EPA-600/2-77-062
March 1977
Environmental Protection Technology Series
X-RAY ANALYSIS OF AIRBORNE ASBESTOS
Interim Report: Sample Preparation
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of.pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-062
November 1976
X-RAY ANALYSIS OF AIRBORNE ASBESTOS
INTERIM REPORT: SAMPLE PREPARATION
by
M. Fatemi, E. T. Johnson, R. R. Whitlock,
L. S. Birks and J. V. Gilfrich
Material Sciences Division
Naval Research Laboratory
Washington, D. C. 20375
Interagency Agreement EPA-IAG-D6-0651
Project Officer
Jack Wagman
Emissions Measurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, N. C. 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences
Research Laboratory, U. S. Environmental Progection Agency, and
approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U. S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.
-------
ABSTRACT
A measurement technique, suitable for submicrogram quantities
of asbestos, using a combination of fiber alignment and x-ray
diffraction was previously introduced. The x-ray measurement
of aligned fibers is a straightforward operation. However, it
has since been found that the sample preparation itself depends
on several parameters which need critical control.
The composition of the alignment medium is established as
10-20 ppm (W/V) of parlodion (cellulose nitrate) in distilled
amyl acetate. The effect of filter pore-size has been shown to
be more significant than previously thought. Fiber losses in the
range of 50-80% of total mass are not unlikely, depending on the
fiber size distribution. Similarly, ultrasonification has been
shown to affect fiber size distribution and subsequent fiber
retention by filters. Ambient relative humidity suitable for
alignment of single drops has been established to be in the range
of 35% to 45% at 22° C to 20° C respectively. Parameters involved
in radio frequency ashing of filters are also discussed. It is
shown that ashing quality and subsequent dispersion depends on
the fiber concentration on the surface of the filter membrane, and
improves as this concentration is increased.
-------
CONTENTS
Abstract iii
1. Introduction 1
2. Summary 2
3. Conclusions 3
4. Recommendations 5
5. Brief Review of the Feasibility Report 7
6. Significant Parameters in Sample Preparation 9
A. Filter Pore-Size and Fiber Retention 10
B. Composition of the Alignment Medium 12
C. Alignment of Large Volumes 13
D. Effect of Humidity on Alignment 15
E. R-F (RADIO FREQUENCY) Plasma Ashing 18
F. Contamination Effects 25
References 29
-------
1. INTRODUCTION
The feasibility of an x-ray diffraction technique for direct
measurement of small masses of asbestos was demonstrated in an
earlier report (Ref. 1). The method involved the electrostatic
alignment of fibers and embedding them in a thin film for x-ray
measurement. The resulting minimum detection limit for pure
chrysotile, 0.15 yg, was nearly two orders of magnitude better
than previously reported by the usual x-ray diffractometry
(Ref. 2). Thus, it showed considerable promise in application
to pollution problems - for example, in the analysis of asbestos
samples obtained near an emission source.
Following the publication of the feasibility report (Ref. 1),
work began on a study of problems associated with the analysis of
real samples. It was during the initial stages of this study,
that several problems were encountered relative to the parameters
involved in the alignment of pure fibers. This work is a des-
cription of additional research on sample preparation, which de-
tails those difficulties and their remedies. In this report, a
brief review of the information covered in Ref. 1 will be pre-
sented, followed by an examination of additional parameters which
were found to be significant in sample preparation. Finally,
suggestions will be offered which are believed to be helpful in
alleviating the remaining difficulties.
1
-------
2. SUMMARY
In the preparation of samples of aligned asbestos fibers by
x-ray diffraction, several parameters have been found to be
critical. The effects of these parameters on sample preparation
are discussed in both qualitative and quantitative terms.
Effects of filter pore size, sonication time, radio frequency
ashing, ambient humidity and contamination sources are
demonstrated.
-------
3. CONCLUSIONS
Preparation of samples of electrostatically aligned
asbestos fibers suitable for measurement in a double-detector
x-ray diffraction system depends on rigid control of several
significant parameters.
The chemical composition of the dispersion/alignment
medium has been established as 10-20 ppm (W/V) of parlodion
(cellulose nitrate) in distilled amyl acetate.
