Structural studies of nanocrystalline SnO2 doped with antimony: XRD and M?ssbauer spectroscopy

Article abstract of DOI:10.1016/S0022-3697(01)00226-8

A series of Sb-doped SnO₂ samples, with doping levels 0, 3.1, 6.2, 11.9 and 14.0 at% Sb, has been hydrothermally prepared and characterized by X-ray powder diffraction. Diffraction lines were broadened, the line broadening being anisotropic. Both the line broadening and line anisotropy were dependent on the Sb doping level. The samples are tetragonal, space group P4₂/mnm and isostructural with TiO₂(rutile). Sb doping of SnO₂ causes the increase of unit-cell parameters. The structure of pure SnO₂ and of samples containing 6.2 and 11.9 at% Sb has been refined by the Rietveld method. Crystal structure indicated that both Sb³⁺ and Sb⁵⁺ are substituted for Sn⁴⁺ in the SnO₂ structure, Sb³⁺ being dominant for the investigated doped samples. The samples were also examined by ¹¹⁹Sn- and¹²¹Sb-M?ssbauer spectroscopy. M?ssbauer spectroscopy confirmed the XRD results. Also, the values of the isomer shifts and quadrupole coupling constants indicated that the configuration around the Sb³⁺ site includes the presence of the stereochemically active lone pair electrons.

Full text of DOI:10.1016/S0022-3697(01)00226-8

Journal of Physics and Chemistry of Solids 63 ;2002) 765±772
www.elsevier.com/locate/jpcs
Structural studies of nanocrystalline SnO2 doped
È
with antimony: XRD and Mossbauer spectroscopy
a,
B. Grzeta , E. Tkalcec , C. Goebbert , M. Takeda ,
b,1
b
c
*
Ï
Ï
c
M. Takahashi , K. Nomura , M. Jaksic
d
Ï Âe
a
Ï Â
Division of Materials Chemistry, Ruder Boskovic Institute, P.O. Box180, 10002 Zagreb, Croatia
Å
Institut fur Neue Materialien, 66123 Saarbrucken, Germany
b
È
È
cDepartment of Chemistry, Faculty of Science, Toho University, Funabashi, Chiba 274-0072, Japan
dSchool of Engineering, University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-8656, Japan
e
Ï
Division of Experimental Physics, Ruder Boskovic Institute, P.O. Box180, 10002 Zagreb, Croatia
Å
Â
Received 18 May 2001; received in revised form 21 August 2001; accepted 7 September 2001
Abstract
A series of Sb-doped SnO2 samples, with doping levels 0, 3.1, 6.2, 11.9 and 14.0 at% Sb, has been hydrothermally prepared
and characterized by X-ray powder diffraction. Diffraction lines were broadened, the line broadening being anisotropic. Both
the line broadening and line anisotropy were dependent on the Sb doping level. The samples are tetragonal, space group P42/
mnm and isostructural with TiO2;rutile). Sb doping of SnO2 causes the increase of unit-cell parameters. The structure of pure
SnO2 and of samples containing 6.2 and 11.9 at% Sb has been re®ned by the Rietveld method. Crystal structure indicated that
both Sb31 and Sb51 are substituted for Sn41 in the SnO2 structure, Sb31 being dominant for the investigated doped samples. The
samples were also examined by 119Sn- and 121Sb-Mossbauer spectroscopy. Mossbauer spectroscopy con®rmed the XRD results.
Also, the values of the isomer shifts and quadrupole coupling constants indicated that the con®guration around the Sb31 site
includes the presence of the stereochemically active lone pair electrons. q 2002 Elsevier Science Ltd. All rights reserved.
È
È
È
Keywords: A. Oxides; B. Sol±gel growth; C. X-ray diffraction; C. Mossbauer spectroscopy; D. Crystal structure
1. Introduction
increased for heavily-doped SnO2 [6]. Also, it was noticed
that heavily-doped SnO2 was deeply coloured [7]. Up to
now many structural studies have been devoted to antimony
doped SnO2 to elucidate the above phenomena, but still
there are uncertainties as to how antimony is incorporated
into the SnO2 structure. Tsunashima et al. [8] found no trace
of any other crystalline phase but cassiterite in the diffrac-
tion pattern of thin SnO2 ®lms containing up to 30 at% Sb,
and heat-treated at 6008C. However, the diffraction lines
were broad; for larger antimony content the lines were
broader. The authors explained this effect by possible incor-
poration of amorphous antimony oxide into cassiterite.
