
95
&100
un
8.0966(5)
8.0468(4)
8.055
15.703(2)
15.761(1)
15.688
1029.9(2)
1020.5(1)
1017.9
0.516
0.510
0.513
0.514
ꢂ
ꢃ
Sr Si H
ꢂ
ꢃ
ꢁꢇꢆ
A
Ref. (23)
Ref. (24)
8.089(4)
15.733(8)
1029 (1)
Ba Si A
dv
sc
dv
2
2
1
90 (B)
un
8.4855(6)
8.4558(4)
8.4726(2)
8.436(6)
8.465(6)
16.421(3)
16.484(2)
16.443(1)
16.53(1)
16.45(1)
1182.4(3)
1178.6(1)
1180.3(1)
1176 (1)
1179 (1)
0.517
0.513
0.515
0.510
0.515
ꢂ
ꢃ
Ba Si H
'95 (B)
ꢂ
ꢃ
ꢁꢇꢆ
ꢁꢇꢆ
Ba Si F
80 (B)
15BaF , un
ꢂ
ꢃ
ꢁ
Ref. (25)
Ref. (26)
Eu Si
Eu Si H
dv
sc
3
2
40
95
20M, un
5CB
7.919(2)
7.9052(4)
7.9188(7)
15.384(6)
15.281(2)
15.322(2)
964.9(5)
954.9(1)
960.8(2)
0.515
0.517
ꢂ
ꢃ
ꢂ
ꢃ
Ref. (27)
Ca Ge
dv
sc
dv
dv
2
2
3
2
95
&100
'90
60
7.7156(5)
7.7218(5)
7.7093(3)
7.7101(3)
7.74
14.885(2)
14.660(2)
14.789(2)
14.787(2)
14.66
886.1(1)
874.2(1)
879.0(1)
879.0(1)
878
0.518
0.527
0.521
0.521
ꢂ
ꢃ
Ca Ge H
ꢂ
ꢃ
ꢁꢇꢆ
Ca Ge F
ꢃ ꢀꢇꢆ
5CB, (5 CaF
30CB, 10CaF
ꢂ
ꢁ
Ca Ge F
ꢂ
ꢃ ꢀꢇꢂ
ꢁ
Ref. (21)
Sr Ge
dv
sc
2
2
95
&100
5CB
8.1675(4)
8.1236(5)
8.134(4)
15.730(3)
15.732(2)
15.771(8)
1049.3(1)
1038.2(2)
1043 (1)
0.519
0.516
0.516
ꢂ
ꢃ
ꢃ ꢁꢇꢆ
Sr Ge H
ꢂ
Ref. (28)
Ba Ge
dv
sc
2
2
'95 (B)
8.557(1)
8.5106(5)
8.519(4)
16.465(1)
16.549(2)
16.554(8)
1205.6(4)
1198.7(2)
1201 (1)
0.520
0.514
0.515
ꢂ
ꢃ
Ba Ge H
95 (B)
5CB
ꢂ
ꢃ
ꢁꢇꢆ
Ref. (29)
Eu Ge
dv
sc
2
2
90
&100
10CB
7.9987(8)
7.9853(5)
7.990(1)
15.380(3)
15.281(2)
15.367(1)
984.0(2)
974.4(1)
981.0(1)
0.520
0.522
0.520
ꢂ
ꢃ
Eu Ge H
ꢂ
ꢃ
ꢁꢇꢆ
Ref. (30)
Sr Sn
dv
sc
2
2
'95
8.5757(8)
8.5694(5)
8.54(1)
8.565
8.567(2)
16.318(2)
16.027(3)
16.06(2)
16.261
1200.1(2)
1176.9(2)
1171 (3)
1192.9
0.525
0.535
0.532
0.527
0.527
ꢂ
ꢃ
Sr Sn H
95
5CB
ꢂ
ꢃ ꢁꢇꢆ
Ref. (31)
Ref. (32)
Ref. (33)
16.254(7)
1192.9(8)
?
Conditions: dv"dynamic vacuum, sc"sealed silica jacket; 1: 13003C for 6 h, 83C/h to 6003C, cool; 2: 11503C for 6 h, 10}123C/h to 6503C; 3: 12003C
for 6 h, 203C/h to 11003C, 6 h, 103C/h to 6003C.
@
Structural types: C, Cr B ; CS, Co Si; B, Ba Si distortion of Cr B ; CB, orthorh. CrB; CS, Co Si; M, Mn Si .
Reduced symmetry.
ꢂ ꢃ
ꢁ
ꢂ
ꢃ
ꢂ ꢃ
ꢁ
ꢂ ꢃ
A
(23) clearly pertained to the hydride, again, if the data were into the probable hydride region. A comparison of the
accurate. Although the cell volume from a 1999 single- positional data (below) make this categorization more
crystal report for Sr Si (24) agrees closely with ours for the certain.
ꢂ ꢃ
binary phase, the cell in fact reapportions appreciably on
The Ba Si compound is well known for its primitive
ꢂ ꢃ
hydrogen uptake, and their individual cell dimensions (re- distortion of the Cr B -type parent (abbreviated B in the
"ned from multiple di!ractometer data) put their sample table) (25), as will be described later for the comparable
ꢂ ꢃ

154
LEON-ESCAMILLA AND CORBETT
Ba Si F . This is another case in which the lattice constants
move in opposite directions on hydrogen uptake. The lattice dride with a 1.0% volume contraction. The recently re-
constants from both the 1966 (25) and the 1999 (26) studies ported single-crystal study of Eu Si (27) may have involved
at face value indicate appreciable hydrogen contamination. some hydrogen although their (Guinier) lattice constants
What was new at the time, Eu Si (34), also forms a hy-
ꢂ ꢃ V
ꢂ ꢃ
ꢂ ꢃ
TABLE 3
Re5ned Positional and Displacement Parameters for Ca5Tt3(H,F)x Phases with Cr5Sn3F-Type Structures
?