Filter pore size and ultrasonication have significant
effects on fiber retention. Significant fiber loss occurs with
smaller fiber size distributions (i.e., by increasing sonication
times) and with larger filter pore sizes. The recommended filter
pore size is reduced to 0.22 ym. A minimum sonication time
dependent on the source of the fibers is emphasized.
Ambient humidity at room temperature has been shown to have
a critical influence on fiber alignment. The established range
is between 45% at 20°C and 35% at 22°C.
Contamination effects may enter at each stage of sample
preparation to affect both dispersion and alignment of fibers.
Some contaminations impede the radio frequency ashing of filters
while others affect the overall fiber dispersion through pro-
duction of secondary non-ashable products. In this respect, the
use of deionized, but undistilled water in the washing stage has
-------
been correlated with poor dispersion of fibers in amyl acetate-
parlodion, whereas no problem seems to arise with deionized,
filtered and distilled water.
The most critical step in sample preparation appears to be
the radio frequency ashing of filters with small quantities of
asbestos, which yields poor dispersion. Preconcentration of
filters through the use of smaller filter disks seems to improve
the ashing quality but increases the filtration time several
fold.
-------
4. RECOMMENDATIONS
The radio-frequency ashing of filters containing only a
few micrograms of material should be further investigated. It
appears necessary to control the parameters of RF ashing much
more rigidly than has been customarily done. For example, the
effects of plasma input power, vacuum level, impurity gas mix-
tures as catalysts should be further studied.
Use of polystyrene-backed cellulose ester filters may prove
to be beneficial in reducing the ashed filter residue, through
increased mechanical support during ashing. Since the backing
itself is not easily ashable, it may be removed from the ash
after dispersing the latter in a liquid by mild sonication.
The use of plain (unbacked) filter is recommended for
normal situations when the deposit concentration is relatively
high. However, with low concentrations it may again prove
advantageous to reduce the filtration area as far as practicable
in order to increase the surface concentration, and to reduce
the quantity of residue ash from the filter.
Finally, since the primary source of the difficulty with
RF ashing appears to be the filter membrane itself, the feasi-
bility of eliminating the ashing step (either partially or
altogether) should be considered. For example, the filter
-------
collection may be dissolved in an organic solvent, and centri-
fuged to separate nearly all of the collected solvent (by
sonication) and re-centrifuged until only an insignificant level
of soluble inorganics remain. One can then determine whether
the remaining deposit is readily alignable or whether it
requires ashing without the filter membrane to eliminate
organic interferences.
-------
5. BRIEF REVIEW OF THE FEASIBILITY REPORT
Calibration standards containing a few micrograms of
aligned fibers were prepared in the following way. About 3 mg
of chrysotile was sonicated in a 1/2% solution of Aerosol OT in
H20 to produce short straight fibrils. 150 microgram aliquots
were filtered through a mixed cellulose ester filter membrane
(such as Millipore HAWP). The filter was then ashed in a radio
frequency (RF) plasma asher to remove the filter material. The
ashed product was dispersed in the "alignment medium", defined
as 0.001% parlodion in amyl acetate. A small aliquot of this
dispersion, equivalent to about 5 yg asbestos, was placed on a
special multielectrode alignment grid and aligned.
After drying, the fibers were sprayed with parlodion
solution (^ 2% in amyl acetate), so that a thin parlodion film
would be formed on the substrate. This film which contained all
the fibers was then floated off on water and mounted on a plastic
ring.
This ring was mounted perpendicular to a broad, collimated
x-ray beam (see Fig. 1). The diffracted beam was measured at the
29 angle corresponding to the (002) planes of chrysotile. The
scattered beam was measured at the same angle but away from the
diffraction direction. The instrument used in our research,
although conceptually similar to that shown in Fig. 1, used only
a single detector. Therefore, instead of a simultaneous measure-
ment, the signal and background were measured sequentially with
the sample rotated 90° between the measurements.
-------
detector H
x-ray tube
detector I
Figure 1. Double-detector instrument for x-ray diffraction
measurement of aligned asbestos fibers. Detector I measures
diffracted beam from aligned fibers; detector II measures the
background corresponding to the same area on the sample.
