Synthesis, Electrical Conductivity, and Crystal Structure of Cu4Sn7S16and Structure Refinement of Cu2SnS3

Article abstract of DOI:10.1006/jssc.1998.7822

Single crystals of Cu₄Sn₇S₁₆and Cu₂SnS₃were prepared by the conventional solid state reaction method. Their crystal structures were determined from single-crystal X-ray data. Cu₄Sn₇S₁₆is of a new structure type with rhombohedral symmetry:R-3m,a=7.372(1),c=36.010(7) A,Z=3,R=0.018 for 522 structure factors and 36 variable parameters. Its crystal structure may be described as a defect variant of the spinel-type. In the unit cell, 3/4 of the tetrahedral sites and 1/8 of the octahedral sites are occupied by Cu⁺, while 7/8 of the octahedral sites are occupied by Sn⁴⁺ions. The compound can therefore be formulated as (Cu_0.75□_0.25)(Sn_1.75Cu _0.25)S₄. Cu₂SnS₃has the unit cell formula Cu_2.665(7)Sn_1.335(7)S₄(Z=2) and it is isotypic with the tetragonal stannite [Cu₂(Fe, Zn)SnS₄] structure:I-42m,a=5.413(1),c=10.824(1) A,R=0.030 for 281 reflections and 16 variables. In the case of Cu₂SnS₃, the positions of the Cu and Sn atoms of stannite are occupied by the composite atoms M1 [43.6(2) at.% Sn+56.4(2) at.% Cu] and M2 [46.3(3) at.% Sn+53.7(3) at.% Cu], respectively, while those of the (Fe, Zn) atoms are replaced by only Cu atoms. Electrical conductivity measurements confirm the expected semiconducting behavior of Cu₄Sn₇S₁₆, and XPS analysis reveals the presence of only Cu⁺in Cu₄Sn₇S₁₆.

Full text of DOI:10.1006/jssc.1998.7822

JOURNAL OF SOLID STATE CHEMISTRY 139, 144—151 (1998)
ARTICLE NO. SC987822
Synthesis, Electrical Conductivity, and Crystal Structure of Cu Sn S
4
7 16
and Structure Refinement of Cu SnS
2
3
Xue-an Chen, Hiroaki Wada,ꢀ Akira Sato, and Masahiro Mieno
National Institute for Research in Inorganic Materials, 1-1 Namiki, Tsukuba-shi, Ibaraki 305, Japan
Received November 20, 1997; in revised form March 9, 1998; accepted March 17, 1998
the compounds Cu SnS , Cu SnS , ‘‘Cu Sn S ,’’ and
ꢃ ꢈ
S correspond to the Cu S mole proportion of
ꢃꢄꢈꢂ ꢅ
.6667, 0.50, 0.25, and 0.1176, respectively. Structural char-
ꢇ
ꢇ
ꢆ
ꢃ
ꢆ
Single crystals of Cu Sn S and Cu SnS were prepared by the CuSn
4
7
16
2
3
ꢆ
conventional solid state reaction method. Their crystal structures
were determined from single-crystal X-ray data. Cu Sn S is of
0
4
7
16
acterizations of Cu SnS and CuSn
S have been done
ꢇ
ꢇ
ꢃꢄꢈꢂ ꢅ
a new structure type with rhombohedral symmetry: R-3m,
a ؍ 7.372(1), c ؍ 36.010(7) A, Z ؍ 3, R ؍ 0.018 for 522 structure
by Jaulmes et al. (7, 8). Cu SnS was found to be dimorphic
ꢆ
ꢃ
s
with a phase transition at 780°C, the high-temperature
phase was reported to be isotypic with the cubic ZnS struc-
ture, and the low-temperature phase was considered as
a triclinic superstructure of sphalerite type (4), while Hahn
et al. (9) thought that the powder data of the low-temper-
factors and 36 variable parameters. Its crystal structure may be
described as a defect variant of the spinel-type. In the unit cell,
3
/4 of the tetrahedral sites and 1/8 of the octahedral sites are
؉
occupied by Cu , while 7/8 of the octahedral sites are occupied by
4
؉
Sn ions. The compound can therefore be formulated as
(
Cu0.75ᮀ0.25)(Sn1.75Cu0.25)S . Cu SnS has the unit cell formula ature phase could be indexed on a tetragonal cell, and
4 2 3
Cu2.665(7)Sn1.335(7)S4 (Z ؍ 2) and it is isotypic with the tetragonal Villars et al. (10) furthermore assigned the space group I-42d
stannite [Cu (Fe, Zn)SnS ] structure: I-42m, a ؍ 5.413(1), c ؍ to its structure. The present work confirmed the tetragonal
2
4
1
0.824(1) A
s
, R ؍ 0.030 for 281 reflections and 16 variables. In
system given by Hahn et al., but with a different space group
(I-42m). The phase ‘‘Cu Sn S ’’ was first reported by Wang
the case of Cu SnS , the positions of the Cu and Sn atoms of
2
3
ꢆ
ꢃ ꢈ
stannite are occupied by the composite atoms M1 [43.6(2) at.%
Sn ؉ 56.4(2) at.% Cu] and M2 [46.3(3) at.% Sn ؉ 53.7(3) at.%
Cu], respectively, while those of the (Fe, Zn) atoms are replaced
by only Cu atoms. Electrical conductivity measurements confirm
the expected semiconducting behavior of Cu Sn S , and XPS
analysis reveals the presence of only Cu in Cu Sn S . ( 1998
(4), who predicted that it crystallizes in a monoclinic system
with the space group Cc or C2/c. In his following work (11),
the formula of this compound was redetermined to be
Cu Sn S . Our single-crystal structure determination con-
4
7
16
ꢆ
ꢃꢄꢂ ꢅ
؉
firmed the revised formula (Cu Sn S ) and found the
4
7
16
ꢆ
ꢃꢄꢂ ꢅ
space group actually to be rhombohedral R-3m rather than
monoclinic Cc or C2/c.
Academic Press
SAMPLE PREPARATION AND CHARACTERIZATION
INTRODUCTION
Starting materials were powders of elements copper, tin,
and sulfur, all with nominal purities greater than 99.9%.
