United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S4-84-048 Aug. 1984
Project Summary
Investigation of Filter Media for
Use in the Determination of
Mass Concentrations of Ambient
Paniculate Matter
Kenneth A. Rehme, C. Frederick Smith, Michael E. Beard, and Terence Fitz-
Simons
Quartz and Teflon® filter media were
evaluated under controlled use conditions
to determine their suitability for ambient
particulate measurements. Weighing
tests and handling tests were conducted
in a laboratory environment. A field
comparison test in which samples were
collected on glass, quartz, and Teflon®
filters was conducted at a typical air
quality monitoring site.
An analysis of weight changes ob-
served during repeated weighings of 8 x
10 inch and 102 mm glass, quartz, and
Teflon® filters showed no evidence of
systematic weight loss during the
weighing process. Weight losses were
observed for both glass and quartz
filters during filter mounting and
handling tests, but the estimated errors
in corresponding mass concentration
measurements due to such weight
losses were always less than 3 /yg/m3.
Teflon® filters generally gained weight
during these tests.
Total suspended particulate, nitrate,
and sulfate concentrations measured
during the field comparison test using
quartz and Teflon® filters on high-
volume samplers were lower than those
measured using glass filters. Observed
differences could be explained reason-
ably well by artifact effects and the
aforementioned handling effects. Teflon®
filters showed a tendency to clog at
ambient total suspended particulate
concentrations around 75 pg/m3.
This Project Summary was developed
by EPA's Environmental Monitoring
Systems Laboratory, Research Triangle
Park. NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering
information at back).
Introduction
The most widely used methods for
measuring the mass concentration of
particulate matter in the atmosphere
involve aerosol collection on a filter
substrate and subsequent gravimetric
mass determination. The current EPA
reference method for determination of
suspended particulates in the atmosphere
(TSP) uses the high-volume sampler and
a glass fiber filter for sample collection. A
major disadvantage of glass fiber filters is
their inability to provide a chemically inert
collection surface. Consequently, artifact
formation resulting from gas-to-particle
conversions on the filter surface often
represents a significant interference in
the desired mass concentration measure-
ment. Sulfate and nitrate artifacts,
formed by the oxidation of acidic gases
(i.e., SOa, N02> and retention of nitric acid
on the surface of these alkaline glass
fiber filters, have been demonstrated by
several investigators in both laboratory
and field studies. Estimates of the most
probable combined errors from sulfate
and nitrate artifacts range from about 7
jjg/m3 (typical sampling locations) to as
high as 11 /ug/m3 (Los Angeles Basin) for
a 24-hour sampling period.
Over the past few years size-specific
.particulate samplers have been used to
-------
collect air quality data in support of
anticipated revisions in the national
ambient air quality standards (NAAQS) for
paniculate matter. These samplers
include the conventional high-volume
sampler equipped with size-selective
inlets designed to collect particles in the
< 15 fjm and < 10 fjm size ranges.
Although glass fiber filters continue to be
used with these samplers, it is clearly
evident that mass concentration errors
resulting from artifact formation pose a
more significant problem for the size-
specific measurements than for TSP
since less mass is collected in the < 15 fjm
or < 10 fjm size ranges.
Consideration of available alternative
filter media for these samplers indicates
that quartz filters may exhibit less artifact
interference. However, their use in
routine sampling will depend on whether
they are sufficiently rugged to withstand
the normal handling operations encoun-
tered in a typical air monitoring applica-
tion. Quartz filters are known to be
extremely fragile and may be prone to
fiber loss during weighing, handling, and
sampling.
Teflon,® used as the filter substrate in
low-volume (dichotomous) samplers and
the recently developed medium-volume
(4 cfm) samplers, is another alternative
filter medium. Positive sulfate and nitrate
artifact formation is not a problem with
Teflon® filters; however, loss of paniculate
nitrate by dissociation or chemical
reaction has been reported. Physical loss
of particles after sample collection and
static charge interferences in the weigh-
ing process can be problems. Teflon®
filters are also prone to more rapid
clogging as mass loadings increase.