ꢁ
(A )
@
Atom
x
y
z
B
;
;
;
;
;
;
%ꢃ
ꢀꢀ
ꢁꢁ
ꢃꢃ
ꢀꢁ
ꢀꢃ
ꢁꢃ
Ca Si
Ca1
Ca2
ꢂ
ꢃ
0
0
0
1.27(2)
1.00(2)
1.13(2)
0.83(2)
0.0108(3)
0.0141(2)
0.0147(3)
0.0109(2)
U
U
U
U
0.0267(4)
0.0099(2)
0.0134(5)
0.0097(3)
0
0
0
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
0.17978(3)
0
0.38693(6)
x#ꢀ
0
x#ꢀ
0.13986(2)
!0.0008(1) !0.0005(1)
U
0
ꢁ
ꢀꢃ
Si1
Si2
Ca Si F
ꢀ
ꢅ
0
0
0
0
!0.0005(2)
0
ꢁ
ꢂ
ꢃ ꢆꢇꢅ
Ca1
0
0
0
1.64(3)
1.34(2)
1.27(3)
0.88(2)
0.9(1)
0.0127(4)
0.0178(3)
0.0167(6)
0.0119(3)
0.015(2)
U
U
U
U
U
0.0371(7)
0.0153(3)
0.0150(7)
0.0096(4)
0.005(3)
0
0
0
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
Ca2
Si1
Si2
F#
0.17407(5)
x#ꢀ
0
0.14386(3)
0.0017(2) !0.0039(1)
0
0.0006(3)
0
U
ꢁ
ꢀꢃ
0
0.38724(8)
0
ꢀ
ꢅ
0
0
0
0
0
0
0
x#ꢀ
ꢁ
ꢀ
ꢀ
ꢁ
ꢅ
Sr Si
ꢂ
ꢃ
Sr1
Sr2
0
0
0
1.27(5)
0.74(2)
1.0(1)
0.0077(7)
0.0106(4)
0.013(2)
0.008(1)
U
U
U
U
0.033(1)
0.0070(5)
0.013(3)
0.006(2)
0
0
0
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
0.18003(7)
0
0.3920(3)
x#ꢀ
0
x#ꢀ
0.14008(6)
!0.0014(3) !0.0010(3)
U
0
0
ꢁ
ꢀꢃ
Si1
Si2
ꢀ
ꢅ
0
0
0
0
0.59(9)
!0.000(1)
ꢁ
Ba Si F
B
ꢂ
ꢃ ꢆꢇꢁ
Ba1
ꢀ
ꢀ
0.2716(1)
0.10612(3)
0.4915(3)
2.11(3)
1.42(3)
1.4(1)
1.00(6)
0.8(5)
0.0203(3)
0.0206(3) 0.0187(3) 0.0147(4)
0.020(1)
0.0157(8)
U
0.0394(8)
0
!0.0008(2)
0
0
0
ꢅ
ꢅ
ꢀꢀ
Ba2
Si1
Si2
0.56191(5)
0.08065(5)
0.0046(2) !0.0043(3)
0
!0.0011(9)
ꢀ
ꢀ
U
U
0.013(3)
0.007(2)
0
ꢅ
ꢅ
ꢀꢀ
ꢀꢀ
0.6481(2)
!x
ꢀ
ꢅ
0
0.002(1)
U
ꢀꢃ
F
3/4
ꢀ
ꢅ
Eu Si H
ꢂ
ꢃ
V
Eu1
Eu2
Si1
Si2
(H
0
0
0
1.98(3)
1.09(2)
1.1(1)
0.0134(4)
0.0157(3)
0.015(1)
0.012(1)
U
U
U
U
0.0485(8)
0.0099(3)
0.011(3)
0.008(2)
0
0
0
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
0.17414(4)
x#ꢀ
0
0.14786(3)
!0.0001(2) !0.0024(1)
U
0
0
ꢁ
ꢀꢃ
0
0.3925(3)
0
ꢀ
ꢅ
0
0
0
0
x#ꢀ
0.86(8)
0.001(1)
ꢁ
ꢀ
ꢀ)
ꢁ
ꢅ
Ca Ge
ꢂ
ꢃ
Ca1
0
0
0
1.18(4)
0.95(2)
1.04(2)
0.70(1)
0.0079(5)
0.0122(3)
0.0124(3)
0.0074(2)
U
U
U
U
0.029(1)
0
0
0
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
Ca2
Ge1
Ge2
0.17807(8)
0
0.38200(6)
x#ꢀ
0
x#ꢀ
0.13998(6)
0.0117(4)
0.0148(5)
0.0117(3)
!0.0008(3) !0.0006(2)
U
0
0
ꢁ
ꢀꢃ
ꢀ
ꢅ
0
0
0
0
!0.0004(3)
ꢁ
Ca Ge (H )
ꢂ
ꢃ
V
Ca1
0
0
0
1.50(5)
1.09(2)
1.08(3)
0.85(2)
0.0116(6)
0.0153(3)
0.0136(3)
0.0114(3)
U
U
U
U
0.034(1)
0
0
0
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
Ca2
Ge1
Ge2
(H
0.16797(8)
x#ꢀ
0
0.14852(8)
0.0108(6)
0.0139(6)
0.0096(4)
!0.0001(3) !0.0018(3)
U
0
0
ꢁ
ꢀꢃ
0
0.38296(6)
0
ꢀ
ꢅ
0
0
0
0
x#ꢀ
!0.0004(3)
ꢁ
ꢀ
ꢀ)
ꢁ
ꢅ
Ca Ge F
ꢂ
ꢃ ꢆꢇꢊ
Ca1
0
0
0
1.69(4)
1.26(2)
1.16(2)
0.81(1)
1.4(2)
0.0118(6)
0.0167(3)
0.0151(3)
0.0108(2)
0.020(3)
U
U
U
U
U
0.041(1)
0.0146(4)
0.0138(4)
0.0091(3)
0.013(3)
0
0
0
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
Ca2
Ge1
Ge2
FC
0.16964(7)
x#ꢀ
0
0.14617(5)
0.0013(3) !0.0036(2)
0
0.0002(2)
0
U
ꢁ
ꢀꢃ
0
0.38224(4)
0
ꢀ
ꢅ
0
0
0
0
0
0
0
x#ꢀ
ꢁ
ꢀ
ꢀ
ꢁ
ꢅ
?Wycko! positions for Ca1, Ca2, Si1, Si2, and F atom positions are 4c, 16l, 4a, 8h and 4b, respectively, and site symmetries for the same atoms are 4/m,
..m, 422, m2m and 42m, respectively.
@¹"exp [!2pꢁ(; hꢁa*ꢁ#; kꢁb*ꢁ#; lꢁc*ꢁ#2; hka*b*#2; hla*c*#2; klb*c*)].
ꢀꢀ
ꢁꢁ
ꢃꢃ
ꢀꢁ
ꢀꢃ
ꢁꢃ
AOccupancy"0.42(1).
BP4/ncc, origin at 1. Occupancy of F"0.16(2).
COccupancy"0.66(2).

HYDROGEN IMPURITY EFFECTS
155
inexplicably di!er from ours only in c, which shows 60% of to have been uniquely hydrogen-stabilized (4, 8, 11). The
the contraction we observe on hydride formation.
1994 result from our labs (33), synthesized from Sr metal
from the same source without any special cleaning, appears
to have been roughly one-third saturated. As an aside, we
have also shown that none of the 3d elements V}Ni intro-
duced as potential interstitials a!ord any alternative to the
3.2.2. Germanides. The Ca}Ge system likewise contains
the binary Ca Ge as well as the stu!ed hydride and #uor-
ꢂ
ꢃ
ide. The cell volume from the 1974 report (21) on Ca Ge
ꢂ
ꢃ
suggests some hydride, but the values are probably not
accurate enough for that judgement. A second #uoride
Cr B -type binary product. The search for a possible stu!ed
ꢂ ꢃ
Mn Si -type Sr Sn M was predicated on the formation of
ꢂ ꢃ
ꢂ
ꢃ
reaction loaded as Ca Ge F
ꢃ ꢀꢇꢂ
(except for excess CaF ) as for Ca Ge F according to the
yielded the same product
ꢂ
this sort of product for Ca Pb M (36).