Kikuchi and Umehara [9] found that the largest amount of
antimony incorporated at about 10008C was 2.3 at% Sb.
They assumed formation of a solid solution SbxSn12xO2,
with antimony in the Sb51 state incorporated into a cassiter-
ite structure. The lattice parameters were constant up to
0.5 at% Sb, while between 0.5 and 2.3 at% Sb the lattice
parameters increased linearly with the antimony content.
SnO2 ;cassiterite) posseses a tetragonal TiO2 ;rutile)-type
structure [1], in the space group P42/mnm. It is a wide-band
gap semiconductor ;Eg ˆ 3.97 eV) with transmittance
cut-off at 333 nm [2]. When doped with F, Sb or Mo it
becomes a conductor, while its optical transmittance is
invariant [3,4]. Because of these properties, SnO2 has been
a
widely used material for various optoelectronic
applications, such as thin ®lm transparent conductors in
electroluminescent devices, solar cells, etc. [3,5].
In the case of antimony doping, it was evidenced that the
resistivity decreased for lightly-doped SnO2, while it
* Corresponding author. Tel.: 1385-1-4561120; fax: 1385-1-
4680098.
Ï
E-mail address: grzeta@rudjer.irb.hr ;B. Grzeta).
1
Present address: Faculty of Chemical Engineering and Technol-
Â
ogy, University of Zagreb, Marulicev trg 20, 10000 Zagreb, Croatia.
0022-3697/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0022-3697;01)00226-8

Ï
B. Grzeta et al. / Journal of Physics and Chemistry of Solids 63 >2002) 765±772
766
Higher amount of antimony caused a separation of the
system into two phases, Sb2O4 and SnO2, respectively.
Studying the crystallization of amorphous Sb±Sn±O thin
®lms sprayed onto glass substrates, Kojima et al. [10]
found a non-linear increase of the lattice parameters of the
crystal ;cassiterite) phase, which they attributed to the
presence of antimony in two different oxidation states,
Sb31 and Sb51 [11]. However, they proved the coexistence
of both antimony oxidation states only by blackening of the
®lms heat-treated at 5008C. Crnjak-Orel et al. [4] also
observed the non-linear dependence of the lattice para-
meters of antimony-doped cassiterite on the antimony
content. Terrier et al. [12] tried to determine the actual
doping level of Sb in the doped SnO2 ®lms, and also the
oxidation state of incorporated antimony, by means of X-ray
photoelectron spectroscopy ;XPS) and secondary ion mass
spectroscopy ;SIMS). They found that in doping the Sb31
overcame the Sb51 content, but they could not ascertain the
exact content of antimony for which this occurred. The
present paper reports the preparation of powder samples
of SnO2 containing antimony in amounts up to 14.0 at%
Sb and their structural characterization. Locating antimony
in the cassiterite structure was achieved by means of
Prepared powder samples were characterized by X-ray
diffraction ;XRD) at room temperature ;RT) using a Philips
MPD 1880 counter diffractometer with monochromatized
Cu Ka radiation. The following data sets were collected:
;i) XRD patterns of the samples S0±S4 mixed with silicon
powder ;99.999%, Koch-Light Lab. Ltd., UK) as an internal
standard, scanned in steps of 0.028 ;2 u) in the 2 u range
from 10 to 1008 with a ®xed counting time of 5 s, for the
purpose of precise determination of unit-cell parameters, ;ii)
XRD patterns of pure samples S0, S2 and S3 scanned in
steps of 0.068 ;2 u) in the 2 u range from 10 to 1408 with
a ®xed counting time of 10 s, for the purpose of the Rietveld
structure re®nement [14].
121
The 119Sn- and Sb Mossbauer spectroscopy was per-
formed for determination of the oxidation state of tin and
È
119
È
antimony in the samples.
Sn Mossbauer spectra were
È
recorded at RT by a conventional transmission Mossbauer
spectrometer ;Topological System Co.) using a Ca119mSnO3
source ;37 MBq) and a NaI;Tl) scintilation detector. 121Sb
È
È
Mossbauer spectra were measured by a Wissel Mossbauer
spectometer system using a Ca121mSnO3 source ;16 MBq)
and a germanium solid-state detector; in these measure-
ments both the source and the samples containing antimony
were kept at 12 K in a cryostat incorporating a closed-cycle
refrigerator [15].