Commercially available quartz and
Teflon® filters were evaluated under
carefully controlled laboratory and field
conditions to determine the suitability of
these filters for particulate mass concen-
tration measurements. Tests included
weighability and handleability tests, as
well as a field comparison test in which
TSP samples were collected on glass,
quartz, and Teflon® filters over a 12-day
period.
Procedure
The filters selected for evaluation were
Gelman microquartz fiber filters (1979
and 1981 production), Whatman quartz
microfibre filters (QM-A), Pallflex tissu-
quartz filters (2500 OAST), and Membra-
na/Ghia Zefluor™ membrane filters
(P5PI). Schleicher and Schuell glass fiber
filters (1981 production) were also
included for comparative purposes.
Ten 8x10 inch filters of each type were
subjected to repeated weighings to
determine whether significant weight
losses occurred during the weighing
process. Filters were conditioned and
then weighed once each day for six
days in a climate controlled weighing
room (< 50% RH, T = 22°± 3° C). Ten addi-
tional filters of each type were also
subjected to a handling test to determine
the magnitude of weight losses due to
placement on a sampler. In this test the
filters were conditioned and then weighed
twice before and twice after placement
on a high-volume sampler. All filter
loading, unloading, and other handling
operations were designed to simulate the
normal handling that filters undergo
during typical air monitoring applications.
Particle free air was drawn through each
filter for 5 minutes during this test.
Similar handling tests were also conducted
using 102 mm circular filters and a
medium-volume (4 cfm) sampler.
A field comparison test using the six
types of filters was conducted at a typical
urban-commercial-industrial air moni-
toring site in Durham, North Carolina.
The six filter types were systematically
alternated among six high-volume samp-
lers so that each filter type was used
twice with each sampler over 12 sampling
days. Total suspended particulate (TSP)
measurements were obtained each day
for each filter type. Sulfate and nitrate
analyses were performed on all collected
samples.
Results and Discussion
The average weight changes between
successive weighings (over six weighings)
of 8 x 10 inch filters and the standard
deviations of the weight changes are
given in Table 1. The average weight
changes over the six weighings are given
in the bottom row of the table.
The magnitude of the changes in weight
for all filter types were similar except for
the Gelman(81) quartz filters, which
exhibited much more dramatic effects
than the other filters. A strong correlation
(r2 = 0.88) was obtained between the
average weights of the ten Gelman(81)
quartz filters and the relative humidity
in the weighing room.
An analysis of covariance that included
humidity as a factor showed no evidence
of systematic weight losses during the
weighing process for any of the filters
tested. It was concluded that some
random effect associated with each
weighing session was the major compo-
nent of the observed variation in the data
over the six weighing sessions. The
results with the Gelman(81) filters
suggest that closer control of relative
humidity in the filter conditioning and
weighing environment might be advan-
tageous for some quartz filters, but
further investigation is recommended.
The average weight changes between
successive weighings before and after
mounting of the filters on a high-volume
sampler and the average weight changes
due to handling and mounting on the
sampler are tabulated in Table 2. During
this filter mounting test the sampler was
turned on and allowed to sample clean air
for 5 minutes. It was assumed that the
major loss of fiber material from the
filters would occur during this sampler
start-up period.
The weight changes between two
successive weighings before (W2-W1)
and after (W4-W3) mounting of the filters
on the sampler and,the filter-to-filter
variabilities were similar (except for the
Table 1. Weight Changes (mg) Between Successive Weighings (8x10 inch Filters)
W2-W1:'
W3-W2:
W4-W3:
W5-W4:
W6-W5:
W6-W1:
Avg.
S.D.
Avg.
S.D.
Avg.
S.D.
Avg.
S.D.
Avg.
S.D.
Avg.
S.D.