8
-------
6. SIGNIFICANT PARAMETERS IN SAMPLE PREPARATION
In the preparation of aligned standards, two potential
problems had been recognized which were discussed in Ref. 1 and
subsequent papers (Refs. 3, 4). One was the presence of ex-
traneous material which accompanies any field-collection. The
other was the fact that one could expect only a small total
amount of asbestos in any reasonable-time collection requiring
that all of the sample (rather than an aliquot) would have to be
processed, aligned and measured. Therefore, it would be necessary
to consider the difficulties associated with the alignment of the
fibers in more than a single drop of fluid, perhaps in the
presence of extraneous particulate material.
In the course of investigating these two parameters we also
encountered additional problems associated with sample preparation
even for pure asbestos standards. For example, the presence of
undispersed lumps of asbestos degraded the alignment. In addition,
even in samples with very good apparent alignment there were un-
explained variations in x-ray sensitivity. These problems and
more were systematically attacked and, with one exception (namely,
the ashing of small samples), were satisfactorily resolved.
The present report is an account of that investigation.
The relevant parameters are discussed roughly in the order in
which they appear in the sample preparation procedure of Ref. 1,
except when it is necessary to change the order to preserve
continuity of subject matter. The parameters discussed are the
following:
-------
A - Filter pore size, fibril size distribution and fiber
retention.
B - Composition of alignment medium.
C - Alignment of large (^ 100 y&) volumes.
D - Effect of humidity.
E - Radio frequency ashing of small samples and blank
filters.
F - Contamination effects.
A. FILTER PORE-SIZE AND FIBER RETENTION
The effect of filter pore-size on fiber retention and sub-
sequent x-ray sensitivity is more significant than initially
thought. When good alignments were obtained (as judged by
observing fibers through optical microscope) this factor alone
could explain most of the fluctuation in x-ray sensitivity. Al-
though our feasibility report suggested the use of 0.45 ym
filters for sample preparation, extensive research on filtration
parameters since that time showed that significant losses could
occur depending on the fiber size distribution.
In the tests performed to relate the x-ray sensitivity to
filter pore size, fiber size distributions were varied by
changing the sonication time, keeping all other sonication
parameters as nearly constant as possible. For each sonication
time (i.e. fiber size distribution) identical masses of asbestos
dispersed in H20-OT were passed through various pore-size filters.
The three filters used were of 0.22 ym, 0.45 ym and 0.80 ym pore
sizes. Each experiment was repeated to obtain statistically valid
data. Otherwise, all samples were prepared in as identical a
fashion as possible.
10
-------
A measure of the relative mass of fibers retained by each
filter is the x-ray sensitivity measured for an aliquot as des-
cribed in the 1975 report (Ref. 1). Here, x-ray sensitivity is
to be interpreted as counts per second per microgram of the
asbestos mass used in the filtration step, rather than the mass
present on the sample. Table 1 shows the variation of measured
x-ray sensitivities for a given sonication time with various pore
sizes. It also shows the variation of sensitivity as a function
of sonication time for a given pore size.
TABLE 1. X-RAY SENSITIVITY VS SONICATION TIME AND FILTER PORE SIZE
Filter
Pore-Size
0
0
0
.22
.45
.80
ym
ym
ym
Sonication Time
40
5.
4.
3.
min.
7 + 0.7
7 + 0.4
8 + 0.5
60
5.
4.
2.
min.
3 + 0.
0 + 0.
5 + 0.
120
5
5
2
4.
2.
1.
2
6
3
min.
+ 0.
+ 0.
+ 0.
3
3
4
The numbers in this table should be considered only relative
to one another. At the time these measurements were made, some
problems still remained due to the effects of contamination on
ashing and the spreading of the aligned drops due to the lower-
than-optimum ambient humidity (both of which are discussed later
in this report). However, the numbers given here remain consistent
among themselves, even though the sensitivity obtained for 0.45 ym
filter and 40-minute sonication time did not approach the pre-
viously reported value of 5.9 c/sec-yg. Note that, following the
remedies discussed later, a well aligned standard chrysotile sample
11
-------
using a 0.22 ym filter is now expected to give a sensitivity of
about 7.5 c/sec-yg using the same x-ray tube parameters.