È
the Mossbauer spectroscopy and the Rietveld structure
re®nement.
2.3. Data processing
2. Experimental procedure
The XRD data of samples with admixed silicon powder
were employed for precise determination of unit-cell para-
meters. The method proposed by Toraya [16] was used.
Bragg angle positions, 2 u, of seven diffraction lines of the
examined sample and of three diffraction lines of silicon
were determined by the individual pro®le ®tting method
;program PROFIT [17]) and taken as input data for the
program UNITCELL [16]. The polynomial model of the
peak shift correction function was applied. Then, the unit-
cell parameters were re®ned by the whole-powder-pattern
®tting method using the program WPPF [17]. The ®tting
was performed using the split-type pseudo-Voigt pro®le
function and the polynomial background model. Aniso-
tropic line broadening was taken into account in the ®tting
procedure. The starting line pro®le parameters for the lines
hk0 and h00 were different from those for h0l and 00l lines
and from those for hkl lines, respectively.
2.1. Sample preparation
The powder samples of pure SnO2 and ones doped with
antimony in the amounts of 3.1, 6.2, 11.9 and 14.0 at% ;the
samples S0±S4) were prepared using a sol±gel technique
followed by hydrothermal treatment. A solution of tin;IV)
chloride in ethanol containing various amounts of SbCl3 was
added dropwise to an aqueous ammonia solution containing
10 wt.% of the surface modifying agent, b-alanine, with
respect to the oxide. The suspensions were treated at
1508C for 3 h. The resulting powders were isolated by
centrifuging and washed with deionized water several
times and after that hydrothermally treated at 2508C and
2.5 MPa. The samples were light yellow coloured except
of pure SnO2 which was white.
2.2. Methods
The crystal structures of the samples S0, S2 and S3 were
re®ned by the Rietveld method [14]. A starting structure
model for S0 was the structure of cassiterite, reported for
the single crystal by Baur [1]: space group P42/mnm, Sn41
on 2 ;a) sites, ;0, 0, 0; 1/2, 1/2, 1/2), and O22 on 4 ; f) sites,
^;x, x, 0; 1/2 1 x, 1/2 2 x, 1/2), with x ˆ 0.307. In this
structure, the six oxygen atoms about Sn form the octa-
hedron in which four Sn±O distances are slightly different
from the other two. For the Sb-doped samples, S2 and S3, a
cassiterite structure model was used in which a portion of
The antimony concentrations in the samples were
determined by means of both the atomic emission spectro-
scopy, using a Jobin Yvon JY24 spectrometer, and PIXE
;Particle Induced X-ray Emission) spectroscopy, using a
nuclear microprobe facility with a 3 MeV proton beam
and a semiconductor Si;Li) X-ray detector [13]. The
K-series of emitted X-ray radiation from thin powder
samples was used for the analysis. The results of atomic
emission spectroscopy and of PIXE analysis were equal
within the experimental error.
Sn41 ions was substituted by antimony ions, Sb31 and Sb51
.

Ï
B. Grzeta et al. / Journal of Physics and Chemistry of Solids 63 >2002) 765±772
767
Sb31 ion were taken from the International Tables for
X-ray Crystallography [19], and for O22 from Tokonami
[20]. Dispersion corrections were also applied. Isotropic
thermal vibration modes were assumed for all atoms.
119
The Sn Mossbauer spectra were computer-®tted with
È
one quadrupole doublet of Sn;IV) using the MossWinn
program [21], and the isomer shifts ;d) were determined
È
121
relative to SnO2 at RT. The Sb Mossbauer spectra were
analyzed using a transmission integral method [22] assum-
ing 12 quadrupole-split lines, while the isomer shifts were
determined relative to InSb at 12 K.
3. Results and discussion
3.1. XRD characterization of samples
XRD patterns indicated that all the prepared samples,
S0±S4, had the TiO2;rutile)-type structure [23]. Only
sample S4 contained impurity, as seen in Fig. 1, namely
~5 wt.% Sb2O3 ;i.e. ~2.7 mol% Sb2O3). Re®ned unit-cell
parameters of the samples are listed in Table 1 along with
the literature data for cassiterite [1,25]. Both a and c
parameters for the samples S1, S2 and S3 increased with
Sb-doping level, while for S4 decreased to the values
which could be situated between those for S0 and S1. If
we consider the ionic radii for the 6-coordinated Sn41
Fig. 1. Characteristic part of the XRD patterns of the samples S0, S1,
S2, S3 and S4. Black triangles denote diffraction lines of Sb2O3.