S&S
Glass
40.37
0.19
-0.06
0.11
+0.13
0.12
-0.43
0.15
+0.33
0.13
+0.30
0.17
Gelman(79)
Quartz
-0.03
0.22
+0.14
0.16
-0.10
0.18
-0.67
0.22
-0.13
0.17
-0.79
0.22
Gelman(81)
Quartz
+ 1.88
0.36
+0.05
0.23
-0.98
0.47
-1.67
0.27
+0.81
0.38
+0.10
0.54
Whatman
Quartz
+0.08
0.36
-0.49
0.43
-0.12
0.13
-0.43
0.16
+0.06
0.12
-0.90
0.29
Pallflex
Quartz
+0.26
0.15
-0.18
0.25
-0.23
0.15
+0.14
0.24
-0.05
0.12
-0.06
0.18
Gh/a
Teflon®
-0.26
0.35
+0.24
0.16
+0.19
0.22
+0.14
0.20
-0.02
0.18
+0.29
0.39
*W1 indicates the first weighing and so on.
-------
Ghia Teflon® filters) to the results
obtained in the earlier weighing tests.
The weight changes (W3-W2) due to
handling and mounting on the sampler
(with clean air flow for 5 minutes) were
substantially higher for the quartz filters
than for the S&S glass filters. The Ghia
Teflon® filters gained weight (+ 0.50 mg)
with a high filter-to-filter variability. An
analysis of variance revealed that moun-
ting was a significant effect for all filter
types except the Ghia Teflon.®
The average of the first two weighings
and the average of the last two weighings
were used to estimate the weight
changes due to mounting the filters.
These estimates appear in Table 3 with
estimates of the corresponding errors in
mass concentration measurements. The
estimated errors in mass concentration
for quartz filters were less than the
estimated errors due to artifact formation
on glass filters for most sampling
locations.
The results of mounting tests for the
102 mm filters were comparable to those
obtained for the 8x10 inch filters after
adjustment for filter size and sampler
flowrates. Once again, the Ghia Teflon®
filters showed a slight weight gain
equivalent to 0.2 fjg/nr.
The measured TSP concentrations for
the 12-day field comparison test when all
six types of filters were used for sampling
are given in Table 4. Each day the highest
TSP concentration was obtained with the
S&S glass filters, except for day 2 when
the Whatman quartz gave the highest
value. The Whatman filters contain a
small amount (5 percent) of borosilicate
glass, added during the filter manufactur-
ing process. This small amount of glass
can apparently cause an increase in
artifact formation over what would be
expected from filters containing only
quartz. Sulfate and nitrate measurements
(not shown) were higher for the S&S and
Whatman filters than for the quartz or
Teflon® filters. On two of the days (days 5
and 6) clogging of the Ghia Teflon® filters
resulted in a significant drop in the
sampler flow rates. Since the TSP
concentrations were only 65 to 75 A/g/m3
on these days, these results suggest that
3.0 fjm pore size Teflon® might not be
suitable for use on high-volume samplers
because of potential overloading prob-
lems.
Conclusions and
Recommendations
1. An analysis of the weight changes
observed during repeated weighings of 8
x 10 inch and 102 mm glass, quartz, and
Table 2. Weight Changes (mg) Between Successive Weighings and Before and After Mount-
ing on Sampler (8x10 inch Filters)
W2-W1:"
W4-W3:
W3-W2:
Avg.
S.D.
Avg.
S.D.
Avg.
S.D.
S&S
Glass
-0.16
0.09
-0.32
0.12
-0.57
0.22
Gelman(79)
Quartz
-0.36
0.09
-0.29
0.16
-3.19
0.97
Gelman(81)
Quartz
-1.13
0.56
-0.31
0.40
-4.31
0.92
Whatman
Quartz
-0.10
0.08
-0.20
0.11
-2.28
0.73
Pallflex
Quartz
+0.08
0.12
-0.19
0.11
-1.08
0.40
Ghia
Teflon®
-1.26
2.11
+0.16
0.16
+0.50
1.74
*W1 indicates the first weighing and so on.