ꢂ
ꢃ
ꢁ
ꢂ
ꢃ ꢀꢇꢆ
two sets of lattice data. This indicates that equilibrium had
been attained at the upper limit of #uoride, which was 3.3. Structures
re"ned to have an occupancy of &0.66 (below).
The structures of eight of the compounds reported here
were re"ned with the aid of single-crystal X-ray di!raction
data. General collection and re"nement parameters were
given in Table 1. Table 3 contains the atomic positional and
displacement ellipsoid data, and Tables 4 and 5 list impor-
tant distances in these. Seven occur as the classical or the
stu!ed Cr B (Ca Sn F) structure type (I4/mcm), whereas
The germanides Sr Ge , Ba Ge , and Eu Ge also show
ꢂ
ꢃ
ꢂ
ꢃ
ꢂ
ꢃ
clear di!erentiation from the respective ternary hydrides
according to our experiments. The recently reported c di-
mension for the Sr Ge (28) falls outside of the range we
ꢂ
ꢃ
observed, and so the comparison is inconclusive. Compari-
son of dimensions reported for Ba Ge (29) (from multiple
ꢂ
ꢃ
ꢂ ꢃ
ꢂ
ꢃ
di!ractrometer data) indicate a substantial hydrogen in-
the eighth, Ba Si F , exhibits the characteristically dis-
ꢂ ꢃ ꢆꢇꢁ
volvement, and for Eu Ge with Guinier data (30), possibly
ꢂ
ꢃ
torted structure known for its parent Ba Si (P4/ncc) (25).
ꢂ ꢃ
The stu!ed Cr B -type tetragonal structure is illustrated
ꢂ ꢃ
but not seriously so. The Ba Ge and its hydride both occur
ꢂ
ꢃ
with distorted Ba Si type (P4/ncc) structures according to
ꢂ ꢃ
for the Ca Ge F example in Fig. 1. The structure type has
ꢂ
ꢃ ꢆꢇꢄ
both our powder data and a single-crystal study of the
two characteristic features: (i) the equal distribution of
main-group tetrel anions (solid spheres) between monomers
former (29).
ꢀ
ꢃ
3.2.3. Stannide. Finally, Sr Sn , which evidently is the (in layers at z" and ) and dimers (in the mirror planes at
ꢂ
ꢃ
ꢅ
ꢅ
ꢀ
only stannide or plumbide that forms both a binary and z"0, ), each with di!erent orientations in its repeat (i.e.,
ꢁ
ꢀ
a ternary hydride with the same structure, has been reported from centering and the mirror plane at z" ); (ii) interstitial
ꢁ
three times before. The "rst in 1977 (31) had clear hydrogen tetrahedral cavities de"ned by Ca2 atoms (shaded) around
ꢀ ꢀ
ꢀ ꢃ
involvement, whereas the 1978 report (32) was possibly so (0, , ), (0, , ), etc. which may be occupied by H, F, or
ꢁ ꢅ
ꢁ ꢅ
troubled, judging solely from the dimensions; however, four nothing. These two features are also shown in normal views
ꢀ
other 5:3 products described in the same article are known looking in ꢀc in Fig. 2: (a) the section around z" that
ꢁ
TABLE 4
Important Interatomic Distances (A) in Cr5B3-like Structures
Atom 1}Atom 2
Ca Si
Ca Si F
Ca Ge
Ca Ge H
Ca Ge F
Sr Si
Eu Si H
ꢂ
ꢃ
ꢂ
ꢃ ꢆꢇꢅ
ꢂ
ꢃ
ꢂ
ꢃ
V
ꢂ
ꢃ ꢆꢇꢄ
ꢂ
ꢃ
ꢂ
ꢃ
V
Tt2}Tt2 1;
¹t1}A1 2;
¹t1}A2 8;
¹t2}A1 2;
¹t2}A2 2;
¹t2}A2 4;
A1}A2 8;
2.447(1)
2.435(2)
3.7073(5)
3.2306(3)
3.0795(5)
3.1383(9)
3.0930(6)
3.5370(4)
3.549(1)
3.759(1)
4.267(1)
4.120(1)
3.9899(3)
2.4516(5)
2.575(1)
3.7213(5)
3.2771(5)
3.0848(4)
3.048(1)
3.1264(8)
3.5212(6)
3.632(2)
3.886(2)
4.167(2)
4.275(2)
4.0143(4)
2.556(1)
2.5678(9)
3.6972(5)
3.2488(4)
3.0835(2)
3.1695(9)
3.1212(7)
3.5875(5)
3.536(1)
3.699(2)
4.323(2)
4.034(1)
4.0489(4)
2.4038(8)
2.474(7)
3.9267(3)
3.276(3)
3.4380(5)
3.293(2)
3.249(2)
3.6987(6)
3.807(2)
4.124(2)
4.401(2)
4.519(2)
4.2046(4)
2.403(7)
3.7088(5)
3.2498(3)
3.0836(3)
3.0539(6)
3.0957(5)
3.4923(3)
3.6039(8)
3.8897(7)
4.149(1)
3.6650(5)
3.2356(6)
3.0922(4)
3.202(1)
3.121(1)
3.6050(8)
3.473(2)
3.669(2)
4.355(2)
3.947(2)
4.0634(5)
3.8203(3)
3.217(2)
3.3116(3)
3.327(3)
3.215(1)
3.6947(4)
3.553(1)
3.894(1)
4.519(1)
4.1622(9)
4.1306(3)
?
A2}A2 1;
@
A2}A2 1;
A
A2}A2 1;
B
A2}A2 2;
4.2714(7)
3.9728(2)
A2}A2 4;
A2}Z
C
C
C
C
C
(2.540)
(2.600)
(2.362)
(2.688)
(2.4952)
?
Unshared side of (A2) .
Top of tetrahedron"top across (¹t2) (viewed along ꢀc).
ꢈ
@
A
ꢁ
Height of trigonal prism.
B
Side of tetrahedron.
To centroid of tetrahedron.
C

156
LEON-ESCAMILLA AND CORBETT
TABLE 5
Important Distances (A) in Ba5Si3F0.2
Atom 1}Atom 2
Atom 1}Atom 2
Si2}Si2
Si1}Ba1
Si1}Ba1
2.442(6)
3.617(5)
4.606(5)
3.549(3)
3.601(2)
3.500(1)
3.378(1)
3.458(1)
3.538(2)
Ba1}Ba2 4;
3.7991(9)
4.055(1)
3.892(1)
4.2887(9)
4.623(1)
4.2526(6)
4.558(1)
2.7646(5)
Ba1}Ba2 4;
@
Ba2}Ba2
A
Si1}Ba2 4;
Si1}Ba2 4;
Si2}Ba1 2;
Ba2}Ba2
A
Ba2}Ba2 2;
B
Ba2}Ba2 2;
?
B
Si2}Ba2 2;
Ba2}Ba2 2;
?
Si2}Ba2 2;
Ba2}F
Si2}Ba2 2;
?