B. COMPOSITION OF THE ALIGNMENT MEDIUM
The liquid in which the fibers are electrostatically aligned
is the same liquid in which they are dispersed after the organic
material has been ashed. In this sense, the primary purpose of
the alignment medium is its capability to disperse the fibers of
asbestos. It is commonly observed that asbestos fibers in general
tend to flocculate when they are suspended in water and then are
allowed to dry. Similarly, when ashed pure asbestos fibers are
stirred in pure amyl acetate with the aid of the small tip of
sonic cell disrupter, they appear to be dispersed during the mix,
but quickly settle when agitation has stopped. However, the
addition of a quantity as small as 10 ppm parlodion (cellulose
nitrate) to amyl acetate will produce a marked difference in the
suspension of fibers and their dispersion. It appears that
parlodion, even at this low level acts as a surface-active agent
suitable for chrysotile and other silicates in amyl acetate, just as
aerosol OT is suitable for chrysotile in water. However, although
aerosol OT is highly soluble in amy;l acetate, this mixture has not
been found effective as a dispersing agent.
Any concentration of parlodion in amyl acetate (up to 1 or
2%) could be used to disperse asbestos fibers. But the second
requirement of the dispersing liquid is to provide the proper
dielectric properties for electrostatic alignment. Many experi-
ments were performed to establish the suitable range of parlodion
concentration in amyl acetate. Additional tests were conducted on
12
-------
a number of other liquids and solutions (such as alcohols, acetone,
freons, polystyrene thinner, etc.) to check for their suitability
both as dispersant and alignment medium. The final results in-
dicated that the proper liquid is a 10-20 ppm solution of parlodion
in distilled amyl acetate. The total amount of parlodion necessary
for any dispersion depends somewhat on the number of fibers in the
liquid. However, as an empirical rule, any standard asbestos con-
centration up to 'v 1 yg/u& can be easily dispersed in the solution
specified above.
When the parlodion concentration is reduced much below
0.001% (10 ppm), many fibers may remain undispersed, while for
concentrations much above 0.002% (20 ppm) the fiber alignment is
destroyed as the liquid dries on the alignment stage.
C. ALIGNMENT OF LARGE VOLUMES
In the case of field samples, the total asbestos content may
be so small that it will be necessary to process and align the
entire collected filters. The ashed residue from such a filter
is usually high, as the total particulate matter in air is
typically about 200 yg/m .
Since complete separation of asbestos from such a mass is
not practical, the ashed residue cannot be dispersed easily in
5-10 y£ of liquid, but may require as much as 100 y&. Alignment
of the fibers present in such a relatively large volume was ex-
pected to be more difficult than the single drop used for the
standards. Two procedures were considered: one was the use of a
container above the alignment electrode, to contain the fluid
while the solvent evaporated with the voltage applied to the grid
13
-------
ALIGNMENT LIQUID IN
GRID TEFLON WASHER
Figure 2. Test for feasibility of large volume dispersions. The
small teflon cup is held in place by means of a clamp (not shown)
to prevent the flow of liquid through the gap between it and the
electrodes.
(see Pig. 2). This procedure was tested using teflon washers of
different sizes. Although alignment was observed in most in-
stances after evaporation, the major difficulty with the tech-
nique was the uncertainty in the amount of fibers lost due to the
seepage of the liquid under the container, or due to their ad-
herence to the walls of the container.
The second alternative was to align single drops sequentially
on the same electrode until all the liquid was exhausted, allowing
each drop to dry before the next is applied. This straightforward
but tedious approach has proved more satisfactory. The reproduci-
bility of sequential alignment can be shown by preparing pure
asbestos dispersions diluted 10:1 and aligning 10 drops rather
than 1. The x-ray sensitivities for three such samples are shown
in Table 2.
14
-------
TABLE 2. MULTI-DROP VS SINGLE DROP ALIGNMENT
TYPICAL X-RAY SENSITIVITIES
Single drop
1
2
3
5 y£
8.
7.
6.
(4 yg)
2
7
9
Single drop
5 y£ (0.4 yg)
7.3
6.4
8.6
Multi drop
50 y£ (4
7.7
7.9
7.7
yg)
Sequential multi-drop alignment is obviously more time
consuming per se than single drop operation. However, with care-
ful planning this step can be performed without significantly
influencing total time necessary for sample preparation when
dealing with multiple specimens.
D. EFFECT OF HUMIDITY ON ALIGNMENT
The effect of humidity on alignment was discovered when it
was noted that samples of asbestos which had been ashed and
aligned in one test did not align at a later date, even though
presumably the same conditions had been used. After considerable
experimentation two important conclusions were drawn which re-
lated the humidity in the ambient environment to the quality of
alignment:
1 - At very low humidities (e.g., 20% RH at 18° C), fiber
alignment is very poor, regardless of the ashing conditions or
parlodion concentration. As the humidity is raised to ^ 40% RH
at 20° C, the quality of alignment (containment of the drop in a
small area, fiber orientation) reaches an optimum level.