The compound Cu Sn S was synthesized by a stoichio-
The Ag- and Cu-containing chalcogenides are of con-
siderable interest because they exhibit novel structure char-
acteristics and unusual physical properties, e.g., incomplete
occupancy of certain Ag or Cu sites causes the order-dis-
order phenomena and phase transitions (1), and some of
these compounds such as Cu PS Hal (Hal"Cl, Br, I) (1)
ꢇ
ꢈ ꢀꢁ
metric mixture of Cu S (0.318g, 2.0 mmol, prepared by
ꢆ
a stoichiometric reaction of Cu with S at 500°C) and SnS
ꢆ
(1.280g, 7.0 mmol, prepared by a stoichiometric reaction of
ꢁ
ꢂ
Sn and S at 450°C). The sample was ground in an agate
mortar and pressed into pellets. The pellets were introduced
into a silica tube and sealed at pressure of less than 10\ꢃ
Torr. The tubes were held at 450°C for three days, then
heated gradually to 850°C, where they were kept for three
days, then cooled at a rate of 5°C/h to 600°C and quenched
in air. The gray crystals with metallic luster were found in
the tubes, and they appear as thin plates of trigonal outlines
with a dimension up to 0.5 mm across. As mentioned by
and Ag Sn S (2) have high ionic conductivities. The ter-
ꢃꢄꢅ ꢃ ꢅ
nary system Cu—Sn—S has been investigated in the past by
Khanafer et al. (3), Wang (4), Moh (5, 6), and Jaulmes et al.
(7, 8). Many phases were proposed, and most of them were
situated on the Cu S—SnS line of this system. For example,
ꢆ
ꢆ
ꢀ
To whom correspondence should be addressed. Fax: 81-298-52-7449.
E-mail: wadah@nirim.go.jp.
1
44
0
022-4596/98 $25.00
Copyright ( 1998 by Academic Press
All rights of reproduction in any form reserved.

Cu Sn S AND Cu SnS
ꢈ ꢀꢁ ꢃ
145
ꢇ
ꢆ
Wang (4), the parallel growth on the surface plane of The diffraction patterns of Cu Sn S showed a rhom-
ꢈ ꢀꢁ
ꢇ
the plates is frequently observed. Semiquantitative EDAX bohedral cell. The only systematic extinctions (reflections
analysis with the microprobe of Hitachi S-5000 scanning hkl were observed only with !h#k#l"3n) were com-
electron microscope confirmed the presence of all three patible with the space groups R3, R-3, R-32, R3m, and R-3m,
elements in a ratio of approximately 4.2 : 6.8 : 16. The of which the centrosymmetric group R-3m was found to be
compound appears to be relatively stable in air and water. correct during the structure refinement. The precession pho-
In an alternative tube with a slightly SnS -rich composition, tographs of Cu SnS show Laue symmetry 4/mmm and the
ꢆ
ꢆ
ꢃ
both Cu Sn S and Sn S crystals were observed. The reflection condition hkl : h#k#l"2n, leading to the pos-
ꢈ ꢀꢁ ꢆ ꢃ
Sn S single crystals are easily recognizable as very thin sible space groups I4/mmm, I-42m, I-4m2, I4mm, and I422.
ꢆ ꢃ
needles.
ꢇ
With analogy to other chalcopyrite-like compounds, the
Cu SnS was prepared by a direct reaction of the space group I-42m of acentric symmetry could be adopted
ꢆ
ꢃ
stoichiometric elements Cu (0.381g, 6.0 mmol), Sn (0.356g, as the most probable one.
.0 mmol), and S (0.289g, 9.0 mmol) at 1050°C for two days. Intensity data for both compounds were collected on an
3
The sample was slowly cooled at a rate of 5°C/h to 700°C Enraf-Nonius CAD4 automatic four-circle diffractometer
and annealed there for two days, then quenched in air. The with graphite monochromated MoKa radiation. Cell con-
crystals had the form of prisms with dimensions of up to stants were obtained from a least-squares refinement with
0
8
2
.4;0.4;1.0 mmꢃ. Microprobe analysis by JEOL JXA- 20 automatically centered reflections in the ranges 30°(2h
600 MX gives an approximate atomic ratio Cu : Sn :S of (50° for Cu Sn S and 30°(2h(55° for Cu SnS .
ꢇ
ꢈ ꢀꢁ
ꢆ
ꢃ
: 1 : 3, in good agreement with that deduced from the struc- Three standard reflections were remeasured after every 200
tural refinement.
reflections with no indication of crystal decay. The intensity
X-ray powder diffraction intensity data were collected data were corrected for Lorentz and polarization effects and
using the graphite-monochromated CuKa radiation of a for absorption by a Gaussian numerical integration using
Rigaku diffractometer (Geigerflex, RAD-2B system). Both the measured dimensions of the crystal (12). The crystallo-
compounds were found to be single phases. X-ray diffrac- graphic data and some results are summarized in Table 1.
tion patterns of Cu Sn S are in good agreement with
ꢈ ꢀꢁ
those reported by Wang (4) for the monoclinic phase SIR-92 program (13) and refined in SHELXL-93 system (14)
Both structures were solved by direct methods in the
ꢇ
‘
‘Cu Sn S ’’. It was also possible to index reflection peaks by full-matrix least-squares methods on Fꢆ. The Cu Sn S
ꢃ ꢈ ꢈ ꢀꢁ
by using a pseudohexagonal cell with dimensions a " crystal structure was first successfully solved in a monoclinic
ꢆ
ꢀ
ꢇ
)
7
.371 A
was finally determined by X-ray single crystal data to be [a"12.772(2) A
trigonal. X-ray diffraction patterns of Cu SnS confirmed 109.53(2)°]. However, the space group was C2/m, R"0.022
s
and c "36.010 A
s
. The symmetry of this compound system based on the lattice constants reported by Wang (4)
)
s
, b"7.371(1) A, c"12.736(2) A, and b"
s
s
ꢆ
ꢃ
the earlier work (9).
for 79 variables and 1226 structure factors. After checking
X-ray photoelectron spectroscopy (XPS) measurements the atomic parameters carefully, we found a missed three-
were performed using a JEOL JPS-9000 MC spectrometer fold axes symmetry. The intensity data was subsequently
with a MgKa X-ray source (hl"1253.6 eV) in a vacuum of transformed to a rhombohedral setting by a matrix
1
;10\ꢈ Pa. Powder samples were pressed onto indium foil (0!1 0!1/2 1/2 0!1 0!3) and the structure was finally
and then mounted in an aluminum sample holder. The refined in a trigonal R-3m space group.
samples were sputtered with an Ar ion beam before the For Cu Sn S , in order to check for deviations from the
ꢇ
ꢈ ꢀꢁ
spectra were recorded, and the electron binding energies ideal compositions all occupancy parameters were allowed
were referred to the C 1s line of impurity carbon, which had to vary along with the positional and thermal parameters.
a value of 284.5 eV.