About waist of dimer.
Unshared edge of (Ba2) .
@
ꢈ
A
Ba F tetrahedron.
Within squares.
ꢅ
B
contains the Ge2 dimers (black) (one site is highlighted) and
ꢀ
both Ca1 and Ca2 atoms, and (b) the section around z"
ꢅ
with the isolated Ge1 atoms, Ca2, and the shaded tetrahed-
ral holes that bind H or F. Note that the squares of Ca2
atoms at the corners and center in each view are common to
both parts, square prisms centered by Ca1 at the top and
distorted square antiprisms centered by Ge1 in the bottom
illustration. The same is true for the top and bottom edges
of the (Ca2) tetrahedra in (b) and the edges above and
ꢁ
below the waist of the dimers in (a).
FIG. 2. [001] sections in the Ca Ge F
* compare with Fig. 1. (a) the slab centered at z"0.5 (and at z"0 with an
altered orientation) with a pair of trigonal prisms surrounding one dimer
(Ca Sn F-type) structure
ꢂ
ꢃ ꢆꢇꢄꢄ
ꢂ
ꢃ
highlighted; (b) the slab centered at z"ꢀ with the characteristic Ca2
ꢅ
tetrahedra surrounding #uoride shaded. The atom legends are the same as
in Fig. 1.
Compounds in this structural family that have small
anions and large cations exhibit a characteristic distortion
to the primitive Ba Si type structure, P4/mcc (25). These
ꢂ ꢃ
versions have recently been re"ned for Ba Si (H) and
ꢂ ꢃ
Ba Ge (H) (26, 29), whereas we saw the characteristic pow-
ꢂ
ꢃ
der pattern changes for both the two binaries and their
hydrides and re"ned the structure of Ba Si F . The last is
ꢂ ꢃ ꢆꢇꢁ
illustrated in Fig. 3 around the former 0,0,z axis, Fig. 1. (The
FIG. 1. An o!-[100] view of the centered tetragonal structure of
ꢀ ꢀ ꢀ
structure change includes an origin shift of , , .) The typi-
Ca Ge F
with vertical. The smaller open and larger solid spheres
ꢅ ꢅ ꢅ
ꢂ
ꢃ ꢆꢇꢄꢄ
cal distortion in this case displaces both Ba1 and Si1 by
represent Ca and Ge, respectively, and the interstitial F is centered in the
shaded (Ca2) tetrahedra.
#0.355(2) and !0.140(5) A out of the former respective
ꢅ

HYDROGEN IMPURITY EFFECTS
157
with ours is useful are those for Sr Si (24). Although the cell
ꢂ ꢃ
dimension alone suggests probable hydrogen contamina-
tion, the fractional coordinates establish this more clearly.
Those for Sr2, which de"ne the tetrahedra, show the princi-
pal di!erences, as would be expected on examination of all
aspects of the structure (the variables for Si2 di!er by (1p
while the other atom types lie in "xed positions). The re-
ported x and z coordinates of Sr2 di!er from ours by
!0.00188 (24p) and #0.00154 (23p), di!erences that are
18.6 and 18.0% of the characteristic shifts observed in single
crystal data for Ca Ge H relative to Ca Ge (Table 3). This
ꢂ
ꢃ
ꢂ
ꢃ
shift makes d(Sr2-centroid) 0.031 A smaller than in the pres-
ent binary. (The results for Sr Ca Si , x"0.38, 0.48, and
ꢂ\V V ꢃ
Sr Si
Ge
(37) are similar but the last is not quite as
ꢂ ꢁꢇꢈꢊ ꢆꢇꢀꢃ
extreme. The reverse analysis, Eu Si H vs Eu Si (27), is
ꢂ ꢃ V
ꢂ ꢃ
not as clear.) Both studies agree that the distorted structure
(I4cm) attributed to Sr Si in 1967 (23) was not attained.
ꢂ ꢃ
The three re"ned #uoride occupancies, 0.42, 0.66, 0.16 in
FIG. 3. A portion of the primitive tetragonal cell of Ba Si F
about
ꢂ
ꢃ ꢆꢇꢀꢄ
the Ca}Si, Ca}Ge, and Ba}Si systems, respectively, are
unusual in their contrast, although the cell volume decre-
ments in forming these from the clean binary components
are reasonably proportional to #uoride content. All were
achieved from A ¹t F loadings, but the same result was
ꢀ, ꢀ, z that shows the characteristic displacements of Ba1 and Si1 from the
ꢅ ꢅ
centers of the former square prismatic and antiprismatic sites de"ned by
Ba2, respectively.
ꢂ
ꢃ ꢀꢇꢆ
ꢀ
planes at z"0 and . As noted before (26), this likely occurs attained for Ca Ge F (Table 2), so equilibrium appears
ꢅ
ꢂ
ꢃ ꢀꢇꢂ
because of crowding in the once-Ba1-centered square prism, to have been attained. Our earlier article also reported on
which distortion now makes the faces of the prism of un- the structures of Ca Sn F and Sr Pb F in systems
ꢂ
ꢃ ꢆꢇꢈꢋ(ꢀ)
ꢂ
ꢃ ꢀꢇꢆ
equal size (Table 5). The proportions of the F-centered in which the binary version does not exist. The best proposi-
tetrahedron as well as separations in (¹t2) remain typical, tion, lacking more information, is that these "ve #uorine
ꢁ
both sites being relatively well removed from the distortion coe$cients represent the compositions in equilibrium with
sites.
AF (s) (plus, presumably, small amounts of other phases).
ꢁ
Binding of H or F in the nominal tetrahedra (42m) always The coe$cients generally increase with the size, and pre-
produces a contraction of the tetrahedral site and the lattice sumed reduced bond strengths, of the heavier tetrelide
(Table 6), evidence of the additional bonding therein. The dimers. In addition, the interstitials are second-nearest
gain from this is presumably responsible for tipping the neighbors (through A2) to, and would with increasing con-
scales of stability for the group of stannide and plumbide centration withdraw more n* electron density from the
examples reported earlier that exist only as ternary hydride heavier weaker dimers (below).
and #uoride phases. The sole instance in which a compari-
son of atom positions from published single crystal studies
3.4. Structural Generalizations
Some characteristic and signi"cant distances and their
TABLE 6
di!erences are summarized in Table 7 for the 12 A ¹t Z
Observed Cell Volume Decrements (%) between A5Tt3
ꢂ
ꢃ
a
b
compounds we have structurally re"ned in this work or
before (11). Although variations among the two or three
di!erent elements that make up the structures could result
in some substantial and complex relationships among them,
the results are remarkable for the regularities that can be
found among them. Distances in the tetrel dimers are quite
consistent within each type. These are 2.40}2.47 A among
four silicides with three cations, and 2.56}2.58 A for three
germanides with the same Ca cation, with or without H or
F (Table 7, col. 1). The dimers in the three stannides are
likewise more similar in length, 2.88}2.91 A, than might be
expected for the cation and interstitial variations. But, the
dimer distances found with europium with tin or silicon are
and A5Tt3Hx or A5Tt3Fx
A!¹t
Si
Ge
Sn
Ca
0.88
0.42
0.92
0.32
0.18
1.36
0.81
1.07
Sr
Ba
1.97
A
A
0.59
A
Eu
1.05
0.98
?