15
-------
The quality of alignment may be judged from the visual
appearance of the aligned drop. Under good conditions, the
fibers remain confined to the area covered by the drop at the
moment of its impact on the grid, although the liquid itself may
spread unevenly over the surface. Under poor conditions (such
as low humidity) the spread of the liquid also carries with it
the suspended fibers, most of which may collect at the boundary
of the drop, unaligned (Figs. 3A, B).
To demonstrate the effect of ambient humidity the following
alignments were conducted. Sample A was a 5 y£ drop aligned in
the ambient humidity of 25% (20° C). Sample B was a drop of the
same specimen, which, after deposition on the grid, was exposed
(A) (B)
Figure 3. Appearance of alignment under (A) poor and (B) proper
relative humidity. The effect of humidity is to "lock in" the
fibers in an area determined by the drop at the moment the power
is turned on.
16
-------
POWER
CONTACT CLIPS
50%
R-.H.
BLOTTER
MULTI-ELECTRODE
GRID
Figure 4. Apparatus to establish correlation between humidity
and alignment. A small inverted cup (approximately 25 mm in
diameter and 30 mm high) over the electrodes contains a circle
of blotter paper, moistened just enough to raise the relative
humidity inside to about 50%. Only a short exposure (1-2 seconds)
to the drop is needed, after the power is turned on, to yield
Figure 3B.
for 1-2 seconds to a small inverted cup (Figure 4) conditioned to
an RH of 50%. The striking difference between the samples was
not only seen qualitatively, as described above, but also in the
x-ray sensitivities: 2-4 c/sec-yg for 25% RH, 6-7 c/sec-yg for
50% RH. Similar values were obtained consistently for many more
samples tested for this variable (Table 3).
TABLE 3. DEPENDENCE OF X-RAY SENSITIVITY ON
AMBIENT RELATIVE HUMIDITY
Sample #
1
2
3
4
RH
20-30%
2.3
3.2
4.3
4.6
RH
40-50%
6.0
6.9
6.4
7.3
17
-------
2 - At very high relative humidities (> 50% RH at 20-22° C)
alignment is still possible, but is accompanied by electrode
corrosion. This phenomenon at times is so severe that it will
destroy the central region of the grid beyond further use. Some
degree of accidental corrosion is tolerable, as long as it leaves
most of the electrodes continuous and intact. Slight non-
uniformities in the electrodes' profile have had no detectable
effect on the future alignments or on the x-ray sensitivities.
Grids seriously corroded during alignment have caused samples
to be more difficult to float off the electrode (even when the
electrode had been corroded slightly during previous use)
possibly because of the adherence of the corrosion product to
the quartz substrate.
The optimum humidity condition for alignment has been deter-
mined to be 40-45% RH at 20-22° C. Extremes, such as 50% RH at
23-25° C, or 25% RH at 18° C should be avoided altogether.
E. R-F (RADIO FREQUENCY) PLASMA ASHING
The conditions required for proper ashing were investigated
on the basis of the observation that the lower the quantity of
ashed residue, the better the overall quality of the final
aligned sample. Proper ashing is dependent on several inter-
related factors.
a - Instrumental factors: RF level, vacuum conditions, gas
flow, geometry.
b - Filter conditions: filter types and "loading", i.e.,
the amount and type of material present; filter exposure to RF
field, etc.
18
-------
c - Contamination effects.
The ashing step is the most critical step in the sample
preparation. Good fiber dispersion or alignment canot be ex-
pected unless each of the factors described above is strictly
controlled.
Instrumental Factors
One of the first parameters to be studied was the geometry -
and the physical placement of the sample in the RF asher, i.e.
the way in which filters and their load should be exposed to the
RF plasma. The ashed material should be contained suitably for
further redispersion in the H20-OT or dispersion in amyl acetate-
parlodion (AA-P). Redispersion of ashed residue in H20-OT is
necessary for field-sample collections which contain a finite
level of soluble impurities, and which need washing and re-
filtration. Dispersion in AA-P is for alignment of clean
standard samples, and for samples ashed after redispersion in
H20-OT.