The resulting occupancies varied between the values 1.04(3)
Conductivity measurements of Cu Sn S were carried for Sn2 and 1.07(3) for S2. The only exception was the Cu1
ꢇ
ꢈ ꢀꢁ
out on rectangular, cold-pressed polycrystalline samples position, which showed significant deviation from the full
(
2.6;5.0;5.0 mmꢃ). The experimental cells were built and occupancy with the value 0.53(2). Therefore, in the final
placed in a special glass container under high purity argon least-squares cycles all positions were assumed to be fully
atmosphere. The alternate current method was employed occupied with exception of this Cu position, for which the
with a Solartron 1286 electrical interface and a Solartron occupancy was fixed at 0.50. Thus, the compound has the
1
255 frequency response analyzer.
composition Cu Sn S , corresponding to the Cu S mole
ꢈ ꢀꢁ
proportion of 0.2222 in the pseudobinary Cu S—SnS sys-
tem. The refinement resulted in the conventional and
ꢇ
ꢆ
ꢆ
ꢆ
STRUCTURE DETERMINATIONS
weighted residuals R"0.018 and wR(Fꢆ)"0.040 for 36
Single crystals were examined in a Buerger precession variables and 522 structure factors. Final difference Fourier
camera to establish their symmetry. Overexposed photo- map showed the highest and lowest electron densities of 0.88
graphs revealed no supercell reflections for both compounds. and !1.86 e A
\ꢃ, respectively, which were too close to the
s

1
46
CHEN ET AL
TABLE 1
Some Data and Results of the Structure Determinations of
Cu Sn SnS
a formula. In separate least-squares cycles, Cu and Sn were
put together in this site with the constraint that their posi-
tional and thermal parameters are equal and the total occu-
pation for this site is 1.0. The refinement of population
parameters leads to a Cu/Sn ratio of 0.997(13)/0.003(13).
4
7
S
16 and Cu
2
3
Chemical formula
Formula weight
Crystal dimension (mm)
Space group
Cu Sn S
ꢈ ꢀꢁ
1597.95
0.12;0.08;0.02
R-3m (No. 166)
Cu SnS
341.95
ꢇ
ꢆ
ꢃ
0.07;0.10;0.08 Therefore, in the final refinement, only the Cu atom is
I-42m (No. 121)
assigned to this site, which gives a reasonable formula.
Lattice constants
For Cu SnS , the space group I-42m does not permit the
a (A
c (A
Volume (A
s
)
)
7.372(1)
36.010(7)
1694.8(5)
3
5.413(1)
ꢆ
ꢃ
s
10.824(1)
317.15(9)
16/6
assignment of copper and tin atoms to different sites within
the limitation of eight metal atoms/unit cell. As no evidence
of a supercell was found, it must be presumed that the
copper and tin atoms are disordered over the metal sites.
Refinement of the atomic occupancies indicates that one
metal site (2a) is fully occupied by the copper atom, and the
other two sites (2b and 4d) are clearly shared by tin and
copper atoms. Therefore, in the final least-squares cycles the
Sn and Cu atoms were allowed to disorder over these two
positions with the constraints that the Sn and Cu occu-
pancies at a given site sum to unity, that they have the same
thermal parameters at a given site, and that the charge
balance is fitted. The final R value is 0.030 for 281 reflections
and 16 variables, and the final difference Fourier map is
nearly flat with the highest and lowest electron densities 0.80
s
ꢃ)
Z
Calculated density (g cm\ꢃ)
Temperature of data collection (K)
Wavelength (jMoKa) (A
Linear absorption coefficient k (cm\ꢀ)
Transmission factors
Scan type
4.697
293
0.71073
127.45
0.463—0.822
u—2h
4.774
293
0.71073
151.90
0.419—0.496
u—2h
s
)
Scan speed
Scan range
h (max)
Variable
(0.8#0.35 tan h)
60°
Variable
(0.6#0.35 tan h)
70°
2
Range in hkl
Total number of reflections
Unique reflections
!10—0, !8—10,$50 0—8, 0—8,$17
3515
672
1672
399
Inner residual
Reflections with Fꢆ'2p(Fꢆ)
Number of variables
R "0.049
R "0.051
*
*
522
281
ꢀ
ꢀ
36
16
Secondary extinction coefficient
S (Goodness-of-fit on Fꢆ)
Weighting scheme
0.00039(2)
1.110
w"1/[pꢆ(Fꢆ)#
0.030(2)
0.974
w"1/[pꢆ(Fꢆ)#
(0.030P)ꢆ]
and !0.95 e A
s
\ꢃ, respectively. The compound has the
(
0.013P)ꢆ#8.485P]
composition Cu
Sn S (Z"2), corresponding
here P"(Fꢆ#2Fꢆ)/3
ꢆꢄꢁꢁꢂ(ꢈ) ꢀꢄꢃꢃꢂ(ꢈ) ꢇ
to Cu SnS (Z"16/6). Refinement of Flack parameter
ꢀ
#
R[Fꢆ'2p(Fꢆ)]
wR(Fꢆ)
*/p)ꢂ!ꢃ
0.018
0.040
0.001
0.030
0.069
0.000
ꢆ
ꢃ
gives a value of !0.09(6), indicating that the absolute struc-
ture is correct. As a further test of this model, an alternative
with complete disorder of the metal atoms over three kinds
of positions (2a, 2b, and 4d) was tested, no final results were
available because the refinement was unstable. Thus the
(
Cu2 position (0.55 and 0.81 A
s
, respectively) to be suitable model with Sn/Cu disorder confined to positions 2b and 4d
for the accommodation of any additional atoms. It should is in good agreement with the diffraction data, and gives
be pointed out that this refinement results in a six-coor- results that appear physically reasonable. The final posi-
dinated Cu3 site, which is uncommon. Refinement of this tional, the equivalent isotropic and anisotropic displace-
Cu position as a Sn atom leads to a formula of ment parameters of both compounds are given in Tables 2
Cu Sn
S
with the Sn occupancy of about 61%. It is and 3, respectively. Listings of the structure factors are
clear that charge neutrality could not be obtained for such available from the authors.
ꢉ
ꢆꢆꢄꢅꢃ ꢇꢅ
TABLE 2
Atomic Coordinates, Equivalent Isotropic and Anisotropic Displacement Parameters (A
s
ꢆ) for Cu
4
Sn S
7 16
Atom
Site type
x
y
z
º%ꢁ
ºꢀꢀ
ºꢆꢆ
ºꢃꢃ
ºꢆꢃ
ºꢀꢃ
ºꢀꢆ
Sn1
Sn2
Cu1
Cu2
Cu3
S1
S2
S3
S4
18h
3b
6c
6c
3a
18h
18h
6c
0.5017(1)
0
0
0
0.4983(1)
0
0
0
0.2513(1)
1/2
0.1884(1)
0.2945(1)
0
0.1244(1)
0.3760(1)
0.1239(1)
0.3756(1)
0.011(1)
0.012(1)
0.016(1)
0.039(1)
0.054(1)
0.010(1)
0.011(1)
0.010(1)
0.010(1)
0.011(1)
0.013(1)
0.017(1)
0.013(1)
0.074(1)
0.010(1)
0.010(1)
0.011(1)
0.010(1)
0.011(1)
0.013(1)
0.017(1)
0.013(1)
0.074(1)
0.010(1)
0.010(1)
0.011(1)
0.010(1)
0.013(1)
0.008(1)
0.014(1)
0.090(1)
0.014(1)
0.009(1)
0.012(1)
0.009(1)
0.010(1)
0.001(1)
0
0
0
!0.001(1)
0.006(1)
0.007(1)
0.009(1)
0.006(1)
0.037(1)
0.006(1)
0.005(1)
0.006(1)
0.005(1)
0
0
0
0
0
0
0
0.4937(1)
0.5074(1)
0
0
0.5063(1)
0.4926(1)
0
0
0.000(1)
0.001(1)
0
0
0.000(1)
!0.001(1)
0
0
6c
Notes. The occupation factor of Cu1 is 0.50. º is defined as one-third of the trace of the orthogonalized º tensor. The anisotropic displacement factor
%
ꢁ
GH
exponent takes the form !2nꢆ(hꢆa*ꢆºꢀꢀ#2#2hka*b*º ).