31
51
Ê
;0.83 A), Sb
Ê
;0.90 A) and Sb
Ê
;0.74 A) [26], such
behaviour of the cell parameters may indicate a possibility
that both Sb31 and Sb51 ions were substituted for Sn41 in the
samples S1±S4, with variable Sb31/Sb51 content ratio.
The ratio c/a was constant for all samples having the
value characteristic for the cassiterite structure, 0.672.
Diffraction lines were broadened ;Fig. 1 and Table 2),
indicating the nanosized crystallites in the samples. The
lines of sample S0 showed the greatest broadening. Incor-
poration of antimony up to 11.9 at% Sb ;sample S3)
lessened the broadening, while for the sample S4 the
broadening was bigger again. Also, the line broadening
Re®nement was performed with the program PFLS written
by Toraya and Marumo [18], using again the split-type
pseudo-Voigt pro®le function and the polynomial back-
ground model. Six background parameters, a zero-point
shift, the unit-cell parameters, three peak width parameters,
three peak asymmetry parameters and two mixing para-
meters for the pseudo-Voigt function were included in the
re®nement, with structural parameters. Atomic scattering
factors for Sn41 ion ;the same value for Sb51 ion) and
Table 1
Composition of Sb-doped SnO2 samples and re®ned values of their unit-cell parameters; space group P42/mnm ;136). Rp and Rwp are the
discrepancy factors which characterize a quality of ®tting result [24]
Ê
a ;A)
Ê
c ;A)
Ê 3
Sample notation
Antimony doping level ;at% Sb)
Rp
Rwp
c/a
V;A )
a
SnO2
4.737;1)
3.185;1)
0.672
0.672
0.672
0.672
0.672
0.672
0.672
71.47
b
SnO2
4.7367;1)
4.7331;4)
4.7399;5)
4.7402;5)
4.7406;4)
4.7379;4)
3.1855;1)
3.1815;4)
3.1854;4)
3.1856;4)
3.1865;4)
3.1830;4)
71.471
S0
S1
S2
S3
S4
0
0.044
0.049
0.042
0.037
0.049
0.063
0.072
0.065
0.053
0.073
71.273;3)
71.565;2)
71.579;2)
71.611;2)
71.451;2)
3.1;4)
6.2;6)
11.9;5)
14.0;7)c
a
Data quoted from Baur [1].
Â
Data quoted from Haines and Leger [25].
b
c
The total amount of antimony found in the sample S4 was 19.4;7) at%. A part of it ;5.4 at% Sb) crystallized as Sb2O3 ;namely 5 wt.%
Sb2O3), as found by XRD.

Ï
B. Grzeta et al. / Journal of Physics and Chemistry of Solids 63 >2002) 765±772
768
Table 2
Full-widths at half-maximum ;FWHM) for the selected diffraction lines of the samples and crystallite sizes in the direction normal to the planes
;110) and ;101), respectively
Sample
FWHM ;8 2 u)
D' ;110)
DÁ;101)
110
211
002
101
;nm)
;nm)
S0
S1
S2
S3
S4
1.892;8)
1.449;7)
1.409;6)
1.356;6)
1.838;8)
1.829;8)
1.405;7)
1.395;6)
1.297;6)
1.813;8)
1.724;8)
1.306;7)
1.018;6)
0.918;6)
1.373;8)
1.647;8)
1.107;7)
0.936;6)
0.813;6)
1.260;8)
4;1)
6;2)
6;2)
6;2)
4;1)
5;1)
8;2)
9;3)
11;3)
7;2)
was anisotropic: the lines hk0 and h00 were systematically
broader, while the lines h0l and 00l were systematically
narrower than the hkl lines. The line 101 was the narrowest
one for all the samples. A possible explanation of such a
character of the line broadening is proposed in Section 3.2.
The crystallite sizes for the samples were determined by a
well-known Scherrer formula [27] from diffraction lines 110
;the broadest one) and 101 ;the narrowest one), and the
values are listed in Table 2.