In either case, it is necessary to contain the ashed product
in a small volume such as a test tube, so that the final dis-
persion volume in AA-P can be small. This is true, because in the
case of the source sample, all of the liquid must be aligned drop
by drop, and hence the total elapsed time for a complete sample
alignment depends on the volume of the liquid.
Confinement of the filter in a test tube, however, raises a
question about the variation in the amount of RF exposure to
different samples. An optimum length for the test tubes was found
suitable for our asher (Perkin-Elmer Coleman 40 - not available
any longer commercially). Ordinary culture tubes (10 x 75 mm)
19
-------
are cut to 25 mm length and cleaned. The filters are then folded
by tweezers and inserted in these tubes. Short tubes were found
to be more satisfactory than the original long tubes, because the
latter increased the time required for complete ashing.
The Coleman RF asher has no provision for measuring the
power being input to the samples. By comparing the results of
ashing samples in other instruments (through the courtesy of the
Electronics Division, NRL and NASA, Goddard), it can be estimated
that ashing six samples requires about 10-20 watts of RF power.
An oxygen flow of 10 cc/min through a 4-cm diameter reaction
chamber has been established to give optimum results under
vacuum of 0.5-1 torr.
One ashing practice for electron microscopy is to"attach the
filter onto a glass slide using a few drops of acetone. The glass
slide is then inserted in the reaction chamber, and exposed to RF
plasma. We found this procedure unsatisfactory; in most cases the
filter began peeling from the slide soon after vacuum pumpdown
started in the chamber. The ashed product lay loosely atop the
glass slide, and sample loss occurred when air was allowed in the
system. Only in those cases where the glass slide was purposely
contaminated with finger dirt, etc., did the filter adhere to the
glass throughout the ashing operation. This, however, would yield
a.product not only unsuitable for alignment but for electron
microscopy as well.
In other instances when the sample is on a slide, it has been
observed that the filter material flashes out spontaneously, scat-
tering fibers irretrievably in the reaction chamber. This is
another reason why containment of the ashed product in a test tube
is the only way to proceed.
-------
Filter Condition: Relationship Between Load Level and Ashing
A peculiar aspect of the ashing process concerns the amount
of ashed filter residue, when filters containing small amounts of
2
pure asbestos (<\, 1 yg/cm ) are compared to filters with large
2
quantities of material (5-70 yg/cm ). In the feasibility report
(1975), a description was given for sample preparation, when
2
quantities as large as 150 yg asbestos (^ 75 yg/cm ) were filtered,
ashed, and dispersed in AA-P. From these samples, small (5 yg)
aliquots were placed directly on the alignment grid and aligned.
Observation of these samples by optical and electron microscopy
revealed very little extraneous matter, so that the ashed residue
of the filter itself was presumed to be negligible. Ashed,
heavily-loaded filters in the test tube have a distinct translucent
appearance, characteristic of the asbestos fibers themselves. The
ashed filter "copies" the out-
line of the area filtered from
H-O-OT dispersion. In general,
there appears to be no visible
residue from the filter itself
(Figure 5).
Figure 5. Appearance of an
ashed filter containing a large
(150 yg) quantity of standard
asbestos. Note in particular
the shape of the asbestos sheet
at the boundary of filtration
area.
i
-------
On the other hand, the preparation of a field sample
(collected near or at an emission source) may involve the ashing
of a "lightly-loaded" filter, i.e., a filter with only 1-2 yg
(0.5-1.0 yg/cm ) asbestos. Here, the entire filter must be ashed
and aligned. Either blank, or lightly loaded filters (with pure
standard asbestos) have shown erratic results in appearance and
amount of ashing residue.
The ashing of a blank filter often results in a solid
plasticized mass at the bottom of the test tube (Figure 6A). No
reasonable level of sonication in AA-P will break or disperse
this mass, which, incidentally, is denser than the liquid and
will not remain suspended (Figure 6B).
(A) (B)
Figure 6-(A). An ashed blank filter in test tube (3x). (B)
Optical micrograph (500x) of the filter dispersed in AA-P and
"aligned" to show undispersed lumps and their chaining in the
field.
22
-------
Figure 7. Lightly-loaded
filter, ashed and dispersed
in AA-P, showing large
particles which distort the
alignment field.
