REACTIONS OF MnO2, Mn2O3, α-Bi2O3, AND Bi12Ti1 – xMnxO20
583
wise, either there will be an uncontrolled pressure rise (Geigerflex diffractometer, Ni-filtered CuKα radiation)
even at moderate temperatures or a gaslike, low-density
fluid will form.
and thermal analysis (DTA + TG) in air (STA-409
Netzsch thermal analyzer, platinum orAl2O3 crucibles).
The samples were heated and cooled at a rate of
5°C/min.
The ampule was placed in a massive steel block with
two windows for illuminating the reaction mixture and
examining it using an optical system. The block was
designed to minimize the temperature gradient and to
protect the operator in the case of explosion. It was
mounted in a resistance furnace, whose heating and
cooling rates could be varied in a wide range.A thermo-
couple was placed in direct contact with the middle part
of the ampule wall and was calibrated from supercriti-
cal phenomena in water. The difference in the wall and
mixture temperatures was normally no greater than 3–
5°C. The assembly could be rotated about an axis per-
pendicular to the ampule axis to drive the liquid and
solid (loose or dense) phases from one ampule end to
the other. Rotation ensured proper stirring of the reac-
tion mixture and, more importantly, enabled us to
observe all the phases present and to determine the color
of the noncrystalline phases throughout the run [9].
In the first autoclave procedure, we used Pyrex
ampules with a volume of 12–15 cm3, diameter of 14–
15 mm, and wall thickness of 0.7–1 mm. A metal oxide
powder (0.5 g) or crystal plate (0.01–0.2 g) and isopro-
panol were sealed in the ampule. The amount of the
alcohol was such that the mean density was as close to
the critical isopropanol density as possible. Accord-
ingly, the liquid and vapor were at equilibrium up to the
critical point and the supercritical fluid had an appropri-
ate density. Six ampules, in a holder with an internal
thermocouple, were placed in a 1-l autoclave. An
appropriate amount of isopropanol was also poured
into the autoclave to counterbalance the pressure in the
ampules and to prevent them from breaking. The auto-
clave was placed in a furnace and held at the preset tem-
perature. After cooling, the ampules were unsealed, and
their content was analyzed.
In the first run, we studied the behavior of pure iso-
propanol between room temperature and 240°C in
order to calibrate the thermocouple and determine the
amount of alcohol that would ensure an appropriate
fluid density. We located the critical point, refined the
temperature calibration plot, and found that heating
causes no observable changes in isopropanol.
This procedure has two drawbacks. First, it is dif-
ficult to charge or empty the ampule. Second, there is
a high probability of the ampule exploding during
unsealing, because the reaction yields gaseous pro-
ducts.
In the second and third runs, the ampule was
charged with bismuth oxide (0.1 g) and isopropanol
(one-third of the ampule volume). As the ampule was
rotated, the liquid and powder moved readily even at
room temperature. The critical phenomena in the liq-
uid–vapor system took place at pressures and tempera-
tures characteristic of pure isopropanol, suggesting that
no considerable contamination of the alcohol occurred.
In both runs, heating the ampule to 300–320°C caused
no changes in the color of the liquid, vapor, and fluid.
The powder, initially bright yellow, turned light gray.
The second autoclave procedure is free of these
drawbacks. We employed small, 15-cm3 autoclaves
made from the EP-943 nickel alloy. A quartz tube con-
taining an oxide powder (0.5 g) or plate (0.01–0.2 g)
was placed into the autoclave. The amount of isopro-
panol to be added was calculated as in the first proce-
dure. Eight autoclaves were placed in an air thermostat
and held at temperature for the required time. The seal
used in the autoclaves was quite safe: it withstood pres-
sures of at least 0.5 GPa, ruling out autoclave explo-
Further studies showed that this change in color was sion, and allowed the autoclave to be easily and safely
due to oxide reduction, yielding metal granules. Thus, unsealed after the experiment.
it was ascertained that the system was heterogeneous
In the first autoclave procedure, the reaction mixture
(contained a solid phase) throughout the reaction.
is in contact with glass only, while in the second, it is in
Unfortunately, the color change could not be
contact with the autoclave wall as well. To check that
observed in situ, because any powder appears black in
these procedures give equivalent results, we carried out
transmitted light. Furthermore, it was practically
special runs in which the autoclave was charged with
impossible to unseal the ampule in order to analyze the
isopropanol only. Chemical analysis demonstrated that
reaction products.
isopropanol had the same composition before and after
For high-pressure experiments, we developed two
autoclave procedures, one using sealed ampules and the
other using open tubes. The system was brought to a
supercritical state by raising the temperature and,
accordingly, pressure. Standard experiments were per-
formed 50–70°C above the critical temperature of iso-
propanol or at even higher temperatures, outside the
instability region, which generally lies near the transi-
those runs.
The reaction temperature was maintained within
1−2°C in both autoclave procedures. The error in the
measured temperature was estimated at ±10–15 and
±5–7°C for the first and second autoclave procedures,
respectively. In view of this, we conducted our experi-
ments either well below or well above the critical tem-
tion point. After the ampule was cooled and unsealed, perature to be certain about the phase composition of
the solid phase was characterized by x-ray diffraction the mixture.
INORGANIC MATERIALS Vol. 38 No. 6 2002