3.2. Rietveld structure re®nement
The observed and calculated powder patterns for the
samples S0, S2 and S3 are presented in Fig. 2. Re®ned
structural parameters are given in Table 3. The re®nement
for pure SnO2, sample S0, con®rmed the TiO2;rutile)-type
structure. In the case of the Sb-doped samples, S2 and S3,
re®nement was started using a cassiterite structure model in
which antimony in its whole content was incorporated as
Fig. 2. Results of the Rietveld re®nement for the samples S0, S2 and S3. The observed pro®le intensity is represented by black squares and the
calculated intensity by the solid line. Differences between the two intensities are plotted at the bottom of the diagrams on the same scale.
Vertical bars are the re¯ection position markers. Rp and Rwp values are given in Table 3.

Ï
B. Grzeta et al. / Journal of Physics and Chemistry of Solids 63 >2002) 765±772
769
Table 3
Structural parameters and their standard deviations for the samples S0, S2 and S3. Rp and Rwp are the discrepancy factors which characterize a
quality of ®tting result [24]
Ê 3
Sample
Rp
Rwp
Ion
Wyckoff position
Occupancy parameter
x
Biso ;A )
a
SnO2
Sn41
O22
2;a)
4;f)
1
1
0.307;1)
0.3070;4)b
S0
S2
0.057
0.054
0.077
0.072
Sn41
O22
Sn41
Sb51
Sb31
O22
Sn41
Sb51
Sb31
O22
2;a)
4;f)
1
0.53;3)
1.43;7)
0.19;2)
1
0.3066;7)
0.3073;6)
0.3074;6)
ꢀ
0.96
2;a)
2;a)
4;f)
0.04;2)
0.94;3)
0.91
0.19;2)
1.09;20)
0.59;3)
S3
0.052
0.069
}
2;a)
}
2;a)
4;f)
0.09;2)
0.91;4)
0.59;3)
1.49;27)
a
Data quoted from Baur [1].
Data quoted from Haines and Leger [25].
b
Â
Sb31, substituting for Sn41. During re®nement a possibility
of Sb51 incorporation was taken into consideration as well.
The occupancy parameter of Sb31 was varied, but the sum of
occupancy parameters for ;Sn41, Sb51) and Sb31 was kept
constant, i.e. equal to unity. Re®nements converged to
Rwp ˆ 0.119 and 0.136, respectively. With the introduction
of oxygen occupancy parameter into re®nements, Rwp
decreased to 0.072 for S2, and to 0.069 for S3. These re®ne-
ment results proved that samples S2 and S3 have the oxygen-
de®cient rutile-type structure in which both Sb31 and Sb51
ions substituted for Sn41, Sb31 being dominant. In S2
;containing 6.2 at% Sb) antimony is incorporated as 4 at%
Sb31 and 2.2 at% Sb51, while S3 ;containing 11.9 at% Sb)
incorporated 9 at% Sb31 and 2.9 at% Sb51. Found oxygen
de®ciency in S2 and S3 is the consequence of a charge
imbalance due to Sb31 substituting Sn41. Table 4 lists
selected metal±oxygen distances derived from the ®nal
atomic parameters in Table 3. In the SnO6-octahedra of
cassiterite six Sn±O distances are equal within the limits
of error. In the Sb-doped samples four Sn/Sb±O distances
remain unchanged, while two others are longer by 0.44%.
However, it is to be noted that this is an overall structural
picture. In the vicinity of an oxygen vacancy there should be
a lone pair of electrons instead of a metal±O bond.
If we look at Fig.1 and the data in Table 2 in the light of
the above results, the line anisotropy can be explained as
follows. In the series from sample S0 to sample S3 the
incorporated Sb31 content increases, and the oxygen
de®ciency increases as well. So, it seems that the oxygen
de®ciency enlarges the line anisotropy, with the line 101
being the narrowest one, and also reduces the line broaden-
ing. This conforms with the work reported by Geurts et al.
[28]. They found that a little deformation in the tin matrix of
SnO2 caused by the oxygen de®ciency can induce the
appearance of a layered structure, which is manifested by
anisotropy in the XRD line widths, with the 101 line being
the narrowest one. Also, the layered structure leads to the
increase in crystallite sizes, and decrease in line broadening.