All loaded as A ¹t H and equilibrated down to 6503C.
ꢁꢇꢆ
Cr B - and Ca Sn F-type structures. The #uoride di!erences are on the
ꢂ
ꢃ
@
ꢂ ꢃ
ꢂ
ꢃ
second line.
A
Ba Si -type.
ꢂ
ꢃ

158
LEON-ESCAMILLA AND CORBETT
TABLE 7
Some Important Distance Parameters (A) in A5Tt3(Z) Phases with Re5ned Structures
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Matrix e!ect Matrix e!ect
on dimer:
*d(¹t2}A2),
end}waist
on dimer:
*d(A2}A2)
side-top
Tetrah.
radius,
A2}Z
Side/top
proportion
in tetrah.
Twist in (A2) :
Top edge unshared} shared d(A2}A2)
Square:
Radius of
ꢈ
¹t1 in
d(Tt}Tt)
of tetrah.
side edges
(4;)
(A2)
ꢈ
Ca Si
Ca Si F
2.447(1)
2.435(2)
2.474(7)
2.403(7)
2.575(1)
2.556(1)
2.568(1)
2.903(1)
2.910(1)
2.884(3)
3.106(3)
2.442(6)
!0.042
0.045
0.044
0.112
!0.078
0.08l
0.048
0.048
0.046
0.038
0.056
0.260
0.508
0.277
0.625
0.281
0.686
0.624
0.829
0.846
0.640
0.837
2.540
2.452
2.688
2.495
2.600
2.362
2.404
2.391
2.397
2.550
2.540
2.765
1.098
1.096
1.096
1.069
1.100
1.076
1.091
1.063
1.073
1.050
1.070
1.078
3.89
3.76
4.12
3.89
3.89
3.67
3.70
3.75
3.73
4.03
3.96
4.29
0.67
0.57
0.71
0.61
0.38
0.47
0.50
0.39
0.38
0.54
0.39
0.73
3.973
3.990
4.205
4.103
4.014
4.063
4.049
4.316
4.310
4.453
4.597
2.11
2.09
1.96
1.91
2.14
2.10
2.11
2.25
2.26
2.21
2.30
ꢂ
ꢃ
ꢃ ꢆꢇꢅ
ꢂ
Sr Si
ꢂ
ꢃ
?
Eu Si H
ꢂ
ꢃ
V
Ca Ge
ꢂ
ꢃ
Ca Ge H
ꢂ
ꢃ
V
Ca Ge F
ꢂ
ꢃ ꢆꢇꢄꢄ
@
ꢃ ꢆꢇꢋ
Ca Sn H
ꢂ
ꢃ
@
Ca Sn F
ꢂ
?ꢌ@
Eu Sn H
ꢂ
ꢃ
@
Sr Pb F
ꢂ
ꢃ
ꢃ ꢆꢇꢁ
A
Ba Si F
0.008, 0.160 0.448
4.253, 4.558 2.06, 2.11
ꢂ
?
@
A
Consistent deviations; see text.
Ref. (11).
Primitive Ba Si -type cell (B).
ꢂ
ꢃ
\
appreciably smaller than those in other examples of each binding H relative to the empty version always decreases
type and should probably be set aside. We will return to this the size of the tetrahedral hole, whereas the relative vari-
aspect later.
ations in the lattice constants, in a/c, and in the proportions
Factors that allow this general constancy of d(¹t2}¹t2) of the tetrahedra are somewhat diverse. Substitution of
\ \
F for H always results in a smaller expansion, consistent
seem worthy of further exploration. The dimers are seen to
run through the face that is common to a pair of trigonal with their comparative radii (37). The volume changes rela-
prisms of A2 cations (one is highlighted in Fig. 2a), and tive to empty phases were summarized in Table 6, and the
a reasonable concern might be whether the dimer distances radii of the approximate tetrahedra are given in Table 7, col.
are in#uenced by matrix e!ects (crowding) from the cations. 4. Variations in the proportions of the tetrahedra are enu-
(Or better, why are they not more a!ected?) The A2}¹t2 merated in col. 5, the ratio of side to top A2}A2 edges of the
distances occur in two categories, four equal values around tetrahedra (viewed along ꢀc ), and col. 6, the smaller top (and
the waist of the dimer (which lies on an inversion center) and bottom) edge lengths alone (normal to ꢀc ), respectively. The
two equal pair between the ¹t2 atoms and the far edge of nominal square antiprismatic surroundings about ¹t1 are
each antiprism. The di!erences between A}¹t distances at imperfect in that the alternating distances between the
the end and waist of the trigonal prism, a conceivable source squares, around the edges of the "gurative crown, are un-
of a matrix e!ect, are given in col. 2 of Table 7. (Of course, equal, the smaller A2}A2 interplanar separation always
the dimer is not electronically isotropic in these two direc- being the side edge of the tetrahedron, Fig. 2b. The di!er-
tions.) In the one example with a large positive value, ence between these, col. 7, Table 7, is identi"ed as the
ꢁ>
ꢄ\
Eu Si H , reduced attractions to Eu at the ends of Si
&&twist,'' the unshared minus the shared edge lengths. This
ꢂ ꢃ V
ꢁ
could conceivably cause the observed reduction in the Si-Si varies with reapportionment of the tetrahedron and be-
bond length, but this questionable attribution is not perti- comes less for larger ¹t.
nent to the other Eu problem, a short Sn}Sn separation.
An interesting and relatively simple coupling between the
A better electronic reason will be suggested later. The four tetrahedra and the sizes of the squares in the distorted A2
A2 atoms about the waist of (¹t2) actually de"ne a rec- antiprisms about ¹t1 can also be found. The latter naturally
ꢁ
tangle, longer alongꢀc, and the di!erence between the verti- increase with the size of A, the size of ¹t at "xed A, and with
cal and the horizontal edge is given in col. 3 for later substitution of F and then H at "xed A and ¹t. The "rst two
reference.
are fairly straightforward size e!ects, while for "xed A and
The means by which not only the isolated ¹t1 but also ¹t sizes there is a simple mechanical coupling of the squares
a potential small interstitial anion are accommodated in the with the top edge of the tetrahedron, as marked in Fig. 4b.
other layer in this structure (Fig. 2b) further illustrate the A decrease in the top and bottom edges on binding H or F is
versatility and speci"city of this structure type. The e!ect of coupled directly to a twist of the antiprism, CCW (counter-

HYDROGEN IMPURITY EFFECTS
159
ꢀ
faces of the antiprisms around z" are of course common
ꢅ
to those of the A1-centered square prisms around z"0,
ꢀ
. As shown in Fig. 4a, transmission of the twist in the
ꢁ
ꢀ
former (z" ) to the further face of the square prisms
ꢅ
ꢀ
centered around z" on interstitial binding changes the
ꢁ
proportions of the face-shared antiprisms around ¹t too.