Filters with small amounts of asbestos have also shown the
same general effect, in contrast to heavily-loaded samples.
Dispersions of these samples under optical microscope show masses
of transformed matter many tens of microns in diameter, each
containing a small fraction of the fibers (Figure 7). Alignment
of such samples almost invariably leads to a chaining together of
the masses, distortion of the field near the few free fibers and
a nearly zero sensitivity.
A comparison of lightly-loaded and heavily-loaded filters
seems to indicate that the large quantity of interwoven asbestos
fibers acts as a supporting grid to allow the filters to ash
slowly and completely. Most of the incomplete or poor ashing
appears to be the result of the filter collapsing and rolling onto
itself - thereby preventing complete interaction with the plasma.
23
-------
This suggests the idea of preloading the filter with an inert
fibrous material which would not interfere with the x-ray
measurements. For example, when the measurement of chrysotile
or amosite is intended, a suitable quantity of the crocidolite
could be added to the samples. The diffraction peak from
crocidolite is at about 14° 29 (Cr Ka radiation) which is
sufficiently removed from both chrysotile (18 ) and amosite (16°).
Also under consideration at the present time is the use of
polyester-reinforced mixed cellulose ester filter membrane.
Polyester is highly resistant to ashing and may therefore support
the filter throughout ashing and prevent it from collapse. After
ashing, it may be shaken off with sonication in H^OOT. During
the second ashing, a much smaller filter could be used to
minimize the volume of the alignment liquid. An advantage of the
smaller filter disks (e.g. 8 mm) is that the load concentration
is increased to the point that the filter disk may be more com-
pletely ashed due to the load support (see discussion below).
In most sampling situations, however, preloading auto-
matically occurs due to the presence of extraneous particulate
matter: If a cubic meter of air is collected, then it is likely
that about 100-200 yg extraneous material may also be present on
the filter. This quantity seems to be just the proper level of
filter support for good ashing.
An attempt has been made to increase the concentration of
asbestos on the filter membrane by reducing the filtration area.
Consider the case of pure chrysotile. Ashing a 16 mm filtration
disk has been observed to be successful for a load level of
2
^ 10 yg (5 yg/cm ). Therefore, it would seem reasonable that a
24
-------
Figure 8. Improvement in the
ashing quality of small (8mm)
filter disks (3x). This test
tube contained about 4 yg
total asbestos, and should be
compared with Figures 5 and 6.
load level of 2.5 yg pure asbestos with no extraneous matter
would require a disk of 8 mm in diameter. To test this idea,
8-mm upright filteration funnels were fabricated and used to filter
2.5-3 yg asbestos. A definite improvement was observed in the
ashing quality of smaller disks (Figure 8) approaching that shown
for larger samples (Figure 5). However, it is clear that such a
disk is not suitable for large volume sonication, and its use may
be limited to the re-ashing of the second filtration.
F. CONTAMINATION EFFECTS
At least part of the problem of poor ashing may be related to
the contamination of the filter itself. For example, the spacer
used in the packaging of the filter membranes may become suspect.
The presence of impurities in the filter may act to at least
locally affect the quality of ashing. To ascertain whether this
25
-------
impurity played a major role in the alignment, several filter
membranes were acid cleaned to dissolve away impurities. Filters
were then washed, dried, and used for filtration of standard
asbestos. (Some warping of the membrane occurs after acid bath
and washing.) Experiments with these filters, however, showed
no significant improvement in the fiber dispersion or ashing of
blank filters. However, with some ashings, it is even possible
to detect the operator's fingerprint on the ashed asbestos sheet.
B
Figure 9A. Appearance of non-alignable bundles of fibrils (500x
photomicrograph). For comparison, fully dispersed, aligned
fibers are shown in 9B.
26
-------
Other tests on the presence of impurities were conducted. For
example, the presence of a very thin (^ 1-2 ym) layer of ferric
sulfate or silver chloride on the filter greatly dimishes the
ashing rate and thus contributes to the later contamination of
the alignment medium, through dissolution of the filter membrane
in amyl acetate. This contamination can destroy electrostatic
alignment.
Another contamination problem, distinct from, but similar
to the incomplete ashing of blank filters, is traced to the use
of undistilled but deionized water. Lumps of asbestos fibrils
appear to remain intact after an otherwise complete ashing.