In the case of S4 appeared an impurity phase Sb2O3. For that
reason, it is obvious that the incorporated Sb31 content in S4
;and oxygen de®ciency) does not follow the pattern noticed
for the above series. Indeed, the line anisotropy for S4 is
smaller than for S3, and lines of S4 are much broader.
Table 4
Selected interatomic distances ;A) and their standard deviations for
Ê
the samples S0, S2 and S3
Ê
Sample Interatomic distances ;A)
È
3.3. Mossbauer spectroscopy
Four distances ;Sn±O)1
Two distances ;Sn±O)2
2.056;7)
119
a
È
Sn Mossbauer spectra of the samples S1±S4
The
SnO2
2.052;5)
2.051;2)
S0
2.052;2)
recorded at RT are shown in Fig. 3, and Mossbauer para-
meters: isomer shift ;d) and quadrupole coupling constant
;e2qQ) are listed in Table 5. The results indicate that there is
no tin valence state else than Sn;IV) present in the samples.
The non-zero quadrupole splitting indicates the existence of
oxygen vacancies around Sn41 ions. The value of isomer
shift which is for all the samples slightly different from that
È
Four distances ;Sn/Sb±O)1 Two distances ;Sn/Sb±O)2
S2
S3
2.051;2)
2.051;2)
2.060;3)
2.061;3)
a
Data quoted from Baur [1].

Ï
B. Grzeta et al. / Journal of Physics and Chemistry of Solids 63 >2002) 765±772
770
Fig. 3. 119Sn Mossbauer spectra of: ;a) sample S1, ;b) sample S2, ;c)
sample S3 and ;d) sample S4, recorded at room temperature.
È
Fig. 4. 121Sb Mossbauer spectra of ;a) sample S1, ;b) sample S2, ;c)
sample S3 and ;d) sample S4, recorded at 12 K.
È
of pure SnO2 ;d ˆ 0 mm s21) suggests a small variation in
121
Ê
three short ;1.977±2.023 A) and two longer ;2.518,
the Sn±O bond length compared to the one in pure SnO2.
Ê
È
Sb Mossbauer spectra of the samples S1±S4
The
2.619 A) Sb±O bonds and a lone pair of antimony electrons
[29]. The values of d and e2qQ for Sb31 sites re¯ect sensi-
tively the stereochemical activity of the lone pair of elec-
trons and indicate the type of oxygen coordination [30]. It is
noted that both d and e2qQ for Sb31 decrease as the anti-
mony content increases from S1 to S4 ;Table 6), indicating
that the stereochemical activity around Sb31 is largest for S1
and decreases with the increase in antimony content. This
suggests that Sn31 ions are forced to substitute the octa-
hedrally coordinated Sb41. The e2qQ values for Sb51 in
S1±S4 are experimentally zero. This suggests that Sb51
ion is substituting for Sn41 in the centre of the octahedral
hole where the electric ®eld gradient eq vanished. This is
measured at 12 K are presented in Fig. 4. The broad asym-
metric maxima at around 212 mm s21 are due to Sb31 and
the symmetric maxima at 0 mm s21 to Sb51. The least
square curve ®tting was performed by assuming one Sb51
site and one Sb31 site in the samples S1±S3. In S4, one
Sb51 site and two Sb31 sites were assumed, namely that of
Sb31 incorporated into SnO2 structure and that of Sb31 in
Sb2O3 since 5 wt.% of Sb2O3 was detected in S4 by XRD.
The Mossbauer parameters for Sb31 site in Sb2O3 were ®xed
È
in calculations. Isomer shift ;d), quadrupole coupling
constant ;e2qQ), asymmetric parameter ;h), experimental
linewidth ;Gexp) and relative peak area ;A) for the examined
samples are summarized in Table 6.
Both d and e2qQ values for Sb31 in S1 are similar to those
for Sb31 in Sb2O3. This suggests that the environment
around Sb31 in S1 is similar to that in Sb2O3, which has
quite natural since the ionic radius for the 6-coordinated
Sb51 ;0.74 A) is slightly smaller than that of Sn
41
Ê
51
;0.83 A) [26]. The d values for the Sb site in S1±S4
Ê
well agree with those for double oxides having the
rutile and trirutile structure: MIIISbO4 ;M ˆ Al, d ˆ
8.70 mm s21) and MIISb2O6 ;M ˆ Mg, Zn, Cu; d ˆ 8.58±
8.70 mm s21) [31].