ꢁ
This increases the distance between the ¹t2 atoms and the
end of the prism (;2), lessening any crowding there, and
decreases the distance between the waist A2 and ¹t2 (;4),
Fig. 4a. This is as observed in the di!erence, col. 2, Table 7.
Note that the A2}A2 pair at the top of the tetrahedron and
that across the waist of the dimer normal to ꢀc are the same
atoms, Figs. 1, 2, accounting for part of the changes just
elaborated.
These changes are also synchronized with a lengthening
ꢀ
of d(A2}A2) around z" and parallel to c, the height of the
ꢁ
trigonal prisms. Thus, a #attening of the tetrahedra and of
the square antiprisms is accompanied by a signi"cant in-
crease in the other variable along ꢀc , the overall A2-A2
ꢀ
heights of the layers around z" , and thence of the trigonal
ꢁ
prismatic pairs seen in projection around the waist of ¹t .
ꢁ
This means the four A2 atoms about the waist are disposed
in an increasingly rectangular array. These edge di!erences
in col. 3, Table 7, are over 0.80 A for the larger ¹t with Ca
and Sr. It is not clear that this increasingly anisotropic
environment around the waist of ¹t has a signi"cant e!ect,
ꢁ
although it will split the degeneracies of the n levels on the
dimers.
We suggest that the instabilities of binary A ¹t phases
ꢂ
ꢃ
with empty tetrahedra for all ¹t"Sn, Pb (except Sr Sn ),
ꢂ
ꢃ
but their stabilities as ternaries with bound hydride and,
where tested, #uoride (Ca}Sn, Ba}Sn, Eu}Sn, Sr}Pb,
Ba}Pb, Eu}Pb) (11), all arise because of the foregoing inter-
relationships. In other words, the absence of interstitial
binding yields an inappropriate sizing of the pair of con-
facial A2 trigonal prismatic sites around the larger ¹t ,
FIG. 4. Relative displacements of Ca2 in Ca Ge Z that accompany the
ꢂ
ꢃ
ꢁ
binding of H or F in a smaller tetrahedron. (a) at z"ꢀ, around the Ge
namely, with too-short A2}¹t distances at the opposite
ꢁ
ꢁ
units; (b) around z"ꢀ in which only displacements of A2 in the top
ends of this "gure, the reverse of the distortion marked in
Fig. 4a. A forecast of this is seen in the large negative value
of this di!erence for Ca Ge in col. 2, Table 7. Of course,
ꢅ
tetrahedral face and of the upper face of the square antiprism are marked.
ꢂ
ꢃ
instability is always with respect to an alterative, which in
clockwise, marked) in the top face and CW in the bottom, these cases is usually binary W Si -, Pu Pt -, or
ꢂ ꢃ ꢃꢀ ꢁꢆ
which simultaneously increase the length of the unshared Yb Sn -type structures.
ꢃꢄ ꢁꢃ
Finally, it is useful to estimate some radii for the tetrelide
side edge of the antiprism, col. 5. Decreases in the side edges
of the tetrahedron (those with components along ꢀc ) and anions. Those for the dimers bound in the confacial trigonal
a #attening of the antiprisms along ꢀc also follow increased prisms are de"ned by the two di!erent values, about the
bonding in the tetrahedron. The last occurs in a less predict- waist and at the ends (Fig. 2a). These distances may not be
able articulated way, although the #attening is naturally very transferable, and there is also the problematic localiza-
\ꢄ
accompanied by increases in the size of the A2 squares for tion of high-lying n* electrons on the ¹t dimers (below).
ꢁ
"xed A and ¹t, col. 8, Table 7, in order to maintain some- On the other hand, distances between the isolated, nominal
\ꢅ
thing near a "xed d(A2}¹t1).
¹t and A2 in the twisted antiprisms seem to represent
These articulated or coupled e!ects also extend to the something real and reproducible. For this purpose, col. 9 of
other section of the structure that contains the dimer, Table 7 contains those radii deduced from d(¹t1}A2) (;8)
a unique feature of this common structure type. The square on subtraction of the six-coordinate crystal radii of

160
LEON-ESCAMILLA AND CORBETT
ꢁ>
A
(39). The values obtained for Si, 1.96}2.11 A, Ge,
2.10}2.14 A, Sn 2.25}2.26 and Pb, 2.30 A are &0.4, 0.3, and
0.2 A larger than radii for the respective isoelectronic and
much less polarizable halide anions, a reasonable trend for
the increased polarizability of heavier ¹t. (The smaller
values for Eu salts are again omitted.)
3.5. Physical Properties
There is ample evidence that the parent A ¹t phases in
ꢂ
ꢃ
the Cr B -type structure with dipositive cations are not the
ꢂ ꢃ
classical Zintl phases suggested by their structural compo-
>ꢁ
\ꢅ
\ꢄ
nents (A ) (¹t )(¹t ), that is, with localized electrons
ꢂ
ꢁ
on the anions as indicated. Instead, these phases exhibit
both resistivities and paramagnetic susceptibilities indica-
tive of metallic characteristics. Not surprising, the oxidized
hydrides and #uorides thereof behave similarly. Figure 5
shows the resistivities measured by the &&Q'' method for
Ca Ge and Ca Ge H , both falling in the range of &30 to
FIG. 6. Molar magnetic susceptibility data (v) as a function of temper-
ature for Ca Ge , (᭿), Ca Ge H (᭢) and Ca Ge F (z).
ꢂ
ꢃ
ꢂ
ꢃ
V
ꢂ
ꢃ V
ꢂ
ꢃ
ꢂ
ꢃ V
&70 k) cm over the range shown with positive temper-
ature coe$cients (0.32(4) and 0.24(4)% K , respectively), the large elevation of the Ca Ge H susceptibility is un-
i.e., reasonable for poor metals. Figure 6 shows the fairly known. Similar Pauli-like results but with slightly larger
typical Pauli-like paramagnetic properties of Ca Ge (189), increases in v on cooling were reported earlier for
\ꢀ
ꢂ
ꢃ V
ꢂ
ꢃ
Sr Ge (163), and Sr Sn (144) as well as of Ca Ge H
Ca Sn H , Ca Sn F , and Sr Pb F with room temper-
ꢂ
ꢃ
ꢂ
ꢃ
ꢂ
ꢃ V
ꢂ
ꢃ V
ꢂ
ꢃ ꢆꢇꢋ
\ꢄ
ꢂ
ꢃ
\ꢀ
(7680), Ca Ge F (585), with the room temperature s value ature s values (10 emu mol ) of 39, 53, and 350, respec-
ꢂ
ꢃ V
\ꢄ
\ꢀ
(10 emu mol ) for each in parentheses. (These results tively (11).
may of course include van Vleck TIP contributions.) The
Ca Ge sample was known to contain &5% CaGe (CrB sizeable temperature dependencies of the susceptibilities are
type), which is presumably metallic, but this does not appear found for the two strontium hydride phases Sr Ge H and
In contrast to the regular behavior of binary Sr Ge ,
ꢂ
ꢃ
ꢂ
ꢃ
ꢂ
ꢃ V
to have had a noticable e!ect at that level. The reason for Sr Sn H , Fig. 7. The k (BM) and h(K), parameters that "t
ꢂ
ꢃ V
%&&
these to the Curie}Weiss function drawn in (without TIP
terms) are 1.67(5)/!161 and 1.59(4)/!64, respectively. The
tin sample was epr silent at liquid N temperature and
above. Such behaviors for main-group element compounds
ꢁ
are unusual and not understood.