ꢀꢆ

Cu Sn S AND Cu SnS
147
ꢇ
ꢈ ꢀꢁ
ꢆ
ꢃ
TABLE 3
Atomic Coordinates, Equivalent Isotropic and Anisotropic Displacement Parameters (A
s
ꢆ) for Cu
2
SnS
3
Atom
Site type
x
y
z
º%ꢁ
ºꢀꢀ
ºꢆꢆ
ºꢃꢃ
ºꢆꢃ
ºꢀꢃ
ºꢀꢆ
M1
M2
Cu
S
4d
2b
2a
8i
0
0
0
1/2
0
0
1/4
1/2
0
0.019(1)
0.019(1)
0.022(1)
0.015(1)
0.019(1)
0.018(1)
0.020(1)
0.015(1)
0.019(1)
0.018(1)
0.020(1)
0.015(1)
0.018(1)
0.021(1)
0.024(1)
0.015(1)
0
0
0
0
0
0
0
0
0
0.2461(1)
0.2461(1)
0.1231(1)
!0.001(1) !0.001(1) !0.001(1)
Notes. M1"43.6(2) at.% Sn#56.4(2) at.% Cu, M2" 46.3(3) at.% Sn#53.7(3) at.% Cu. º% is defined as one-third of the trace of the orthogonalized
ꢁ
º
t
e
n
s
o
r
.
T
h
e
a
n
i
s
o
t
r
o
p
i
c
d
i
s
p
l
a
c
e
m
e
n
t
f
a
c
t
o
r
e
x
p
o
n
e
n
t
t
a
k
e
s
t
h
e
f
o
r
m
!
2
n
ꢆ
(
h
ꢆ
a
*
ꢆ
º
ꢀ
ꢀ
#2#2hka*b*º
).
G
H
ꢀꢆ
DISCUSSION
in a narrow range 2.5095(7)—2.5956(7) A
s
and the 90° octa-
hedral angles for SnS units are in the range 86.19(5)—
ꢁ
Cu Sn S crystallizes in a new structure type of trigonal
system (Pearson symbol hR81). Its crystal structure is char-
acterized by a three-dimensional framework, which is close-
9
2.79(3)°, indicating that the octahedral environments are
slightly distorted. The average Sn—S bond distances are
.564(1) A for Sn1 and 2.552(1) A for Sn2, respectively. They
ꢇ
ꢈ ꢀꢁ
2
s
s
ly related to that of spinel (MgAl O type) (15). The strong
are in excellent agreement with those found in other Snꢇ>-
ꢆ ꢇ
X-ray powder diffraction peaks of this compound could be
containing sulfides such as: 2.570(4) A in La SnS (17),
s
ꢆ
ꢂ
indexed as a spinel structure with a pseudocubic unit cell of
2
.566(2) A
The tetrahedron around Cu1 is almost regular, with the
"2; S—Cu—S angles being near the tetrahedral value and the
3;a . In fact, a is one-half of the [110] vector of the cubic Cu—S distances exhibiting a narrow range [2.322(2)—
s
in Ag Sn S (2), and 2.56(1) A
s
in CuSn
S (8).
ꢃꢄꢅ ꢃ ꢅ
ꢃꢄꢈꢂ ꢅ
about a +10.4 A
s
. The approximate dimensional relations
#
are: a"b"7.372 A"(2;a /2, and c"36.010 A
s
s
#
(
#
s s
cell and c corresponds to the double of its [111] vector. The 2.330(1) A]. The average Cu—S distance of 2.328(2) A com-
space group R-3m of Cu Sn S is actually a translationen- pares well with those generally found in ternary or quater-
ꢇ
ꢈ ꢀꢁ
gleiche subgroup of the group Fd-3m adopted by spinel. nary copper chalcogenides with similar coordination
A comparison of the atomic parameters of Cu Sn S
around Cu>: 2.343(2) and 2.333(1) A
s
in Cu TiS (18),
ꢇ
ꢈ ꢀꢁ
ꢇ
ꢇ
(
Table 2) with those of MgAl O (15) indicates that the 2.320(1) A in Cu FeSn S (19), and 2.307(1) A in KCuZrS
s
s
ꢆ ꢇ
ꢆ
ꢃ ꢅ
ꢃ
tetrahedral sites of the spinel structure are split into two (20).
nonequivalent groups, 6c and 6c, with 3m symmetry, and
they are occupied by Cu2 and statistically by Cu1, respec-
tively. The octahedral sites are split into three nonequi-
valent sets, 18h, 3b, and 3a, with symmetry m, !3m, and
!
3m, and accommodated by Sn1, Sn2, and Cu3, respec-
tively. Similarly, the sulfur atoms are divided into four
different groups, 18h, 18h, 6c, and 6c, with symmetry m or
3
m. The 12-layer sulfur atoms form a slightly distorted cubic
close-packed arrangement along the c-axis in sequence
..ABC.... The Cu and Sn atoms are distributed in an ordered
.
manner over the tetrahedral and octahedral voids between
the S-layers and some of them are displaced from the centers
of their coordination polyhedra, which results in a lowering
of the space-group symmetry. Figure 1 shows a portion of
the structure, according to the method of representation for
spinel used by Baltzer et al. (16), and Fig. 2 gives a perspect-
ive view of the unit cell. As can be seen, the Sn- and
Cu3-centered octahedra are interconnected by edge sharing
and they are connected to Cu1- or Cu2-centered tetrahedra
by corner sharing. The Cu1- and Cu2-centered tetrahedra
are separated from each other.