Table 5
119
È
Sn Mossbauer parameters for the samples S1, S2, S3 and S4
The atomic content ratio of each species is almost equal to
the relative peak area, A, of each species listed in Table 6.
Atomic contents of Sb31 and Sb51 in the samples S1±S4
calculated on the basis of relative areas are listed in Table 7
along with the values obtained from XRD examination and
the Rietveld re®nement. The atomic contents obtained from
Sample
Site
d ;mm s21
)
e2qQ ;mm s21
)
S1
S2
S3
S4
Sn41
Sn41
Sn41
Sn41
0.09
0.09
0.10
0.10
0.58
0.56
0.63
0.53
È
Mossbauer data well agree with those from XRD. Also, the

Ï
B. Grzeta et al. / Journal of Physics and Chemistry of Solids 63 >2002) 765±772
771
Table 6
121Sb MoÈssbauer parameters for the samples S1, S2, S3 and S4
a
Sample
Site
d ;mm s21
)
e2qQ ;mm s21
)
h
Gexp ;mm s21
)
A ;%)b
S1
Sb31
Sb51
Sb31
Sb51
Sb31
Sb51
Sb31
Sb51
Sb31
23.64
8.69
17.7
0.0
0.62
0c
3.15
3.15
3.23
3.23
3.00
3.00
2.92
2.92
2.92
66
34
70
30
74
26
49
19
32
S2
S3
S4
24.01
8.63
16.7
0.0
0.53
0c
24.21
8.67
16.5
0.0
0.76
0c
24.47
8.69
23.32c
15.8
0.58
0c
0.36c
0.0
18.3c
Sb2O3
a
Relative to InSb.
b
c
Relative area.
Fixed.
Table 7
Sb31 and Sb51 content in the samples S1, S2, S3 and S4 as obtained by the Rietveld re®nement ;RR) and Mossbauer spectroscopy ;MS)
È
Sample Total
Sb content
Incorporated
Sb31 content
;at%)
Incorporated
Sb51 content
;at%)
Average Sb31
content/Sb51
content
Impurity
Sb2O3;mol%)
;at%)
RR
MS
RR
MS
RR, MS
XRD
2.7
MS
3.1
S1
S2
S3
S4
3.1
6.2
±
2.0
4.3
8.8
9.5
±
1.1
1.9
3.1
3.7
1.8
2.0
3.0
2.6
4.0
9.0
±
2.2
2.9
±
11.9
19.4
value of 6.2 at% Sb31 forming Sb2O3 ;i.e. 3.1 mol% Sb2O3),
4. Conclusions
È
obtained from Mossbauer relative peak area of 32% for
È
Preparation, XRD and Mossbauer spectroscopy examina-
tions of the sample SnO2 and the ones doped with Sb in content
up to 14.0 at% Sb have been reported. The prepared samples
were nanocrystalline. All the samples were tetragonal with
TiO2;rutile)-type structure. Unit-cell parameters increased
with Sb content up to the doping level of ~12 at% Sb, after
which they decreased. The level of Sb doping in¯uenced the
broadening of diffraction lines, and the anisotropy of the line
widths. In the Sb-doped samples both Sb31 and Sb51 ions are
substituted on the Sn41 sites in the SnO2 lattice, Sb31 being
dominant. Above doping level of about 12 at% Sb a second
phase appeared in the system, namely Sb2O3.
sample S4 containing in total 19.4 at% Sb, agrees well
with that of 5 wt.% Sb2O3 ;i.e. 2.7 mol% Sb2O3) from
XRD within the experimental error. It is seen that the
Sb31/Sb51 content ratio increases with the increase in anti-
mony incorporated level up to sample S3, while for sample
S4 it is smaller than for S3. For all the samples the value of
ratio is greater than unity, i.e. the Sb31 content is larger than
Sb51 content. The fact that Sb31 favourably substitutes Sn41
implies that the charge imbalance due to this process should
be compensated by the formation of oxygen vacancies.
Indeed, the oxygen occupancy parameters as obtained by
the Rietveld re®nement decreased with the increase in the
antimony content doped ;Table 3).
Furthermore, cell volume increases with the increase in
Sb31 content ;in our case up to sample S3 as seen in Table 1).
