|
B. Oxide Powder Characteristics
Both bulk samples of the mixed oxides and caps from
the stainless steel wells were analyzed to determine the
surface area and particle size for various compositions.
Bulk samples of the mixed oxide compositions were de-
composed and reacted following the heat treatment
schedule used for all of the actual libraries, except that
the samples were calcined at different final temperatures
(500, 600, and 800 °C) to observe the effects on particle
size and surface area. Figure 4 gives a plot of these
surface areas and their dependence on composition. A
dependence of surface area on composition is expected
because of the different crystal structures of the oxides.
Because this system is composed of mixed oxides and no
new crystal structure is formed, the surface areas of the
compositions in the library are mixtures of the two oxide
surface areas, which can be seen in Fig. 4. The effects of
pyrolysis are shown in Fig. 4 by the labeled Points A and
B. In the drying and slow decomposition stages of the
heat treatment these samples ignited and pyrolyzed. Sev-
eral attempts were made to avoid pyrolysis of these
compositions and are included in this figure. The samples
FIG. 4. Plot of experimentally determined BET surface areas of bulk
Cu
1−x
Ce
x
O
3
compositions calcined at 500, 600, and 800 °C and sur-
face areas of caps from combinatorial library calcined at 600 °C.
Points A and B were calcined at 800 °C, but pyrolysis occurred to
increase the surface area.
FIG. 2. Schematic representation of the (a) original and (b) final well
design used during powder processing and catalytic testing.
FIG. 3. Photograph of the dried resin caps in the Cu–Ce–O combina-
torial library. The composition at the upper left corner is 100% CeO
2
,
while to the right of it is the Cu
0.067
Ce
0.923
O
3
composition. The copper
content increases successively in each cap following a serpentine pattern.
H.M. Reichenbach
et al.:
Combinatorial synthesis of oxide powders
J. Mater. Res., Vol. 16, No. 4, Apr 2001
970
http://journals.cambridge.org
Downloaded: 15 Mar 2015
IP address: 138.251.14.35
(indicated by points A and B) clearly have surface areas
that are much greater than the neighboring compositions,
but because the conditions that occur during pyrolysis
cannot be controlled, the benefit of the high surface area
does not outweigh the need for reproducibility. Finally,
for comparison, the fourth data set on this plot shows the
surface area of powders from the oxide caps heated to a
final temperature of 600 °C. The particle sizes for the
oxide powder caps and bulk oxides processed at 600 °C
calculated from BET measurements are listed in Table I.
In addition to determining the particle size dependence
on composition, several samples were analyzed to ascer-
tain the variability of the process from one sample to
another. Table II shows the surface area for the three
compositions (x
⳱ 0.2, 0.5, 0.8) as well as the standard
deviation among the samples of each composition. The
samples were heated to a final temperature of 600 °C and
soaked for 2 h. The surface area measurements of these
caps confirm that reproducible samples can be generated
using the modified Pechini process with combinatorial
synthesis techniques.
Scanning electron microscopy was performed on sev-
eral of the caps to further examine the structure after the
600 °C soak. Figure 5(a) shows one of the powder oxide
caps (Cu
0.8
Ce
0.2
O
3
) at low magnification; several holes
and cracks are visible, that are the results of the decom-
position of the organic resin. Most of the cracks are ap-
proximately 50 to 100
m in length, as seen in Fig. 5(b),
and occur throughout the cap are easily discernible in this
figure. Particle agglomerates and individual particles are
also discernible in Fig. 5(c) and have a diameter of ap-
proximately 40–60 nm, similar to the values predicted by
calculations from surface area measurements. These par-
ticles are representative of the oxide powder found else-
where in the cap.
Do'stlaringiz bilan baham: | 