TableS. Weight Changes (mg) Due to Mounting on Sampler (8x10 inch Filters)
Filter Type
S&S Glass
Gelman(79) Quartz
Gelman(81) Quartz
Whatman Quartz
Pallflex Quartz
Ghia Teflon®
Wt. Changes (mg)
-0.82
-3:51
-5.03
-2.43
-1.14
+0.054]
Corresponding Error in Mass
Concentration (fjg/m3)*
-0.5
-2.0
-2.8
-1.4
-0.6
+0.03]
'Assuming 24-hour high-volume sample (1800 m3 sample volume).
t Weight gain observed.
Table4. TSP Concentrations (fjg/m3) Measured with High-Volume Samplers Using Glass,
Quartz, and Teflon® Filters
Day No.
1
2
3
4
5
6
7
8
9
10
11
12
Avg.
S&S
Glass
43.60
46.38
95.60
76.02
89.11
79.92
76.51
36.71
39.44
42.82
36.11
29.52
57.64
Gelman(79)
Quartz
34.59
43.42
83.75
63.94
74.57
65.31
60.12
26.94
33.88
33.Q9
25.55
23.63
47.40
Gelman(81)
Quartz
38.64
41.92
82.63
66.42
86.88
58.07
71.49
25.63
30.39
34.86
27.01
26.47
49.20
Whatman
Quartz
41.46
50.60
87.69
68.27
85.46
76.87
68.85
32.80
37.82
37.64
31.89
28.84
54.02
Pallflex
Quartz
36.16
45.54
82.84
60.00
76.67
68.24
63.22
29.02
30.29
33.66
26.10
24.06
47.98
Ghia
Teflon®
37.84
44.11
84.37
65.71
72.49*
67.40*
63.80
26.26
32.57
34.11
26.38
24.09
48.26
^Plugging of Teflon® filter resulted in significant drop in flowrate.
Teflon® filters showed no evidence of
systematic weight loss during the weigh-
ing process for any of the filters tested.
2. Weight losses due to mounting of
the 8x10 inch and 102 mm filters on
particle samplers were observed for both
glass and quartz filters. An analysis of the
test data revealed that mounting was a
significant effect for all filter types tested
except the Ghia Teflon®. The estimated
errors in corresponding mass concentra-
tion measurements were less than 3
//g/m3.
3. Ambient TSP, sulfate, and nitrate
concentrations measured using quartz
and Teflon® filters on high-volume
samplers were lower than those measured
using glass fiber filters. The observed
differences in TSP measurements could
be explained in part by the increased
artifact nitrate and sulfate on filters
containing glass fibers (S&S glass and
Whatman quartz).
4. The humidity in the filter condition-
ing and weighing environment has an
apparent effect on filter weights, more so
for the quartz filters than for the glass
filters. The correlation between filter
weight and humidity for the Gelman(81)
quartz filters suggests that closer control
of humidity during conditioning and
weighing might be advantageous for this,
and perhaps other quartz filters, but
further investigation is recommended.
5. The Ghia Teflon® filters exhibited a
tendency to gain weight upon repeated
weighings, even after mounting on the
samplers. Although the observed weight
-------
gains were small over the time periods
involved in this study, this phenomenon
warrants further investigation. The 8x10
inch Ghia Teflon® filters also had a
tendency to clog during ambient sampling
at TSP concentrations around 65 to 75
Aig/m3.
6. Based on the results of this study,
the use of 8 x 10 inch and 102 mm quartz
filters as collection substrates on high-
and medium-volume particle samplers
appears to be feasible. Special care
during weighing, handling, and mounting
operations is necessary when using
quartz filters. The use of filter cassette
holders, designed to facilitate installation
of filters at field sites, is recommended.
The EPA authors, Kenneth A. Rehme (also the EPA Project Officer, see below),
C. Frederick Smith, Michael E. Beard, and Terence Fitz- Simons are with
Environmental Monitoring Systems Laboratory, Research Triangle Park, NC
27711.
The complete report, entitled "Investigation of Filter Media for Use in the
Determination of Mass Concentrations of Ambient Paniculate Matter," (Order
No. PB 84-139 876; Cost: $8.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
U.S. GOVERNMENT PRINTING OFFICE; 1984 — 759-015/7769
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
------- |