4. GENERAL CONSIDERATIONS
The source of the metal-like properties of some, and
probably all, of the binary A ¹t phases considered here is
ꢂ
ꢃ
\ꢄ
not hard to imagine. The formal closed-shell ¹t ions are
ꢁ
isoelectronic with the following dihalogens, which are com-
ꢁ
ꢅ
*ꢅ
monly described in terms of the p-orbital states (p )(p )(n ).
'
ꢄ
'
Occupancy of localized antibonding states in a condensed
solid salt would be problematic, and certainly less favored
\ꢅ
energetically than the idealized ¹t octet states (in a cation
matrix, of course). At least some n* electrons must be
delocalized into the conduction band, so that the observed
ready oxidations by hydrogen or #uorine are not really
unusual given the clearly suitable means of accommodating
a small interstitial in the structure. Possible oxide and ni-
tride interstitials have not been explored in these systems
except for negative results for Ca Ge O , but these are
FIG. 5. Resistivity data for Ca Ge (᭡) and Ca Ge H (᭿) as a func-
tion of T (K).
ꢂ
ꢃ V
ꢂ
ꢃ
ꢂ
ꢃ
V
known in two other Cr B -type families, the electron-richer
ꢂ ꢃ

HYDROGEN IMPURITY EFFECTS
161
spill over into the conduction band. The presence of some
'''
Eu would not need to be contradictory in terms of n*
occupancies if the cation band were low enough, and this
would also be consistent with the &0.05}0.10 A smaller
values observed for the apparent radius of ¹t1 (Table 7, col.
9) for those two examples (as well as in Eu Ge (H (30)).
ꢂ
ꢃ
V
Magnetic data are not available for any of these europium
salts.
A conceivable consequence of a real depopulation of the
¹t n* states is that the lengths of the bonds should ideally
ꢁ
decrease. However, the complexities of the close structural
interrelationships already considered generate some caution
that it could be this simple. Notwithstanding, the three cases
for which we have ¹t}¹t bond distances for both empty and
stu!ed examples with a "xed cation show small consistent
decreases on oxidation (Table 7): Ca Si to Ca Si F , by
ꢂ ꢃ
ꢂ ꢃ ꢆꢇꢅ
0.012(2) A; Ca Ge to Ca Ge H , 0.019(1) A, and to
ꢂ
ꢃ
ꢂ
ꢃ V
Ca Ge F
, 0.007(1) A. Of course, we know nothing about
ꢂ
ꢃ ꢆꢇꢄꢄ
the actual degree of n* depopulation and whether, or how,
this changes with cation or Z; it is also possible that the
di!erentiating electrons in these cases come from the con-
duction band. The picture is not so simple when d(¹t2}¹t2)
in this structure type from other sources are included, even
with impurity uncertainties. However, all reported Si}Si and
Ge}Ge distances in this structure (except for Eu and those
with lattice constant problems) fall in small ranges only 0.04
and 0.02 A wide, respectively. In addition to this, we must
have what are generally recognized as the e!ects of charge
(and electron pair) repulsion in lengthening bonds in
dimeric or other oligomeric anions (41). We have noted
earlier that the Sn}Sn distances in two such phases (Table 7)
are less than those for the dimers in the semiconducting
Zintl phase Ca Sn
by 0.15 to 0.28 A, whereas the
ꢃꢀ ꢁꢆ
homoatomic bonds in the element or in anionic network
structures with lower charge densities and smaller repul-
sions are still shorter than in the present dimers (11).
Such a body of comparative information for the dimers is
harder to "nd for the present silicides and germanides. The
Si}Si and Ge}Ge bonds in the binaries reported here,
&2.45 and &2.57 A, are naturally longer than in the
4-bonded elements, 2.34 and 2.44 A (42). The similarly
FIG. 7. Molar magnetic susceptibility data as a function of temper-
ature. (top) Sr Ge and Sr Ge H ; (bottom) Sr Sn and Sr Sn H . Data
ꢂ
ꢃ
ꢂ
ꢃ
V
ꢂ
ꢃ
ꢂ
ꢃ
V
for the two hydrides are also shown as 1/s along with the best
Curie}Weiss-type "ts.
La Pb (O, N) in which the lead dimers no longer exist (13) bonded clathrate II network structures A Aꢁ ¹t
ꢀꢄ ꢀꢃꢄ
and in the isotypic KBa Sb O in which K substitutes for dently have some if not all of the electrons from a small
evi-
ꢂ
ꢃ
ꢈ
ꢅ
ꢃ
Ba1 (40). Extended Huckel band calculations for Ca Sn H, number of cations in a conduction band, and those distan-
ꢂ
ꢃ
Ca Ge Z, etc., show, not surprisingly, that the Fermi energy ces are slightly longer than in the elements, 2.36}2.39 A and
ꢂ
ꢃ
lies within the antibonding n* band (11, 34). However, the 2.49}2.51, respectively (43). Another concordant example is
band results did not exhibit an appropriately low gap be- 2.50 A between the only fully occupied Ge}Ge atom pair in
tween valence and conduction bands because only default Ba Ge (44). Di!erences from our results presumably arise
ꢄ
ꢁꢅ
H values were used that are more suitable for neutral or because there are smaller charge and negligible lone pair
GG
cationic triels. We surmise that the shorter distances ob- repulsion e!ects in this last group. Likewise, the well-known
ꢅ\
served for the Si}Si and Sn}Sn dimers when Eu is the cation tetrahedral anions ¹t
presumably exhibit the e!ects of
ꢅ
(Table 7) (but not for Eu Ge (H ) (30)) arise because this smaller charge repulsions, reduced electron pair repulsions,
cation-dominated conduction band lies lower than those for and perhaps angular strain, whereas the e!ects of cation
Ca, Sr, Ba, and more electrons from the anions' n* states alterations appear to be small. Distances in Si
ꢂ
ꢃ
V
ꢅ\
in the
ꢅ

162
LEON-ESCAMILLA AND CORBETT
alkali-metal salts, BaSi , and Na BaSi all fall in the range 13. A. M. Guloy and J. D. Corbett, Z. Anorg. Allg. Chem. 616, 61 (1992).
ꢅ
ꢁ
ꢅ
14. P. Bottcher, Th. Doert, Ch. Druska, and S. Bradtmoller, J. Alloys
Compd. 246, 209 (1997).
of 2.40}2.46 A, actually in the lower half of the range except
in RbSi and CsSi (45), and these di!er surprisingly little from
15. C. B. Magee, in &&Metal Hydrides,'' (W. M. Mueller, J. P. Backledge, and
G. G. Libowitz, Eds.), Chap. 6. Academic Press, New York, 1968.
16. P. Villars and L. D. Calvert, &&Pearson's Handbook of Crystallographic
Data for Intermetallic Phases,'' 2nd ed. American Society for Metals
International, Metals Park, OH, 1991.