FIG. 1. Portion of the structure of Cu Sn S with the atomic num-
ꢈ ꢀꢁ
bering scheme. The dashed lines represent weak Cu2...S4 interactions of
Table 4 lists the selected bond distances and angles con-
cerning the coordination geometries of the cation-centered
ꢇ
2
.922(3) A and the symmetry codes here are the same as those in Table 4.
s
sulfur polyhedra in Cu Sn S . The Sn—S bond lengths vary Displacement ellipsoids are shown at the 50% probability level.
ꢇ
ꢈ ꢀꢁ

1
48
CHEN ET AL
TABLE 4
Selected Bond Lengths (A) and Angles (°) for Cu
s
4
Sn S
7 16
Sn1—S2*
2.5095(7)
2.5095(7)
2.582(1)
2.590(1)
2.5956(7)
2.5956(7)
2.552(1)
2.552(1)
2.552(1)
2.552(1)
2.552(1)
2.552(1)
Cu1—S3
2.322(2)
2.330(1)
2.330(1)
2.330(1)
2.227(1)
2.227(1)
2.227(1)
2.549(1)
2.549(1)
2.549(1)
2.549(1)
2.549(1)
2.549(1)
Sn1—S2**
Sn1—S4***
Sn1—S3*ꢄ
Sn1—S1ꢄ
Sn1—S1ꢄ*
Sn2—S1***
Sn2—S1ꢄ**
Sn2—S1**
Sn2—S1ꢄ***
Sn2—S1*ꢃ
Sn2—S1ꢃ
Cu1—S1*ꢄ
Cu1—S1ꢄ
Cu1—S1ꢃ*
Cu2—S2***
Cu2—S2*ꢃ
Cu2—S2**
Cu3—S2*ꢄ
Cu3—S2ꢃ**
Cu3—S2ꢃ*
Cu3—S2ꢃ***
Cu3—S2ꢄ
Cu3—S2ꢃ*ꢄ
S2*—Sn1—S2**
S2*—Sn1—S4***
S2**—Sn1—S4***
S2*—Sn1—S3*ꢄ
S2**—Sn1—S3*ꢄ
S4***—Sn1—S3*ꢄ
S2*—Sn1—S1ꢄ
S2**—Sn1—S1ꢄ
S4***—Sn1—S1ꢄ
S3*ꢄ—Sn1—S1ꢄ
S2*—Sn1—S1ꢄ*
S2**—Sn1—S1ꢄ*
S4***—Sn1—S1ꢄ*
S3*ꢄ—Sn1—S1ꢄ*
S1ꢄ—Sn1—S1ꢄ*
S1***—Sn2—S1ꢄ**
89.14(5)
92.79(3)
92.79(3)
88.58(3)
88.58(3)
178.09(5)
178.31(3)
92.33(3)
87.97(3)
90.64(3)
92.33(3)
178.31(3)
87.97(3)
90.64(3)
86.19(5)
180.0
88.05(3)
91.95(3)
91.95(3)
88.05(3)
180.0
88.05(3)
91.95(3)
88.05(3)
91.95(3)
91.95(3)
88.05(3)
S1**—Sn2—S1ꢃ
S1ꢄ***—Sn2—S1ꢃ
S1*ꢃ—Sn2—S1ꢃ
S3—Cu1—S1*ꢄ
S3—Cu1—S1ꢄ
S1*ꢄ—Cu1—S1ꢄ
S3—Cu1—S1ꢃ*
91.95(3)
88.05(3)
180.0
108.57(5)
108.57(5)
110.36(5)
108.57(5)
110.36(5)
110.36(5)
119.62(1)
119.62(1)
119.62(1)
180.0
87.40(3)
92.60(3)
92.60(3)
87.40(3)
180.0
87.40(3)
92.60(3)
87.40(3)
92.60(3)
92.60(3)
87.40(3)
92.60(3)
87.40(3)
180.0
S1*ꢄ—Cu1—S1ꢃ*
S1ꢄ—Cu1—S1ꢃ*
S2***—Cu2—S2*ꢃ
S2***—Cu2—S2**
S2*ꢃ—Cu2—S2**
S2*ꢄ—Cu3—S2ꢃ**
S2*ꢄ—Cu3—S2ꢃ*
S2ꢃ**—Cu3—S2ꢃ*
S2*ꢄ—Cu3—S2ꢃ***
S2ꢃ**—Cu3—S2ꢃ***
S2ꢃ*—Cu3—S2ꢃ***
S2*ꢄ—Cu3—S2ꢄ
S2ꢃ**—Cu3—S2ꢄ
S2ꢃ*—Cu3—S2ꢄ
S2ꢃ***—Cu3—S2ꢄ
S2*ꢄ—Cu3—S2ꢃ*ꢄ
S2ꢃ**—Cu3—S2ꢃ*ꢄ
S2ꢃ*—Cu3—S2ꢃ*ꢄ
S2ꢃ***—Cu3—S2ꢃ*ꢄ
S2ꢄ—Cu3—S2ꢃ*ꢄ
FIG. 2. View of the Cu Sn S structure along [110] with the unit cell S1***—Sn2—S1**
ꢈ ꢀꢁ
outlined. The open circles are Cu atoms, the cross-hatched circles are Sn S1ꢄ**—Sn2—S1**
ꢇ
atoms, and the partly shaded circles are S stoms.
S1***—Sn2—S1ꢄ***
S1ꢄ**—Sn2—S1ꢄ***
S1**—Sn2—S1ꢄ***
S1***—Sn2—S1*ꢃ
S1ꢄ**—Sn2—S1*ꢃ
Two different coordination geometries may be used to
characterize the environment of an Cu2 atom: (i) The Cu2 S1**—Sn2—S1*ꢃ
site may be described as a triangular site in which the copper S1ꢄ***—Sn2—S1*ꢃ
S1***—Sn2—S1ꢃ
S1ꢄ**—Sn2—S1ꢃ
atom is displaced slightly out of the plane through three S2
atoms by 0.139(2) A
and S2—Cu2—S2 bond angles 119.62(1)°. These values are
closely comparable to those for CuS coordination found in x!y#1/3, x!1/3, !z#2/3; (iii) !x#1/3, !y#2/3, !z#2/3;
s s
. The Cu2—S2 distances are 2.227(1) A
Note. Symmetry codes: (i) y#1/3, !x#y#2/3, !z#2/3; (ii)
ꢃ
Cu U S (21) (2.219(2) A
ꢆ ꢃ ꢈ
distances are slightly shorter than those usually found for
s
and 119.84°). However, the Cu2—S (iv) !x#2/3, !y#1/3, !z#1/3; (v) y!1/3, !x#y#1/3,
!
