POWER July 2017 - 38

ENVIRONMENTAL ISSUES
Table 3. Size matters. This table
breaks down the data into PM10 and PM2.5
percentages. Courtesy: MVA Scientific Consultants
Diameter
PM2.5
PM10
>10.0
Mass
(%)
8.8
62.1
37.9
electron imaging mode, is shown in Figure
3. In this mode, particles are clearly defined
against the background.
Figure 4 shows the particles that have
been recognized by the image analysis system.
Once the particles are recognized, automated
image analysis systems can determine
the average particle size, as well as any other
morphological parameters that might be of
interest, such as aspect ratio, circularity, and
perimeter.
In fully automated systems, the prepared
filter is stepped through the field of view of
the microscope until a sufficient number of
particles have been identified and analyzed,
which typically totals around 20,000 individual
particles in MVA's lab. Table 1 shows
a sample output, listing the requested particle
size and morphology parameters for the first
20 particles analyzed. The full dataset consists
of similar data for all 20,000 particles.
The accumulated data from the analysis
of the individual particles are submitted to
appropriate software for the PSD calculation
and other sample parameters, as desired.
Table 2 presents the PSD for the full dataset
in terms of both particle number and particle
mass percentages. Table 3 shows the data in
terms of PM10
and PM2.5
fractions.
When particulate matter has been collected
on quartz or glass fiber filters and
cannot be removed to a different substrate,
automated SEM methods cannot be used. For
these samples, SEM images of the particulate
(such as is shown in Figure 2) are analyzed
in image processing software, but the individual
particles in each image are identified
by a human operator, and the measurement
of individual particle sizes and any required
morphological parameters is done by the image
analysis software, with the parameters
for each individual particle saved in a manner
similar to that for fully automated analysis.
In those cases, the typical number of particles
analyzed is a few thousand, rather than a few
tens of thousands.
Scanning Electron Microscopy
Offers Options
Microscopical methods offer the ability to determine
PSDs on particulate matter samples
of as little as 0.25 mg of particulate collected
in Method 5 (or Method 17) sample trains,
allowing for significantly reduced sample
collection times when compared to cascade
impactor and cyclone methods of particle
sizing. These methods allow for the use of
completely customizable size fractions, and
the reported fractions can be adjusted after
sample collection.
Automated scanning electron microscopy
PSD determination is based on the measurement
of the physical dimensions of several
tens of thousands of individual particles.
For samples where fully automated analysis
is not possible, several thousand individual
particles are analyzed. Average diameter,
volume, perimeter, aspect ratio, and other
morphological parameters that may be of interest
in engineering and compliance studies
are recorded for each of the particles, and the
appropriate particle size and/or morphology
distributions are calculated. ■
-Tim B. Vander Wood, PhD
(tvanderwood@mvainc.com) is executive
director of MVA Scientific Consultants.
38
www.powermag.com
POWER | July 2017
http://www.powermag.com
POWER July 2017

Table of Contents for the Digital Edition of POWER July 2017