Under optical microscopic examination, they appear as granular
portions of unashed filter (Figure 9). Even after extensive
sonication and experimental changes in parlodion concentration,
they remain fairly intact, and the alignment of fibrils as well
as x-ray sensitivity is affected by their presence. However,
electron microscopy shows that each "granule" is composed of
straight, short fibrils held together with a minute amount of
adhesive-like material (Figure 10). The exact nature of the
adhesive has not been determined. However, it is interesting to
observe the sharp boundary between fiber and no-fiber region.
It indicates that the nature of this impurity is local and is
probably brought about as a result of the decomposition of the
impurity into a non-ashable,insoluble compound. The impurity
finds its way through the ion-exchanger column, and may be
trapped only by the small pores of the filter membrane.
27
-------
The use of distilled water, or deionized, distilled and
filtered water seems to eliminate this difficulty, and again is
emphasized throughout the sample preparation.
Figure 10. Electron micrograph (SOOOx) from bundles in sample
of Figure 9A. Unlike the photomicrograph, the boundaries of
lumps appear quite sharp.
28
-------
REFERENCES
1. Birks, L. S., M. Fatemi, J. V. Gilfrich, and E. T. Johnson.
Quantitative Analysis of Airborne Asbestos by X-Ray
Diffraction. EPA-650/2-75-004, 1975, U. S. Environmental
Protection Agency, Research Triangle Park, N. C., 11 pp.
2. Rickards, A. L. Estimation of Trace Amounts of Chrysotile
Asbestos by X-Ray Diffraction. Analytical Chemistry 44(11):
1872-3, 1972.
3. Birks, L. S., and M. Fatemi, Quantitative Measurement of
Pollutant Asbestos: I-A New X-Ray Diffraction Technique.
In: 68th Annual Meeting of the Air Pollution Control
Association, Boston, Massachusetts, 1975.
4. Fatemi, M., E. T. Johnson, L. S. Birks, and J. V. Gilfrich.
Quantitative Measurement of Pollutant Asbestos: II-Sample
Preparation and Typical Results. In: 68th Annual Meeting of
the Air Pollution Control Association, Boston, Massachusetts,
1975.
29
-------
TECHNICAL REPORT DATA
(Please read Instructions on Me reverse before completing)
1. REPORT NO.
EPA-600/2-77-062
3. RECIPIENT'S ACCESSION-NO.
4, TITLE AND SUBTITLE
X-RAY ANALYSIS OF AIRBORNE ASBESTOS
Interim Report: Sample Preparation
5. REPORT DATE
March 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOH(S)
M. Fatemi, E. T. Johnson, R. R. Whitlock,,
L. S. Birks, and J. V. Gilfrich
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Material Sciences Division
Naval Research Laboratory
Washington, D. C. 20375
10. PROGRAM ELEMENT NO.
1AD605
11. CONTRACT/GRANT NO.
EPA-IAG-D6-0651
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Research Triangle Park. N. C. 27711
- RTF, NC
13. TYPE OF REPORT AND PERIOD COVERED
Interim - 10/74-10/76
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT •
A measurement technique, suitable for submicrogram quantities of asbestos, using
a combination of fiber alignment and x-ray diffraction was previously introduced.
The x-ray measurement of aligned fibers is a straightforward operation. However,
it has since been found that the sample preparation itself depends on several
parameters which need critical control.
The composition of the alignment medium is established as 10-20 ppm (W/V) of
par.lodion (cellulose nitrate) in distilled amyl acetate. The effect of filter
pore-size has been shown to be more significant than previously thought. Fiber
losses in the range of 50-80% of total mass are not unlikely, depending on the
fiber size distribution. Similarly, ultrasonification has been shown to affect
fiber size distribution and subsequent fiber retention by filters. Ambient
relative humidity suitable for alignment of single drops has been established
to be in the range of 35% to 45% at 22°C to 20°C, respectively. Parameters
involved in radio frequency ashing of filters are also discussed. It is shown
that ashing quality and subsequent dispersion depends on the fitler membrane,
and improves as this concentration is increased.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
^IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
*Asbestos
*Air pollution
*X-ray diffraction
*Alignment
*Quantitative analysis
Chrysotile
Crocidolite
Amosite
HE
13B
20F
07D
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)'
UNCLASSIFIED
21. NO. OF PAGES
36
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
30
------- |