z#1/3; (vi) x!y#2/3, x#1/3, !z#1/3; (vii) x!1/3, y!2/3,
z#1/3; (viii) !x#y!1/3, !x#1/3, z#1/3; (ix) y!2/3,
!x#y!1/3, !z#2/3; (x) !y#2/3, x!y#1/3, z#1/3; (xi)
x!y!1/3, x!2/3, !z#1/3; (xii) x!2/3, y!1/3, z!1/3; (xiii)
such a coordination: 2.33(1) A in TlCu S (22), 2.321(9) A in
s
s
ꢃ ꢆ
Cu SnS (7). (ii) The Cu2 site may also be described as
ꢇ
ꢇ
a distorted tetrahedron site made up of three S2 and one S4 !x#y#1/3, !x#2/3, z!1/3; (xiv) !y#1/3, x!y!1/3, z!1/3;
atoms, but in this case one Cu—S bond length [Cu2—S4 of (xv) x!1, y!1, z; (xvi) !x#1/3, !y!1/3, !z#2/3; (xvii) !y,
x!y!1, z; (xviii) !x#y, !x, z; (xix) x!y!2/3, x!4/3, !z#2/3;
2
s
.922(3) A] is very long compared to the others. The exist-
(
xx) x#2/3, y#1/3, z#1/3; (xxi) x!1/3, y#1/3, z#1/3; (xxii)
ence of such an elongated Cu—S bond may be responsible
for the shortening of the Cu—S distances on the opposite
side.
!
x#4/3, !y#2/3, !z#2/3.
Cu3 has an unusual octahedral environment with Cu—S
distances of 2.549(1) A
s
. Indeed, with a 3dꢀꢊ4sꢊ electronic measurements were done on the powder samples of this
configuration, Cu> favors a linear or tetrahedral surround- compound. Figure 3 illustrates the Cu 2p core level spec-
ing, the latter being common in sulfides. In order to examine trum. The observed values of the binding energies for Cu
the oxidation state of copper ions in Cu Sn S , XPS 2p and Cu 2p are in good agreement with those in the
ꢇ
ꢈ ꢀꢁ
ꢃꢋꢆ
ꢀꢋꢆ

Cu Sn S AND Cu SnS
ꢈ ꢀꢁ ꢃ
149
ꢇ
ꢆ
explained not as corresponding to strong vibrations, but to
a shift of Cu> or Ag> from the octahedral center to off-
center positions. For instance, in CuVP S , three different
ꢆ ꢁ
identifiable sites were found, one near the center of the Cu>
octahedron, another close to the triangular face of this
octahedron, and the last one in a van der Waals tetrahedral
position. In the AgScP S structure, only one octahedral
ꢆ ꢁ
Ag> site was found, and its abnormally large thermal para-
meter indicates a tendency to move away from the center of
the octahedra, which was attributed to a second-order
Jahn-Teller coupling between the filled silver e manifold
'
and the empty s orbital. The Cu3 atom in Cu Sn S is very
FIG. 3. Cu 2p core level XPS spectrum of Cu Sn S . The binding
energy positions (with the corresponding FWHM in parentheses) are
indicated in the figure.
ꢇ
ꢈ ꢀꢁ
ꢇ
ꢈ ꢀꢁ
similar to the Ag atom in AgScP S and a similar explana-
ꢆ ꢁ
tion may also be reasonable for our case.
In the pseudobinary Cu S—SnS system, CuSn
ꢃꢄꢈꢂ ꢅ
a defect cubic spinel-type structure with space group F-43m
S has
ꢆ
ꢆ
literature (23). The full width at half maximum (FWHM) (8). Its formula can be written as (Cu
ᮀ
)(Sn
ꢊꢄꢂ ꢊꢄꢂ ꢀꢄꢅꢈꢂ
)S . Thus, half of the tetrahedral sites and 15/16 of the
ꢊꢄꢀꢆꢂ ꢇ
calculated for Cu 2p
and Cu 2p
are 1.7 and 2.3 eV, ᮀꢀ
ꢃꢋꢆ
ꢀꢋꢆ
respectively, which are also in good agreement with the octahedral sites are occupied by Cu> and Snꢇ> ions, respec-
reported values for Cu> (23). Besides, the Cu 2p satellite tively. Although Cu Sn S has a rhombohedral symmetry,
ꢃꢋꢆ
ꢇ
ꢈ ꢀꢁ
peaks, which are usually centered at about 942 eV and its formula can also be similarly represented as
characteristic of Cuꢆ> (24), are missing in Fig. 3. Therefore, (Cu
ᮀ
)(Sn
Cu
)S . In the unit cell, both 3/4 of
ꢊꢄꢈꢂ ꢊꢄꢆꢂ
ꢀꢄꢈꢂ ꢊꢄꢆꢂ ꢇ
we may conclude that only Cu> is present in this com- the tetrahedral sites and 1/8 of the octahedral sites are
pound. XPS analysis also revealed the normal valence states occupied by Cu>, while 7/8 of the octahedral sites are
ꢂꢋꢆ
for Sn and S atoms. The Sn 3d
energies, 486.8 and 495.2 eV, respectively, compared well be described as a defect variant of the spinel-type.
and Sn 3d
ꢃꢋꢆ
binding occupied by Snꢇ> ions. Therefore, its crystal structure may
with those observed in SnS (25), and S core level spectrum
Cu SnS has the unit cell formula Cu
Sn
ꢆꢄꢁꢁꢂ(ꢈ) ꢀꢄꢃꢃꢂ(ꢈ) ꢇ
(Z"2) and crystallizes with the tetragonal stannite
S
ꢆ
ꢆ
ꢃ
shows the binding energies of 161.8 and 162.9 eV for 2p
ꢃꢋꢆ
and 2p states, respectively, very similar to those observed [Cu (Fe, Zn)SnS ] structure (31). The composite atoms M1
ꢀꢋꢆ
ꢆ
ꢇ
in chalcopyrite CuFeS (26).