17. A. M. Guloy and J. D. Corbett, Inorg. Chem. 32, 3532 (1993).
18. E. A. Leon-Escamilla, W.-M. Hurng, E. S. Peterson, and J. D. Corbett,
Inorg. Chem. 36, 703 (1997).
19. &&TEXSAN, Single Crystal Structure Analysis Software,'' Version 5.0.
Molecular Structure Corp., The Woodlands, TX, 1989.
20. J. T. Zhao and J. D. Corbett, Inorg. Chem. 34, 378 (1995).
21. B. Eisenmann and H. Schafer, Z. Naturforsch. 29B, 460 (1974).
22. P. Manfrinetti, M. L. Fornasini, and A. Palenzona, Intermetallics
8, 223 (2000).
23. G. Nargosen, G. Rocktaschel, H. Schafer, and A. Weiss, Z. Naturforsch.
22B, 101 (1967).
24. R. Nesper and F. Zuercher, Z. Kristallogr., New Cryst. Struct. 214,
19 (1998).
25. K. Janzon, H. Schafer, and A. Weiss, Z. Naturforsch. 21B, 287
(1966).
26. R. Nesper and F. Zuercher, Z. Kristallogr., New Cryst. Struct. 214,
20 (1999).
the distance in Ca Si , 2.45 A, in spite of all the changes. The
same is true for the analogous AGe and BaGe phases,
2.56}2.61 A with a similar dependence on the nature of
ꢂ ꢃ
ꢁ
A versus 2.58 A in Ca Ge (46).
ꢂ
ꢃ
We will report later on a brief study of some analogous
tetragonal A Tr phases for the electron-poorer triels (Tr)
Ga, In, and Tl (8). Signi"cant in the present context, the
reported thallide is actually stable only as Sr Tl H . In this
case, the n* band would ideally be empty for x"1, and the
Tl}Tl distance is indeed shorter than any reported before.
The much lower n* populations possible in the A Tr
ꢂ
ꢃ
ꢂ
ꢃ V
ꢂ
ꢃ
examples may be important in the unusual stability of these
triel phases if n* populations in the dimers are at all ger-
mane to their description. There is also another group of
tetrels for Sm ¹t and Yb ¹t that occur only with
ꢂ
ꢃ
ꢂ
ꢃ
Mn Si -type structures, usually for both the binaries and
ꢂ ꢃ
ternary hydrides. These compounds perhaps contain some
trivalent cations since all of the anions in this structure type
are isolated, and electron de"cient phases are generally not
found among tetrelides, but this situation has not been
veri"ed.
27. P. Pottgen, R.-D. Ho!mann, and D. Kussmann, Z. Anorg. Allg. Chem.
624, 945 (1998).
28. R. Nesper and F. Zuercher, Z. Kristallogr., New Cryst. Struct. 21, 410
(1999).
29. R. Nesper and F. Zuercher, Z. Kristallogr., New Cryst. Struct. 214,
22 (1999).
ACKNOWLEDGMENTS
30. R. Pottgen and A. Simon, Z. Anorg. Allg. Chem. 622, 779 (1996).
31. W. Dorrscheidt, A. Widera, and H. Schafer, Z. Naturforsch. 32B, 1097
(1977).
32. G. Bruzzone, E. Franceschi, and F. Merlo, J. ¸ess-Common Met. 60,
59 (1978).
33. J. Harp, M. S. thesis, Iowa State University, Ames, Iowa, 1994.
34. E. A. Leon-Escamilla, Ph.D. dissertation, Iowa State University,
1996.
35. M. Notin, J. Mejbar, A. Bouhajib, J. Charles, and J. Hertz, J. Alloys
Compd. 220, 62 (1995).
We thank J. Ostenson and D. K. Finnemore for the magnetic susceptibil-
ity measurements.
REFERENCES
1. S. Kauzlarich (Ed.), &&Chemistry, Structure, and Bonding of Zintl
Phases and Ions.'' VCH, Boca Raton, FL, 1996.
2. T. Hughbanks, in &&Inorganometallic Chemistry'' (T. Fehlner, Ed.),
p. 291. Plenum Press, New York, 1992.
36. A. M. Guloy and J. D. Corbett, Inorg. Chem., to be submitted.
3. E. A. Leon-Escamilla and J. D. Corbett, J. Alloys Compd. 206, L15 37. H. S. Marek and J. D. Corbett, Inorg. Chem. 22, 3194 (1983).
(1994).
38. R. Nesper and F. Zuercher, Z. Kristallogr., New Cryst. Struct. 214,
21, 432, 413 (1999).
39. R. D. Shannon, Acta Crystallogr. Sect. A: Found Crystallogr. 32, 751
(1976).
40. B. Eisenmann, C. Gieck, and U. Rossler, Z. Anorg. Allg. Chem. 625,
1331 (1999).
41. R. Nesper, Prog. Solid State Chem. 20, 1 (1990).
42. L. Pauling, &&The Nature of the Chemical Bond,'' 3rd ed., p. 403. Cornell
University Press, Ithaca, NY, 1960.
43. S. Bobev and S. C. Sevov, J. Solid State Chem. 153, 92 (2000).
44. S.-J. Kim, S. Hu, C. Uher, T. Hogan, B. Huang, J. D. Corbett, and
M. G. Kanatzidis, J. Solid State Chem. 153, 321 (2000).
45. B. Huang and J. D. Corbett, Solid State Sci. 1, 555 (1999).
4. E. A. Leon-Escamilla and J. D. Corbett, J. Alloys Compd. 265, 104
(1998).
5. W.-M. Hurng and J. D. Corbett, Chem. Mater. 1, 311 (1989).
6. C. E. Messer, J. Solid State Chem. 2, 144 (1970).
7. J. D. Corbett, E. Garcia, A. M. Guloy, W.-M. Hurng, Y.-U. Kwon, and
E. A. Leon-Escamilla, Chem. Mater. 10, 2824 (1998).
8. E. A. Leon-Escamilla and J. D. Corbett, to be submitted.
9. R. W. Henning, E. A. Leon-Escamilla, J.-T. Zhao, and J. D. Corbett,
Inorg. Chem. 36, 1282 (1997).
10. B. Huang and J. D. Corbett, Inorg. Chem. 37, 1892 (1998).
11. E. A. Leon-Escamilla and J. D. Corbett, Inorg. Chem. 40, 1226
(2001).
12. A. K. Ganguli, A. M. Guloy, E. A. Leon-Escamilla, and J. D. Corbett, 46. J. T. Vaughey, G. J. Miller, S. Gravelle, E. A. Leon-Escamilla, and
Inorg. Chem. 32, 4349 (1993). J. D. Corbett, J. Solid State Chem. 133, 501 (1997).