1
30
PRITUZHALOV et al.
the cation sublattice but also in the anion one, generat-
Conductivity measurements were carried out with a
ing tysonite-like α-BiO F
tions of O for F ions (O
through partial substitu- two-point probe using a Solarton HF FRA 1255 fre-
y
3 – 2y
2–
2–
–
–
6
F ) [14].
quency-response analyzer (frequency range, 5 to 10 Hz)
on disk samples (d = 8 mm, h = 1–2 mm) with copper
or gold electrodes from 300 to 523 K under an argon
atmosphere. Gold electrodes were sputtered in vacuum.
Copper electrodes were formed during molding as fol-
lows: to a mold (d = 8 mm), a thin layer of powdered
copper, a layer of a powdered test sample, and another
layer of powdered copper were placed. The test disks
prepared in this manner (under a load of 1.0 t) with cop-
This work was intended to search for and study tyso-
3
+
nite solid solutions containing Bi with the goal of elu-
cidating relationships between their structural features
and ionic conductivity.
To achieve the above goal, we used simultaneous
aliovalent cation and anion doping: in the solid solution
α-BiOyF3 – 2y, bismuth was partially substituted by
sodium, strontium, and neodymium cations. In choos- per leads on their ends were sintered at 723 K in welded
ing these dopant ions, we were guided not only by their copper ampoules under a dry argon atmosphere for sev-
different oxidation numbers but also by the closeness of eral hours. X-ray diffraction did not signify any reac-
their sizes to those of bismuth and fluorine ions: RF
–
=
tion between the lead material and test samples or any
change in the composition of test samples. Test disks
were transferred to a silica cell; the cell was degassed in
dynamic vacuum (423 K, 2 h) and filled with dry argon.
The resulting impedance loci were processed using the
Zview2.1 program (1998 Scribner Associates, Inc.,
written by Derek Johnson).
1
.17 Å, RO2– = 1.24 Å (CN = 4), RBi3+ = 1.31 Å, RNa
+
=
1.32 Å, RSr2+ = 1.40 Å, and RNd3+ = 1.25 Å (CN = 8) [15].
EXPERIMENTAL
The starting chemicals used were commercially
available samples of strontium, sodium, and neody-
mium fluorides, and bismuth oxides (all of pure for
analysis grade) and bismuth fluoride (chemically pure
grade). All starting chemicals were pretreated to
remove trace water and other impurities.
RESULTS AND DISCUSSION
BiF –BiOF–NaF and BiF –BiOF–SrF System
3
3
2
X-ray powder diffraction data for annealed samples
of the system BiF –BiOF–MF (where M = Na or Sr;
3
n
n = 1 or 2) show α-BiOyF3 – 2y-base tysonite solid solu-
Sodium and strontium fluorides were dried to con-
tions Bi1 – xå (O, F)3 – d: Bi1–xNa O F
(Fig. 1a)
x
x
y
3 – 2x – 2y
stant weight at 423 K; NdF and BiF were exposed to
3
3
and Bi1 – xSr O F (Fig. 1b). The X-ray diffraction
x
y 3 – x – 2y
flowing gaseous HF at 573 K for 3 h and degassed in
dynamic vacuum (~10 Pa) at 473 K for 2 h.
patterns for all single-phase tysonite samples were
indexed assuming the hexagonal crystal system (phase I)
by analogy with the α-BiOyF3 – 2y tysonite phase [6].
The unit cell parameters are listed in Tables 1 and 2.
In addition to the commercial bismuth fluoride sam-
ple, we used a sample prepared in a laboratory from bis-
muth hydroxide (chemically pure grade), as in [16]. At
Figures 1a and 1b display the experimentally deter-
mined homogeneity regions with account for the litera-
ture data on α-BiOyF3 – 2y and β-BiOyF3 – 2y [14, 16, 17].
the first stage of this synthesis, Bi(OH) was concen-
3
trated with 40% HF in a Teflon beaker; next, the precip-
itate was dehydrated in a copper-and-nickel reactor at
73 K for 3 h in flowing gaseous HF.
5
Figure 1a shows the triangulation scheme for the
system BiF –BiOF–NaBiF at 873 K suggested pro-
3
4
X-ray diffraction identification of the starting chem-
icals verified the absence of impurity phases. Ready-
ceeding from the results of this work and [18]. There
are three solid solutions (ss) in this system, namely flu-
orite solid solutions (F ss), tysonite solid solutions
T ss), and β-BiOyF3 – 2y-base ones, as well as five two-
phase fields and four three-phase ones.
for-use starting chemicals were stored over P O in a
2
5
desiccator.
(
Test samples were prepared by solid-state synthesis.
Homogenized and pelletized blends of powdered
chemicals were transferred to copper ampoules.
Figure 1b shows our suggested triangulation scheme
1
Ampoules with test samples were degassed in dynamic for the system BiF
–BiOF–SrF
at 873 K. There are
3
2
vacuum at 473 K for 1 h, filled with dry argon, and her- three solid solutions (ss) in this system, namely F ss,
metically welded. Ampoules were annealed (873 ± 10 K, T ss, and β-BiOyF3 – 2y ss, as well as six two-phase
3
h) and quenched to cold water.
fields, and three three-phase ones. The fields designed
in terms of the phase rule and unverified experimentally
are indicated by dashed lines in both phase diagrams.
The phase composition of test samples was studied
by X-ray powder diffraction. X-ray diffraction patterns
were recorded in a Guinier monochromator camera
with the effective diameter 228 mm using CuK radia-
1
The authors thank E.N. Dombrovskii, a student of the Chemistry
α1
Department, Moscow State University, for participation in the
experimental refinement of some fields of the phase diagram.
tion.
RUSSIAN JOURNAL OF INORGANIC CHEMISTRY Vol. 53 No. 1 2008