[43.6(2) at.% Sn#56.4(2) at.% Cu] and M2 [46.3(3) at.%
ꢆ
In this structure, two Sn and four S positions are well Sn#53.7(3) at.% Cu] occupy the positions of the Cu and
defined, as revealed by their full local site occupancies and Sn atoms of stannite, respectively, and the rest of Cu atoms
the smaller equivalent thermal factors (Table 2). Of three Cu in Cu SnS occupy those of the corresponding (Fe, Zn)
ꢆ
ꢃ
atoms, the Cu1 site has almost a spherical displacement atoms. As Cu and Sn have different sizes, disorder between
ellipsoid, but it is half-occupied. This is a common occur- them is unique. However, the crystal of Cu SnS was ob-
ꢆ
ꢃ
rence in such a system. For example, the half-occupied tained by quenching from 700°C. At this temperature, such
tetrahedral Cu atom in CsCuCe S (27) has the comparable a copper distribution may be possible. Moh (5) previously
ꢆ ꢁ
%
equivalent thermal parameter [B of 1.7(2) A
s
ꢆ] and Cu—S observed under the microscope a complete solid solution
ꢁ
bond lengths [average value of 2.353(4) A
s
]. The displace- range between this phase and stannite (Cu FeSnS ) and
ꢆ
ꢇ
ment ellipsoid of the Cu2 atom is far from spherical, the concluded that this phase crystallizes in a tetragonal lattice
values of º being much bigger than those of º and º . similar to that of stannite. Our structural refinement con-
ꢃꢃ
ꢀꢀ
ꢆꢆ
This may be explained by its coordination geometry. As firms his conclusion. The crystal of Cu SnS showed narrow
ꢆ
ꢃ
mentioned above, the Cu2 site may be regarded as a tet- and symmetric diffraction peaks during reflection scans and
rahedral site in which the Cu2 atom bonds with three S2 the tetragonal lattice parameters obtained by the diffrac-
atoms [average distances 2.227(1) A
s
s
] along the xy plane and tometer [a"5.4126(1), c"10.8239(4) A] are reliable with
one S4 atom [2.922(3) A
s
] along the z direction. Thus Cu2 very small standard deviations. The fact that the diffraction
atom is strongly anisotropic along the z direction. The Cu3 peaks were not split and the systematic absences were con-
site has an octahedral S environment, which is unexpected. sistent with I-42m symmetry excludes the possibility of
The larger thermal factors of this atom may be associated twinning. Attempts to refine the structure in lower-sym-
with its greater mobility in the xy plane, which can be seen metry space groups with an ordered distribution of Cu and
from the relatively large º and º values in Table 2. Sn atoms were unsuccessful.
ꢀ
ꢀ
ꢆꢆ
A similar situation was previously observed in the com-
In the structure, all metal atoms are fixed on special
pounds CuCrP S (28), CuVP S (29), and AgScP S (30), positions and each is surrounded by four sulfur atoms
ꢆ ꢁ
ꢆ ꢁ
ꢆ ꢁ
where the high B%ꢁ values of closed shell Cu> and Ag> were forming a distorted tetrahedron. Each sulfur atom is

1
50
CHEN ET AL
TABLE 5
Selected Bond Lengths (A) and Angles (°) for Cu
unit cell structure and atomic coordination arrangements.
s
2
SnS
3
A comparison of this figure with the corresponding ones in
stannite indicates that the configurations of the coordina-
tion tetrahedra are slightly different in both structures. For
Cu SnS , the smaller copper atom at the origin of the unit
M1—S*
M1—S**
M1—S***
M1—S
2.3562(6)
2.3562(6)
2.3562(6)
2.3562(6)
2.356(1)
M2—Sꢄ*
M2—S**
Cu—S
Cu—Sꢄ**
2.356(1)
2.356(1)
2.307(1)
2.307(1)
2.307(1)
ꢆ
ꢃ
cell gives rise to a displacement of the coordinating sulfur
M2—S*ꢄ
Cu—S*ꢃ
toward this position, thereby forming almost an ideal CuS
tetrahedron [angles 109.46(2)—109.49(5)° in comparison
ꢇ
M2—Sꢄ
2.356(1)
Cu—Sꢄ***
2.307(1)
S*—M1—S**
S*—M1—S***
S**—M1—S***
S*—M1—S
S**—M1—S
S***—M1—S
S*ꢄ—M2—Sꢄ
S*ꢄ—M2—Sꢄ*
Sꢄ—M2—Sꢄ*
108.65(4)
S*ꢄ—M2—S**
Sꢄ—M2—S**
Sꢄ*—M2—S**
S—Cu—Sꢄ**
S—Cu—S*ꢃ
Sꢄ**—Cu—S*ꢃ
S—Cu—Sꢄ***
Sꢄ**—Cu—Sꢄ***
S*ꢃ—Cu—Sꢄ***
108.64(2)
111.15(5)
108.64(2)
109.46(2)
109.49(5)
109.46(2)
109.46(2)
109.49(5)
109.46(2)
with 107.0(1)°—110.7(1)° for Fe/Zn site in stannite]. Sim-
ilarly, the coordination tetrahedron around M1 is less dis-
torted than it is in stannite [angles 108.65(4)°—109.88(2)° as
opposed to 108.1(1)°—112.3(1)° for Cu site in stannite] and
the coordination tetrahedron around M2 is more distorted
with a wide angle range of 108.64(2)°—111.15(5)°. The ther-
109.88(2)
109.88(2)
109.88(2)
109.88(2)
108.65(4)
108.64(2)
111.15(5)
108.64(2)
mal parameters of Cu [B "8nꢆº "1.74(8) A
%ꢁ
ꢆ] atoms are quite comparable to those
s
ꢆ] and
%ꢁ
S [B "1.18(8) A
s
%
ꢁ
Note. Symmetry codes: (i) !y#1/2, x#1/2, !z#1/2; (ii) y!1/2, observed in the isotypic compound Cu CdSnS (1.64 and
ꢆ
ꢇ
!
x#1/2, !z#1/2; (iii) !x, !y#1, z; (iv) !x#1/2, !y#1/2,
z#1/2; (v) !y#1/2, x!1/2, !z#1/2; (vi) x!1/2, y!1/2, z#1/2;
vii) !y, x, !z; (viii) y, !x, !z; (ix) !x, !y, z.
1
.16 A
s
ꢆ, respectively) (33), but they are higher than those in
stannite. This probably reflects the overall differences in
their diffraction data. Although we could not compare the
thermal parameters of the mixed sites M1 and M2 with
those in the literature, the relative magnitudes º (Cu)'
(
%
ꢁ
tetrahedrally bonded to two M1, one M2, and one Cu º (Sn, Cu) in Cu SnS follow the general pattern
%ꢁ
ꢆ
ꢃ
atoms. Selected bond lengths and angles are collected in B (Cu)'B (Sn) observed in stannite and the other chal-
%ꢁ
is in good agreement copyrite-like structures.
%ꢁ
Table 5. The Cu—S distance 2.307(1) A
s
with that observed in the chalcopyrite structure [2.302(1) A
s
]
There are no S—S bondings in Cu Sn S and Cu SnS ;
ꢇ
ꢈ ꢀꢁ
ꢆ
ꢃ
(
[
32). The M1—S distance 2.3562(6) A is equal to that of M2—S the shortest S...S nonbonded contacts are 3.522(2) and
s
2.356(1) A] within experimental error, and both are inter- 3.768(2) A, respectively, significantly longer than the 2.07 A
s
s
s
,
mediate between 2.320(2) A of Cu—S and 2.411(2) A of Sn—S expected for S—S single bonds (34). Therefore, both
s
s
bond lengths observed in stannite (31). Figure 4 shows the compounds may be formulated as Cuꢀ>Snꢇ>Sꢆ\ and
ꢇ
ꢈ
ꢀꢁ
FIG. 4. Unit cell structure and atomic coordinations of Cu SnS . The interatomic distances are in angstroms and angles in degrees. The symmetry
ꢆ
ꢃ
codes here are the same as those in Table 5 and displacement ellipsoids are shown at the 50% probability level.