Replacement of Sn41 by Sb31 thereby induces deformation
in the lattice, so substitution can occur only at the small Sb31
contents. At a certain limit the host lattice excludes Sb31
introduced, and Sb2O3 is formed as the second phase. This
occurs between 10 and 15 at% Sb. When it occurs, the
Sb31/Sb51 content ratio in the doped sample decreases and
the cell volume decreases consequently ;as found for sample
S4).
References
[1] W.H. Baur, Acta Cryst. A 9 ;1956) 515.
Ï Ï
[2] B. Orel, U. Lavrencic-Stangar, Z. Crnjak-Orel, P. Bukovac,
M. Kosec, J. Non-Cryst. Sol. 167 ;1994) 272.
[3] K.L. Chopra, S. Major, D.K. Pandya, Thin Solid Films 102
;1983) 1.
ÏÏ Ï Ï
[4] Z. Crnjak-Orel, B. Orel, M. Hodescek, V. Kaucic, J. Mater.
Sci. 27 ;1992) 313.

Ï
B. Grzeta et al. / Journal of Physics and Chemistry of Solids 63 >2002) 765±772
772
[5] B.O. Seraphin ;Ed.), Solar Energy Conversion, Springer
Verlag, Berlin, 1979.
[19] J.A. Ibers, W.C. Hamilton, International Tables for X-ray
Crystallography, Vol. IV, Kynoch Press, Birmingham, 1974.
[20] M. Tokonami, Acta Cryst. 19 ;1965) 486.
[6] A. Rohatgi, T.R. Viverito, L.H. Slack, J. Am. Ceram. Soc. 57
;1974) 278.
[21] Z. Klencsar, E. Kuzman, A. Vertes, J. Radioanal. Chem. Lett.
210 ;1996) 105.
[22] G.K. Shenoy, J.M. Friedt, H. Maletta, S.L. Ruby, in: I.J.
[7] D. Elliot, D.L. Zellmer, H.A. Laitinen, J. Electrochem. Soc.
117 ;1970) 1343.
È
Gruverman, C.W. Seidel, D.K. Dieterly ;Eds.), Mossbauer
[8] A. Tsunashima, H. Yoshimizu, K. Kodaira, S. Shimada, T.
Matsushita, J. Mat. Sci. 21 ;1986) 2731.
Effect Methodology, Vol. 9, Plenum Press, New York, 1974,
[9] T. Kikuchi, M. Umehara, J. Mater. Sci. Lett. 4 ;1985) 1051.
[10] M. Kojima, H. Kato, M. Gatto, Philosoph. Mag. B 68 ;1993) 215.
[11] M. Kojima, H. Kato, M. Gatto, J. Non-Cryst. Sol. 218 ;1997)
230.
p. 277.
[23] C. Legrand, J. Delville, C. R. Acad. Sci. Paris 236 ;1953) 944.
[24] R.A. Young, E. Prince, R.A. Sparks, J. Appl. Cryst. 5 ;1982)
357.
Â
[25] J. Haines, J.M. Leger, Phys Rev. B 55 ;1997) 11144.
[12] C. Terrier, J.P. Chatelon, J.A. Roger, R. Berjoan, C. Dubois, J.
Sol±Gel Science and Technology 10 ;1997) 75.
Ï Â Â Â Â
[13] M. Jaksic, I. Bogdanovic, D. Dujmic, S. Fazinic, T. Tadic,
[26] R.D. Shannon, Acta Cryst. A 32 ;1976) 751.
[27] P. Scherrer, Gottin Nachricht 2 ;1918) 98.
[28] J. Geurts, S. Rau, W. Richter, F.J. Schmitte, Thin Solid Films
121 ;1984) 217.
Strojarstvo 38 ;1996) 279.
[14] H.M. Rietveld, J. Appl. Cryst. 2 ;1969) 65.
[15] M. Takeda, Radioisotopes 24 ;1985) 628.
[16] H. Toraya, J. Appl. Cryst. 26 ;1993) 583.
[17] H. Toraya, J. Appl. Cryst. 19 ;1986) 440.
[18] H. Toraya, F. Marumo, Rep. Res. Lab. Eng. Mater. Tokyo
Inst. Technol. 5 ;1980) 55.
[29] C. Svensson, Acta Cryst. B 30 ;1974) 458.
[30] M. Takeda, M. Takahashi, R. Ohyama, I. Nakai, Hyper®ne
Interactions 28 ;1986) 741.
[31] Y. Kajitani, M. Takahashi, M. Takeda, unpublished
results.