Results and discussion
3.1. FE SEM macroscopic study
In a macroscopic study, a randomly selected area of
the support is analyzed to get a statistically significant
view of the particle size and morphology. Observations
of a sufficient number of particles must be made to get
meaningful particle size distributions (PSD). For this
part of our study, we used a field emission SEM (FEI
DB235) operating at low voltages with in-lens detection of backscattered electrons (BSE) to minimize charging
and improve image contrast to permit automated particle
counting. Figure 3 shows the contrast enhancement
available with BSE imaging over conventional secondary
electron (SE) imaging. Figure 3a was obtained at
15 kV in a conventional SEM (JEOL 6300 V LaB6)
using an Everhart-Thornley (ET) detector. The sample
was tilted to emphasize topographic contrast. Artifacts
due to charging are visible throughout this image and
the contrast for detection of nanoparticles is not very
good. Figure 3b shows an image of the same sample by
obtained using the FEI DB235 at 5 kV but with an
in-lens backscattered electron detector. The BSE images
make it possible to detect particles as small as 10 nm.
This image was obtained without any carbon coating,
but subsequent images in this section were all obtained
after a thin film of carbon was deposited on the sample.
Images and particle size distributions from isothermal
sintering at 900 C in 600 Torr of N2, with 4 nm of
palladium film on quartz and sapphire supports can be
seen in figures 4, 5 at time t = 0 and t = 24 h. The
amount of Pd on both these samples is identical, but we
see significant differences in the mean particle size and
the size distribution. For the Pd on quartz, the mean size
was 36 nm and the particle size distribution was lognormal.
The Pd on sapphire on the other hand a broader
size distribution with a mean diameter of 46 nm, considerably
larger than that on the quartz. These initial
differences persisted as the sample was aged at 900 C in
N2. After 24 h of aging, the mean sizes were 47 nm and
61 nm respectively. The Pd on sapphire was aged further
for an additional 48 h [14]. The rate of particle growth
slowed considerably and at the end of 72 h of aging the
mean particle size was 72 nm [14]. There is some suggestion
of particle coalescence in the images (see
figure 4c). Furthermore, the log normal distribution has
been suggested to be an indication that the predominant
sintering mechanism is particle migration and coalescence
[15]. However our recent work has led us to the
conclusion that log normal size distributions are seen
both under conditions where Ostwald ripening as well as
particle migration are implicated [4].
Results and discussion3.1. FE SEM macroscopic studyIn a macroscopic study, a randomly selected area ofthe support is analyzed to get a statistically significantview of the particle size and morphology. Observationsof a sufficient number of particles must be made to getmeaningful particle size distributions (PSD). For thispart of our study, we used a field emission SEM (FEIDB235) operating at low voltages with in-lens detection of backscattered electrons (BSE) to minimize chargingand improve image contrast to permit automated particlecounting. Figure 3 shows the contrast enhancementavailable with BSE imaging over conventional secondaryelectron (SE) imaging. Figure 3a was obtained at15 kV in a conventional SEM (JEOL 6300 V LaB6)using an Everhart-Thornley (ET) detector. The samplewas tilted to emphasize topographic contrast. Artifactsdue to charging are visible throughout this image andthe contrast for detection of nanoparticles is not verygood. Figure 3b shows an image of the same sample byobtained using the FEI DB235 at 5 kV but with anin-lens backscattered electron detector. The BSE imagesmake it possible to detect particles as small as 10 nm.This image was obtained without any carbon coating,but subsequent images in this section were all obtainedafter a thin film of carbon was deposited on the sample.Images and particle size distributions from isothermalsintering at 900 C in 600 Torr of N2, with 4 nm ofpalladium film on quartz and sapphire supports can beseen in figures 4, 5 at time t = 0 and t = 24 h. Theamount of Pd on both these samples is identical, but wesee significant differences in the mean particle size andthe size distribution. For the Pd on quartz, the mean sizewas 36 nm and the particle size distribution was lognormal.The Pd on sapphire on the other hand a broadersize distribution with a mean diameter of 46 nm, considerablylarger than that on the quartz. These initialdifferences persisted as the sample was aged at 900 C inN2. After 24 h of aging, the mean sizes were 47 nm and61 nm respectively. The Pd on sapphire was aged furtherfor an additional 48 h [14]. The rate of particle growthslowed considerably and at the end of 72 h of aging themean particle size was 72 nm [14]. There is some suggestionof particle coalescence in the images (seefigure 4c). Furthermore, the log normal distribution hasbeen suggested to be an indication that the predominantsintering mechanism is particle migration and coalescence[15]. However our recent work has led us to theconclusion that log normal size distributions are seenboth under conditions where Ostwald ripening as well asparticle migration are implicated [4].
การแปล กรุณารอสักครู่..