Cu Sn S AND Cu SnS
ꢈ ꢀꢁ ꢃ
151
ꢇ
ꢆ
3
4
5
. M. Khanafer, J. Rivet, and J. Flahaut, Bull. Soc. Chim. Fr. 2670 (1974).
. N. Wang, Neues Jahrb. Mineral. Monatsh. 424 (1974).
. G. H. Moh, in ‘‘Sulphide Systems Containing Sn,’’ Year book 62,
p. 197. Carnegie Inst., Washington, 1963.
6
7
. G. H. Moh, Chem. Erde 34, 1 (1975).
. S. Jaulmes, J. Rivet, and P. Laruelle, Acta Crystallogr. Sect. B 33, 540
(1977).
8
9
. S. Jaulmes, M. Julien-Pouzol, J. Rivet, J. C. Jumas, and M. Maurin,
Acta Crystallogr. Sect. B 38, 51 (1982).
. H. Hahn, W. Klingen, P. Ness, and H. Schulze, Naturwissenschaftea 53,
1
8 (1966).
1
0. P. Villars and L. D. Calvert, ‘‘Pearson’s Handbook of Crystallographic
Data for Intermetallic Phases.’’ Materials Information Society, Mater-
ials Park, OH, 1991.
1
1
1. N. Wang, Neues Jahrb. Mineral. Monatsh. 166 (1975).
2. B. A. Frenz and Associates, ‘‘SDP for Windows Reference Manual.’’
College Station, TX, 1995.
1
1
3. A. Altomare, G. Cascarano, C. Giacovazzo, and A. Guagliardi, J. Appl.
Crystallogr. 26, 343 (1993).
4. G. M. Sheldrick, ‘‘SHELXL93, a Program for the Refinement of
Crystal Structures.’’ University of G o¨ ttingen, Germany, 1993.
1
1
5. N. G. Zorina and S. S. Kvitka, Soviet Phys. Crystallogr. 13, 599 (1969).
6. P. K. Baltzer, P. J. Wojtowicz, M. Robbins, and E. Lopatin, Phys. Rev.
1
51, 367 (1966).
1
1
7. S. Jaulmes, Acta Crystallogr. Sect. B 30, 2283 (1974).
8. K. O. Klepp and D. Gurtner, J. Alloys Compd. 243, 19 (1996).
FIG. 5. Electrical conductivity of Cu Sn S versus temperature.
ꢈ ꢀꢁ
ꢇ
19. J. C. Jumas, E. Philippot, and M. Maurin, Acta Crystallogr. Sect. B 35,
195 (1979).
0. M. F. Mansuetto, P. M. Keane, and J. A. Ibers, J. Solid State Chem.
01, 257 (1992).
21. A. Daoudi, M. Lamire, J. C. Levet, and H. Noe
23, 331 (1996).
2. K. Klepp and K. Yvon, Acta Crystallogr. Sect. B 36, 2389 (1980).
rez, J. Alloys Compd. 234,
2
2
1
Cuꢀ>Snꢇ>Sꢆ\, the closed-shell system that one would ex-
ꢆ
ꢃ
¨ l, J. Solid State Chem.
pect to exhibit insulating or semiconducting properties. The
electric conductivity measurements of Cu Sn S were per-
1
2
ꢇ
ꢈ ꢀꢁ
formed in the cell arrangement Pt/sample/Pt. Experimental 23. J. Llanos, A. Buljan, C. Mujica, and R. Ramı
´
results were plotted in Fig. 5 as a function of temperature.
Data indicate clearly a typical semiconductor behavior. The
electric conductivity of Cu Sn S is log p+!5.01 at 22°C
40 (1996).
2
4. L. D. Partain, R. A. Schneider, L. F. Donaghey, and P. S. McLeod,
J. Appl. Phys. 57, 5056 (1985).
ꢇ
ꢈ ꢀꢁ
25. A. R. H. F. Ettema and C. Haas, J. Phys. Condens. Matter 5, 3817
1993).
energy of 0.42 eV. The supplemental DSC measurement 26. I. Nakai, Y. Sugitani, K. Nagashima, and Y. Niwa, J. Inorg. Nucl.
and increases linearly to !3.56 at 100°C with an activation
(
revealed no phase transition in this temperature range. The
electrical properties of Cu SnS have been investigated pre-
viously by Khanafer et al. (35), and it also exhibits the
Chem. 40, 789 (1978).
7. A. C. Sutorik, J. Albritton-Thomas, T. Hogan, C. R. Kannewurf, and
M. G. Kanatzidis, Chem. Mater. 8, 751 (1996).
8. P. Colombet, A. Leblanc, M. Danot, and J. Rouxel, J. Solid State Chem.
2
2
2
3
3
ꢆ
ꢃ
expected semiconducting behavior.
4
1, 174 (1982).
9. E. Durand, G. Ouvrard, M. Evain, and R. Brec, Inorg. Chem. 29, 4916
1990).
0. S. Lee, P. Colombet, G. Ouvrard, and R. Brec, Inorg. Chem. 27, 1291
1988).
1. S. R. Hall, J. T. Szymanski, and J. M. Stewart, Can. Mineral. 16, 131
1978).
(
ACKNOWLEDGMENTS
(
(X.-A.C.) thanks JST for an STA fellowship. The authors thank Mr. M.
Tsutsumi for carrying out the EDAX analyses, Mr. K. Kosuda for his help
with EPMA, and Dr. H. Nozaki for his guidance in electrical conductivity
measurement.
(
3
3
3
2. S. R. Hall and J. M. Stewart, Acta Crystallogr. Sect. B 29, 579 (1973).
3. J. T. Szymanski, Can. Mineral. 16, 147 (1978).
4. S. A. Sunshine, D. Kang, and J. A. Ibers, J. Am. Chem. Soc. 109, 6202
(
1987).
5. M. Khanafer, O. Gorochov, and J. Rivet, Mater. Res. Bull. 9, 1543
1974).
. W. F. Kuhs, R. Nitsche, and K. Scheunemann, Mater. Res. Bull. 14, 241 36. G. M. Sheldrick, ‘‘SHELXTL PC Version 4.1, an Integrated System for
REFERENCES
3
(
1
2
(
1979).
. O. Amiel, D. C. Frankel, and H. Wada, J. Solid State Chem. 116, 409
1995).
Solving, Refining, and Displaying Crystal Structures from Diffraction
Data.’’ Siemens Analytical X-Ray Instrument Inc., Madison, Wiscon-
sin, 1990.
(